Freshwater Biology (1988) 20, 199-210

Natural community bioassays to determine the abiotic factorsthat control phytoplankton growth and succession

E. VAN DONK. A. VEEN* and J. RINGELBERG* Provincial Waterboard ofUtrecht. Utrecht, and *Department of Aquatic Ecology, University of Amsterdam,The Netherlands

SUMMARY. 1. The successional patterns of the dominant phytoplanktonspecies in Lake Maarsseveen {The Netherlands) were very similar in 1980and 1981. In December/January the diatoms Stephanodiscus hanlzschiiGrun., Stephanodiscus astraea (Ehr.) Grun. and Asteriotiella formosaHass. dominated the algal community [A. formosa had several furtherpopulation increases during the year). Fragilaria crotonensis Kitt. came todominance in March/April, followed by the chrysophyte Dirtohryondivergens Imhof and the diatom Cyclotella comta (Ehr.) Klitz in May/June. A second appearance of D. divergens was observed in July/August,followed in 1980 by F. crotonensis and a third small increase of D.divergens. Inbothyears5. aslraeadwdS. /iflrt?25cft« started to grow again inNovember/December. Cryptophyceae and Chlorophyta were presentthroughout the year, but did not show a distinct succession.

2. Natural community bioassays, performed under natural light andtemperature conditions in a newly developed outdoor bioassay apparatus,showed that phosphate was the major nutrient limiting the growth rate ofthe phytoplankton. From January till June, during the decline in phos-phorus concentration, the diatoms became successively phosphate limitedin the sequence: S. hanlzschii, S. astraea. F. crotonemis, A. formosa andC. comta. Light limitation was probably the major cause of the relativelylate start of F. crotonensis in early spring.

3. D. divergens. Increasing after the diatoms from June till September,was stimulated by the addition of a chclator (EDTA). The cheiator mightstimulate the formation of trace metal species favouring their uptake (e.g.iron).

4. The patterns of succession of the diatoms observed from January tillJune and from July till December were to a large extent symmetrical. Thecontrolling factors followed opposite trends: declining phosphorus con-centrations with increasing irradiance from winter till spring and increasingphosphorus concentrations with decreasing irradiance from summer tilllate winter.

Correspondence: Dr E. van Dotik, Provincial Waterboard of Utrecht, Postbox 80300, 3508 TH Utrecht, TheNetherlands.

199

200 E. van Donk, A. Veen and J. Ringelberg

5. A. formosa was sometimes heavily infected by the parasitic chytridfutigus Zygorhizidium planktonicum Canter. The infection may have beenresponsible for the cyclic wax and wane of Asterionella with intervals of 2-3months, reductions in Asterionella perhaps favouring increases in otherdiatom species.

Introduction

In studying the factors regulating the growth andsuccession of the dominant phytoplanktonspecies in Lake Maarsseveen (The Nether-lands) , attention has been paid to both biotic andabiotic factors using a combination of field andlaboratory studies (Van Donk. 1983). The roieof biotic factors (e.g. parasitism) has been dis-cussed in Van Donk & Ringelberg (1983). In thispaper we present the results of natural com-munity bioassay experiments, carried out over a2-year period, to assess the abiotic factor(s)limiting the growth rates of the differentphytoplankton species. At the same time,nutrient concentrations and algal populationdensities were monitored in the lake.

Bioassays. by means of nutrient enrichmentexperiments, have been widely applied to studyeffects of nutrient supply and limitation onphytoplankton growth in natural waters. Twomajor variants to the method can be distin-guished: (1) nutrients are added to filtered lakewater into which laboratory cultured species areinoculated (e.g. Paasche. 1978; Reynolds &Butterwick. 1979; Dc Vries. 1983. 1985). and(2) nutrients are added to lake water containingthe natural phytoplankton community (e.g.Schelske etal.. 1974; Frey & Small. 1980; Vander Does & Klapwijk. 1987). Since elucidationof natural succession was our major objective,the second approach was used here.

Experiments were performed in a newly con-structed bioassay apparatus, situated outside thelaboratory, in which growing conditions simul-ated those of the natural situation as closely aspossible.

The exclusion of nutrient fluxes from sedi-ments and other sources may produce conditionsthat are not comparable with those in the lakeitself. Only large tubes, as used by Reynolds &Butterwick (1979). Lund & Reynolds (1982) andIstvanovics et al. (1986). include the seditnent asa possible nutrient source. Whether the exclu-

sion of nutrient fluxes in enclosures presents aserious problem depends mainly on the morpho-logical, physical and chemical characteristics ofthe lake in question. Exclusion of nutrient fluxesin the bioassay experiments is not a major pro-blem in the case of Lake Maarsseveen because ithas no inflow apart from groundwater seepageand rainfall (Van Donk. 1987). Moreover, theepilimuion is well separated from the hypolim-nion for a long period (Van Donk, 1983) and thesediment is well oxygenated for 10 months of Iheyear (Swain. Lingeman & Heinis. 1987). Sincethe iron content of the sediment is relativelyhigh, its binding capacity will be large, par-ticularly for phosphate.

The use of bioassays may give problems whentwo nutrients are simultaneously in short supply.We obviated this problem by counting the num-ber of algal cells and by employing both themultiple and single addition technique (seeMethods). The growth rate for each of the domi-nant species was measured, because two (ormore) species, growing under identical nutrientconditions can be simultaneously limited bydifferent nutrients (Glooscheusko & Avis. 1973;Titman, 1976, 1977).

Materials and Methods

De.scription of Lake Maarsseveen

Lake Maarsseveen. situated in the centre ofThe Netherlands near the city of Utrecht, wasformed around 1960 by excavation of sand Iu apeat-bog area. The oligo-mesotrophic, trough-shaped lake is replenished mainly by precipita-tion and ground-water and drained via an outletat the northern shore. The lake is subrectangularelongate in outline with a surface area of 70 haand a maximum depth of 30.8 m. Annually, thelake stratifies thermally and an oxycline isformed. A more comprehensive description ofthe lake and its drainage area is given by Swain etut. (1987) and Van Donk (1987).

Natural community bioa.ssays 201

Sampting and analyses

To determine densities of the dominantphytoplankton species in Lake Maarsseveensamples were taken once a week iu 1980 and1981 with a 3-litre vau Dorn sampler at I mintervals to a depth of 10 m. Dominantphytoplankton species were counted accordingtoDorgelo. Van Donk & De Graaf Bierbrouwer(1981). From these data in situ reproductiverates of the algal species were calculated.

Water samples for chemical analyses weretaken weekly at four depths (0.5,5.10 and 15 m)and filtered through a 0.45 //m filter presoakedin diluted HCl and rinsed with double-distilledwater. Soluble reactive phosphorus (SRP). sili-cate, nitrate, nitrite and ammonium were deter-mined according to the methods described inVan Donk & Ringelberg (1983). Data on thetotal daily irradiance were obtained from theRoyal Netherlands Meteorological Institute atDe Bilt. situated 7 km from the lake.

Natural community bioassays

Two basins of stainless steel (1.40x1.20 m)with a depth of 0.30 m were dug into the groundat an open spot near the laboratory (40 km fromthe lake). The temperature of the water wasmaintained at the same level as in the epilimnionof the lake by a cryomat (Lauda TK 30D) and athermomix (Braun). Eight perspex cylinders(length 68 cm. diameter 15 cm), with a capacityof 12 litres each could be placed horizontally ineaeh basiu. The cylinders were rotated con-tinuously at a slow rate by a motor. Iron bars, aslong as the cylinder and coated with tephlon,were put inside the cylinder to prohibit algalgrowth on tbe inner walls (Fig. 1). Natural lightconditions in the cylinders were simulated by ablue perspex plate (Plexiglass 607933). servingas a filter. It absorbed mainly red and ultravioletligbt. reducing the light intensity by 45%. Thealgae within the cylinder received about 40C ofthe incident light intensity, which corresponds toa depth of about 2 m in Lake Maarsseveen. Thealgae in the lake, however, most probablyexperience a gradient of light intensities, due tocirculation through the epilimnion during thetime of the bioassay. It was not possible to simul-ate these conditions in the experiments.

Mixed water samples of the upper 7 m of LakeMaarsseveen were conveyed to the laboratorywithin 1 h of sampling, and filtered through

c

FIG. I. A plan of the bioassay apparatus.A = Cy!indcr; B = Cliain. driven by ;i motor; C = Irt)nbar. coated with tephlon; D = Opening in the cylinder,which can be closed; i = Innow of water in the basin(tetnperature regulated); o^Outfiow of water frottiihc biisin.

150/(m gauze to remove most of thezooplankton before filling the cylinders (en-closures). Single nutrient addition bioassayswere performed (oniy one nutrient added perenclosure), and, in addition, we applied :i mul-tiple nutrient addition bioassay technique(developed by Maslin & Boles. 1978) in whichlake water was enriched with all the nutrients butone. the test nutrient. The omitted nutrienttherefore will limit algal growth in the bioassayat the level of its natural concentration. Theexperiments were carried out once a month.initially in one basin (from March 1980 till April1981) and later in two basins (April 1981 tiilSeptember 1981). Two basins offered thepossibility of sixteen nutrient combinations.Table 1 gives the concentrations of the addednutrients and Table 2 gives the nutrient com-

202 £. van Donk, A. Veen and J. Ringelberg

TABLE 1. Range and concetitration of nutrientsadded to the bioassay enclosures

Nutrients

PNSiB

Vitamin mixBiotinB,:B,

Trace metalsCuZnCoMnMo

FeEDIA

Compound

K:HP04NaNO,Na,SiO,.9H,0H,BO.,

CyanocobaiaminThiatiiine-HCI

CuSOj 5H2OZnS0r7H,0CoCl.-6H,6Mna,-4H.ONa:Mo04'2H,0

FeCIrftH^ONa.EDTA 2H.O

Concentration

3.20 (HM)

71.0035.70

O.IO

()-05(/,g|-i)5.0

100.0

0.039 (/VM)0.0770.0420.9140.026

16.00!3.(K)

TABLE 2. Nutrient combinations used in the bioassayexperiments. (All) indicates the lake walcr (LW)enriched with complete list of nutrients given in Table1.

March 1980 toApril 1981

AllAII-PAll-P-SiAll-SiLWLW+PLW+SiLW+P+Si

May 198! toSeptember 1981

AllAII-PAll-P-SiAll-SiLWLW+PLW+SiLW+P+SiAll-NAll-FeAll-FeEDTAAll-Trace metalsLW + FeLW + FeEDTALW+EDTALW + FeEDTA + P

binations in the different cylinders during 1980and 1981. In the cylinders the pH equalled them.y/m value (8.0). Because of the low algal densityand the short duration of the experiments pHnever increased by more than 0.5 unit.

The growth of the different algal species wasfollowed for 6-8 days every second day. bycounting the cells in three subsamples. Weevaluated responses of nutrient enrichmentfrom the exponential phase of the growthcurves. Growth rates («) with their 95% eonfi-deuee intervals were calculated by a leastsquares linear regression analysis (Sokal &Rohlf. 1969).

Results

Sueeessional patterns in the two years were verysimilar (Fig. 2). In December/January thediatoms Stephanodi.scus hantzschii Grun.,Stephanodiscus astraea (Ehr.) Grun. andAsterionella formosa Hass. dominated the algalcommunity. A. formosa had several furtherpopulation increases during the year. The twoStephanodiscus species were followed by anincrease in Fragilaria crotonensis Kitt. inMarch/April. In May/June the chrysophyteDinobryon divergens Imhof and the diatomCyclotetta comta (Ehr.) Kutz were dominant. Asecond increase by D. divergens was observed inJuly/August, followed in 1980 by F. crotonensisand a small third increase in D. divergens. Inboth years S. astraea and S. hantzschii started togrow again in November/December. Cryp-tophyceae and Chlorophyta were presentthroughout the year, but did not sbow a distinctsuccession. A comprehensive description of thesuccession over a period of 5 years is given inVan Donk (1983).

Nitrite and ammonium concentrations in thelake were always negligible. Nitrate concentra-tions remained high throughout the 2 years.varying between 18 and 37 ,HM (Fig. 3). The SRPconcentration was very low throughout the year.highest concentrations being measured duringDeeember (approx. 0.15 UM P ) . and decliningduring the spring to undetectable values(<0.03;m). In August/September the SRPconcentration rose again above the detectionlimit and increased slightly until Deeember (Fig.3). Silieon concentration remained relativelyhigh, the lowest silicon concentration (7 /m Si)being measured in the epilimnion during July1981 (Fig. 3B).

Examples of the bioassays performed inAugust 1980 are given in Fig. 4. At that date, A.formosa, F. crotonensis and D. divergens werepresent in the lake (>10 cells ml"'). For y4. for-mosa and F. crotonensis phosphorus was thelimiting nutrient. With phosphorus added (LakeWater+P. Lake Water+P+Si. All and All-Si),the growth rates were significantly higher than incombinations without phosphorus (Lake Water.,Lake Water+Si. AU-P. All-P-Si). The growthrate of D. divergens was limited by a nutrientother than phosphorus or silicon. A significantstimulation of growth was present in the com-binations All. All-P, All-P-Si and All-Si. Addi-

Natural community bioassays 203

1960

A

Jan Feb Mar Apr May Jun JuL Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov DecI t ( t 4 t I t ( t I t t 4 4 ( 4

FIG. 2. Successional pattern of the dominatit phytoplankton species in 1980 and 1981. The arrows beneath thehorizontal axis indicate the start of the bioassay experiments. A=A. formosa', B=5. hantzschii: C = 5. astraea;D = F. crotonensis; E = D. divergens: F=C. comta.

tion of phosphorus alone or together with silicon(Lake Water+P and Lake Water+P+Si)resulted in an even lower growth rate of D.divergens eompared to that in lake water. In1981, using sixteen nutrient combinations, thiseffect was analysed further. The chelating agentEDTA proved to be the stimulating factor (seebelow).

For the diatoms A. formosa, F. crotonensis, S.astraea and S. hantzschii phosphorus proved tobe the only nutrient that limited growth ratethroughout the year (Table 3). For each speciesthe growth rates, as measured in all nutrientcombinations with phosphorus added, did notdiffer significantly from one another {regressionanalysis; Sokal & Rohlf. 1969). The same holdsfor the growth rates in combinations withoutphosphorus. Therefore, for each group of com-binations these growth rates were pooled andaveraged. Mean growth rates from the seven-teen bioassay experiments, performed during1980 and 1981, are compared in Table 3. The insifu rates of increase of the algae, are also given.

From October till April, A. formosa did notdemonstrate a significant stimulation in growthwith nutrient addition (Table 3a). In nearlyevery experiment this species was to a greater orlesser extent infected by a parasitic ehytridfungus, Zygorhizidium planktonicum (VanDonk & Ringelberg, 1983). Infection of morethan S()% of the cells was observed in March1980 and in the period February to the end ofMarch 1981. During these months the rate ofincrease of A. formosa in the lake was oftenlower than the growth rate measured in acylinder with pure lake water. Most probablyhigh sinking rates of the infected cells in the lake(we observed clotting of the colonies) wasresponsible for the observed difference inincrease.

S. astraea and S. hantzschii, only presentabove the detection level during winter and earlyspring, were continuously limited in their growthby phosphorus. In December 1980, when theSRP concentration in the lake reached its max-imum, P-addition had minimal effect on the

204 E. van Donk, A. Veen and J. Ringelberg

CL5mt5mJ A N ' f E B ' M A H I A P R ' M A Y ' J U N ' J U L ' A U G ' S E P ' O C T N O V D E C

JAN ' FEB I MAR ' APR ' WAV ' j u N ' JUL ' AuG " " ^ 7 ^ ' B c T " " r ^ ^ T ^NOV DEC

FIG. 3. Concentrations of silicate, soluble reactive phosphorus (SRP) atid nitrate at different depths during (A)1980 and (B) 1981- In Ihe case of SRP, mean concentration in the upper 10 m is given.

Natural community bioassays 205

B

FIG. 4. The growth responses of (A) A. formosa, (B) F. crotonensis. and (C) D. divergens in the bioassayexperiments performed in August 1980. AM nutrients. D; All-Si, *; Lake Water (LW)-)-P, • : LW + P+Si, A;LW + Si. O; All-P-Si, A;LW~ x; All-P. • .

growth rate (Tables 3b and 3c). At the end of itsgrowth period, only a small proportion of S.astraea cells were infected by Z. planktonicum(30%) (Table 3b).

In December 1980 and January 1981 thegrowth rate of F. crotonensis could not be stimul-ated by addition of nutrients (and no fungalparasites were observed). However, fromMarch til! October, F. crotonensis was stronglyphosphorus limited (Table 3d). After the onsetof thermal stratification (end of April) lowerrates of increase were found in the lake thanin the cylinders without phosphorus addition(^(-P)) . At this time the sinking of diatoms inthe lake may have been accelerated due to thehigher water temperature and the lower tur-bulence (Table 3d).

For D. divergens and C. comta the artificialchelator EDTA proved to play a role in com-bination with phosphorus. Therefore, growthrates were averaged. differentiating according tonutrient combinations with and without EDTA(Table 4). In May 1980 and 1981, at the start ofthe growth period, the growth rates of bothspecies were limited by phosphorus. In June,only a small stimulation by a single addition ofphosphorus was observed for C comta. Max-imum stimulation of growth only occurred whenEDTA was added as well. In June D. divergensshowed no growth stimulation after addition of P

alone, whereas a single addition of EDTAincreased growth rate. In July/August, duringthe second population Increase, the growth of D.divergens was inhibited when phosphorus with-out EDTA was added compared to growth in thecylinders without phosphorus addition (Table4a). In August 1981 the growth of D. divergenswas again enhanced by a single addition ofEDTA.

Discussion

Succession during winter and spring

During winter and early spring, phosphateappears to be the only nutrient that limits growthrate in Lake Maarsseveen. Generally, phos-phate is the major limiting nutrient in non-eutrophic freshwater environments (e.g.Schindler, 1977; Lin & Schelske. 1979) thoughoccasionally silicon limitation has been reported(Youngman. Johnson & Farley. 1976; Sommer& Stabel. 1983; Dokulil & Skolaut. 1986).

The two Stephanodiscus species were phos-phorus limited throughout their period of occur-rence (Fig. 5), even in winter when the otherdiatoms were not limited by any nutrient. Inboth 1980 and 1981 A. formosa was heavilyinfected by a parasitic fungus. Z. ptanktonicum.This may have reduced the competitive ability of

206 E. van Donk, A. Veen and J. Ringelberg

-o.£:u•c

c•a

T3CM

_2JSU

*-"C

uu(J

"I? a.

2 J£ u

o .=

4= ^ IJ

1

O O O I I

o o o o+1 +1 +1 +1

CD C: C O-l-l -fl -H +1

I I i I i I I I I I I I I I

n

o o

ooo

ooo1

oO 11

OS r--n —1 d d

i

q

Trn

qd£ddd

i

07)

o+1

o

s0+

2303

)

o-HI

f,

o

—o+1

r-l

X

-HI

-o

ocoo+1rnm

03)

o+1

06)

o+1

r-l I--

°

o+1r-

°

o

TlIf-)

02)

o+1

Ti

o+1

r-fN

o+1

m

o

Ti

(90

o+1

o

°

o+1

(20

o-HI

r-lO

moo+1

(20

o-HI

o

so+1

o

03)

Tl

°

oo-HI

rlo

Tl

ri

°

-HI

rn

(01

Tl

°

o

-HI

oc

o

+\ac

°

d

+1OSfN

01)

o+1

;=;o

C:!o+1nc

o

-HI

o

r--oo+1r-l

01)

Ti

o

n

o+1in

r o o r i o o r o O— - H o o r j o

rl rl O u", C!j c t

I I

—

(i.1

"'

•a

a.+

09)

±0

.

°

00o

±0

61

07)

±0

.

sor-i

o

±0

om

06)

±0.

ocoo

o

±0

fN^^

02)

±0

£o

o

TifN

(

+ 0

c

r-

+

T

-c*

(20

±0

rlOO

o-HI

r-lt

15)

±0

sD

mo

Ti

(01

Tl

oo

o

•n rn00 \C-

06)

±0

oc•n-

o

o

Ti1---

03)

Tirio

oo+1

fN'-^

T,

±0

o

s+1

l / " l

o

05)

±0

o

oc

o+1O S

o

+1

o

o

-HIr,nr-

O

Tinoo

o

Ti

07)

±0

o

oo

o+1

oo ocr) r-—

08)

±0

r-l

oo

o+1

f,

01)

±0

s

+\<x

o

+1

oo

Ti

•/-•. •/".

±0

;:^

O

O

O

+1

±005

o

07

o+1

n rn rn

I I I I IV V

I OI

o

15 (

±0

oOT

) 2(

o

02)

OT

) 01

' — •

±0

.03

±0.0

3±

0.02

21 14 2401

))3

(±0

±0.

0123

03)

)4(±

0±

0.02

I I I l O O O O O l I I

oocm

•4

(a)

0

O -^-^cc ac

socqr-lfN

opdO woc OOq c

O k ^ 'DC OCDC

r^ oc fNs30O

Op6[ri

op

DCo

op5

oc oc

SSrn rnO O

oc

d

m

oc

O

00

o

00

-08-

r-lQ 2 i£

Natural community bioassays 207

TABLE 4 The mean growth rates of (:i) Dinobryon divergens and (b) Cydotella comta enriched with phosphorus(^( + P)) and nut enriched with phosphorus (j"i(-P)), with or without a simultaneous EDTA addition. I he 95%confidence intervals arc given in parentheses. Also given are the rate of increase in the lake and the percentage ofcells infected by parasitic fungi C JI) at the beginning of the experiments.

Date

(a) 04-03-8008-04-KO22-04-8006-05-8017-06-8018-07-8002-08-800X-(W-8027-10-8008-12-8006-01-8103-02-8103-03-8131-«3-Sl11-05-8110-06-8111-08-81

(b) 04-03-8008-04-8022-04-8006-05-KO17-06-8018-07-8002-08-8008-09-8027-10-8008-12-8006-01-SI03-02-8103-03-8131-03-8111-05-8110-06-Sln-08-81

M+P)(t l ' )

+ EDTA

---

-EDTA

---

0.33 (±0.03)0.41 (±0.06)0.50 (±0.12)0.69 (±0.10)0.05 (±0.09)------

0.07 (±0.03)-0.02 (±0.02)-0.21 (±0.08)-0.79 (±0.08)

------

0.57 (±0.06)0.30 (±0.05)0.38 (±0.09)

---0.36 (±0.02)0.42 (±0.07)--

--

----0.42 (±0.()6)-

0.12 (±0.03)-0.14 (±0.01)

---0.24 (±0.04)0.14(±0.(M)--------_-0.24 (±0.03)-

/>(-P)(d ')

-fEDTA

---

-EDTA

---

0.28 (±0.02)0.32 (±0.02)0.42 (±0.09)0.61 (±0.11)0.03 (±0.10)-----

0.10 (±0.03)O.IO (±0.04)0.21 (±0.08)

-0.29 (±0.02)------

0.47 (±0.03)0.20 (±0.01)0.40 (±0.08)

---0.18 (±0.03)-----------0.07 (±0.01)-

0.13 (±0.04)0.16 (±0.02)

u(lake)r V r

(d-o---0.28

-0.350.030.320.03------0.26

-0.220.08

---0.14

-0.09-----------—

-----------------

----5----------

10—

A. formosa for phosphorus giving an advantageto other species. In the previous yeiirs (1978 and1979), when parasitism of A. formosa was ofminor importance due to inhibition of fungalactivity by low temperatures, A. formosa seemsto have outcompeted F. crotonensis and bothStephanodisctis species, because of its higheraffinity for phosphorus (Van Donk &Ringelberg. 1983; Van Donk. 19S.1).

F. crotonensis started to grow around the mid-dle of February. In Deeember and January thisspeeies was present above the detection level,but no signifieant growth stimulation withnutrient addition could be induced till February.Apparently, the growth of ihis speeies waslimited by a factor other than nutrients in this

period. Although compared to the otherdiatoms F. crotonensis grows slowly at Iow temp-eratures (Van Donk. 1983). temperature cannotbe taken as a key factor, because F. crotonensisachieved a faster rate of growth inFebruary/March 1981 (temperature of 4''C)than in December 1980 (8°C) and January 1981(4°C) (Tables 3d and 5). Light was probably thedetermining factor. The mean total daily irra-diance (in Joules cm"-), averaged over theexperimental periods, were 125 (December198(1). 183 (January 1981), 585 (February 1981)and 440 (March 1981) (Table 5). The fastergrowth rate of this species in February/March19SI was probably due to an increase in light.Reynolds (1973) found that F. crotonensis was

208 E. van Donk. A. Veen and J. Ringelberg

FC

Dd

Cc

JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV DEC

^ ^ Limiied tiy phosphorus

|o°op| Limited by parosites

|£| ; | j Limited by phosD^Ofus and parasites

Ug^ j Limited by light

I : ; ; | Limited by trace metal

I ^ I Not present aboue the detection level

FIG, 5, Schematic survey of Ihe factors limiting thegrowth rate of the different dominani phytoplanktonspecies in Lake Maarssevcen in 1980, Sh = .V,hantzschii. Sa= S. astraea. M=A. formosu. Viz=F.crotonensis. Dd = O, divergens. and Cc = C, comta.

less able to grow under low light conditions thanA. formosa.

D. divergens and C. comta dominated in Mayand June. In May D. divergens and C. comtawere slightly phosphorus-limited. In JuneEDTA proved to play a role. C. comta waslimited in its growth rate by phosphorus, but

TABLE 5. The water temperature and light conditions(mean total daily irradiance) during the bioassayexperiments

DateWatertemperature ("C)

04-03-8008-04-8022-04-8006-05-8017-06-8018-07-80O2-O8-8t)08-(W-8027-10-8008-12-8006-01-8103-02-810: -0.3-81.31-03-8111-05-8110-06-8!11-08-81

47«

101816ly171384447

l.>1718

Light{Joules

40913761124181211871174127010314101251835854405{)6

16481270I0I3

d ' )

even faster growth resuited when both EDTAand phosphate were added. Perhaps EDTAfacilitated the uptake of phosphorus. Like otherchelators, EDTA is able to modify membranepermeability and osmoregulation by complexingof cations (Wetzel, 1975). The possibility thatEDTA acted as a detoxifieator. by reducing theconcentration of a toxic trace metal, is not verylikely because addition of trace metals, withoutEDTA. produced no toxic or inhibitory effect onalgal growth. Also the possibility that a tracemetal was the limiting nutrient for C. cotrxta andbecame more available after addition of EDTAdoes not seem very likely, because phosphatehad already been established as the limitingnutrient. We observed no growth stimulation ofC. comta in AU-P. Lake Water-t-EDTA andLake Water-l-trace metals.

The addition of EDTA alone, however,stimulated the growth of D. divergens from Junetill September. The inhibitory effect on thegrowth of this species of the single addition ofphosphate alone was a striking phenomenon(Table 4a). Rodhe (1948) was the first todemonstrate this effect, which has also beenobserved by Schelske & Stoermer (1972) andLehman (1976). However, Lehman (1976)found no inhibition of D. cylindricum and D.sociaie when Na,PO4 instead of K^HPOj wasadded. He concluded that potassium and notphosphorus was the inhibiting factor. We foundno difference using NaiPOj or K2HPO4; additionof either inhibited the growth of D. divergetis,but not when EDTA was given simultaneously.Reynolds (1986) suggested a high pH (>8.8) tobe the major factor inhibiting Chrysophyceaein phosphorus-deficient, oligotrophic lakes.However, in our experiments pH neverexceeded 8.5.

Chelators, like EDTA, reversibly combinewith cations in such a way that equilibriabetween free ions and soluble ion-chelate com-plexes are set up in the water (Stumm & Morgan,1981). In the absence of these chelators, cationsreadily precipitate as hydroxides, carbonates orphosphates. A soluble reservoir of chelator-cation complexes ensures a steady supply ofthese ions for the phytoplankton. Iron, in par-ticular, forms colloidal Fe'+ hydroxides and Fe'^phosphates of low solubility. Therefore, infor-mation on total iron concentration in lake water(0.03-0.21 mg 1"' of Fe in Lake Maarsseveen) isof doubtful value in relation to iron demands.

Natural communily hioassays 209

The availability of the element seems to dependon its chemical form (speciation). In addition,iron is one of the most important trace metals forphytoplankton, as it is a constituent of manyenzymes, cytochromes and some otherporphyrins. it is also important in chlorophyllbiosynthesis. Therefore the inhibited growth ofD. divergens after addition of phosphate mighthave been caused by iron becoming less availa-ble through the formation of insoluble FePO4.However, we are not sure that it is iron, ratherthan another trace metal, which is the limitingnutrient for D. divergens. Addition of iron with-out EDTA did not enhance growth rate. Proba-bly the iron is immediately inorganicallyprecipitated and inactivated. An absolute needof Dinobryon for iron was found by Lehman(1976). Addition of other trace metals, withoutEDTA, had no effect, A stimulation of algalgrowth in natural waters by addition of EDTAor Fe-EDTA was also reported by Lund,Jaworski & Butterwick (1975), Lin & Schelske(1979), Reynolds & Butterwick (1979) andDe Haan, Wanders & Moed (1982). However,De Haan et al. (1982) suggested in LakeTjeukemeer that the stimulating effect ot EDTAon phytoplankton growth might have been dueto detoxification of a heavy metal, most proba-bly copper.

In our bioassay experiments we excluded anyinfluence by zooplankton. In winter and spring,grazing was probably not an important factor,because of small zooplankton populations{Butter, 1981).

Succession during summer and autumn

In August/September of 1980 and 1981, A.formosa increased together with D. divergens.A.formo.sa was stimulated in its growth by phos-phate and D. divergens by EDTA. So, duringthe same period, the two species were limited bydifferent factors, A. fortnosa was attacked againby the fungus, though not as severely as duringthe spring. However, in summer about 70% ofthe colonies of A. fortnosa were covered with aVorticella species.

In September 1980, A. formosa and D.divergens were followed by growth of F. cro-tonensis and another, smaller growth of D.divergens and A. formosa. At that time, A. for-mosa and F. crotonensis were limited by phos-phorus and the growth rate of O. divergens was

again stimulated by EDTA addition. A secondappearance of F. crotonensis and a third of D.divergens were not observed in 1981. InAugust/September 1981, A. formosa reachedan abundance 4 times higher (600 ceils ml"')than in I980(l50ccllsml ') and dominated overa longer period. Perhaps competition for phos-phorus (both A. formosa and F. crototiensiswere phosphorus limited) was the reason why F.crotonensis did not come to dominate in Septem-ber/October 1981. In November/December, inboth years, A. formosa and S. astraea startedto grow again while F. crotonensis probablybecame light limited (we found no stimulationin growth with addition of any nutrient). InDecember, the two Stephanodiscus species werealready phosphorus limited, while A. formosadid not react to addition of phosphorus.

The patterns of succession observed fromJanuary 1980 till May 198U and from July 1980till F'ebruary 1981 were to a large extent sym-metrical. This is not surprising when we realizethat the controlling factors followed oppositetrends: declining phosphorus concentration withincreasing irradiance from winter to spring andincreasing phosphorus concentration anddecreasing irradiance from summer till late win-ter. Fungal parasitism on A. formosa was proba-bly an important controlling factor throughoutthe year. The fungus may be responsible for thecycles in abundance of this species with intervalsof 2-3 months (Van Donk & Ringelberg, 1983),

Acknowledgments

This investigation was financially supported bythe Foundation of Fundamental BiologicalResearch (BION), which is subsidized by TheNetherlands Organization of Advancement ofPure Research (ZWO), We thank Dr H, deHaan for reading tht manuscript critically andLeny M. Matulessya for typing the paper.

References

Butler M.E, (1981) Estimation of zooplanktonpopulations in Lake Maarsseveen: problems, pro-cedures and results, Hydrobiological Bulletin. 15,51-W.

Dc Haan H,. Wanders J.B,W. & Moed J,R, (1982)Multiple addition bioassay of Tjeukemeer water.Hydrobiotogia. 88, 233-244.

210 E. van Donk. A. Veen and J. Ringelberg

De Vries P.J,R, (19S3) Bioassays wiih StifieocloniiimKiitz (Chlorophyceae) to identify nitrogen andphosphorus limitations. Aquatic Botanv, 17.95-105.

Dc Vries P,J.R, (1985) Effm of phosphorus andnitrogen enrichment on Ihe yield of some strains ofStigeoctonium Kiitz (Chlorophyceae), FreshwaterBiology. 15. 95-103.

Dokulil M, & Skolaui C. (198(>) Succession ofphytoplankton in a deep stratifying lake; Mondsee.Austriii. Hydrobiologia. 138, 9-24,

Dorgelo J,. Van Donk E, & Dc Graaf BierhrouwerJ,M, (1981) The lake winter/spring hloom and suc-cession of di;itoms during lour years in LakeMaarsseveen, Verhcindlutigen der InterncitionalenVereinigung fiir Theoretische und AngewandteLimnologie. 21,938-947.

Frey B,E, & Small L,F, (1980) Effects of micro-nutrients and major nutrients on naturalphytoplankton popukitions. Journal of PlanktonResearch. I. 1-22,

Glooschensko W,A, & Avis C, (1973) Changes inspecies compositioEi of phytoplankton due toenrichment of N, P and Si of water from a NorthFlorida Lake. Hydrobiotogia, 42. 285-294,

Istvanovics V,, Vorbs L,, Herodck S,. Toth L.G. &Tatrai J, (1986) Changes of phosphorus and nitro-gen concentration and of phytoplanklon inenriched lake enclosures. Limnology andOceanography. 31. 798-811,

Lehman J,T, (1976) Ecological and nutritional studieson Dinohryon Ehrenb: Seasonal perodicity and thephosphate toxicity problem. Limnology andOceanography. 21. 646-658,

Lin C.K, & Schelske C L , (1979) Effects of nutrientenrichment, light intensity and temperature ongrowth of phytoplankton from Lake Huron, U,S.EPA. Duluth. Minnesota, Rep, EPA-6()0/3-79-049, 40. 231-293,

Lund J.W.G,, Jaworski G,H,M, & Butlerwick C-(197.S) Algal bioassay of waler from BlclhamTarn.English Lake District and the growth of pliinktonicdiatoms. Archiv fiir Hvdrobiologie Supplement.49, 46-^9.

Lund J.W,G, & Reynolds C,S, (1982) The develop-ment of operation of large limnetic enclosures inBlelham Tarn. English Lake District, and theircontribution to phytoplankton ecology. In; Pro-gress in Phycological Research. Vol, I (Eds E, E,Round and D, J. Chapman), Elsevier BiomcdiealPress N,V,

Maslin P.E, & Boles G.L, (1978) Use of multipleaddition bioassay to determine limiting nutrients inEagle Lake California, Hvdrobiologia, 58, 261-69,

Moss B, (1980) The plankton. In; Ecotogy ofFreshwa-ters. Blackwell Scientific Publications. Oxford.

Paasche E, (1978) Growih experiments with marineplankton algae; the role of "water quality" in speciessuccession. Mttteilungen der Internationale Ver-einigttng fiir Limnologie, 21, 521-527,

Reynolds C S , (1973) The seasonal periodieiiy ofplanktonic diatoms in a shallow eutrophic lake.Freshwater Biology. 3, 8^^110,

Reynolds C,S, (1986) Experimental manipulations ofthe phytoplankton periodicity in large limnetic

enclosures in Blelham Tarn, English Lake District.Hxdrobiologia. 138, 43-64.

ReynoldsC.S. & Butterwick C, (1979) Algal bioassaysof unfertilized and artificially fertilized lake water,maintained in Lund Tubes. Archiv fiir Hydro-biologie Supplement. 56. 166-83.

Rodhe W, (1948) Environmental requirements offreshwater plankton algae. Experimental studiesof the ecology of phytoplankton, Symbolae Botani-cal Upsaliensew 10, 1-49,

Schelske C L , , Rotham E D , . Stoermer E.E. & San-tiago M,A. (1974) Responses to phosphoruslimited Lake Michigan phytoplankton to factorialenrichments with nitrogen and phosphorus. Lim-nology ctnci Oceanogruphv. 19, 409-419,

Schelske C L , & Stoermer E,F, (1972) Phosphorus,silica and eulrophication of Lake Michigan. In:Nutrients and Eutrophication (Ed, G. E. Likens),American Society for Limnology and Oceanogra-phy Special Symposium I,

Schindler D,W, (1977) Evolution of phosphoruslimitation in lakes. Science, 195, 260-262.

Sokal R,R. & Rohlf F,J. (1969) Biometry. The princi-ples and practice of statistics in biological research.Freeman. San Francisco,

SommerU, & Stabel H,H. (19S3) Silicon consumptionand populaiion density changes of dominantplanktonic diatoms in Lake Constance, Journal ofEcology. 73. 119-1.30,

Stumm W, & Morgan J.J, (1981) Aquatic Chemistry.A n introduction emphasizing chemical equilibria innatural waters. Wiley. New York,

Swain W.R,. Lingeman R. & Heinis F. (1987) A char-acterizalion and description of the Maarsseveenlake system, Hydrobiotogicat Bulletin. 1\, 5-16.

Titman D. (1976) Ecological competition betweenalgae: experimental confirmation of resource-based competition theory. Science. 192, 463-465.

Titman D. (1977) Resource competition betweenplanktonie algae: an experimental and theoreticalapproach. Ecology. 58, 338-48,

Van der Does J, &'Klapwijk S, (1987) Effects ofphosphorus removal on the maximal algal growthin bioassay experiments with water from fourDutch lakes. Internationale Revue fur die GesamteHydrobiologie. 11, 27-39,

Van Donk E. (1983) Factors influencingphytoplankton growth and succession in lakeMaarsseveen. Thesis. University of Amsterdam,

Van Donk E, (1987) The water quality of the twoMaarsseveen Lukes in relation to their hydro-dynamics, Hydrobiotogicat Bulletin. 21, 17-24.

Van Donk E, & Ringelberg J. (1983) The effects offungal parasiiism on the suecession of diatoms inLake Maarsseveen (The Netherlands), FreshwaterBiology. 13,241-251,

Wetzel R. (1975) Seasonal suecession ofphyloplankton. In; Limnology (Ed, R, Wctzcl),W. B, Saunders Company,

Youngman R.E,. Johnson D, & Farley M R , (1976)Factors influencing phytoplankton growth and sue-cession in Farmoor Reservoir. Freshwater Biology.6, 253-263.

(Manuscript accepted 25 March 1988}