Hydrobiologia 415: 243–247, 1999.J. M. Caffrey, P. R. F. Barrett, M. T. Ferreira, I.S. Moreira, K. J. Murphy & P. M. Wade (eds),Biology, Ecology and Management of Aquatic Plants.© 1999Kluwer Academic Publishers. Printed in the Netherlands.

243

Succession ofEgeria densain a drinking water reservoirin Morbihan (France)

A. Dutartre1,∗, J. Haury2 & A. Jigorel31Cemagref, Unit´e Qualite des Eaux, 50, Avenue de Verdun, F-33612 Cestas Cedex, France2ENSA-INRA, 65, Route de Saint Brieuc, F-35042 Rennes Cedex, France3INSA, 20, Avenue des Buttes de Coesmes, CS 14315, F-35043 Rennes Cedex, France

Key words: Egeria densa, succession, water quality, management, reservoir, France

Abstract

Located in western France, the Pen Mur dam of Muzillac (Morbihan) is mainly used for the production of drinkingwater and for fishing. Over the past few years, the upstream part of this eutrophic waterbody has been invaded byEgeria densa. To define suitable methods for the management of this plant, a 1996 study focused on the morpho-metric characteristics of the reservoir, the quality of its water and sediments, the diversity of its vegetation andvarious aspects of colonisation byE. densa. Analyses of usage and pollution were also made, as well as a review ofthe techniques available to control the development of this species. Based on the results of this study, a managementplan was put forward. Studies in 1997 and 1998 showed a significant decline in the level of colonisation. Thiswas probably due to the substantial floods which uprooted submerged plants in winter, followed by massive iceformation during winter and warm temperatures in spring leading to a development of cyanobacteria which aredetrimental to macrophytes. Monitoring the competition betweenE. densaand the cyanobacteria could lead todetermination of the causes of the predominance of one or the other.

Introduction

Excessive growth of imported species is becoming anincreasing problem in France (Dutartre et al., 1996).Among these,Egeria densaPlanchon (Hydrochar-itaceae) is extremely widespread in tropical and sub-tropical parts of the world (Cook & Ürmi-König,1984). This underwater plant with its dense foliageand ramified stems can reach a length of three metres.It can grow on various substrates but prefers organicsilt.

It was first observed in France in 1960 : Feuillade(1961) reported it in a reservoir in the Manche (West-ern Normandy). Since then, the species has spreadalong the entire Atlantic coast.

The Pen Mur Reservoir in Muzillac belongs to theMorbihan County Council. It is mainly used for theproduction of drinking water, for fishing and leisureactivities.

∗ Author for correspondence

Its waters receive large inputs of nutrients fromintensive agriculture in the catchment area. Thus, eu-trophication causes the development of cyanobacteriawhich damage the water quality during the summer.Over the past few years, the reservoir has also beeninvaded byE. densa.

The purpose of the study was 1. to identify thepattern of its colonization in Pen Mur reservoir, 2.to define the methods of managing this plant in thebasin, and 3. to use this example to assess the necessityof integrated management of aquatic plants in otherreservoirs and lakes.

Site, material and methods

The Pen Mur reservoir is approximately 3 km long andgenerally less than 120 m wide. The maximum waterdepth is approximately 3 m, and the mean volume is900 000 m3.

The bathymetry and distribution of the plants weremonitored in 1996 at contact points distributed over

244

Table 1. Sucession ofE. densaon some transects (frequency in%)

Transect 1996 1997 1998

Upstream part of dam (no 4) 100 5 0

Larger part of dam (no 7′) 90 57 11

Immediately downstream of 90 55 0

larger part of dam (no 8)

Downstream part of dam (no 13) 9 0 0

17 cross sections. In 1997 and 1998, four referencetransects were studied in different parts of the dam.Observations were also made over the whole area ofthe reservoir to specify the biological diversity, checkthe distribution ofE. densaand assess the negativeeffects which it brings about.

Physical measurements and chemical sampleswere taken in the water during each site survey, includ-ing analyses of nutrients (nitrogen and phosphorus).Cores samples made it possible to determine the depthof the silt and to take samples for chemical and particlesize distribution analyses. Chemical analyses of theplants (organic matter, phosphorus, nitrates, Kjeldahlnitrogen) were also performed.

Results

The study confirmed the ecological value of the site:over 100 Spermatophyta taxons were observed, in-cluding species of special interest such asLuroniumnatans(L.) Rafin, which is now protected through-out Europe,Potentilla palustris(L.) Scop.,Littorellauniflora (L.) Ackerson,Eleocharis multicaulis(Sm.)Desv., E. acicularis (L.) Roemer & Schultes,TrapanatansL. andLudwigia palustris(L.) Elliott (Dutartreet al., 1997).

Colonization of the reservoir byE. densahadalready been quite substantial before 1996. In 1996,the species was found at 49% of 456 contact points(Figure 1a). Nevertheless, in relation to the depth(Figure 1b), this high frequency varied considerablybetween the upstream and downstream parts of thereservoir (Figure 1c). The first upstream section hada very small population, which rapidly increased toexceed 80% over approximately one quarter of thelength of the reservoir. It then dropped considerably,remaining generally below 20% in the middle part,and decreasing to 2–3% in the downstream part. The

maximum depth of occurrence ofE. densais 2.5 m,which could enable it to colonise practically the entirewaterbody. In 1997 and 1998,E. densahad decreasedconsiderably in the reservoir, probably due to highfloods and severe frost (15 cm of ice for several weeks)in winter 1997. Comparisons of results in some tran-sects on about 170 contact points show this importantdecrease (Table 1). The mean frequency ofE. densaon the dam decreased from 49% in 1996, 35% in 1997to 6% in 1998.

The decrease was much faster in the upstream andnarrowed part of the dam under direct influence ofwinter floods. In the larger part,E. densaremainedonly on old root crowns and small stems. In the down-stream and deeper part of the dam,E. densahadalready disappeared in 1997.

Based on literature (Cook & Ürmi-König, 1984;Getsinger & Dillon, 1984) and from analysing dataobtained in the current study (frequency and abond-ance ofE. densaon points contacts in the reservoir),estimates of the biomass ofE. densaindicated 50 tonsof dry weight for 1996, just over 13 tons for 1997 andbelow 1 ton in 1998.

The thickness of the reservoir bottom increasedmoving from upstream (20–40 cm) to downstream(about 1.5 m). The total volume of deposits was estim-ated to be 170 000 m3. The particle size distribution ofsediments indicated a distinct upstream/downstreamevolution. The predominantly fine sandy fraction up-stream was progressively replaced by a fine clay andsilt fraction downstream. Observations under the scan-ning electron microscope made it possible to differen-tiate the detrital fraction, due to erosion of the catch-ment area, from the biogenic fraction, produced by thereservoir’s aquatic flora (Jigorel & Bertru, 1993).

The mineral detrital fraction decreased from theupstream to the downstream part. The vertical vari-ations in the deposits indicated that the biogenic frac-tion had increased over the years, resulting in a recentrise in the trophic level of the reservoir, related to anincrease in the input of nutrients. The phosphorus con-tent of the sediments was higher in the upstream partof the reservoir, with heavy colonization ofE. densa(Figure 1f). The submerged plant community trappeda considerable amount of fine sediments, which wasconfirmed by the fact that over the last few yearsthe fluctuations in the suspended solids content atthe water treatment plant had been noticeably smaller(Compagnie Générale des Eaux, pers. comm.).

The nitrate content (approximately 30 mg/l) andphosphate content (up to nearly 0.06 mg/l) of the water

245

Figures 1 a–f.Environmental parameters in transects and characteristics of colonization and nutrients contents ofEgeria densa(1996). (a)Location of transects and characteristics ofE. densacolonization, (b) mean depth (m) of transects, (c) frequency ofE. densaon differentstransects (%), (d) chemical composition ofE. densa(P in% of dry matter NO3 in mg of wet plants /kg), (e) chlorophyll a (mg/m3)andoxidizability (mg/l) f - phosphorus in sediment (% of dry matter). (d) chemical composition ofE. densa(P in% of dry matter NO3 in mg of wetplants /kg), (e) chlorophyll a (mg/m3)and oxidizability (mg/l) (f) phosphorus in sediment (% of dry matter).

246

were relatively high and explained the summer de-velopment of cyanobacteria. The predominant specieswereMicrocystis flos-aquae(Wittr.) Kirchner andM.aeruginosaKütz. There were high concentrations ofchlorophylla which exceeded 120 mg/m3in the sum-mer. They resulted in a major drop in the transparencyof the water, which decreased from over 2 m in thespring to less than 0.4 m in summer. During the surveyin September 1996, the cyanobacteria were developingimmediately downstream of the dense communities ofE. densawith high concentrations of chlorophylla andincrease oxidizability (dissolved organic matter) (Fig-ure 1e). Over a distance of a few hundred metres, thechlorophyll a content in the water increased from 5–15 mg/m3 (content where dense plant communities ofE. densaoccurred) to 120–130 mg/m3 in the part ofthe reservoir whereE. densawas scarce.

Analyses of the plants (Figure 1d) indicated theclose relationships between the nutrient content of thespecies and of the water and sediments of the site.A sharp decrease in the phosphorus and nitrate con-tents downstream indicated the purifying capacity ofEgeria. The phosphorous concentration of the plantdecreased from 0.8 upstream to 0.56 downstream (inpercentage of dry matter), and the concentration innitrates dropped from 35 to less than 4 (in mg/kg)(Figure 1d).

Discussion

It is obvious that the dynamics of the colonization ofthe reservoir byE. densawere strongly influenced bywinter climatic events – rise in the waters of waterlevel in the tributaries and freezing of the open stretchof the watercourse.

In 1996 and 1997, the reservoir presented a sig-nificant upstream–downstream gradient which con-cerned the colonization ofE. densa, the chemical com-position and particle size distribution of the sediments,the phosphorus and nitrate content of the macrophytes,and the summer development of cyanobacteria.

Observations made in 1996 and 1997 seemed toindicate that competition between the macrophyte andthe planktonic populations resulted in a predomin-ance ofE. densain the upstream part of the reservoir,whereas the planktonic algae predominated down-stream. This longitudinal distribution of vegetal popu-lations was the result of the input of nutrients at theupstream end of the reservoir and of the elongatedmorphology of the basin.

The observations made in 1998 showed a veryclear decrease ofE. densacolonization. Only smallparts of plants remained in the larger part of the damunder weak influence of winter floods.

Greater depth and low transparency values weremajor limiting factors ofE. densadistribution on thedam. The organic matter content in superficial layer ofsediment were relatively high, up to 15–25%. In theupstream part of the reservoir, though the mineral de-trital fraction is upper than in downstream, the organicmatter content (about 15–20%) were not a limitingfactor forE. densagrowing.

Winter floods and ice cover in 1997 seemed to havealso injuredE. densapopulations.

In 1998, early development of cyanobacteria due tohigh water temperature in spring had probably causedadditional damage toE. densa.

After the observations of 1996, management pro-posals were prepared. At that time, eventual colon-ization of the entire reservoir byE. densaseemedpossible and would have created problems for anglersand biodiversity. However, no intervention was pro-posed at that time because the annual fluctuationswere unknown. Regular surveillance of colonizationby anglers according to a clearly defined protocol wasproposed. For betterin situ application of this pro-tocol, this surveillance was undertaken with help fromanglers during the 1997 and 1998 summers.

Harvesting and dredging ofE. densawould havebeen possible. Nevertheless, in such a eutrophic riverbasin with high nutrients concentrations in the wa-ter, it appeared to be preferable to useE. densabedsfor water cleaning in upper part of the dam, particu-larly for trapping fine sediment particles, in particularphosphorus and for restricting cyanobacterial bloomsby competition for nutrients. Consequently, it wasprobably advantageous to keep it as a natural filter inthe upstream part of the reservoir and to restrict anyintervention to a local level, by mechanical means,according to the requirements of anglers, over themiddle and downstream parts. At some later date, par-tial dredging of the upstream part of the pond wouldmake it possible to remove large quantities of phos-phorus contained in the sediments and in the plantsthemselves. Regular monitoring is recommended toimplement possible interventions at the right time.

Observations of the fastE. densadecrease in 1997and 1998 have changed our problem analysis, al-though the surveillance is going on. Could thisE.densaregression lead to total disappearance or is it

247

only a lower stage of a biological cycle as describedby Dr Jacoby (pers. comm.)?

The survival of this species is possible under icecover (Cook & Ürmi-König, 1984), but the optimaltemperature of growth is between 15 and 25◦C (e. g.Haramoto & Ikusima, 1988). This species is also adap-ted to various water qualities and can grow in nutrientenriched waters (Reddy et al., 1987). In laboratoryconditions, its nutrient uptake was relatively high. Itslight requirements are also relatively weak (Clayton J.,National Institute of Water and Atmospheric Research,Hamilton, New Zealand. pers. comm.). In laborat-ory experiments, Tanner et al. (1993) showed thatE.densais likely growing up to 25–35 g/m3of suspendedsolids. In New Zealand, this species is frequent andoften dominant in meso and eutrophic water bodies(Howard-Williams, 1993).

It seems that ice cover and winter flooding in 1997could not totally explain the fast decrease ofE. densain the Pen Mur dam. In this ecosystem, many othersecological conditions, such sediment and water qual-ity could be considered as influencing factors of thisdecrease.

The main phenomenon in this evolution was prob-ably the competition between cyanobacteria and themacrophyte. This competition vary between years andthe successive damage toE. densain 1997 (ice cover,flooding, high temperature in spring) may have al-lowed cyanobacteria to become dominante. On theother hand, allopathic substances from cyanobacteriacould have negative effects onE. densagrowth. Al-ternatively or additionally, the biological cycle ofE.densais perhaps also responsible for its decrease.

The future management proposals will depend onthe relative nuisance ofE. densaand cyanobacteriain the reservoir. For drinking water production fromthe reservoir, a water treatment station could take outcyanobacteria. However, the great variations of cy-anobacteria concentrations and high concentrations ofsuspended solids create some difficulties in treatment.According to the treatment manager, the winter con-centrations of suspended matter were relatively stableduring the years of the macrophyte invasion. Themacrophyte decrease could allow greater variationsof these suspended matter and subsequently increasetreatment difficulties.

On the other hand, harvesting ofE. densawhenmacrophyte beds were very dense and dredging reser-

voir bottom could reduce amounts of nutrient in thedam, particularly phosphorus. The actual decrease ofE. densaprevented this operation.

In the future, management of this reservoir couldintegrate the cyanobacteria/macrophyte competitionand regular surveillance will be continued to verifythis functional hypothesis.

Acknowledgements

This study, commissioned by the Morbihan CountyCouncil, was granted by the Morbihan County Wa-ter Board, the Brittany Regional Council and theLoire-Brittany Water Board. Gilles Boulande, fromthe local Anglers Association, Mr Dayon, a localfarmer, Myriam Jaffré, Aline Moreau, Gaëlle Royer,Christophe Laplace, Paul Simon and Olivier Touzotall participated in the site surveys.

References

Cook, C. D. K. & K. Urmi-König, 1984. A revision of the genusEgeria(Hydrocharitaceae). Aquat. Bot. 19: 73–96.

Dutartre, A., J. Haury, A. Jigorel & C. Laplace, 1997. Possibilités degestion de l’invasion de la retenue de Pen Mur (Muzillac, Morbi-han) par une plante aquatique exotique:Egeria densa.Cemagref,ENSA/INRA, Bordeaux Rapport pour le Conseil Général duMorbihan: 142 pp.

Dutartre, A., J. Haury & A. M. Planty-Tabbachi, 1997. Macro-phytes aquatiques et riverains introduits en France. Bull. Fr.Pêche Piscic. 344–345 (1–2): 407–426.

Feuillade, J., 1961. Une plante aquatique nouvelle pour la FranceElodea densa(Planch.) Casp. Bull. Soc. Linnéenne Normandie10 (2): 47–51.

Getsinger, K. D. & C. R. Dillon, 1984. Quiescence, growth andsenescence ofEgeria densain lake Marion. Aquat. Bot. 20: 329–338.

Haramoto, T. & I. Ikusima, 1988. Life cycle ofEgeria densaPlanch., an aquatic plant naturalized in Japan. Aquat. Bot. 30:389–403.

Howard-Williams, C., 1993. Processus of aquatic weed invasions:the New Zealand example. J. Aquat. Plant Mgmt 31: 17–23.

Jigorel, A. & G. Bertru, 1993. Endogenic development of sedimentsin a eutrophic lake. Hydrobiologia 268: 45–55.

Reddy, K. R., J. C. Tucker & W. F. Debusk, 1987. The role ofEgeria densain removing nitrogen and phosphorus from nutrientenriched waters. J. Aquat. Plant Mgmt. 25: 14–19.

Tanner, C. C., J. S. Clayton & R. D. S. Wells, 1993. Effects of sus-pended solids on the establishment on growth ofEgeria densa.Aquat. Bot. 45: 299–310.