Zebra mussels (Dreissena polymorpha): a new perspective for water quality management

Hvdrobiologia 200/201: 437—450, 1990. R. D. Gulati, E. H. R. R. Lammens, M.-L. Meijer & E. van Donk feds.), Biomanipulation © 1990 Kluwer Academic Publishers. Printed in Betgium.

—

loot for Water Management.

437

Zebra mussels (Dreissena potymorpha): a new perspective for water qnality management H. H. Reeders & A. Bij de Vaate Institute for Inland Water Management and Waste Water Treatment, F.0. Box 17, 8200 AA Lelystad, The Netherlands

Key words: biomanipulation, lake restoration, filtration rate, Bivalvia, Dreissena polymorpha

Abstract In the evaluation of the role of lake restoration programmes in situ measurements of the filtration rate of the freshwater mussel Dreissena poÏymorpha have been performed in Lake Wolderwijd, The Nether lands. The filtration rate mainly depends on the suspended matter content of the water, and shows an inverse exponential relationship with this factor. The filtration activity is temperature indifferent between approx. 5 and 20 °C. At low temperatures the filtration rate drops abruptly, at high temperatures gradual inhibition occurs. The filtration rate shows a sigmoidal relation with the length of the mussel. The largest D. polymorpha have a diminished filtration rate compared to animals of smaller size. This might be a degenerative feature of the oldest mussels. In Lake Wolderwijd a population density of 675 per m2 is required to compensate phytoplankton growth by grazing. Manipulation of the size of the population can be executed by adding suitable substrates for the mussel.

Introduction The zebra mussel Dreissena poÏymorpha (Pail.) is filter-feeder. Particle retention is 1OO0/ for particles > 1 îim (Jorgensen et al., 1984). Food selection takes place by ciliary action inside the mussel (Ten Winkel & Davids, 1982). Rejected particles, e.g. silt and non-edible algae, are agglu tinated witli mucus and excreted as pellets, called pseudofaeces. When occurring in large densities a D. polymorpha population can play an important role in the circulation of matter in a lake eco system. Wiktor (1963) describes the siltation of the Szczecin Lagoon (Poland) due to the large quantities of pseudofaeces formed by the D. poly morpha population. In Lake Mikolajskie (Poland) this species filtered 8 0/0 of the annual primary production and contributed to 13°/s of the yearly

sedimentation by deposition of pseudofaeces (Stanczykowska et al., 1976). The turnover fre quency of the Lake Stregiel epilimnion (Poland) by the filtration activity of D. polymorpha amount ed to once per five days (Stanczykowska, 1968). Lake IJsselmeer and Lake Markermeer are fl1tered by the D. polymorpha population at least once or twice a month (Reeders etaÏ., 1989). Hence, a D. polpmorpha population can play an important role in the self purification of a lake (Lvova-Katchanova, 1971). Klee (1971) calls this species the largest purifying plant in the Bodensee, but after the collapse of the population its clearing impact dirninished (Walz, 1978k’). The possible use of D. potymorpha in bio manipulation programmes was suggested only recently (Richter, 1986; Reeders et al., 1989). The filtering capacity of a sufficiently large population

438 can accomplish a reduction in algal biomass and hence increase transparency. Stanczykowska et al. (1975) and Kryger & Riisgârd (1988) provide an overview of measurements of the filtration rate of D. polj’morpha in literature, to which the investi gations of Dorgelo & Smeenk (1988) and Reeders et al. (1989) can be added. The values for the filtration rate of mussels of 22 mm in size in the literature are derived from equations in Kryger & Riisgârd (1988) range from 2—287 ml mus sel 1 h Experimental conditions and setup can adversely affect the resuits, generally under estimating filtration rates (Møhlenberg & Riisgârd, 1979). Graphite suspensions reveal very low values (Hinz & Schei!, 1972; Benedens & Hinz, 1980). The kind of alga! cu!ture used in the experiment influences the fi!tration rate (Morton, 1971). Certain flage!!ates can have a toxic effect on mussels (Shumway & Cucci, 1987). The in experiments widely used Chioreïla caused reduced filtration rates of Mytitus edulis (Davids, 1964). Previously it was demonstrated that D. poly morpha showed unnatura! ifitration rates under laboratory conditions (Reeders et al., 1989). for a proper translation to the practice of water quality management in situ measurements of fi!tration rate are advocated. These are on!y known to have been performed by Stanczykowska et al. (1975), Mikheev (1967) and Reeders et al. (1989). Since va!ues for fi!tration rate measured with a pre vious!y used method (Reeders et al., 1989) needed a correction, measurements in present research were conducted with an improved version. The re!ation of filtration rate with temperature, particle content, alga! composition of the water, and size of the musse! were studied in the hypertrophic Lake Wolderwijd, The Nether!ands (Fig. 1), one of the polder border !akes. —

-

.

Methods The measurements of the fi!tration rate were carried out in Lake Wo!derwijd, in a dammed off experimental pond (25 x 25 m). Since exchange of water through the dam was possib!e the condi tions in the pond reflect those in Lake Wolderwijd.

Wave action, that would disturb the measure ments, was prevented by the dam. In the pond a working platform was constructed. Figure 2 shows the apparatus used for the measurements, fixed to the rai!ing of the platform. The position to the water level in the pond can be adjusted. At the start of an experiment the perspex tube is lowered and pressed to the rubber bottom plate (fig. 2B). An in situ enclosure of the centra!ly !ocated mussels results. Transportation through the air, severely disturbing the mussels (Reeders et al., 1989), was prevented in this way. By adding pond water up to the overfiow of the tube an experimental volume of 16.0! was attained. The rubber p!ate prevented exchange of the enc!osed water with pond water. A simplified version of the apparatus without musse!s was used as a blank (in dup!o). Every 10 minutes, for a period of 1 h, a water samp!e of 50 ml was taken to which a few drops of forma!in were added to stop phytoplankton growth. Before sampling the contents of the perspex tube were thoroughly mixed. Extinction of the samples was measured the same day or the day after the experiment at 665 nm (chlorophyl) in a 10 cm cuvet. Extinction is a measure for the algal concentration C. Rewriting the common!y used equation for the calculation of filtration rate (Coughlan, 1969) in terms of the regression of !nC against time (lnC = b + at) yields (Reeders etal., 1989): fr.

=

v*

—

n

t— a

+

a’)

in which V = volume (16.0 1), n number of mus sels, a and a’ = regression coefficients with and without (blank) musse!s respective!y. Measure ment ‘flutter’ is leve!led out by this method, increasing the reproducibility. Measurements were conducted from April—November 1988. During each experiment the temperature, oxygen and suspended matter content (dry matter, ashes, ch!orophyl) were measured. Phytoplankton samp!es were taken from the experimen tal pond and Lake Wo!derwijd every fortnight. The zebra mussels were collected

439

fig. 1. The Ijsselmeer-area, showing the polders and border lakes. The measurements of the filtration rate of Dreissena polymorpha were conducted in Lake Wolderwijd.

1’-

H

g

d

b

h —

Iz

0e0

h

fig. 2. The apparatus used for measuring filtration rates, fixed to the railing of the working platform; (A) in resting position, (B) during an experiment. a = cage with mussels; b, c = stainless steel screw-threaded bars; d = perspex tube; e = top lid; f = rubber bottom plate; g = churn; h = pond water level; i = overfiow, incubated water level; k = sampling tube.

a

-‘

o o

e

iL

_

441 in Lake IJsselmeer (Enkhuizerzand) (Fig. 1). from April—September the measurements were carried out with 4 size classes in triplicate, in October—November with 6 classes in duplo. The number of mussels in an experiment varied between 70—200, depending 011 size. Two size classes, 18 mm and 22 mm, were maintained during each experiment.

Resuits and discussion Figure 3 shows the phytoplankton composition of Lake Wolderwijd and the experimental pond during 1988. The patterns are very similar: the conditions in the pond reflect those of the lake. In early spring diatoms, mainly Diatoma elongatum, dominated the phytoplankton spectrum. Green algae were dominant in the rest of the year. These were mainly Ankistrodesmus falcatus and, to a lesser extent, Scenedesmus spp. Cyanobacteria showed a distinct peak in spring of Oscillatoria agardhii and Oscillatoria redekei, and a smaller peak at the end of the summer of Merismopedia glauca. Aphanizomenon flos-aquae is the most common cyanobacterium in autumn. figure 4 shows the relation between filtration rate (ml mus sel 1 h 1) and dry matter content of the water (mg 1- 1) as found for D. polymorpha of 18 mm and 22 mm. In accordance with previous research (Reeders et al., 1989) distinction can be made between measurements in the summer period (April—October, temperature 10 °C) and the winter period (November, temperatures 5 °C): at equal dry matter content lower filtra tion rates are measured in winter than in summer. Apparently, temperature sets a certain gross level to the filtration rate in both periods, so that summer and winter periods should be distin guished and treated separately in analyses with respect to this factor. Figure 5 shows the relation between the filtration rate of D. poÏymorpha (18 mm and 22 mm) and temperature. Despite the considerable range in temperature (10—21 °C)no significant relation could be found between the filtration rate and water temperature in the sum mer period (April—October). Although tempera —

ture determines the level of filtration rate between seasons, it does not significantly effect the filtra tion rate within a season. No significant relation could be found between the filtration rate of D. polymorpha and dry matter content (summer period). Combination of temperature and dry matter content in a multiple regression analysis of filtration rate neither provided a significant model. The number of data in the winter period is too small to allow analysis. A comparative experiment leamed that filtra tion rates obtained with the present method can be directly compared to the, corrected (f= 2.17), results from the previously operated method (Reeders et al., 1989). fig. 6 shows the relation of filtration rate of D. polyrnorpha (22 mm) and dry matter content for the previous (1985) and present (1988) measurements combined. The range in dry matter content in 1988 is small (17—36 mg 1—1) compared to 1985 (5—79 mg l1). The presently measured filtration rates fail in the flat section of the curve, based on data from 1985, which explains why no significant relation can be found. For the measurements during the summer period of 1985 and 1988 (temperature > 10 °C) the curve is described by the equation: fr.

=

187.1 eo037’

(R2=O.70;p

fig. 6. The relation between dry matter content of the water and the ifitration fate of Dreissenapolymorpha of 22 mm length for the measurements of 1985 (Reeders ei al., 1989) and 1988 combined. For the summer period the equation yields: fr. = 187.1 e°°37’.

regulating its pseudofaeces-production (Foster Smith, 1975). D.poÏymorpha however shows a combination of both strategies. Besides regulating its filtration rate the pseudofaeces production linearly increases with higher concentrations (Reeders, 1989). D. poly morpha is an opportunistic species colonising ver2 different habitats, like rivers, lakes and ditches. The ability to exhibit both regulatory mechanisms enables a flexible response to a variety of environ mental conditions, and explains why it can thrive in the silty river Rhine as well as the dear Lake Constance. Between 10 and 20 0C temperature hardly affects the filtration activity of D. polymorpha. Temperature dependence varies between bivalve species: the filtration rate of Crassostrea and Mytilus is likewise hardly affected by temperature in the range 10—20 °C, while Venerupis decussata and Mercenaria mercenaria are very sensitive to temperature differences (Walne, 1972). In Novem ber 1988 the temperature dropped sharply to 5 °C

within a week, a common feature for the shallow Dutch lakes. Figure 4 and especially fig. 5B in dicate that below approximately 5 °C the filtra tion rate decreases rapidly. The pseudofaeces production activity of D. polymorpha revealed the same lower threshold (Reeders, 1989). Observa tions on growth confirm this feature: growth starts at ± 6 °C (Bij de Vaate, in prep.). As observed before, different seasons can thus be distinguished, in which temperature sets a gross level to filtration rate. Above 20 °C temperature appears to become inhibitory for the filtration activity of D. polymorpha: the filtration rate at 21 °C tends to be lower. Higher temperatures were not recorded in 1988. These features give rise to a model for the relation of temperature and filtration rate of D. polymorpha, shown in Fig. 7. This resembles the optimum model for ingestion rate described by Walz (197$”), but differs with respect to the temperature-indifferent section of approx. 5—20 °C. The relation between size of the mussel, as shell

.46 Dreissena polymorpha

1-.

0

t!)

E E G)

4-

0 4-

cu

4-

0

1

1

1

1

5

10

15

20

25

30

Temperature (°C) fig. 7. A model for the relation between water temperature and the filtration rate of Dreissena polymorpha.

mgth L (mm), and fittration rate is shown in ig. 8. All measurements from April to October 988 are combined, which is justified, since filtra [on rate shows no significant relation with either mperature or dry matter content in this period. ie filtration rate of D. polymorpha 22 mm) how a diminished filtration rate. This gives rise o a sigmoidal shape of the curve, which yields: 15.43 0.293

+

=

O.59;p

DW= 6.47 10 L2875

(R2

=

0.99)

ADW= 6.35 10 L2806

(R2

=

0.98)

The relations between DW(mg) resp. ADW(mg) and filtration rate of D. poÏymorpha take the form off.r. = aW”, and yield:

52.38 e_O367L (R2

because analysis of the mussels, originating from a relatively clean area, showed low accumulation levels op pollutants (Reeders, 1989). The relations of shell length L (mm) with dry weight DW (mg) and ash free dry weight ADW (mg) of D. polymorpha were determined in May 1988, and yield:

<

0.001)

ie reduction in filtration rate of the largest, and ldest, mussels is most probably a degenerative ature of age. This was found for the pseudo aeces production activity as well (Reeders, 1989), nd has not been observed sofar. It is known that .ccumulation of pollutants can show similar ffects (P. de Kock, pers. comm.). In this case Dxical interference can 5e excluded, however,

f.r.

=

5.132 DW608 (p