LAKE EUTROPHICATION MODEL BASED ON THE IMPACT OF THE ZOOPLANKTON COMMUNITY ON PHYTOPLANKTON SUCCESSION

To estimate the long-term effects of ecosystem-manipulating lake restoration methods such as biomanipulation, we need a numerical model which can predict the succession of plankton community with sufficient accuracy. An accurate numerical model can be prepared based on an accurate data background. In this sense, two series of mesocosm (outdoor open-air pond) experiments that mimicked conditions in shallow and eutrophic water ecosystem were performed to clarify the effects of the zooplankton community on phytoplankton succession quantitatively. The experiments were carried out in both summer and winter periods. The results revealed that three types of zooplankton, namely rotifers, copepods, and cladocerans, have different effects on phytoplankton succession. Rotifers did not graze blue-green algae. Cladocerans suppress the growth of phytoplankton even for some blue-green algae effectively by their predation. The results of these experiments suggest that the amount of crustaceans, i.e., copepods and cladocerans, should be increased for the successful enforcement of biomanipulation. A model based on these quantitative outcomes, and its prediction of a biomanipulation for an eutrophic shallow lake is also mentioned

LAKE EUTROPHICATION MODEL BASED ON THE IMPACT OF THE ZOOPLANKTON COMMUNITY

ON PHYTOPLANKTON SUCCESSION

* Institute of Industrial Science, University of Tokyo, 4-6-1 Komaba, Meguro, Tokyo, 153-8505, Japan

**The University of the Air, 2-11, Wakaba, Mihama, Chiba City,Chiba, 261-8586, Japan,

***United Nations University,53-70, Jingu-Mae 5-chome, Shibuya, Tokyo 150-8925, Japan

ABSTRACT

To estimate the long-term effects of ecosystem-manipulating lake

restoration methods such as biomanipulation, we need a numerical model

which can predict the succession of plankton community with sufficient

accuracy An accurate numerical model can be prepared based on an

accurate data background In this sense, two series of mesocosm (outdoor

open-air pond) experiments that mimicked conditions in shallow and

eutrophic water ecosystem were performed to clarify the effects of the

zooplankton community on phytoplankton succession quantitatively The

experiments were carried out in both summer and winter periods The

results revealed that three types of zooplankton, namely rotifers, copepods,

and cladocerans, have different effects on phytoplankton succession

Rotifers did not graze blue-green algae Cladocerans suppress the growth

of phytoplankton even for some blue-green algae effectively by their

predation The results of these experiments suggest that the amount of

crustaceans, i.e., copepods and cladocerans, should be increased for the

successful enforcement of biomanipulation A model based on these

quantitative outcomes, and its prediction of a biomanipulation for an

eutrophic shallow lake is also mentioned

KEYWORDS

lake ecosystem model, mesocosm, biomanipulation

INTRODUCTION

People have always needed clear and safe water Lakes supply most of the water for human communities, so this desire appears in the growing concern with the purification of lakes In most cases, lake purification is equivalent to the prevention of lake eutrophication and consequent algal-bloom, which can cause many problems in water use Interest in the "biomanipulation" that utilizes zooplankton as algal predators to effect the sustainable purification of lakes has been

growing (Wright and Shapiro 1984; Hosper 1989; Shapiro 1990; Perrow et al 1997; Hosper and

Meijer 1999) In biomanipulation, the zooplankton-eating fish (planktivorous fish, e.g., bream) was removed to keep the high biomass of zooplankton, or fish-eating fish (piscivorous fish, e.g., pike)

was introduced to keep the low biomass of planktivorous fish

Some biomanipulation trials have resulted in clear water (e.g., Driessen et al 1993; Meijer et al 1994), while in other cases the suppression of algal growth was not observed (e.g., Pekkarinen

1990; Van Donk et al 1990; Van der Vlugt et al 1992; Bratil 1994) Moreover, the long-term

stability of biomanipulation remains questionable (Shapiro, 1990) The lake ecological model is one

of the most powerful tools to predict the long-term effects of biomanipulation Studies have tried to predict the effects of biomanipulation on the basis of lake ecological model calculations (e.g., Janse

et al.,1992; Scheffer,1989,1991) These calculations outlined some features of biomanipulation's

stability However, the rate process parameters of these models, especially for the dynamics of zooplankton and fish, were estimated on the basis of a limited amount of data The problem that we have to consider next is to estimate these model parameters more accurately

It is obvious that the succession of phytoplankton is strongly affected by the zooplankton species that exist in the habitat, and this effect is defined by the food preference of the zooplankton For successful modeling of the lake ecosystem, the impact of zooplankton on phytoplankton succession should be clarified The semi-large scale experiments (so called mesocosm experiments) are suitable for estimating the intra-species relationship because they show a lower variability of observation data than that of actual lakes and are more similar to the actual lakes than are the laboratory scale experiments

In this paper, we describe the results of our mesocosm observations especially focused on the impact of zooplankton on algal succession, and dominant material flow in water body estimated from the results Furthermore, an application of our eutrophication model for an eutrophic lake (Lake Balaton, Hungary) is also outlined

MATERIALS AND METHODS

The mesocosm experiments were

carried out using four open-air

circular ponds (4.0 m in diameter,

0.85 m in depth; Ponds #1 ~ #4)

located in Shida County, Shizuoka,

Japan A vertical sectional view of

one of the ponds is found in Fig 1

To prevent the water temperature from rising to undesirable levels, all ponds were covered with shading nets, which cut off 70% of sunlight Highly eutrophic sediment (total nitrogen (T-N): 3.1-3.9 mg·g-1-sediment, total phosphorus (T-P): 6.2-9.4 mg·g-1-sediment) and well water were introduced to the four ponds After five months of pre-observation, we used these ponds as experimental mesocosms

Experimental observations were carried out in summer (June 11, 1997 ~ Oct 13, 1997), and in winter (Nov 4, 1997 ~ Jan 12, 1998) At the start of the summer experiment, 60 g-wet·m-3 of

crucian carp (Carassius sp., body length: 8-10 cm, body weight: 15-30 g-wet) were stocked as

planktivorous fish in Ponds #1 and #3 During the experimental period, no water was supplied except rainfall The water temperature and irradiance of these two experiments are given in Fig 2

A little amount of macrophytes was founded on June 9, 1997, in Ponds #3 and #4 The species of

macrophytes were Hydrocaryaceae in Pond #3 and Hydrocaryaceae and Lemnaceae in Pond #4

During the pre-observation period, these ponds were rendered stable as regards the lack of inorganic

Fig 1 Vertical sectional view of the pond used in this study

Overflow

Sediment

Water

0.45 m

0.40 m 1.00 m

2.00 m 4.00 m

nitrogen Observations were thus initiated by loading inorganic nitrogen (ammonium sulfate) into the ponds The loading was performed in pulses (rise 0.56 mgN·L-1 in a load) every few days

0 50 100 150 200 250 300

0.0

10.0

20.0

30.0

0 20 40 60 80 100 120

Time after the start of experiment [day]

irradiance water temp.

Jun 11 Oct 13

100 200 300

0.0 10.0 20.0 30.0

0 10 20 30 40 50 60

Time after the start of experiment [day]

Winter Exp.

Fig 2 The variation of temperature and irradiance during each experiment

The sampling methods and its analysis were described in our previous paper (Suzuki et al., 2000) The water temperature was continuously measured in situ using a maximum-minimum thermometer (Digi MIII, NIHON KEIRYOKI KOGYO, Japan) The observed daylight time at the Omaezaki weather station (about 20 km away from the pond) was utilized for calculating the daylight fraction The latitude of the observation site (34˚47') was used to calculate the daily irradiance These calculations were done according to Stewart (1975), Groden (1977), and Jayaweera and Asaeda (1996)

RESULTS AND DISCUSSION

The observation results of

the summer experiments are

depicted in Fig 3 Rapid

growth of green algae was

observed immediately after

the start of inorganic nitrogen

loading in each pond The

dominant species of green

algae were Ankistrodesmus,

Scenedesmus, and/or Cruci-

genia Three types of

blue-green algae were

observed, namely, Micro-

Merismopedia The prompt

growth of rotifers was

observed in the ponds

without macrophytes (#1 and

#2) followed by the growth

of green algae The dominant

species of rotifers were

Fig 3 Dynamics of plankton communities in the summer experiment.

(Upper side: phytoplankton, Lower side: zooplankton)

0 100 200 300 400

0 100 200 300 400

0 100 200 300 400 500

Euglena s p.

Cryptomonas sp Merismopedia s p Phormidium s p Microcystis sp.

diatoms green algae

0 100 200 300 400

I-N Load

I-N Load

I-N Load

I-N Load

Time after Start of the Experiment [ day ]

Pond 1 with fish;

without macrophytes

Pond 3 with fish;

with macrophytes

Pond 4 without fish;

with macrophytes Pond 2

without fish;

without macrophytes

#1;

w/ Fish w/o Macrophyte

-#3;

w/ Fish w/ Macrophyte

#2;

w/o Fish w/o Macrophyte

#4;

w/o Fish w/ Macrophyte

0.0 0.5 1.0 1.5 2.0 2.5 3.00 20 40 60 80 100 120

0.0 0.5 1.0 1.5 2.0 2.5

0.0 0.5 1.0 1.5 2.0 2.5 3.0

Cladocera Copepodite Copepoda nauplii Rotatoria

0.0 0.5 1.0 1.5 2.0 2.5

I-N Load

I-N Load

I-N Load

I-N Load

Pond 1 with fish;

without macrophytes

Pond 3 with fish;

with macrophytes

Pond 4 without fish;

with macrophytes Pond 2

without fish;

without macrophytes

Time after Start of the Experiment [ day ]

#1;

w/ Fish w/o Macrophyte

-#3;

w/ Fish w/ Macrophyte

#2;

w/o Fish w/o Macrophyte

#4;

w/o Fish w/ Macrophyte

Trichocerca, Brachionus, Keratella, and Asplanchna A decrease in rotifers was observed in Pond

#1, probably the result of fish predation, while the zooplankton was dominated by rotifers in Pond

#2, in which the crucian carp had not be stocked In this pond, Microcystis dominates the phytoplankton In Pond #3 and #4, a temporal growth of Microcystis and Phormidium was observed,

respecitvely However, these species was eliminated after that, and the phytoplankton was

dominated by green algae In both ponds, large amount of cladoceran (Daphnia sp.) was observed

On the whole, some important relations between phytoplankton and zooplankton were clarified

The biomass of rotifers was correlated with Microcystis On the other hand, the biomass of Daphnia, which is abundant especially in Pond #4, was negatively correlated with Microcystis and

Phormidium Considering some literature (e.g., Matveev et al.,1994; Hlawa and Heerkloss 1994), it

was concluded that the Daphnia seems to graze not only green algae but also Microcystis and

Phormidium While, rotatoria seems not to graze Microcystis and Phormidium This was due to its

size In fact, the large size of blue-green algae was not grazed by Daphnia However, the large colonies of Microcystis were not observed in these experiments, and the filamentous of Phormidium

in this study seemed to be small enough to be grazed by Daphnia

The observation results of the

winter experiment are depicted

in Fig 4 The difference of the

phytoplankton succession

among each pond was more

obvious than that in the

summer experiment In Pond

#1, a relatively high biomass of

phytoplankton dominated by

green algae was observed The

biomass of phytoplankton in

Pond #2 was also high,

however, it was dominated by

Phormidium The growth of

phytoplankton in Pond #4 was

completely suppressed The

large number of Daphnia was

observed in Pond #4, indicating

that the strong suppression of

phytoplankton growth in this

pond might have been caused

by high grazing pressure of

Daphnia Almost only rotifers were grew in Pond #2, in which the domination of Phormidium was

observed These results consistent with the correlation observed in the summer experiment

Furthermore, the temporary growth of Microcystis was observed in Ponds #1 and #2, however, it

was eliminated gradually, probably as a response to low water temperature (Fig 2)

In the winter experiment, the water temperature was low and the true growth rate of phytoplankton was slow As a result, the grazing rate of zooplankton became a dominant factor to determine an apparent growth rate of phytoplankton That is why we can observe the impact of zooplankton community on phytoplankton succession was more obvious in the winter experiment than in the

Fig 4 Dynamics of plankton communities in the winter experiment.

(Upper side: phytoplankton, Lower side: zooplankton)

0 250 500 750

0 250 500 750

0 250 500 750 1000

Euglena Cryptomonas Merismopedia Phormidium Microcystis

diatom green

0 250 500 750

I-N Loading

I-N Loading

I-N Loading

I-N Loading

Time after Start of the Experiment [ day ]

Pond 1 with fish

Pond 3 with fish

Pond 4 without fish Pond 2

without fish

#1;

w/ Fish

#3;

w/ Fish

#2;

w/o Fish

#4;

w/o Fish

0.0 0.5 1.0 1.5 2.0 2.5 3.00 10 20 30 40 50 60

0.0 0.5 1.0 1.5 2.0 2.5

0.0 0.5 1.0 1.5 2.0 2.5 3.0

0.0 0.5 1.0 1.5 2.0 2.5

Cladocera Copepodite Copepoda nauplii Rotatoria

Time after Start of the Experiment [ day ]

I-N Loading

I-N Loading

I-N Loading

I-N Loading

P ond 1 with fish

P ond 3 with fish

P ond 4 without fish

P ond 2 without fish

#1;

w/ Fish

#2;

w/o Fish

summer experiment On the other hand, the selectivity of copepods on their food was not clarified

in this experiment

NUMERICAL MODEL

Differences of the characteristics of predation in zooplankton were found in the experiments described above From these results, it is clear that the functional aspects of zooplankton communities are different based on which species is dominant Coupled with some literature (e.g., Matveev et al., 1994; Price, 1988; DeMott 1989,1990; Vanderploeg 1990; Burns and Hegarty, 1994)

we concluded that the structure of model depicted in Fig 5 can describes the impact of zooplankton

on algal succession sufficiently

N IN

N DON

2

2

2

2

2

Di ssolve d

O rgani c Carbon

4

1

3

3 1

Zoobe nthos

P DOP P IP

Inorgan i c

C arbon

C RO

C CO

C CL

C PF

•

•

•

•

•

•

•

•

•

• Surface Water

Sediment

1

4

4

Fig 5 Flow scheme of the model Arrows with solid line mean the flow of carbon, nitrogen and phosphorus

Material or organism which was depicted in square with solid line means the state variable in the model And material or organism depicted in square with dashed line was not considered as variable in the model

On the other hand, each state variable (box) described in Fig 5 consists many species and growth stages of organisms Therefore, the kinetic parameters of them have a certain ranges of statistical variations Considering this feature, the parameters should be determined not as a certain values but

as a values with ranges The set membership approach is one of the most feasible methods to determine such parameter sets The detail description of this method was found in our previous work (Sagehashi et al., 2000)

One example that shows the

effectiveness of such lake model for the

prediction of the long-term effect of

biomanipulation is the application for the

Keszthely Basin, Lake Balaton (Hungary)

Figures 6 and 7 shows the model structure

and the calculation results Note that the

calculation results have some bands As

mentioned above, the calibrated

parameters based on actual observation

data have some statistical ranges

Therefore, we performed ten times of

model calculation with random sampling

of the parameters within the calibrated

rages The lines depicted in Fig 7 are thus

the average and standard variation of the

calculations From this figure, the effective

algal growth suppression, especially

blue-green algae, by biomanipulation (90%

removal of planktivorous fish) will take place

within first 2 years However, the algal growth

become initial states after 10 years This means

that only once biomanipulation is not sufficient

for the long-term stabilized lake restoration It

needs at least every two years removal of

planktivorous fish More information about this

modelling is shown in our previous paper

(Sagehashi et al., 2001)

As mentioned, the numerical model that

includes the impact of zooplankton community

on phytoplankton succession is a powerful tool

to predict the long-term effect of lake ecosystem

manipulating restoration method such as

biomanipulation The concern with the

symbiosis of human beings and nature has been

growing, therefore, such restoration methods

play more important role in conservation of lake

ecosystem It is also true that the dynamics

concerned with fish were not sufficient In

future works, the clarification of fish behavior

with high accuracy using mesocosms or such

controllable experimental apparatuses will be

desired

Fig 6 Schematic diagram of the Keszthely Basin model

Water

Sedi- ment

Algae Zoo- plankton

Dissolved

Interstitial

Pikeperch

Bream

P

P

P P

E&D

E&D E&D

E&D

E&D

M

M El

D:non-predatory Death E:Excretion

El:Elution M:Mineralization

P:Predation R:Resuspension S:Sedimentation U:Uptake

Fig 7 Fate prediction of the ecosystem in the

Keszthely Basin after 90% removal of bream Thick lines: average calculations; thin lines: average±SD (n =10)

0 10 20 30 40 50

0 100 200 300 400

0 200 400 600 800

0 20 40 60 80

0 20 40 60 80

0 200 400 600 800

0 50 100 150

0 0.05

0.

1 0.15

90% removal of bream

Calculation Year

CONCLUSION

The characteristics of three types of zooplankton communities, i.e., rotifers, copepods, and cladocerans, were revealed on the basis of mesocosm experiments The rotifers select against

blue-green algae (Microcystis and Phormidium) Cladocerans (Daphnia) suppress the growth of algae including small size Microcystis and Phormidium strongly by their predation These results

suggested that the biomass of cladocerans should be increased for successful biomanipulation On the basis of these results, a material flow structure required for the lake numerical model used for the prediction of ecosystem manipulating restoration methods (e.g biomanipulation) was proposed

ACKNOWLEDGMENT

The authors wish to thank Metocean Environment Inc (Tokyo, Japan) for their helpful cooperation

in our pond experiment Thanks are also due to Mr Yosuke Mochizuki and Mr Yugo Nomura for their assistance in the experiments The authors wish to thank Dr Károly Kutics at the Balaton Regional Cooperation and Development Co., (Hungary) for his helpful advice about the characteristics of Lake Balaton This research was supported in part by a Grant from the Kurita Water and Environment Foundation

REFERENCES

Bratli, J.L (1994) Water quality, phosphorus input reductions, analytical methods and lake internal/self-purification measures: a case study of lake Froylandsvatn, Norway Marine Pollution Bulletin 29(6-12), 435-438

Burns, C.W., and Hegarty, B (1994) Diet selection by copepods in the presence of cyanobacteria Journal of Plankton Research 16(12), 1671-1690

DeMott, W.R (1989) Optimal foraging theory as a predictor of chemically mediated food selection

by suspension-feeding copepods Limnology and Oceanography 34, 140-154

DeMott, W.R (1990) Retention efficiency, perceptual bias, and active choice as mechanisms of food selection by suspension-feeding zooplankton In Hughes, R.N (ed.) Behavioural Mechanisms of Food Selection Springer-Verlag, Berlin, 569-594

Driessen, O., Pex, B., and Tolkamp, H.H (1993) Restiration of a lake : First results and problems Verh Internat Verein Limnol 25, 617-620

Groden, T.W (1977) Modeling temperature and light adaptation of phytoplankton Report No.2, Center for Ecological Modeling, Rensselaer Polytechnic Institute, Troy, NY., 17

Hlawa, S and Heerkloss, R (1994) Experimental studies into the feeding biology of rotifers in brackish water Journal of Plankton Research 16(8), 1021-1038

Hosper, S.H (1989) Biomanipulation, new prospectives for restoration of shallow, eutrophic lakes

in the netherlands Hydrobiol Bull 23, 5-10

Hosper, S.H., and Meijer, M.L (1999) Biomanipulation in shallow lakes: Results of nine long-term case studies in the Netherlands Journal of Japan Society on Water Environment 22(1), 18-24 Janse J H., Aldenberg T and Kramer P R G (1992) A mathematical model of the phosphorus cycle

in Lake Loosdrecht and simulation of additional measures Hydrobiologia 233, 119-136

Jayaweera, M., and Asaeda, T (1996) Modeling of biomanipulation in shallow, eutrophic lakes: An application to Lake Bleiswijkse Zoom, the Netherlands Ecological Modelling 85, 113-127

Matveev, V., Matveev, L., Jones, G.J (1994) Study of the Ability of Daphnia carinata King to

control phytoplankton and resist cyanobacterial toxicity: Implications for biomanipulation in Australia Aust J Mar Frashwater Res 45, 889-904

Meijer, M.L., Jeppensen, E., Van Donk, E., Moss, B., Scheffer, M., Lammens, E.H.R.R., van Nes, E.H., van Berkum, J.A., de Jong, G.J., Faafeng, B.A., and Jensen, J.P (1994) Long-term responses to fish-stock reduction in small shallow lakes: interpretation of five-year results of four biomanipulation cases in The Netherlands and Denmark Hydrobiologia 275/276, 458-466 Pekkarinen, M (1990) Comprehensive survey of the hypertrophic lake tuusulanjarvi - nutrient loading; water quality and prospects of restoration Aqua Fennica 20(1), 13-25

Perrow, M.R., Meijer, M.L., Dawidowicz, P., and Coops, H (1997) Biomanipulation in shallow lakes: State of the art Hydrobiologia 342/343, 355-365

Price, H.J (1988) Feeding mechanisms in marine and freshwater zooplankton Bull Mar Sci., 43, 327-343

Sagehashi, M., Sakoda, A., and Suzuki, M., (2000) A predictive model of long-term stability after biomanipulation of shallow lakes Water Research 34(16), 4014-4028

Sagehashi, M., Sakoda, A., and Suzuki, M., (2001) A mathematical model of a shallow and eutrophic lake (the Keszthely Basin, Lake Balaton) and simulation of restorative manipulations Water Research 35(7), 1675-1686

Scheffer, M (1989) Alternative stable states in eutrophic, shallow freshwater systems: a minimal model Hydrobiological Bulletin 23, 73-83

Scheffer, M (1991) Fish and nutrients interplay determines algal biomass: a minimal model Oikos

62, 271-282

Shapiro, J (1990) Biomanipulation: the next phase - making it stable Hydrobiologia 200/201, 13-27

Stewart, D.C (1975) Mathematical modelling of the ecosystem of Lough Neagh Ph D dissertation, Queens University, Belfast

Suzuki, M., Sagehashi, M., and Sakoda, A., Modelling the structural dynamics of a shallow and eutrophic water ecosystem based on mesocosm observations Ecological Modelling 128, 221-243

Van Donk, E., Grimm, M.P., Gulati, R.D., Heuts, P.G.M., De Kloet, W.A., and Van Liere, E (1990) First attempt to apply whole-lake food-web manipulation on a large scale in the Netherlands Hydrobiologia 200/201, 291-302

Van der Vlugt, J.C., Walker, P.A., Van der Does, J., and Raat, A.J.P (1992) Fisheries management as

an additional lake restoration measure : biomanipulation scaling-up problems Hydrobiologia

233, 213-224

Vanderploeg, H.A (1990) Feeding mechanisms and particles selection in suspension-feeding zooplankton In Wotton, R.S (ed.) The Biology of Particles in Aquatic Systems CRC Press, Boston, Ann Arbor, Boca Raton, 183-212

Wright, D.I and Shapiro, J (1984) Nutrient reduction by biomanipulation: an unexpected phenomenon, and its possible cause Verh int Verein Limnol 22, 518-524

Full Contact Details

Masaki SAGEHASHI, Ph D (Engineering), Research Associate

Institute of Industrial Science, University of Tokyo, 4-6-1, Komaba, Meguro-ku, Tokyo, 153-8505, Japan

Telephone & Facsimile : +81 3 5452 6348, e-mail : sage@iis.u-tokyo.ac.jp