Sj. P. KLAPWIJK EUTROPHICATION of SURFACE WATERS in the \ ^ JÏ ** DUTCH POLDER LANDSCAPE

EUTROPHICATION OF SURFACE WATERS IN THE DUTCH POLDER LANDSCAPE

Sj. P. KLAPWIJK EUTROPHICATION OF SURFACE WATERS IN THE \ ^ JÏ ** DUTCH POLDER LANDSCAPE

Yfjl /■» ,U' EUTROPHICATION OF SURFACE WATERS IN THE DUTCH POLDER LANDSCAPE

PROEFSCHRIFT ter verkrijging van de graad van doctor aan de Technische Universiteit Delft, op gezag van de Rector Magnificus, Prof. Dr. J. M. Dirken, in het openbaar te verdedigen ten overstaan van een commissie aangewezen door het college van Dekanen, op 23 juni 1988 te 14.00 uur

door SJOERD PIETER KLAPWIJK geboren te Sappemeer, doctorandus in de Wiskunde en Natuurwetenschappen

1988 HOOGHEEMRAADSCHAP VAN RIJNLAND LEIDEN Dit proefschrift is goedgekeurd door de promotor Prof. Dr. M. Donze.

PROMOTIECOMMISSIE :

Plv. Rector Magnificus: Prof. Drs. P.J. van der Berg

Promotor : Prof.Dr. M. Donze

Overige leden : Prof.Dr.Ir. J.C. van Dam (Technische Universiteit Delft) : Prof.Ir. J.H. Kop (Technische Universiteit Delft) : Prof.Dr. W.H.O. Ernst (Vrije Universiteit Amsterdam) : Prof.Dr. M. Vroman (Vrije Universiteit Amsterdam) Gasten : Dr. R.M.M. Roijackers (Landbouwuniversiteit Wageningen) : Dr. P.J.R. de Vries (Unie van Waterschappen) STELLINGEN

1. De stelling van Schmidt-van Dorp, dat "niet is aangetoond, dat de toename van de eutrofiëring van het Nederlandse oppervlakte water, die in de laatste decennia plaatsvond, veroorzaakt is door een grotere fosfaattoevoer aan het water", wordt door historische gegevens tegengesproken.

Stelling bij A.D. Schmidt-van Dorp, 1978. De eutrofië ring van ondiepe meren in Rijnland (Holland). Proef schrift R.U. Utrecht.

Dit proefschrift.

2. Chemische extractiemethoden kunnen bioassays met sediment niet vervangen om het voor algen beschikbaar sedimentfosfaat te be palen .

Dit proefschrift.

3. Ten onrechte menen Riegman en Mur dat opnamekinetiek-experi menten defintief uitsluitsel geven over de aard van de nutriënt- limitatie van de natuurlijke fytoplanktonpopulatie.

R. Riegman & L.R. Mur, 1986. Phytoplankton growth and phosphate uptake (for P limitation) by natural phytoplankton populations from the Loosdrecht lakes (The Netherlands). Limnol. Oceanogr. 31: 983-988.

Dit proefschrift.

4. Voor een grootschalige en effectieve bestrijding van de eutrofië ring in het westen van Nederland moet het fosfaatgehalte in de Rijn niet met de helft maar tot minder dan een kwart worden te ruggebracht. Verder dient, naast het verwijderen van fosfaten op zuiveringsinstallaties, de fosfaatuitstoot uit de land- en tuinbouw drastisch te worden verminderd.

Aktionsprogramm "Rhein", 1987. Internationale Kommis- sion zum Schutze des Rheins gegen Verunreinigung, Strassburg.

Notitie Fosfaatbeperkende maatregelen Nederlandse op pervlaktewateren. Tweede Kamer, vergaderjaar 1987- 1988, 20342, nr. 2.

Dit proefschrift.

5. De aanwezigheid van het pigment chorofyl-b bij de in zakpijpen levende Prochloron didemni R.A. Lewin en de door Burger-Wiers- ma et al. in de Loosdrechtse plassen gevonden draadalg hoeft niet te betekenen dat beide organismen fylogenetisch nauw ver want zijn.

T. Burger-Wiersma, M. Veenhuis, H.J. Korthals, C.C.M. van de Wiel & L.R. Mur, 1986. A new prokaryote containing chlorophylls a and b. Nature 320: 262-264. 6. Het is onjuist om in het laboratorium aangetoonde effecten van lage zuurstofgehalten op eieren en larven van vissen zonder meer te vertalen naar absolute normen voor het minimum zuurstofge halte in oppervlaktewater gedurende het gehele jaar.

J.S. Alabaster & R. Lloyd, 1980. Water quality criteria for freshwater fish. Butterworths, London.

Indicatief Meerjaren Programma Water 1985-1989. Tweede Kamer, vergaderjaar 1984-1985, 19153, nr. 2.

7. In veel polders in het westen van ons land wordt meer organi sche stof geproduceerd door kroos en kroosvaren dan door de mens.

8. De aanleg van drinkwaterleiding heeft op veel plaatsen in ons land geleid tot een sterke achteruitgang in de kwaliteit van het oppervlaktewater.

H. van Zon, 1986. Een zeer onfrisse geschiedenis. Studies over niet-industriële vervuiling in Nederland, 1850-1920. Proefschrift R.U. Groningen.

S.P. Klapwijk & C.J. Smit, 1988. Gouda en de water kwaliteit van Rijnland. In: Ludy Giebels (red.): Water beweging rond Gouda van ca. 1100 tot heden: geschie denis van Rijnlands waterstaat tussen IJssel en Gouwe, Leiden.

9. De filosofie van voortschrijdende normstelling, neergelegd in het IMP-Milieubeheer, waarbij normen steeds moeten worden ver scherpt om druk op de sanering te houden, werkt ontmoedigend voor degenen die de saneringsmaatregelen moeten uitvoeren en bekostigen en heeft daardoor een averechts effect.

Indicatief Meerjaren Programma Milieubeheer 1986-1990. Tweede Kamer, vergaderjaar 1985-1986, 19204, nr. 2.

10. De complexiteit van de voorgestelde structuur voor normering van waterbodems maakt dat men de daaruit voortvloeiende normen voorlopig het beste naast zich neer kan leggen.

Interimrapport van de werkgroep Normering, 1986. Onderwaterbodem overleg RWS-DGMH.

11. Ten behoeve van de zogenaamde multifunctionele bodemkwaliteit ware het beter volkstuinen te gebruiken als referentie in plaats van veldgegevens uit landelijke gebieden en als schoon beschouw de waterbodems.

Voortgangsrapportage Milieuprogramma 1988-1991. Tweede Kamer, vergaderjaar 1987-1988, 20202, nr. 2.

12. Om beleidsmakers, bestuurders en politici er van te doordringen dat zeer lage fosfaatgehalten nodig zijn om de algengroei te be perken, verdient het aanbeveling om fosfaatconcentraties voortaan in microgrammen in plaats van in milligrammen uit te drukken. 13. Het aanprijzen van fosfaatvrije wasmiddelen met het toevoegsel "Groen" lijkt op anti-reclame.

14. Het uitzetten van graskarpers zou vergunningsplichtig moeten zijn in het kader van de Wet verontreiniging oppervlaktewateren.

15. Als waterschappen een voortrekkersrol moeten gaan vervullen bij integraal waterbeheer, zal hun politieke basis moeten worden ver breed.

Naar een samenhangend oppervlaktewaterbeheer, 1987. Unie van Waterschappen, 's-Gravenhage.

16. Gezond boerenverstand vormt de kracht èn de zwakte van het waterschapsbestel.

17. De stormvloedkering in de Oosterschelde laat zien dat waterstaat kundige (kust)werken geen kunstwerken maar -soms- kunstige werken zijn.

18. Alleen al om bibliografische redenen kunnen vrouwelijke auteurs beter onder hun eigen naam en niet onder die van hun echtge noot publiceren.

19. De gewoonte om bij echtscheidingen niet alleen de boedel maar ook de ouderlijke macht te scheiden is vaak niet in het belang van de kinderen, de ouders en de gemeenschap.

20. Bij alle rechten aan de doctorstitel verbonden behoort blijkens het promotieprotocol van de T.U. Delft ook de plicht voor de pas gepromoveerde om aan het slot van de promotieplechtigheid niets terug te zeggen. Dit staat in te schril contrast met de verwach ting bij de daaraan voorafgaande verdediging van het proef schrift .

Stellingen behorende bij het proefschrift "Eutrophication of surface waters in the Dutch polder landscape".

Sjoerd Pieter Klapwijk 23 juni 1988. Aan mijn vader ds. P. Klapwijk (1906-1972)

Voor Floor CIP-DATA Koninklijke Bibliotheek, Den Haag Klapwijk, Sjoerd Pieter Eutrophication of surface waters in the Dutch polder landscape / Sjoerd Pieter Klapwijk ; [ill. Anne Post toe Slooten]. - Leiden : Hoogheemraadschap van Rijnland. - 111. Thesis Delft. - With ref. - With summary in Dutch. ISBN 90-72381-02-5 SISO 568 UDC 504.45.058(492)(043.3) Subject heading: eutrophication ; surface waters ; The Netherlands. -7-

CONTENTS Chapter Page

1. General introduction 9 PART A: PHYTOPLANKTON 17

2. Introduction phytoplankton 19 3. Biological assessment of the water quality in South Holland (The Netherlands) 23 4. Comparison of historical and recent data on hydrochemistry and phytoplankton in the Rijnland area (The Netherlands) 59 5. Dose-effect relationships between phosphorus concentration and phytoplankton biomass in the Reeuwijk Lakes (The Netherlands) 83

PART B: SEDIMENTS 91

6. Introduction sediments 93 7. Available phosphorus in lake sediments in the Netherlands 99 8. Application of derivative spectroscopy in bioassays estima ting algal available phosphate in lake sediments 111 9. Available phosphorus in the sediments of eight lakes in the Netherlands 119

PART C: BIOASSAYS 129

10. Introduction bioassays 131 11. Effects of phosphorus removal on the maximal algal growth in bioassay experiments with water from four Dutch lakes 137 12. Algal growth potential tests and limiting nutrients in the Rijnland Waterboard area (The Netherlands) 151 13. Bioassays using Stigeoclonium tenue Kütz. and Scenedesmus quadricauda (Turp.) Bréb. as testorganisms; a comparative study 169 14. Comparison of different methods to determine growth limi ting factors for phytoplankton in the Reeuwijk Lakes (The Netherlands) 179 15. Summary and conclusions 203 16. Samenvatting en conclusies 211

Curriculum vitae 220

Publikaties 220

Dankwoord 225

Colofon 227 -9-

CHAPTER 1: GENERAL INTRODUCTION

"Eutrophication, which may be natural or man-made, is the response in water to overenrichment by nutrients, particularly phosphorus and nitrogen."

OECD, 1982. Eutrophication of waters; monitoring, assessment and control, p. 17. Organisation for Economic Co-operation and Development, Paris. -10-

GENERAL INTRODUCTION Eutrophication, defined by Parma (1980) as "the process in water during which the factors stimulating autotrophic production becomes optimal" but mostly known as "the overenrichment of surface water with nutrients, having strong impact on abiotic and biotic factors in aquatic ecosystems (algal blooms, anaerobiosis, fish kills)", is essen tially a natural phenomenon. Naturally the loading of surface waters with nutrients happens through erosion, through deposition of animal faeces, as from bird colonies, plant debris, etc. In this century the process of eutrophication was accelerated through human activities. The human population increased, the construction of drinking-water and sewage systems led to large-scale discharges of wastewater (van Zon, 1986), agricultural methods changed (e.g. by the use of fertili zers and by not recycling all manure) and after 1950 the use of phos phate-containing detergents increased rapidly. As in other parts of the world (well-known examples are the Great Lakes in North Ame rica), most lakes in Europe deteriorated, like the Swiss and Italian Alp lakes (Thomas, 1953; Vollen weider, 1968), many East-European lakes (Straskraba & Straskrabova, 1969), the Grosser Plöner See (Ohle, 1955), several Danish lakes (Berg et al. , 1958; Johnsen et al., 1962) and the previous oligotrophic Swedish lake Trummen (Andersson et al., 1973). The water in the delta of the river Rhine was probably always moderately rich in nutrients the river imported. The description of dominant phytoplankton species in the Rhine delta from Lauterborn (1918) at the beginning of this century shows that the water was al ready rather eutrophic then. Peelen (1975) comparing the old plankton data of several investigators with recent observations found that the plankton composition of the river Rhine between the beginning of the century and ca. 1973 had changed somewhat, but that the saprobic level had not moved. However he did find an increase in the amount of plankton organisms, probably caused by increasing eutrophication and/or lengthening of the residence times in the Rhine branches. In the Netherlands especially Golterman (1965, 1970a,b, 1971, 1972, 1973a,b) called attention to the rapidly increasing eutrophication in the sixties and early seventies. He initiated a special steering com mittee of the Royal Dutch Chemical Society to study this problem. As a result a report was compiled on the causes and consequences of the eutrophication problem in the Netherlands. It suggested measures to be taken to reduce the phosphate-loading of the Dutch, surface waters (Golterman, 1976). This study gave occasion to the so-called "Fosfa- tennota" (Phosphate Report) in 1979 by the ministries of Public Health & Environmental Hygiene and Public Transport & Public Works (1979), explaining the governmental policy concerning eutrophication control. It stated that the reduction of the phoshate levels had to be achieved by phosphate-removal at sewage treatment plants and by replacement of phosphates in detergents. In the Rijnland waterboard area the eutrophication problem has been studied from 1973-1976 by Schmidt-van Dorp (1975, 1978), who found that due to the high levels of total and inorganic phosphate in most of the lakes phosphorus was not a limiting nutrient for algal growth any more. She suggested that in this area more attention should be paid to nitrogen reduction. Nevertheless it was generally -1.1- concluded (Klapwijk, 1977; Hosper, 1978; Golterman, 1979) that eutro phication could be better combatted by P-reduction than by N-reduction for the following reasons, as indicated in the postscriptum (p. 241-242) to Schmidt-van Dorp (1978):

1. Nitrogen is discharged more diffuse than phosphate e.g. by agri culture. It is therefore more difficult to control.

2. In nearly all surface waters nitrogen-fixing blue-green algae oc cur, capable to use free nitrogen for their growth.

It was feared that N-reduction would stimulate a change from nonfixing blue-greens to nitrogen fixing blue-green algae, which is also confirmed by laboratory experiments. This would not contribute very much to the solution of the eutrophication problem. Moreover, it should always be kept in mind that nitrogen has become the primary limiting nutrient in surface waters because the phosphorus load has relatively more increased than the nitrogen load (Thomas, 1953). It seems more sensible to reduce the nutrient which is relatively most increased by human activities. Therefore the Waterboard of Rijnland carried out a large-scale phosphate-removal experiment at three sewage treatment plants (Gouda, Bodegraven and Nieuwveen) from 1979 to 1982 in order to reduce eutrophication in the lakes in the south-east of its area (Klap wijk, 1977). Because of the complexity of the eutrophication problem, extensive limnological research was accompanying this experiment to monitor the effects of phosphate-removal at the three sewage treat-1 ment plants (Klapwijk, 1981; Hoogheemraadschap van Rijnland, 1984; van der Does & Klapwijk, 1985, 1987). Parts of this research are en closed in this thesis (Chapters 7, 8, 9 and 11). Since it had to be concluded from this experiment that phosphate- removal at sewage treatment plants will not lead to an immediate de crease of algal biomass in the lakes, which form part of the basin system of Rijnland, later the attention has been focussed on the more isolated polder lakes in which the execution of complete phosphorus reduction measures can be more rigorous and might be more effective. Therefore lake restoration projects have been developed for the Reeu- wijk, the Nieuwkoop and the Langeraar lakes (Geerplas), in which all phosphate sources will be reduced simultaneously. The Reeuwijk lakes were studied from 1983-1985 to collect back ground information on the water quality in the lakes. Parts of this investigation are presented in the Chapters 5 and 14 (cf. van der Vlugt et al., 1986; van der Vlugt & Klapwijk, 1987). Phosphate-remo val was started at the local sewage treatment plant of the village of Reeuwijk in 1986. So far, no clear results have appeared. As supple mentary measures fish population management and in-lake treatment with a precipitant will be carried out in parts of the lakes. For the Nieuwkoop lakes area a complex project has been developed including among others the diminution of water transit, separation of the agri cultural part from the lakes, and dephosphating of the intake water. For the Geerplas project the following measures are planned: isolation of the Geerplas of the remaining Langeraar lakes, dredging of the up per phosphate-rich sediment layer, dephosphating of the intake water and polishing of this water by a macrophyte-swamp. -12-

This thesis deals with the eutrophication research carried out from 1977-1987 at the Rijnland Waterboard laboratory. It consists of three parts: Part A deals with phytoplankton, especially its use in assessing water quality in an ecological way (Chapter 3), while in Chapter 4 a comparison is made between historical and recent phytoplankton and chemical data with the intention to find ecological objectives for com batting eutrophication. In Chapter 5 a description of the phosphate- phytoplankton relationships in the Reeuwijk lakes is presented. Part B is focussed on sediments, especially the availability of se diment phosphates for algal growth. This availability is determined in four lakes with a bioassay technique and compared with two chemical extraction techniques (Chapter 7), while in Chapter 8 a new method is proposed, which is applied to the sediments of eight lakes in the Rijnland area (Chapter 9). Part C treats bioassays, which are used to assess the algal growth potential and to determine the limiting nutrient(s) for algal growth in lakes and canals. This is done with the aid of the indige nous phytoplankton population (Chapter 11) and with testalgae like Scenedesmus quadricauda (Chapter 12). In Chapter 13 bioassays with two different testalgae and with different procedures are compared, while in Chapter 14 a comparison is made between two bioassay tech niques and several other methods to assess growth limiting factors in the Reeuwijk lakes.

In this thesis the policy development of the Rijnland Waterboard with respect to the eutrophication problem in the last ten years is also recognizable. In the first instance the attention was directed to the assessment of the water quality and the sediments (Chapters 3, 7, 8, 9), followed by large-scale phosphate-removal at sewage treat ment plants (Chapter 11). Since these source-directed measures did not lead to results, the policy has been changed in a water quality approach of specific lakes (Chapters 5, 12, 14). Based on the re search described in this thesis lake restoration projects could be de veloped for the Reeuwijk, the Nieuwkoop and the Langeraar lakes, which are being carried out now. Research is often a cooperative effort as is apparent from the fact that scientific papers are seldom written by one author. In this thesis Chapters 5, 8, 11 and 13 are written in the first instance by collegues in cooperation and responsibility with this author. Chapters 7, 9, 12 and 14 are mainly written by me with support by several co authors. The remaining Chapters 1, 2, 3, 4, 6, 10, 15 and 16 are entirely my responsibility.

REFERENCES Andersson, G., G. Cronberg & C. Gelin, 1973. Planktonic changes following the restoration of lake Trummen, Sweden. Ambio 2: 44-47.

Berg, L.K., K. Andersen, T. Christensen et al., 1958. Fures0 undersögelser 1950-1954. Folia Limnologica Scandinavica 10: 1-189. -13-

Does, J. van der & S.P. Klapwijk, 1985. Phosphorus removal and effects on waterquality in Rijnland. H20 18: 381-387 (in Dutch with an English summary).

Does, J. van der & S.P. Klapwijk, 1987. Effects of phosphorus removal on the maximal algal growth in bioassay experiments with water from four Dutch lakes. Int. Revue ges. Hydrobiol. 72: 27-39.

Golterman, H.L., 1965. Hydrobiologische aspecten van de Vechtplassen. Akademiedagen 17: 23-36.

Golterman, H.L., 1970a. Eventual consequences of the phosphate-eutrophication of fresh water. H20 3: 209-215 (in Dutch with an English summary). Golterman, H.L., 1970b. De invloed van het menselijk handelen op de biocoenosen in het water. In: Biosfeer en Mens, p. 80-103. Pudoc, Wageningen.

Golterman, H.L., 1971. De vervanging van polyfosfaten in wasmiddelen door NTA. H20 4: 557-559. Golterman, H.L., 1972. De zegepralende Vecht. H20 5: 33-34. Golterman, H.L., 1973a. The influence of phosphate on aquatic life. H20 6: 430-438 (in Dutch with an English summary).

Golterman, H. L. , 1973b. Natural phosphate sources in relation to phosphate budgets: A contribution to the understanding of eutrophication. Water Res. 7: 3-17.

Golterman, H.L. (ed.), 1976. Fosfaten in het Nederlandse oppervlaktewater. Rapport van de 'Stuurgroep Fosfaten' van de K.N.CV. Sigma Chemie.

Golterman, H.L. , 1979. Removal of phosphate from sewage waters; the only possible mea sure to decrease algal blooms, even in the lakes in Rijnland. H20 12: 40-44 (in Dutch with an English summary). Hoogheemraadschap van Rijnland, 1984. Rapport betreffende het onderzoek naar de effecten van fosfaat verwijdering op de a.w.z.i.'s Gouda, Bodegraven en Nieuwveen. Rapport technische dienst van Rijnland, Leiden.

Hosper, S.H., 1978. Nitrogen, phosphorus and eutrophication. H20 11: 385-387 (in Dutch with an English summary). -14-

Johnsen, P., H. Mathiesen & U. Rtfen, 1962. The Sor0 lakes, lake Lyngby S0 and lake Bagsvaerd S0, limnoli- gical studies in five culturally influenced lakes in Sjaelland (Zea land). Dansk Ingeni0rforening, Spildevandskorn. 14: 1-135.

Klapwijk, S.P., 1977. Experimentele fosfaatverwijdering op praktijkschaal in Rijnland. Waterschapsbelangen 62: 284-289.

Klapwijk, S.P. , 1981. Limnological research on the effects of phosphate removal in Rijnland. H20 14: 472-483 (in Dutch with an English summary). Lauterborn, R., 1918. Die geographische und biologische Gliederung des Rheinstroms. Teil III. Sitzungsber. Heidelberg. Akad. Wissensch. , Math. -nat. Klasse Abt. B, 1918: 1-87.

Ministries of Public Health & Environmental Hygiene and Public Trans port & Public Works, 1979. Fosfatennota: Maatregelen voor het terugdringen van de fosfaat belasting van het Nederlandse oppervlaktewater. Staatsuitgeverij, 's-Gravenhage.

Ohle, W., 1955. Die Ursachen der rasanten Seeneutrophierung. Verh. internat. Ver. Limnol. 12: 373-382.

Parma, S. , 1980. The history of the eutrophication concept and the eutrophication in the Netherlands. Hydrobiol. Bull. 14: 5-11.

Peelen, R. , 1975. Changes in the composition of the plankton of the rivers Rhine and Meuse in the Netherlands during the last fifty-five years. Verh. internat. Verein. Limnol. 19: 1997-2009.

Schmidt-van Dorp, A.D., 1975. Phosphate and eutrophication in the Waterboard of Rhineland. H20 8: 254-258 (in Dutch with an English summary).

Schmidt-van Dorp, A.D., 1978. Eutrophication of shallow lakes in Rijnland. Report Technical Service, Hoogheemraadschap van Rijnland, Leiden (in Dutch with an English summary).

Straskraba, M. & V. Straskrabova, 1969. Eastern European Lakes. In: Eutrophication: causes, consequen ces, correctives, p. 65-97. Proc. Symp. Nat. Acad. Sci. Was hington, D.C., 1969.

Thomas, E.A., 1953. Zur Bekampfung der See-Eutrophierung: Empirische und experimentelle Untersuchungen zur Kenntnis der Minimum stoffen in 46 Seen der Schweiz und angrenzender Gebiete. Schweiz. Ver. Gas- Wasserfachm. Monatsbull. 33: 25-32; 71-79. -15-

Vollenweider, R.A., 1968. Scientific fundamentals of the eutrophication of lakes and flowing waters, with particular reference to nitrogen and phosphorus as factors in eutrophication. OECD-report, Paris.

Vlugt, J.C. van der, S.P. Klapwijk & J.A.A.M. van Eijk, 1986. Waterkwaliteitsonderzoek Reeuwijkse plassen WOR 1983-1985. Report nr. 840156001 National Institute for Public Health and Environmental Hygiene, Bilthoven. Vlugt, J.C. van der & S.P. Klapwijk, 1987. Water quality research in the Reeuwijk lakes 1983-1985. H20 20: 86-91 (in Dutch with an English summary).

Zon, H. van, 1986. A very dirty affair - Studies in non-in dus trial pollution in the Netherlands, 1850-1920. Ph. D. Thesis, Univ. Groningen (in Dutch with an English summary). - 16-

Foto pag. 17: Massaal voorkomen van draadvormige blauwalgen in het fytoplankton van de Nieuwkoopse plassen (Noordeinder- plas d.d. 30.03.88). PART A: PHYTOPLANKTON

/ Ï **^

\-/ \jC>~ yf\A .//—{""" * AV/: ;' ',f\ . A

-19-

CHAPTER 2:

INTRODUCTION PHYTOPLANKTON

"Intusschen is het echter gebleken, dat de samenstelling van het wa ter zoodanig inwerkt op de planten en dieren, die er in leven, dat in de ontwikkeling, die zij vertoonen, niet alleen de gevolgen van cata- strophale veranderingen tot uitdrukking komen, maar dat ook minder acuut verloopende veranderingen in de samenstelling van het water hierop onmiskenbare gevolgen teweeg brengen."

G. Romijn, 1924. Hydrobiologische toestand van Rijnland. Ver slagen en mededelingen betreffende de Volksgezondheid, no. 2, p. 110. -20-

INTRODUCTION PHYTOPLANKTON Eutrophication in surface waters is most often expressed in an increase in phytoplankton biomass, resulting sometimes in severe algal blooms (Golterman, 1975, 1976; Parma, 1980; Barica & Mur, 1980). Not only the algal biomass is increasing, the algal species composition changes too as a result of eutrophication. Phytoplankton assemblages rich in species from different groups (e.g. diatoms, green algae and flagellates) disappear and blue-green algae begin to dominate the plankton association. • Not only eutrophication, but other factors, e.g. the intensity of supply of organic matter and the speciation of chemical elements, in fluences phytoplankton composition and abundance also. The species composition reflects the different conditions in the water and can therefore be used to assess the water quality of surface waters. Kolk- witz and Marsson (1902) developed a saprobity system that is based on the presence of species and that is appropriate for classifying the impact of organic pollution in running waters. This system is later extended by Liebmann (1960-1962) and Sladecek (1963, 1973). About twenty years ago Caspers & Karbe (1966, 1967) presented a water quality system based on physiological and ecological views, that tried to combine the leading concepts on saprobity and trophism. They showed that both are running parallelly (Caspers & Karbe, 1966, 1967; Caspers, 1977; Sladecek, 1973). This was and still is im portant to the practical water management in the western part of the Netherlands, where both phenomena play a leading role in water qua lity. Caspers and Karbe's system seems to be valid at least for stag nant bodies of water (Sladecek, 1973). Applying the system in the Netherlands seemed to be very promising, since most of the Dutch surface waters are stagnant and the relation between discharges of organic material, causing saprobity, and inorganic material, causing eutrophication, is often very intricate. Caspers and Karbe (1967) stated that classifying waters in one of their six classes can occur in three separate ways: (1) according to the bioactivity, which means the ratio between the intensity of pri mary production and respiration; (2) according to oxygen regime cri teria, especially the extent of the daily fluctuations and (3) according to the structure of biological communities in the water, for instance the diversity and biomass of the phytoplankton community and the re lation between producers, consumers and decomposers. Caspers and Karbe never worked out their scheme into a system for water quality assessment of specific waters. Therefore, we tried to elaborate their scheme in order to use it for the quality assessment of large stagnant surface waters in the western part of the Netherlands (Chapter 3). Since phytoplankton compositions reflect very well the physical and chemical conditions in a water, old plankton data can be compared to recent data to see if anything has changed as a consequence of in creasing eutrophication. Therefore, in Chapter 4 a comparison is pre sented between historical and recent data on water chemistry and phytoplankton in two canals and two lakes in the Rijnland Waterboard area in order to determine the differences and to find realistic objec tives for eutrophication parameters such as chlorophyll-a and transpa rency . The employed historic data set (from 1941 and 1942) is parti cularly interesting since it contains besides qualitative data on phyto plankton composition quantitative data too in addition to several time series of hydrochemical analyses. -21-

Finally, in Chapter 5 dose-effect relationships between phospho rus concentration and phytoplankton biomass are described for the Reeuwijk lakes. The aim of this study was to collect basic information on the water quality in the lakes, which can be compared with possi ble changes occurring when the external phosphorus load is reduced, e.g. by dephosphating at the local sewage treatment plant of Reeuwijk.

REFERENCES Barica, J. & L.R. Mur (eds.), 1980. Hypertrophic ecosystems. Developments in Hydrobiology 2: 1-348. Junk, The Hague - Boston - London.

Caspers, H., 1977. Qualitat des Wassers - Qualitat der Gewasser. Die Problematik der Saprobiensysteme. In: Sladecek, V., 1977 (ed.), Symposium on Saprobiology. Ergebn. Limnol. 9: 3-14.

Caspers, H. & L. Karbe, 1966. Trophie und Saprobitat als stoffwechsel-dynamischer Komplex. Gesichtspunkte fiir die Definition der Saprobitatsstufen. Arch. Hydrobiol. 61: 453-470.

Caspers, H. & L. Karbe, 1967. Vorschlage für eine saprobiologische Typisierung der Gewasser. Int. Revue ges. Hydrobiol. 52: 145-162.

Golterman, H.L., 1975. Physiological Limnology. Elsevier Scientific Publishing Company, Amsterdam.

Golterman, H.L.(ed.), 1976. Fosfaten in het Nederlandse oppervlaktewater. Rapport van de stuurgroep fosfaten van de K.N.CV. Sigma Chemie.

Kolkwitz, R. & M. Marsson, 1902. Grundsatze für die biologische Beurteilung des Wassers nach seiner Flora und Fauna. Mitt. Prüfungsanst. Wasserversorgung. Abwassereinigung 1: 33-72.

Liebmann, H. (ed.), 1960-1962. Handbuch der Frischwasser- und Abwasserbiologie. Band I & II. R. Oldenbourg, München.

Parma, S., 1980. The history of the eutrophication concept and the eutrophication in the Netherlands. Hydrobiol. Buil. 14: 5-11.

Sladecek, V., 1963. A guide to limnosaprobical organisms. Science Papers Inst. Chem. Techn. Prague, Techn. Water 7: 543-612.

Sladecek, V., 1973. System of water quality from the biological point of view. Er gebn. Limnol. 7: 1-218.

-23-

CHAPTER 3: BIOLOGICAL ASSESSMENT OF THE WATER QUALITY IN SOUTH HOLLAND (THE NETHERLANDS)

Accepted for publication in: Int. Revue ges. Hydrobiol.

"The proposal of Caspers & Karbe is no doubt modern, elegant and attractive, but the engineer asks for numbers and how to obtain them. For this reason Casper's scheme is more a programme than a functio ning system."

Sladecek, V., 1973. System of water quality from the biological point of view. Ergeb. Limnol. 7, p. 44. -24-

BIOLOGICAL ASSESSMENT OF THE WATER QUALITY IN SOUTH HOLLAND (THE NETHERLANDS).

Sjoerd P. Klapwijk

Key words: biological assessment, water quality, phytoplankton, biological indicators, saprobity, trophism.

ABSTRACT A hydrobiological study was carried out to elaborate the water quality classification system of Caspers and Karbe (1966, 1967) for the large stagnant bodies of water in the western part of the Nether lands. Significant correlations have been established between different physical and chemical parameters and phytoplankton community struc tures such as diversity and saprobity. A proposal is developed to quantify the water quality classes in Caspers and Karbe's scheme on parameters measuring the bioactivity, oxygen regime and phytoplank ton community structure. The scheme is now incorporated in the prac tical water quality assessment in North Holland and South Holland both for routine monitoring as well as for special water quality stu dies.

1. INTRODUCTION

Both from a scientific point of view as well as in the applied wa ter management practice classification of waters is useful and necessa ry and is widely applied since the beginning, of this century. Practi cal water management requires preferably a classification system which has a solid scientific base and is simple and cheap to execute. The classic saprobity system developed by Kolkwitz and Marsson (1902) and later extended by Liebmann (1960-1962) and Sladecek (1973) is providing this base on the presence of species and is especially ap propriate for classifying the impact of organic pollution in running waters. Twenty years ago Caspers and Karbe presented a water quality system based on physiological and ecological principles (Caspers, 1966; Caspers & Karbe, 1966; 1967). This system tried to combine the leading concepts on saprobity and trophism. It was shown that both are running parallelly (Caspers & Karbe, 1966; Caspers, 1977a; Sla decek, 1977). This was and still is important to the practical water management, where both phenomena play a leading role in water qua lity. Essential in the system is also that it combined functional as well as structural aspects of the aquatic ecosystem. Caspers and Karbe's system with six water quality classes seems to be valid at least for stagnant bodies of water (Sladecek, 1973). Applying the system in the Netherlands seemed very promising, since most of the Dutch surface waters are stagnant and the relation be tween discharges of organic material, causing saprobity, and inorga nic material, causing eutrophication, is often very intricate. Caspers and Karbe (1967) stated that classifying waters in one of their six classes can occur in three separate ways (cf. Hovenkamp et al., 1982): Table 1. Scheme of water quality classes according to Caspers & Karbe (Modified after Caspers & Karbe, 1967 and Sladecek, 1973).

Intensity of Bioactivity Oxygen regime Structure of community of organisms supply of organic matter Intensity Intensity of Total Type of organisms of primary respiration biomass production

negligible oxygen content inde poor in individuals; balanced relationship between producers, consu pendent of bioacti moderate diversity mers and decomposers vity, determined by hydrographic factors

oxygen content deter large rich in individuals; balanced relationship between producers, consu mined more or less high diversity mers and decomposers equally by assimila tion, respiration and hydrographic factors

high high daily oxygen content rich in individuals; substantially balanced relationship between pro shows a distinct high diversity ducers , consumers and decomposers; a relative dependence upon the increase of decomposers and consumers activity of organisms; I frequently supersaturation during daytime and oxygen deficit at I night

very high high very high oxygen content large rich in individuals; many consumers and decomposers, few producers; usually below the low diversity mass development of bacteria and ciliates saturation value; anaerobic conditions prevail at night; variations in the supply of poliutant material are clearly recognizable from the oxygen content

extremely high low extremely high anaerobic conditions large extremely rich in producers drastically decline; only a few macro- predominate even individuals; fauna species are present in great abundance; during the daytime low diversity mass development of bacteria and ciliates

VI extremely high approx. O extremely high permanently anaerobic total biomass is formed practically solely by anaerobic bacteria and fungi; conditions producers and macro-organisms are absent -26-

1. According to the bioactivity, which means the ratio between the primary production and respiration (cf. Odum, 1956 and Slade- cek, 1973). 2. According to oxygen regime criteria, especially the extent of the daily oxygen fluctuations. 3. According to the structure of biological communities in the water, for instance the diversity and biomass of the community and the relationship between producers, consumers and decomposers.

Their ideas about the classification into six classes have been summarized in Table 1. Caspers and Karbe never worked out this scheme into a system for water quality assessment of specific waters. As Sladecek (1973) already remarked the scheme is "more a program me than a functioning system". Therefore the aim of this study was to elaborate the scheme in order to use it for the water quality as sessment of large stagnant surface waters in the western part of the Netherlands. The study aimed to find suitable chemical, physical or biological parameters, which can be used to measure the aspects bio activity (production and respiration), oxygen regime and community structure in Caspers and Karbe's scheme in an ecologically based and consistent way. Since the scheme has to be applied in practical water quality management, special emphasis has been paid to the relation between the phytoplankton in the waters and the physical and chemi cal water quality parameters, which are normally determined in rou tine monitoring programs.

2. MATERIALS AND METHODS

During 1977 and 1978 samples for physical and chemical. water analyses were taken before noon every month at more than sixty sam pling sites at a depth of 0.5 m in six different morphological and hydrological water types in the province of South-Holland (Table 2).

Table 2. Summary of the considered water types, the number of sam pled sites and the frequency of sampling.

Water types Number of sites for

physical and chemical analysis plankton analysis

year: 1977 1978 1977 1978

Canals 26 21 6 3

Lakes 24 26 6 5

Peat lakes 13 12 .6 3 Deep sandpits 2 4 - 2 Deep pools 1 2 - 2 Dune lakes 0 2 - 0

Total 66 67 18 15

Frequency of sampling per year: 12x 12x -27-

In 1977 at eighteen and in 1978 at fifteen of the sampling sites also plankton samples were taken respectively five and twelve times per year. An extensive description of the sampling method and sampling sites is given by Klapwijk (1982). In this study 1085 water analyses and 260 plankton analyses were used. The sampling sites were chosen in such a way that both polluted and unpolluted stations were incorporated in the study. The following water analyses were carried out according to standard methods of the Netherlands Normalisation Institute: pH, temperature, oxygen concen tration (02), oxygen saturation (02%), chemical oxygen demand (COD), biological oxygen demand (BOD), Kjeldahl-N, NH4-N, N02+ N03-N, ortho-P, total-P, Cl , Secchi disk transparency, specific con ductivity and chlorophyll-a (CHL-a). One liter plankton samples were also taken at 0.5 m depth, transported to the laboratory in glass jars, preserved with 10 ml 37% formalin and poured into one liter high measure glasses to let the plankton settle in a quiet and dark place. After two to six weeks the supernatant water was siphoned off and the settled plankton was brought into 80 ml test-tubes to settle again. After a few weeks the supernatant water again was poured off so that the plankton was con centrated in a volume less of than 30 ml. In those samples that con tained many floating blue-green algae (e.g. Microcystis aeruginosa Kütz) the supernatant was not discarded but filtrated through a 20 p plankton gauze and the filtrated plankton was added. Random plankton countings up to at least hundred individuals or colonies per sample were carried out on an inverted plankton micro scope (Olympus IMT) with magnification of 40x, lOOx, 200x and 400x using a 2 cm3 cuvette with a height of 2 mm. Every single cell, colo ny of cells or filament is counted as one individual with the exception that ten loose cells of Microcystis aeruginosa were counted as one. Colonies of this species nearly always fell apart by the formalin pre servation. Remnants of algae or empty cells (e.g. diatom thecae) were excluded from the countings. The phytoplankton species were identified using mainly the keys in Huber-Pestalozzi (1950-1962), Pascher (1913-1930), Prescott (1964), Schroevers (unpublished), Streble and Krauter (1974), Uherkovitz (1966) and van der Werff and Huls (1957-1974). Apart from the phy toplankton countings, the number of ciliate species was determined for the quotient of Dresscher and van der Mark (1976) by concentrating an unpreserved one liter water sample over a 20 p plankton gauze to a small volume (5-10 ml) and counting the number of ciliate species in a few drops of the living concentrate under a normal microscope (Olympus BHA) with 40x, lOOx, 400x and lOOOx magnification. This microscope was also used to check uncertain identifications made with the inverted microscope and to identify the diatom species in the sam ples by means of separate diatom preparations. With the aid of a DEC PDP 14/34 computer the following ecolo gical characteristics were calculated and derived from the plankton countings.

2.1. Similarity:

Species composition on the sampling sites was compared in two different ways: -28- a. the qualitative resemblance, based on presence of species on the sampling sites, is computed by the similarity index of Jaccard (Mueller-Dombois & Ellenberg, 1974) according to the following formula: n., J index of Jaccard (%) = + _ x 100% j k jk where n. = the number of species occurring- in sample j, n, = the number of species occurring in sample k, n., = the number of species occurring in both sam ples j and k. b. the quantitative resemblance, based on abundance of species at the sampling sites, was computed by the similarity index of Ellen- berg (Mueller-Dombois & Ellenberg, 1974) according to the follo wing formula: S., : 2 x 100 index of Ellenberg (%) = g + £ + s —^ % j k jk where S. = the sum of the percentage individuals of the J species occurring only in sample j, S, = the sum of the percentage individuals of the species occourring only in sample . k, S., = the sum of the percentage individuals of the species occurring in both samples j and k.

2.2. Biomass:

The total number of algae per ml (I/ML) was calculated from the number of counted algae and the volume of the original and concen trated plankton sample and the number and volume of the counted viewing fields under the microscope by the following formula:

Nx . Vt . 103 I/ML = N2 . V2 . V3

where Nx = the number of counted algae, Vj = the volume after concentration (in ml),

V2 = the volume of the sample before concentration (in ml),

N2 = the number of counted viewing fields,

V3 = the volume of the counted viewing fields (in mm3).

2.3. Diversity:

The species diversity at the sampling sites was measured in three ways (Klapwijk, 1980; Klapwijk et al., 1983): -29-

by counting the number of species (S) in the samples (mostly apprbx. 100 individuals or colonies counted),

by computing species diversity indices with Margalef's formula (Odum, 1971): S - 1 D In N where D = index of Margalef, S = the number of species, N = the number of organisms or cells counted to find S species. c. by computing indices of species diversity and equitability with the Shannon-Wiener function (Krebs, 1972):

H = - I (p )(log2p ) i=l

H = 2iog s max ° E = H/H max where • H = index of species diversity, S = number of species, p. = proportion of total sample belonging to ith species, E = equitability or evenness (range: 0-1), H = maximum species diversity max ^ J 2.4. Saprobity: The saprobity of the plankton communities was measured in three different ways (Klapwijk, 1978, 1980; Klapwijk et al., 1983). a. by calculating the saprobic quotient of Dresscher and van der Mark (1976, 1980) using the following formula: C + 3D - B - 3A SQ A + B + C + D

where SQ = saprobic quotient (range - 3 to + 3), A = the number of Ciliata, indicating very severe pollution (polysaprobity), B = the number of Euglenophyceae, indicating considerable pollution (a-mesosaprobity), C = the number of Chlorococcales and Diatomeae, indicating moderate pollution (p-mesosaprobity), D = the number of species of the Peridineae, Chrysophyceae and Conjugatae, indicating very slight pollution (oligosaprobity). -30- b. by calculating the saprobic index of Pan tie and Buck (1955) using Sladecek's (1973) and Mauch's (1976) extended lists of saprobity indicating organisms according to the following formula: I (n . s.) I n. l where SI = saprobic index of Pan tie and Buck, n. = abundancy e.g. number of individuals of ith species, s. = saprobic value of individual species i. c. by calculating the saprobic valencies according to Zelinka and Marvan (1961) using also Sladecek's (1973) and Mauch's (1976) extended lists of saprobity indicating organisms according to the following formula: 1 n. g x X = i=1 n 2 n. g i=l l

where X saprobity level, n. abundancy e.g. number of individuals of ith species, g indicative weight of species (range 1-5), x the share from the whole saprobic valency given to the Xth degree or the number of points in the respective degree.

Besides these ecological characteristics, the following statistical tech niques were used to evaluate and test the results of the water che mistry and plankton analyses: Average linkage cluster analyses (Everitt, 1974) were used to cluster the similarity indices of Jaccard and Ellenberg. Linear correlation and regression analyses were used according to Sokal and Rohlf (1969) to compute the relationships between different chemical and ecological parameters.

3. RESULTS The physical and chemical measurements as well as the species composition of the planktonic communities cannot be presented here in detail. The complete data are given by Klapwijk (1982). Some data are summarized in Table 3, showing that both the means as well as the minima and maxima of the various parameters differ in the diffe rent water types. Especially the canals showed rather low oxygen concentrations and relatively high nutrient concentrations in compari son with the other water types. The peat lakes showed the highest, and the deep sand pits the lowest chlorophyll-a concentrations and algal biomass (I/ML). Table 3. Summarized data on the water analyses in the sampled water type. Explanation of abbreviations: COD : chemical oxygen demand; BOD : biological oxygen demand; I/ML : number of individuals per ml; S = diversity index of Margalef; H = diversity index of Shannon & Wiener; SQ : saprobic quotient of Dresscher & van der Mark; SI - saprobic index of Pan tie & Buck; SI* : saprobic index of Pantle & Buck with changed saprobic valencies for dominating bluegreens; : number of sampling stations; : not measured.

Canals Lakes Peat lakes Deep sand pits Deep pools Dune lakes

min. mean max. min. mean max. min. mean max. min. mean max. min. mean max. min. mean max. (n = 25) (n = 29) (n = 13) (n = 5) (n = 2) (n = 2)

PH 6.90 7.69 9.05 7.15 8.11 9.20 7.40 8.27 9.35 7.10 8.03 8.90 7.50 8.04 8.60 7.30 7.93 9.10

Temperature CO 0.0 10.8 22.0 0.5 11.0 24.5 0.5 11.0 22.0 0.0 10.6 21.5 0.0 10.6 19.0 0.0 10.7 21.0

Oxygen (mg l"1 ) 0.0 7.1 26.6 1.4 10.0 22.1 1.3 10.4 20.7 1.4 10.1 15.7 3.2 9.5 15.5 2.6 8.4 13.6

Oxygen saturation (%) 0 63 221 13 88 218 12 92 180 13 90 140 28 84 120 23 73 101

COD (mg 1" ) 13 53 151 7 45 114 19 64 119 7 32 71 11 37 79 20 58 138

BOD (mg f ) 1 7 75 1 5 33 1 9 28 1 3 15 1 3 13 1 4 40

Kjeldahl-N (mg 1~ ) 0.6 4.2 25.0 0.5 2.5 10.0 1.2 2.9 7.0 0.5 2.2 8.3 0.8 1.6 3.3 0.6 2.0 4.8 I CO NH«-N (mg 1" ) 0.1 2.1 21.0 0.1 0.6 7.4 0.1 0.3 4.9 0.1 0.8 6.2 0.1 0.2 1.1 0.1 0.2 0.7

N02tN03-N (mg f ) 0.02 3.3 22.4 0.01 3.14 17.8 0.01 0.62 14.4 0.01 2.64 14.4 0.01 0.43 1.47 0.01 0.09 0.43

Ortho-P (mg l" ) 0.03 0.77 7.4 0.01 0.50 5.9 0.01 0.21 1.6 0.01 O.U 0.5 0.01 0.48 1.7 0.01 0.17 0.71

Total P (mg f ) 0.08 1.02 8.2 0.04 0.68 6.5 0.04 0.39 1.7 0.01 0.20 1.6 0.06 0.58 1.9 0.02 0.31 1.50

Chloride (mg 1-1 ) 63 286 3500 73 198 450 129 195 675 59 129 224 57 84 125 42 98 145

Transparency (m) 0.1 0.9 7.7 0.1 0.8 3.2 0.1 0.4 1.2 0.1 1.6 6.0 0.6 1.2 1.8 0.2 0.31 0.4

Conductivity 25 C(mS cm "') 0.65 1.63 12.1 0.66 1.36 2.68 0.73 1.15 3.14 0.58 0.92 1.81 0.50 0.77 1.16 0.46 0.76 1.84

Chlorophyll-a (mg m" 3) 1 44 516 1 54 506 14 93 375 1 8 55 1 27 258 3 22 103

(n = 9) (n = 11) (n = 9) (n = 2) (n = 2) I/ML 156 9400 83800 91 29200 142000 870 54800 292000 115 4400 21000 594 12500 67800 - - - S 7 22 39 4 20 40 3 16 29 7 17 36 8 19 35 - - - D 1.3 4.5 7.9 0.6 3.9 6.4 0.4 3.0 5.9 1.3 3.3 7.1 1.1 3.8 7.2 - - - H 0.7 3.2 4.7 0.6 3.0 4.3 0.3 2.3 3.8 0.8 2.6 4.4 0.3 2.8 4.6 - - - SQ -1.9 -0.1 1.2 -1.5 0.1 1.4 -3.0 -0.6 0.7 -1.2 0.5 2.3 -0.9 0.3 1.1 - - - SI 1.6 2.3 3.0 1.5 2.1 2.7 1.4 2.0 3.5 1.1 1.8 2.5 1.4 2.1 2.6 - - - SI* 1.6 2.4 3.0 1.7 2.4 3.5 1.4 2.9 3.5 1.1 1.8 2.5 1.4 2.2 2.9 _ _ _ -32-

The highest diversity values (S, D and H) were observed in the canals, the lowest in the peat lakes. Dresscher and van der Mark's (1976) saprobic quotient (SQ) was highest in the deep sand pits and lowest in the canals and the peat lakes. In contrast, because of the opposite scale of Pantle and Buck's (1955) saprobic index, the highest SI values were observed in the canals, lakes and peat lakes and the lowest in the deep sand pits, probably due to their isolated situation. The plankton species with the highest presence and abundance at the sampled sites are listed in Table 4. The bluegreen alga Lyng- bya limnetica, the diatoms Stephanodiscus hantzschii, Stephanodiscus as-traea var, minutula and Nitzschia acicularis and different chlorococcal green algae,of the genera Scenedesmus, Monoraphidium, Kirchne- riella, Dictyosphaerium and Tetrastrum are present in more than 87% of all the sampling sites. The highest mean abundance was reached by the blue-greens Lyngbya limnetica (more than 7700 filaments per ml), Oscillatorja redekei, Oscillatoria agardhii, the diatoms Stephano discus hantzschii, Stephanodiscus astraea var, minutula, Diatoma elongatum, Nitzschia acicularis and chlorococcal green algae of different genera (all more than 200 individuals per ml). The plankton in the sampled waters can therefore be characterized generally as a green algae (Chlorococcales) and diatom community (especially Centrales of the genus Stephanodiscus). In some lakes the plankton community is dominated by a permanent bloom of filamentous blue-greens like Lyng bya limnetica, Oscillatoria agardhii, Oscillatoria redekei or a periodic bloom of the blue-green Microcystis aeruginosa.

Table 4. Algal species with the highest presence (in percentage pre sent at all sample sites) and the highest abundance (in mean numbers of individuals per ml) at the sampled sites.

Species Presence Species Mean abundance (%) (individuals . ml 1)

Stephanodiscus hantzschii GRUN. 100 Lyngbya limnetica LEMM. 7700 Dictyosphaerium pulchellum WOOD 100 Stephanodiscus hantzschii GRUN. 2900 Monoraphidium contortum (THUR.) KOM.-LEGN. 100 Oscillatoria redekei van GOOR 2400 Scenedesmus costato-granuiatus SKUJA 96 Planctonema lauterbornii SCHMIDLE 2000 Tetrastrum staurogeniaeforme (SCHRÖD.) LEMM. 95 Oscillatoria agardhii GOM. 1600 Lyngbya limnetica LEMM. 95 Monoraphidium contortum (THUR.) KOM.-LEGN. 1000 Stephanodiscus astraea v. minutula (KG.) GRUN. 95 Dictyosphaerium pulchellum WOOD 600 Monoraphidium komarkovae NYG. 95 Scenedesmus quadricauda (TÜRP.) BREB. 400 Scenedesmus spec. 17 95 Monoraphidium komarkovae NYG. 400 Nit2schia acicularis W. SMITH 87 Planctomyces bekefü GIMESI 400 Kirchneriella.obesa (W. WEST) SCHMIDLE 87 Chlorophyt spec. 3 400 Scenedesmus quadricauda (TÜRP.) BREB. 87 Cryptomonas caudata SCHILLER 300 Scenedesmus tenuispina CHOD. 87 Scenedesmus falcatus CHOD. 300 Monoraphidium minutum (NAG.) KOM.-LEGN. 87 Scenedesmus armatus CHOD. 300 Scenedesmus falcatus CHOD. 87 Skeletonema potamos (WEBER) HASLE 300 Synura uvella E./petersenii KORSCHIK 300 Diatoma elongatum (LYNGB.) AG. 200 Nitzschia acicularis W. SMITH 200 KirchnerieUa obesa (W. WEST) SCHMIDLE 200 Tetrastrum staurogeniaeforme (SCHRÖD.) LEMM. 200 Stephanodiscus astraea v. minutula (KG.) GRUN. 200 Scenedesmus costato-granulatus SKUJA 200 Scenedesmus indeterminanda 200

The mean species composition of the sampled sites was compared by computing similarity coefficients of Jaccard and Ellenberg (Mueller- Dombois & Ellenberg, 1974), while a grouping of the sampled stations was made by applying average linkage cluster analysis (Everitt, 1974) on the calculated similarity coefficients (Fig. 1 and 2). -33-

.262

X .297 □LU z .332 1/1 CD CC < .36' l_> <

13 ,137 ra rr o .u* i J i_j

< .507 >- i— ca < ,512 co n 1. Average linkage dendrogram of the similarity coefficients accor ding to Jaccard between the algal composition on the sampling sites. Explanation of the sampling site number: canals (1-7), lakes (8-13), peat lakes (14-19), deep pools (20-21), deep sandpits (22-23).

.373

CD' Z

CC LU CO .583

UJ ,653 O I— i o cr o < i >- -J62 (X 1

2. Average linkage dendrogram of the similarity coefficients accor ding to Ellenberg between the algal composition on the sampling sites. For explanation of sampling site numbers see Fig. 1. -34-

Figure 1 shows the cluster dendrogram of the Jaccard indices. Dis tinct clusters can be distinguished: From left to right a group of four canals can be seen (no. 1, 7, 3, 6), then a small group of three canals (no. 2, 4, 5), a group of three lakes (no. 8-10) followed by a deep lake (no. 22). Next to that a cluster of three shallow peat lakes (no. 12, 13 and 19) and one deep pool (no. 20) and two shallow peat lakes (no. 14 and 16) can be observed. Then a large lake (no 11), two shallow peat lakes (no. 15 and 17) and finally to the right two isolated deep waters (no. 21 and 23) and an isolated polder lake (no. 18). Similar results showed the clusteranalysis of Ellenberg's indices, which also takes the numbers of the different species into account (Fig. 2). The clusters seem to be mainly determined by the. dominant species on the sampling sites. From left to right in Figure 2 a group of canals mainly dominated by Chlorococcales and Centrales (no. 1, 2, 5) and lakes (no. 8, 10) can be seen, then a large group of mainly lakes dominated by filamentous bluegreen algae (no. 7, 16, 15, 12, 14, 19, 13, 20, 17), followed by two more polluted canals (no. 3 and 4). Next to that a small cluster of a canal (no. 6) and a polderlake (no. 18) and two lakes dominated in the late summer by Microcystis aeruginosa (no. 9, 11). Finally to the right a cluster of two deep waters (no. 21, 22) and totally separated an isolated and unpolluted deep sand pit (no. 23), which seems to display a quite different plankton composition. Figures 1 and 2 show also that the qualitative resemblance in species composition is not very high (approximately 50%), but that the quantitative resemblance is generally much, higher (approx. 80%) between the sampling sites. In both aspects, the species composition seems to be correlated to a large extent with the geomorphology of the sampling sites. However, it is not clear whether this is caused by the geomorphology of the sampling site or by its chemical characteris tics, since it has been shown that the latter is depending on the wa ter type (Table 3). Apart from that it is obvious that the degree of pollution and eutrophication is also determining the species composi tion.

3.1. Suitable parameters for the Caspers and Karbe scheme

The aim of this study was to find suitable parameters, which can be used to fill in the scheme of Caspers and Karbe in an ecologically based and consistent way. Since the system has to be applied in the practical water management, simple physical and chemical parameters, which are normally determined in a routine monitoring programme, are considered in this respect. By comparing these parameters with more sophisticated ecological characteristics derived from plankton countings, such as biomass, diversity and saprobity. indices, an attempt has been made to find cheap and simple determinable parameters, suitable for use in the Caspers and Karbe scheme for the aspects of bioactivity (production and respiration) and oxygen regime. Therefore, with the aid of product-moment correlation analyses between the measured physical and chemical parameters and the ecolo gical plankton characteristics, it became evident if and in what way the parameters were correlated with each other and with specific pa rameters indicating organic pollution, e.g. BOD, 02, COD, or eutro phication, e.g. N03-N, total P, chlorophyll-a (Table 5). In this table can be seen that the correlation between several parameters is some times highly significant. Table 5. Matrix of product-moment correlations between several physical and chemical parameters and plankton characteristics. Ecologically significant correlations, mentioned in the text, are underlined.

Explanation of abbreviations: CHL-a = chlorophyll-a; 02 = oxygen content; 02% = oxygen saturation; E = equitability index of Shannon-Wiener; for other abbreviations see Table 3.

parameters: Temp. pH Kj-N NH4-N N02+N03-N o-P t-P CHL-a BOD COD • 02 02% I/ML Trans- S D H E SQ SI SI* parency *** * *** *** '*** *** *** * ** ** Temperature -1.00 0.20 -0.08 -0.16 -0.21 0.02 0.04 0.12 0.12 0.03 -0.34 0.03 0.16 -0.05 0.19 0.19 0.10 0.04 0^02 -0.12 0.00 ** *** ************** *** *** *** *** ** ****** *** *** ** *** *** pH 1.00 -0.11 -0.36 -0.31 -0.15 -0.11 0.42 0.25 0.29 0.57 0.68 0.46 -0.09 -0.35 -0.37 -0.34 -0.25 -0.23 -0.21 0J!2 *** *** *** *** *** *** *** *** *** *** *** *** Kjeldahl-N - 1.00 0.85 0.02 0.68 0.73 0.32 0.68 0.55 -0.26 -0.29 -0.01 -0.31 0.10 0.08 0.04 0.02 -0.33 0.43 0J!S *** *** *** *** *** *** *** ** *** ** * * * *** NH,-N - 1.00 .0.12 0.67 0.66 -0.05 0.37 0.16 -0.38 -0.45 -0.18 -0.11 0.17 0.16 0.13 0.10 -0.17 0.38 0.05 *** *** *** ****** ****** **** N02*N03-N 1.00 0.02 0.01 -0.25 -0.19 -0.20 0.02 -0.06 -0.30 0.15 0.14 0.15 0.19 0.17 0.08 0.32 -0.15 *** * *** *** *** *** * *** *** *** * *** Ortho-P —• — 1.00 0.97 0.07 0.34 0.23 -0.33 -0.33 -0.13 -0.12 0.22 0.21 0.15 0.10 -0.07 0.29 0.03 *** *** *** *** *** *** *** *** * *** Total-P - - 100 0..1B 0.44 0.32 -0.30 -0.30 -0.08 -0.18 0.22 0.21 0.14 0.08 -0.11 0.33 0.07 *** *** *** *** *** *** *** *** *** *** *** ** *** Chlorophyll-a - 100 (L68 0.54 0.23 0.28 0_j>6 -0.37 -0.21 -0.25 -0.31 -0.30 -0.48 0.19 0.52 *** *** *** * * *** *** *** BOD - 100 0.66 -0.02 0.03 0.34 -0.43 -0.06 -0.08. -0.14 -0.15 -0.53 0.37 OJ53 * ** *** *** * * ** **. *** *** *** , COD - 100 0.07 0.10 0.34 -0.45 -0.13 -0.15 -0.19 -0.17 -0.54 0.27 0.48 to en *** *** *** *** *** *** ** | 0 -- 1.00 0.92 0.22 0.05 -0.42 -0.44 -0.37 . -0.28 -0.10 -0.18 0.06 2 *** *** *** *** *** *** 0 -saturation - — 1.00 0.31 -0.03 -0.40 -0.41 -0.37 -0.27 -0.10 -0.24 0.08 2 *** ** *** *** *** *** *** I/ML 1.00 -0.29 -0.18 -0.22 -0.33 -0.36 -0.24 0.04 0.48 * *** *** *** Transparency - - 100 -0.13 -0.12 -0.11 -0.08 0.51 -0.46 -0.56 *** *** *** *** *** S - 1.00 0.99 0.82 0.57 0.37 0.09 -0.22 *** *** *** *** D 1.00 0.86 0.63 0.37 0.08 -0.24 *** *** *** H - - - - 1.00 0.93 0.28 0.03 -0.25 *** E - - 1.00 0.12 -0.01 -0.19 SQ - -■ 1 00 -CL34 -0.65 *** SI > - 1.00 0.45 SI* — — - - — 1.00

No. of observation pairs 1085 1085 1085 1085 1085 1085 1085 1085 1085 1085 1085 1085 258 1085 258 258 258 258 189 260 258 * P <0.05; ** P <0.01; ***~ïr~Ö".Ö0T""~ -36-

In this respect it is interesting to stress the moderate correla tion between CHL-a and BOD (r = 0.68), indicating that the degra- dable organic matter in the sampled waters consists for the most part of algae instead of unpurified wastewater. Therefore, in most waters the signs of saprobity are less clear than the signs of eutrophication. Furthermore it is conspicuous that the negative correlation between CHL-a and NH4-N (r = -0.05) or N02+N03-N (r = -0.25) in compari son with the positive correlation between CHL-a and ortho-P indicates that the algae in most of the waters are more likely to be limited by nitrogen than by phosphorus (cf. van der Does & Klapwijk, 1987; de Vries & Klapwijk, 1987). Note further that the correlation between the saprobic quotient SQ and the Si-index of Pantle and Buck, using Sladecek's (1973) indicator values, is low (r = -0.34; n = 189) (N.B.: The sign is negative due to the opposite scales of the two indices). On the basis of these correlation analyses the following parame ters were chosen to use in the scheme of Caspers and Karbe for the different aspects: CHL-a (for production), BOD (for respiration), 02 or 02% (for the oxygen regime), I/ML (for biomass), D (for diver sity) and SI or SQ (for saprobity).

3.2. Proposal for filling in Caspers and Karbe's scheme

The above mentioned parameters (CHL-a, BOD, 02%, I/ML, D and SI or SQ) were critically reviewed and mutually compared in or der to achieve a consistent filling in of Caspers and Karbe's scheme.

3.2.1. Production According to Caspers and Karbe (1967) the intensity of pri mary production is increasing from class I to class IV, while it is very moderate in class V and absent in class VI (Table 1). Since class I, representing extremely clean waters like mountain springs, is probably lacking in South Holland, the CHL-a measurements, simula ting primary production, should fit in class II to IV with the highest concentrations (e.g. >100 mg m 3) in class IV and the lowest (e.g. <25 mg m 3) in class II. Due to the high correlation between CHL-a and BOD it was necessary to tune both parameters precisely to each other, because otherwise discrepancies could arise in the scheme. In our data the maximal CHL-a/BOD ratio was approximately 20 : 1, BOD expressed as mg 1 1 and chlorophyll-a as mg m 3 (Fig. 3). Due to the fact that the highest algal production is found in the summer months, the average summer concentration of CHL-a is chosen as most suitable for use in the scheme.

3.2.2. Respiration According to Caspers and Karbe (1967) decomposition is in creasing from class I to class VI (Table 1). Since class I is unlikely in South-Holland and class VI is meant for untreated wastewaters with 1 BOD values from 100-600 mg 02 l" , the measured BOD values should fit into the classes II to V. Sladecek (1973) already mentioned BOD standards for stagnant waters (Table 6). Since BOD values in South- Holland mostly appeared to fall into the classes III and IV, a subdivi sion of these classes is considered. Due to the occurring fluctuations in the BOD, also by the different seasons, class limits based on percentile values will be preferable to year averages or maxima, although the correlation between the 75 percentile and the year average of the BOD values was very high (r = 0.98; n = 13). -37-

BOD versus CHL0R0PHYLL-a

500 - 1 a 1 ■ E

er 1 a 400 - 1 a ii IB a 1 ■ _Ii 300 - la 2B IB _i >- 2a T 2a 2a SB 6B n 21 21 6B 1 B 1 B o 200 - 3a 4a 3B 4B 1 B a: 2a 3a 3B IB 1 B 1 a o la 3a 14a 10a 3B 1B _i 2B 3 B 12» 20B 9a 1B IB X 100 - 6B 7 8 28B 33a 10 a IB IB IB o 16* 43 ■ 53 B 16 a 10 a 2B 1 a BSB 75 ■ 39 B 13 a 1 a IB IB 1 B la 1 a 422» 68* 40a IS a 4a 2B 3B 3B la la 0 - i 1 1 1 1 1 1 1 1 10 20 30 40 50 BOD (mg/l)

16 a - NUMBER OF PAIRED OBSERVATIONS

Fig. 3. Relation between Biological Oxygen Demand (BOD) and Chlo- rophyll-a (CHL-a) values.

Table 6. Saprobity levels and corresponding BOD and oxygen values for standing water bodies according to Sladecek (1973).

level ]BO D . o2 8 11 oligosaprobity < 5 > 6 III 0-mesosaprobity < 10 > 4 IV a-mesosaprobity < 15 > 2 V polysaprobity < 100 > 0.1 VI isosaprobity < 600 0

After Sladecek (1973), Table 63. -38-

3.2.3. Oxygen regime

According to Caspers and Karbe (1967) the dynamics in the oxygen regime, especially the intensity of the diurnal fluctuation, is increasing from class I to class IV. In class V oxygen depletion es pecially at night but also during the day is found, while class VI is characterized by a permanent oxygen deficit (Table 1). Sladecek (1973) also gave standards for oxygen concentrations in the various saprobity levels (Table 6). Since the ideas of Caspers and Karbe are mainly focussed on the daily oxygen fluctuations, the concept of "favourable" and "very good" oxygen observation coupled with percentage occurrence is developed. Both a very low oxygen concentration as well as a supersaturation of oxygen are considered as undesirable. Therefore "favourable" and "very good" 02-observation are those observations lying respectively between 60 and 140% saturation and between 80 and 120% saturation. Apart from that, a good tuning of the criteria for oxygen and BOD is required. However, the oxygen concentration itself is absolutely not correlated with BOD (Table 5), probably because BOD reflects both production and decomposition and the relations of both processes with oxygen are opposite.

3.2.4. Biomass

Caspers and Karbe (1967) stated that the biomass is increasing from class I to IV, while in class V a mass development of bacteria and bacteria-eating ciliates occurs. Class VI can be characterized by a lack of C-autotrophic organisms and the total biomass would consist entirely of anaerobic bacteria and fungi. In the literature (Sladecek, 1973) some data can be found on the number of (psychrophylic and coliform) bacteria in the various saprobic classes, but not on algal densities. Biomass standards for phytoplankton could therefore not be based on literature data. On the other hand the correlation between CHL-a and I/ML was rather moderate (r = 0.56; n = 258) and CHL-a was already proposed for use as a production parameter in the scheme. Besides, the number of individuals per ml takes only the number of the present algae into account but not their volume. Therefore it is questionable whether it renders a better biomass measure than CHL-a already does.

3.2.5. Diversity

According to Caspers and Karbe (1967) the diversity of spe cies is decreasing from class II to class V, while class I is considered as moderately rich in species. Since class I cannot be expected in South Holland it is necessary to fit the observed diversity values in the classes II to IV (maybe V). The literature mentions no specific class limits for diversity in the Caspers and Karbe scheme, so the limits have to be set on the basis of this study by comparing the di versity values with other parameters. The diversity index of Margalef (D) shows no significant relationship with BOD but a negative signifi cant correlation with CHL-a and the oxygen content (Table 5). D also shows a fair correlation with SQ and SI* (resp. r = 0.37 and -0.24; n = 189 and 258) which made it possible to derive tentative class li mits for D by comparing these parameters. -39-

3.2.6. Saprobity

As for the saprobic quotient SQ and the saprobic index SI relation is sought with known class limits in the literature. Dresscher and van der Mark (1976) translated their SQ values into the classic saprobity classes of Kolkwitz and Marsson (1902), which in turn can easily be translated into the saprobic classes of Caspers and Karbe's scheme. Sladecek (1973) did already the same for the SI values of Pantle and Buck (1955). We could not assume that both SQ and SI were without doubt appropriate to the larger waters in South Holland. The saprobic quotient of Dresscher and van der Mark (1976) has not been used in the Netherlands for a long time and still has to prove its suitability on a larger scale. The SI index of Pantle and Buck was drafted for Mid-European, mostly running waters and can possibly not be.. applied without changes to the stagnant waters in South Holland. Although SQ and SI were mutually significantly correlated (r = -0.34; n = 189), SQ was better correlated with BOD (r = -0.53; n = 189) than SI (r = 0.37; n = 260), as given in Table 5. Moreover the majority of the SI values lies between 1.5 and 2.5, the p-mesosaprobic zone. This is mainly because the dominant blue- green algae in South Holland Oscillatoria agardhii, Oscillatoria limner . tica, Oscillatoria redekei and Microcystis aeruginosa. have in the indi cator lists of Sladecek (1973) and Mauch (1976) oligo- to p-mesosapror bic indicator values. This may be true for Mid-European rivers and deep lakes, in the Netherlands these species and also Lyngbya limne tic a and Microcystis viridis seem to represent the ultimate stage in the èutrophication process. An a-mesosaprobic (class IV) to polysaprobic (class;V)■■.value- would therefore be-more appropriate in Cas pers and Karbe's line of thought. Also Schuurmans (1970) and Ach terstraat et al. (1973). concluded :in ;studies;,in respectively the Kager Lakes and several waters in North Holland and South Holland that an a-mesosaprobic to polysaprobic indicator value for these species would be more appropriate for 'the situation 'in the "western part of the Netherlands. Therefore, the values indicating saprobity for these spe cies are changed, as is indicated in Table 7.( As. a consequence of this the saprobic indices for samples., dominated by the above mentioned blue-green algae are increased and shifted to the a-mesosaprobic zone. In Figure 4 the relation between BOD and Pantle and Buck's index with changed valencies for dominant blue-greens is shown. Correlation coefficients with the changed index (= SI* with saprobic values for blue-green algae derived from Table 7) prove that SI* is positively correlated with pH, COD, BOD, N-Kjeldahl and CHL-a and negatively with N02+N03-N and transparency (Table 5). The height of the correlation coefficients with COD, BOD and CHL-a is clearly increased. Also the correlation between SI* and SQ is clearly impro ved (r = -0.65; n = 189). From this is concluded that the calculation of SI* with changed saprobic values for blue-green algae is more ap propriate for the biological assesment of the water quality in Dutch stagnant bodies of water. Based on the above mentioned considerations and conclusions the following proposal is presented to make the scheme of Caspers and Karbe relevant to the Dutch situation (Table 8). The following re marks have to be made: -40-

Table 7. Saprobic valencies of several dominant blue-green species according to Sladecek (1973) and to this study.

Saprobl ic values according to

Name of species: Sladecek (1973) This study

s X 0 P o P G S s X (> P a P G S Lyngbya limnetica LEMM. a-p 6 4 3 3.4 Microcystis aeruginosa KÜTZ. P 3 6 1 3 1,75 a-p 6 4 3 3.4 Microcystis viridis (A.BR.) LEMM. a-p 6 4 3 3.4 Oscillatoria agardhii GOM. P 8 2 4 2,2 a-p 5 5 3 3.5 Oscillatoria limnetica LEMM. o-p 1 5 4 2 2,35 a-p 6 4 3 3.4

Oscillatoria redekei van GOOR p-o 4 6 3 1,6 a-p 6 4 3 3.4 s = indicator of saprobity by simple letter x = xenosaprobity o = oligogaprobity B - beta-mesosaprobity a = alpha-mesosaprobity p = polysaprobity G = indicative weight of species (range 0-5) S = saprobic index (range 0-8)

BOD versus SAPROBIC INDEX

3.5 - 1 ■ 21 2a 1 a i a 21 41 1 a 1 ■ 21 1 a 3B 2B 4a 5a SB 2a 21 1 a 1 a i a 1 a 3 - 3B 61 3a 1 a 1 a 1 a 4a 6a 4a 4a i a i a 2 a 1 a 3" 2" 3a 2B 1 ■ 1 a 2" 2" X 2.5 - 7* S« 6B 4a 7B 1 a 1 a 1 a 1 a 1 a 1 a Lü 3» 5" 21 1 a 1 B 1 a 1 a Q na 16» ioa 3B 1 a 21 i a 1 a 1 a Z 2 - ioa 9B SB 1 a 1 a Ü 21 ■ 5a 1 a 3m 4a m 4» 1 ■ 1 a o 1.5 - on 2" QL 3" < 2 ■ CO 1 -

0.5 -

0 - 1 r i l i i 1 i i 1 l l 1 12 16 20 24 28 BOD (mg/l) 16a NUMBER OF PAIRED OBSERVATIONS Fig. 4. Relation between Biological Oxygen Demand (BOD) and Pantle and Buck's (1955) saprobic index with changed valencies for dominant blue-greens (SI*). Table 8. Filling in of Caspers & Karbe's scheme with measurable parameters and class limits.

Water BIOACTIVITY OXYGEN f REGIME STRUCTURE OF COMMUNITY OF ORGANISMS quality class Production Respiration o2-observations 02-minimum Biomass Diversity Saprobity

2) 3) 4) 5) 6 7 8) Chlorophyll-a 1) BOD very good I/ML D > SQ > SI favourable (mg m"3) (mg 1"1) (%) (%) (mg l'1) (x 1000)

II < 25 < 3 > 85 and > 50 and 7 1 ± 0.7 2.25 ± 1.24 1.51 - 3.00 0.51 - 1.50

III A < 50 < 6 > 70 and > 40 and 5 4 ± 5 3.89 ± 1.32 0.76 - 1.50 1.51 - 2.00

III B < 100 < 9 > 55 and > 20 and 3 14 ± 27 4.26 ± 1.38 0.01 - 0.75 2.01 - 2.50

IV A < 150 < 13 > 40 45 ± 56 3.83 + 1.24 -0.75 - 0.00 2.51 - 3.00

IV B i 150 < 20 > 25 72 ± 64 2.84 ±1.20 -1.50 - -0.76 3.01 - 3.50

V p.m. < 30 S 25 and > 0.5 -3.00 - -1.51 3.51 - 4.50

VI i 30 S 25 and i 0.5 4.51 - 5.50

1) The summer mean, (april up to and including september). 2) 75 percentile of a year series (without allylthioureum addition).

4) "very good" 02-observation: within 80 and 120% saturation. 5) The mean number of individuals . ml ' * standard deviation. 6.) The mean diversity according to Margalef ± standard deviation. 7) Saprobic quotient according to Dresscher & van der Mark. 8) Saprobic index according to Pantle & Buck. -42- a. The table has to be seen as the elaboration of Caspers and Karbe's scheme for the large stagnant bodies of water in the western part of the Netherlands. Therefore, it is not suitable to smaller water bodies, such as ditches. This was also stated by de Hoogh (Provinciale Waterstaat van Noord-Holland, 1983) who verified the scheme in several watertypes in North Holland. b. The classes III and IV are subdivided into 2 subclasses A and B to get more discrimination in the mesosaprobic zone, where most of the waters in South Holland belong. c. Since class I" and class VI in the South Holland surface waters hardly can be expected, no class limits are given for all the parameters in class I and for several parameters in class VI. d. The BOD limits of class IV are somewhat relieved in comparison with Sladecek's (1973) levels (Table 6) to get a better agreement between the class limits of chlorophyll-a and to prevent that waters with high algal densities fall - based on their correspon ding high BOD values - automatically into the classes V and VI, where they do not belong according to the philosophy of Caspers and Karbe (Table 1). e. As can be seen the scheme is extended with, class limits for CHL-a as a production parameter. Since in class V a very mode rate production can occur mainly by mixotrophic species (Slade- cek, 1973) this possibility is stated by a p.m. f. With respect to the oxygen regime the criteria are based on "fa vourable" and "very good" 02-observations and the minimum oxy gen concentration. This is done because in the philosophy of Caspers and Karbe both the minimum oxygen content as well as the daily fluctuations play an important role. r g. For the aspects biomass (I/ML) and diversity (D,) no exact class limits have been filled in for earlier mentioned reasons. Instead of this the means and standard deviations of these parameters are given in the various classes according to the saprobic index of Pan tie and Buck. As can be seen the standard deviations are very large, which supports the statement that exact class deter mination based on these parameters is almost impossible. h. The class boundaries for SQ and SI are derived respectively from Dresscher and van der Mark (1976) and Sladecek (1973), while the p-mesosaprobic (class III) and a-mesosaprobic (class IV) zone is divided in two equal parts. See b. i. For the routine water quality assessment in South-Holland the criteria for BOD, oxygen and CHL-a of this scheme are used. Apart from that, investigation of the plankton community and determination of I/ML, D, SQ and SI is carried out more inciden tally at a limited number of sampling stations in a lower frequen cy, since plankton analyses are too expensive and time-consu ming to carry out frequently in large scale monitoring programs. The class determination takes place for BOD, 02 and CHL-a se parately according to the scheme and the highest calculated class determines the ultimate class of the sampled water. See Table 9 for some examples of the class determination. -43-

Table 9. Three examples of class determination according to the as pects bioactivity and oxygen regime in the proposed scheme (Table 8). For explanation of notes see Table 8.

°2" observations 02minimum Ultimate BOD 2) fa vourable 3) very good 4) class site: determi (mg m"3) (mg l"1) (%) (%) (mg l"1) nation :

5 21 5 75 8 5.3

16 132 13 100 75 7.2

20 42 6 83 42 3.2

4. DISCUSSION

4.1. Species composition

The algal species found in this study are characteristic for the surface waters in South Holland, since as much as possible different water types with differing degrees of pollution were sampled. De Lange and de Ruiter (1977) regard some of the species found in this study, e.g. Stephanodiscus hantzschii, Nitzschia acicularis, Oscillato- ria agardhii and Scenedesmus quadricauda more typical for the rivers Rhine and Meuse than for fresh basin- and polderwaters in the wes tern part of the Netherlands, although they state that river species also occur in polder and basin areas. Perhaps this illustrates the large influence of the River Rhine on the waters in South Holland, whieh cH=e for the greater part also fed by water from this river. The diatom species Centronella reicheltii VOIGT is regarded as a typical fluviatüe speciei (de Lange & dê~ Ruiter, 1977), characteristic for deep pools. This species was indeed found in one deep pool named "Kruithof Wiel". The pyrrhophyt Ceratium hirundinella O.F. MULLER SCHRANK was found both in deep pools as well as in deep sand pits and may be regarded as characteristic for deeper waters (>10 m). The very common and sometimes abundant central diatom Skeletonema potamos (WEBER) HASLE, synonym with Stephanodiscus subsalsus (Hasle & Evensen, 1976) is not very often reported from the Netherlands (Dresscher, 1976). The presumption exist however that this species is rather common in the Netherlands but easily overlooked or identified as a Tribonema species. In special diatom slides this species is hardly recognizable due to the very week silicated thecae. -44-

The very abundant filamentous blue-green alga in the Nieuwkoop and Reeuwijk lakes is called Lyngbya limnetica LEMM., in imitation of Leentvaar (1978), although a mucilaginous sheath was not always visable and the form therefore also could be named Oscillatoria limnetica LEMM. We found all transitions between forms with and without dis tinct sheats and even with forms identified as Oscillatoria redekëi van GOOR. Other authors doubt also if this "species" should be regarded as a distinct species (Moed & Hoogveld, 1982). The concerning spe cies is presumably the same as the chlorophyll-b containing prokaryote form in the Loosdrecht lakes (Burger et al. , 1985).

4.2. Algal typology

The similarity computations revealed a relative large measure of resemblance in algal composition between the sampling stations, espe cially when the algal composition was regarded in a quantitative sense. Only the more of less isolated surface waters diverge from the rest of the sampling sites. Apparently the mutual connections between the waters cause a levelling. Possibly the followed method, i.e. calculation of similarity between the mean algal composition on sampling sites over one or two years also contributed to this levelling. Furthermore the clustering of the similarity indices showed more distinction between different water types than between different degrees of pollution. This proves that similarity coefficients not necessarily contain infor mation on the pollution of a water as has been discussed by Chutter (1972, 1978), Pinkham and Pearson (1976) and Brock (1977) and re viewed by Washington (1984).

4.3. Bioactivity

The best parameter for measuring production in an ecosystem is primary production. Because none of the known techniques for measu ring primary production is suitable for routine analysis of a large number of sampling stations, this study used the actual chlorophyll-a concentrations instead. However, these are strongly influenced, not only by the primary production, but also by other* f actors, e.g. the retention time of a water and zooplankton grazing. This could perhaps be met by using so-called algal growth potential tests, in which the maximal growth potential of a test alga under laboratory conditions is measured (Environmental Protection Agency, 1971; Bolier et al., 1981; Forsberg et al., 1978; Ryding, 1980). Caspers (1977a) however stated expressly that the condition of a water does not depend on the poten tial, but on the actual processes. Therefore it seems more appropriate to use the chlorophyll-a concentration as a production parameter in the scheme. On the other hand Faegri (1954) defends, also on logical grounds, that the trophic state of a water can better be estimated on the potential productivity. In this study the parameter BOD is used as a decomposition parameter, although several authors e.g. Wilderer et al. (1977) and van Stralen and Kersting (1977) cast serious doubts on the use of this parameter. We found a rather good correlation between BOD and COD (r = 0.66; n - 1085), but no correlation at all between BOD and 02, probably because BOD reflects both decomposition processes (by measuring the 02-consumption) and - indirectly - also production (by measuring the respiration of algae). Since both processes play an -45- important role in South-Holland it is explicable that a correlation be tween BOD and 02 is absent. Wüderer et al. (1977, 1978) state that in the BOD parameter several oxygen demanding processes, such as substrate respiration, nitrification and endogenous respiration of bac teria and other organisms are expressed together without revealing the share each process has in the BOD determination. This is true, but does not implicate that BOD cannot be used as a measure for the decomposition, which means the total oxygen demand in a water. On the contrary, in surface waters also all above mentioned processes can simultaneously play a role. However it is clear that the conditions for the BOD measurement could be adapted more to reality. Although the BOD determination during five days at 20°C has been generally adopted in the water quality management and has proven its use in practice, it seems ad visable to investigate how laboratory conditions can influence the results. This applies especially for the temperature, since all mentio ned processes, for instance the nitrification, are temperature depen dent. Caspers (1975; 1977b; 1978) investigated successfully the ef fects of thermal pollution in the Elbe using BOD determinations at the prevailing water temperature. In many cases it was also possible to reduce the time for the determination to one or two days.

4.4. Oxygen

The concepts of "favourable" and "very good" oxygen observa tions was introduced to meet the ideas of Caspers and Karbe, that both the daily fluctuations as well as the mean or minimum concentra tion has to be taken into account with respect to the oxygen regime. In the coming years this has to be supported by continuous oxygen observations in a number of sampling stations in the larger waters of South Holland, after which the class limits for favourable and good oxygen observations and for the minimum oxygen concentration have to be evaluated. The continuous oxygen observations in larger waters suggest that the daily fluctuation is much smaller than sometimes is suggested in the literature (Ringelberg, 1976). A clear supersaturation during the daytime caused by algal blooms does not always cause an immediate oxygen deficit during the night.

4.5. Biomass

Since organisms differ widely in size and volume, it should be investigated to change and possibly improve the biomass parameter I/ML by taking into account the biovolume per ml instead of the num ber of individuals per ml. However, this would require data on the mean volume per organism. These can be found partly in the litera ture (e.g. Ringelberg, 1976) but most of them have to be estimated or calculated based on time consuming microscopical measurements.

4.6. Diversity

Although the different diversity parameters (S, D and H) are comparable fairly well, no clear relation is found between these para meters and the chemical water quality parameters (Table 6). In Table 8 therefore no exact class limits for D are given but only means and standard deviations per class. Especially in the lower classes II and -46-

IIIA diversity shows a large variation. This is in agreement with the findings of Archibald (1972) and van Dam (1982) who found that there exists no simple and consistent relation between species' rich ness and pollutional stage (cf. also Washington, 1984). The results of this study seem to confirm their findings. However, it is possible that also the applied method of plankton counting had a large influ ence on the calculated diversity values. Since the diversity calculation is based on approximately hundred random counted individuals in each plankton sample, this sample size might be large enough to give a fair view of the diversity on the polluted and eutrophicated sampling sta tions, but not of the diversity on the relatively unpolluted and clean sites. On the last sampling stations, with mostly a lower algal density, a diversity calculation based on a counting of hundred individuals may give an underestimation of the real species richness (cf. Washing ton, 1984).

4.7. Saprobity

In this study three saprobity parameters are determined (SQ, SI and Zelinka and Marvan's 10 point system (ZM)) in order to compare the results of the three methods. The fundamental difference between SQ on the one hand and SI and ZM on the other hand is that SQ supposes the indicative value of organism groups, while SI and ZM use organism species as indicators. So, already on theoretical grounds greater accuracy may be expected from the last two parameters, since higher taxa have presumably not the same narrow ecological amplitude as species. Proper identification of the species is of course necessary.

4.7.1. Quotient of Dresscher and van der Mark

The saprobity quotient SQ (Dresscher & van der Mark, 1976, 1980) is presented as a simplified method, based on the fact that some taxa surely have indicative value in general. Moreover the determina tion is easier and less time consuming than determination of the clas sic saprobity indices, because not every species has to be identified to species level. However, there are some disadvantages: one practi cal disadvantage is that the number of ciliate species must be coun ted on the day of sampling (Dresscher & van der Mark, 1976), since some ciliates die quickly. Furthermore Dresscher and van der Mark only partly standardized the sampling and the method of analysis. For instance, they did not prescribe the mesh width of the plankton gauze, through which the plankton is concentrated, and the amount of the concentrated material to be analysed. I think that especially the latter can have a rather large influence on the numbers of coun ted forms in the different groups. Moreover Dresscher and van der Mark (1976, 1980) state that the quotient cannot be applied during an algal bloom. In this study the quotient is applied to all situations, even those with severe algal blooms, because otherwise a large num ber of samples would be dropped. Moreover it is our impression that the determination of SQ, expecially during algal blooms, has led to very satisfactory results, since during blooms of blue-green algae the number of ciliates was rather high. In this study the determination of SQ is not exactly carried out as Dresscher and van der Mark described. The number of forms of different algal groups is namely derived from the random countings to -47-

hundred individuals in preserved and settled plankton samples, while the number of ciliate forms is determined in one liter living plankton samples concentrated through a 20 p plankton gauze. Rightly Dres- scher (personal communication) stated that this can cause an over- estimation of the number of ciliates with respect to the remaining groups. We opted for this method because one of the aims of this study was to compare SQ with other saprobity parameters such as SI and ZM, which have to be based on quantitative plankton countings. Dresscher and van der Mark (1980) suggest that SQ never should be applied in wintertime only when a larger number of ciliates could be found. Our results however did not show a seasonal effect: Both SQ as well as SI* are absolutely not correlated with water temperature (r = 0.02 and 0.00 resp.; n = 189 and 258), so the in other studies (cf. Heuss, 1976; Ziemann & Tümpling, 1981) established dependency of saprobity on temperature and season is not confirmed by this stu dy. SQ however does show a larger variation over the year than SI* (Fig. 5).

LANGERAAR LAKES (NR 16)

CLASS , SQ SI I , SI -05

-15 MA

HB - 00 2.5 EA

EB J5 35

.30 45

J ' F ' M ' A ' M ' J ' J ' A ' S ' 0 ' N ' D

PUT VAN BROEKHOVEN (NR.231 CLASS SQ , SQ i SI - 30 n - \S HA

HB _ 00 25 EA

SB - 45 35

2 -JO 45 m

J ' F ' M ' A ' M ' J ' J ' A ' S ' 0 ' N ' D

Fig. 5. The saprobity according to Pantle and Buck's index (SI*) and Dresscher and van der Mark's (1976) quotient (SQ) at two sampling sites (nrs. 16 and 23) during 1978. -48-

4.7.2. Index of Pantle and Buck

Pantle and Buck's (1955) index is calculated in this study using the saprobity indicating values presented by Sladecek (1973) or Mauch (1976). This study again has proven that indicator values can not be transferred from one water type to the other. Therefore the dominant blue-green species which showed the greatest discrepancy between saprobity value in the literature and the real situation in South Holland are changed (Table 7). The reason why foreign sapro bity values cannot always be applied to the Dutch situation is presu mably partly due to the fact that those values are conferred to run ning waters and not necessarily apply to stagnant waters. It is also possible that it is a result of the existence of different geographical races with varying tolerance or preference to pollution or eutrophication. Moreover Francke and ten Cate (1980) showed that algal species can show ecotypic differentiation over a relatively short distance as a response to nutritional factors. In my opinion also the saprobity value of other plankton species have to be carefully examined and, if neces sary, changed to the Dutch situation. This could be done by compa ring the presence and abundance of particular species with the che mical data of the sampling stations where the species occurred, as is done in this study. Several authors (Beer, 1961; von Tiimpling, 1962; von Tiimpling & Ziemann, 1962) tested already statistically the reliability of Pantle and Buck's saprobity index. We did not repeat this but checked whether or not the index is very sensitive to the numbers of indivi duals of several indicator species. This seemed not to be the case. In a hypothetical sample of twenty species the SI value changed only 0.08 when the numbers of a particular indicator species changed from ten to hundred and ten individuals. We did not use the indicative weight value G in Pantle and Buck's formula, because the authors themselves did not prescribe it. However, introducing an indicative weight factor in Pantle and Buck's formula, as sometimes has been done (cf. Waterschap Zuiveringschap Limburg, 1980), can increase the reliability and sensitivity of the index.

4.7.3. Zelinka and Marvan's 10 point system

What is said before about Pantle and Buck's index holds more or less true for the 10 points procedure of Zelinka and Marvan (1961) also. The results of this procedure are not enclosed extensively in this paper, because the method produces not one parameter but a distribution of ten points over five classes. Of course, a kind of weighed average per sample could be calculated, as van Nuland and Meis (1980) did, but it is the question whether this renders basically something different from Pantle and Buck's (1955) index. For the es sential part of Zelinka and Marvan's method is that the weighed eco logical amplitude of the total species composition is presented, which immediately is lost by averaging out the different classes.

4.7.4. Comparison of the saprobity parameters

Comparison of the SI and SQ values shows that generally SI gives a slightly more positive view. This concurs with the results of van Nuland and Meis. (1980), who concluded the same after comparing -49-

SQ with the index of Zelinka and Marvan averaged by them to one figure and based on the same indicator list of Sladecek. Also the by Dresscher and van der Mark (1976) given example of the river Rhine at Lobith shows slightly better values for SI than for SQ, although in their example both methods reveal the same pattern. The yearly fluc tuations in SQ by Dresscher and van der Mark in the river Rhine and by van Nuland and Meis (1980) in the Haarsteegse Wiel are not greater than the yearly fluctuations in the index of Pantle and Buck, which was so striking in this study (Fig. 5). On this ground we concluded that the advantage of a more simple and quicker method, which SQ is supposed to be, does not weigh up against the disad vantages of a less stable and more fluctuating parameter, which has therefore to be determined more frequently. We prefer therefore the index of Pantle and Buck over the index of Dresscher and van der Mark, especially when quantitative plankton countings are necessary anyway to get insight in the species composition and structure of the plankton community.

4.8. Trophism

In this study the estimation of the trophic stage is not based on the plankton communities because of the lack of an adequate method for this aspect in the waters in het western part of the Netherlands. The classical plankton quotient of Nygaard (1976) is unsatisfactory due to the lack of Conjugales. Also the method for biological typology and evaluation of fresh waters developed in the Netherlands by Coesel (1975) is only applicable to the oligotrophic and mesotrophic situation. The waters in South Holland belong without doubt to the eutrophic to hypertrophic zone, perhaps with the exception of a few isolated wa ters like some deep pools or sandpits, which could possibly be called mesotrophic. The difference between eutrophic and hypertrophic wa ters seems to be expressed more in quantitative aspects (higher pro duction and algal densities) than in qualitative aspects. Since the majority of the sampled waters in our province answers to the seven criteria for hypertrophy, stated by Leentvaar (1980), it seemed not very useful to look for qualitative criteria in the plankton associa tions. The degree of (hyper)trophy can presumably be better des cribed by aspects of the oxygen regime and bioactivity as we did in this study.

4.9. Biological assessment of water quality

With an objective "biologically healthy" water it is inevitable to take into account the biology of waters in water quality assessment. This does not imply that biological assessment of the water quality should take the place of the routine assessment based on physico- chemical measurements. To the contrary, both methods are comple mentary. The framework of Caspers and Karbe has the advantage that it is based on an ecosystem philosophy, in contradiction with other water quality assessment systems, such as presented in both Dutch Water Action Programmes (1975-1979 and 1980-1984), in which the choice of parameters is more or less subjective and arbitrary and in principle unlimited (Ministry for Transport and Public Works, 1975, 1980). In this manner the second Water Action Programme (1980-1984) presents a list of 39 parameters and objectives, needed to determine whether a water meets the description "basic water quality" or not. -50-

It is almost impossible to get a quick overall impression of the water quality from such a list. The three ways of class determination in the system of Caspers and Karbe, namely on the basis of bioactivity, the oxygen regime and the structure of the biotic communities make the system very suitable for the practical water quality management. The functional parameters to measure the bioactivity and oxygen regime (BOD, CHL-a, 02) are very suitable to "influence" water systems over a relatively short time and to follow the short term effects of certain water purification me thods. The structural parameters, based on the algal communities (I/ML, D, SI*, SQ), can serve to follow water quality changes over a longer period of time and to check whether or not certain objectives are achieved.

5. CONCLUSIONS a. The water quality system of Caspers and Karbe (1966, 1967) is chosen as a suitable framework for the biological assessment of the water quality in the western part of the Netherlands. Accor ding to the authors, water quality can be assessed in three in dependent ways:

1. according to the bioactivity (ratio production/respiration); 2. according to the oxygen regime; 3. according to the structure (biomass, diversity, saprobity) of the communities in the water. An elaboration of this framework for the large stagnant waters in the province of South-Holland was carried out using the phytoplankton composition and its relation with the physical and chemical water characteristics. b. The plankton in the South Holland waters can be generally cha racterized as a green algae (Chlorococcales) and diatom (especi ally Centrales of the genus Stephanodiscus) community. In some lakes the plankton community is dominated by a permanent bloom of filamentous blue-greens (Lyngbya limnetica, Oscillatoria agardhii, Oscillatoria redekëi) or a periodic bloom of Microcystis aeruginosa. c. Morphological and hydrological watertypes (canals, lakes, peat lakes, deep sand pits, deep pools) are reflected in the species composition, which is also determined further by the degree of pollution and eutrophication. d. Physical, chemical and ecological parameters, that are measured in this study, are investigated for possible use in Caspers and Karbe's scheme. By means of correlation and regression analyses the following parameters were selected: chlorophyll-a (CHL-a) for measuring the production; BOD for measuring the decomposition; oxygen-content and saturation for measuring the oxygen regime; number of individual algae per ml (I/ML) for measuring the biomass; Margalef's index (D) for measuring the diversity and Dresscher and van der Mark's quotient (SQ) or Pantle and Buck's index (SI) for measuring saprobity. -51- e. By changing the saprobic valencies of the dominant bluegreen algae Lyngbya limnetica, Microcystis aeruginosa, Microcystis viridis, Oscillatoria agardhii, Oscillatoria limnetica, Oscillatoria rede- këi from oligo/B-mesosaprobity to a-meso/polysaprobity a rather good correlation was achieved between most of the selected para meters. Only D and I/ML showed a rather large variation in the different classes of the scheme and could therefore not or less well be used for water quality assessment. f. A new scheme with measurable water quality criteria was propo sed based on the framework of Caspers and Karbe's scheme (Table 8). The new scheme is meant for the large waters in the western part of the Netherlands. For other watertypes (e.g. dit ches) and regions it should be adapted. g. Both SQ and SI* (with changed saprobity valencies for blue- greens) can be applied in the scheme for measuring the sapro bity, although SQ showed a larger variation throughout the year and therefore has to be determined more frequently. h. The proposed scheme and criteria can be applied both in routine water monitoring programs, using CHL-a, BOD and 02 as simple and cheap functional parameters, as well as in special water qua lity studies, using also the more complicated structural parame ters I/ML, D and SI or SQ. i. The results, especially with respect to the use of saprobity in dices in water quality assessment, are compared with literature data and discussed.

6. ACKNOWLEDGEMENTS

This research was financially supported by the province of South Holland. The author is greatly indebted to Prof. Dr. H. Caspers, who encouraged him strongly to publish the results in an internatio nal journal. The author is grateful to Drs. J. van der Does, Prof. Dr. M. Donze and Prof. Dr. W.H.O. Ernst for critical reading of the manuscript, to Mr. W..G. Hey for performing the statistical analyses, and to Miss A. Honnef for correcting and Miss C. van Dijk for typing the English text.

7. REFERENCES

Achterstraat, J.A., M. Burggraaff, B.J. Flach-de Geus & S.P. Klap wijk, 1973. Beoordeling van het oppervlaktewater van het Groene Hart vol gens biologische maatstaven, gebaseerd op inventariserend onder zoek aan fytoplankton. Vrije Universiteit, Instituut voor Milieu vraagstukken, Verkenningen, serie B no. 3.

Archibald, R.E.M., 1972. Diversity in some South African diatom associations and its rela tion to water quality. Water Res. 6: 1229-1238. -52-

Beer, W.D., 1961. Methodologische Untersuchungen zur biologische Fliessgewasser- analyse. Int. Revue ges. Hydrobiol. 46: 5-17.

Bolier, G., A.N. van Breemen & G. Visser, 1981. Eutrophication tests: a possibility to estimate the influence of sewage discharge on the biological quality of the receiving water. H20 14: 88-92 (in Dutch with an English summary). Brock, D.A., 1977. Comparison of community similarity indexes. J. Wat. Pollut. Contr. Fed. 49: 2488-2494.

Burger-Wiersma, T., M. Veenhuis, H.J. Korthals, C.C.M, van de Wiel & L.R. Mur, 1986. A new prokaryote containing chlorophylls a and b. Nature 320: 262-264. Caspers, H. , 1966. Stoffwechseldynamische Gesichtspunkte zur Definition der Sapro- bitatsstufen. Verh. internat. Verein. Limnol. 16: 801-808.

Caspers, H. , 1975. B.O.D. experiments on the effect of thermal pollution in rivers and estuaries. Verh. internat. Verein. Limnol. 19: 3192-3198.

Caspers, H. , 1977a. Qualitat des Wassers - Qualitat der Gewasser. Die Problematik der Saprobiensysteme. In: Sladecek, V. (ed.). Symposium on Saprobiology. Ergebn. Limnol. 9: 3-14.

Caspers, H., 1977b. Experiments for a prediction of the effects from artificial heating in an estuarine system. Prog. Wat. Techn. 9: 631-636.

Caspers, H. , 1978. Experimentelle Untersuchungen zur Abschatzung der Auswirkung einer Warmebelastung von Flüssen. Warme-einleitungen in Gewas ser und deren Auswirkungen. Workschop 27/28 September 1977, Karlsruhe.

Caspers, H. & L. Karbe, 1966. Trophie und Saprobitat als stof f wechsel- dynamischer Komplex. Gesichtspunkte für die Definition der Saprobitatsstufen. Arch. Hydrobiol. 61: 453-470.

Caspers, H. & L. Karbe, 1967. Vorschlage für eine saprobiologische Typisierung der Gewasser. Int. Revue ges. Hydrobiol. 52: 145-162.

Chutter, F.M. , 1972. An empirical biotic index of the quality of water in South African streams and rivers. Water Res. 6: 19-30. -53-

Chutter, F.M., 1978. Discussion to "Applications of a new coefficient of similarity to pollution surveys" van C.F. Pinkham & J.C. Pearson, 1976. J. Wat. Pollut. Contr. Fed. 50: 791-792.

Coesel, P.F.M., 1975. The relevance of desmids in the biological typology and evalua tion of fresh waters. Hydrobiol. Bull. 9: 93-101.

Dam, H. van, 1982. On the use of measures of structure and diversity in applied diatom ecology. In: H. Hakansson & J. Gerloff (eds.). Fest schrift Niels Foged. Diatomaceae III. Beiheft zur Nova Hedwigia 73.

Does, J. van der & S.P. Klapwijk, 1987. Effects of phosphorus removal on the maximal algal growth in bioassay experiments with water from 4 Dutch lakes. Int. Revue ges. Hydrobiol. 72: 27-39.

Dresscher, Th.G.N., 1976. Index van de namen en vindplaatsen die betrekking hebben op in Nederlandse wateren aangetroffen algen en enige groepen van micro-organismen. North-Holland, Amsterdam.

Dresscher, Th.G.N. & H. van der Mark, 1976. A simplified method for the biological assessment of the quality of fresh and slightly brackisch water. Hydrobiologia 48: 199-201.

Dresscher, Th.G.N. & H. van der Mark, 1980. Experience with a simple method for the biological evaluation of surface water quality. Hydrobiologia 71: 169-173.

Environmental Protection Agency (E.P.A.), 1971. Algal Assay Procedure, National Eutrophication Research Program. Corvallis, Oregon.

Everitt, B., 1974. Cluster analysis. Heinemann Educational Books Ltd, London. Faegri, K., 1954. Some reflections on the trophic system. Nytt. Magasin for Bota- nik 3: 43-49.

Forsberg, C, S.-O. Ryding, A. Claesson & A. Forsberg, 1978. Water chemical analyses and/or algal assay? - Sewage effluent and polluted lake water studies. Mitt, internat. Verein. Limnol. 21: 352-363.

Francke, J.A. & H.J. ten Cate, 1980. Ecotypic differentiation in response to nutritional factors in the algal genus Stigeoclonium Kiitz. (Chlorophyceae). Br. phycol. J. 15: 343-355. -54-

Hasle, G.R. & D.L. Evensen, 1976. Brackisch water and fresh water species of the diatom genus Skeletonema II. Skeletonema potamos comb. nov. J. Phycol. 12: 73-82.

Heuss, K., 1976. Untersuchungen zur Bewertung von Verfahren der biologischen Gewasser-Beurteilung. Schriftenreihe Landesanstalt Wasser Abfall Landes Nordrhein-Westf. H. 36.

Hovenkamp, I.R.M., S.P. Klapwijk & J.E.F. Landman, 1982. Biological assessment of the water quality in Noord-Holland and Zuid-Holland. H20 15: 406-412 (in Dutch with an English sum mary).

Huber-Pestalozzi, G., 1950-1962. Das Phytoplankton des Süsswassers. Teil 1-5. E. Schweizerbart'- sche Verlagshandlung, Stuttgart.

Klapwijk, S.P., 1978. Saprobiological evaluation of Dutch ditches using benthic algae on artificial substrates. Verh. internat. Verein. Limnol. 20: 1811- 1815.

Klapwijk, S.P., 1980. Effect of laundry wastewater on benthic algae in ditches in the Netherlands. Hydrobiol. Bull. 14: 142-152.

Klapwijk, S.P., 1982. Hydrobiologisch onderzoek naar de uitwerking van het waterkwa liteitsklassensysteem van Caspers en Karbe voor grotere wateren in Zuid-Holland. Rapport Hoogheemraadschap van Rijnland, Tech nische Dienst, Leiden.

Klapwijk S.P., T.F. de Boer & M.J. Rijs, 1983. Effects of agricultural wastewater on benthic algae in ditches in the Netherlands. In: R.G. Wetzel (ed.): Periphyton of freshwa ter ecosystems, p. 322-319. Junk, The Hague.

Kolkwitz, R. & M. Marsson, 1902. Grundsatze fiir die biologische Beurteilung des Wassers nach sei ner Flora und Fauna. Mitt. Priifungsanst. Wasserversorgung. Abwassereinigung 1: 33-72.

Krebs, C.J., 1972. Ecology. The experimental analysis of distribution and abundance. Hasper & Row, New York - Evanston - San Francisco - London.

Lange, L. de, M.A. de Ruiter & Werkgroep Biologische Waterbeoorde ling, 1977. Biologische waterbeoordeling. Methoden voor het beoordelen van Nederlands oppervlaktewater op biologische grondslag. Instituut voor Milieuhygiëne en Gezondheidstechniek TNO, Delft. -55-

Leentvaar, P., 1978. De hypertrofiëring van onze binnenwateren. Een hydrobiologisch onderzoek van de Reeuwijkse plassen in 1976/1977. Rapport Rijks instituut voor Natuurbeheer, Leersum.

Leentvaar, P., 1980. Comparison of hypertrophy on a seasonal scale in Dutch inland waters. In: J. Barica & L.R. Mur (eds.). Hypertrophic ecosys tems. Developments in Hydrobiology 2: 45-55. Junk, The Hague.

Liebmann, H. (ed.), 1960-1962. Handbuch der Frischwasser- und Abwasserbiologie. Band I & II. R. Oldenbourg, Miinchen. Mauch, E., 1976. Leitformen der Saprobitat fiir die biologische Gewasseranalyse. Cour. Forsch. Inst. Senckenberg, 21. Frankfurt am Main.

Ministery for Transport and Public Works, 1975. The combat against surface water pollution in the Netherlands. Prospective multiannual Program 1975-1979.

Ministry for Transport and Public Works, 1980. Water Action Programme: 1980-1984. The Netherlands. The prin ciples and main outline of national policy to maintain the quality of Dutch surface waters.

Moed, J. & H.L. Hoogveld, 1982. The algal periodicity in Tjeukemeer during 1968-1978. Hydrobio- logia 95: 223-234.

Mueller-Dombois, D. & H. Ellenberg, 1974. Aims and methods of vegetation ecology. J. Wiley & Sons, New York.

Nederlands Normalisatie Instituut. Tests methods for waste water. NEN 3235 (in Dutch). Nuland, G.J. van & J.F.G.M. Meis, 1980. Comparison of a few systems for the determination of saprobic and trophic degree on the basis of plankton data. Hydrobiologia 70: 251-256.

Nygaard, G., 1976. Tavlerne fra Dansk Planteplankton. Gyldentalske Boghandel, Nordisk Forlag A. S., Copenhagen.

Odum, E.P., 1971. Fundamentals of ecology. 3rd ed. W.B. Saunders Co, Philadelphia - London -Toronto.

Odum, H.T., 1956. Primary production in flowing waters. Limnol. Oceanogr. 1: 102- 117. -56-

Pantle, E. & H. Buck, 1955. Die biologische Überwachung der Gewasser und die Darstellung der Ergebnisse. Gas- Wasserfach 96: 604.

Pascher, A., 1913-1930. Die Siisswasserflora Deutschlands, Oesterreichs und der Schweiz (ab 1930: Mitteleuropas). Heft 1-12. Gustav Fischer Verlag, Jena.

Pinkham, C.F.A. & J.G. Pearson, 1976. Applications of a new coefficient of similarity to pollution surveys. J. Wat. Pollut. Contr. Fed. 48: 717-723.

Prescott, G.W. , 1964. How to know the fresh water algae. 2d ed. Wm. C. Brown Co Publ., Dubuque, Iowa, USA.

Provinciale-Waterstaat van Noord-Holland, 1983. Waterkwaliteitsbeoordelingsmethodiek. Een verdere evaluatie op basis van in 1982 door de provincie Noord-Holland verricht wa terkwaliteitsonderzoek, Haarlem.

Ringelberg, J. , 1976. Inleiding tot de aquatische oecologie in het bijzonder van het zoete water. Bonn, Scheltema & Holkema, Utrecht.

Ryding, S-O. (ed), 1980. Monitoring of inland waters. Report from the working group for eutrophication research. Nordforsk publication 1980: 2, Helsing- fors, Finland.

Schuurmans, R.R., 1970. Hydrobiologisch onderzoek aan de Kagerplassen. Doctoraal ver slag Zoölogisch Laboratorium, afdeling Milieubiologie, Rijksuni versiteit Leiden / Rijksinstituut voor Natuurbeheer (R.I.N.), Zeist.

Sladecek, V., 1973. System of water quality from the biological point of view. Ergebn. Limnol. 7: 1-218.

Sladecek, V. , 1977. Zum Verhaltnis Saprobitat Trophie. In: V. Sladecek (ed.). Symposium on Saprobiology. Ergebn. Limnol. 9 : 79-93.

Sokal, R.R. & F.J. Rohlf, 1969. Biometry. The principles and practice of statistics in biological research. W.H. Freeman & Co, San Francisco.

Stralen, M. van & K. Kersting, 1977. BOD§° -method, not suited for measuring quality of surface wa ters. H20 10: 329-331 (in Dutch with an English summary).

Streble, H. & D. Krauter, 1974. Das Leben in Wassertropfen. Mikroflora und Mikrofauna des Süsswassers. 2 nd ed. Franckh'sche Verlagshandlung, Stuttgart. -57-

Tümpling, W. von, 1962. Statistische Probleme der Biologischen Gewasserüberwachung. Wasserwirtschaft/Wassertechnik 12: 353-357.

Tümpling, W. von & H. Ziemann, 1962. Beitrag zur Kritüc der Einheitsmethodik in der biologischen Ge wasserüberwachung. Mitt. Inst. für Wasserwirtschaft 12: 119-130.

Uherkovich, G., 1966. Die Scenedesmus-arten Ungarns. Akadémiai Kiadó - Verlag der Ungarischen Akadémie der Wissenschaften, Budapest.

Vries, P.J.R. de & S.P. Klapwijk, 1987. Bioassays using Stigeoclonium tenue Kütz. and Scenedesmus quadricauda (Turp.) Bréb. as testorganisms; a comparative study. Hydrobiologia 153: 149-157.

Waterschap Zuiveringschap Limburg, 1980. Biologische waterbeoordeling op grond van waterfauna-onderzoek met behulp van diverse saprobie-systemen. I. De Duits-Limburgse grenswateren in 1980. Roermond.

Werff, A. van der & H. Huls, 1957-1974. Diatomeeënflora van Nederland. Abcoude. Wilderer, P., G. Engelmann & H. Schmenger, 1977. Kritik am BSB5 als Verschmutzungsparameter. Gas- Wasserfach 118: 357-364.

Wilderer, P., L. Hartman & G. Keser, 1978. Der Plateau - BSB als Mass für die Konzentration an biologisch abbaubarer Substanz. Z.f. Wasser- und Abwasser-Forschung 11: 82-86.

Zelinka, M. & P. Marvan, 1961. Zur Prazisierung der biologischen Klassification der Reinheit fliessender Gewasser. Arch. Hydrobiol. 57: 389-407.

Ziemann, H. & W. von Tümpling, 1981. Untersuchungen über jahreszeitliche Veranderungen der Sapro- bitat in einem Fliessgewasser. Acta Hydrochim. Hydrobiol. 9: 131-146.

Washington, H. G., 1984. Diversity, biotic and similarity indices. A review with special relevance to aquatic ecosystems. Water Res. 18: 653-694. -59-

CHAPTER 4: COMPARISON OF HISTORICAL AND RECENT DATA ON HYDRO- CHEMISTRY AND PHYTOPLANKTON IN THE RIJNLAND AREA (THE NETHERLANDS)

Submitted to: Hydrobiologia.

"Ten aanzien van de technische betekenis van dit werk zij opgemerkt dat die betekenis, zowel voor Rijnland als voor andere waterbeheren- de instanties, sedert 1942 onveranderd is gebleven, De waarde dier gegevens is in hoofdzaak hierin gelegen, dat zij vergelijkingsma teriaal leveren en of men later, bijvoorbeeld over 25 of 100 jaar een vergelijking treft met 1941 of met 1950 is van geen belang."

P. de Gruyter & E.L. Molt, 1950. Rijnlands boezem. Deel III. De hoedanigheid van het boezemwater. Rapport Hoogheemraadschap van Rijnland, p. II, Leiden. -60-

COMPARISON OF HISTORICAL AND RECENT DATA ON HYDRO- CHEMISTRY AND PHYTOPLANKTON IN THE RIJNLAND AREA (THE NETHERLANDS)

S.P. Klapwijk

Key words: eutrophication, phytoplankton, waterchemistry, saprobity, ecological objectives, canal, lake.

ABSTRACT In two canals and two lakes in the western part of the Nether lands a comparison is made between data on water chemistry and phy toplankton from 1942/1942 with recent data. Especially orthophosphate but also inorganic nitrogen has increased tremendously, especially in the Gouwe canal where Rhinewater is entering the area. The inorganic N/P mass ratio decreased in the last 45 years indicating that the limi ting nutrient has been changed from phosphate in 1941 to nitrogen in 1987. The average sestonvolume, measured by filtrating 100 1 water through a planktonnet (50 um), has generally been doubled. In the early 1940's the blue-green alga Microcystis aeruginosa regularly for med waterblooms, as it does now. The plankton composition seems to have impoverished in the last decennia, since several taxa have dis appeared e.g. Anabaena circinalis, Synura spp., Eudorina elegans, Pandorina morum, Ceratium hirundinella, Dinobryon spec, Fragilaria crotonensis and Tabellaria fenestrata. Some others (e.g. Aphanizo- menon flos-aquae, Pediastrum spp., Asterionella formosa, Diatoma elongatum) are strongly reduced in number. Microcystis aeruginosa and Melosira italica seemed to have maintained their number, while M. granulata (var. angustissima) seems to have increased. The saprobic index according to Pantle and Buck (1955) has not been changed. This concurs with the results of Peelen (1975), who compared the changes of plankton composition of the Rhine and Meuse during the last hundred years. Based on the relations between chlorophyll-a and biological oxy gen demand (BOD) and between transparency and seston volume, BOD, dry weight and ash free dry weight (AFDW) in 1987 chlorophyll-a and transparency in 1941 are estimated. It is concluded that the average chlorophyll-a concentration in the lakes has doubled or tripled in the last 45 years and that mean transparency in the Gouwe canal is redu ced from 75 to 50 cm. At the other sampling sites this is not so ob vious, although the submerged higher plants at some sites have dis appeared in the last decennia. The results can be used to develop ecological objectives for combatting eutrophication in canals and lakes and possibly even the river Rhine. -61-

INTRODUCTION

A comparison of historical and recent plankton data is interesting in many respects. It gives information if changes have occurred over time. Moreover it can indicate the effects of general pollution, like acidification or eutrophication (cf. Osborne & Moss, 1977; van Dam et al., 1978; Higler, 1979; Moss, 1979; van Dam, 1980, 1987). Therefore in history can be looked for a "natural" situation as a reference, as suming the situation was better then, that is to say less polluted or in other ways influenced than nowadays. This is important too in view of standardization (cf. Klapwijk, 1987). A problem of comparing older with recent plankton data is, that older research often has not been quantitative. Mostly the abundance is noted in estimates of the numbers as "many", "few" and "sporadic". In addition to quantitative counts most often measurements like the chlorophyll-a concentration are missing, which could give an idea of the total algal biomass. Peelen (1975), in a comparison of the plankton composition of the rivers Rhine and Meuse from the beginning of this century, was wrestling with this problem too. Nevertheless data which could give an impression of quantitative consequences of increasing eutrophication, are very welcome and useful, also in the view of standardization of eutrophication parameters such as phosphate, nitro gen and chlorophyll-a. Data being available in literature and the archives about com position and quantity of plankton in lakes, pools and canals in the Waterboard of Rijnland area (Fig. 1) are mostly also presented quali tatively (cf. Lauterborn, 1918; Redeke, 1923; Romijn, 1924; van der Werf f, 1943; Leentvaar, 1956, 1963; Hofstede & Blok, 1964; Salomé, pers. comm. 1964, 1967). In most cases they concern analyses of net- plankton samples with semi-quantitative estimates of the numbers in five categories. Only since 1970 quantitative plankton counts have been done in settled samples and/or the, total algal biomass has been quantified with the aid of chlorophyll-a measurements (cf. Schuurmans, 1970; Achterstraat et al., 1973; Schmidt-van Dorp, 1978; Klapwijk, 1981; Hoogheemraadschap van Rijnland, 1984). However, in an investigation in the canals and lakes in south east Rijnland, van der Werff (1943) has during 1941 and 1942 used some techniques to measure the total quantity of seston. He filtrated 100 1 water samples through a planktonnet with meshwidth of 50 urn and measured the seston volume in cm3. In addition he carried out some physical and chemical analyses, which give also an impression of the total plankton biomass and the degree of eutrophication. In 1986 and 1987 four of the sampling sites used by van der Werff have been reinvestigated as far as possible in the same manner. In addition the variables important for eutrophication like chloro phyll-a and transparency have also been determined. Preliminary re sults have already been published elsewhere (Klapwijk et al. , 1988). The purpose of this investigation is a comparison between the change of water quality and its effects on plankton organisms to get an im pression of the qualitative and quantitative consequences of eutrophi cation that has been increasing as a result of population growth and the use of detergents with phosphate (Golterman, 1976; C.B.S., 1985). -62-

MATERIAL AND METHODS

Van der Werff has during 1941 and 1942 done a hydrobiological investigation on nine places in the Rijnland area. The complete results of his research, among others counting lists with semi-quantitative estimates of amounts of different species of plankton, are kept in the Rijnland archives. For a comparison of water quality and plankton quality and quantity four of his sample places were selected (Fig. 1).

Fig. 1. The area of the Waterboard of Rijnland and the position of the sampling sites.

Table 1 gives some characteristics of the sampling sites. Two sampling sites (31C and 106) are situated in lakes, two (32 and 116) in canals. Both the canals and lakes form part of the basin system of the Rijnland Waterboard, which is an interconnected system of canals, lakes and ditches with the same water level (0.6 m below mean sea level). Fresh water can be taken in the system near Gouda, from the Hollandsche Ussel, a branch of the Rhine, while excess water can be pumped out to the sea by various pumping stations. -63-

Table 1. Some hydrological characteristics of the sampling sites.

Site Name Width Depth Average surface nr. On) (m) (106 m2) 31C Westeinder lakes 2.9 8.9 32 Ringvaart canal 45 3.0 106 Lake Braassem 3.5* 4.6 116 Gouwe canal 50 3.4

* with the exception of deep sandpit

In table 2 the amount of water let in the basin system at Gouda and discharged at the pumping stations of Spaarndam, Halfweg, Gouda, Katwijk and Leidschendam is summarized for the years of con cern. A rather great amount of Rhine water had to be let in at Gouda in the Rijnland area during 1941 and 1942. Winter of 1941/1942 was severe, so from 29 December 1941 till 14 April 1942 the sampling had to be abandonned because of heavy ice. Sampling therefore took place in the period of June through December 1941 and April through Sep tember 1942.

Table 2. Intake of water at Gouda and discharge of water at the pumping stations Spaarndam, Halfweg, Gouda, Katwijk and Leidschendam in 1941, 1942, 1986 and 1987.

1941 1942 1986 1987

Intake at Gouda (x 106 m3) 173 111 88 45 Discharge at: Spaarndam ) 93 116 143 165 Halfweg ) 224 194 187 194 Gouda ) 80 69 78 78 Katwijk ) 202 155 207 272 Leidschendam ) - - 24 7 Total discharge (x 106 m3) 599 534 639 716

The samples were taken as follows: according to circumstance 5-10 buckets of water, each holding ten liters, were filtered through a finely mazed planktonnet (meshwidth 50 urn); the filtrate was put in a glass tube and fixed with a few drops of formalin. The samples were concentrated by centrifuge (2000 r.p.m.; 15 sec.) in the labora- -64- tory. The seston volume was measured in cm3. Floating blue-green algae were left out of the measurement because of non settling and therefore could not be centrifuged. The organisms occurring in the water were grouped in tables and the frequency of occurrence was in dicated by: ccc - extremely common cc - very common c - common + - present r - rare rr - very rare

In 1941/1942 samples were taken to determine watertemperature, pH, chloride, chemical oxygen demand (COD) by KMn04 consumption, total and bicarbonate hardness, ammonium, sulfate, .color (in mg Pt), iron, nitrite, nitrate, C02, BOD5 (at room temperature), phosphate, dry weight and loss on ignition (LOI).

At the four places measurements were made in 1986 and 1987, as far as possible in the same way. To calculate the saprobic degree ac cording to Pantle. and Buck (1955) the old and recent estimates (rr, r, +, c, cc, ccc) are translated into a scale (1, 2,. 3, 5, 7, 9) cor responding with Sladecek (1973). The saprobic valencies of various phytoplankton species given by Sladecek (1973) and Mauch (1976) were used here, while the saprobic valencies of some dominant blue- green algae have been adjusted conform Klapwijk (1982, 1988). Watersamples were taken at nearly 0.5 m depth in which the pH, N03+N02-N, NH4-N, P04-P, total-P, Cl", BOD§° , dry weight, LOI, ash free dry weight (AFDW), COD (by KMn04 consumption) and chlorophyll-a has been determined. Transparency has been determined with the aid of a Secchi-disk. Because of the_ comparison the histori 3 cal data about ammonium (NH , ), nitrite (N02) and nitrate (N03) on the one side and ortho-phosphate (POf ) on the other' side, the data were recalculated in both milligrams N and P. Inorganic-N has been calculated as the sum of nitrite, nitrate and ammonium values. The in organic-N/ortho-P ratio has been calculated by division of the weights of inorganic-N and ortho-P. Historical data about dry weight and loss on ignition have been recalculated in AFDW. Correlation and regres sion analyses between various variables have been done according to Sokal and Rohlf (1969).

RESULTS The results of the physical and chemical measurements, as well as the results of the plankton counts are summarized in Table 3. In comparing the nutrient values in 1941/1942 with that of 1986/1987 it is evident that the orthophosphate values have strongly increased at all sampling sites. The nitrogen values have increased in the Gouwe canal and lake Braassem but remained at the same level in the West- einder lakes (Fig. 2). Inorganic nitrogen, however, has increased less than orthophosphate. This is evident from the decrease of the inorganic-N/ortho-P ratio; for example at the Rijnland inlet pumping station in the Gouwe canal near Gouda this ratio was around 40 in 1941/1942 and has decreased since to values around 15 (Fig. 3A). Table 3. Summary of physical, chemical and biological variables (means i's.d.) during 1941/1942 and 1986/1987 at four sampling stations. The number of observations are given in parentheses. "-" = not determined.

Sampling stations: Westeinder lakes Ringvaart ; canal Lake Braassem Gouwe canal variables year: 1941/1942 1986/1987 1941/1942 1986/1987 1941/1942 1986/1987 1941/1942 1986/1987 PH 8.3 ± 0.5 8.5 ± 0.2 - 8.0 ± 0.3 8.3 ± 0.6 8.3 ± 0.3 7.6 ± 0.3 7.6 ± 0.2 (31) (14) (41) (31) (14) (31) (34) Transparency (m) - 0.62 ± 0.21 - 0.73 ± 0.2 - 0.77 ± 0.40 - 0.47 1 0.18 (11) (39) (11) (22)

Chloride (mg f') 119 ± 16 141 ± 5 125 ± 20 140 t 16 112 ± 20 143 * 11 109 1 40 152 1 40 (31) (14) (26) (54) (31) (14) (31) (41) NH„-N (mg f') 0.26 ± 0.13 0.10 1 0.03 - 0.70 i 0.46 0.28 ± 0.15 0.25 t 0.20 0.81 ± 0.49 1.73 1 1.38 (31) (21) (22) (31) (20) (31) (22)

(N0 +N0 )-N 3 2 (mg f') .0.61 t 0.42 0.62 t 0.79 - 2.25 t 1.27 1.04 1 0.61 1.94 ± 1.35 1.94 ± 0.83 3.87 1 0.79 (31) (21) (20) (31) (20) (31) (21) Inorganic-N (mg 1"') 0.87 ± 0.47 0.72 ± 0.80 - 2.83 ± 1.60 1.32 ± 0.64 2.19 t 1.46 2.75 ± 0.87 5.40 1 1.49 (31) (21) (22) (31) (20) (31) (20) Ortho-P (mg f') 0.07 ± 0.05 0.14 ± 0.09 - 0.47 t 0.19 0.10 ± 0.06 0.35 ± 0.09 0.10 t 0.06 0.48 t 0.26 (31) (21) (21) (31) (20) (31) (21)

Total-P (mg f') 0.26 ± 0.08 0.65 ± 0.20 0.45 ± 0.16 0.76 1 0.35 - " - - I (21) (21) (20) (21) a> en Inorganic-N/Ortho- P 16.9 1 13.8 10.8 ± 14.3 - 7.4 ± 4.9 16.7 + 10.7 6.6 ± 4.7 43.2 t 30.1 14.1 1 5.2 i (31) (21) (21) (31) (20) (31) (20) BOD (mg 1"') 3.1 ± 1.1 4.6 ± 1.5 - 4.9 t 2.0 3.0 ± 0.8 4.3 ± 3.3 3.8 ± 1.2 5.8 1 3.2 (31) (14) (35) (31) (14) (31) (34) COD (mg l">) 58 ± 10 58 ± 22 - 55 ± 13 48 t 13 53 ± 14 48 t 18 56 1 26 (31) (14) (14) (31) (11) (31) (15) Dry weight (mg f') 42 ± 58 16 ± 7 - 17 ± 11 15 1 8 14 ± 14 20 ± 8 52 t 79 (16) (14) (34) (16) (14) (31) (20) LOI (%) 70 ± 12 55 ± 20 - 44 ± 19 88 1 9 47 ± 23 60 1 19 38 ± 29 (16) (13) (13 (16) (11) (29) (12) AFDW (mg f') 26 ± 30 9 14 - 6 t 4 13 ± 7 10 t 13 12 1.4 22 1 42 (16) (13) (13) (16) (11) (29) (12) Chlorophyll-a (mg m 3) - 62 ± 17 - 51 t 29 - 84 1 151 - 33 1 29 (14) (29) (14) (28)

Sestonvolume (cm^.lOO 1~' ) 1.6 ± 1.8 2.4 1 1.8 1.3 ± 1.0 2.1 t 1.0 1.2 ± 1.2 2.0 1 1.8 1.2 1 1.4 3.0 ± 2.8 (25) (23) (28) (23) (28) (23) (28) (23)

Number of species 8.7 ± 4.1 10.7 ± 4.7 10.7 ± 4.5 10.4 1 3.8 10.2 ± 3.1 7.5 ± 3.4 14.2 1 4.6 10.4 1 4.3 (25) (23) (28) (23) (28) (23) (28) (23)

Saprobic index 2.20 1 0.29 2.03 ± 0.20 2.10 ± 0.19 2.05 ± 0.25 2.10 ± 0.24 2.09 1 0.33 1.69 1 0.09 1.82 1 0.16 . (25) (23) (28) (23) (28) (23) (28) (23) -66-

WESTEINDER LAKES WESTEINDER LAKES

LAKE BRAASSEM LAKE BRAASSEM

1.0 -

0.9 -

0.8 -

0.7 -

0.6 -

0.5 -

0.4 -

0.3 -

0.2 -

0.1 - D O D' ..°'

' 11 2 3 4 5 6

GOUWE CANAL GOUWE CANAL

' ' I i r- 9 10 11 12 1 9 10 11 12 't 2 3 4 5 6 7 8 9 MONTHS MONTHS -D 1941/42 ° 1986/87 1941/42 1986/87

Fig. 2. Monthly averages of concentrations of orthophosphate as mg P 1 * (left) and inorganic nitrogen as mg N 1 1 (right) in the Westeinder lakes, lake Braassem and Gouwe canal, in 1941/1942 and 1986/1987. -67-

In the Westeinder lakes and lake Braassem also a distinct decrease of this ratio is perceptible. Several authors (cf. Forsberg et al. , 1978; Claesson & Forsberg, 1980; de Vries & Klapwijk, 1987) accept 12-15 as critical range for the inorganic N/P ratio, above which phosphate and below which nitrogen is limiting for algal growth. Also the chlo ride value has risen by some tens of milligrammes since 1942, espe cially in the Gouwe canal, in which water from the Hollandsche IJssel is let in the Rijnland area in case of low precipitation (Fig. 3B).

GOUWE CANAL GOUWE CANAL A. 220 - B.

200 -

^1M- : '> CTM60 - >—'140 - / \ / 'n..., \

Q v.. .> \ l ó - ...« \ 1 'D'" CHLORID E

a *D fl - V '* '' 40 - :--v^-— 20 - 1 1 1 ' ' ', 1 ' 1 10 11 12 1 ' ' ' ' ' ' 10 11 12 1 MONTHS MONTHS 1941/42 o 1986/87 1941/42 1986/87

Fig. 3. Ratio of inorganic nitrogen to orthophosphate (A) and chlo ride concentration (B) in the Gouwe canal in 1941/1942 and 1986/1987.

Moreover Table 3 shows that the sestonvolume at the sampling sites has about doubled since 1941/1942. This is also evident when we set up graphically the sestonvolume determinations at the different sampling sites (Fig. 4), although the data do show a rather large fluctuation in time. Especially in the lakes the sestonvolume has gone up drastically. It is striking that the increase in planktonbiomass is not so clearly expressed in the remaining variables, that can give an indication of the biomass, namely BOD, COD, dry weight and AFDW, although BOD and COD do show the same tendency. Dry weight and AFDW have only increased in the Gouwe canal. The LOI has decreased by about a factor of two at the different sites. The qualitative comparison of plankton shows that the total num ber of species has not changed substantially. Neither can one decide on a change in the saprobic grade according to Pantle and Buck (1955). All selected waters belong to the p-mesosaprobic class, where by the Gouwe canal inclines mostly to the oligosaprobic class. This probably is because blue-green algae are less dominant (cf. Klapwijk, 1982). In Table 4 the estimated mean and maximum amounts of the ob served phytoplankton species are presented. During the last 45 years certain species have disappeared from the plankton picture, e.g. Ana- baena circinalis, Eudorina elegans, Pandorina morum, Volvox aureus, -68-

Ceratium hirundinella, Peridinium spp., Coscinodiscus rothii, Dino- bryon spp., Fragilaria crotonensis, Synura spp. and Tabellaria fenestrata. Other species like Aphanizomenon flos-aquae, Pediastrum spp., Asterionella formosa, Diatoma elongatum, Fragilaria construens, Suri- rella robusta and Synedra acus have been decreased in number. Only Microcystis aeruginosa and Melosira italica seem to have held their frequency in these waters for the last 45 years. Some taxa are only found in the recent years and not in 1941/1942, e.g. Lyngbya limnetica, Phormidium mucicola, Coscinodiscus subsalsus, Fragilaria pinnata, Gyrosigma spp., Nitzschia romana, Sceletonema potamos, Stephanodis- cus dubius. Melosira granulata and M. granulata var, angustissima seemed to have increased in number.

WESTEINDER LAKES RINGVAART CANAL

6 7 8 9

LAKE BRAASSEM GOUWE CANAL i 110 - o O 9 - —*\a -

E 7- o > _->i »- 9 O 4- / \ z> J, hn- c/i 2- j/ LÜ - 1/1 1 - Q. \ J-* T^* "~^\y ,-■■■:

»..v ,."■ 0 - 1 1 1 1 1 1— "D fl' i i r , i i 1 1 1 1 1 1 1 r 2 3 4 5 6 7 a 9 10 11 12 1 234567B9 9 10 11 12 1 6 7 8 9 MONTHS MONTHS 1986/87 o 1941/42 ■ 1986/87 1941/42

Fig. 4. Sestonvolume (cm3 per 100 1 x) in the Westeinder lakes, Ringvaart canal, lake Braassem and Gouwe canal in 1941/ 1942 (D) and 1986/1987 (<>). Table 4. Assessed mean and maximum amounts of observed phytoplankton taxa at four sampling stations during 1941/1942 and 1986/1987. Values from the 1941/1942 samples are derived from van der Werff (1943).

Cyanobacteria/. Aphanocapsa spp. rr (+) rr (r) (rr) Anabaena circinalis Rabenhorst <+) (cc) Anabaena spiroides Klebahn (D rr (r) rr (rr) rr (r) Aphanizomenon flos-aquae (L.) Ralfs (cc) rr (r) (ccc) rr (rr) rr (+) Chroococcus spp. (+) r (+) (r) rr (+) (r) rr c+) rr (rr) Coelosphaerium kuetzingianum Nageli (rr) Hormogonales spp. rr (+) rr (c) rr (c) rr (♦) Merismopedia glauca (E.) Nageli (r) rr (rr) Microcystis aeruginosa Kützing (ccc) c (ccc) (ccc) (ccc) (ccc) c (ccc) rr (rr) rr (rr) Lyngbya limnetica Lemmermann rr (r) (r) rr (r) rr (r) Oscillatoria spp. (r) (r) rr (r) rr (r) rr (r) Phormidium mucicola Naumann et Huber-Pestalozzi rr (c) (+) rr (r) rr (r) Chlorophyta Actinastrum hantzschii Lagerheim rr (+) Closterium spp. rr (rr) (rr) rr (r) (rr) rr (rr) rr (r) rr (r) I Coelastrum microporum Nageli (r) rr (rr) (r) rr (r) rr (rr) rr (rr) en Cosmarium spp. rr (rr) CO Crucigeniella rectangularis (Nageli) Kom. (r) Dictyosphaerium spp. (r) rr (rr) rr (r) rr (c) Eudorina elegans Ehrenberg rr (rr) (rr) rr (+) Geminella spp. (cc) rr (r) Kirchneriella spp. (rr) (rr) Lagerheimia ciliata (Lagerheim) Chodat (r) Micractinium pusillum Fresenius Microspora spp. (cc) rr (rr) rr (r) rr (+) Monoraphidium spp. rr <+) rr (rr) Oedogonium spp. rr (r) ■ Oocystis spp. rr (r) rr (r) Pandorina morum (Muller) Bory (D rr (*) rr (+) rr (rr) r (c) Pediastrum, boryanum (Turp.) Menegh. (c) * (+) r (c) r <♦) r (r) r (♦> r (♦) rr (r) Pediastrum clathratum (Schrot.) Lemm. rr (rr)

Pediastrum duplex Meyen (c) rr (r) rr (r) rr (♦) rr (r) rr (r) r Cc) rr (r) Scenedesmus acuminatus (Lagerheim) Chodat (rr) rr (r) rr (rr) rr (r) rr (rr) rr « ■ rr (rr)

Scenedesmus quadricauda (Turp.) de Brébisson (c) r (♦) rr (r) . r (r) rr (r) rr (♦) rr (r) rr (r) Scenedesmus spp. rr (♦) . rr (r) rr (rr) rr (r) rr (rr) Siderocelis ornata (Fott) Fott (rr) Table 4. (continued)

Sphaerocystis schroeteri Chodat rr (rr) Spirogyra spp. rr (r) Staurastrum spp. rr (rr) rr (rr) Tetrastrum staurogeniaeforme(Schrod.) Lemmermann rr (r) Ulothrix spp. rr (r) rr (r) rr (rr) Volvox aureus Ehrenberg rr (rr) rr (+) rr (+) rr (rr) Westella botryoides (W. West) De-Wild. rr (rr) Euglenophyta Euglena acus Ehrenberg rr (rr) Phacus longicauda (E.) Dujardin rr (rr) rr (rr) rr (r) Trachelemonas volvocina Ehrenberg rr (rr) rr (r) Pyrrhophyta Ceratium hirundinella (Muller) Schrank rr (r) rr (rr) rr (rr) Peridinium spp. rr (+) rr (rr) Chrysophyta Actinoptychus splendens (Shadbolt) Ralfs rr (rr) Actinoptychus undulatus (Bailey) Ralfs rr (rr) rr (r) rr (rr) rr (rr) rr (rr) rr (rr) rr (rr) Amphora ovalis Kiitz. rr (rr) Asterionella formosa Hass. r (+) r (cc) rr (c) rr (rr) c (ccc) r (♦) Aulacodiscus argus (Ehrenb.) Schmidt rr (rr) rr (rr) rr (rr) Bacillaria paradoxa Gmelin rr (r) rr (rr) rr (r) rr (rr) rr (r) Biddulphia spp. rr (rr) rr (rr) rr (rr) Caloneis amphisbaena (Bory) Cleve rr (rr) Campylodiscus spp. rr (r) rr (r) rr (rr) rr (r) rr (rr) Cerataulus smithii Ralfs rr (rr) Chrysococcus spp. rr (r) Coscinodiscus excentricus Ehrenberg rr (rr) rr (rr) rr (rr) Coscinodiscus lacustris Grunow rr (rr) rr (rr) rr (r) rr (rr) rr (r) rr (r) rr (r) Coscinodiscus rothii (Ehrenberg) Grunow rr (rr) rr (rr) rr (+) Coscinodiscus subsalsus Juhlin-Dannfelt rr (rr) rr (r) rr (r) Cyclotella meneghiniana Kiitzing rr (r) rr <+) rr (r) Cyclotella striata (Kützing) Grunow rr (r) Cymatopleura elliptica (de Bréb.) Smith rr (+) rr (r) rr (r) rr (+) rr (r) rr (rr) Cymatopleura solea (de Bréb.) Smith rr (rr) rr (rr) rr (rr) rr (r) Cymbella spp. rr (r) Diatoma elongatum (Lyngbye) Agardh rr (rr) rr (r) rr (r) rr (r) rr (+) rr (r) + (.cc) rr (+) Diatoma vulgare Bory rr (rr) rr (+) rr (r) rr (r) rr (r) rr (r) rr (rr) Table 4. (continued)

Sampling stations: Lake Westeinder Ringvaart canal Lake Braassem Gouwe canal Years: 1941/1942 1986/1987 1941/1942 1986/1987 1941/1942 1986/1987 1941/1942 1986/1987 Number of samples: 25 23 28 23 28 23 28 23 mean (max.) mean (max.) mean (max.) mean (max.) mean (max.) mean (max.) mean (max.) mean (max.) Dinobryon sertularia Ehrenberg rr (r) rr (rr) rr (rr) Dinobryon sociale Ehrenberg rr (-O rr (r) rr (+) rr (rr) Fragilaria capucina Desmazières rr (r) rr <+) rr (+> rr <+) rr (+) Fragilaria construens (Ehrenb.) Grunow rr (r) rr (+) rr (rr) rr (+> rr (rr) rr (rr) Fragilaria crotonensis Kitton rr (rr) rr (c) rr (+) c (ccc) Fragilaria pinnata Ehrenberg r (+) rr (r) rr Gyrosigma spp. rr (r) rr (r) rr (r) Melosira granulata (Ehrenb.) Ralfs rr (+) rr (cc) rr (+) + (ccc) rr (c) r (ccc) rr (c) r (c) Melosira granulata var. angustissima Muller r (cc) r (cc) r (cc) rr (+) r (c) Melosira italica (Ehrenb.) Kützing + (cc) r (c) + (ccc) + (cc) + (ccc) r (c) + (ccc) c (ccc) Melosira varians Agardh rr (rr) rr (r) rr (+) r (-0 rr (r) rr (r) rr <+) rr (+) Melosira spp. rr (r) rr (rr) Navicula spp. rr (c) rr' (r) Nitzschia acicularis Smith rr (rr) Nitzschia holsatica Hustedt rr (+) Nitzschia romana Grunow r (c) r (cc) rr (r) -*0 Nitzschia sigma (Kützing) Smith rr (r) rr (rr) rr (rr) rr (r) rr (r) I Nitzschia sigmoidea (Ehrenb.) Smith rr (rr) rr (rr) rr (r) rr (rr) rr (r) rr (r) Pinnularia nobilis Ehrenberg rr (rr) Podosira stelliger (Bailey) Mann rr (rr) Rhoicosphenia curvata (Kützing) Grunow rr (r) rr (rr) Sceletonema potamos (Weber) Hasle rr (r) rr (r) rr (r) rr (c) Stephanodiscus astraea (Ehrenb.) Grunow rr (+) rr (r) S. astraea var. minutula (Kützing) Grunow rr (r) rr (r) rr (r) Stephanodiscus dubius (Fricke) Hustedt rr (r) rr (r) Stephanodiscus hantzschii Grunow rr (+) rr (r) rr (r) rr (r) Surirella robusta Ehrenberg rr (r) r (+) rr (r) r (+) rr (rr) r (c) r (+) Surirella spp. rr (rr) rr (rr) rr (rr) rr (rr) rr (+) rr (rr) rr (rr) rr (r) Synedra acus Kützing rr (rr) rr (r) rr (rr) rr (rr) + (cc) rr (rr) Synedra ulna (Nitzsch) Ehrenberg rr (rr) rr (r) rr (+) rr (rr) rr (r) rr (rr) Synedra spp. rr (rr) rr (r) rr (r) Synura spp. rr (r) rr (c) rr (c) rr (c) rr (rr) Tabellaria fenestrata (Lyngbye) Kützing rr (rr) rr (r) rr (+) (cc) Thalasstosira spp. rr (+> Triceratium favus Ehrenberg rr (r)

: -rr = very rare; r = rare; present; c = common; cc = very common; ccc = extremely common. -72-

Table 5. Matrix of product-moment correlation coefficients between se veral variables connected with plankton biomass.

Seston BOD COD Dry AFDW Chl-a Transpa volume weight rency

Sestonvolume 1.00 0.19 0.04 0.31** -0.02 -0.00 -0.44* (138) (135) (110) (101) (57) (41)

BOD 1.00 0.19 0.15 0.14 0.53** -0.41** (145) (184) (109) (111) (116)

COD 1.00 -0.14 -0.04 0.04 -0.10 (116) (108) (52) (36)

Dry weight 1.00 0.85** 0.13 -0.51** (110) (104) (99)

AFDW 1.00 0.24 -0.65** (48) (34)

Chlorophyll-a 1.00 -0.32* (92)

Transparency -

* V <0.01; ** P <0.001; the number of paired observations is given in parentheses.

To be able to assess the transparency values and chlorophyll-a concentrations in 1941/1942, correlation and regression calculations are executed between sestonvolume, BOD, COD, dry weight, AFDW, chlorophyll-a and transparency. The results of the correlations are shown in Table 5. No correlation between sestonvolume and chloro phyll-a appears, but there is a significant negative correlation be tween sestonvolume and transparency (Fig. 5A for a graphic presen tation). A positive significant relation between sestonvolume and dry weight has been established, which indicates that the sestonvolume presents a fairly good image of the transparency of the water and the amount of suspended matter in it. Transparency shows also a signifi cant (negative) correlation with BOD, dry weight, AFDW and chloro phyll-a. Fig. 5JB and 5C are showing the relationships with BOD and dry weight. Transparency in 1941/1942 can therefore be assessed on the basis of its relation with sestonvolume, BOD, dry weight or AFDW with the following regression formulas:

Y = - 0.098 Xx + 0.863 Y = - 0.063 X2 + 1.059 Y = - 0.020 X3 + 1.100 Y = - 0.017 X4 + 0.779 in which Y = transparency (in m) Xi = sestonvolume (in cms 100 1 0 1 X = BOD (in mg 1 ) 2 l X3 = dry weight (in mg 1 ) 1 X4 = AFDW (in mg 1 ) -73-

TRANSPARENCY - SESTONVOLUME RELATIONSHIP TRANSPARENCY - BOD RELATIONSHIP

- A. 3.2 - B. 3.0 - - ■ 2.B - f ■ v"- ■ ■ U2.o- z . LÜ 1.»- ■ & 1.6 J ■ ■ ■ ■ Q<_ 1.«- ■ £,.2- ■ : Y=-0.0625X+1.059 m < 1.0- B . R=-0.40**; N-112 ■ P ■ PP -j 1 ■ • ^^^ Y=-O.098X+0.863 £o..- ■ ■ ■ ■■ ■ ^^^^^ R=-0.J5; N=39 : ■ 1 ~r 0.6 - ■ ■ ■ 1 - ■ ^^-^^ 0.4 - ■ ! ' ■ ■ ■ ■ ^~~~~~~~~-^_^^ 0.2 - a 0 □ ■ ~~n"~— 1 1 1 1 1 1 1 r ■ ->*--i i ' 5 7 9 II 13 15 SESTONVOLUME [cm3/! 001] BOD [mg/l]

TRANSPARENCY - DRY WEIGHT RELATIONSHIP CHLOROPHYLL - BOD RELATIONSHIP

■ 190 - ' D. c. 180 - 1—'170 - ""e 160 - \150- p a> 140 - E t30 - 1 ' 120 - -/Y=12.2X-7.2 m O 110 - m ■ R-0.53**; N=111 1 100 - p p m —S 90 - ■ ■ m p p Y--0.020X+1.1 V 80- 1 ■ ■■ R=-0.48**; N=98 1 ■ ■ ■ P ■ ft" "60" 1 i O ' ^m ■ p cr so - ■ ■ ■ ■ ■ ■ ■-•■■«*■—- P O M- 1 1 III "■■• ppa_"p ■ P PR P ■!■•" P~"BJ-—~__ 1 i p P P ■ ■ P^ p O* »201 - P 1 p p 10 - //i 1 0 - -f- T 1 1 1 1 1 ■ a -+- DRY WEIGHT [mg/l] BOD [mg/l] D TRANSPARENCY < 0.2m

Fig. 5. Regression plots between transparency and sestonvolume (A), transparency and BOD (B), transparency and dry weight (C) and between chlorophyll-a and BOD (D) based on data from 1986 and 1987. Note: Very low transparency values (<0.2 m) in Fig. 5A-C are not included in the calculations.

Very low transparency values (<0.2 m), probably caused by floating blue-green algae, are regarded as unreliable and are not included in the calculations (Fig. 5A-C). Likewise the chlorophyll-a concentration can be estimated from the positive linear correlation between chloro phyll-a and BOD (Fig. 5D) according to the regression formula:

Y = 12.209 X2 - 7.206 in which Y = chlorophyll-a (in mg m 3) and 1 X2 = BOD (mg l" ) -74-

With the aid of these linear regression formulas the average transparency and chlorophyll-a concentrations on the various sampling sites are estimated for 1941/1942 (Table 6). This table shows that the transparency in the Westeinder and Braassem lakes (31C and 106) and in the Ringvaart canal (32) has not decreased significantly, but that the transparency in the Gouwe canal (116) must have fallen from about 75 cm to about 50 cm. The average chlorophyll-a concentration has doubled in the Westeinder lakes to almost tripled in lake Braassem.

Table 6. Calculated and measured transparency and chlorophyll-a con centrations (means ± s.d.) at four sampling sites in 1941/ 1942 and 1986/1987 respectively.

Table 6. Calculated and measured transparency and chlorophyll-a concentrations (means ± s.d.) at four sampling sites in 1941/1942 and 1986/1987 respectively.

TRANSPARENCY (m)

1941/1942 values calculated from: Measured Site name (1986/1987) number Sestonvolume BOD Dry weight AFDW

31C Westeinder lakes 0.70 + 0.17 0 .87 ± 0.07 0.58 ± 0.30 0.45 ±0.20 0.62 ± 0.21 (n 25) (n = 31) (n = 16) (n = 16) (n = 11)

32 Ringvaart canal 0.74 + 0.09 - - - ' 0.73 + 0.20 (n 28) (n = 39)

106 Lake Braassem 0.75 ± 0.12 0 .87 ± 0.05 0.81 ± 0.16 0.56 ± 0.11 0.77 ± 0.40 (n —28 ) (n = 31) (n = 16) (n = 16) (n = 11) 116 Gouwe canal 0.75 ± 0.14 0 .82 ± 0.07 0.71 ± 0.15 0.58 ± 0.06^ 0.47 + 0.18 (n —28 ) (n = 31) (n = 31) (n = 29) (n = 22)

CHLOROPHYLL-a (mg m 3)

1941/1942 values calcijlate d from: Measured BOD (1986/1987)

31C Westeinder lakes 30 ± 13 62 + 17 (31) (14)

31 Ringvaart canal - 51 ± 29 (29)

106 Lake Braassem 30 ± 10 84 ± 151 (31) (14)

116 . Gouwe canal 39 ± 14 33 ± 29 (31) (28)

= not determined; n = number of observations.

DISCUSSION

From the data presented it is evident that the nutrient concen trations and the plankton biomass in the Rijnland basin waters have risen quite strongly. The mean orthophosphate concentrations have increased by a factor of* two (Westeinder lakes) to five (Gouwe canal). Inorganic nitrogen concentrations in lake Braassem and the Gouwe -75-

canal have about doubled. Assuming that P04-P was about 2/3 of total-P, as is the case now (cf. Table 3), implies that the mean total- P concentration in the Gouwe canal, where Rhinewater was and still is entering the area, was about 0.15 mg P 1 1 in 1941/1942, compared to 0.75 mg p-l"1 nowadays. De Wit (1980) and van Acht and de Jong (1983) described that P04-P in the river Ussel, another branch of the river Rhine in the Netherlands, has been increased from <0.05 in 1932 up to about 0.4 mg P f1 in 1978. However, in lake Ussel the P04-P concentration has risen less and is largely dependent on the distance to the inflowing river Ussel. Differences in the weather between the years compared will pro bably have weakened rather than strengthened the conclusions, be cause 1941 and 1942 were rather dry years with a great need for wa ter intake, while 1986 and 1987 were rather wet years (cf. Table 2). Therefore the intake of eutrophic Rhine water has been relatively small in the previous two years, and weather conditions have been relatively unfavourable for algal growth. At first glance the plankton composition seems not to have changed drastically. Likewise there occurred in the basin waters of Rijnland regularly (that is in summer) blooms of the blue-green alga Microcystis aeruginosa in 1941/1942 and before that (cf. Lauterborn, 1918; Romijn, 1924), but the algal composition was rather more varied as is evident from the comparison in Table 4. Next to Microcystis there occurred regularly Aphanizomenon f los- aquae and Anabaena circinalis as dominant blue-green algae, with in addition flagellates and diatoms that do not occur or do so much less nowadays. It is unlikely that differences like this would be caused because the estimates have been done by two different researchers. Also in the index of the names and finding places of microorganisms by Dres- scher (1976) for example Anabaena circinalis and Fragilaria crotonensis are not mentioned anymore for lake Braassem and the Westeinder lakes, while these lakes have been researched by among others Leent- vaar (1956) and Salomé (pers. comm. 1964, 1967). Only with respect to some species, e.g. Lyngbya limnetica, Phormidium mucicola, Nitz- schia romana, Skeletonema potamos it might be possible that discre pancies between 1941/1942 and 1986/1987 are caused by differences in conception of two researchers, since these species might have been overlooked or identified as an other species (Klapwijk, 1988). A qualitative judgement according to the saprobic index by Pan tie and Buck (1955) did not produce a significant difference. Peelen (1975) too concluded in his study of plankton in the rivers Rhine and Meuse, that the plankton structure between the beginning of the century and the seventies has changed somewhat but that the saprobic level has not moved. He makes probable that since 1916 an increase in the amount of plankton took place, probably caused by increasing eutrophication and by lengthening of the residence times in the Rhine and Meuse branches. In the polder lakes in the Rijnland area, which are situated in polders at about 2 m below main sea level separated from the basin system, the change in species structure probably has been much greater. Lauterborn (1918) described that the plankton structure in the Nieuwkoop lakes in the beginning of this century consisted of Microcystis aeruginosa and Anabaena spiroides with in addition Pedias- trum spp., filamentous algae (Oedogonium) and flagellates like Dino- bryon sertularia, D. stipitatum and Ceratium hirundinella. Redeke -76-

(1923) detected mainly Synura uvella, Peridinium spec, Pediastrum boryanum, P. duplex, Scenedesmus quadricauda, Microcystis aeruginosa, Oscülatoria spp. and various naviculoids in the intestins of breams from the Langeraar lakes. Scheygrond (1933) mentions that the Reeuwijk lakes were well known among Dutch hydrobiologists in the thirties because here the natural succession of various groups, like those of diatoms and coloured flagellates in spring, the coherent appearence of Copepoda and Cladocera and the dominance of Rotifera in summer and fall could be nicely observed. At this moment all men tioned polder lakes are considerably hypertrophied and the algal structure consists one-sidedly of filamentous blue-green algae of the genera Lyngbya and Oscülatoria (Klapwijk 1981; Hoogheemraadschap van Rijnland, 1984). These polder lakes are naturally rather oligotrophic to mesotrophic while the basin canals and lakes probably al ways have been eutrophic because of the greater influence of Rhine water and the wastewater of cities like Gouda, Leiden and Haarlem (Klapwijk & Smit, 1988). Therefore the consequences of eutrophication have problably been smaller in the basin waters than in the more iso lated polder lakes. Because of the absence of a correlation between chlorophyll-a and sestonvolume it was not possible to estimate the chlorophyll-a percentage in 1941/1941 this way. The absence of a relation with chlo rophyll-a is probably connected with the fact that floating blue-green algae are not included in the sestonvolume determination. By way of the established positive correlation between chlorophyll-a and BOD it has however been possible to estimate the chlorophyll-a content in 1941/1942. The same holds for the relationship between transparency and sestonvolume, BOD, dry weight and AFDW. On the basis of these relationships the conclusion can be made that the transparency in the Rijnland basin waters was 45 years ago not very different from the present except in the Gouwe canal, where the mean transparency is reduced from approximately 75 to 50 cm and that the summer average of the chlorophyll-a concentration in the lakes was about two to three times as low as nowadays. Yet, the light climate in these waters was probably more favoura ble in those years for submerged waterplants than nowadays. This can be deduced from the description of the submerged vegetation in the canals and lakes as given by van der Werff (1943). At site 32 (Ringvaart canal) he found on 10 July 1941 Potamogeton pectinatus, P. perfoliatus and Lemna trisulca and at site 106 (lake Braassem) P. crispus and P. pectinatus. A year later he found no vegetation in the open water of lake Braassem, but along the banks a moderate growth of P. perfoliatus and Enteromorpha intestinalis. At this moment even this moderate bank vegetation has vanished there. The above findings can be used for setting ecological standards (cf. C.U.W.V.O., 1986). So can the observed nitrogen and phosphate concentrations and ratio of inorganic nitrogen/orthophosphate in 1941/ 1942 be used as ecological objectives for canals and lakes in the wes tern part of the Netherlands. Moreover, the observed concentrations in 1941/1942 at sampling site 116 near Gouda can probably also be used to set standards for the lower course of the river Rhine. In this respect it would be interesting to compare and evaluate the expected effects of the Rhine Action Programme (IKSR, 1987) with these data. Also standards of transparency and chlorophyll-a can critically be viewed in the light of the results of this research. -77-

Moreover it is evident from the results that a basjc quality level, i.e. a summer average for chlorophyll-a of 100 mg m 3, as described in the Dutch Water Action Programma 1985-1989 (Ministry for Trans port and Public Works, 1985) is very generous for these lakes domi nated by Microcystis. Even very hypertrophic lakes as lake Braassem and the Westeinder lakes, showing an enormous blooming of Microcys tis every year, conform largely to this norm. On one side this is be cause the blooming is very temporary (August-September). On the other hand chlorophyll-a measurements during blooms of Microcystis often give an underestimation of the real biomass, because this alga can float or sink, so routine samples taken at 0.5 m depth give not a realistic image of the amount of algae present, if no other sampling techniques are being used.

CONCLUSIONS

1. Orthophosphate but also inorganic nitrogen and chloride concen trations have increased significantly from 1941 to 1987 in canals and lakes in the Rijnland Waterboard area, especially in the Gouwe canal where Rhinewater is entering the area.

2. The ratio of inorganic N/orthophosphate has decreased in the last 45 years indicating that the limiting nutrient has probably been changed from phosphate in 1941/1942 to nitrogen in 1986/ 1987.

3. The sestonvolume, measured by filtrating 100 1 water through a planktonnet (50 um), has generally been doubled.

4. The plankton composition seemed to have changed in the last de cennia, since several taxa e.g. Anabaena circinalis, Eudorina elegans, Pandorina morum, Volvox aureus, Ceratium hirundinella, Dinobryon spp., Fragilaria crotonensis, Synura spp. and Tabel- laria fenestrata have almost disappeared and some others, e.g. Aphananizomenon flos-aquae, Pediastrum spp., Asterionella formosa, Diatoma elongatum are strongly reduced in numbers. Only Microcystis aeruginosa and Melosira italica seemed to have main tained their numbers, while M. granulata (var. angustissima) seemed to have increased in number.

5. As in the rivers Rhine and Meuse, the saprobic index according to Pantle and Buck (1955) has not changed during the last 45 years in the canals and lakes from the basin system of Rijnland.

6. From the relations between chlorophyll-a and BOD and between transparency and seston volume, BOD, dry weight and AFDW it is concluded that the average chlorophyll-a content in the lakes is doubled to tripled in the last 45 years and that the mean trans parency in the Gouwe canal is reduced from about 75 to 50 cm.

7. In the Ringvaart canal and lake Braassem the submerged vegeta tion, consisting of Potamogeton spp., Lemna trisulca and Entero- morpha intestinalis has disappeared in the last decennia.

8. The results can be used to develop ecological objectives for ca nals and lakes and indirectly probably also for the river Rhine. -78-

A C KNOWLEDGEMËNT S Miss L. van" den Hove carried out the recent plankton identifica tions and estimations, Mr. R. Westhoek assembled the historical and recent hydrological data and Mr. P. Nieuwpoort determined the seston volumina and compiled the data for statistical analysis and graphical presentation. The author is greatly indebted to Mr. A. van der Werff for critical and inspiring discussions with him on the methods applied in 1941/1942 and nowadays. I thank Prof. Dr. M. Donze and Prof. Dr. W.H.O. Ernst for valuable suggestions with respect to the manus cript and Miss A. Honnef for correcting and Miss C.A.L.M. van Dijk for typing the English text.

REFERENCES

Acht, W.N.M, van & J. de Jong, 1983. Ecological aspects of the Zuiderzee Project. Werkdocument, Rijks dienst voor de IJsselmeerpolders, Lelystad.

Achterstraat, J.A., M. Burggraaff, B.J. Flach-de Geus & S.P. Klap wijk, 1973. Beoordeling van het oppervlaktewater van het Groene Hart vol gens biologische maatstaven, gebaseerd óp inventariserend onder zoek aan fytoplankton. Vrije Universiteit, Instituut voor Milieu vraagstukken, Verkenningen, serie B rio. 3, Amsterdam.

C.B.S., 1985, Fosfor in Nederland. Centraal bureau voor de statistiek. Staats uitgeverij, 's Gravenhage.

Claesson, A. & A. Forsberg, 1980. Algal assay studies of polluted lakes. Arch. Hydrobiol. 89: 208- 224.

C.U.W.V.O., 1986. Ecological quality objectives for surface waters in the Nether lands . Coördinatiecommissie Uitvoering Wet Verontreiniging Opper vlaktewateren (in Dutch).

Dam, H. van, 1980. Veranderingen in de vennen bij Oisterwijk tussen 1840 en 1976. Natura 77: 98-111.

Dam, H. van, 1987. Verzuring van vennen: een tijdsverschijnsel. Thesis Agricultural University, Wageningen.

Dam, H. van & H. Kooyman-van Blokland, 1978. Man-made changes in some Dutch moorland pools, as reflected by historical and recent data about diatoms and macrophytes. Int. Revue ges. Hydrobiol. 63: 587-607.

Dresscher, Th.G.N'., 1976. Index van de namen en vindplaatsen die betrekking hebben op in Nederlandse wateren aangetroffen algen en enige groepen van micro-organismen. North-Holland Publishing Company, Amsterdam. -79-

Forsberg, C, S.-O. Ryding, A. Claesson & A. Forsberg, 1978. Water chemical analyses and/or algal assay? - Sewage effluent and polluted lake water studies. Mitt, internat. Verein. Limnol. 21: 352-363.

Golterman, H.L. (ed.), 1976. Fosfaten in het Nederlandse oppervlaktewater. Rapport van de "Stuurgroep Fosfaten" van de K.N.C.V. Sigma Chemie.

Higler, L.W.G., 1979. Limnological data on a Dutch moorland pool trough sixty years. Hydrobiol. Bull. 13: 138-143.

Hofstede, A.E. & E. Blok, 1964. Visserij-biologische waarnemingen in de Langeraarse plassen. Documentatierapport nr.4. Inspectie der Visserijen, Utrecht.

Hoogheemraadschap van Rijnland, 1984. Rapport betreffende het onderzoek naar de effecten van fosfaat verwijdering op de a.w.z.i.'s Gouda, Bodegraven en Nieuwveen. Rapport technische dienst van Rijnland, Leiden.

IKSR, 1987. Aktionsprogramma "Rhein", ausgearbeitet von der Internationa len Kommission zum Schutze des Rheins gegen Verunreinigung (ISKR), Strassburg.

Klapwijk, S.P., 1981. Limnological research on the effects of phosphate removal in Rijnland. H20 14: 472-483 (in Dutch with an English summary). Klapwijk, S.P., 1982. Hydrobiologisch onderzoek naar de uitwerking van het waterkwa liteitsklassensysteem van Caspers en Karbe voor grotere wateren in Zuid-Holland. Rapport Hoogheemraadschap van Rijnland, Tech nische Dienst, Leiden.

Klapwijk, S.P., 1987. Setting ecological standards for fresh waters. Landschap 4: 28-40 (in Dutch with an English summary).

Klapwijk, S.P., 1988. Biological assessment of the water quality in South-Holland (The Netherlands). Int. Revue ges. Hydrobiol. (in press).

Klapwijk, S.P. & C.J. Smit, 1988. Gouda en de waterkwaliteit van Rijnland. In: Ludy Giebels (ed.): Waterbeweging rond Gouda van ca. 1100 tot heden: Geschiedenis van Rijnlands Waterstaat tussen IJssel en Gouwe. Hoogheemraad schap van Rijnland, Leiden.

Klapwijk, S.P., L. van den Hove & P. Nieuwpoort, 1988. Een vergelijking tussen historische en recente gegevens van hy- drochemie en fytoplankton in het gebied van Rijnland. In: R.M.M. Roijackers (ed.): Hydrobiologisch onderzoek in Nederland. Fun damentele en toegepaste aspekten. Publ. no. 6 Hydrobiological Society, Amsterdam: 93-103. -80-

Lauterborn, R., 1918. Die geographische und biologische Gliederung des Rheinstroms. Teil III. Sitzungsberichte der Heidelberger Akademie der Wissen schaften, Math.-naturwiss. Klasse Abt. B, 1918: 1-87.

Leentvaar, P., 1956. Voorjaarsplankton van enige Nederlandse meren. De Levende Natuur 59: 132-137.

Leentvaar, P., 1963. Resultaten van het hydrobiologisch onderzoek van oppervlakte water in 1960. Water 47: 203-207.

Mauch, E., 1976. Leitformen der Saprobitat für die biologische Gewasseranalyse. Cour. Forsch. Inst. Senckenberg, 21. Frankfurt am Main.

Ministry for Transport and Public Works, 1985. Water Action Programme: 1985-1989. The Netherlands. Moss, B., 1979. Algal and other fossil evidence for major changes in Strumpshaw broad, Norfolk, England in the last two centuries. Br. phycol. J. 14: 263-283.

Osborne, P.L. & B. Moss, 1977. Paleolimnology and trends in the phosphorus and iron budgets of an old man-made lake, Barton Broad, Norfolk. Freshwat. Biol. 7: 213-233.

Pantle, E. & H. Buck, 1955. Die biologische Überwachung der Gewasser und die Darstellung der Ergebnisse. Gas- Wasserfach 96: 604.

Peelen, R., 1975. Changes in the composition of the plankton of the rivers Rhine and Meuse in the Netherlands during the last fifty-five years. Verh. internat. Verein. Limnol. 19: 1997-2009.

Redeke, H.C., 1923. Rapport over onderzoekingen aangaande den groei van den Bra sem in verschillende wateren. Verhandelingen en Rapporten Rijks instituten voor Visscherij-onderzoek 1: 221-254.

Romijn, G., 1924. Hydrobiologische toestand van Rijnland. Verslagen en medede lingen betreffende de Volksgezondheid, no. 2: 107-132.

Scheygrond, A., 1933. De Reeuwijksche en Sluipwijksche plassen. 3rd ed. A. Brinkman & Zoon, Gouda.

Schmidt-van Dorp, A.D., 1978. Eutrophication of shallow lakes in Rijnland. Report Technical Service, Hoogheemraadschap van Rijnland, Leiden (in Dutch with an English summary). -81-

Schuurmans, R.R. , 1970. Hydrobiologisch onderzoek aan de Kagerplassen. Doctoraal ver slag Zoölogisch Laboratorium, afdeling Milieubiologie, Rijksuni versiteit Leiden / Rijksinstituut voor Natuurbeheer (R.I.N.), Zeist.

Sladecek, V., 1973. System of water quality from the biological point of view. Ergebn. Limnol. 7: 1-218.

Sokal, R.R. & F.J. Rohlf, 1969. Biometry. The principles and practice of statistics in biological research. W.H. Freeman & Co, San Francisco.

Vries, P.J.R. de & S.P. Klapwijk, 1987. Bioassays using Stigeoclonium tenue Kütz. and Scenedesmus quadricauda (Turp.) Bréb. as testorganisms; a comparative study. Hydrobiologia 153: 149-157.

Werf f, A. van der, 1943. Het biologisch onderzoek van het boezemwater van Rijnland van begin juni 1941 tot eind september 1942 met 15 bijlagen. In: P. de Gruijter & E.L. Molt. Rijnlands Boezem, deel III. De hoeda nigheid van het boezemwater. Hoogheemraadschap van Rijnland, Leiden.

Wit, J.A.W. de, 1980. Aspects of the waterquality in the IJsselmeer area. H20 13: 251- 256 (in Dutch with an English summary). -83-

CHAPTER 5: DOSE-EFFECT RELATIONSHIPS BETWEEN PHOSPHORUS CONCENTRA TION AND PHYTOPLANKTON BIOMASS IN THE REEUWIJK LAKES (THE NETHERLANDS)

J. C. van der Vlugt & S. P. Klapwijk.

Published in: Verh. internat. Verein. Limnol. 23: 482-488 (1987).

"In dit isolement der plassen schuilt hun beteekenis als object voor hydrobiologisch onderzoek. Immers hier wordt de jaarlijksché periodi citeit der levensverschijnselen van de in het water voorkomende orga nismen niet of nagenoeg niet verstoord door den invloed van naburige waterbewegingen en kan de natuurlijke opeenvolging van de biocoenosen in de loop der seizoenen het best worden bestudeerd. De opbloei der Diatomeeën en gekleurde Flagellaten in het vroege voorjaar, het daarmee samenhangende optreden van Copepoden en Cladoceeren, het overheerschen van Rotatoren in zomer en najaar, al deze bekende verschijnselen kunnen in het plankton der Reeuwijksche en Sluipwijk- sche plassen buitengemeen fraai worden waargenomen".

A. Scheygrond, 1933. De Reeuwijksche en Sluipwijksche plassen, 3d ed. A. Brinkman & Zoon, Gouda, p. 27. -84-

Verh. Internat. Wretn. Limnol. Stuttgart, Januar 1988

Dose-effect relationships between phosphorus concentration and phytoplankton biomass in the Reeuwijk Lakes (The Netherlands)

J. C. VAN DER VLUGT and Sj. P. KLAPWIJK

With 5 figures and 1 table in the text

Introduction The Reeuwijk Lakes area consists of a series of smaller and larger lakes situated at 52 ° 2' North latitude and 4 " 45' East longitude in the Western part of The Netherlands, near Gouda. The lakes originate from peat-digging in the 16th and 17th century and have a total surface of 7 km2 and an average depth of about 2 m. The lakes are all interconnected except for Lake Broek- velden-Vettenbroek, which is completely isolated from the other lakes. The shallow and peaty Reeuwijk Lakes have become highly eutrophic during the last decades due to the inlet of very nutri ent rich water from the adjacent Breevaart canal (Fig. 1). This canal receives its water from the river Rhine and passes the village of Reeuwijk that loads the canal with secondary treated waste water. During the last decades the lakes have become less suitable for conservational and recreational purposes like surfing, swimming and sailing due to algal blooms that have decreased the transpar ency to less than 0.5 m. Therefore, it is attempted to restore the lake water quality by reducing the phosphorus load (cf. VAN DER DOES & KLAPWIJK 1986). This paper presents a description of the water quality of the Reeuwijk Lakes in the years 1983—1985, prior to the introduction of P removal (April 1986) at the sewage treatment plant of the village of Reeuwijk. The aim of this study was to collect basic information on the water quality in the lakes, which can be compared with changes likely to occur when the external P load is reduced.

Material and methods At two stations in the lake system water samples were collected weekly during 1983—1985. The stations are situated in Lake Elfhoeven (ROP 13408) and in Lake Nieuwenbroek (ROP 13409)

.uauot ."ƒƒ// / *Kl£w N&

Fig. 1. The Reeuwijk Lakes area with Lake Elfhoeven and Lake Nieuwenbroek, the x>. )h>'-J~l&*L^~. Breevaart canal and the sewage treatment plant of the village of Reeuwijk. -85-

J. C. van der Vlugt et al., Phosphorus and phytoplankton biomass

with respectively a surface of 1.1 km2 and 1.3 km2 and mean hydraulic residence times of about two months and 1.5 years. Samples were taken with a 0.5 m plexiglass tube sampler of 31 from three depths (0—0.5 m, 0.5—1.0 m, 1.0—1.5 m) and mixed to a composite sample from which subsamples for the chemical and biological analyses were taken. Analysis of P and N compounds was per formed by standard methods on a Technicon AAII autoanalyzer. For the estimation of the zooplankton biomass and species distribution a modified UTERMÖHL microscopical method (VAN HEUSDEN 1972) was used. Chlorophyll-a was measured spectrophotometrically after extraction with 80% ethanol (75 °C). Seston and particulate organic matter were determined gravimetrically as dry weight (75 °C) and ash free dry weight (600 °C). A SECcm-disc was used for measuring transparency. The values presented here are monthly arithmetic means of weekly observations.

Results and discussion

Table 1 gives the average values of the water quality parameters measured over three years. It is clear that there is a considerable amount of total phosphorus in both lakes consisting of about 75 % particulate P and less than 15 % inorganic P. The seasonal variations of the most relevant parameters are presented in Fig. 2. In both lakes the lowest concentrations of total phosphorus are observed in the first quaner of the year. At that time there is little difference between the P levels in both lakes. In Elfhoeven, the lake which is directly connected with the Breevaart canal, the total phos phorus concentration reaches a level of 200/ig/l whereas in Lake Nieuwenbroek at some distance of the Breevaart canal a level of 100/ig/l is observed (Fig. 2). This is in accordance with the differences in P-load between both lakes, ■ respectively 1.56gP • m"2 • year"1 in Elfhoeven and of 0.5gP • m~2 ■ year"1 in Nieuwenbroek. However, chlorophyll concentrations are inversely coupled with the P concentrations (Table 1). Therefore P and chlorophyll concentrations in these two lakes do not show a similar ratio. The chlorophyll/phosphorus ratio in Lake Nieuwenbroek is 1.4, whereas the ratio in Lake Elfhoeven is as low as 0.6. Since chlorophyll is an often criticized biomass parameter it is better to discuss it in combination with other biomass parameters, like ash free dry weight (AFDW). Regres sion analysis of AFDW against chlorophyll shows a significant correlation between these

Table 1. Yearly averages and three years averages of some water quality parameters in the Reeuwijk Lakes in 1983-1985. Lake Elfhoeven Lake Nieuwenbroek 1983 1984 1985 '83-'85 1983 1984 1985 '83-'85 PÜ4-P O**) 26 16 11 18 14 10 6 10 Particulate-P (Kg/1) 90 80 110 100 60 60 60 60 Total P OVO 140 120 140 130 80 80 80 80 NH4-N (mg/l) 0.06 0.05 0.07 0.06 0.13 0.13 0.23 0.16 NO3-N (mg/1) 0.10 0.09 •0.02 0.07 0.08 0.07 0.05 0.07 Particulate N (mg/l) 0.81 0.87 0.82 0.84 1.49 1.51 1.39 1.46 Total N (mg/1) 1.96 2.05 1.89 1.96 2.83 2.82 2.76 2.81 Chlorophyll (Kg/1) 77 87 77 80 120 116 96 111 Dry weight (mg/l) 20.3 17.5 17.3 18.4 29.8 24.6 23.2 26.0 Ash free DW (mg/l) 12.9 10.9 10.3 11.3 25.8 21.4 18.9 22.2 Zooplankton (mg/l) 5.6 7.4 2.9 5.3 1.2 1.7 1.8 1.6 Temperature (°Q 12.2 11.8 12.2 12.0 12.5 11.7 12.3 12.1 Secchi-disc (cm) 61 60 54 58 43 41 45 43 transparency -86-

?w> — ELFHOEVEN -o> a --- NIEUWENBROEK 200 z> o 1M> CxL O 100 x 50

* 30_ I 20. u. " ML o IO- IA IO lo

Fig. 2. The SECCHI-CÜSC transparency and the concentrations of total phosphorus, chlorophyll-a, ash free dry weight and zooplankton in the Reeuwijk Lakes over the years 1983—1985. two biomass parameters (Fig. 3). Both in intercepts and slopes considerable differences appear between the two lakes. Especially in Nieuwenbroek there is obviously an amount of AFDW that is not immediately connected with phytoplankton. Another point of in terest in this respect is the carbon, estimated as half of the AFDW concentration, to chlo- -87-

J. C. van der Vlugt et al., Phosphorus and phytoplankton biomass

50.0 ELFHOEVEN 50.0 NIEUWENBROEK 1983-1985 1983-1985

40.0 40.0

3O0 B3O0 / 5 20.0 £ 2O0 * . * ' " / ft 10.0 * . • !^-* 100 y = 5.52 + 0.07X y 11.87- 0.09X r = 0.68 — r 0.54-" 0.0 _ao lo l50 Il00 1150 I200 I250 I300 lO l50 Il00 !l50 !200 '250 1300 CHLOROPHYLL-a (pg/l) CHLOHOPHYU-a (pgï) Fig. 3. Relationship between chlorophyll and AFDW in the Reeuwijk Lakes over the years 1983-1985.

150 125 O H 100 <

y 50 — ELFHOEVEN o 25 — NIEUWENBROEK

J IA IJ lo U IA IJ lo IJ IA IJ lo 1983 1984 198S Fig. 4. The C/Chlorophyll ratio in the Reeuwijk Lakes over the years 1983—1985.

rophyll ratio in the course of the year (Fig. 4). The seasonal fluctuation is about the same in both lakes, with the lowest values in winter and very high values up to 150 in summer. The yearly average C/Chl ratios of 70 in Elfhóeven and 100 in Nieuwenbroek are in the range of C/Chl ratios in other Dutch lakes like the Loosdrecht Lakes (VAN LIBRE et al. 1986), but are much higher than values in the literature (e.g. JORGENSEN 1979). From the regression analyses of AFDW against chlorophyll and from the seasonal fluctuations in the C/Chl ratio, it can be concluded that the difference in phytoplankton biomass be tween the two lakes is even greater than based on the chlorophyll concentrations alone -88-

3.0 ELFHOEVEN NIEUWENBROEK' 1983-1985 1983-1985

\ 20 _20_ >- z UJ cc ^" ■' ' 1Q_ ft

3.0 ELFHOEVEN 1983 19 v

E 2 0 o> - UzJ cc ft< 1.0 z

Fig. 5. Relationship between both chlorophyll and AFDW and the Seccm-disc transparency in the Reeuwijk Lakes over the years 1983—1985.

(Fig. 2). So the inverse relationship between phytoplankton biomass and P concentration in the two lakes is greater than stated before. The question, however, why there is more phytoplankton in Nieuwenbroek than in Elfhoeven is still unanswered. A possible explanation may be provided by another differ ence between the two sampled lakes, namely the concentration of herbivorous zooplank ton. The zooplankton biomass in Elfhoeven is three times the biomass in Nieuwenbroek (Table 1, Fig. 2). It is obvious that such a great difference must have a major influence on the development of the phytoplankton biomass and seston concentration in the two lakes. Moreover, the phytoplankton in Lake Nieuwenbroek consists mainly of fil amentous bluegreen algae, which are not readily eaten by zooplankton species, while in Lake Elfhoeven the phytoplankton population consists also of chlorococcal green algae and diatoms. Therefore a plausible explanation is that grazing controls the algal growth in Elfhoeven. There are several other data supporting this assumption (e.g. VAN DER VLUGT 1976, POSTMA 1980, BANNINK et al. 1980, GULATI 1984, GULATI et al. 1985, LAMPERT & TAYLOR 1985). The next question is why there is not enough zooplankton in Nieuwen broek for controlling algal biomass. Restoring an ecosystem after a forty-year-period of P overloading can presumably not simply be done by lowering the P load alone. Supplementary measures like restoring the balance in the fish population are necessary. At present the fish population in the Reeu- -89-

J. C. van der Vlugt et al., Phosphorus and phytoplankton biomass wijk Lakes is dominated by bream resulting in a low zooplankton biomass in Lake Nieu wenbroek. This is probably due to the lack of hiding places for fish of prey like pike or perch pike. Because of the minor transparency of the water (Table 1), the submerged macrophyte vegetation has disappeared. It is not easy to predict if an amelioration of the undèr-water light climate in the next years can be expected after the reduction in P-loading (started in April, 1986). For that reason regression analyses are made of the reciprocal SECCHi-disc transparency against both chlorophyll and AFDW (Fig. 5). Both in inter cepts and slopes considerable differences between the two lakes can be observed. It may be expected that a decreasing growth of algae in Elfhoeven improves the transparency more effectively than in Nieuwenbroek. Anyhow it will take much time to reach the Dutch water quality objective of 1 m SEC- CHi-disc transparency for swimming water. Possibly a rough remedy is necessary to solve the problem of turbidity much quicker. Perhaps the in-lake addition of a precipitant, like iron salts (FeS04 or FeCl3), can result in a rapid increase in transparency and thus in the growing of submerged macrophytes and in the restoration of the whole ecosystem. At this point it deserves attention that the in-lake treatment with iron salts in the "Grote Rug" reservoirs as a long-term measure has proven to be an effective means to the inactivation of P (BANNINK & VAN DER VLUGT 1978 a, b, BANNINK et al. 1980, VAN DER VLUGT & ALDENBERG 1982). Iron removed phosphorus and paniculate matter from the water column in an effective way, which resulted in lower chlorophyll levels and a higher transparency.

Acknowledgements This study is part of an extensive limnological research program on the effects of phosphorus re moval on the water quality of the Reeuwijk Lakes, carried out in cooperation by the Technical Service of the Waterboard of Rijnland and the National Institute of Public Health and Environ mental Hygiene. This program is supported by the Directorate General of Environmental Hygiene of the Dutch Ministry of Housing, Physical Planning and Environment.

References

BANNINK, B. A. & VLUGT, J. C. VAN DER, 1978 a: Erfahrungen mit dem Zusatz von Fallmitteln im „Grote Rug" Speicherbecken und vergleichende Untersuchungen in 3 Modell-Reservoiren nach LUND. - DVGW-Schnftenreihe Wasser 16: 216-243. _ _ 1978 b: Hydrobiological and chemical response to the addition of iron and aluminium salts, studied in three LuND-type butylrubber reservoirs. — Verb. Internat. Verein. Limnol. 20: 1816-1821. BANNINK, B. A., MEULEN, J. H. M. VAN DER, PEETERS, J. C. H. & VLUGT, J. C. VAN DER, 1980: Hydro- biological consequences of the addition of phosphate precipitants to inlet water of lakes. — ' Hydrobwl. Bull. 14 (1/2): 73-89. DOES, J. VAN DER & KLAPWIJK, SJ. P., 1987: Effects of phosphorus removal on the maximal algal growth in bioassay experiments with water from four Dutch lakes. — Int. Rev. ges. Hydro bwl. 72 (1): 27-39. GULATI, R. D., 1984: The zooplankton and its grazing measures of trophy in the Loosdrecht Lakes. — Verb. Internat. Verein. Limnol. 22: 863—867. GULATI, R. D., SIEWERTSEN, K. & POSTEMA, G., 1985: Zooplankton structure and grazing activities in relation'to food quality and concentration in Dutch Lakes. - Arch. Hydrobtol. Suppl. 21: 91-102. HEUSDEN, G. P. H. VAN, 1972: Estimation of the biomass of plankton. - Hydrobwlogia 39: 165-208. -90-

JBRGENSEN, S. E. (ED.), 1979: Handbook of environmental data and ecological parameters. — Int. Soc. ■ Ecol. Modell. Copenhagen. LAMPERT, W. & TAYLOR, B. E., 1984: In situ grazing rates and partiele selection by zooplankton: Ef fects of vertical migration. — Verb. Internat. Verein. Limnol. 22: 943—946. LIERE, L. VAN, BALLEGOOIJEN, L. VAN, KLOET, W. A. DE, SIEWERTSEN, K., KOUWENHOVEN, P. & ALDEN- 8ERG, T., 1986: Primarv production in the various parts of Loosdrecht Lakes. — Hydrobiol. Bull. 20 (1/2): 77-85. ' POSTMA, L., 1980: Modelling as a tool. - Hydrobiol. Bull. 14 (1/2): 55-63. VLUGT, J. C. VAN DER, 1976: Comparative limnological research in the "Grote Rug" and model re servoirs. — Hydrobiol. Bull. 10: 136—144. VLUGT, J. C. VAN DER & ALDUNBERG, T., 1982: Blooms of algae and their control by phosphorous

precipitation/inactivation. — (In Dutch with English summary). H20, 15 (4): 73—79.

Authors' addresses:

Drs. J. C. VAN DER VLUGT, Rijksinstituut voor Volksgezondheid en Milieuhygiëne, Postbus 1, NL-3720 BA Bilthoven, The Netherlands. Drs. Sj. P. KLAPWIJK, Hoogheemraadschap van Rijnland, Postbus 156, NL-2300 AD Leiden, The Netherlands.

Foto pag. 91: Opgedroogd sediment. PARTB: SEDIMENTS

fo i

CHAPTER 6:

INTRODUCTION SEDIMENTS

"Before radical and expensive management measures are taken for the reduction of eutrophication, the contribution of the internal nutrient source to the sustainment of algal growth should be considered." L. Lijklema, 1980. Eutrophication; the role of sediments. Hydro- biol. Bull. 14: p. 98. -94-

INTRODUCTION SEDIMENTS

Sediments play an important role in the cycle of elements such as phosphorus and nitrogen in fresh waters (Mortimer, 1941, 1942; Hutchinson, 1957; National Academy of Sciences, 1969; Golterman, 1975). They serve not only as a sink for elements but also as a po tential source for nutrients, which can be released from the sediments (Golterman, 1966; van Kessel, 1976; Golterman (ed.), 1977; Sly, 1986) Therefore, before management measures are taken for the reduction of eutrophication, the contribution of the internal nutrient source to the sustainment of algal growth should be considered (Lijklema, 1980). Several ways to estimate the phosphorus release from sediments are applied in limnology:

1. The phosphate exchange technique: The desorption of adsorp tion of phosphate is measured in water overlying an undisturbed sediment core (cf. van Liere & Mur, 1982; Boers et al. , 1984; van Raaphorst & Brinkman, 1984; Boers, 1986; Brinkman & van Raaphorst, 1986; van Raaphorst et al., 1987). With this techni que theoretical fluxes can be measured in the laboratory of phos phate release from the sediments, ignoring every interaction by worms (Tubifex), fishes etc.

2. The chemical extraction technique: Sediment samples are extrac ted by different chemical solutions, and the amount of extracted phosphate is used to estimate potential phosphorus release from the sediments (cf. Williams et al., 1971; Hieltjes & Lijklema, 1980; Golterman, 1982; Pettersson, 1986).

3. The model approach: The release of phosphorus from sediments is directly measured in physical models, such as enclosures (cf. Kou we & Golterman, 1976). Sometimes the results are used to construct mathematical models by which the phosphate release in other lakes is predicted (cf. de Rooij, 1980a,b; Brinkman & van Raaphorst, 1986; van Eek & Smits, 1986; Brinkman et al., 1987a,b).

4. The bioassay technique: Algae are grown in a medium with sedi ment as the sole source of phosphate, and all other nutrients in excess, to determine the amount of phosphorus available to algae (cf. Golterman et al., 1969; Golterman, 1977; Grobler & Davies, 1979, 1981; Williams et al., 1980; Klapwijk et al., 1982; Bruning & Klapwijk, 1984; Klapwijk & Bruning, 1986).

The last method has a special advantage, since it can directly estimate the availability of sediment phosphate to algae. All other methods measure extracted or released phosphate, which does not have to be entirely available for algal growth. See also Hegemann et al. (1983) for a critical review and analysis of different methods to determine algal-available phosphorus. All these techniques, however, ignore the activity of animals, local movement of water bodies, etc. In order to see whether several methods, which claim to predict the amount of phosphorus available for algal growth, give the same results, we compared two extraction techniques, viz. the NTA column method (Golterman, 1982) and the stepwise NH4Cl-NaOH-HCl shaking -95- method (Hieltjes & Lijklema, 1980), with the bioassay method using the testalga Scenedesmus quadricauda (Chapter 7). Since the coun ting of algal cells in such bioassays is very laborious, especially when the heterogeneity in sediment composition in a lake requires a rather great number of sampling stations and therefore bioassays, we devel oped a new biomass parameter in bioassays which are disturbed by sediment, applying derivative spectroscopy (Chapter 8). With the aid of this quick, sensitive and reliable technique the sediment composi tion of eight lakes in the Rijnland Waterboard area has been investi gated to estimate the release of phosphorus from the sediments and to search for relations between the available P and P-binding compo nents in the sediments, such as Fe, CaCÖ3, clay, and organic matter (Chapter 9).

REFERENCES Boers, P.CM., 1986. Studying the phosphorus release from the Loosdrecht lakes se diments, using a continuous flow system. Hydrobiol. Bull. 20: 51-60.

Boers, P.CM., J.W.Th. Bongers, A.G. Wisselo & Th.E. Cappenberg, 1984. Loosdrecht Lakes Restoration Project: Sediment phosphorus dis tribution and release from the sediments. Verh. internat. Verein. . Limnol. 22: 842-847.

Brinkman, A.G. & W. van Raaphorst, 1986. De fosfaathuishouding in het Veluwemeer. Ph.D. Thesis, Tech nical University of Twente.

Brinkman, A.G., W. van Raaphorst, L. Lijklema & G. van Straten, 1987a. Mathematical description of sediment-water phoshorus exchange processes. H20 20: 658-663 (in Dutch with an English summary). Brinkman, A.G. , W. van Raaphorst, L. Lijklema & G. van Straten, 1987b. Experimental techniques for investigating sediment-water ex change processes. H20 20: 664-668 (in Dutch with an English summary).

Bruning, C. & S.P. Klapwijk, 1984. Application of derivative spectroscopy in bioassays estimating algal available phosphate in lake sediments. Verh. internat. Ver ein. Limnol. 22: 172-178.

Eck, G.Th.M. van & J.G.C. Smits, 1986. Calculation of nutrient fluxes across the sediment-water interface in shallow lakes. In: P.G. Sly (ed.): Sediments and water inter actions, pp. 289-301. Springer Verlag, New York. -96-

Golterman, H.L. , 1966. Influence of the mud on the chemistry of water in relation to productivity. Proc. I. B .P.-symposium, Amsterdam-Nieuwersluis, October 1966, pp. 297-313. Kon. Ned. Akad. Wetenschappen, Amsterdam.

Golterman, H.L., 1975. Physiological Limnology. Elsevier Scientific Publishing Company, Amsterdam.

Golterman, H.L. (ed.), 1977. Interactions between sediments and fresh water. Junk, The Hague & Pudoc, Wageningen.

Golterman, H.L., 1977. Sediments as a source of phosphate for algal growth. In: Golter man, H.L. (ed.). Interactions between sediments and fresh water Proceedings of an international symposium held at Amsterdam, the Netherlands, September 1976. Junk, The Hague & Pudoc, Wageningen.

Golterman, H.L., 1982. Differential extraction of sediment phosphates with NTA solution. Hydrobiologia 92: 683-687.

Golterman, H.L., Bakels, C.C. & Jacobs-Mbgelin, J., 1969. Availability of mud phosphates for the growth of algae. Verh. internat. Verein. Limnol. 19, 39-58.

Grobler, D.C. & E. Davies, 1979. The availability of sediment phosphate to algae. Water SA 5: 114- 122.

Grobler, D.C. & E. Davies, 1981. Sediments as a source of phosphate: a study of 38 impoundments. Water SA 7: 54-60.

Hegemann, D.A., Johnson, A.H. and Keenan, J.D., 1983. Determination of algal-available phosphorus on soil and sediment: A review and analysis. J. Environ. Qual. 12: 12-16.

Hieltjes, A.H.M. & L. Lrjklema, 1980. Fractionation of inorganic phosphates in calcareous sediments. J. Environ. Qual. 9: 405-407.

Hutchinson, G.E. , 1957. A treatise on limnology I. Geography, physics and chemistry. John Wiley, New York.

Kessel, J.F. van, 1976. Influence of denitrification in aquatic sediments on the nitrogen content of natural waters. Ph.D. Thesis, Agricultural University of Wageningen. Centre for Agriculture Publishing and Documen tation , Wageningen. -97-

Klapwijk, S.P., J.M.W. Kroon & M-L. Meijer, 1982. Available phosphorus in lake sediments in the Netherlands. Hy- drobiologia 92: 491-500.

Klapwijk, S.P. & C. Bruning, 1986. Available phosphorus in the sediments of eight lakes in the Netherlands. In: P.G. Sly (ed.): Sediments and water interac tions, pp. 391-398. Springer Verlag, New York.

Kouwe, F.A. & H.L. Golterman, 1976. The role of sediment phosphates in the eutrophication process. H20 9: 84-86 (in Dutch). Liere, L. van & L.R. Mur, 1982. The influence of simulated ground-water movement on the phos phorus release from sediments, as measured in a continuous flow system. Hydrobiologia 92: 511-518.

Lijklema, L., 1980. Eutrophication; the role of sediments. Hydrobiol. Bull. 14: 98-105. Mortimer, C.H., 1941. The exchange of dissolved substances between mud and water in lakes. I. J. Ecol. 29: 280-329.

Mortimer, C.H., 1942. The exchange of dissolved substance between mud and water in lakes. II. J. Ecol. 30: 147-201.

National Academy of Sciences, 1969. Eutrophication: causes, consequences, correctives. Washington, D.C.

Pettersson, K., 1986. The fractional composition of phosphorus in lake sediments of different characteristics. In: P.G. Sly (ed.): Sediments and water interactions, pp. 149-155. Springer Verlag, New York.

Raaphorst, W. van & A.G. Brinkman, 1984. The calculation of transport coefficients of phosphate and calci um fluxes across the sediment-water interface, from experiments with undisturbed sediment cores. Water Sci. Techn. 17: 941-951. Raaphorst, W. van, A.G. Brinkman, L. Lijklema & G. van Straten, 1987. The internal phosphorus load in Lake Veluwe. The Netherlands. H20 20: 669-674 (in Dutch with an English summary). Rooij, N.M. de, 1980a. A chemical model to describe nutrient dynamics in lakes. In: J. Barica and L.R. Mur (eds.). Hypertrophic Ecosystems. Devel opments in Hydrobiology 2: 139-149. Junk, The Hague-Boston- London . -98-

Rooij, N.M. de, 1980b. Application of a chemical model to combat eutrophication. Hydro- biol, Buil. 14: 106-115.

Sly, P.G. (ed.), 1986. Sediments and water interactions. Springer Verlag, New York. Williams, J.D.H., J.K. Syers, R.F. Harris, & D.E. Armstrong, 1971. Fractionation of inorganic phosphate in calcareous lake sediments. Soil Science American Proc. 35: 250-255.

Williams, J.D.H., H. Shear & R.L. Thomas, 1980. Availability to Scenedesmus quadricauda of different forms of phosphorus in sedimentary materials from the Great Lakes. Lim- nol. Oceanogr. 25: 1-11. -99-

CHAPTER 7: AVAILABLE PHOSPHORUS IN LAKE SEDIMENTS IN THE NETHERLANDS S.P. Klapwijk, J.M.W. Kroon & M-L. Meijer.

Published in: Hydrobiologia 92: 491-500 (1982).

"In algal cultures the sediments proved to be an excellent source of phosphate, if sediments and algal cells are mixed without a physical barrier between the two." H.L. Golterman, 1977. Sediments as a source of phosphate for algal growth. In: Golterman, H.L. (ed.): Interactions between sediments and fresh water, p. 286. Junk, The Hague & Pudoc, Wageningen. -100-

Available phosphorus in lake sediments in The Netherlands

S. P. Klapwijk, J. M. W. Kroon & M-L. Meijer Hoogheemraadschap van Rijnland, Breesiraal 59, 2311 CJ Leiden. The Netherlands

Keywords: phosphate, sediment, bioassay, algae, phosphate fractionation, NTA extraction

Abstract

The amount of phosphorus available to algae in the sediments of four lakes in the western part of the Netherlands has been assessed by means of chemical extraction and bioassay techniques. In addition to direct chemical sediment analyses, extractions were carried out with a NTA column method and a stepwise NH4 Cl-NaOH-HCl shaking method, the latter supposedly separating the weakly bound, the Fe- and Al-bound and the Ca-bound phosphates in the sediments. Bioassays, with sediment as the sole source of P, were made with Scenedesmus quadricauda in modified Skulberg's Z8 medium to determine the amount of phosphates available to algae. The average total P concentration of the sediments varied from 0.8 to 3.6 mg P g"' dry wt and correlated well with the net external P loading of the lakes. Uptake of P by algae in the bioassays varied from 0.4 to 36% while NTA extracted 36-69% of the total P. The ratio NH4C1 extracted/ NaOH extracted/ HC1 extracted phosphates is different from lake to lake, although in all lakes the highest extractions (27-62% of total P) are found in the NaOH fraction. However, in the peaty sediments of these lakes, the NaOH step extracted not only the Fe- and Al-bound phosphates but, also, large amounts of humus compounds. Hence, this fraction also contains non-available organic P. The results are related to soil type and chemical characteristics of the sediments, and compared with data from other authors. A positive correlation was found between phosphate available to algae and NTA- and NaOH-extractable P, but the correlation with total phosphorus was higher. Moreover, algal-extractable P proved to be positively correlated with total iron and clay content and negatively with the amount of organic matter. It is concluded that the sediments in the investigated lakes show great variability and that the chemical extraction techniques cannot replace the bioassays to assess the amount of phosphorus available to algae.

Introduction nr2 y~' remained in the sediments. In the area of the Rijnland Water Authority, situated on the delta of Due to the growing population and the increas the Rhine (Fig. 1), the values are even higher: A ing nutrient load of the river Rhine, eutrophication gross P-load of 14-15 g P m~2 y_l is estimated, of in the Netherlands has progressed so far in the last which between 3 and 7 g P nr2 y"1 deposits in the decades that most of the inland lakes should now be area, depending on climatic and hydrological con considered hypertrophic (Leentvaar 1980). ditions (Schmidt-van Dorp 1978; Klapwijk 1981). In 1970 the gross P-load of the surface waters in Large differences in P-loading exist between the the Netherlands was estimated at an average of 6 g lakes within the area. About 10% of the P-load Prrt 2 y1 (Golterman 1976, 1977), of which 4 g P originates from the Rhine, while 70% comes from

Hydrobii>logia92.49l 500(191(2). 0018 8158/82,0922 0491,$02.00. c, l)r W. Junk Publishers. The Hague. Printed in The Netherlands. -101-

HAARLEM

rWESTEINXR LAKES

rBRAASSEM J) LAKE

NIEUWKOOP LAKES

THE HAGUE'

■* REEUWIJK LAKES GOUDA

Okm s 10 km .

TTTT"r BORDER OF THE RIJNLAND WATER AUTHORITY HOLLANDSE USSEL

Fig. I. The area of the Rijnland Water Authority, its location in the Netherlands (inset), and the location of the sampled lakes. -102-

purified and unpurified wastewater. wijk Lakes situated in polders at about 2 m below Asa lake restoration experiment, P-removal was main sea level, are separated from the boezem but implemented in 1979 at three sewage treatment are also dependent on it for their water supply. plants, in the south-east of the Rijnland area. The Lake characteristics, including morphometry, effects on the lakes are being followed by limnolog- hydrology and phosphate loading, are given in Ta ical studies, which include plankton analysis and ble 1. The top layer of the sediments from a number bioassays. of sampling stations in the lakes was collected with It is expected that the effects of the reduced P- a Lenz bottom grab (Schwoerbel 1972). After sepa load will be delayed by the release of phosphates ration into an upper (0-5 cm) and a lower (5-15 cm) from the sediments (see also Golterman 1977). Gol- layer, the samples were transported to the laborato terman el al. (1969) proved that algae can use sedi ry in glass jars and stored at 4 "C. Dry weight was ment phosphate for their growth. Therefore it estimated by drying duplicate samples at 103 °C. seemed necessary to investigate how much phos The remaining analyses and the bioassays and phate is accumulated in the sediments and how chemical extractions were carried out after par much of it is potentially available for algal growth. tial dewatering of the samples by squeezing (30-40 In this paper, the first results of an investigation min; up to 600 kPa), or by slight drying (30-40 °C; in four lakes, using bioassay and chemical extrac 48 h). tion techniques, are presented. This research pro Loss on ignition (LOI) was determined after gram will be continued and extended to the remain heating to 600 °C. Particle size (<2 jum, <16 jum, ing lakes in Rijnland. <50 /im) was determined using the pipette method. Percentage CaC03 was measured volumetrically after shaking with hydrochloric acid. Anion ex Sampling area, materials and methods change capacity (AEC) was assessed by treating the sediment with ammonium phosphate and subse The Rijnland Water Authority encompasses a quent percolation with ammonium chloride, and densely populated area (about 10' inhabitants in measuring the P content in the eluents. Chemical I 000 km2) situated between the cities of Amster oxygen demand (COD) was measured by titration dam, Haarlem, The Hague and Gouda (Fig. 1). with ferrous ammonium sulphate, after oxidizing In the 'boezem' of Rijnland, an interconnected with potassium dichromate. Kjeldahl-nitrogen(Kj- system of canals and lakes and ditches, the water N) was measured after digestion with sulphuric acid level is kept constant at 0.6 m below mean sea level. + Wieninger selenium mixture and total phospho Water can be pumped in near Gouda, from the rus (t-P) was measured after digestion with sulphur Hollandse IJssel, a branch of the Rhine, while ex ic acid + persulphate. Subsequently, determina cess water can be pumped out to the sea. The boe tions of N Ht and ortho-P were made on a Techni- zem supplies a great number of inlying polders with con AA II autoanalyser. Total iron (Fe), total water. Some lakes, such as the Braassem Lake and calcium (Ca) and total aluminium (Al) were deter the Westeinder Lakes, form part of this boezem mined by atomic absorption spectroscopy on a system. Others, such as the Nieuwkoop and Reeu- Perkin-Elmer S400 spectrophotometer. Particle

Tahle I. General characteristics of morphometry. hydrology and isphate loading of Braassem Lake, Westeinder Lakes. Reeuwijk Lakes and Nieuwkoop Lakes.

Lake Surface Mean depth General Situation Residence Gross P-load Net P-load (X 10* w2) (m) sediment oflake time (gPm^y"1) (gPm"2y"') composition (y)

Braassem 4.6 4.0 Clay Boezem 0.13 20.6 3.7 Westeinder 9.3 2.7 Sand/Peal Boezem 4.8 0.9 0.3 Reeuwijk 8.2 1.9 Peat Polder 8.2 1.0 0.5 Nieuwkoop 10.9 1.5 Peat Polder 7.2 0.7 0.5 Table 2. Physical and chemical analyses of the sediment at the sampling stations in the investigated lakes.

Sampling station Dry weight <2 am

size, % CaC03 and AEC were measured according were strongly coloured by humus compounds, or- to Al and Holland (1977), and all other analyses tho-P could not be determined by the available were made according to the standard, methods of methods. Therefore, in these cases, calculation of the Nederlands Normalisatie Instituut. the extracted P was based on total P analyses, Bioassays with sediment as sole phosphate source departing from the method described by Hieltjes & were carried out in duplicate using Scenedesmus Lijklema (1980). The amount of P in the HC1 ex quadricauda (Turp.) de Brébisson as test-organism, tract was sometimes measured as total P (Braassem and a modified Skulberg's Z8 50% medium (see Lake and Reeuwijk Lakes) because ortho-P could Bolier el al. 1981) with an equimolar amount of KC1 not be measured, for unknown reasons. instead of K2HP04. A 1-week-old culture of starved Scenedesmus quadricauda, on 10% Z8 with 5% P, was used to avoid cells which could have taken up Results luxury-P. The cultures were incubated on a shaking table (100 rpm) in 150 ml Jena erlenmeyers with In Table 2 the physical and chemical analyses of 100 ml medium. About 200 mg of partly dried sed the sediment at 16 sampling stations and at two 6 iment and 6.10 cells were used as inoculum under depths in the investigated lakes are summarized. the following conditions: 20 ± 1 °C; light: 70 jiE ' From this table it can be seen that the clay content 2 m~ s~'; day/night regime: 12/12 h. Growth of the (% <2 ftm) in Braassem Lake is much higher than in algae was followed by counting the cells in approx the other lakes. The highest values for CaC0 were 3 3 imately 1 mm , in a Bürker-Turk counting cham observed in the Westeinder Lakes. LOI, COD and ber. After reaching the maximal yield, 1 mg P1"' as Kj-N are relatively low in Braassem Lake, but high K2HP04-P was added to one of the duplicate cul in the other lakes, especially in the Nieuwkoop tures in order to test whether the maximum growth Lakes. The clay content decreases and the amount was limited by phosphate. Available phosphate was of organic matter increases in the following order: calculated by comparing the maximum yield with a Braassem Lake, Westeinder Lakes, Reeuwijk La- (linear) standard, established from the maximum yields in bioassays with S. quadricauda cultured in

the same medium and under the same conditions, REEUWIJK LAKES l2W03:0-Scml but with known amounts of K2HPO„ (0-30-50-100- torn. 250-500-1000 Mg P 1"') instead of sediment. Extraction with nitrilotriacetic acid (NTA) was performed essentially according to Golterman (1977): 1-1.5 g partly dried sediment was suspended in 50 ml 0.01 M neutralized NTA solution in a co lumn (2.5 cm diameter). After settling of the sedi ment, 450 ml NTA solution was added and perco lated at a rate of 100 ml h~'. Total P was determined in the extract. Selective fractionation of phosphates was per formed according to Hieltjes & Lijklema (1980), using the following stepwise extraction scheme: Loosely bound + CaC03 adsorbed phosphates, iron + aluminium bound phosphates, calcium bound phosphates were selectively extracted re spectively by NH4C1, NaOH and HC1. Extraction was carried out on a shaking table (120 rpm) at a temperature of 20 ± 1 °C. After each extraction step the solution was filtered through a 0.45 fim ■Fig. 2. Growth of Scenedesmus quadricauda (means ± s.d.)ina bioassay with sediment from Reeuwijk Lakes as only source of P-free membrane filter, the NaOH step being pre phosphate. On the 23th day I mg P l"1 is added to one of the ceded by centrifugation. As the NaOH extracts duplicate cultures (arrow). -105-

Table 3. Algal-available and chemical-extractable phosphates determined by different methods. All values expressed as percentage of total sediment phosphorus. Values for bioassay calculated by comparing with standard; values for NH4C1 extraction based on ortho-P, NTA and NaOH extractions based on total-P, H Cl extraction of Braassem and Reeuwijk sediments based on total-P, of Westeinder en Nieuwkoop sediments on ortho-P determinations. See text.

Sampling station Bioassay NTA extraction Selective extraction with

NH„CI NaOH HC1 Braasem Lake 271 (0- 5 cm) 31 42 21 83 16 (5-15 cm) 24 41 8 58 10 273 (0- 5 cm) 41 46 12 50 17 (5-15 cm) 46 68 7 58 18 Mean ± s.d. 36 ± 10 49 ± 13 12 ± 6 62 ±14 15 ±4 Westeinder Lakes 279 (0- 5 cm) 0 51 40 50 10 280 (0- 5 cm) 5.5 71 40 67 9 (5-15 cm) 9.8 103 43 69 ■ - 28! (0- 5 cm) 0 58 30 45 23 282 (0- 5 cm) 4.7 60 32 51 4 (5-15 cm) 0 76 35 60 8 283 (0- 5 cm) 6.2 71 34 36 7 (5-15 cm) 3.4 50 36 45 4 284 (0- 5 cm) 0 80 58 47 - Mean ± s.d. 3.3 ± 3.6 69± 17 39 ± 8 52 ± II 9±6 Reeuwijk Lakes 119.14 (0- 5 cm) 16 23 13 37 47 (5-15 cm) 6 49 17 18 63 134.08 (0- 5 cm) 10 36 9 55 36 (5-15cm 9 33 5 17 42 218.03 (0- 5 cm) 13 12 4 22 20 (5-15 cm) 2 40 7. 10 25 Mean ± s.d. 9±5 36 ±13 9±5, . 27 ±5 39 ± 16 Nieuwkoop Lakes 94.09 (0- 5 cm) 0 54 14 48 4 (5-15 cm) 0 42 33 61 6 94.11 (0- 5 cm) 0 29 22 41 5 (5-15 cm) • 3.4 47 21 34 5 94.12 (0- 5 cm) ' 0 27 19 60 10 (5-15 cm) 0 36 28 48 9 94.19 (0- 5 cm) 0 28 14 53 9 (5-15 cm) 0.3 33 26 57 10 94.20 (0- 5 cm) 0 26 17 47 6 (5-15 cm) 0 41 28 40 II Mean± s.d. 0.4± 1.1 36 ±9 22 ±6 46 ±9 8±3 kes and Nieuwkoop Lakes. The total P content is of phosphate on the 23rd day clearly shows that 4-5 times higher, and the total Fe is about 2 times growth was limited by phosphate. higher in Braassem Lake than in the other three In Table 3 the results of the bioassays and the lakes. Most parameters show no clear difference chemical fractionation of P forms are summarized. between the top 0-5 cm layer and the lower 5-15 cm The highest percentage of P available to algae was layer; total-P and Kj-N, however, show slightly found in bioassays with Braassem sediment (about higher values in the top layer. 36%), and Reeuwijk sediment (about 9%). The frac IJI Fig. 2 a growth curve of Scenedesmus quadri- tions of available phosphate in bioassays with West cauda is shown for a bioassay with Reeuwijk sedi einder and Nieuwkoop sediment appeared to be ment as sole source of P. The increase after addition extremely low (3.3 and 0.4%, respectively). The -106-

Broassem Lake • Westeinder Lakes 2000 Reeuwijk Lakes A Nieuwkoop Lakes -

1500

woa

5QQ_

- A A

-*—! 1 r ~~I—i—l—|—r 1 I 1500 2000 NTA-P.figlg dr.wl hifi. J. The relationship between algal-extractable phosphate (algal-P) and NTA-extractable phosphate (NTA-P) in sediment samples of lour lakes. amount of phosphate extracted by NTA is much In Fig. 3 the amount of P extracted by NTA is higher in all lakes, ranging from 36% (Reeuwijk and plotted against the amount taken up by the algae in Nieuwkoop) to 69% (Westeinder) of the total P- bioassays. For Braassem sediments the relationship content. In the selective extractions, the NaOH step between algal-extractable and NTA-extractable P extracted the highest percentage from all sedi is fairly good, but the Braassem samples deviate ments. However, the ratio between the NH4C1, radically from the other lakes. Because of the lack NaOH and HCI fractions varied with each lake. of normality in the data-set, due to the values of the Neither the NTA extractable phosphate, nor the Braassem sediment, the computation of product- NaOH fraction (or the NH4CI +NaOH fraction) is moment correlations (Grobler & Davies 1979) is quantitatively comparable to the amount of algal- not possible with our sediment data. Therefore, we extractable phosphate. In all lakes the chemical computed Spearman's non-parametric rank-corre treatments extracted much more phosphate than lation coefficients (Sokal & Rohlf 1969) between did the algae. the different extraction methods and between algal- No clear difference was seen in phosphate availa extractable phosphate and some sediment proper bility, between the top layer and the underlying ties; these are summarized in Table 4. sediment, assessed by different methods. This table shows that both the NTA-extractable -107-

Table 4. Spearman's rank correlation coefficients (rs) between algal extractable phosphates (algal-P) and different chemical extractable phosphates and some sediment properties.

Chemical-extractable P by:

NTA NH4C1 NaOH NH„Cl+NaOH HCI Algal-P/. 0.45" 0.13 0.38* 0.36 0.14

Sediment properties:

total-P Fe

*P < 0.05; "P < 0.02; *"P < 0.01; *"*P < 0.001. In all cases n = 29, except for algal-P/ HCI extr.-P, in which n = 7. and the NaOH-extractable phosphates are positive experimental conditions may influence the P con ly correlated (a = 0.05) with algal-extractable tent per cell. phosphates. However, neither the NH4C1- and Our results on the percentage of available phos HCl-extractable phosphates nor the sum of NH4C1 phate, established by means of bioassays, are low. + NaOH-extractable phosphates show any signifi The results from Westeinder and Nieuwkoop Lakes cant correlation with algal-extractable P. The a- are especially low when compared to the data of mount of algal-extractable P is very positively Golterman (1969, 1977, 1980), Grobler & Davies correlated with total phosphorus, total Fe and clay (1979) and Williams el al. (1980). The difference content, but CaC03 is not correlated with algal-ex between methods for calculating the phosphate tractable P. A negative correlation is found be availability to algae may be a reason for this, as tween algal-extractable P and the parameters COD previously mentioned. and LOI, indicating organic matter. In the bioassays with sediment the pH sometimes rose to 11; especially in fast growing cultures. This might promote the extraction of Fe- and Al-phos- Discussion phates (Hieltjes 1980) and therefore increase the amount of P available to algae relative to the natu The average total P content in the four lakes ral situation. In the investigated lakes pH values up corresponds fairly well with their net P-load (Table to 9 are sometimes measured, due to the occurrence 1). From a comparison of the Pcontent of our lakes of algal blooms. with data from other authors, e.g. Williams et al. The % P extracted by NTA in our sediments is (1971, 1980), Grobler&Davies (1979) and Hieltjes generally much higher than the % NTA-extractable (1980), it appears that the P content of Braassem P recorded by Golterman (1977), Grobler & Davies sediment is rather high and that the P content iri (1979, 1981) and Williams el al. (1980). Moreover Westeinder, Nieuwkoop and Reeuwijk sediments is our data from four lakes suggest, just like the results relatively low. from Boström and Pettersson (1981), that the % In this study the amount of algal-extractable P extractable P varies from lake to lake, according to was calculated by comparing the maximal yield of a the type of sediment (see Table 3). sediment bioassay with a calibration standard (es Our findings are not in agreement with Golter tablished by the maximal yields in bioassays from man (1977), who found that the amount of phos known amounts of ortho-P, and cultured simul phate extracted with NTA, from sediments of sev taneously under the same conditions). We consider eral Dutch lakes, was similar in quantity to that this to be a better procedure than calculating the P extracted by algae such as Scenedesmus sp. Only in available to algae from experimentally determined the case of Braassem sediments did NTA extract P fixed coefficients, between algal biomass and P amounts to the same order of magnitude as that in content (Golterman 1977; Golterman el al. 1969; the bioassays. NTA extracted 4-90 times more P and Grobler & Davies 1979), because different than algae, in the sediments of our other lakes. Our -108-

results disagree even more strongly with those of used by algae, also fits in very well with our find Grobler & Davies (1979, 1981), using Selenastrum ings. On the other hand, little algal available P capricornutum, who found that algal available seemed to be adsorbed onto carbonates, in this phosphate in sediments in South African reservoirs study. was generally 4-5 times higher than the NTA-ex- Insufficient data were available to warrant any tractable P. From Fig. 3, it can be seen that the further statistical analysis. A continuation of this relationship between algal-extractable and NTA- research over the remaining lakes in Rijnland, with extractable P seems to depend on the type of sedi more sampling stations, is necessary to make better ment. For example, the ratio algal-P/NTA-P is statistical analyses. much lower in the Westeinder Lakes than in Braas- sem Lake. The NaOH fraction of the Hieltjes and Lijklema Conclusions fractionation procedure probably contained organ ic P, due to the extraction of brown humus com 1. The P content of the sediments in the four inves pounds and the necessity to measure total P instead tigated lakes in Rijnland varied from about 1 mg of ortho-P (previously stated). According to Gol- P g_l dry wt in Westeinder, Nieuwkoop and terman (1973) and Williams el al. (1980), organic P Reeuwijk Lakes, to about 4 mg P g~' dry wt in is not available to algae. The fractionation scheme Braassem Lake. It correlates very well with the of Hieltjes & Lijklema (1980) therefore required an net P-load of the lakes. adaptation for use in peaty sediments. Siebers el al. 2. The amount of algal extractable P in bioassays (1981), using Mössbauer spéctroscopy, also proved varied from 0.4% (Nieuwkoop Lakes) to 36% the lack of selectivity of an NaOH extraction for (Braassem Lake) of the total P content. iron-phosphates in a sludge sample. 3. Chemical extraction methods with NTA and In none of the four lakes is either the NaOH stepwise fractionation with HN4Cl-NaOH-HCl fraction or the NH4C1 + NaOH fraction in the same extracted more phosphorus, and probably other order of magnitude as the amount of phosphate forms, than algae. Therefore, these methods extracted in bioassays, as Hieltjes (1980) found with cannot be used to predict the amount of P avail sediment from 'De Grote Rug' reservoir. In our able to algal growth in our lakes. experiments NaOH extracted 2-10 times more 4. The percentage extractable P (both algal and phosphorus than algae from most sediments. This chemical) varied from lake to lake, probably ratio was least in the Braassem sediments, and cor depending on the type of sediment. responded well with their relative low peat content. 5. Although a positive correlation has been estab Although calculated correlations between algal- lished between algal extractable P vs. NTA-P extractable P and NTA-P, and NaOH-P, are signif and NaOH-P, the correlation between algal-P icant, the correlation coefficients are not very high and total P is better. The NTA and NaOH ex and in both cases much lower than the correlation traction methods did not show a specific selectiv between algal-P and total P (Table 4). The advan ity for algal-available phosphorus. tages of applying an NTA extraction (or a selective 6. The chemical extraction method modified by extraction according to Hieltjes & Lijklema 1980) Hieltjes & Lijklema (1980) cannot be applied for to these sediment samples are not clear, because of use with peaty sediments without revision. the uncertain predictions of phosphorus available 7. Algal-extractable P was positively correlated to algae. Even a normal total P determination with total P, total Fe.and clay content, and nega showed better correlation with algal-extractable P. tively with the amount of organic matter. Our findings with respect tó the nature of algal- extractable phosphates are in. fairly good agree ment with the literature. Hieltjes (1980) stated that Acknowledgements most of the algal extractable P is associated with iron or is adsorbed onto clay, and this is supported The authors are greatly indebted to Ir. M. A. by the data from Table 4. The statement of Goiter- Heinsdijk, Prof. Dr. M. Donze and Prof. Dr. man (1973) that organic phosphates are not easily L. Lijklema for valuable suggestions with respect to -109-

the manuscript, to Drs. J. P. Al and Mr. A. M. B. Grobler, D, C. & Davies, E., 1979. The availability of sediment Holland for practical advice and technical assist phosphate to algae. Wat. S.Afr. 5: 114-122. ance, to Mr. P. van Doorn for drawing the figures, Grobler, D. C. & Davies, E., 1981. Sediments as a source of phosphate: a study of 38 impoundments. Wat. S.Afr. 7: to Paul Gutteridge for correcting and Miss C. van 54-60. Dijk for typing the English text and to the Board of Hieltjes, A. H. M., 1980. Eigenschappen en gedrag van fosfaat in the Rijnland Water Authority for permission to sedimenten. Thesis, Twente University of Technology. 302 publish these preliminary results. pp. (in Dutch with an English summary). Hieltjes, A. H. M. & Lijklema, L., 1980. Fractionation of inor ganic phosphates in calcareous sediments. J. envir. Qual. 9: 405-407. Klapwijk, S. P., 1981. Limnological research on the effects of

References phosphate removal in Rijnland. H20 14: 472-483 (in Dutch with an English summary). Al, J. P. & Holland, A. M. B., 1977. Geochemische bemonster- Leentvaar, P., 1980. Comparison of hypertrophy on a seasonal ings- en analysemethodieken. Rijkswaterstaat, Deltadienst, scale in Dutch inland waters. In: Barica, J. & Mur, L. R., hoofdafd. Milieu en Inrichting. Nota 76-60 (in Dutch). (Eds.) Hypertrophic Ecosystems. Developments in Hydro- Bolier, G., van Breemen, A. N. & Visser, G., 1981. Eutrophica- biology 2: 45-55. Junk, The Hague. tion tests: a possibility to estimate the influence of sewage Nederlands Normalisatie Instituut. Tests methods for waste wa discharge on the biological quality of the receiving water. ter. NEN 3235.

H20 14; 88-92 (in Dutch with an English summary). Schmidt-van Dorp, A. D., 1978. Eutrophication of shallow Boström, B. & Pettersson, K., 1982. Different patterns of phos lakes in Rijnland. Report Technical Service, Hoogheem phorus release from Jake sediments in laboratory batch raadschap van Rijnland, 254 pp. (in Dutch with an English experiments. (These proceedings). summary). Golterman, H. L., Bakels C. C. & Jakobs-MÓgelin, J., 1969. Schwoerbel, J., 1972. Methods of Hydrobiology, 2nd edn. Per- Availability of mud phosphates for the growth of algae. gamon Press, Oxford. 200 pp. Verh. int. Verein. Limnol. 17: 467-479. Siebers, H. H., van der Kraan, A. M. & Donze, M.. 1982. The Golterman, H. L., 1973. Natural phosphate sources in relation fractionation of iron-phosphorus compounds in sediments to phosphate budgets. A contribution to the understanding studied by Mössbauer spectroscopy. (These proceedings). of eutrophication. Wat. Res. 7: 3-17. Sokal, R. R. & Rohlf, F. J., 1969. Biometry - the Principles and Golterman, H. L., (Ed.) 1976. Fosfaten in het Nederlandse op Practice of Statistics in Biological Research. W. H. Freeman, pervlaktewater. Rapport van de Stuurgroep Fosfaten der San Francisco, 776 pp. KNCV. 133 pp. Sigma Chemie (in Dutch). Williams, J. D. H., Syers, J. K., Shukla, S. S.. Harris, R. F. & Golterman, H. L., 1977. Sediments as a source of phosphate for Armstrong, D. E., 1971. Levels of inorganic and total phos algal growth. In: Golterman, H. L. (Ed.). Interactions be phorus in lake sediments as related to other sediment pa tween Sediments and Fresh Water. Junk, The Hague & rameters. Envir. Sci. Technol. 5: 1113-1120. Pudoc, Wageningen. Williams, J. D. H., Shear H.&Thomas, R. L., 1980. Availability Golterman, H. L., 1980. Studies on phosphate release from lake to Scenedesmus quadricauda of different forms of phospho bottoms in artificial enclosures in a shallow eutrophic lake. rus in sedimentary materials from the Great Lakes. Limnol. HjO 13: 513-515 (in Dutch with an English summary). Oceanogr. 25: 1-11. -111-

CHAPTER 8: APPLICATION OF DERIVATIVE SPECTROSCOPY IN BIOASSAYS ESTIMATING ALGAL AVAILABLE PHOSPHATE IN LAKE SEDIMENTS

C. Bruning & S.P. Klapwijk.

Published in: Verh. internat. Verein. Limnol. 22: 172-178 (1984).

"Derivative spectroscopy is a simple yet powerful technique for mag nifying the fine structure of spectral curves." J.E. Cahill, 1979. Derivative spectroscopy:' understanding its ap plication. Amer. Lab. 11, p. 79. -112-

Verh. Internat. Verein. Limnol. 22 172-178 Stuttgart, Juli 1984

Application of derivative spectroscopy in bioassays estimating algal available phosphate in lake sediments

C. BRUNING and S. P. KLAPWIJK

With 4 figures and 1 table in the text

Introduction Bioassays have been applied by several authors to assess the amount of sediment phosphate available to algae, e.g. GOLTERMAN et al. (1969), CHIOU & BOYD (1974), FITZGERALD & UTTORMARK (1974), GOLTERMAN (1977), SMITH et al. (1978), GROBLER & DAVIES (1979, 1981), HIELTJES (1980), WILLIAMS et al. (1980), KLAPWIJK et al. (1982). In this technique the maximum algal biomass growing on a medium with sediment as sole phosphate source is used as a measure of the available sediment-P. However, the method used to follow the algal growth in those experiments varies widely: GOLTERMAN et al. (1969), CHIOU & BOYD (1974), GOLTERMAN (1977), SMITH et al. (1978), WILLIAMS et al. (1980), and KLAPWIJK et al. (1982) followed the growth by counting algal cells, HIELTJES (1980) by chlorophyll-* measurements, GROBLER & DAVIES (1979, 1981) by the WALKLEY-BLACK method for organic carbon determination, and FITZGERALD & UTTORMARK (1974) by optical density meas urements at 750 nm. All these methods have certain disadvantages. Counting algal cells is very laborious, especially when the heterogeneity in sediment composition in a lake requires a rather great number of sampling stations and therefore bioassays. Chlorophyll-* measurements are also time-consuming and require large culture volumes. Organic carbon determinations are inappropriate in bioassays with organic soils, such as peat, and the normal spectrophotometric optical density measurement can possibly be interfered by the sediment addition. We used derivative spectroscopy to measure algal growth in bioassays troubled with sediment. Since the technique is unfamiliar in limnology and, as far as we know, has never been applied to measure algal growth in turbid samples, in this paper the principle of the method is elucidated and the sensibility and reliability of the method is tested and compared to the normal chloropyll-a ex traction procedure. The application of the technique is demonstrated in bioassays with sediments from a lake in the western part of the Netherlands and the results are discussed, especially with respect to pH.

Derivative spectroscopy In our bioassays growth of the algae was measured by derivative spectroscopy on a Phi- lips/Pye Unicam SP 8800 spectrophotometer. The principle of this technique consists of calculat ing the first (1st), second (2nd) (or even higher) order derivative of a spectrum with respect to wa velength and plotting this derivative rather than the spectrum itself. A comprehensive description of the technique and a review of the relevant literature is given by CAHILL (1979) and COTTRELL (1980). To illustrate the possibilities Fig. 1 shows the normal absorption spectra and 1st derivative spectra between 600 and 750 nm of a Scenedesmus culture, a suspension of sediment and a mixture of a Scenedesmus culture + sediment suspension. Fig. 1A represents the normal absorption spectrum of a Scenedesmus culture. The peak near 680 nm is typical for chlorophyll-^, the height of which can be used as a biomass parameter. -113-

0.20r

0.15-

z 0.1 Oh 3 ce ° 0.05-

i cc Fig. 1. Absorption spectra and 1st derivative < uu spectra of a Scenedesmus culture, a sediment su Q spension and a mixture of both. A, C, E: normal absorption spectra of respectively Scenedesmus 650 750 650 750 650 750 culture, sediment suspension and mixture. B, D, F: corresponding 1st derivative spectra. WAVELENGTH (nm)

Fig. 1B shows the 1st derivative spectrum of the same Scenedesmus culture. In this the 1st de rivative of the spectrum in Fig. 1A is plotted as a function of the wavelength. Therefore, the value of the 1st derivative at a certain point is a measure of the slope of the absorption spectrum at that specific point. It is positive as the normal spectrum rises and negative as it declines. At the curve's maximum and minimum it goes through zero. With increasing algal concentra tions the slopes before and after the peak near 680 nm become steeper. As a consequence the maxi mum in the 1st derivative spectrum moves upwards and the minimum moves down. Therefore, the distance between maximum and minimum in the 1st derivative spectrum can also be used as a measurement of the algal concentration. The advantage of derivative spectroscopy becomes evident from Fig. 1C and 1D, showing re spectively the absorption spectrum and the 1st derivative spectrum of the sediment suspension. The absorption spectrum of the sediment suspension (Fig. 1C) descends with almost a con stant slope, due to the fact that absorption is especially caused by scattering and scattering declines with inclining wavelength. The 1st derivative spectrum (Fig. 1D) therefore is negative and almost constant. Finally, Fig. 1E and 1F represent normal absorption and 1st derivative spectra of a mixture of the Scenedesmus culture + the sediment suspension in the same concentrations as in Fig. 1 A—1D. The absorption peak round about 680 nm lies higher than in the pure Scenedesmus culture due to absorption (scattering) by the sediment and its height cannot be used as a biomass parameter. The 1st derivative spectrum lies lower than that of the Scenedesmus culture due to the negative slope of the sediment absorption spectrum. But the distance between maximum and minimum is almost unchanged, since the derivative spectrum of the sediment suspension has neither minimum nor maximum. Therefore, this distance — in this paper called "1st derivative signal" — can also be used'as a biomass parameter in an algae/sediment mixture. We measured this derivative signal in ' cm, obtained with a scan speed of 10 nm • s~', a band width of 2 nm and a response time of 10 s.

Testing the derivative spectroscopy technique

Linearity and detection level of the technique were tested by measuring the 1st de rivative signal in a dilution series of a Scenedesmus culture (Fig. 2 open circles). The fig ure shows a good linear relationship (represented by the solid line) between the deriva tive signal and the cell concentration in the range of 2 x 104—106 cells • ml"1. Even at -114-

1000 s 1ST DERIVATIVE SIGNAL WITHOUT SEDIMENT • 1ST " WITH ■ 2ND

100

i/i 10

Fig. 2. Relation between cell concentration and deriva tive signal in Scenedesmus cultures with and without 011 J, u_ sediment (the bars are 95% confidence limits to means s 10' 10 10' (A i • •> SCENEDESMUS CONCENTRATION (CELLS/ml) 0t * ODServatlons;.

low cell concentrations of 3 x 103 cells • ml-1 a fairly reliable estimation of the biomass seems possible. The highest sediment/algae ratio which leads to inaccuracies due to interference by the sediment, is determined by measuring the 1st derivative signal of a Scenedesmus dilu tion series, where to each dilution the same amount of sediment is added. The results (means of 4 observations) are shown in Fig. 2 (closed circles) with the 95 % confidence limits (as bars). The maximum biomass in our bioassays lay between 105 and 2 x 106 cells • ml"1. In this range the deviation from the solid line (for the sediment-free samples) is neglectible and the confidence limits are small. Only at concentrations below 3 x 104 cells • ml-1 the presence of the sediment introduces considerable acci dental and systematic errors. From this can be calculated, that in algae/sediment mix tures, where only 6 % of the total absorption at 680 nm is due to the presence of algae a reliable algal biomass measurement is still possible with the derivative technique. In our bioassays the share of the Scenedesmus absorption in the total absorption at 680 nm was at least 10%. Such low values appeared only in bioassays with strong (by humic substances) coloured sediments, which were poor in phosphates. In most bioassays the contribution of algae to the total absorption was much higher. By measuring a derivative signal of higher (2nd, 3rd, 4th) derivative order a reduction in the chance of systematic errors by interference of the sediment, but an increase of ac cidental errors due to an unfavourable signal/noise ratio can be expected on theoretical grounds. See CAHILL (1979). The squares in Fig. 2 show the 2nd derivative signal of a Scenedesmus dilution series with added sediment. Though the means lie fairly well on a straight line, the 95 % confidence limits prove that the lower signal/noise ratio leads to increasingly inaccurate estimates of the algal biomass in the relevant range. Therefore, the 1st derivative signal is considered a better biomass parameter than the 2nd and for that reason the higher derivative orders are left out of account. In the usual algal biomass determination methods (see e. g. Standard Methods 1980, GOLTERMAN & CLYMO 1971, and Nederlands Normalisatie Instituut) chlorophyll-d is extracted in ethanol or acetone and separated from other cell components by filtration to eliminate interference caused by scattering. Since derivative spectroscopy is not very -115-

C. Bruning & S. P. Klapwijk, Derivative spectroscopy

uuo r -. 0.0006» -00017 t' = 0.9951 p 0.001 006 - / / 004 / /

002 Fig. 3. Relation between measurements of chlorophyll-a in extracted cultures and of the 1st de / / 1 i rivative signal in living cultures of Scenedesmus 25 50 75 100 (points represent means of 3 observations). 1ST DERIVATIVE SIGNAL IN LIVING CULTURE (cm)

sensitive to scattering, it seemed interesting to compare the derivative technique meas uring chlorophyll-*? in living cells to the normal procedure of measuring extracted chlo rophyll-^. Fig. 3 shows the results of parallel determinations (in triplo) of the algal biomass by both techniques from 8 bioassay samples during their stationary growth phase. The results of both techniques proved to be positively correlated. Therefore, from these experiments can be concluded that the 1st derivative signal at approximately 680 nm is a reliable biomass parameter, at least for algae growing under comparable conditions.

Application of the technique

Bioassays with sediment as sole phosphate source and Scenedesmus quadricauda (TURP.) DE BRÉBISSON as test organism were carried out with a total of 55 sediment samples from 8 shallow lakes in the western part of the Netherlands. A full report on the applied methods and the obtained results will be published elsewhere (KLAPWIJK & BRUNING, in prep.). High sediment doses make the cultures turbid, both by the sediment itself and by the dense algal population growing on the available phosphate. This can lead to elevated internal chlorophyll-** concentrations in the algae, due to light interference (GOLTER- MAN 1975). Moreover, in dense populations the pH can rise to pH > 10, causing extra desorption of iron- and aluminium-bound phosphates. Since such high pH values do not occur in the lakes, their appearance in bioassays should be anticipated. Therefore, the amount of sediment dosed in the bioassays was chosen in such a way that the cul tures contained 100—150/tg • l_l P. If during an experiment the optical density at 680 nm rose above 0.240 cm"1 or the pH rose above 8.6, the bioassay was repeated with less than half of the sediment dose. To show the suitability of the derivative technique, Fig. 4 presents the growth curves of 2 bioassays (in duplo) with different additions of sediment from sampling station Ml in lake Mooie Nel. From this figure can be concluded that the derivative technique produces fairly regular growth curves. The differences between the duplo's are rather small and presumably not influenced by the growth measurement, but mainly caused -116-

1000

! ***' Ti+P % 100

£ ■r*~tJ. T +P

10 Fig. 4. Growth curves of 2 bioassays (in duplicate) with sediment from sampling station Ml in lake Mooie Nel. A: bioassays with high sediment addition 10 20 30 40 and high maximum pH (9.6). B: bioassays with low DAYS sediment addition and low maximum pH (8.1).

Table 1. Available phosphate and corresponding highest pH values in bioassays with different se diment additions from 4 sampling stations in lake Mooie Nel (values are means of duplicate experi ments). Low sediment addition High sediment addition Sampling Available P Highest PH Available P Highest pH station (% of total P) (% of total P) Ml 33.9 8.1 81.4 9.6 M2 35.3 8.0 72.4 9.9 M3 34.4 8.1 84.9 10.3 M6 45.3 8.2 76.2 9.5 Lake average 37.2 78.7

by small differences in sediment addition. Adding 0.25 mg KH2PO4 (arrows in figure) always resulted in extra growth, proving that the maximum growth was limited by phosphate. From the growth curves the amount of available phosphate can be calculated by means of the constant internal P-concentration of the Scenedesmus cells (see KLAPWIJK & BRUNING, in prep.). Table 1 summarizes the calculated availability of sediment phosphates and the corre sponding pH levels of double (with 2 sediment doses and different pH levels) executed bioassays with sediment from 4 sampling stations in lake Mooie Nel. As can be seen from Table 1 the availability of the sediment phosphates is more than doubled at higher pH levels and increased from on average 37 % to 79 % in this lake. This pH effect has also been established in other lakes.

Discussion and conclusions

The derivative spectroscopy technique proved to be a very quick, sensitive and re liable biomass parameter in bioassays which are disturbed by sediment. It can even be used in sediment/algae mixtures, where only 6 % of the total absorption (at 680 nm) is -117-

C. Bruning & S. P. Klapwijk, Derivative spectroscopy

caused by the presence of algae. Therefore, the technique can possibly also be applied to determine the presence of algae in natural sediments (cf. KAPPERS 1976, 1977). Though the method is well known in chemical science (see CAHILL 1979 and COT- TRELL 1980), its application in water chemistry and other limnological areas is not very common. It can, for instance, also be used to analyse phosphate in turbid samples (HARMSEN 1984), which can be extremely useful in sediment extraction procedures (cf. KLAPWIJK et al. 1982). We believe, therefore, that derivative spectroscopy offers many possibilities for different limnological fields and we hope that its application will be en couraged. This study has also demonstrated that the rising pH in bioassays with sediment can greatly increase the availability of sediment phosphates. This is in conformity with AN DERSON (1975) who showed the influence of pH on the release of phosphorus from Dan ish lake sediments in exchange experiments. Clearly, in our bioassays, the potentially occurring precipitation of calcite with coprecipitation of phosphate did not balance the release of Fe- and Al-bound phosphates from the sediments at higher pH levels. The pH in bioassays with sediment should thus be controlled in order to get com parable and interpretable results (cf. HIELTJES 1980). However, in some, hypertrophic lakes in the Netherlands the pH of the surface waters can rise to above pH 9. It is beyond doubt, that particularly in such shallow lakes this can result in an enhanced re lease of phosphates from the sediments.

Acknowledgements The authors are greatly indebted to Drs. J. P. AL and Ir. M. A. HEINSDIJK for critical reading of the manuscript, to Mr. P. VAN DOORN for drawing the figures, to PAUL GUTTERIDGE for correcting and Miss C. VAN DIJK for typing the English text and to the Board of the Rijnland Water Author ity for permission to publish these preliminary results.

References ANDERSEN, J. M., 1975: Influence of pH on release of phosphorus from lake sediments. — Arch. Hydrobiol. 76 (4): 411-419. CAHILL, J. E., 1979: Derivative spectroscopy: understanding its application. — Amer. Lab. 11 (11): 79-85. CHIOU, C.-J. & BOYD, C. E., 1974: The utilization of phosphorus from muds by the phytoplank- ter, Scenedesmus dimorpbus, and the significance of these findings to the practice of pond fer tilization. — Hydrobiologia 45 (4): 345—355. COTTRELL, C. T., 1980: Derivative and log spectrophotometry. — Pye Unicam SP8 Series, Pan No. 7064.36.2887.21. FITZGERALD, G. P. & UTTORMARK, P. D., 1974: Applications of growth and sorption algal assays. — EPA report -660/3-73- 023, Washington. GOLTERMAN, H. L., 1975: Physiological limnology. An approach to the physiology of lake ecosystems. — Elsevier Sci. Publ. Co., Amsterdam. — 1977: Sediments as a source of phosphate for algal growth. — In: GOLTERMAN, H. L. (ed.), Interactions between Sediments and Fresh Water. Junk, The Hague & Pudoc, Wageningen. GOLTERMAN, H. L., BAKELS, C. C. & JAKOBS-MÖGELIN, J., 1969: Availability of mud phosphates for the growth of algae. — Verb. Internat. Verein. Limnol. 17: 467—479. GOLTERMAN, H. L. & CLYMO, R. S., 1971: Methods for chemical analysis of fresh waters. 3rd ed. — I. B. P. Handbook 8. Blackwell Sci. Publ., Oxford/Edinburgh. GROBLER, D. C. & DAVIES, E., 1979: The availability of sediment phosphate to algae. — Wat. S. Afr. 5:114-122. -118-

GROBLER, D. C. & DAVIES, E., 1981: Sediments as a source of phosphate: a study of 38 impound ments. - Wat. S. Afr. 7: 54-60. HARMSEN, J., 1984: Analysis of phosphate in turbid aqueous samples by derivative spectrophotometry. — Analytica Chimica Acta (in press). HIELTJES, A. H. M., 1980: Eigenschappen en gedrag van fosfaat in sedimenten. — Thesis, Twente Univ. of Technology, 302 pp. (in Dutch with English summary). KAPPERS, F. I., 1976: Blue-green algae in the sediment of the lake Brielse Meer. — Hydrobiol. Bull. 10: 164-171. — 1977: Presence of blue-green algae in sediments of lake Brielle. — In: GOLTERMAN, H. L. (ed.), Interactions between sediments and fresh water. Junk, The Hague & Pudoc, Wageningen. KLAPWIJK, S. P., KROON, J. M. W. & MEIJER, M.-L., 1982: Available phosphorus in lake sediments in the Netherlands. — Hydrobiologia 92: 491—500. Nederlands Normalisatie Instituut: Spectrophotometric determination of chlorophyll-a content. NEN 6520 (in Dutch). SMITH, E. A., MAYFIELD, C. I. & WONG, P. T. S., 1978: Naturally occurring apatite as a source of orthophosphate for growth of bacteria and algae. — Microb. Ecol. 4: 105—117. Standard Methods, 1980: Standard methods for the examination of water and wastewater. 15th ed. — Amer. Public Health Assoc, Washington. WILLIAMS, J. D. H., SHEAR, H. 8C THOMAS, R. L., 1980: Availability to Scenedesmus quadricauda of different forms of phosphorus in sedimentary materials from the Great Lakes. — Limnol. Oceanogr. 25: 1—11.

Authors' address:

Waterboard of Rijnland, Breestraat 59, 2311 CJ Leiden, The Netherlands -119-

CHAPTER 9: AVAILABLE PHOSPHORUS IN THE SEDIMENTS OF EIGHT LAKES IN THE NETHERLANDS

S.P. Klapwijk & C. Bruning. Published in: P.G. Sly (ed.): Sediments and water interactions, pp. 391-398. Springer Verlag, New York (1986).

"Mud is also a source of phosphates. Muds from several different ori gins when added as the only source of phosphates to a Rodhe culture solution gave excellent algal growth. Of the muds which were used, the only one which was unable to support algal growth was a pure "unpolluted" clay Another source of unavailable phosphate is peat."

H.L. Golterman, 1975. Physiological Limnology. An approach to the Physiology of Lake Ecosystems, pp. 95-96. Elsevier, Amster dam. -J.20-

Available Phosphorus in the Sediments of Eight Lakes in the Netherlands

S.P. Klapwijk and C. Bruning

The sediment composition of eight lakes in the Netherlands has been investigated to estimate the release of phosphate (P) from sediments. The amount of phosphate available to algae from sediments has been assessed by means of bioassays, using derivative spectroscopy to measure algal growth. The average total P concentration of the top 5-cm sediment layer varied from 1.1 -5.7 mg g '' dw. Based on the bioassays, available P varied between 10-45% of total P. The uptake by algae increased strongly with rising pH. Among the P-binding components in the sediment (Fe, CaCO,. clay, and organic matter), only organic matter was strongly negatively correlated with the available P, indicat ing that a high organic matter content tends to diminish the availability of the sediment P. The amount of available P was highly positively correlated with the ortho P concentration in the interstitial water and in the water immediately above the sediment. This provided a quick estimate of the immediately available sediment P. The results will be used to estimate the effect of P released from sediments and applied to lake restoration plans.

Introduction construct mathematical models by which the P release in other lakes is predicted; for example, de The amount of phosphate (P) that is released from Rooij (1980). lake sediments both before and after restoration is 4. The bioassay technique: Algae are grown in usually unknown. Measurements that allow a a medium with sediment as the sole source of P, prediction of these amounts are urgently needed. and all other nutrients in excess, to determine the Several ways of estimating the P release from amount of phosphorus available to algae; for exam sediments have been tried. ple, Golterman et al. (1969), Golterman (1977), Grobler and Davies (1979, 1981), Williams et al. 1. The P exchange technique: The desorption or (1980). Klapwijk et al. (1982), and Hegemann adsorption of P is measured in water overlying an etal. (1983). undisturbed sediment core; for example, van Liere and Mur (1982). In our opinion, the last method has a special 2. The chemical extraction technique: Sediment advantage, since it can directly estimate the availa samples'are extracted by different chemical solu bility óf sediment P to algae. There is a further tions, and the amount of P is used to estimate poten benefit to this method, since the growth of algae in tial phosphorus release from the sediment; for turbid samples can be measured easily and example, Williams et al. (1971}, Hieltjes and Lij- accurately by means of the derivative spectroscopy klema (1980), Golterman (1982), and Hegemann et technique (Bruning and Klapwijk, 1984). al. (1983). We applied this technique to sediments from eight 3. The model approach: The release of phospho lakes in the Rijnland Water Authority that have been rus from sediments is directly measured in physical influenced for decades by P-rich water from the models, such as enclosures; for example, Kouwe and Rhine River and from discharges of treated and Golterman (1976). Sometimes the results are used to untreated wastewater in the area. -121-

Materials and Methods using a Lenz bottom grab (Schwoerbel, 1972). The samples were transported to the laboratory in glass In the area of the Rijnland Water Authority, situated jars and stored at 4°C. No special precautions were in the delta of the Rhine River between the cities taken to prevent oxygen from reaching the muds. A of Amsterdam, Haarlem, Gouda, and The Hague, Jenkin sampler (Ohnstad and Jones, 1982) was used the sediments of eight shallow lakes were inves to sample the water immediately above the sedi tigated (Fig. 1). Some of the lakes form part ment, and the sample tubes with undisturbed sedi of the "boezem" system of the Rijnland, which ment and water were transported in a vertical is an interconnected system of canals, lakes, and position to the laboratory. The following day, a ditches with the same water level (0.6 m below mean water layer of 13 cm in the tubes was carefully sea level). The others are situated in polders (about siphoned off a few cm above the sediment. 2 m below mean sea level) that are separated from The interstitial water was obtained by squeezing the boezem, but are dependent on it for their water (30-40 min at 600 kPa) 2-cm thick layers of sedi supply. Some characteristics of the lakes are given in ment under oxic conditions through a 0.45-/*m Table 1. membrane filter (A1 and Holland, 1977). Dry weight Four to nine samples of the usually oxic top layer was estimated by drying duplicate samples to 105°C. (0-5 cm) of sediments were taken in each lake by Loss on ignition (LOI) was determined after heating

BORDER OF THE RIJNLAND WATER AUTHOR! Fig. .1. The location of the sam pled lakes. -122-

Available Phosphorus in Lake Sediments

Table 1. General characteristics of the eight investigated lakes.

Mean General Gross Surface area depth sediment Situation Residence Pload 6 2 Lake (10 m ) ■ (m) composition of lake time (mo) (g P m~! yr~') Braassem 4.61 3.5 Clay/silt Boezem 2.2 13.0 Kaag 3.23 2.8 Clay/silt Boezem 0.6 37.0 Westeinder 8.90 2.9 ' Sand/peat Boezem 4.6 4.5 Nieuwe Meer 1.35 5.0 Sand/silt Boezem 4.1 12.0 Mooie Nel 0.87 4.5 Sand/silt Boezem 1.2 54.0 Reeuw ijk 7.61 1.9 Peat Polder 24.0 0.9 Nieuwkoop 10.90 1.5 Peat Polder Unknown Unknown Langeraar 1.74 2.0 Peat Polder 18.0 0.9 these dried samples to 600°C, and was used as an The sediment samples were homogenized in an indication of the organic matter content. Particle ultraturrax apparatus to obtain a representative sub- size (< 2 /*m; > 50 ^m) was determined using the sample. To prevent the development of other algae pipette method. species, the sediment dose was suspended in 1 ml The CaC03 was measured volumetrically after aquadest and heated for 1 h at 75°C. shaking with hydrochloric acid. Total phosphorus Derivative spectroscopy was used to follow the (total P) was measured after digestion with sulfuric growth of the algae. This method is particularly- acid + persulfate and subsequent determination of appropriate for the measurement of algal growth in ortho P on a Technicon AA II autoanalyzer. Total turbid samples: it compensates for the loss of sensi iron (total Fc) was determined by atomic absorption tivity due to turbidity caused by the sediment. This spectrometry (AAS) on a Perkin-Elmer S400 spec- technique was extensively described by Bruning and trophometer. Particle size and percentage CaC03 Klapwijk (1984). We measured the derivative signal were measured according to Al and Holland (1977), in cm obtained at a scan speed of 10 nm/s, a band and all other analyses were measured according to width of 2nm, and a response time of 10s. After standard methods (NEN, 1969). reaching the maximum yield. 0.25 mg KH2P04 was Bioassays with sediment as the sole P source were added to each culture to test whether the maximum carried out in duplicate by using Scenedesmus quad- growth was limited by P. From the maximum yields. rirauda (Turp.) de Brébisson as the test organism available P was calculated using the internal P con and a modified Skulberg's Z8 50% medium (Bolier centration of the Scenedesmus cells. This was deter et al., 1981) with an equimolar amount of KC1 mined by measuring the maximum yields in bioassays with S. quadrkauda cultured in the same instead of KH2P04. A 4-6-week-old culture of starved S. quadrkauda, on Z8 medium with l/10th medium and under the same conditions, but with of the normal P concentration, was used as an inocu known amounts of KH2P04 (50. 100, 200. and 300 lig PI"1) instead of sediment. lum to avoid cells that could have taken up luxury P. The inoculum volume was chosen to start the bioas says at concentrations of 5-10 x 104 cells/ml. The tests were carried out on a shaking table (150 rpm) in Results 150-ml Jena Ehrlenmeyers with 100 ml medium at 20 ± )°C. Six Philips TLD tubes (I8W, color 54) Sediment Analyses were installed 25 cm above the shaking table with day/night regime: 12/12 h. The sediment dose was chosen so that each culture Selected physical and chemical characteristics of the contained 100-150 ng total P-l"1. The turbidity in sediments, the ortho P concentrations of interstitial our bioassays d id not exceed the natural levels of tur water, and the water immediately above the sedi bidity in our lakes during moderate wind action. If ment in the lakes are summarized in Table 2. 1 the optical density at 680 nm rose above 0.240 Total P in sediments ranged from 1.1 mg g" dw cm-1 or the pH rose above 8.6 during an experi (Lakes Reeuwijk and Wcsteinder) to 5.7 mg g~' dw ment, the bioassay was repeated with less than half (Lake Braassem). The LOI. indicating organic mat of the sediment dose (Bruning and Klapwijk, 1984). ter content, was relatively low in Lake Mooie Nel Table 2. Physical and chemical analyses of the sediment (upper 5 cm), of the interstitial water, and of water immediately above the sediment." Interstitial Water above Sediment . water sediment (mg g ' dw) (%) (mg PI -') Sand Clay

Lake n Total P Total Fe CaCOj LOl (>50M) (<2^) Ortho P Ortho P Braassem 9 5.7 ± 43% 30.7 ± 32% 156 ± 65% 27.2 ± 10% 6.9 + 65% 28.5 ± 24% 10.3 ± 32% 0.37 + 32% Kaag 8 1.7 ± 54% 23.2 ± 20% 144 ± 47% 24.5 + 12% 18.7 ± 33% 20.6 ±24% 5.4 + 84% 0.48 ± 15% Westeinder 9 1.1 + 35% 12.3 ± 13% 241 ± 45% 38.4 ± 9% 8.3 ± 60% 11.4 ± 35% 4.3 + 70% 0.22 ± 14% Nieuwe Meer 4 2.0 ± 83% 22.4 ± 60% 68 ± 51% 40.0 + 80% 11.6 ± 93% 22.1 ± 89% 7.9 ± 122% 0.40 + 12% Mooie Nel 6 5.1 ± 29% 19.9 + 43% 65 ± 13% 17.5 ± 8% 36.8 ± 74% 23.3 ± 45% 42.0 ± 49% 1.83 + 18% Reeuwijk 9 1.1 ± 21% 13.6 ± 68% 101 + 109% 52.2 ± 15% 10.3 + 57% 22.0 ± 20% 0.63 + 68% <0.02 ± 50% Nieuwkoop 6 1.2 ± 28% 26.8 ± 25% 37 ± 190% 63.2 + 15% 5.8 ± 74% 19.0 ± 20% 0.26 ± 81% 0.02 + 50% Langeraar 5 1.8 ± 19% 10.5 ± 46% 68 + 57% 52.6 ± 32% 13.9 + 133% 13.6 ± 68% 4.2 + 71% 0.26 + 58% "Means ±95% confidence intervals expressed as % of the means; n = number of sampling stations per lake. Co -124-

and high in the peaty lakes, such as Lakes Reeuwijk, 500 Nieuwkoop, and Langeraar. The sand (> 50 ft) con tent ranged from 6-37% and the clay (< 2 jt) content from 11-28%. The total Fe content of the lakes ranged from 10.5 mg g~' dw (Lake Langeraar) to

30.4 mg g-' dw (Lake Braassem). The CaC03 con tent was very variable and ranged from 4% (Lake Nieuwkoop) to 24% in Lake Westeinder. The ortho P concentrations in interstitial water differed even more between the lakes; they ranged from 0.26 mg PI-1 in Lake Nieuwkoop up to 42 mg PI-1 in Lake Mooie Nel. The ortho P concentra tions in the water immediately above the sediment were relatively low and ranged from 0.02-1.83 mg P I"1 (Lake Mooie Nel). The ortho P concentrations in interstitial water were 10-20 times higher than in the water overlying the sediment. Generally,, the composition of bottom sediments in the lakes was rather variable. In Lake Nieuwe Meer. the variability in the sediment composition is extremely high, largely due to the dumping of canal rtiud in the deeper parts of the lake. Fig. 2. Growth curves of a bioassay (in duplicate) with sediment from sampling station Wl in Lake Westeinder; the arrows (+P) indicate the moment of P addition. Internal P Concentration of Seenedesmus quadricauda the experiment, these secondary effects did not sig The internal P concentration of the Seenedesmus nificantly affect the uptake P concentrations. There cells was determined in order to calculate the fore, to calculate the amount of available P from our amount of available P from the maximum yields in bioassay results, we used an average uptake P con the bioassays. This was done by measuring the maxi centration of 0.95 iig P 1~' cm-'. This is approxi mum algal yields in bioassays with known and limit mately 10~7 ng P per cell, which corresponds fairly ing amounts of P. In Table 3, the average well with values found by Golterman et al. (1969) maximum algal yields (expressed as cm-derivative with respect to the internal P concentration of S. signals) and the calculated internal Pconcentrations obliquus under P limitation. of S. quadricauda cells are shown at different P levels. Although changing P concentrations in the Bioassays medium influenced pH and light conditions during Figure 2 presents the growth curves of a bioassay (in duplicate) with sediment from sampling station Table 3. Average maximum algal yield and internal Wl in Lake Westeinder to show the course of the (uptake) P concentration of Seenedesmus quadricauda at different P levels. algal growth during an experiment. The difference in maximum algal yield between the duplicate Internal experiments is due to small differences in the sedi P concentration P concentration Maximum with 95% ment dose. Maximum biomass was reached in 2-4 in medium algal yield confidence limits weeks, after which P was added to check whether (/^gPI1) (cm) GxgPI-' cm-') n this growth had been limited by P. In all cases, a 50 56.7 0.88 ± 0.06 2 renewed growth appeared, indicating that the maxi 100 102.5 0.98 ± 0.09 8 mum yield was limited by phosphate. 200 200.2 1.00 ± 0.19 4 The results of the bioassays with sediment are 300 335.7 0.89 ± 0.06 2 summarized in Table 4. The uptake of P by algae n = number of replicates. varied from 0.1 P g~' dw (10% of total P) in Lake -125-

S.P. Klapwijk and C. Bruning

Table 4. Means and 95% confidence intervals of total A very high correlation was also found between P and available P in the sediments (upper 5 cm) of eight the P availability and the ortho P concentration in lakes in the Waterboard of the Rijnland. the interstitial water and the water immediately Total P Available P above the sediment. Thus, ortho P concentrations (% of may be applied as a reasonable and quick estimate of Lake n (mgP g ' dw) total P) the P availability using the following regression Braassem 8 5.7 ± 2.4 1.2 ± 0.4 22 ± 5 equations: Kaag 8 1.7 ± 0.9 0.6 ± 0.3 35 ± 11 Y = 0.03 X, + 0.43, and Westeinder 9 1.1 ± 0.4 0.3 ± 0.2 26 ± 7 Nieuwe Meer 4 2.0 ± 1.7 0.7 ± 0.9 30 ± 23 Y = 1.02 X2 + 0.28, Mooie Nel 5 5.1 ± 1.5 2.2 ± 1.8 45 ± 23 Reeuwijk 9 1.1 ±0.2 0.2 ± 0.05 14 ± 3 in which: Nieuwkoop 6 1.2 ± 0.4 0.1 ± 0.05 10 ± 2 Langeraar 5 1.8 ± 0.4 0.6 ± 0.2 33 ± 4 Y = available P (in mg P g~'dw), n = number of sampling points per lake. X, = ortho P concentration of the interstitial water (in mg PI"1),

Nieuwkoop to 2.2 P g~' dw (45% of total P) in Lake X2 = ortho P concentration of the water immedi Mooie Nel. ately above the sediment (in mg P 1_1).

Relation Between Sediment Composition and P Release Discussion and Conclusions

The availability of sediment phosphate is not only The sediment P concentrations obtained in this dependent on the total P concentration in the sedi study correspond fairly well with previous work ment, but upon the amount of P-binding compo (Klapwijketal., 1982), and minor discrepancies can nents, such as organic matter, Fe, CaC03, and clay. be explained by differences in sampling locality and Furthermore, natural algal uptake is modified by the heterogeneity in the sediments. The percentages of gross P load to thelakes and their residence times. available P in our earlier study differed from the In fact, some of these factors can interfere with each present results, most probably due to changes in the other; this is undoubtedly the case in the Rijnland method óf bioassay. The derivative spectroscopy lakes, where lakes with the most organic sediments also have the lowest P load and the longest residence time. To establish the relationships between sedi Table .5. Correlation coefficients between available P ment composition and the distribution of sediment P (as mg P g~' dw and as % of total.P) and different sedi with P in the available and nonavailable fractions, ment characteristics. we computed simple correlation coefficients Correlation coefficient with between different parameters (available P as mg P Available P Available P g-' dw and as percentage of total P) and several (mg Pg-' dw) (% of total P) characteristics of the sediments (Table 5). These Total P 0.71r 0.17 coefficients are slightly different from those we Total Fe 0.34" -0.10 presented earlier (Klapwijk and Bruning, 1984); CaC03 -0.03 0.07 earlier coefficients were based on lake averages (n = LOI -0.55c -0.53r 8), and the figures in Table 5 are based on data Clay « 2 „) 0.24 -0.04 r c from 50 separate sampling stations. Sand(> 50^) 0.50 0.48 Ortho P (interstitial f The amount of available P is strongly positively water) 0.50r 0.39* correlated with total P (p < 0.001). Among the P- Ortho P (water above r c binding components—Fe, CaC0 , organic matter, sediment) 0.65 0.57 3 n 50 50 and clay—only organic matter (as indicated by LOI) showed a strong negative correlation with P availa n = number of paired observations. ap < 0.05. bility. Thus, much of the sediment P occurs in forms hp < 0.01. bound to organic matter. cp < 0.001. -126-

Available Phosphorus in Lake Sediments

technique replaced the need to count algal cells respect to the manuscript; to P. van Doorn for draw (Bruning and Klapwijk, 1984), and pH in the bio- ing the figures; to W. G. Hey for executing the assays was kept below 8.6 in the present study. This statistical analyses; to Paul Gutteridge for correct is important, since we have shown previously (Brun ing, and Miss C.A.L.M. van Dijk for typing the ing and Klapwijk, 1984) that the amount of available English text; and to the Board of Rijnland Water P can be more than doubled when pH in the bio- Authority for permission to publish these results. assays rises above 9.0. We believe this explains why we found 36% available P in Lake Braassem in our earlier study (Klapwijk et al., 1982), while only 22% References appeared to be available in the present study. From the high correlations between available P in Al. J.P. and Holland. A.M.B.. 1977. Geochemische the top sediment layer of all our different lakes and bemonsterings- en analyse methodieken. Deltadienst. hoofdafdeling Milieu en Inrichting |in Dutch). Nota. the ortho P concentrations in the interstitial water 76-60, 38 pp. and the immediately overlying water, we concluded Bolier. G., van Breemen, AN. and Visser. G.. 1981. that these ortho P concentrations can be used as a Eutrophication tests: A possibility to estimate the reasonable and quick estimate of immediately avail influence of sewage discharge on the biological quality able sediment P in the upper part of our sediment of the receiving water (in Dutch). H20. 14(4):88-92. cores. Bruning, C. and Klapwijk. S.P., 1984. Application of Since the amount (per unit of volume) of available derivative spectroscopy in bioassays estimating algal P in the sediment is at least 10-20 times higher than available phosphate in lake sediments. Verh. Internal!. in the interstitial water, this implies that immediate Verein. Limnol., 22:172-178. P release cannot simply be explained by the amount de Rooij. N.M.. 1980. A chemical model to describe of dissolved Pin the interstitial water—but the ortho nutrient dynamics in lakes. In: J. Barica and L.R. Mur (Editors). Hypertrophic Ecosystems. Developments in P concentrations in the interstitial water and the Hydrobiology. vol. 2. pp. 139-149. Junk. The Hague, immediately overlying water are in equilibrium with The Netherlands. the amount of available sediment P. Golterman, H.L.. 1977. Sediments as a source of phos The large differences in total and available P in the phate for algal growth. In: H.L. Golterman (Editor). sediments from the various lakes seem to be not only Interactions Between Sediments and Fresh Water. a result of the P loading, but also of the sediment Proceedings of an International Symposium Held at characteristics such as the organic matter content. In Amsterdam, The Netherlands. September 6-10. 1976. the Rijnland Waterboard area, sediments with low Junk, The Hague. Pudoc. Wageningen, The Nether organic matter content and high P loading contain lands. pp. 286-293. high P concentrations and high percentage availabil Golterman, H.L , 1982. Differential extraction of sedi ment phosphates with NTA solution. Hydrobiologia, ity (Lake Mooie Nel); organic sediments with low P 92:683-687. loading contain low P concentrations and low per Golterman, H.L., Bakels. C.C. and Jakobs-Mögelin. J., centage availability (Lakes Reeuwijk and 1969. Availability of mud phosphates for the growth of Nieuwkoop). In the Lake Langeraar internal loading algae. Verh. Internall. Verein. Limnol., 17:467-479. dominates the eutrophication process. Before dredg Grobler, DC. and Davies. E., 1979. The availability of ing, however, bioassays will be carried out with sedi sediment phosphate to algae. Whter S.A., 5:114-122. ment material from the deeper layers to establish the Grobler, D.C. and Davies, E., 1981. Sediments as a source availability of phosphates. of phosphate: A study of 38 impoundments. WjterS.A.. Lakes Nieuwkoop and Reeuwijk have the lowest P 7:54-60. availability from sediment, and can be identified as Hegemann, D.A., Johnson, A.H. andKeenan, J.D., 1983. Determination of algal-available phosphorus on soil and areas where measures to reduce the external P load sediment: A review and analysis. J. Environ. Qua!., arc most likely to succeed. 12(1):12-16. Hieltjes, A.H.M, and Lijklema, L.. 1980. Fractionation of inorganic phosphates in calcareous sediments. J. Em-iron. Qua!., 9(3):405-407. Acknowledgments. The authors are greatly indebted Klapwijk. SP. and Bruning, C, t984. Available phospho to Ir. M. A. Heinsdijk, Professor Dr. M. Donze, and rus in the sediments of eight lakes in the Netherlands. Dr. J. van der Does for valuable suggestions with In: Proc. 3rd Internal!. Symp. on Interactions Between -127-

Sediments and Water. CEP Consultants, Edinburgh, Ohnstad, F.R. and Jones, J.G., 1982. The Jenkin Surface pp. 76-79. Mud Sampler User Manual. Freshwater Biological Klapwijk, S.P., Kroon, J.M.W. and Meijer, M.L., 1982. Association, Occasional Publication no. 15, Amble- Available phosphorus in lake sediments in the Nether side, U.K., 45 pp. lands. Hydrobiologia, 92:491-500. Schwoerbel, J., 1972. Methods of Hydrobiology, ed. 2. Kouwe, F.A. and Golterman, H.L., 1976. The role of Pergamon Press, Oxford, United Kingdom, 200 pp. sediment phosphates in the eutrophication process (in Williams, J.D.H., Syers, J.K., Shukla, S.S., Harris, R.F. Dutch). H20, 9(5):84-86. and Armstrong, D.E., 1971. Levels of inorganic and Liere, L. vanandMur, L.R., 1982. The influence of simu total phosphorus in lake sediments as related to other lated groundwater movement on the phosphorus release sediment parameters. Environ. Sci. and Technol., from sediments, as measured in a continuous flow sys 5:1113-1120. tem. Hydrobiologia, 92:511-518. Williams, J.D.H., Shear, H. and Thomas, R.L., 1980. NEN, 1969. Test Methods for Waste Water (in Dutch). Availability to Scenedesmus quadricauda of different Nederlands Normalisatie Instituut, NEN 3235, Rij forms of phosphorus in sedimentary materials from the swijk, The Netherlands. Great Lakes. Limnol. Oceanogr., 25:1-11. -128-

Foto pag. 129: De in de bioassays gebruikte testalg Scenedesmus quadricauda (Turp.) Bréb.. PARTC: BIOASSAYS

«ft

•^

\ >> K***** -131-

CHAPTER 10:

INTRODUCTION BIOASSAYS

"The significance of measuring the algal growth potential of water is that a differentiation can be made between the nutrients that are in the sample (as determined by chemical analysis) and the nutrient forms that are actually available for algal growth."

W.E. Miller, J.C. Greene and T. Shiroyama, 1978. The Selenastrum capricornutum Printz algal assay bottle test. Environmental Protection Agency EPA-600/9-78-018, p. 1. Cor- vallis, Oregon, U.S.A. -132-

INTRODUCTION BIOASSAYS

In studies related to the eutrophication of surface waters it is important to know the possible algal growth (the so-called algal growth potential) in these waters and to establish those factors which are limiting maximum growth of phy toplankton. This applies especially for lake restoration programs, where the effects of phosphorus load reduction have to be predicted with some precision. One of the most popular techniques to measure algal growth potential and to establish the limiting factors is the bioassay technique. The principle of the bioassay method is based on von Liebig's Law, which states that the yield (biomass) of a species (or popula tion) is determined by the nutrient, which is present in the least available quantity if not other factors, such as light, temperature etc. are growth limiting (cf. Miller et al., 1978; Rodhe, 1978). Addi tion of this nutrient will raise the yield. With such bioassays the growth potential and the limiting nutrient(s) can be detected. This is mostly done under standarized conditions in the laboratory (cf. E.P.A., 1971; Nordforsk, 1973; Miller et al. , 1978; Marvan et al., 1979; Ryding, 1980). According to several authors (e.g. Miller et al., 1978; Forsberg et al. , 1978; Raschke & Schultz, 1987) the signifi cance of such assays is that they can differentiate between the nu trients that are present in a sample and the availability of these nu trients for algal growth. Algal bioassays have been applied in limnology since the early fifties (Thomas, 1953; Bringmann & Kiihn, 1965; E.P.A., 1971) and are used nowadays worldwide (e.g. Nordforsk, 1973; Chiaudani & Vighi, 1974, 1975, 1976; Miller et al. , 1974, 1978; E.P.A., 1975; S.I.L., 1978; Marvan et al. , 1979; Ryding, 1980). In the Netherlands bioassays were used in the first study of eutrophication of the shal low lakes in the Rijnland area more than 10 years ago. With their aid Schmidt-van Dorp (1975, 1978) found that most lakes in the Rijnland area were nitrogen limited_ and not phosphorus limited due to high P concentrations (> 0.2 mg 1 1) and low N/P ratios (< 16). In the Netherlands bioassays are applied to measure the growth potential of surface waters and sewage effluents (Bolier et al. , 1981; Veenstra, 1981; de Vries & Hotting, 1985; de Vries & Klapwijk, 1987) or to assess the limiting nutrients for algal growth (Schmidt-van Dorp, 1975, 1978; de Haan et al. , 1982; van Donk, 1983; de Vries et al., 1983), but also to investigate the effects of phosphorus removal at sewage treatment plants (Klapwijk, 1981; Hoogheemraadschap van Rijnland, 1984; van der Does & Klapwijk, 1985, 1987). Bioassays can be carried out both with the indigenous phytoplankton population (Thomas, 1953; Goldman, 1978; Schmidt-van Dorp, 1975, 1978;' de Haan et al., 1982; van der Does & Klapwijk, 1987) as well as with unialgal cultures. In Chapter 11 bioassays with the natu ral phy toplankton population are described, carried out from 1980- 1982 with water from four lakes in the Rijnland Waterboard area. In Chapter 12 the results of 440 algal growth potential tests with the testalga Scenedesmus quadricauda are presented, carried out in 24 surface waters (lakes, canals) also in the Rijnland area from 1983- 1986. Special attention has been paid in this chapter to the effects of pretreatment of the sampled water and to the relation between algal growth potential and the nutrient concentrations in the water. -133-

In order to know whether different test organisms showed the same reaction to nutrient concentrations and/or ratios or not, a com parative bioassay study was carried out using Stigeoclonium tenue Kütz., a filamentous greenalga that occurs frequently in polder dit ches, and Scenedesmus quadricauda (Turp.) Bréb., a chlorococcal greenalga for phytoplankton populations in larger bodies of water, as testorganisms (Chapter 13). Finally, in Chapter 14 other methods to assess growth limiting factors for phytoplankton, nutrient uptake rates, nutrient ratios and statistical relationships between chlorophyll-a and nutrient concentra tions are compared with bioassay results from the Reeuwijk lakes.

REFERENCES Bolier, G., A.N. van Breemen & G. Visser, 1981. Eutrophication tests: a possibility to estimate the influence of sewage discharge on the biological quality of the receiving water. H20 14: 88-92 (in Dutch with an English summary).

Bringmann, P.G. & R. Kiihn, 1965. Nitrat oder Phosphat als Begrenzungsfaktor des Algenwachstums. Gesundheitsing. 7: 210-214.

Chiaudani, G. & M. Vighi, 1974. The N.P ratio and tests with Selenastrum to predict eutrophica tion in lakes. Water Res. 8: 1063-1069.

Chiaudani, G. & M. Vighi, 1975. Dynamics of nutrient limitations in six small lakes. Verh. inter nat. Verein. Limnol. 19: 1319-1324.

Chiaudani, G. & M. Vighi, 1976. Comparison of different techniques for detecting limiting or sur plus nitrogen in batch cultures of Selenastrum capricornutum. Water Res. 10: 725-729.

Does, J. van der & S.P. Klapwijk, 1985. Phosphorus removal and effects on waterquality in Rijnland. H20 18: 381-387 (in Dutch with an English summary).

Does, J. van der & S.P. Klapwijk, 1987. Effects of phosphorus removal on the maximal algal growth in bioassay experiments with water from four Dutch lakes. Int. Revue ges. Hydrobiol. 72: 27-39.

Dorik, E. van, 1983. Factors influencing phytoplankton growth and succession in lake Maarsseveen. Ph.D. Thesis, University of Amsterdam.

Environmental Protection Agency (E.P.A.), 1971. Algal Assay Procedure, National Eutrophication Research Program. Corvallis, Oregon. -134-

Environmental Protection Agency (E.P.A.), 1975. Biostimulation and nutrient assessment. Proceedings Workshop, October, 1973. CorvaUis, Oregon. EPA-660/3-75-034.

Forsberg, C, S.-O. Ryding, A. Claesson & A. Forsberg, 1978. Water chemical analyses and/or algal assay? - Sewage effluent and polluted lake water studies. Mitt, internat. Verein. Lünnol. 21: 352-363.

Goldman, 1978. The use of natural phytoplankton populations in bioassay. Mitt. internat. Verein. Limnol. 21: 364-371.

Haan, H. de, J.B.W. Wanders & J.R. Moed, 1982. Multiple addition bioassay of Tjeukemeer water. Hydrobiologia 88: 233-244.

Klapwijk, S.P. , 1981. Limnological research on the effects of phosphate removal in Rijnland. H20 14: 472-483 (in Dutch with an English summary). Marvan, P., S. Pfibil & O. Lhotsky (eds), 1979. Algal assays and monitoring eutrophication. E. Schweizerbart'sche Verlagsbuchhandlung, Stuttgart.

Miller, W.E., T.E. Maloney & J.C. Greene, 1974. Algal productivity in 49 lake waters as determined by algal as says. Water Res. 8: 667-679.

Miller, W.E., J.C. Greene & T. Shiroyama, 1978. The Selenastrum capricornutum Printz algal assay bottle test. Environmental Protection Agency (EPA). Corvallis, Oregon.

Nordforsk, 1973. Algal assays in water pollution research. Proceedings from a Nordic symposium, October 1972, Oslo. Nordforsk secretariat of Environmental Sciences Publ. , 1973: 2.

Raschke, R.L. & D.A. Schultz, 1987. The use of the algal growth potential test for data assessment. J. Wat. Pollut. Contr. Fed. 59: 222-227.

Rodhe, W. , 1978. Algal in culture and nature. Mitt, internat. Verein. Limnol. 21: 7-20.

Ryding, S-O. (ed.), 1980. Monitoring of inland waters. Report from the working group for eutrophication research. Nordforsk publication 1980: 2, Helsing- fors, Finland. -135-

Schmidt-van Dorp, A.D., 1975. Phosphate and eutrophication in the Waterboard of Rhineland. H20 8: 254-258 (in Dutch with an English summary). Schmidt-van Dorp, A.D., 1978. Eutrophication of shallow lakes in Rijnland. Report Technical Service, Hoogheemraadschap van Rijnland, Leiden (in Dutch with an English summary).

S.I.L., 1978. Symposium: Experimental use of algal cultures in limnology. Mitt. internat. Verein. Limnol. 21: 1-607. E. Schweizerbart'sche Ver- lagsbuchhandlung, Stuttgart.

Thomas, E.A., 1953. Zur Bekampfung der See-Eutrophierung: Empirische und experimentelle Untersuchungen zur Kenntnis der Minimumstoffen in 46 Seen der Schweiz und angrenzender Gebiete. Monatsbull. Schweiz. Ver. Gas- Wasserfachm. 33: 25-32; 71-79.

Veenstra, S., 1981. De algengroei-potentie-toets. Een nieuw instrument ter beoorde ling van het eutrofiërend karakter van een water. Landbouw hogeschool Wageningen, Vakgroep Waterzuivering, sektie Hydro- biologie. Doktoraal verslagen serie nr. 81-1.

Vries, P.J.R. de, M. Torenbeek & H. Hillebrand, 1983. Bioassays with Stigeoclonium Kütz (Chlorophyceae) to identify nitrogen and phosphorus limitations. Aquat. Bot. 17: 95-106.

Vries, P.J.R. de & E.J. Hotting, 1985. Bioassays with Stigeoclonium tenue Kütz on waters receiving se wage effluents. Water Res. 19: 1405-1410.

Vries, P.J.R. de & S.P. Klapwijk, 1987. Bioassays using Stigeoclonium tenue Kütz. and Scenedesmus quadricauda (Turp.) Bréb. as testorganisms; a comparative study. Hydrobiologia 153: 149-157. -137-

CHAPTER 11: EFFECTS OF PHOSPHORUS REMOVAL ON THE MAXIMAL ALGAL GROWTH IN BIOASSAY EXPERIMENTS WITH WATER FROM FOUR DUTCH LAKES J. van der Does & S.P. Klapwijk.

Published in: Int. Revue ges. Hydrobiol. 72: 27-39 (1987).

"Die anthropogene Düngstoffzufuhr bewirkte in manchen Seen, dass an Stelle oder neben Phosphorverbindungen die Stickstoffverbindun- gen zu Minimumstoffen wurden. "

E.A. Thomas, 1953. Zur Bekampfung der See-Eutrophierung: Empirische und experimentelle Untersuchungen zur Kenntnis der Minimumstoffe in 46 Seen der Schweiz und angrenzender Gebiete. Monatsbulletin Schweiz. Verein von Gas- und Wasserfachmannern 33, p. 78. -138-

1 Int. Revue ges. Hydrobiol. 72 1987 1 27-39

J. VAN DER DOES and S. P. KLAPWIJK

Waterboard of Rijnland, Leiden, The Netherlands

Effects of Phosphorus Removal on the Maximal Algal Growth in Bioassay Experiments with Water from Four Dutch Lakes

key words: algal growth potential,-bioassay. limiting factor, eutrophication, phosphorus removal, Rijnland

Abstract

This paper reports the effects of phosphorus removal at three sewage wastewater treatment plants on the state of eutrophication of four shallow lakes in the south-eastern part of the Rijnland Waterboard area during the years 1980—1982. With chemical analyses and bioassay experiments using the natural phytoplankton population no significant lowering could be detected of respec tively the phosphate concentration and the maximal algal growth potential. All lakes proved to be principally nitrogen limited except the Reeuwijk Lakes, which showed clearly, after a primary nitrogen limitation, a secondary phosphorus limitation. Therefore the main attention with respect to phosphorus reduction should be concentrated on the Reeuwijk Lakes in the first place. For the other lakes in the investigated area phosphorus removal will, when it is the only measure taken, presumably not lead at short notice to a decrease of the algal biomass.

1. Introduction

Due to the increasing nutrient load of the river Rhine and the growth of the human population, eutrophication in the Netherlands has progressed so far in the last decades 1 hat most of the inland lakes may he considered hypertrophic (LEENTVAAB, 1980). In the area of the Rijnland Water Authority, situated in the delta of the Rhine (Fig. 1) the phosphorus loading is so high, that nitrogen became the primary limiting factor for the growth of algae in most of the lakes (SCHMIDT-VAN DORP, 1978). How ever, within this area large differences exist between the mean total P values, ranging from 0.2 nig P/l in the south-east up to 2.5 mg P/l in the north-west (KLAPWIJK, 1981). The aim of the Dutch government is to reduce the external phosphorus load to those surface waters that are sensitive to eutrophication (Ministry for Transport and Public Works, 1980). The south-eastern part of Rijnland can be considered to be an area where the reduction of the external phosphorus loading may be a successful method in low ering algal biomass. As a lake restoration experiment the Waterboard of Rijn land started phosphorus removal at the sewage wastewater treatment plants of Gouda, Bodegraven and Nieuwveen (Fig. 1) in the autumn of 1979, resulting in a P-reduction of approximately 00 tons of phosphorus per year (KLAPWIJK, 1981). The effects of this P removal on the lakes were followed by hydrochemical and limnological studies, which included plankton analyses, bioassays and sediment in vestigations. A full report of this research project has been published elsewhere (Waterboard of Rijnland, 1984). The bioassays to determine the availability of nutrients for algal growth were carried out with both the natural phytoplankton population and with test-algae. -139-

Kigure 1. The area of the Rijnland Water Authority, its location in the Netherlands (inset), and the location of the sampled lakes.

The principle of the bioassays is based on Von Liebig's law, which states that the yield (biomass) of a species (or population) is determined by the nutrient, which is present in the least available quantity. Addition of this nutrient will raise the yield. With such bioassays the growth potential and the limiting nutrient factor (N, P orN + P) can be detected (BOLIEH, VAN BEEEMEN and VISSER, 1981; EPA, 1971; FORSBERG et al., 1978; Nordforsk, 1973). This paper presents the results of chemical investigations over a six year period (1977—1982) and the results of bioassay ex periments with the natural phytoplankton population from four lakes during the period of reduction of the external phosphorus loading (1980-1982). -140-

2. Sampling Area, Materials and Methods

The Rijnland Water Authority encompasses a densely populated area (about 10° inhabitants in 1000 km2) situated between the cities of Amsterdam, Haarlem, The Hague and Gouda (Fig. 1). In the "boezem" of Rijnland, an interconnected system of canals, lakes and ditches, the water level is kept constant at 0.6 m below mean sea level. Water can be pumped in near Gouda, from the Hollandse Ijssel, a branch of the Rhine, while excess water can be pumped out to the sea. The boezem supplies a great number of polders with water. Some lakes, such as the Braassem Lake and the Westeinder Lakes form part of this boezem system. Others, such as the Nieuwkoop and Reeuwijk Lakes, situated in polders about 2 in below mean sea level, are separated from the boezem but also depend ent on it for their water supply. The investigated lakes (Braassem Lake, Westeinder Lakes, Nieuwkoop Lakes and Reeuwijk Lakes) are rather small (5—11 km2) and shallow (1.5-4 m mean depth). General characteristics of morphometry, hydrology and phos phate loading are given bv KLAPWIJK, KROON and MEIJER (1982) and Waterboard of Rijnland (1984). Water for chemical analyses was sampled once or twice every month for 6 years in 1 litre milk bottles at 0.5 m depth. Nitrate + nitrite nitrogen (XO.^ +NOJ-N), ammonium nitrogen (NH,+ -N) and orthophosphate (o-P) were measured on a Teehnicon AA II autoanalyser. Kjeldahl-nitrogen (Kj-N) and total Phosphorus (t-P) were measured after digestion with respectively sulphuric acid-fWieninger selenium mixture and sulphuric acid + persulphate. Subsequently NH^"-N and o-P were deter mined on the autoanalyser. Total nitrogen was calculated by addition of the weights of Kj-N and NO^ + NOj-N. Inorganic nitrogen (inorg. N') was calculated by addition of the weights of NHJ-N and NOj +NOJ-N. Water samples for bioassays were taken in March, June, September and December with a Kriedinger sampling apparatus which is 45 cm long and has a volume of 3.5 litres. A vertical water column was taken just above Sccchi depth. If this depth was less than 45 cm, a water column of 0-45 cm was sampled. The water was transported in dark polyethylene flasks. Zooplankton was removed in the laboratory with an 80 u.m plankton gauze. Water containing the natural phytoplankton population was divided into eight 2-litre Erlenmeyer flasks filled up to I I. The experiments were carried out in duplo. Two Erlenmeyers had no addition of nutrients. The others were enriched with respectively 2 mg N/l (as KNO:1) or 0.1 mg P/l (as K2HPO,,/KH2PO,,) or 2 mg N/l-■•0.1 mg P/l. The culture vessels, closed with cotton plugs, were placed in a shaking water bath at a temperature of 20 + 1 °C, a light intensity of about 5000 lux, and a light/dark rhythm of 12/12 hours (lamp type Philips TL 40 W/75, 4 above and 4 beneath the cultures). The amount of chlorophyll « measured after extraction with ethanol, according to a standard method of the Nederlands Normalisatie Instituut (MEN 0250), was used as a parameter for growth. .Maximal biomass was calculated from no more than 4 data out of the stationary growth phase. When the growth was less than 5 % per day twice, these two and the preceding data were used to calculate maximal biomass after rejection of outliers (KPA, 1971). The difference in mean maximal yield (of duplo experiments) with and without nutrient enrichment was tested for significance' applying the /-test for means of two samples (SOKAL and ROHLF, 1969). Product-moment correla tions (SOKAI. and Rom.F, 1909) were computed between the algal growth potential and the in organic and total P and X concentrations in ever\' lake.

3. Results and Discussion

The results of the chemical analyses over the period 1977—1982 are presented in Figures, 2, 3, 5, 6, 8, 9, 12 and 13. The results of the bioassays are shown in Figures 4, 7, 10 and 11. The chlorophyll a concentrations in the lakes at the moment of sampling are indicated in these figures with crosses (chl. a, Z = 0). -141-

J. VAN DEB DOES and S. P. KLAPWIJK

a) Braassem Lake

Braassem Lake had high concentrations of nitrogen in spring (6-8 mg N/1). In sum mer a strong decrease in the inorganic and total nitrogen concentrations was observed (Fig. 2). The phosphorus concentrations were also very high. The total phosphorus concen tration in winter exceeded 0.5 mg P/1. The o-P concentration did not decrease below 0.2 mg P/1 (Fig. 3). It is remarkable that the oscillations in the nitrogen and phos phorus concentration were out of phase. The increase in phosphorus concentration

01JAN77 01JAN7B 01JAN79 01JAN80 01JANS1 01JAN82 31DECB2 Figure 2. Total (♦)- and inorganic (•*) nitrogen (mg N/l) in Braassem Lake (1977—1892).

Braeaaem Lake

0.2-

01JAH77 01JAN7B 01JAN79 01JAN80 01JANB1 01JANS2 3106082 Figure 3. Total (♦)- and ortho (*) phosphate (mg P/1) in Braassem Lake (1977-1982). -142-

Effects of Phosphorus Removal in Dutch Lakes started in July/August, whereas the decrease began round about January. On the other hand, the increase in the nitrogen concentration started a few months later and continued while the phosphorus concentration was already decreasing. The start of the increase in the phosphorus concentration coincided with the rise of the pH- values. This phenomenon can be related to the short residence time of Braassem Lake. In summer the lake is filled with phosphate-rich water from the river Rhine. In winter and spring the lake is filled with superfluous rainwater pumped out of the polders with relatively low phosphate and high nitrogen concentrations. The high concentrations of inorganic nitrogen in spring originate from leaching of the agri cultural polderland. The reason of the steady decrease in mean nitrogen values over the period 1977— 1982 is not known. The bioassay results show that the growth potential in Braassem Lake (Fig. 4, no

chl.a.[mg/m3]

900

800 x chl.a,t=o o no addition 700 . *N m + P 600 ■ + N+P

500

400

300

200

100

0 J FMAHJ J A S ON DU F HAMJ J ASO N DU F MA M J J ASO N DI 1980 1981 1982 months >

Figure 4. Maximal algal yield in bioassays (means of duplo experiments) with water from Braas sem lake (1980-1982).

addition) varied between 20 and 740 nig chlorophyll a/m\ The maxima were reached in March, whereas the minima were measured in September (Fig. 4). The growth po tential seems to diminish between 1980 and 1982. The enrichment experiments indi cate that nitrogen is the limiting nutrient factor for algal growth. This is supported by the fact that there is a significant (P< 0.001) correlation between the growth poten tial and t-N and inorg. N (respectively r = 0.88 and r = 0.91). No significant correlation is found between the growth potential and o-P or t-P. -143-

J. VAN DER DOES and S. P. KLAPWIJK

It is remarkable that the chlorophyll a concentration in Braassem Lake was low compared to the growth potential. This was probably not caused by a toxic effect, because the natural phytoplankton population grew well under laboratory conditions.

b) Westeinder Lakes

The Westeinder Lakes have, like Braassem Lake, high inorganic nitrogen concen trations in winter (Fig. 5). The very high inorganic nitrogen concentration in March 1981 coincided with an excess of rainwater pumped out of the polders. The phosphorus

Westelnder Lakes

t 1 2-

01JAN77 01JAN7B . 01JAN79 01JAN80 01JAN81 01JAN82 310ECB2 .KiKiinT). Total (♦)- and inorganic (*) nitrogen (mg N/l) in the Westeinder Lakes (1977—1982). graphs presented in Figure (5 show t-P concentrations between 0.1—0.6 mg P/1 and o-l' concentrations between 0.02 and 0.5 mg P/1. In spring (March, April), o-P is sometimes depleted coinciding with a bloom of diatoms. The growth potential in the Westeinder Lakes varied between 5 and 400 mg chlorophyll a/m3 (Fig. 7, no addition). All maxima were measured in March. Increase or decrease of the growth potential was not perceptible during the years 1980—1982. Enrichment experiments indicate that nitrogen remained the primary limiting nu trient factor for the maximal algal biomas. This is supported by calculations of the correlation coefficients between the growth potential and t-N or inorg. X, respectively 0.92 (P<0.001) and 0.71 (F<=0.01). The coefficients with t-P and o-P were respectively-0.10 and 0.17. The difference between the chlorophyll a values and the growth potential was much smaller in the Westeinder Lakes than in Braassem Lake. -144-

Effects of Phosphorus Removal in Dutch Lakes

Westainder Lakes

c o 0.4-

n 0.3- t

0.0-

01JAN77 01JAN7B 01JAN79 01JAN80 01JAN81 01JANS2 31DEC82

Figure 6. Total («>)-and ortho(#) phosphate (mg P/l) in the Westeinder Lakes (1977-1982).

chla.[mg/m3] '800 X Chl.Q,t=0 o no addition 700 ■ + N a + p 600 ■ + N + P

500

400

300

200

100 0 J J FMAMJ J A SON DU FMAMJ J A SON Ob FMAM J J AS ON D] 1980 1981 1982 months ,.

Figure 7. Maximal algal yield in bioassays (means of duplo experiments) with water from the Westeinder Lakes (1980-1982). 3 Int. llevue ges. Hydrobiol. 72 (1987) 1 -145-

J. VAN DER DOES and S. P. KLAPWIJK

e) Nieuwkoop Lakes

The results of the hydrochemical analyses of the Nieuwkoop Lakes are shown in Figures 8 and 9. The nitrogen and phosphorus concentrations were higher than in the Reeuwijk Lakes. A remarkable periodicity was observed in the nitrogen and phos-

Nieuwkoop Lakes

01JAN77 01JAN7B 01JAN79 01JAN80 01JAN81 01JAN82 31DEC82 Figure 8. Total (♦)- and inorganic (*) nitrogen (mg N/l) in the Nieuwkoop Lakes (1977—1982).

phorus concentrations. High total nitrogen concentrations in summer synchronized with low concentrations of inorganic nitrogen. The total phosphorus concentrations were minimal in winter, whereas the maxima were reached during the summer months. These phenomena were the opposite of the results obtained from the Braassem Lake. The algal growth potential varied between 80 and 380 mg chlorophyll a /m3 (Fig. 10, no addition). The highest growth potentials were measured in September and the lowest in March or December, again as contrasted with Braassem Lake. Instead of a decreasing algal growth potential during the years 1980—1982 a remarkable increase was observed. Moreover, nitrogen remained the primary limiting nutrient factor for the maximal algal biomass. Correlation computations between the growth potential and both the chemical parameters and chlorophyll a showed a significant correlation only with t-N and chlorophyll a, respectively r = 0.69 (P<0.01) and r = 0.94 (P-=0.001). The results of the hydrochemical analyses and bioassays in the Nieuwkoop Lakes were influenced primarily by the water supply to the lakes. In winter total nitrogen and phosphorus concentrations decreased due to dilution with rainwater. Because of evaporation in summer the lakes take in water from the boezem with high nitrogen and phosphorus concentrations. However, the strong increase of the growth potential cannot sufficiently be explained by the increase of the total nitrogen concentration. Possibly N2-fixation by Aphanizomenon flos-aquae also contributed to -146-

Effeets of Phosphorus Removal in Dutch Lakes

Nieuwkoop Lakes

0.4-

a 0.2- t 1

0.0-

01JAN77 01JAN78 01JAN79 01JAN8O 01JANB1 01JAN82 31DEC82

Figure 9. Total (♦)- and ortho (*) phosphate (mg P/1) in the Nieuwkoop Lakes (1977-1982)

chl.a [mg/m3l x chla, t=o D no addition ■> + N n + p - ♦ N+P

J FMAMJJAS 6'NDIJ FMAMJ J A S 0 ND! j F M A M J J A S 0 N "51 1980 19811 1982 months

Figure 10. Maximal algal yield in bioassays (means of duplo experiments) with water from the Nieuwkoop Lakes (1980-1982). -147-

J. VAN DEK DOES and S. P. KLAPWIJK the increase of the growth potential. The possibility that N2-fixation is dependent on the phosphate concentration cannot he dismissed. Therefore the algal growth potential may also be influenced by phosphorus released from the lake sediments. The bioavailability of the sediment phosphates proved to be pH dependent (BEUNING and KLAP WIJK, 1984). In summer algal growth results in an increase of the pH values with an extra increase of available phosphorus released from the sediment and a consequent increase in algal biomass. In the Nieuwkoop Lakes maximum pH-values (up to 9.3) and ortho-phosphate concentrations were higher than in the Reeuwijk Lakes. Data on the availability of phosphorus from sediments for algal growth in the four lakes will be published elsewhere (KLAFWUK and BEUNING, 1986).

d) Reeuwijk Lakes

The Reeuwijk Lakes had relatively low nitrogen and phosphorus concentrations (Figs. 12, 13). Total nitrogen varied between 2-3 mg N/l, whereas inorganic N stayed ai. about 0.2 mg N/l with maxima (=-0.5 mg N./l) in winter. The o-P concentration was

x chl.a. t=o chlalmg/m'] . D no addition k m + M BrP

300 :

200

100 -

° jj F"M"A MT J AT0"N~D|7 FMAMJJASONDUFMAMJ JASON D| 1980 1981 1982 months t Kimirc. 11. .Maximal algal yield in Ijioassays (means of duplo experiments) with water, from the Reeuwijk'Lakes (1980-1982). often measured at the detection level (O.Ol mg P/l). The exceptionally high eoncen- I rat ions of phosphorus and nitrogen in 1980 were most probably due to the temporary drainage of the effluent from the sewage wastewater treatment plant at Reeuwijk via the lakes between August 1979 and January 1981. The algal growth potential in the Reeuwijk lakes varied between 50 and 115 mg chlorophyll «/m^ (Fig. 11, no addition). A decrease in growth potential as a result of the reduction of phosphorus loading was not found. Calculation of correlation coeffi cients did not show any significant positive correlation (a =0.05) between the growth potential and the nitrogen and phosphorus parameters. However, the enrichment experiments showed very clearly that the growth potential is limited by primarily nitrogen but also by phosphorus. -148-

Effects of Phosphorus Removal in Dutch Lakes

Reeuwijk Lakes

*-\

01JAN77 01JAN78 01JAN79 01JAN80 01JAN81 01JAN82 31DEC82

Figure 12. Total (♦)- and inorganic (*) nitrogen (mg N/1) in the Reeuwijk Lakes (1977—1982).

ReeuHljk Lakes

'—»*■»—•—■—1—■— 01JAN77 01JAN78 01JAN81 Figure 13. Total (>)- and ortho (*) phosphate (mg/1) in the Reeuwijk Lakes (1977-1982). ^•149-

J. VAN DER DOES and S. P. KLAPWIJK

4. General Discussion and Conclusions

Eutrophication in surface waters in the Netherlands has progressed so far that an easy and rapid success cannot be expected from the removal of phosphorus at wastewater treatment plants. Other authors (e.g. BJÖBK, 1978; FRUH, 1967) also stated with reference to different countries and situations that lake restoration is time-consuming and complicated, because of the different sources of the phosphorus loading. However, the south-east of Rijnland was chosen to make researches into the effects of phosphorus removal be cause of the relatively low phosphorus concentrations. Except for the Braassem Lake there was no reduction of the algal growth potential observed during the period of phosphorus removal (1980—1982). Yet the decrease of algal growth potential in Braas sem Lake is probably not due to the phosphorus reduction, because nitrogen remained the limiting nutrient factor in the lake arid the phosphorus concentration did not decrease below 0.2 nig P/l, a value beyond which SCHMIDT-VAN DORP (1978) did not expect a phosphorus limitation for the investigated lakes. The results are not influenced by the use of the natural phytoplankton population in the bioassay method, because the use of cultures with the test alga Scenedesmuts quadricauda led to the same conclusions (Waterboard of Rijnland, 1984). Moreover, HANNA and DAUTA (1983) investigated 3 parameters for phosphorus bioavailability in bioassays. They came to the conclusion that the yield (algal biomass) is the most reliable parameter for phosphorus bioavailability. Furthermore, the hydrocheniical analyses showed no decrease in the P concen trations during the period of phosphorus removal. The observation that the algal growth potential and the phosphate concentration in the lakes is not lowered by the phosphorus removal at the wastewater treatment plants of Gouda, Bodegraven and Nieuwveen, may have several causes. The hoezem- lakes (Braassem and Westeinder) are strongly influenced by the inlet water from the river Rhine. Agriculture also contributes to the eutrophication of the surface waters as does the discharge of unpurified wastewater which still occurs in some places. The reduction of the phosphorus loading was expected to be most effective in the Reeuwijk Lakes, because of the relatively low phosphorus concentrations in 1978 and the first half year of 1979. However, the effectiveness of the phosphorus removal at the three wastewater treatment plants could not be measured in the Reeuwijk Lakes because of the temporary drainage of the effluent of the wastewater treatment plant of the village of Reeuwijk via the lakes. Summarizing we can conclude that the phosphorus removal at the wastewater treatment plants of Gouda, Bodegraven and Nieuwveen did not contribute signifi cantly to the diminishing of eutrophication in the lakes in the south-eastern part of Rijnland. Chemical analyses and bioassay experiments showed no significant decrease in the phosphate concentration and the algal growth potential in the lakes. All lakes proved to be principally nitrogen limited except for the Reeuwijk Lakes which showed clearly, after a primary nitrogen limitation, a secondary phosphorus limitation. The various other sources of phosphorus loading insure that large-scale phosphorus removal at wastewater treatment plants will probably not lead to an immediate de crease of the algal biomass in the lakes when it is the only measure taken. Study of local situations and the execution of phosphorus reduction measures on a smaller scale may be more appropriate and effective. Therefore in the future the main effort with respect to phosphorus reduction should be concentrated on the Reeuwijk Lakes. -150-

Effects of Phosphorus Removal in Dutch Lakes

5. Acknowledgements

The authors are greatly indebted to Prof. Dr. M. DONZB, Ir. M. A. HEINSDIJK and Prof. Dr. F. BIANCHI for critical reading of the manuscript, to Mr. P. VAN DOORN for drawing Figs. 1,4,7, 10 and 11, to Mr. B. DE GROOT (Delft Hydraulics Laboratory) for plotting the remaining figures, to PAUL GUTTERIDGE and WIES VAN DEN BRIEL for correcting and Mrs. C. VAN DIJK for typing the English text and to the director of the Rijnland Water Authority for permission to publish these results.

6. References

BOLTER, G., A. N. VAN BREEMEN, and G. VISSER, 1981: Eutrophication tests: a possibility to esti mate the influence of sewage discharge on the biological quality to receiving water.—H-.0 14: 88—92 (in Dutch with an English summary). BJÖRK, S., 1978: Restoration of degraded lake ecosystems.—Regional Workshop Land Use Im pacts on Lake and Reservoir Ecosystems. Warsaw, Poland. BRUNING, C, and S. P. KLAPWIJK, 1984: Application of derivative spectroscopy in bioassays estimating algal available phosphate in lake sediments.—Verh. Internat. Verein. Lininol. 22: 172-178. EPA (Environmental Protection Agency), 1971: Algal Assay Procedure.—National Eutrophica tion Research Program, Corvallis, Oregon. FORSBERG, C, S. RYDING, A. CLAESSON and A. FORSBERG, 1978: Water chemical analyses and/or algal assay?—Sewage effluent and polluted lake water studies.—Mitt. Internat. Verein. Lininol. 21: 352-363. FRUH, E. G., 1967: The overall picture of eutrophication.-J. Water Poll. Control Fed. 39: I 449- 1463. HAXNA, M., and A. DAUTA, 1983: Bioassays: a comparative study of three parameters related to phosphorus bioavailability (yield, growth rate and intracellular concentration of phosphorus).— Annls. Limnol. 19: 59-66. KLAPWIJK, S. P., 1981: Limnological research on the effects of phosphate removal in Rijnland. —H20 14: 742—483 (in Dutch with an English summary). KLAPWIJK, S. P., J. M. W. KROON and M.-L. MEIJER, 1982: Available phosphorus in lake sedi ments in The Netherlands.—Hydrobiologia 92: 491-500. KLAPWIJK, S. P., and C. BRUNING, 1986: Available phosphorus in the sediments of eight lakes in the Netherlands. In: P. G. SLY (ed.), Sediments and Water Interactions: 389—396. Springer Verlag, New York. LEENTVAAR, P., 1980: Comparison of hypertrophy on a seasonal scale in Dutch inland waters.— In: J. BARICA and L. R. MUR (eds), Hypertrophic Ecosystems = Developments in Hvdrobio- logy Vol. 2: 45—55. Junk, The Hague. Ministry for Transport and Public Work, 1980: Water Action Programme 1980-1984. The Nether lands. Nordforsk, 1973: Algal assays in water pollution research.—Proceedings from a Nordic Sympo sium, Oct. 1972, Oslo. Nordforsk secretariat of Environmental Sciences Publ., 1973, 2. SCHMIDT-VAN DORP, A. D., 1978: Eutrophication of Shallow Lakes in Rijnland.—Report Technical Service, Waterboard of Rijnland, Leiden (in Dutch with an English summary). SOKAL, R. R., and F. J. ROHLP, 1969: Biometry—the Principles and Practice of Statistics in Biological Research.—W. H. Freeman, San Francisco. Waterboard of Rijnland, 1984: Final Report on the Effects of Phosphorus Removal in the South eastern Part of Rijnland.—Report Technical Service, Waterboard of Rijnland (in Dutch).

J. VAN der DOES S. P. KLAPWIJK Hoogheemraadschap van Rijnland Technische Dienst Postbus 156 2300 AD Leiden, The Netherlands Manuscript accepted: January 31st, 1986 -151-

CHAPTER 12: ALGAL GROWTH POTENTIAL TESTS AND LIMITING NUTRIENTS IN THE RIJNLAND WATERBOARD AREA (THE NETHERLANDS)

S.P. Klapwijk, G. Bolier & J. van der Does. Prepared for: Internat. Conf. on Environmental Bioassays, Univ. of Lan caster, England, 11-14 July 1988.

"Thus, algal assays are of great value for explaining situations which cannot be indicated by chemical analysis. On the other hand, many results presented today by algal assays can equally well be calculated by water chemical data."

C. Forsberg, S-O Ryding, A. Claesson & A. Forsberg, 1978. Water chemical analyses and/or algal assay? Sewage effluent and polluted lake water studies. Mitt, internat. Vérein. Limnol. 21, p. 361. -152-

ALGAL GROWTH POTENTIAL TESTS AND LIMITING NUTRIENTS IN THE RIJNLAND WATERBOARD AREA (THE NETHERLANDS).

S.P. Klapwijk*, G. Bolier** and J. van der Does*.

Key words': bioassays, algal growth potential (AGP), limiting nutrient factor, nitrogen, phosphorus, eutrophication.

ABSTRACT Four hundred forty bioassays with Scenedesmus quadricauda (Turp.) Bréb. as testorganism have been carried out with samples from canals and lakes in the western part of the Netherlands. The results are used to assess algal growth potential (AGP) and to deter mine the limiting nutrient(s) for maximum biomass production. Special attention has been paid to the effects of deep-freezing and autoclaving as pretreatment of water samples on pH and nutri ent concentrations. pH and particulate P increased, while N-Kjeldahl, NH4-N, particulate N, total-N, ortho-P and total-P decreased signifi cantly by the pretreatment. The algal growth potential of the various samples ranged from very low in the relatively isolated polder lakes to very high in canals and lakes, which form part of the basin system of Rijnland. From the bioassays with nitrogen and phosphorus enrichments it is concluded that the lowest yields are generally observed in the nitrogen and phosphorus co-limited waters, while the highest yields are found in waters limited by nitrogen alone. Very high correlation coefficients have been found between algal growth potential and inorganic and total nitrogen concentrations (r= 0.91 and r=0.85 resp.; n=440). AGP is primarily determined by the amount of nitrogen, especially nitrate, in the samples and only secon darily by the amount of phosphorus. From a graphical presentation of the yields with different limiting nutrients against the nitrogen/phosphorus ratios in the samples criti cal ranges for nitrogen and/or phosphorus limitation could be derived. The ranges indicating phoshorus limitation, >50 and >30 respectively for inorganic and total N/P ratios, lie considerably higher than the ranges reported for P-limitation in the literature so far. This means that phosphorus limitation in these waters can only be achieved at relatively high N/P ratios and low phosphorus concentrations. It is concluded that, once the relations between AGP and nutri ent concentrations are established, as in this study, AGP tests do not have to be carried out on a routine basis to monitor algal growth potential of surface waters. They still can be very useful in special studies, e.g. in lake restoration projects and to detect possible toxi cants in effluents and wastewater discharges.

* Waterboard of Rijnland, Technical Service, P.O. Box 156, 2300 AD Leiden, The Netherland.

** Delft University of Technology, Department of Civil Engi neering, P.O. Box 5048, 2600 GA Delft, The Netherlands. -153-

INTRODUCTION Algal growth potential tests (AGP-tests) have been applied in limnology since the sixties (Skulberg, 1964; Bringmann & Kühn, 1965; E.P.A., 1971). Now they are used over the whole world (Nordforsk, 1973; Gargas & Pedersen, 1974; Chiaudani & Vighi, 1974, 1975, 1976; Miller et al., 1974, 1978; E.P.A., 1975; S.I.L., 1978; Marvan et al., 1979; Youngman, 1980; Ryding, 1980; Raschke & Schulz, 1987). In they Netherlands AGP-tests have been introduced in water quality studies in 1981 for testing surface waters and effluents of sewage treatment plants (Bolier et al. , 1981; Bolier & van Breemen, 1982). They are applied in special water quality studies, e.g. to in vestigate the effects of phosphorus removal at sewage treatment plants (Klapwijk, 1981; Hoogheemraadschap van Rijnland, 1984; van der Does & Klapwijk, 1985, 1987), but also to measure the growth potential as a general water quality parameter and to determine the limiting nu trient factors (Veenstra, 1981; de Haan et al., 1982; de Vries & Klap wijk, 1987). Recently, Bolier and coworkers standardized an AGP me thod for Dutch conditions (Anonymus, 1986). The application of AGP-tests is not without problems and ques tions. Here we evaluate the results of a large number of AGP-tests with the following aims:

1. To check on a large scale what the effects are of the pretreatment applied to the water sample and how this affected the re sults of the bioassays. 2. To elaborate advantages of AGP-tests compared to chemical ana lyses . 3. To correlate AGP results with nutrient and chlorophyll-a concen trations. 4. To evaluate in what way the prediction of limiting factors, based on AGP-tests with additions of nitrogen and/of phosphorus, is consistent with data on nutrient concentrations and ratios.

Results of 440 AGP-tests carried out on 24 sampling locations in the central western part of the Netherlands over a time period of four years (1983-1986) have been used in this study.

MATERIALS AND METHODS

Water samples were collected at 0.5 m depth on 24 sampling loca tions with different trophic degree in the Rijnland Waterboard area (Fig. 1), generally four times per year in March, June, September and December from 1983-1986. At two sampling locations in the Nieuw koop lakes (94.11 and 94.12) and two in the Langeraar lakes (95.04 and 95.18) samples were collected six or seven times per year in 1985 and 1986. At two locations in the Reeuwijk lakes (134.08 and 134.09) sampling took place every four weeks from 1983-1986. General charac teristics of the sampling locations are given in Table 1. The first fif teen sites represent canals and lakes in the basin system of Rijnland, an interconnected system of canals, lakes and ditches with a constant water level of 0.6 m below mean sea level. The last nine locations are situated in polder lakes at about 2 m below mean sea level. They are separated from the basin system but also dependent on it for their watersupply, except for sampling location 18.03, which depends only on precipitation. -154-

Fig. 1. The area of the Rijnland Waterboard and the position of the sampled locations (location numbers conform the Waterboard scheme).

Nitrate + nitrite nitrogen (N03+N02-N), ammonium nitrogen (NH4-N) and orthophosphate (o-P) were measured directly after samp ling on a Technicon AA II autoanalyser. Kjeldahl-nitrogen (Kj-N) and total phosphorus (t-P) were measured after digestion with respectively sulphuric acid + Wieninger selenium mixture and sulphuric acid + per sulphate. Subsequently NH4-N and o-P were determined. Total nitro gen (t-N) was calculated by addition of Kj-N and N02+N03-N. Dissol ved inorganic nitrogen (inorganic-N) was calculated by addition of NH4-N and N02+N03-N. Particulate nitrogen (part-N) and phosphorus (part-P) concentrations were calculated by subtraction of respectively Kj-N minus NH4-N and t-P minus o-P. Chlorophyll-a was spectrophotometrically measured after extraction with 80% ethanol (75°C). Algal growth potential tests were carried out to determine the availability of nutrients for algal growth and to identify the growth limiting nutrients. The principle of such assays is based on von Lie- big's Law, which states that the maximum yield of a species (or popu lation) is determined by the nutrient which is present in the least available quantity. Addition of this nutrient will raise the yield. With such bioassays the algal growth potential and the limiting nutrient -155-

Table 1. General characteristics of the sampled locations.

Loca Name: Type: Mean Surface tion depth: nr. (m) (km2)

1 Noorder Buiten Spaarne canal 3.9 21A Canal at Halfweg pumping station canal 3.0 32 Canal Haarlemmermeerpolder canal 3.0 37 Canal at Katwijk pumping station canal 6.0 58 Kaag lakes lake 3.0 3.2 77 Rijn-Schie canal canal 3.6 116 Canal at Gouda pumping station canal 3.4 272 Lake Braassem lake 3.5 4.6 275 Lake Nieuwe Meer lake 5.0 1.4 284 Lake Westeinder lake 2.9 8.9 296 Lake Oosterduin deep sand pit 12.0 0.1 299 Lake Vlietlanden deep sand pit 34.5 0.6 375 Oude Rijn canal canal 2.9 379 Lake Zeegerplas deep sand pit 30.5 0.4 391 Lake Mooie Nel lake 3.4 0.1 18.03 Lake Broekvelden/Vettenbroek deep sand pit 27.0 0.9 94.11 Nieuwkoop lakes (south) peat lake 2.7 1.0 94.12 Nieuwkoop lakes (north) peat lake 2.5 1.5 95.04 Langeraar lakes (north) peat lake 2.0 0.8 95.18 Langeraar lakes (Geerplas) peat lake 2.5 0.2 134.08 Reeuwijk lakes (Elfhoeven) peat lake 2.4 1.1 134.09 Reeuwijk lakes (Nieuwenbroek) peat lake 1.7 1.1 217.06 Lake Amstelveense Poel peat lake 1.5 0.6 403.02 Lake Sloterplas deep sand pit 25.0 0.8

(N, P or N+P) can be identified, (Thomas, 1953; E.P.A., 1971; Chiau- dani & Vighi, 1974, 1975; Forsberg et al., 1978; Rodhe, 1978). AGP- tests do not provide information on light as possible growth limiting factor, since the tests are run in the laboratory with an excess of light. All water samples except those from the sampling locations in the Nieuwkoop, Langeraar and Reeuwijk lakes, were preserved by deep-freezing untill the bioassays were conducted. The samples were thawed in a waterbath at 45°C and poured into 150 ml Erlenmeyer flasks filled up to 75 ml during 1983 and 1984. In 1985 and 1986 a larger volume of 100 ml was poured into 500 ml Erlenmeyer flasks to improve the C02-supply (Anonymus, 1986). After autoclaving the flasks to kill the natural phytoplankton population and to liberate nu- .-156-

trients (Miller et al., 1978), some flasks were enriched with 10 mg 1 1 1 N l" (as KNO,) or 0.5 mg P f (as K2HP04/KH2P04) or 10 mg N l" + 0.5 mg P l-*. Due to the chemical form of the nitrogen and phos phorus additions, also an application of respectively 27.9 or 1.23 mg K l"1 has been added simultaneously. After autoclaving the water was analyzed for pH, Kj-N, NH4-N, N03+N02-N, o-P and t-P in order to measure possible side-effects of the pretreatment. The experiments were run in duplicate (1983, 1984) or triplicate (1985, 1986). Two or three Erlenmeyers without addition served as controls. All Erlenmeyers were inoculated with approximately 10,000 cells per ml of Scenedesmus quadricauda (Turp.) Bréb., from a batch cul ture derived from the Limnological Institute, Nieuwersluis; which were nitrogen and phosphorus starved for about 12 days. The culture ves sels, closed with cotton plugs, were placed on a shaking table (120- 150 rpm) at a temperature of 20 ± 1°C, at a permanent light intensity of 70-100 (JE m"2 s"1. The optical density (OD) at 680, corrected with the OD at 750 nm, was measured on a Philips/Pye Unicam PU 8800/2 Spectrophotometer as biomass parameter. Maximal biomass was calculated from three measurements during the stationary growth phase. When the growth was two times less than 5% per day these last and the preceding data were used to calculate maximal biomass after rejection of outliers (E.P.A., 1971). A scheme of the assay technique is given in de Vries and Klapwijk (1987). - The difference in mean maximal yield (of duplo or triplo experi ments) with and without nutrient enrichment was tested for signifi- cancy applying the t-test for means of two samples (Sokal & Rohlf, 1969). Differences between pH and nutrients in the original water samples and after deep-freezing and autoclaving were tested statisti cally with a paired Student's t-test. Product-moment correlation be tween yields and nutrient concentrations were also calculated accor ding to Sokal and Rohlf (1969), while a multiple regression analysis was carried out according to King (1969).

RESULTS In Table 2 the results of the pH measurements and chemical ana lyses in the original water samples are summarized. The complete data set is available in a LOTUS 123 spreadsheet. A great variety in nu trient concentrations exist between the different sampling locations. Average inorganic-N and t-N concentrations for instance are ranging from respectively 0.2 and 2.1 mg N 1 1 at sampling location 134.08 (Reeuwijk lakes; Elf hoeven) to 5.8 and 8.1 mg N 1 1 at sampling loca tion 391 (Lake Mooie Nel). Average o-P _and t-P concentrations are ranging from appr. 0.01 and 0.04 mg P 1 1 in the deep lake Broek- velden/Vettenbroek (loc. nr. 18.03) to 2.08 and 2.25 mg P f1 in the Noorder Buiten Spaarne (loc.nr. 1). This variety is dependent on the degree of isolation of the various lakes from the basin system of Rijn land and on the local degree of pollution, especially from discharges of effluents of sewage treatment plants. Chlorophyll-a is highest in the Nieuwkoop (loc. nrs. 94.11, 94.12), Langeraar (loc. nrs. 95.04, 95.18) and Reeuwijk lakes (loc. nrs 134.08, 134.09) and in lake Amstelveense Poel (loc. nr. 217.06), which are all shallow peat lakes, relatively isolated from Rijnland's basin system and having rather long residence times (mostly >1 year). -157-

Table 2. Averaged values (± s.d.) over a four years period (1983- 1986) of pH, nutrients and chlorophyll-a concentrations at different sampling locations (n = number of observations).

Loca pH Inorg-N Total-N Ortho-P Total-P Chl-a n tion nr. (-) (mg f1) (mg 1"1) (mg f') (mg f') (mg m"3;1

1 7.9 ± 0.2 6.9 ± 2.4 9.7 ± 2.6 2.08 + 0.26 2.25 ± 0.27 97 ± 85 8 21A 7.8 ± 0.2 4.6 ± 2.0 8.4 + 5.5 0.51 ± 0.12 0.83 ± 0.14 40 + 22 8 32 8.0 ± 0.2 3.2 ± 1.8 5.3 + 1.9 0.54 ± 0.13 0.73 ± 0.15 40 + 31 15 37 7.8 ± 0.2 5.7 ± 2.2 7.7 + 2.3 1.08 ± 0.37 1.23 ± 0.40 30 + 24 15 58 8.2 ± 0.3 2.9 ± 1.8 5.0 + 1.8 0.66 ± 0.55 0.77 ± 0.15 34 ± 32 16 77 7.7 ± 0.2 5.1 ± 1.2 7.2 + 2.0 0.54 ± 0.27 0.80 ± 0.18 30 ± 33 8 116 7.7 ± 0.2 6.0 ± 1.9 7.9 + 2.2 0.60 ± 0.19 1.15 ± 0.74 27 ± 22 16 272 8.3 ± 0.4 2.5 + 1.4 4.3 + 1.5 0.43 ± 0.11 0.52 ± 0.08 40 ± 46 16 275 8.0 ± 0.3 2.8 ± 0.8 4.4 + 1.0 0.42 ± 0.06 0.48 ± 0.08 22 ± 24 16 284 8.5 ± 0.3 0.9 + 0.8 2.8 + 0.8 0.18 ± 0.05 0.31 ± 0.06 52 + 36 16 296 8.4 ± 0.3 0.6 ± 0.4 2.2 + 0.3 0.69 ± 0.21 0.79 ± 0.18 26 ± 28 16 299 8.0 ± 0.2 5.4 + 1.5 7.2 + 1.7 0.22 ± 0.10 0.28 ± 0.11 9 + 8 16 375 7.6 ± 0.2 4.1 + 1.6 6.2 + 1.9 0.69 ± 0.25 0.92 ± 0.21 39 + 30 16 379 8.1 + 0.5 5.1 ± 1.3 6.8 + 1.4 0.91 ± 0.24 0.98 + 0.23 30 ± 37 16 391 8.1 ± 0.2 5.8 ± 2.2 8.1 + 2.2 1.81 ± 0.31 1.92 + 0.30 58 ± 54 16 18.03 8.1 ± 0.2 1.1 ± 0.6 2.0 + 0.6 0.01 ± 0.01 0.04 ± 0.01 3 + 2 16 94.11 8.5 ± 0.6 0.2 ± 0.2 3.4 + 0.6 0.03 + 0.03 0.26 ± 0.10 223 + 123 21 94.12 8.3 ± 0.5 0.2 ± 0.1 2.6 + 0.3 0.01 ± 0.01 0.08 ± 0.02 102 + 34 21 95.04 8.7 + 0.4 0.3 ± 0.3 3.6 + 0.7 0.12 ± 0.11 0.37 ± 0.11 177 ± 76 21 95.18 8.7 ± 0.4 0.2 + 0.1 3.6 + 1.0 0.23" ± 0.13 0.47 ± 0.12 199 + 60 12 134.08 8.1 ± 0.4 0.2 ± 0.1 2.1 + 0.5 0.02 ± 0.02 0.14 ± 0.07 75 + 45 51 134.09 8.2 ± 0.4 0.3 ±0.3 3.0 + 0.4 0.01 + 0.01 0.09 + 0.03 98 + 37 51 217.06 8.7 ± 0.5 0.5 ± 0.8 4.3 + 1.1 0.09 ± 0.10 0.29 ± 0.07 220 + 115 16 403.02 8.4 ± 0.4 1.1 + 0.9 2.2 + 0.9 1.23 + 0.41 1.31 ± 0.36 27 ± 29 16

Comparison of the analyses in the original water samples with the results after deep-freezing and autoclaving shows that at most sampling stations the pH has increased and that the t-N, o-P and t-P concentrations have decreased. This is tested over all samples with a paired Student's t-test (Table 3). Also Kj-N, NH4-N and part-N de creased significantly, while part-P increased. N02+N03-N and inorganic-N did not change significantly, nor did the inorganic N/P and the total N/P ratio, although the mean inorganic-N/P ratio has about hal ved. Obviously by the pretreatment pH is increased due to the loss of C02 and about 4% total nitrogen and 16% total phosphorus is has vanished. -158-

Table 3. Results of t-tests (means ± s.d.) between pH and nutrient concentration before and after deep-freezing and autoclaving.

before after diffe pretreatment pretreatment rence n r t P

PH ( -) 8.23 ± 0.48 8.72 + 0.42 0.49 313 0.15 14.75 <0.001 N-Kj (mg l"1) 2.94 ± 1.70 2.74 + 1.47 -0.20 416 0.95 -7.64 <0.001 1 NH4 0.68.1 1.20 0.62 ± 1.02 -0.06 .418 0.95 -3.14 0.002

1 N02+N03-N 1.47 + 1.73 1.49 + 1.74 0.02 418 0.98 1.42 0.157 ns Inorg-N t 2.14 ± 2.51 2.10 + 2.36 -0.04 418 0.98 -1.57 0.117 ns Part-N ! 2.26 ± 1.10 2.12 + 0.89 -0.14 416 0.88 -5.30 <0.001 Total-N ' 4.41 ± 2.67 4.23 + 2.45 -0.18 416 0.98 -6.08 <0.001 Ortho-P ' 0.42 + 0.53 0.18 + 0.20 -0.24 417 0.83 -13.13 <0.001 Part-P 1 0.15 + 0.19 0.30 + 0.32 0.15 417 0.16 8.75 <0.001 Total-P 1 0.57 ± 0.56 0.48 + 0.47 -0.09 417 0.91 -8.13 <0.001 Inorg-N/P ( -) 38.3 ± 267.2 19.7 + 40.5 -18.60 417 0.10 -1.42 0.155 ns Total-N/P ( •) 16.2 ± 16.3 17.5 + 19.8 1.30 417 0.68 1.78 0.075 ns

n - number of observations; r = product-moment correlation coefficient; t = Student's t-value; p = two-tailed probability; ns = not significant.

It is not possible to show in detail all the results of 440 bioas says. Fig. 2 shows two typical examples of frequently occurring re sults . In both examples the addition of N resulted in a higher yield than in the control (without additions), while the addition of P did not. In the example of Fig. 2A addition of N plus P resulted in an even higher yield than the addition of N alone. In the example of Fig. 2B the combined addition (N+P) resulted in about the same yield as the N addition alone. In both bioassays the algal growth potential (=yield without addition) is primarily N-limited, but the yield with N-addition in Fig. 2B is not limited by P but by another factor, while the yield with N-addition in Fig. 2A is presumably limited by P. Therefore the yield in the example of Fig. 2A is called to be secon darily limited also by P (van der Does & Klapwijk, 1987; de Vries & Klapwijk, 1987; Klapwijk et al., unpublished). Table 4 shows the mean values (± s.d.) of the bioassay results from the different sampling locations ranked in increasing order of yield. The mean algal growth potentials (controls) range from 0.0015 to 0.046 (O.D. 680-750 nm); they are lowest in the polder lakes. From the bioassays with nitrogen and phosphorus enrichments can be seen that in most lakes nitrogen clearly stimulated the algal yield. Only in lake Broekvelden/Vettenbroek (loc. nr. 18.03) the yield is increased by addition of phosphorus alone. In several other lakes, e.g. location nrs. 299, 94.11, 94.12, 95.04, 134.08, 134.09 and 217.06, the yield is stimulated by the addition of nitrogen, but even more by the enrichment with N+P, indicating that the algal growth in these -159-

SITE NR. 134.08 SITE NR. 272 120 110 A.

o" 100 o ° 90

*E 80 o 70 m 7 60 o 8 50 g 40 o 30 _i SÜ 20 10 0. l C +N +P +N+P +N +P +N+P NUTRIENT ADDITIONS NUTRIENT ADDITIONS

Fig. 2. Two examples of yields (± s.d.) in bioassay experiments with nitrogen and phosphorus enrichments carried out with water from the Reeuwijk lakes (Fig. 2A; site nr. 134.08; date 860624) and lake Braassem (Fig. 2B; site nr. 272, date 840306).

C control (no enrichment); +N with nitrogen addition; +P with phosphorus addition; +N+P with nitrogen and phosphorus addition.

lakes is limited primarily by nitrogen but secondarily also by phos phorus. At the other sampling locations, including all the canals, the enrichment with N+P did not generally produce a significant higher yield than the enrichment with N alone. The algal growth therefore is limited by nitrogen only. At these sampling stations phosphorus is obviously present in excess. By the ranking of the bioassay results in increasing order of yield it can be seen that the bioassays with the lowest yields are mostly phosphorus or nitrogen and phosphorus limited, while the highest algal yields are generally limited exclusively by nitrogen. In order to see if and to which degree the bioassay results are related with nutrient and chlorophyll-a concentrations and nutrient ratios in the original water samples correlation coefficients have been calculated (Table 5). Very high correlations have been found between the bioassays without enrichments (controls) and inorganic-N and t-N (resp. 0.91 and 0.85; n= 440). From this may be concluded that the algal growth potential in these waters is predominantly determined by the amount of nitrogen in the samples. -160-

Table 4. Mean values (± s.d.) as optical density (680-750 nm) x 1000 of bioassay results from the different sampling locations ranked from low to high yield.

Loca CONTROL ttflTH ADDITIONS OF GENERALLY tion LIMITING nr. (n) NITROGEN (n) PHOS (n) N + P Cn) NUTRIENT(S) PHORUS

134.08 1.5 + 1.7 (51) 19 + 12 (50) 1.5 + 1.9 (50) 65 + 21- (47) N,p 94.11 2.2 + 1.6 (21) 32 + 12 C 8) 1.7 + 1.1 ( 8) 54 + 11 ( 8) N,p 94.12 2.3 + 2.2 (21) 16 + 5.9 ( 8) 2.2 + 0.8 ( 8) 48 + 13 ( 8) N,p 134.09 2.5 + 2.0 (51) 8.9 + 5.6 [50) 2.2 + 1.9 (50) 62 + 25 (48) N,p 95.18 2.6 + 1.5 (12) 95.04 3.1 + 2.6 (21) 50 + 18 ( 8) 3.8 + 2.6 ( 8) 70 + 8.5 ( 8) N,p 18.03 4.3 + 3.1 (16) 5.2 + 3.9 ( 8) 8.8 + 3.8 ( 8) 36 + 14 , 8) P,n 217.06 4.3 + 2.8 (16) 46 + 13 C 8) 2.5 + 1.6 ( 8) 56 + 16 8) N,p 296 5.3 + 3.0 (16) 88 + 29 ; 8) 6.1 + 2.8 ( 8) 86 + 11 , 8) N 284 6.9 + 5.3 (16) 46 + 23 k 8) 7.5 + 5.5 ( 8) 46 + 24 8) N 403.02 7.0 + 6.4 (16) 68 + 18 8) 10 + 8.1 ( 8) 68 + 16 8) N 272 19 + 11 (16) 85 + 20 8) 23 + 12 ( 8) 96 + 26 8) N 275 20 + 6.5 (16) 69 + 15 8) 19 + 4.3 ( 8) 65 + 22 8) N 58 21 + 13 (16) 101 + 23 8) 26 + 15 ( 8) 111 + 29 8) N 299 25 + 13 (16) 53 + 22 8) 33 + 7.0 ( 8) 83 + 29 8) N,p or P,n 32 27 + 18 (15) 113 + 12 7) 34 + 20 ( 7) 119 + 19 7) N 77 32 + 10 ( 8) 375 32 + 15 (16) 138 + 14 8) 33 + 14 ( 8) 134 + 18 ( 8) N 21A 33 + 17 ( 8) 379 39 + 13 (16) 98 + 25 ( 8) 37 + 7.9 ( 8) 104 + 21.' ( 8) N 379 40 + 15 (16) 106 + 23 ( 8) 41 + 13 ( 8) 100 + 17 ( 8) N 37 41 + 17 (16) 125 + 16 ( 8) 43 + 18 ( 8) 130 + 24 ( 8) N 1 46 + 15 ( 8) 116 46 + 15 (16) 147 + 33 ( 8) 47 + 13 ( 8) 155 + 35 ( 8) N

(n) = number of observations; N,p = primarily nitrogen, secondarily phosphorus limited; P,n = primarily phosphorus, secondarily nitrogen limited; N = nitrogen limited.

Table 5. Correlations coefficient between AGP results and chlorophyll- a with nutrient concentrations and ratios in the original wa ter samples.

inorg- t-N o-P t-P inorg- total (n) N N/P N/P

Control (no addition) 0.91* 0.85* 0.63* 0.65* -0.07ns -0.31 440

Nitrogen addition 0.73* 0.70* 0.69* 0.71* -0.23* -0.60* 243

Phosphorus addition 0.98* 0.92* 0.63* 0.63* -0.12ns -0.33* 243

N + P addition 0.64* . 0.65* 0.45* 0.52* -0.05ns -0.40* 238

Chlorophyll-a -0.38* -0.11* -0.29* -0.20* -0.04ns 0.01ns 434

* - P <0.001; ns = not significant; (n) = number of paired observations. -161-

Phosphorus is only a secondarily limiting factor as can be con cluded from the even higher correlation between the yields in the bioassays with P-enrichment and inorganic-N and t-N (resp. 0.98 and 0.92; n=243). This can be illustrated with Fig. 3, in which the yields without enrichments and with P-addition are plotted against inorganic- N. Fig. 3A shows a positive correlation between yield measured as op tical density and inorganic nitrogen of the location water, with a rela tively high variability between 3 and 10 mg inorg-N l1. Addition of P (Fig. 3B) diminishes the variability probably by removing the.P-limitation. Applying multiple regression analysis to "explain" the algal growth potential in the bioassays (controls) by the inorganic-N and o-P concentrations gives a highly significant multiple correlation coef ficient of R=0.914 (n=439; P<0.001) with the following regression for mula:

Y = 5.43 XI + 6.50 X2 + 3.06 X3 + 0.72 where: Y = algal growth potential (as O.D. 680-750 nm * 1000), J XI = concentration of NH4 (in mg N 1 ), 1 X2 = concentration of N02+N03 (in mg N l" ), X3 = concentration of o-P (in mg P 1 1).

INORGANIC-N versus YIELD , INORGANIC-N versus YIELD (+P-ADDITION) (1983-1986) (1983-1986)

INORGANIC-N [mg/l] INORGANIC-N [mg/l]

Fig. 3. Inorganic nitrogen versus yield without additions (A) and with phosphorus additions (B) in bioassays during 1983-1986.

According to the standard partial regression coefficients (not shown here) N02+N03-N is explaining most of the total variance in the algal growth potential, followed by NH4-N and o-P. With this formula the AGP can be calculated merely on the base of the inorganic nutrient concentrations with great precision. Not only the AGP but also the yields with N- and P-additions can be predicted reasonably well (N- addition) or extremely well (P-addition) from the nutrient concentra tions . -162-

Also notice in Table 5 that the chlorophyll-a concentrations in the field samples are inversely related with inorganic-N, t-N, o-P and t-P. Obviously the actual algal biomass in the sampled canals and lakes is generally more dependent upon other factors, such as turbi dity, residence time and zooplankton grazing, than upon nutrient con centrations .

INORGANIC N/P versus YIELD (1983-1986)

»ï ^ 240\ »->810 + >450 ■**,*?+ ** +•* **■ |»>- r^->5oo

40 60

INORGANIC-N/ORTHO-P RATIO

TOTAL N/P versus YIELD (1983-1966) 90

GO O tl

40 - J, 30 - 1 " + t 20 - i 0"n

0 |-> 76 -t;>11 2 -10 1 0 60 TOT*L-N/P RATIO N-limited N,p—limited P,n limited

Fig. 4. Inorganic (A) and total (B) nitrogen/phosphorus ratios ver sus yield (without addition) in 243 bioassays during 1983- 1986. The limiting nutrient factor(s) are indicated in the figure as follows: N = nitrogen limited; N,p = primarily nitrogen, secondarily phosphorus limited; P,n = primarily phosphorus, secondarily nitrogen limited. -163-

The negative correlations with N/P ratios, especially the nega tive correlation between the N-enriched yields with total N/P ratio, indicate that lower growth potentials are found at higher N/P ratios, which mostly implicates low P-concentrations. The correlation coeffi cients with the nutrient ratios are not very high due to the fact that the relations are not linear at all. In Fig. 4 the yields (without en richments) are plotted against inorganic and total N/P ratios. The factors by which the yield was limited in the different bioassays are indicated too in this figure. From these figures critical ranges for nitrogen and/or phosphorus limitation are derived (Table 6). On the basis of these critical nutrient ratios the limiting nutrient factors (N or primarily N, secondarily P or primarily P, secondarily N) for algal growth can roughly be assessed.

Table 6. Critical ranges of N/P ratios indicating different nutrient limitations according to this study.

Limitation: Inorg- N/P ratio Total N/P ratio

Nitrogen 0-20 0-20 Primarily nitrogen, sec. phosphorus 10-70 10-50 Primarily phosphorus, sec. nitrogen >50 >30

DISCUSSION AND CONCLUSIONS The measurements before and after deep-freezing and autoclaving show clearly that the chemical composition of the water samples is al tered by the pretreatment. This is consistent with the findings of Filip and Middlebrooks (1975), Eloranta and Laitinen (1981), de Vries and Ouboter (1985) and Anonymus (1986). A decrease of t-P by 16% in autoclaved samples agrees with the results of Filip and Middle- brooks (1975), Eloranta and Laitinen (1981) and de Vries and Ouboter (1985). Obviously, the speciation of phosphorus, which can be very important for the availability for algae, is altered by pretreatment due amongst others to the pH changes. Whether this affects the bioassay results seriously is not certain. The decrease in especially the o-P concentration is presumably reversible at least for a part of it, when the pH is decreased again during the bioassays. We found with sam ples from lake Mooie Nel (loc. nr. 391) a reversible reaction at pH levels <8.5 in the original samples and irreversible reactions at pH levels >8.5 (unpublished results). Anyhow, it is obvious that the pretreatment by autoclaving does not solubilize all nutrients, as is sometimes suggested (Miller et al., 1974, 1978). The results have shown that AGP is a good parameter for the potential algal growth in surface waters. It is not directly related with field measurements of chlorophyll-a. On the contrary, AGP yields are inversely related to actual chlorophyll measurements in the canals and lakes (r=-0.33; n=437). Both parameters should therefore be con sidered as supplementary. -164-

With the nitrate enrichments in the form of KN03 also an amount of 27.9 mg K l"1 was added to the test waters, whereas only 1.23 mg K l"1 was added with the phosphorus additions. The question there fore can be raised of the growth stimulation observed by the nitrate additions are really due to the N-addition and not to the addition of potassium. Because of the high concentration of potassium in het surface waters of the Rijnland Waterboard area (approtimately 30 mg K f1) and the relative low K concentrations needed for optimal growth (1.1 - 2.3 mg K l'1) of algal species (Leenvaar, 1980), it is verly unlikely that growth stimulation caused by K-addition occurs in these waters. Moreover, in two bioassay experiments with water from the Reeuwijk lakes in June, 1983 we found no difference in the maximal algal growth after addition of 10 mg N as KNOs or NH4C1 (unpublished results).

The ranges found in this study for N/P ratios indicating a pri mary phosphorus limitation, >50 and >30 for respectively inorganic and total N/P ratios, he considerably higher than the ranges reported for P-limitation in the literature so far (cf. Forsberg et al., 1978; Schmidt-van Dorp, 1978; Claesson & Forsberg, 1980; Healey & Hend- zel, 1980; de Vries & Klapwijk, 1987). This means that in the Rijn land surface waters phosphorus limitation can only be achieved at re latively high N/P ratios and low P-concentrations. Correlation and regression analyses with AGP and nutrients, measured in the original water samples, prove that the algal growth potential measured in bioassays can be calculated with high precision on the ground of only nutrient concentrations. In our opinion this means that:

1) toxins against algae do not play an important role in the Rijnland surface waters, at least not after the applied pretreatment of deep-freezing and autoclaving;

2) once the relations between AGP and nutrient concentrations are established, as is the case here, AGP test do not have to be carried out on a routine basis to monitor the algal growth poten tial of surface waters. See also Forsberg et al. (1978). They still can be very useful in special studies, e.g. to follow the ef fects of phosphorus reduction programs and in lake restoration studies (Klapwijk, 1981; van der Does & Klapwijk, 1985, 1987). Besides that, they may be useful to detect possible toxicants in effluents and wastewater discharges (Gargas, 1978; Forsberg et al., 1978; Laage, 1978, Bolier et al., 1981; de Vries & Hotting, 1985).

ACKNOWLEDGEMENTS

The authors thank Prof. Dr. M. Donze and Prof. Dr. W.H.O. Ernst for valuable suggestions with respect to the manuscript, Mr. P. Nieuwpoort for compilation of the data and graphical presentations and Miss A. Honnef for correcting and Miss C. van Dijk for typing the English text. -165-

REFERENCES Anonymus, 1986. Ontwikkeling van een algengroeiprotentietoets voor oppervlakte water en afvalwater. Rapport Stichting Toegepast Onderzoek Reiniging Afvalwater, Rijswijk.

Bolier, G., A.N. van Breemen & G. Visser, 1981. Eutrophication tests: a possibility to estimate the influence of sewage discharge on the biological quality of the receiving water. HzO 14: 88-92 (in Dutch with an English summary). Bolier, G. & L.W.C.A. van Breemen (eds.), 1982. Standaardisatie van de algengroeipotentietoets voor Nederland: verkenning en afbakening. Laboratorium voor Gezondheidstech- niek, TH-Delft. Rapport nr. 82-03.

Bringmann, P.G. & R. Kühn, 1965. Nitrat oder Phosphat als Begrenzungsfaktor des Algenwachstums. Gesundheitsing. 7: 210-214.

Chiaudani, G. & M. Vighi, 1974. The N: P ratio and tests with Selenastrum to predict, eutrophica tion in lakes. Water Res. 8: 1063-1069.

Chiaudani, G. & M. Vighi, 1975. Dynamics of nutrient limitation in six small lakes. Verh. internat. Verein. Limnol. 19: 1319-1324.

Chiaudani, G. & M. Vighi, 1976. Comparison of different techniques for detecting limiting or sur plus nitrogen in batch cultures of Selenastrum capricornutum. Water Res. 10: 725-729.

Claesson, A. & A. Forsberg, 1980. Algal assay studies of polluted lakes. Arch. Hydrobiol. 89: 208- 224.

Does, J. van der & S.P. Klapwijk, 1985. Phosphorus removal and effects on waterquality in Rijnland. H20 18: 381-387 (in Dutch with an English summary).

Does, J. van der & S.P. Klapwijk, 1987. Effects of phosphorus removal on the maximal algal growth in bioassay experiments with water from four Dutch lakes. Int. Revue ges. Hydrobiol. 72: 27-39.

Environmental Protection Agency (E.P.A.), 1971. Algal Assay Procedure, National Eutrophication Research Program. Corvallis, Oregon.

Environmental Protection Agency (E.P.A.), 1975. Proceedings: Biostimulation and nutrient assessment. Workshop, Oct. 16-17, 1973. Corvallis, Oregon. EPA-660/3-75-034. -166-

Eloranta, V. & O. Laitinen, 1981. Evaluation of sample preparation for algal assays on waters re ceiving cellulose effluents. Verh. internat. Verein. Limnol. 21: 770-775.

Filip, D.S. & E.J. Middlebrooks, 1975. Evaluation of sample preparation techniques for algal bioassays. Water Res. 9: 581-585. Forsberg, C, S.-O. Ryding, A. Claesson & A. Forsberg, 1978. Water chemical analyses and/or algal assay? - Sewage effluent and polluted lake water studies. Mitt, internat. Verein. Limnol. 21: 352-363.

Gargas, E., 1978. The effect of sewage (mechanically, biologically and chemically treated) on algal growth. Mitt, internat. Ver. Limnol. 21: 110- 124.

Gargas, E. & J.S. Pedersen, 1974. Algal Assay Procedure Batch Technique. Contribution from the Water Quality Institute. Danish Academy of technical Science, 1 (2nd ed.). Haan, H. de, J.B.W. Wanders & J.R. Moed, 1982. Multiple addition bioassay of Tjeukemeer water. Hydrobiologia 88: 233-244.

Healey, F.P. & L.L. Hendzel, 1980. Physiological indicators of nutrient deficiency in lake phytoplankton. Can. J. Fish. Aquat. Sci. 37: 442-453.

King, L.L., 1969. Statistical analysis in geography. 2nd ed. Prentice Hall. Inc., Englewood Cliffs, N.J.

Klapwijk, S.P., 1981. Limnological research on the effects of phosphate removal in Rijn land. H20 14: 472-483 (in Dutch with an English summary). Klapwijk, S.P., U.G. Dijkstra-Stam, A.K. Fleuren-Kemüa & J.C. van der Vlugt, in prep. Comparison of different methods to determine growth limiting fac tors for phytoplankton in the Reeuwijk lakes (The Netherlands). Submitted to Water Research.

Laake, M., 1978. Monitoring the effects of chemical and biological wastewater treat ment in situ by dialysis cultures of freshwater algae. Mitt, inter nat. Verein. Limnol. 21: 453-472. -167-

Leentvaar, P., 1980. Eutrophication, nature management and the role of potassium. Hydrobiol. Buil. 14: 22-29.

Marvan, P., S. Pfibil & Lhotsky, 1979. Algal assays and monitoring eutrophication. E. Schweizerbart'sche Verlagsbuchhandlung, Stuttgart.

Miller, W.E., T.E. Maloney & J.C. Greene, 1974. Algal productivity in 49 lake waters as determined by algal as says. Water Res. 8: 667-679.

Miller, W.E., J.C. Greene & T.Shiroyama, 1978. The Selenastrum capricornutum Printz algal assay bottle test. Environmental Protection Agency (EPA). Corvallis, Oregon.

Nordforsk, 1973. Algal assays in water pollution research. Proceedings from a Nordic symposium, Oct. 1972, Oslo. Nordforsk secretariat of Environmental Sciences Publ., 1973: 2.

Raschke, R.L. & D.A. Schulz, 1978. The use of algal growth potential test for data assessment. J. Wat. Pollut. Contr. Fed. 59: 222-227.

Rodhe, W., 1978. Algae in culture and nature. Mitt, internat. Verein. Limnol. 21: 7-20.

Ryding, S-O. (ed), 1980. Monitoring of inland waters. Report from the working group for eutrophication research. Nordforsk publication 1980: 2, Helsing- fors, Finland.

Schmidt-van Dorp, A.D., 1978. Eutrophication of shallow lakes in Rijnland. Report Technical Service, Hoogheemraadschap van Rijnland, Leiden (in Dutch with an English summary).

S.I.L., 1978. Symposium: Experimental use of algal cultures in limnology. Mitt. internat. Verein. Limnol. 21: 1-607. E. Schweizerbart'sche Ver lagsbuchhandlung, Stuttgart.

Skulberg, O.M., 1964. Algal problems related to the eutrophication of European water supplies, and a bioassay method to assess fertilizing influences of pollution on inland water. In: D.F. Jackson (ed.): Algal and Man, pp. 262-269. Plenum Press, New York.

Sokal, R.R. & F.J. Rohlf, 1969. Biometry. The principles and practice of statistics in biological research. W.H. Freeman & Co, San Francisco. -168-

Thomas, E.A., 1953. Empirische und experimentelle Untersuchungen zur Kenntnis der Minimumstoffen in 46 Seen der Schweiz und angrenzender Gebiete. Schweiz Ver. Gas- Wasserfachm. Monatsbulletin 33: 25-32; 71-79.

Veenstra, S., 1981. De algengroei-potentie-toests. Een nieuw instrument ter beoorde ling van het eutrofiërend karakter van een water. L.H. Wage- ningen, Vakgroep Waterzuivering, sektie Hydrobiologie. Dokto raal verslagen serie nr. 81-1 (in Dutch).

Vries, P.J.R. de & E.J. Hotting, 1985. Bioassays with Stigeoclonium tenue Kütz. on waters receiving sewage effluents. Water Res. 19: 1405-1410.

Vries, P.J.R. de & P.S.H. Ouboter, 1985. Watersample treatments and their effects on bioassays using Sti geoclonium helveticum Vischer. Aquat. Bot. 22: 177-185.

Vries, P.J.R. de & S.P. Klapwijk, 1987. Bioassays using Stigeoclonium tenue Kütz. and Scenedesmus quadricauda (Turp.) Bréb. as testorganisms; a comparative study. Hydrobiologia 153: 149-157.

Youngman, R.E., 1980. The prediction of algal crops by bioassay. Technical Report TR 148, Water Research Centre, Environmental Protection, Medmen- ham, U.K. -169-

CHAPTER 13:

BIOASSAYS USING STIGEOCLONIUM TENUE KÜTZ. AND SCENEDESMUS QUADRICAUDA (TURP.) BREB. AS TESTORGANISMS; A COMPARATIVE STUDY

P.J.R. de Vries & S.P. Klapwijk.

Published in: Hydrobiologia 153: 149-157 (1987).

"Knowledge of variability of the experimental strain and of its reac tions to different cultivation conditions is necessary for a correct in terpretation of the results obtained." J. Komarek & P. Marvan, 1979. Selection and registration of strains of algae as assay organisms. In: P. Marvan, S. Pfibil & O. Lhotsky (eds.): Algal assays and monitoring eutrophication, p. 87. E. Schweizerbart'sche Verlags- buchhandlung, Stuttgart. -170-

Bioassays using Stigeoclonium tenue Kütz. and Scenedesmus quadricauda (Turp.) Bréb. as testorganisms; a comparative study

Pierre J. R. de Vries ' & Sjoerd P. Klapwijk2 1 Biological Laboratory, Vrije Universiteit, De Boelelaan 1087, 1007 MC Amsterdam, The Netherlands 2 Walerboard of Rijnland, Technical Service, Breestraat 59, 2311 CJ Leiden, The Netherlands

Received 4 July 1986; in revised form 2 January 1987; accepted 20 January 1987

Key words: bioassays, algal growth potential, eutrophication, limiting nutrients, Stigeoclonium, Scenedesmus

Abstract

Three different bioassays, two culture tube test methods with respectively Stigeoclonium tenueor Scenedesmus quadricauda and one bottle test with S. quadricauda, were compared. The yields obtained in the various tests were linearily correlated (r = 0.86, P < 0.001). The same primary limiting nutrient was indicated by the bio assays in most cases. However the algal growth in the tube test using Stigeoclonium was more often P- limited. In the case of S. quadricauda both test methods (tube and bottle) were nearly equally effective. The yields of N-limited samples were significantly correlated with the inorganic-N as well as total-N concentra tion of the water sample. A significant correlation of the ortho-P as well as total-P concentration with the yield of the P-limitcd assays was only found for Stigeoclonium tenue. The critical total N/P ratio (by weight) for N or P limitation was approximately 17:1 for Stigeoclonium tenue and 22:1 for Scenedesmus quadricau da.

Introduction Lhotsky, 1979). This alga, however, is not very com mon in The Netherlands. Neither is Scenedesmus Eutrophication of aquatic environments often subspicatus Chod., recommended in the ISO test. results in phytoplankton blooms in lakes, or exten Therefore, Bolier & Van Breemen (1982) proposed sive growth of macrophytes, periphyton and the planktonic green-alga Scenedesmus quadricau blanketing algal mats in smaller water bodies da (Türp.) Bréb., which occurs commonly in shal (Klapwijk, 1978; Barica & Mur, 1980; Hillebrand, low lakes and canals in The Netherlands, as test or 1983; Klapwijk et al. 1983). Algal assays are a suita ganism in an algal assay based on the A. A. P. ble tool for monitoring eutrophication. They are bottle test technique (US Environmental Protection used to determine water fertility and to identify Agency, 1971; Miller et al. 1974). For other typical growth limiting nutrients (Thomas, 1953; US En Dutch aquatic environments, i.e. those with an ex vironmental Protection Agency, 1971; Forsberg et tensive littoral such as polder ditches and canals a al. 1978; Ryding, 1980). Most bioassays use unicel bioassay procedure has recently been developed lular green algae or short filamentous blue-green with attached filamentous algae of the genus algae as test organism. Selenastrum capricornutum Stigeoclonium Kütz. as test organisms, using a tube Skulberg has become the major bioassay alga for assay technique (De Vries et al. 1983, 1985; De testing eutrophication in freshwaters (Komarek & Vries & Hotting, 1985). -171-

In the present study the results of bioassays in culture tubes with Stigeoclonium tenue or Scenedesmus quadricauda as test organisms, and results obtained using the bottle test with Scenedes mus quadricauda, are compared to evaluate the ef fectiveness and sensitivity of both techniques and algal species.

Material and methods

Water samples were collected at 14 locations with different trophic status in the central western part of The Netherlands, in March, June, September and December 1984 (Fig. 1). Thirteen sites were lo cated in smaller or greater lakes, one (site 32) was located in a canal (Table I). The nutrient concentrations of the natural test waters were determined as soon as possible. Kjeldahl-nitrogen, ammonium, nitrite, nitrate, total phosphorus and orthophosphate were analysed on a Technicon AA II autoanalyser, according to stan dard methods of the Nederlands Normalisatie In stituut (1975). Bioassays with two test algae and two test Fig. I. The area of the Rijnland waterboard, its location in The methods (tube and bottle test) were carried out to Netherlands (inset) and the location of the sampling sites. determine the availability of nutrients for algal growth and to identify the growth limiting

Table I. Numbers, names and morphomelrical characteristics of the sampling locations.

Site number* Name of sampling site Surface (km2) Mean depth (m)

32 Circular canal of the Haarlemmermeerpolder 3.3 272 Lake Braassem 4.6 3.5 275 Lake Nieuwe Meer 1.4 5.0 284 Westeinder lakes 8.9 2.9 .179 Lake Zeegerplas 0.6 25.0 391 Lake Mooie Nel 0.7 8.0 402 Lake Sloterplas 0.9 30.0 18.03 Reeuwijk lakes (Broekvelden/Vettenbroek) 1.5 20.0 94.11 Nieuwkoop lakes (Zuideinderplas) 1.0 3.0 94.12 Nieuwkoop lakes (Noordeinderplas) 1.5 3.0 95.04 Langeraar lakes 1.7 2.0 134.08 Reeuwijk lakes (Elfhoeven) 1.1 3.0 134.09 Reeuwijk lakes (Nieuwenbroek) 0.9 2.2 217.06 Amstelveen lake 0.5 1.5

* The numbers are sample site numbers of the Rijnland Waterboard. ■172-

Table 2. Scheme of the applied assay techniques.

Methodological aspects

Test-organism Stigeoclonium tenue Kütz; Scenedesmus quadricauda Scenedesmus quadricauda (Turp.) Bréb. (Turp.) Bréb.

Preculture 7 days in NP-deficient Wood's Hole medium (Stein, 12 days in 50«/o Z8 medium (Skulberg, 1966) with

1973, slightly modified by Francke & Ten Cale 1980) 5% KH2P04, 5% NaNO, and 5% Ca(NO,):4H20

Inoculum 0.2 ml of a suspension with an optical density at 10 4 cells ml ', approximately 0.5 ml, to 75 798 nm of 0.4 to 20 ml test water (De Vries & lest water Hotting, 1985)

Incubation 35 ml culture tubes, 0 15 mm containing 20 ml lest 150-ml Erlenmeyer flasks containing 75 ml test water, without addition (algal growth potential), water, without addition (algal growth potential) and with additions of nitrogen, 20 mg N l~' as and with addition of nitrogen, 10 mg N I ', as 1 NaN03 or phosphorus 1.5 rngPI" as KH2P04 or a KNO„ or phosphorus. 0.5 mg P 1 ' as KH,P04/

combination of 20 mg N and 1.5 mg P I"' KH2P04, or a combination of 10 mg N and 0.5 mg P I '

Culture The assays were run in triplicate at 16°C in a 12:12 The assays were run in duplicate at 20°C in a conditions light-dark regime of ± 50 nE m'2 s~'. The tubes, 24 hour light regime at ± 100 pli m - s '. The closed with cotton plugs, were placed in racks bottles were closed with cotton plugs and placed on 10 degrees of horizontal. Cultures of Scenedesmus a shaking table (120 rpm) were shaken daily

Biomass The yield was measured after 3 weeks of incubation. The yield was measured 3-5 times per week with Then the maximum standing crop was attained for a Philips/Pye Unicam PU 8800/02 Speclropholo- Stigeoclonium. The Stigeoclonium cultures were meter by substracting the optical density at 750 nm sonified ultrasonically for 30 sec, whereafter the from the optical density at 680 nm. When the optical density at 798 nm was measured on a Hitachi maximum standing crop was reached (less than 5% 100-40 Spectrophotometer (De Vries & Kamphof, growth day '), mostly within 3-4 weeks, the test 1984) was ended (U.S. Environmental Protection Agency, 1971). nutrients. The principle of such assys is based on & Ouboter (1985). Afterwards the pH was adjusted Von Liebig's Law, which states that the maximum to pH 7.00 ± 0.25 with I N KOH. Schemes of the yield of a species (or population) is determined by tube and bottle test assay techniques are presented the nutrient, which is present in the least available in Table 2. quantity. Addition of this nutrient will raise the Stigeoclonium tenue Kütz culture code 40101 (De yield. With such bioassays the growth potential and Vries & Kamphof, 1984) was one test organism used the limiting nutrient factor (N, P or N + P) can be in the present study. The second was Scenedesmus detected (Thomas, 1953; US Environmental Protec quadricauda (Turp.) Bréb., which was isolated from tion Agency, 1971; Chiaudani & Vighi, 1974; 1975; Lake Tjeukemeer and supplied by the Limnological Forsberg el al. 1978). Institute at Nieuwersluis (The Netherlands). The water samples were preserved by deep The Student t-test (Sokal & Rohlf, 1969) was ap freezing until the bioassays were conducted. For plied to identify the growth limiting nutrients in the this purpose the samples were thawed in a water- bioassays. A nutrient was considered as limiting, if bath at ±45°C and autoclaved for 20 min at addition of the nutrient increased the maximum al 120°C: See for the effects of autoclaving De Vries gal yield significantly (P < 0.05). The product- -173-

moment correlations between yields and nutrient tials are ranked in increasing order of mean yield concentrations were also calculated according to per site (Fig. 2). Comparative low growth poten Sokal & Rohlf (1969). tials were measured at the relatively unpolluted sites 18.03, 134.08 and 134.09, whereas the highest yields were observed in samples from sites 272, 32, Results 379 and 391, being more or less influenced by dis charges of sewage treatment plants and water inlet The maximum algal yields of the bioassays without from the river Rhine. nutrient addition, the so-called algal growth poten Generally the maximum yields obtained in the

Site Stigeoclonium tube test Scenedesmus tube test Scenedesmus bottle test

Mirct> «pp Jurw September

»pp IP[~I June IP[X December

:& March PN[~1 Jon* September P«H

NPI p^ NP i

1 !pf= p 1 June September P i Deteinber March "1 I ' P| 1 September

H 1

i P 1 '

1 P -i—' > 1

1 F r1,

O.D. W»-T5ön

Fig 2 Mean yields of the assays without nutrient addition (algal growth potential) and primary and, if detectable, secondary limiting nutrient in the different tests.

• .• #/ .• I 006

Y^O.73 X.0.0S8 |N^39| 00b X-0.OO6S(N=39l

•. >/ I ^

0 0.3 0.6 0~/ 00. 3 0.6 yield Scenedesmus in tube test {O.D. 798nm) yield Scenedesmus in tube test (O.D. 798nm)

Fig. 3a~b. Correlation plots and mathematic equations between yields in the tube assays using Sligeoclonium and Scenedesmus (A) and between yields in the bottle and tube test using Scenedesmus (B). assays with Sligeoclonium agree with the maximum 94.11 in June and sites 95.04 and 284 in March. Ad yields in the assays with Scenedesmus in the' tube dition of phosphorus could increase the yield in and bottle test (Fig. 2). The graphical presentation samples from site 18.03 in June. However, the of the correlations between the yields obtained with production was less than with phosphorus and the different species and test methods are given in nitrogen added together, which indicates that Fig. 3. Although a very high positive correlation, phosphorus was the primary and nitrogen the r = 0.86; P< 0.001, was found between the yields secondary limiting nutrient. obtained in the tube tests and bottle test, the The observed limiting nutrients per site were highest correlation, r = 0.94; p < 0.001, was found identical in the thus far discussed cases. In some as between the two tube tests. This implicates that the says, however, the three tests revealed different algal growth potentials of both Stigeoclonium growth-limiting nutrients per water sample. For in tenue and Scenedesmus quadricauda were similar stance Stigeoclonium was primary limited in its regardless the applied test technique. growth by phosphorus and secondary by nitrogen The bioassays were also used to identify the algal in the water samples from site 272 in June, whereas growth-limiting nutrients. The primary and, if de some unknown factor (X) determined the yield in tectable, secondary growth-limiting nutrients are the Scenedesmus tube test and finally nitrogen was presented in Fig. 2. Nitrogen growth-limitation oc the only growth-limiting nutrient in the bottle test curred frequently in all three tests, especially in inoculated with Scenedesmus. water samples from site 402 and 391. In most assays The number and percentage of the assays that the algal yield increased even more when both revealed an identical primary growth-limiting nutri nitrogen and phosphorus were added. This indi ent according to the various tests is presented in Ta cates that in waters growth-limited by nitrogen ble 3. The primary growth-limiting nutrient in 87% phosphorus became the next (secondary) growth^ of the assays with Scenedesmus in the tube test was limiting nutrient when sufficient nitrogen was conform to that observed in the bottle test. The provided, i.e. in water samples from sites 134.08 and limitations of the tube tests for both algal species -175-

Table 3. Percentage of assays with an identical primary limit at sites 95.04, 275 and 32 in September (see ing nutrient according to the various tests, in each combination Fig. 2). Also from Fig. 2 it becomes clear that a (n = 39). greater percentage of assays with Stigeoclonium reveals a primary phosphorus limitation compared Combination Percentage of identical orimary to the two Scenedesmus tests i.e. sites 134.08, 94.11, limiting nutrients 284, 275 and 275 in March. However some assays of the tube test inoculated with Scenedesmus rev Stig. tube test - Seen. tube test - Seen. ealed phosphorus as a secondary growth-limiting bottle test 67 nutrient, whereas Scenedesmus in the bottle test Seen. tube test 79 Stig. tube test - showed no secondary limitation i.e. site 217.06. Stig. tube test - Seen. bottle test 69 Seen, tube test -- Seen bottle test 87 Each test water can be characterized by its amount of nitrogen and phosphorus. The total nitrogen and total phosphorus concentrations are agreed with each other in 79% of the assays. The plotted against each other in Fig. 4, together with results obtained with both Scenedesmus test sys the growth limiting nutrients determined on the ba tems were more uniform than those of the sis of the bioassays. In general the Stigeoclonium Stigeoclonium and Scenedesmus tests. tube test revealed no primary phoshorus growth- In some assays, using Stigeoclonium, phospho limitation when the total-N to total-P ratio (by rus could be detected as the secondary growth- weight) was lower than 17. Phosphorus was limiting nutrient while in assays inoculated with primary growth-limiting at ratios above 17. The Scenedesmus, both in the tube and bottle test no critical N to P ratio for the Scenedesmus tube test secondary limiting nutrient could be observed i.e. was approximately 22. The critical ratios calculated

Stigeoclonium tube test Scenedesmus tube test Scenedesmus bottle test

1.4 2.1 2.0 1.5 2.U 2.0 1.5 mgPI- II I Ë Hi

0.5 *. o °l-ï-

10 0 s mgNI-1

Fig. 4a-c. Total-N and total-P concentration of the test waters and the limiting nutrients according to the various tests. The symbols indicate the limiting nutrients: a = nitrogen-, ■ = primary nitrogen-, secondary phosphorus-, O = phosphorus, • = primary phosphorus-, secondary nitrogen limitation. The lines represent the critical N/P ratios for Stigeoclonium (17) and Scenedesmus (22). -176-

Table 4. Correlation between the yields in the assays and con retical expected correlation of r = 1.0. This may be centrations of the primary limiting nulrient in the samples, caused by differences in method of biomass meas urements and test period. The biomass in the tube Nutrient Stigeo- Scene- Scene- test is measured as the optical density at 798 nm, clonium desmus desmus tube test tube test bottle test which measures the turbidity of the suspension and as such all living and dead cells. The yield in the n r n r n r bottle test, however, is measured as the optical den sity at 680 nm minus the optical density at 750 nm Tolal-N 24 0.88*** 31 0.93"" 48 0.86***. (by which chlorophyll-a is measured) thus only the Inorganic-N 24 0.90*** 31 0.89*** 48 0.97*** living cells. This difference in biomass measure Tolal-P 15 0.76** 7 0.72 4 0.26 Ortho-P 15 0.93*** 7 0.57 4 -0.72 ment together with the difference in test period, 3 weeks for the tube test versus a more variable dura **P<0.OI, ***P<0.O0l. tion depending on the 5% growth criterion (US En vironmental Protection Agency, 1971) for the bottle test, may be responsible for the high growth poten as the inorganic nitrogen (N02 + N03 + NH4-N) tials in the Scenedesmus tube test compared with to inorganic phosphorus (P04-P) ratio (by weight) varied from approximately 9 in the the relatively low potentials in some assays in the Stigeoclonium tube test to 12 — 15 in the Scenedes Scenedesmus bottle test i.e. sites 272, 379 and 391 mus tube test. Critical N/P values for the Scenedes in June (Fig. 2). mus bottle test could not be estimated accurately by The lowest cell yields were found in P-limited lack of sufficient (only 4) P-limitations. The differ (tube test) and/or NP limited waters (bottle test) i.e. ences in critical N.P ratios for algal growth impli sites 18.03, 134.08 and 134.09 in the Reeuwijk lakes. cates that Stigeoclonium requires more phoshorus With increasing yields nitrogen became the most for growth than Scenedesmus in water samples important limiting nutrient in the assays. See for with an equal amount of nitrogen. Therefore, the example sites 32, 379 and 391. These findings agree Stigeoclonium tube test reveals more often a P- with the results of Toerien el al. (1975). limitation. The primary growth-limiting nutrient of the The yields of primary nitrogen growth-limited Stigeoclonium and Scenedesmus tube tests were assays were highly significant correlated (P < 0:001) identical in 79% of the assays. In 20%, 8 assays, the with the total-N and inorganic-N concentration, primary limiting nutrients were not identical. The whereas the yields in primary phosphorus growth- growth of Stigeoclonium was primary growth- limited assays were only significant correlated with limited by phosphorus, whereas Scenedesmus was the ortho-P and total-P concentrations in the growth-limited by nitrogen or an unknown factor Stigeoclonium tube test (Table 4). (X). This implicates that Stigeoclonium requires relatively more phoshorus for its growth than Scenedesmus. The nitrogen to phosphorus ratios of Discussion the lake waters and the most growth-limiting nutrients determined in the algal assays indicate the The algal growth potentials obtained in the 3 test critical ratio of these nutrients for the applied test systems were highly significantly correlated. The organisms. Schmidt-van Dorp (1978) using natural lowest yields were found in samples with relatively phytoplankton observed that mass ratios greater low concentrations of nitrogen and phosphorus, than 16 indicated phosphorus limitation. Forsberg whereas the highest yields were obtained in waters et al. (1978) presented critical total N to P mass ra with the highest amounts of nitrogen and phospho tios for Selenastrum capricornutum. Nitrogen was rus regardless the applied test technique. However; growth-limiting at ratios below 10 and phosphorus the calculated correlation between both Scenedes was limited above 17. The critical mass ratios for mus tests (r = 0.86) deviates slightly from the theo Stigeoclonium tenue determined in this study cor- -177-

responds with the ratio reported by Forsberg el al. yield significantly, but the addition may not be (1978). However, the critical total-N to total-P value high enough to eliminate the primary limiting nu (by weight) for Scenedesmus quadricauda based on trient. Therefore additions of both nitrogen and the results in. the tube test were slightly higher, 22, phosphorus together will have the same effect on the than Forsberg el al. (1978) reported for Selenastrum yield as the addition of nitrogen apart and only the capricornulum. This difference in critical value be primary limiting nutrient is detected in the bottle tween the two test species implicated that in test test, whereas in the tube test frequently also a waters with a (total-N : total-P) ratio of approxi secondary limiting nutrient could be detected. mately 20 (by weight), Stigeoclonium could be The attached filamentous alga Stigeoclonium limited in its growth by phosphorus while lenue has the advantage that it occurs throughout Scenedesmus could be nitrogen limited. The critical the year. It can be considered as a fair representa ratios (by weight) found in this study for the inor tive of algal species found in small water bodies ganic phosphorus concentrations correspond very with an extensive littoral, like ditches, whereas well with the ratio of 10 reported by Chiaudani & Scenedesmus quadricauda is a representative of the Vighi (1974, 1975) and of 12 reported by Forsberg planktonic algal community in shallow lakes and el al. (1978) for Selenastrum capricornulum, canals. Both species are easily isolated and cultivat although again in the Scenedesmus tube tests a ed and their reproduction is predominantly a- slightly higher ratio (12-15) was found than in the sexual, which limits the genetic variation of the test Stigeoclonium tube tests (9). clone. The two test methods using Scenedesmus (tube Culture tube experiments require little space, 200 and bottle test) showed that the primary growth- tubes m~2, and the biomass has only to be meas limiting nutrients were identical in 87% of the as ured after 3 weeks of incubation. A disadvantage says. The results from 5 sites differed; samples from of Stigeoclonium is that no measurements can be sites 134.09 and 94.12 in March and September were made during the growth phase without disturbing primary P-limited and site 272 was limited by an the culture. In consequence, the stationary growth unknown factor (X) according to the tube test in phase cannot be determined exactly. The bottle test June, whereas Scenedesmus in the bottle test assays requires much more space, 115 bottles m"2, and was primary limited by nitrogen. No clear explana constant culture shaking. The determination of the tion concerning the factor(s) which underlay the stationary growth phase is exact but laborious due difference in primary limiting nutrients in the 5 as to frequent inbetween measurements of the yield. says can be given. Finally, as contrasted with the culture tube tech The tube tests inoculated with Scenedesmus rev nique with Stigeoclonium, results in the bottle test ealed phoshorus as the secondary growth-limiting with Scenedesmus are mostly obtained after a nutrient more often than the bottle test i.e. site shorter testperiod. 217.06 (Fig. 2). This may be explained by the differ It can be concluded that in the case of S. quad ence in the nutrient concentrations that were added ricauda both test methods are nearly equal. The to the water samples. Nitrogen and phosphorus ad deviation in critical N/P ratios (23% between ditions raised the concentrations of the test waters Stigeoclonium tenue and Scenedesmus quadricau in the tube tests by 20 mg N I"1 and 1.5 mg P 1~' da should be kept in mind, while applying one of respectively, whereas the concentrations in the bot these algae in field studies to evaluate the effects of tle test were raised only by 10 mg N 1"' and 0.5 mg P-reduction measures. PI"1. Therefore, after addition of sufficient nitro gen or phosphorus in the tube tests, the primary growth-limiting nutrient can be eliminated which Acknowledgement enabled the detection of the next (secondary) limit ing nutrient. The addition of 10 mg N 1~' to N- The authors express their appreciation to Prof. Dr. limited samples in the bottle test will increase the M. Donze, Prof. Dr. W. H. O. Ernst, Dr. H. -178-

Hillebrand and Prof. Dr. M. Vroman for critical clusters of filamentous algae. In R. G. Wetzel (ed.), Periphy- reading of the manuscript and to Paul Gutteridge ton of Freshwater Ecosystems. Dev. Hydrobiol. 17: 31-39- for correcting the English text. Junk, The Hague. Klapwijk, S. P., 1978. Saprobiological evaluation of Dutch ditches using benthic algae on artificial substrates. Verh. Int. Ver. Limnol. 20: 1811-1815. Klapwijk, S. P., T. F. de Boer, & M. J. Rijs, 1983. Effects of References agricultural waste water and benthic algae in ditches of the Netherlands. In: R. G. Wetzel (ed.), Periphyton of freshwater Barica, J. & L. R. Mur, 1980 (eds>. Hyperirophic Ecosystems, Ecosystems. Dev. Hydrobiol. 17:311-319. Junk, The Hague. Dev. Hydrobiol., 2. Junk, The Hague, 348 pp. Komarek, J. & O. Lhotsky, 1979. Review of algal assay strains. Bolier, G & L. W. C. A. Van Breemen, 1982. Standaardisatie In P. Marvan, S. Pybril & O. Lhotsky (eds). Algal Assays and van de algengroeipotentie-toets voor Nederland: Verkenning Monitoring Eutrophication. pp. 103-118.E. Schweizer- en afbakening. T. H. Delft, Vakgroep Gezondheidstechniek. bart*sche Verlagsbuchhandlung, Stuttgart. Rapport No. 82.03 (in Dutch). Miller, W. E., T. E. MaloneyA J. C. Green, 1974. Algal produc Chiaudani, G. & M. Vighi, 1974. The N:P ratio and tests with tivity in 49 lake waters as determined by algal assays. Wat. Setenasfrum lo predict eutrophication in lakes. Wat. Res. 8: Res. 8:667-679. 1063-1069. Nederlands Normalisatie Instituut (1975). Test methods for Chiaudani, G. & M. Vighi, 1975. Dynamics of nutrient limita waste water. NEN 3235 (in Dutch). tions in six small lakes Verh. int. Ver. Limnol. 19: 1319-1324. Ryding, S-O. (ed.), 1980. Monitoring of Inland Waters. Report De Vries, P. J. R., S. J. M. De Smet & J. Van der Heide, 1985. from the working group for eutrophication research. Nord- Effects of phosphorus and nitrogen enrichment on the yield forsk publication 1980: 2, Helsingfors Finland, 207 pp. of some strains of Stigeoclonium Kütz. (Chlorophyceae). Schmidt-van Dorp, A. D, 1987. Eutrophication of shallow lakes Freshwat. Biol. 15:95-103. in Rijnland. Technische Dienst Hoogheemraadschap Rijn De Vries, P. J. R. & E. J. Hotting, 1985. Bioassays with land, Leiden, 254 pp.(in Dutch, with a summary in english). Stigeoclonium tenue Kütz on waters receiving sewage ef Skulberg, O. M. 1966. Algal cultures as a means to assess the fluents. Wat. Res 19: 1405-1410. fertilizing influence of pollution. Proc. 3rd Int. Conf. Water De Vries, P. J. R. & G. J. Kamphof, 1984. Growth of some Poll. Res., Sec. 1 paper 6. 113-137. strains of Stigeoclonium (Chlorophyta) on nitrate, ammoni Sokal, R. R. & F. J, Rohlf, 1969. Biometry - the Principles and um, ammonium nitrate and urea. Br. Phycol. J. 19:349-356. Practice of Statistics in Biological Research. W. H. Freeman, De Vries, P. J. R. & P. S. H. Ouboter, 1985. Water sample treat San Francisco. ments and their effects on bioassays using Stigeoclonium hel- Stein, J. R. (ed.), 1973. Handboek of phycological methods: cul veticum Vischer. Aquat. Bot. 22: 177-185. ture methods and growth measurements. Cambridge Univer De Vries, P. J. R., M. Torenbeek, & H. Hillebrand, 1983. Bioas sity Press, Cambridge, 448 pp. says with Stigeoclonium Kütz (Chlorophyceae) to identify Thomas, E. A., 1953. Empirische und experimentelle Unter- nitrogen and phosphorus limitations. Aquat. Bot. 17:95 -106. suchungen zur Kenntnis der Minimumstoffe in 46 seen der Forsberg, C, S. O. Ryding, A Claesson & A. Forsberg. 1978. Schweiz und angrenzender Gebiete. Schweiz. Ver. Gaz. Wasser Water chemical analysis and/or algal assay? Sewage effluent Fachm. 2: 25-32; 71-79. and polluted lake water studies. Verh. int. Ver. Lim Toerien, D. F, K. L. Hyman, & M. J. Bruwer, 1975. A prelimi nol. 21:352-363. nary trophic status classification of some South African im Francke, J. A. & M. J. Ten Cate, 1980. Ecotypic differentiation poundments. Water, S-A 1:15-23. in response to nutritional factors in the algal genus Stigeoclo US Environmental Protection Agency, 1971. National Eutrophi nium Kütz. (Chlorophyceae). Br. phycol. J. 15: 343-355. cation Research Program: Algal Assay Procedure Bottle Test. Hillebrand, H., 1983- Development and dynamics of floating U S. E.P. A. Corvallis, Oregon, 82 pp. -179-

CHAPTER 14: COMPARISON OF DIFFERENT METHODS TO DETERMINE GROWTH LIMITING FACTORS FOR PHYTOPLANKTON IN THE REEUWIJK LAKES (THE NETHERLANDS)

S.P. Klapwijk, U.G. Dijkstra-Stam, A.K. Fleuren-Kemüa, J.C. van der Vlugt.

Submitted to: Water Research.

"It would be interesting to compare through a year's time the res ponse of natural phytoplankton populations to nutrient additions with some standard test organism such as Selenastrum capricornutum which is maintained under uniform test conditons."

C.R. Goldman, 1978. The use of natural phytoplankton popula tions in bioassay. Mitt, internat. Verein. Limnol. 21, p. 365. -180-

COMPARISON OF DIFFERENT METHODS TO DETERMINE GROWTH LIMITING FACTORS FOR PHYTOPLANKTON IN THE REEUWIJK LAKES (THE NETHERLANDS)

S.P. Klapwijk 1), U.G. Dijkstra-Stam 1), A.K. Fleuren-Kemüa 2) & J.C. van der Vlugt 2).

Key words: eutrophication, nutrients, phosphorus, nitrogen, limiting factors, bioassays, uptake rates, Reeuwijk lakes.

ABSTRACT

Five different methods to assess growth limiting factors for al gae were compared and applied to data from the Reeuwijk lakes over the years 1983-1985: (1) bioassays with the natural phytoplankton population, (2) bioassays with testalgae, (3) ammonium and phosphate uptake experiments, (4) nutrient ratios and (5) statistical relation ships between chlorophyll-a and possible limiting factors. The different methods did not always lead to the same results. Both types of bioassays and the N/C ratios predicted in general pri marily a N-limitation for the lakes, while the other methods in general predicted a P-limitation. The reason for this was not always obvious. Several possibilities are discussed and the various methods are eva luated. It is suggested to apply in lake restoration studies aimed at reducing the P-load several methods simultaneously. Furthermore it is concluded that the bioassay method with the natural phytoplankton population is probably the most direct and closest to the field situa tion with respect to assessing the chemical growth limiting factors in the water. Despite the variable results of the different methods it is evi dent that lake Nieuwenbroek with the lowest P-concentrations but with the highest algal biomass is probably closer to a P-limitation than lake Elfhoeven with the highest P-concentrations and the lowest algal bio mass . Therefore P-reduction measures will probably show more and quicker results in the first mentioned lake.

1) Waterboard of Rijnland, P.O. Box 156, 2300 AD Leiden, The Netherlands.

2) National Institute of Public Health and Environmental Pro tection, P.O. Box 1, 3720 BA Bilthoven, The Netherlands. -181-

INTRODUCTION

It is important to know the growth limiting factor(s) for phytoplankton in surface waters, especially in relation to lake restoration programs, where the effects of phosphorus load reduction on algal biomass have to be predicted with some precision. Different methods are used to assess the growth limiting factors, such as bioassays with nutrient enrichments, physiological indicators and nutrient ratios. So far, it is not clear which of these methods produces the best and the most reliable indication for the Limiting factor(s). The few studies in which several methods are compared show that the results are not al ways the same and sometimes contradictory (Gerhart & Likens, 1975; O'Brien & Denoyelles, 1976). In this study five methods of assessing growth limiting factors are compared: (a) bioassays with the natural population, (b) bioas says with testalgae, (c) physiological indicators, (d) nutrient ratios, and (e) statistical relationships between chlorophyll-a and possible limiting factors. a. Bioassays with the natural population

Bioassays with the natural phytoplankton population are in long standing use to assess growth limiting nutrient factors (Thomas, 1953; Schmidt-van Dorp, 1978; Goldman, 1978; Javornicky, 1979; L0vstad, 1984; van der Does & Klapwijk, 1987). The method is based on von Liebig's Law, which.states that the maximum yield of a species (or population) is determined by the nutrient which is present in the least available quantity. Addition of this nu trient will raise the yield of the population. The procedure of the assays is straightforward: water is collected from the field, brought to the laboratory and filtered to remove zooplankton. Some water samples are enriched with nutrients (single or multi ple). Then the cultures are kept under standarized conditions (light, temperature, C02 supply) and algal growth in the en riched samples is compared with water without additions. From this the limiting nutrient(s) for the indigenous phytoplankton population can be derived, which can be easily translated to the field situation. The method is simpler and more direct than the bioassay method with testalgae, because there is almost no pretreatment of the sampled water, which excludes the risk of side- effects of such pretreatments. But differences in plankton com position and abundance between several waters or different sea sons make comparison of results sometimes troublesome. b. Bioassays with testalgae

Bioassays with testalgae are also widely used to assess growth limiting nutrients in water (EPA, 1971; Chiaudani & Vighi, 1974, 1975, 1976; Miller et al., 1974, 1978; Forsberg et al., 1978; Ryding, 1980; Bolier et al., 1981; STORA, 1986; de Vries & Klapwijk, 1987). The principle is the same as the previous ex cept for the use of an unialgal strain of a testalga instead of the natural phytoplankton population. After a pretreatment to kill the indigenous plankton population and to liberate nutrients bound to the dead biomass (effects on chemical speciation have -182- so far been disregarded), the water samples are inoculated with a testalga and enriched with nutrients (single or multiple). Then the cultures are kept under standarized conditions (light, tempe rature) and algal growth in the enriched samples is compared with growth in the controls. From this the limiting nutrient factor(s) for the specific testalga can be derived, which still has to be translated to the field situation. As a result of the pretreatment of the water and the standardization of the method it is quite possible to compare several waters or the results of one water over a period of time (cf. de Vries, 1986).

Physiological indicators

Physiological indicators are more recently developed to discover growth limiting factors. The methods are mostly based on spe cific physiological properties, such as phosphatase activity or phosphate uptake rate, from which it is known that they can provide information about the nature of the growth limiting nu trient, including light (Healey & Hendzel, 1979; 1980; Pettersson, 1980; Lean & Pick, 1981; Zevenboom et al., 1982; Riegman, 1985; Riegman & Mur, 1986). Most indicators are based on growth or uptake rates determined in continuous culture experiments in stead of yields as in most bioassays. The results can therefore not readily be translated to the field situation.

Nutrient ratios

Nutrient ratios are used already a considerable time to assess growth limiting nutrients (e.g. Chiaudani & Vighi, 1974; 1975; Forsberg et al., 1978; Schmidt-van Dorp, 1978; Healy & Hendzel, 1979; Smith, 1982; de Vries & Klapwijk, 1987). The method is based on the fact that the ratio of N, P and C in algae is rather constant. The nutrient that is proportionally the least present relative to the demand of a species is presumably limiting the maximum algal growth. Critical ranges for nutrient mass ratios have been presented in the literature (cf. Forsberg et al., 1978; Schmidt-van Dorp, 1978; Claesson & Forsberg, 1980; Healey & Hendzel, 1980; de Vries, 1986; de Vries & Klapwijk, 1987). The method is very simple but also rather rough and indirect, since nutrient ratios in water can fluctuate considerably, also over the course of the year. Moreover, algal limitation is not only depen dent on nutrient ratios but also on nutrient concentrations.

Statistical relationships

Statistical relationships describing average or maximum chlorophyll-a/total-P ratios are reviewed among others by Nicholls and Dillon (1978) and OECD (1982). Most of these relationships refer to a large number of mainly phosphorus limited lakes and were obtained by regression analysis. In the Netherlands comparable studies have been carried out, which resulted in relationships between mean and maximum chlorophyll-a and the mean total P, total N and the corrected Secchi depth in summer (Hosper, 1980; CUWVO, 1980; Janse et al., 1987). From these relationships the most probable limiting factor (N, P, light) for a summer period can also be inferred. -183-

These five methods have been applied to water samples collected in the Reeuwijk lakes during 1983-1985. The results are used to eva luate the various methods and to predict the effects of phosphorus load reduction measures for these lakes, which have started in 1986 (van der Vlugt and Klapwijk, 1987 a,b).

MATERIALS AND METHODS

Description of sampling area

Samples were taken from the Reeuwijk lakes, situated at 52° 2' N, 4° 45' E in the Reeuwijk polder near Gouda in the western part of the Netherlands (Fig. 1). The lakes originate from peat mining in the 16th and 17th century. They are shallow (mostly < 2.5 m) and rich in humic substances. A large part of the Reeuwijk polder is used for agriculture purposes, mostly cattle breeding. All the lakes, except for lake Broekvelden/Vettenbroek, are interconnected having the same water level and so are the ditches in the polder.

Fig. 1. The Reeuwijk lakes area with the location of the sampling sites in lake Elfhoeven and lake Nieuwenbroek and the se wage treatment plant of the village of Reeuwijk. -184-

The water level in the polder is kept constant at 2.2 m below mean sea level by inlet of water into the Breevaart canal from the city canals of Gouda, which are fed with water from the Hollandse Ussel, a branch of the river Rhine. Excess water in the polder is pumped out near Gouda and Bodegraven. The sewage treatment plant of the village of Reeuwijk (with 6,800 inhabitants) discharges its nu trient-rich effluent into the Breevaart canal, through which the level of the lakes is regulated. Sampling took place at two stations: lake Elfhoeven and lake Nieuwenbroek. In Table 1 some morphological, hydrological and phos phate loading characteristics are summarized for these lakes. The plankton composition in lake Elfhoeven consists mainly of chlorococcal greenalgae, diatoms and sometimes blue-green algae, whereas the phytoplankton in lake Nieuwenbroek is dominated permanently by filamen tous blue-green algae of the genera Lyngbya and Oscillatoria (van der Vlugt & Klapwijk, 1987b).

Table 1. Some morphological and hydrological characteristics of the two sampled lakes in the Reeuwijk Lakes area.

Lake Elfhoeven Lake Nieuwenbroek

Water surface area (km2) 1.1 1.1 Mean water depth (m) 2.4 1.6 Maximum water depth (m) 8.0 2.2 Residence time (year) 0.5 2.5 Gross P-load (g P m 2 y 1) 1.56 0.5 General sediment type peat peat

Sampling and chemical analyses

Water for chemical analyses was sampled every week for three years at three depths (0-0.5 m, 0.5-1.0 m, 1.0-1.5 m) with a verti cal water sampler. The water taken from the three depths was mixed and transported to the laboratory in dark polyethylene jerrycans. Nitrate + nitrite nitrogen (N02+N03-N), ammonium nitrogen (NH4-N) and orthophosphate (PO4-P) were measured with a Technicon A A II autoanalyser. Kjeldahl-nitrogen (Kj-N) and total phosphorus (total-P) were measured after digestion with respectively sulphuric acid + Wieninger selenium mixture and sulphuric acid + persulphate. Subsequently NH4-N and PO4-P were determined. Dissolved inorganic nitrogen was calculated by addition of the weights of NH4-N and N02+ NO3-N. Total nitrogen was calculated by addition of Kj-N and N02+ NO3-N. Chlorophyll-a (Chi) was spectrophotometrically measured after extraction with 80% ethanol (75°C). Seston and particulate organic matter were determined gravimetrically as dry weight at 75°C and ash free dry weight (AFDW) at 600°C. Transparency was measured with a Secchi disk. -185-

Bioassays Bioassays were carried out every month from 1983-1985. For bioassays with, the natural phytoplankton population, zooplankton was removed in the laboratory with 80 um plankton gauze. In 1983 and 1984 water with the natural phytoplankton population was divided into eight 2 liter Erlenmeyer flasks filled to 1 liter, while in 1985 the test were carried out in twelve 500 ml Erlenmeyer flasks filled to 100 ml. The experiments were run in duplicate (1983, 1984) or tripUcate (1985). Two or three Erlenmeyers had no addition of nutrients and served as controls. The others were enriched with 10 mg N 1 * (as 1 KNO,) or 0.5 mg P l" (as K2HP04/KH2P04) or 10 mg N 1 * + 0.5 mg P 1 *. See for the simultaneous addition of respectively 27.9 or 1.23 mg K l"1 Klapwijk et al. (in prep). The culture vessels closed with cotton plugs were placed in a shaking waterbath at a temperature of 20 ± 1°C, a PAR-light intensity of ± 70 uE m"2s_1, and a light/dark rhythm of 12/12 hours (lamp type Philips TL 40W/57 or Philips TL 18W/33). The optical density (O.D.) at 680 nm minus the O.D. at 750 nm was used as a measure of biomass (de Vries & Klapwijk, 1987). Bioassays with a testalga were also carried out in duplicate or triplicate. During 1983 and 1984 the sampled water was poured with out filtration into 150 ml Erlenmeyer flasks filled to 75 ml, while in 1985 100 ml was poured into 500 ml Erlenmeyer flasks. The natural population was killed by autoclaving, which also should liberate nutri ents bond in the organisms (Miller et al., 19_78). Some samples were enriched with the same spiking of 10 mg N 1 ' and/or 0.5 mg PI1. All samples were inoculated with about 10,000 cells per ml of Scene- desmus quadricauda (Turp.) Bréb., from a batch culture derived from the Limnological Institute, Nieuwersluis, which had been starved for nitrogen and phosphorus for about twelve days. The assay tech nique is described in more detail by de Vries and Klapwijk (1987).

Uptake experiments

In 1983 and 1984 phosphate, ammonium and nitrate uptake kine tics of the natural phytoplankton population from the lakes were mea sured essentially according to Healey and Hendzel (1979), Zevenboom et al. (1982) and Riegman and Mur (1986). One hour after collection nine 1 plankton samples were concentrated by centrifugation (15 min.; 3,000 rpm) and resuspended in a medium with dissolved inorganic nu trients (Riegman & Mur, 1986) without phosphate or nitrogen, depen ding on which uptake rate was measured. Samples were incubated at incident irradiance of 30 W m 2 and at a temperature of 20°C. After fifteen minutes of preincubation 10 ml algal suspension was added to 100 ml medium pH = 8.0 (approximately 30 mg dry weight 1 *) and phosphate as K2HP04, ammonium as NH4C1 or nitrate as NaNp3 was ad 1 ded in order to get initial concentrations of 310 yg P04-P l" , 280 |jg 1 1 NH4-N l" and 280 ug N03-N l" . These concentrations were chosen to ensure that, during the uptake experiments, the external nutrient concentrations were sufficient for unlimited uptake (Riegman & Mur, 1986). At different time intervals (t = 0, 5, 10, 15 and 30 minutes for P-uptake; t = 0, 15, 30 and 60 minutes for NH4-uptake; t =0, 1, 1.5, 2 and 24 hours for N03-uptake) 10 ml subsamples were taken and directly filtered with 0.45 M membrane filters and extracellular phosphate, ammonium or nitrate concentrations were determined. The -186- maximum initial uptake rate (V in mg N or P per mg Chl.h *) was calculated as follows: V = A S- I m A t x

where S = extracellular concentration of P (mg 1 x) t = time in hours x = chlorophyll biomass (mg 1 1).

V is divided by the chlorophyl concentration of the lake water to provide normalized estimates. By Riegman and Mur (1986) the natural phytoplankton population was arbitrarily defined to be P-limited, when V exceeded 0.0. 3 mg P mg"1 Chl.h * m & & Nutrient ratios

Total and inorganic N/P and particulate N/P, P/C, N/C and Chl/C ratios were calculated by weight. For the last three ratios the carbon content was estimated as half of the AFDW. Critical ranges for N/P mass ratios (total, particulate or inorganic), indicating limiting nutrients, are derived from the literature, mostly being obtained from empirical data compared with bioassay results (cf. Forsberg et al., 1978; Schmidt-van Dorp, 1978; Claesson & Forsberg, 1980; Healey & Hendzel, 1980; de Vries, 1986; de Vries & Klapwijk, 1987).

Statistical relationships

Simple empirical relationships for predicting the major limiting factor(s) were used according to a study on 90 shallow lakes in the Netherlands during 1977 (Hosper, 1980; CUWVO, 1980). With these relationships the mean and maximal algal biomass can.'be calculated on the basis of the summer (April-September) means of total N, total P and the corrected Secchi depth (So) according to the following formu las:

Chi = -15.53 + 776.32 * t-P Chlmax = -33.87 + 112.9 * t-N Chi = -89.6 + 89.6 * t-N Chlmax = 1206.9 * t-P Chi = 530/H - 40.38/So Chlmax = 1985/H - 94.71/So

where: - Chi, t-N and t-P are respectively average chlorophyll-a, total-N and_ total-P-^concentratiqns in sum mer (in resp. mg m 3, mg 1 x and mg 1 1); Chlmax is maximum chlorophyll-a concentration in summer (in mg m 3); H is the mean depth (in m); So is the average Secchi depth in summer without occurrence of algae (in m), assessed according to 1.9/S = 1.9/So + 0.016 Chi;

where: - S is the average Secchi depth in summer (in m).

The factor that gives the lowest Chi or Chlmax concentrations accor ding to these formulas is considered limiting for algal biomass. -187-

RESULTS AND DISCUSSION

Physical and chemical analyses

In Table 2 physical and chemical data over 1983-1985 of the sam pling stations in the Reeuwijk lakes are summarized. The concentra tion of nitrogen compounds, especially NH4-N, part-N and total-N are higher in lake Nieuwenbroek than in lake Elfhoeven, whereas all the phosphorus compounds are higher in Elfhoeven.

Table 2. Summarized data (yearly averages) of water analyses at the sampling stations in the Reeuwijk lakes.

Lake Elfhoeven Lake Nieuwenbroek

1983 1984 1985 1983 1984 1985

1 NH4-N (mg 1" ) 0.06 0.05 0.07 0.13 0.13 0.23

1 N02 + N03-N (mg f ) 0.10 0.09 0.02 0.08 0.07 0.05

Part-N (mg l"1) 0.81 0.87 0.82 1.49 1.51 1.39

Total-N (mg l"1) 1.96 2.05 1.89 2.83 2.82 2.76

1 PO4-P (mg 1" ) 0.026 0.016 0.011 0.014 0.010 0.006

Part-P (mg l"1) 0.09 0.08 0.11 0.06 0.06 0.06

Total-P (mg l"1) 0.14 0.12 0.14 0.08 0.08 0.08

Chlorophyll-a (Mg 1'1) 77 87 77 120 116 96

Dry Weight (mg 1"1) 20.3 17.5 17.3 29.8 24.6 23.2

1 AFDW (mg.f ) 12.9 10.9 10.3 25.8 21.4 18.9

Secchi depth (m) 0.61 0.60 0.54 0.43 0.41 0.45

Especially total-P in lake Elfhoeven is higher in summer than during winter due to the inlet of Rhinewater, while in both lakes inorganic nitrogen is normally higher in winter due to leaching and run-off from the agricultural polderland (Fig. 2). The concentration of the biomass parameters (chlorophyll-a, dry weight and AFDW) however are on average much higher in lake Nieuwenbroek than in lake Elf hoeven, consistent with the Secchi depth. This is probably due to the difference in plankton composition and the resulting difference in zooplankton biomass and grazing between both lakes (van der Vlugt & Klapwijk, 1987b). -188-

ORTHO- AND TOTAL PHOSPHORUS INORGANIC AND TOTAL NITROGEN

LAKE ELFHOEVEN LAKE NIEUWENBROEK LAKE ELFHOEVEN LAKE NIEUWENBROEK a TOTAL P a TOTAL P o TOTAL N A TOTAL N x ORTHO-P + ORTHO-P X INORGANIC N + INORGANIC N

Fig. 2. Monthly averages of phosphorus and nitrogen concentrations in the Reeuwijk lakes over the years 1983-1985.

Bioassays with natural population

The results of bioassays with the natural populations are presen ted in Figure 3. According to this method the algal growth in both lakes over the three years was limited primarily by nitrogen and se condarily by phosphorus. In lake Elfhoeven the algal growth was sti mulated in some months (e.g. February, 1983) only by the addition of nitrogen, while in all other months a strong stimulation was observed by the addition of N + P. In lake Nieuwenbroek the stimulation by N was generally very weak and only the stimulation by N + P addition gave a strong increase of algal growth, indicating that lake Nieuwen broek is closer to a P-limitation.

BIOASSAYS WITH NATURAL POPULATION BIOASSAYS WITH NATURAL POPULATION

LAKE ELFHOEVEN 1983-1905 LAKE NIEUWENBROEK 1983-1985

O NO ADDITION ft N+P ADDITION a NO ADDITION & N+P ADDITION

Fig. 3. Maximal algal yield in bioassays (means of duplo or triplo ex periments) with the natural phytoplankton population (1983- 1985). The experiments lasted for about two weeks. -189-

According to Goldman (1978) the use of the natural phytoplankton population gives a good indication of what is likely to occur in a lake when one or more nutrients are altered, experimentally. Another advantage is that bioassays with the natural population do not require an acclimation period to the nutrient levels of the water being tested (Goldman, 1978). The relatively long duration of the tests (several weeks) and the use of a yield parameter, such as chlorophyll-a, also contributes to the easy translation of the test to the field situation. Nevertheless, this type of assays had also some disadvantages. Be sides the disadvantages that are also true for bioassays with testalgae, such as the isolation from the natural environment (O'Brien & Denoyelles, 1976; Riegman, 1985; Wurtsbaugh et al., 1985) and the rather time-consuming method, a shifting of the natural plankton po pulation during a bioassay can appear (L,0vstad, 1984), which can confuse the results. On the other hand a shifting in the algal compo sition during the bioassay, e.g. towards nitrofixing bluegreens as was found frequently in this study, may also have some predictive value for the field situation (cf. Javornicky, 1979).

Bioassays with Scenedesmus quadricauda

In Figure 4 the results of the bioassays with Scenedesmus qua dricauda are presented. This figure shows that algal growth in both lakes is limited primarily by nitrogen and secondarily by phosphorus also, although in lake Nieuwenbroek in 1984 and especially 1985 the N-limitation was sometimes very weak and both N and P seemed to be equally limiting.

BIOASSAYS WITH SCENEDESMUS BIOASSAYS WITH SCENEDESMUS

LAKE ELFHOEVEN 1983-1985 LAKE NIEUWENBROEK 1983-I985

1984 1985 1984 1985

o P ADDITION 6 N+P ADDITION N ADDITION » P ADDITION & N+P ADDITION

Fig. 4. Maximal algal yield in bioassays (means of duplo or triplo ex periments) with the testalga Scenedesmus quadricauda (1983- 1985). The experiments lasted for about three weeks.

The bioassay method with testalgae is simple and reproducible (Goldman, 1978) and several nutrients can be tested simultaneously. O'Brien & Denoyelles (1976) showed the validity of this method by -190- comparing the results of bioassays with the testalga Pandorina morum on pond water with large-scale bioassays on the entire pond. The method also has some disadvantages: in the first place, the pretreatment, e.g. autoclaving of the water sample may influence its chemical characteristics (Miller et al., 1978; de Vries & Ouboter, 1985). Se condly, because of luxury uptake of P and the storage of essential elements, cultured organisms, even after various preparations and cell divisions, may carry with them significant quantities of essential nu trients into the new media (Goldman, 1978). In the third place, the method is rather laborious and the results only become available after some time.

Uptake experiments

In Figure 5 the initial ammonium uptake rate is plotted over 1983, whereas in Figure 6 the initial phosphate uptake rates in both lakes in 1983 and 1984 are presented. Very low P-uptake rates and moderate ammonium uptake rates are measured in lake Elfhoeven in 1983, which indicate a moderate N-limitation. Nitrate uptake rates (not shown here) were absent or only present after a lag-period of 24 hours. This means that the phytoplankton in the lakes used only am monium as N-source. See also Kappers (1984). In lake Nieuwenbroek a constant P-uptake rate of about 2 mg P mg 1 Chl.h 1 was measured during 1983 and generally a very low N-uptake rate, indicating a constant but moderate P-limitation and no N-limitation. In 1984, when only the P-uptake rates were measured, the phytoplankton in lake Elfhoeven showed fluctuating and sometimes very high P-uptake rates, up to 10 mg P mg 1 Chl.h * in April 1984, indicating severe P-limited growth, whereas in the other months e.g. September, October and December the P-uptake rate was rather low or even zero, indicating no P-limitation. The phytoplankton in lake Nieuwenbroek showed in 1984 generally the same or slightly higher P-uptake rates than in 1983, indicating again a moderate to strong P-limitation. According to Riegman (1985) the advantages of uptake experi ments are the use of the natural phytoplankton population and the quick, sensitive and accurate results. The method, however, also has some serious disadvantages: Firstly, the technique is calibrated in chemostat experiments with specific algal clones. Deviations can occur among different species and natural assemblages and may also be due to non-equilibrium conditions in the field. Therefore, the in the liter ature (e.g. Riegman & Mur, 1986) mentioned limits for nutrient-limited growth are rather arbitrary. Secondly, in some cases the uptake rate is strongly dependent on the light/dark regime under which the algae are growing (Lean & Pick, 1981) and therefore these rates can fluc tuate considerably during one or several days. Besides that, this method is based on growth or uptake rates instead of yields as in most bioassays. It is not self-evident that the growth rate limiting nutrient automatically has to be identical with the maximum growth (=yield) limiting nutrient. See also Hanna and Dauta (1983). This applies all the more for the phosphate uptake capacity as physiological indicator, since most algae have the possibility to store phosphate but nitrogen to a lesser degree. Nevertheless uptake rate measurements can be very useful in ecophysiological studies in order to understand uptake kinetics of populations of specific species (Riegman, 1985). -191-

AMMONIUM UPTAKE RATE 1983

1983

777?i LAKE ELFHOEVEN LAKE NIEUWENBROEK

5. Maximum initial ammonium uptake rates measured in 1983 on natural phytoplankton populations from the Reeuwijk lakes.

PHOSPHATE UPTAKE RATE

1983 - 1984 10

0 T—I—I I I J F M A J J SON D J F M A M J J A ! 1983 1984 LAKE ELFHOEVEN LAKE NIEUWENBROEK

6. Maximum initial phosphate uptake rates measured in 1983 and 1984 on natural phytoplankton populations from the Reeuwijk lakes. Values above 0.3 mg P mg 1 Chl.h"1 indicate phos- phoshorus limited growth according to Riegman and Mur (1986). -192-

Nutrient ratios The average nutrient ratios and nutrient or chlorophyll to car bon ratios are presented in Table 3. Comparison of the inorganic N/ ortho-P ratios with the critical values for limitation according to the literature shows that lake Elfhoeven is on the average nitrogen-(1983), nitrogen- or phosphorus- (1984) or phosphorus- (1985) limited, while lake Nieuwenbroek is mostly limited by phosphorus. In both lakes the highest inorganic N/ortho-P ratios are found during the winter season and the lowest during the growing season in summer (Fig. 7). There fore, in winter the algal biomass was probably limited by phosphorus and in the summers of 1983 and 1984 by nitrogen. In 1985 the ratios also indicate a moderate P-limited algal growth. The particulate N to P ratios indicate on the average no or a very moderate P-limitation for lake Elfhoeven and a severe P-limitation for lake Nieuwenbroek.

Table 3. Yearly averages of nutrient and chlorophyll to carbon ratios in the two sampled lakes and the critical values for limitation according to the literature.

Lake Elfhiseve n Lake Nieuwer ibroek Critical values for limitation References * ** 1983 1984 1985 1983 1984 1985 N N or P P

Inorganic N/P 3.8 9.8 14.9 16.8 19.0 68.1 < 5 5-12 > 12 a < 2.5 > 13 b < 12 12-15 > 15 c

Particulate N/P 9.1 11.0 7.8 30.3 30.4 23.6 > 10 (20) d

Total N/P 14.4 20.0 15.2 42.7 41.8 98.5 < 10 10-17 > 17 a < 10 > 18 b < 16 > 16 e < 22 > 22 c

Paniculate P/C 13.9 16.2 24.6 4.6 5.9 7.0 < 20 (10) d

Particulate N/C 121.7 162.3 168.4 118.8 143.3 162.3

Particulate Chl/C 11 .5 15.7 ■ 16.4 9.2 10.9 11.4 < 20 (10) d

* The values without brackets indicate moderate limitations; the values between brackets ( ) indicate severe limitations. ** References: a= Forsberg et a). (1978); b= Claesson & Forsberg (1980); c= de Vries & Klapwijk (1987); d= Healey & Hendzel (1980); e= Schmidt-van Dorp (1978).

Also the total N/P ratios of the lakes indicate no or a moderate P-limitation. for lake Elfhoeven and a clear P-limitation for lake Nieu wenbroek. Also these ratios tend to have higher values in winter and lower values in summer (not shown here), although the differences are smaller than with the inorganic N/P ratios. The particulate P to carbon ratios indicate sometimes (e.g. in 1983 and 1984) a moderate P-limitation for lake Elfhoeven and nearly always a severe P-limitation for lake Nieuwenbroek, whereas the particulate N to C ratios indicate sometimes, especially in summer, a moderate N-limitation for both lakes. Finally, the chlorophyll to particulate C ratios, which decrease especially in lake Nieuwenbroek in summertime to extremely low values (< 10; see also van der Vlugt & Klapwijk, 1987b) indicate a moderate (N or P) limitation for lake Elfhoeven and a strong (N or P) limitation for lake Nieuwenbroek. -193-

INORGANIC N / ORTHO-P RATIO 1983 - 1985 uu -

90 -

80 -

70 -

60 -

50 -

40 -

. STRONG 30 - _AP LIMITATION

20 - -AP UUITAT10N 10 -; i-iv s£j ~*N LIMITATION

1 1 1 1 1 1 T 1 1 1 1 I 1 1 I 1 1 I I 1 1 1 1 | 1 1 1 1 1 1 1 III 1983 1984 1985

■ LAKE ELFHOEVEN + LAKE NIEUWENBROEK

Fig. 7. The inorganic-N/ortho-P ratio in the Reeuwijk lakes during 1983-1985.

Although nutrient ratios are very simple to establish, their use in determining limiting factors is also subject to certain disadvantages. In the first place the critical ranges, established by comparison with bioassay results, have shown to be dependent on the algal species used in the bioassays (de Vries & Klapwijk, 1987). This can also ex plain the sometimes rather wide range of critical values for these ra tios in the literature (cf. Table 3). Therefore, using critical values for nutrient ratios from the literature, to determine the limiting factor(s) always leaves open the question whether the natural phytoplankton association is limited by the same nutrient or not. Further more, measurements of P/C, N/C and chlorophyll-a/C ratios have in common that they are related to carbon. Short term (i.e. several hours) changes in the light climate can cause considerable changes in carbohydrate content of algal cells, influencing these ratios dramati cally (Riegman & Mur, 1986). This applies even more where the car bon content is estimated as half of the AFDW. Therefore, N/P and other ratios are evidently only rough and general parameters to indi cate a limiting factor.

Statistical relationships

In Figure 8A-C the relationships between the summer averages of the total-P and chlorophyll concentrations, the total-N and chloro phyll concentrations and _the underwater light climate (H/So) and the amount of chlorophyll m~2 are presented over the years 1983-1985. The lines drawn in the figures represent the maximum values for these relationships according to the 1980 model (CUWVO, 1980; Hos- per, 1980). From Figure 8A it can be seen that the relationships in -194-

TOTAL-P - CHLOROPHYLL RELATIONSHIP (SUMMER MEANS)

140 - 130 - A. Y= -15.53+776.32X// ♦ 1983 120 - + 1984 110 -

100 - ♦ 1985 / Ü 1983 "VT984 90 - / o 1985 80 -

70 -

60 -

50 -

40 -

30 -

20 -

10 -

0 - 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.t6 0.18 0.2 TOTAL-P (mg/l) TOTAL-N - CHLOROPHYLL RELATIONSHIP (SUMMER MEANS)

140 - 130 - B. Y- -B9.6+89.6X / / 1983* 120 - / 1984* 110 i

100 - / D 1983 ♦ / ■> 1984 1985 90 - / ■> 1985 BO -1

70 -

60 - 50 -

40 -

30 -

20-

10 -

0 -

TOTAL-N (mg/l) H/So - CHLOROPHYLL RELATIONSHIP (SUMMER MEANS) c.

Y-530-40.38X

H / So (-) LAKE ELFHOEVEN ♦ LAKE NIEUWENBROEK

Fig. 8. Graphical presentation of the relationships between the total- P and chlorophyll concentration (A), the total-N and chloro phyll concentration (B) and the underwater light conditions (H/So) and the amount of chlorophyll per m2, averaged over the period April to September. The drawn lines respresent the maximum values for these relationships according to Hos- per (1980) and CUWVO (1980). -195- lake Elfhoeven in 1983 and 1984 occupy a position close to the 1980 line, representing the upper limit of the average chlorophyll concen tration at a certain P-concentration. Therefore, it can be assumed that the average algal biomass especially in 1983 but probably also in 1984 is limited largely by the phosphorus concentration. Lake Nieuwen broek occupies positions on the left side of the 1980 line, which illus-. trates the large amount of algal biomass in this lake in relation to the phosphorus concentration (van der Vlugt & Klapwijk, 1987a,b). The algal biomass in the lake therefore has to be severely limited by phos phorus. Figure 8B. shows that the relationships between total-N and chlorophyll in both lakes lie rather close to the line representing the upper limit of the average chlorophyll concentration at a certain N- concentration. Therefore it can be assumed that the average algal biomass in both lakes, especially in lake Elfhoeven is limited also by the nitrogen concentration. Finally Figure 8C, showing the relation ships between the depth divided by the corrected Secchi depth and the chlorophyll concentrations m~2 in both lakes illustrates that the average algal biomass in neither of the lakes is limited by light. This can also be concluded from Table 4 which gives the calcula ted average and maximum chorophyll concentrations in the two lakes over the years 1983-1985 based on the formulas describing the dif ferent relationships. Since the lowest calculated concentration per summer season indicates the most probable limiting factor (N, P or light) it can be concluded that the algal biomass in lake Elfhoeven was limited over the years 1983 to 1985 by P and/or N, while the biomass in lake Nieuwenbroek was limited by phosphorus alone.

Table 4. Measured and calculated average and maximum chlorophyll-a concentrations in summer according to statistical relation ships (Hosper, 1980; CUWVO, 1980) and the most probable limiting factor for the average algal biomass (N.B. the lo west calculated concentrations per year are underlined).

Lake Elfhoeven Lake Nieuwenb roek

1983 1984 1985 1983 1984 1985

Measured average Chi (mg m 3) 102 96 89 124 114 103

Calculated average Chi (mg m 3): based on total-P !2i 92 121 53 54 63 based on total-N 116 120 99 168 166 168 based on H/So 192 192 189 294 292 290

Probable limiting factor (s) •'■' P/n P n P P P

Measured maximum Chi (mg m 3) 168 285 146 222 230 164

Calculated maximum Chi (mg m 3): based on total-P 205 168 212 106 109 122 based on total-N 225 230 204 291 288 290 based on H/So 343 342 335 533 528 524

Probable limiting factor (s) "'•' P P n/p P P P

problably P-, resp. N-limited; P = clearly P-limited. -196-

The advantage of these statistical relationships is their simplici ty. Only summer concentrations of total-P, total-N, chlorophyll-a and the average Secchi depth in summer are needed to predict the maximal and mean algal biomass in summer and the limiting factors (including light). However, this method also has certain disadvantages. In the first place the method is very rough, because the assessment of the limiting factor is based on summer averages (mostly only six analyses.) and does not take into account the variations during the year. Se condly, the relationships are determined on a large number of lakes and are not made for any particular lake with its own specific chemi cal and algal species composition. Moreover, certain lakes, such as lake Nieuwenbroek (Fig. 8a), do not fit in the 1980 model, having chlorophyll-a concentrations far above the maximum Chl-a levels pre dicted by the Chl/P relationship. For such lakes, dominated by fila mentous blue-green algae, only special relationships are appropriate, as Janse et al. (.1987) showed. Anyway, the predictive value of such models is limited, since they cannot predict a change in the algal composition from filamentous bluegreens to other phytoplankton and vice versa. Therefore the results of such models have to be consi dered only as tentative.

Comparison of different methods

In order to evaluate the results obtained by the different me thods we assembled in Table 5 our conclusions in regard of the limi ting factors. The different methods do not seem to give rise to consis tent conclusions at first glance. However, on further inspection some interesting points can be seen from this table. In the first place, most of the methods, for example the nutrient ratios and the CUWVO rela tionships, revealed a more severe P-limitation for lake Nieuwenbroek than for lake Elfhoeven. Except for the particulate N/C ratio in both lakes and the statistical relationships in lake Elfhoeven in 1985, the remaining methods indicated generally a moderate P-limitation for lake Elfhoeven and a more severe P-limitation for lake Nieuwenbroek. This is in accordance with the P-load and P-concentrations in the lakes (cf. Table 1 and 2; van der Vlugt'& Klapwijk, 1987a, b). Also the P-uptake rates pointed to a more significant P-limitation in lake Nieuwen broek during 1983 and 1984 than in lake Elfhoeven. The results of both types of bioassays showed primarily a nitrogen limitation for both lakes, although from Figure 3 and 4 can be seen that both lakes, especially lake Nieuwenbroek also is secondarily phosphate limited. Secondly, when looking at the results of the N/P ratios in the course of a year it is interesting to see that in lake Elfhoeven a (moderate) .P-limitation is more significantly expressed in wintertime. In lake Nieuwenbroek such a phenomenon can also be observed when looking at the inorganic N/P ratio, which predicts even a (moderate) N-limitation in summer and a strong P-limition in winter (Fig. 7). Also the bioassay results show less pronounced N-limitations during the winter season (Fig. 3 and 4). This can easily be explained by the higher leaching and run-off of nitrogen from the agricultural polderland, resulting in higher inorganic nitrogen concentrations in the lakes (Fig. 2). See also van der Does and Klapwijk (1987). Whether the observed tendency to a more pronounced P-limitation in winter has to be regarded as real is not certain, since it is obvious that es pecially in winter algal growth can also be limited by physical factors such as temperature and light. Table 5. Summarizing presentation of limiting factors during 1983-1985 in lake Elf hoeven and lake Nieuwenbroek assessed by different methods.

LAKE ELFHOEVEN Year: 1983 1984 1985 Methods: Month: JFMAMJJASOND JFMAMJJASOND JFMAMJJASOND

Bioassays with-nat. population n n n n n n + + n + + X n - + - + + n N n N N N n N 'N + + N N N + Bioassays with Scenedesmus n n n n n n n n n n n n n n n N n n N N N n N n n N N N-■ N N n N N + N Ammonium uptake rate ' n n n n Phosphate uptake rate P P P P P P P Inorganic N/P ratio * P n n ~ n n P P P n P n n n P P P n P P Paniculate N/P ratio P " P " " P P P - P P P P Total N/P ratio - P P P P P P - P P - - - - n - - - - Particulate P/C ratio - P P P P P P P P P p P - P P P P P P P P P P - - - P P P - P Particulate N/C ratio n - - n n n n - n n n n - n - n n n n n - n n - - - - Chl/C ratio X X X X X X X X X X X X X X X X X X X X X X X X X X X - - - Statistical relationships (averages) X X X X X X P P P P P P n n n n n n Statistical relationships (maxima) P P P p P P P P P P P P X X X X X X (£> LAKE NIEUWENBROEK -3 Year: 1983 1984 1985 1 Methods: Month: J F M A M J J A s O N D J F M A M J J A s O N D J F M A M J J A s O N D

Bioassays with nat. population n n + n n n n n n + n n + + + n n + n N + N + N + n N n n N + + + + Bioassays with Scenedesmus n n n n n n n n n n n n n n n + n n N n n + + N n + N + + + + + + N + Ammonium uptake rate Phosphate uptake rate P P P P P P P P P P P P P P P P P P Inorganic N/P ratio P P P - n n n n - - - P P P P P - - n n n n -- P P P P P - - P P p p P P Particulate N/P ratio P P P P P P P P P P P P P P P P P P P P P P P P P P P P P P P P p p P P Total N/P ratio P P P P P P P P P P P P P P P P P P P P P P P P P P P P P P P P p p P P Particulate P/C ratio P P P P P P P P P P P P P P P P P P P P P P P P P P P P P P P P p p P P Particulate N/C ratio n n - n n n n n n n n n - n' n n n n n - - - - - n - - - n n n - - - - Chl/C ratio X X X X X X X X X X X X X X X X X X X X X X X X X X X X 'X X X X X X X X Statistical relationships (averages) P P P P P P P P P P P P p P P P P P Statistical relationships (maxima) P P P P P P P P P P P P p P P P P P n = nitrogen limited', p = phosphorus limited; + = nitrogen plus phosphorus limited; x = nitrogen or phosphorus limited; - = no limitation detected. Capital letters indicate strong limitation. -198-

When looking at the results of the various methods over the three years no consistent pattern towards more P or more N limitation appears. So the inorganic N/P ratios indicate an increase in P-limita tion in both lakes in 1985, while the same cannot be concluded from the other ratios. The statistical relationships show a tendency from P towards N limitation in lake Elf hoeven in 1985. In contrast, the bioassays from lake Nieuwenbroek reveal a tendency towards a less pronounced N-limitation and more P-limitation over the years (Fig. 3 and 4). Table 5 clearly demonstrates also that the applied methods did not always predict the same limiting nutrient(s). Especially striking are the results of the bioassay methods and the particular N/C ratios on the one side, which predict in general a primary N-limitation, while on the other hand all other methods predict generally a P-limi tation . This cannot be due to the pretreatment of autoclaving in the bioassay method, because the bioassays with the natural population, which have no such pretreatment showed primarily a N-limitation also, although sometimes less pronounced. From all the applied methods the bioassay method - especially that with the natural population - is in our opinion most directly in giving an answer to the question which nutrient in the water phase is limiting maximal algal yield. In this way it is probably also most rele vant to water management. All other methods e.g. the uptake rates, the ratio calculations and the statistical relationships are more indirect with respect to this question. The uptake rate method for instance measures rates rather than yields, while for the water quality mana gement yields (e.g. mean summer concentrations of chlorophyll-a) are more relevant than rates.

CONCLUSIONS In conclusion, the five methods used to assess growth limiting factors did not lead to the same results. Probably this is due for the most part to the fact that each method has its own specific goal and application and renders a more or less indirect measurement of the growth limiting factor(s). It has been shown that the bioassay me thods indicate mostly a N-limitation for the Reeuwijk lakes in compa rison with the other methods, which indicate generally a P-limitation. Therefore, in lake restoration studies aimed at reducing the P-load in order to achieve again a P-limited lake system, e.g. the Loosdrecht lakes project (van Liere, 1986) and the Reeuwijk lakes project (van der Vlugt & Klapwijk, 1987b), the bioassay method is rather conser vative in predicting a N-limitation, while the other methods indicate already a P-limited system. For that reason it seems advisable to ap ply several methods concurrently. It is not easy to say which of the methods has to be considered as most reliable and realistic. For that reason these methods should be compared to large-scale experiments, such as O'Brien and Denoyelles (1976) and Schindler (1975) did, which we could not do in the Reeu wijk lakes area. However based on this comparison and the discussion of the various methods, our conclusion is that from the applied me thods the bioassay method with the natural population is probably the most direct and closest to the field situation with respect to assessing the yield limiting factors. Moreover, it can be concluded that lake Nieuwenbroek is closer to a P-limitation than lake Elfhoeven is. -199-

ACKNOWLEDGEMENTS This study is part of an extensive limnological research program on the effects of phosphorus removal on the water quality of the Reeuwijk lakes, carried out in cooperation between the Technical Ser vice of the Waterboard of Rijnland and the National Institute of Public Health and Environmental Protection. This program is supported by the Directorate General of Environmental Hygiene of the Dutch Minis try of Housing, Physical Planning and Environment. The authors thank Dr. W. Admiraal, drs. J. van der Does', Prof. Dr. M. Donze, Prof. Dr. W.H.O. Ernst and Prof. Dr. L. Lijklema for valuable suggestions with respect to the manuscript, Mr. M. van der Putten and Mr. P. Kuijt for preparing the figures and Miss A. Honnef for correcting and Miss C. van Dijk for typing the English text.

Chiaudani, G. & M. Vighi, 1974. The N:P ratio and tests with Selenastrum to predict eutrophica tion in lakes. Water Res. 8: 1063-1069.

Chiaudani, G. & M. Vighi, 1975. Dynamic of nutrient limitations in six small lakes. Verh. internat. Verein. Limnol. 19: 1319-1324.

Chiaudani, G. & M. Vighi, 1976. Comparison of different techniques for detecting limiting or sur plus nitrogen in batch cultures of Selenastrum capricornutum. Water Res. 10: 725-729.

Claesson, A. & A. Forsberg, 1980. Algal assay studies of polluted lakes. Arch. Hydrobiol. 89: 208- 224.

CUWVO, 1980. Results of a second enquiry into eutrophication in the Nether lands. Report RIZA, Lelystad, the Netherlands (in Dutch).

Does, J. van der & S.P. Klapwijk, 1987. Effects of phosphorus removal on the maximal algal growth in bioassay experiments with water from four Dutch lakes. Int. Revue ges. Hydrobiol. 72: 27-39.

EPA, 1971. Algal assay procedure bottle test. National Eutrophication Re search Program. U.S. Environmental Protection Agency, Cor- vallis, Oregon. -200-

Forsberg, C, S-O. Ryding, A. Claesson & A. Forsberg, 1978. Water chemistry and/or algal assay? - Sewage effluent and pollu ted lake water studies. Verh., internat. Verein. Limnol. 21: 352- 363.

Gerhart, D.Z. & G.E. Likens, 1975. Enrichment experiments for determining nutrient limitation: Four methods compared. Limnol. Oceanogr. 20: 649-653.

Goldman, C.R., 1978. The use of natural phytoplankton populations in bioassay. Mitt. internat. Verein. Limnol. 21: 364-371.

Hanna, M. & A. Dauta, 1983. Bioassays: a comparative study of three parameters related to phosphorus bioavaüabüity (yield, growth rate and intraceUular concentration of phosphorus). Annals Limnol. 19: 59-66.

Healey, F.P. & L.L. Hendzel, 1979. Indicators of phosphorus and nitrogen deficiency in five algae in culture. J. Fish. Res. Board Can. 36: 1364-1369.

Healey, F.P. & L.L. Hendzel, 1980. Physiological indicators of nutrient deficiency in lake phytoplank ton. Can. J. Fish. Aquat. Sci. 37: 442-453.

Hosper, S.H., 1980. Development and practical application of limiting values for the phosphate concentration in surface waters in the Netherlands. Hydrobiol. BuU. 14: 64-72.

Janse, J.H., L. Lijklema & R.M.M. Roijackers, 1987. Vergelijkend onderzoek naar de eutrofiëring in Nederlandse me ren en plassen. Report Agricultural University Wageningen, Dept. Water Poll. Control.

Javornicky, P. 1979. Bioassays using a natural phytoplankton assemblage. In: P. Marvan, S. Pribü, O. Lhotsky. Algal assays and monitoring eutrophication, pp. 119-130. Schweizerbart'sche Verlagsbuch- handlung, Stuttgart.

Kappers, F.I., 1984. On population dynamics of the cyanobacterium Microcystis aeruginosa. Thesis, University of Amsterdam.

Klapwijk, S.P., G. Bolier & J. van der Does, in prep. Algal growth potential tests and limiting nutrients in the Rijnland Waterboard area. Prepared for: Int. Conf. Env. Bioassays. Lan caster, July 1988.

Lean, D.R.S. & F.R. Pick, 1981. Photosynthetic response of lake plankton to nutrient enrichment: A test for nutrient limitation. Limnol. Oceanogr. 26: 1001-1019. -201-

Liere, L. van, 1986. Loosdrecht lakes, origin, eutrophication, restoration and research programme, Hydrobiol. Bull. 20: 9-15. Ltfvstad, O. , 1984. Competitive ability of laboratory batch phytoplankton populations at limiting nutrient levels. Oikos 42: 176-184.

Miller, W.E., T.E. Maloney & J.C. Greene, 1974. Algal productivity in 49 lake waters as determined by algal as says. Water Res. 8: 667-679.

Miller, W.E., J.C. Greene & T. Shiroyama, 1978. The Selenastrum capricornutum Printz algal assay bottle test. EPA-600/9-78-018, U.S. Environmental Protection Agency, Cor- vallis, Oregon.

NichoUs, K.H. & P.J. Dillon, 1978. An evaluation of phosphorus-chlorophyll-phytoplankton relation ship for lakes. Int. Revue ges. Hydrobiol. 63: 141-154.

O'Brien, W.J. & F. Denoyelles, 1976. Response of three phytoplankton bioassay techniques in experi mental ponds of known limiting nutrients. Hydrobiologia 49: 65-76.

OECD, 1982. Eutrophication of waters. Monitoring, assessment and control. Organisation for economic co-operation and development, Paris.

Pettersson, K., 1980. Alkaline phosphatase activity and algal surplus phosphorus as phosphorus deficiency indicators in lake Erken. Hydrobiologia 89: 54-87.

Riegman, R., 1985. Phosphate-phytoplankton interactions. Thesis Univ. Amsterdam. Riegman, R. & L.R. Mur, 1986. Phytoplankton growth and phosphate uptake (for P-limitation) by natural phytoplankton populations from the Loosdrecht lakes (The Netherlands). Limnol. Oceanogr. 31: 983-988.

Ryding, S-O. (ed.), 1980. Monitoring of inland waters. Reports from the working group for eutrophication rechearch. Nordforsk publ. 1980: 2, Helsingfors, . Finland.

Schindler, D.W., 1975. Whole lake eutrophication experiments with phosphorus, nitrogen and carbon. Verh. internat. Verein. Limnol. 19: 3221-3231.

Schmidt-van Dorp, A.D., 1978. Eutrophication of shallow lakes in Rijnland. Report Technical Service, Waterboard of Rijnland, Leiden (in Dutch with an Eng lish summary). -202-

Smith, V.H., 1982. The nitrogen and phosphorus dependence of algal biomass in lakes: An empirical and theoretical analysis. Limnol. Oceanogr. 27: 1101-1112.

STORA, 1986. Ontwikkeling van een algengroeipotentietoets voor oppervlakte water en afvalwater. Report Stichting Toegepast Onderzoek Rei niging Afvalwater, Rijswijk. Thomas, E.A. , 1953. Zur Bekampfung der See-Eutrophierung: Empirische und experimentelle Untersuchungen zur Kenntnis der Minimumstoffe in 46 Seen der Schweiz und angrenzender Gebiete. Schweiz. Ver. Gas- Wasserfachmannern Monatsbull. 33: 25-32; 71-79.

Vlugt, J.C. van der & S.P. Klapwijk, 1987a. Water quality research in the Reeuwijk Lakes 1983-1985. H20 20: 86-91 (in Dutch with an English summary).

Vlugt, J.C. van der & S.P. Klapwijk, 1987b. Dose-effect relationships between phosphorus concentration and phytoplankton biomass in the Reeuwijk lakes (The Netherlands). Verh. internat. Verein. Limnol. 23: 482-488.

Vries, P.J.R. de, 1986. Bioassays on water quality using the attached filamentous alga Stigeoclonium Kiitz. Thesis Free Univ., Amsterdam.

Vries, P.J.R. de & S.P. Klapwijk, 1987. Bioassays using Stigeoclonium tenue Kiitz. and Scenedesmus quadricauda (Turp.) Bréb. as testorganisms; a comparative study. Hydrobiologia 153: 149-157.

Vries, P.J.R. de & P.S.H. Ouboter, 1985. Water sample treatments and their effects on bioassays using Stigeoclonium helveticum Vischer. Aquat. Bot. 22: 177-185.

Wurtsbaugh, W.A., W.F. Vincent, R. Alfaro Tapia, C.L. Vincent & P.J. Richerson, 1985. Nutrient limitation of algal growth and nitrogen fixation in a tro pical alpine lake, Lake Titicaca (Peru/Bolivia). Freshwater Biol. 15: 185-195.

Zevenboom, W., A. Bij de Vaate & L.R. Mur, 1982. Assessment of factors limiting growth rate of Oscillatoria agardhii in hypertrophic Lake Wolderwijd, 1978, by use of physiological indicators. Limnol. Oceanogr. 27: 39-52. -203-

CHAPTER 15:

SUMMARY AND CONCLUSIONS

"A summary, if provided, is for people who have already read the whole paper."

M. O'Connor & F.P. Woodford, 1976. Writing Scientific Papers in English, p. 24. Elsevier, Amsterdam. -204-

SUMMARY AND CONCLUSIONS This thesis deals with different limnological studies related to the eutrophication problem and carried out from 1977-1987 at the Rijnland Waterboard (The Netherlands). Eutrophication is the enrichment of surface water with nutrients to such a degree which may have serious consequences like algal blooms, decrease in transparency and some times even cause anaerobiosis and fish kills. In the Rijnland Water- board area eutrophication also has deteriorated the water quality of the lakes, canals and ditches. For that reason investigations were made into the factor(s) determining the trophic state of the surface waters in the area and to follow the effects of specific management measures like phosphate removal at sewage treatment plants. In the area the eutrophication problem has been studied already from 1973-1976 by Schmidt-van Dorp, who found that due to the high levels of total and inorganic phosphate in most of the lakes phospho rus was not a limiting nutrient for algal growth any more. She sug gested that in this area more attention should be paid to nitrogen re duction. Nevertheless it was generally concluded that eutrophication could be better combatted by phosphorus reduction than by nitrogen reduction for a number of reasons. Therefore the Waterboard of Rijnland carried out a large-scale phosphate-removal experiment at three sewage treatment plants (Gouda, Bodegraven and Nieuwveen) from 1979 to 1982 in order to re duce eutrophication in the lakes in the south-east of its area. Exten sive limnological research was accompanying this experiment to monitor the effects of the phosphate reduction on the algal growth in the lakes. Parts of this research are enclosed in this thesis (Chapters 7, 8, 9 and 11). Since it had to be concluded from this experiment that phosphate- removal at sewage treatment plants will not lead to an immediate de crease of algal biomass in the lakes, which form part of the basin system of Rijnland, later the attention has been focussed on the more isolated polder lakes, such as the Reeuwijk lakes, in which the exe cution of complete phosphorus, reduction measures may be more appro priate and effective. Part of this investigation are also presented in this thesis (Chapter 4 and 14).

The different studies included in this thesis are divided into three parts. They are prefaced with an General Introduction (Chapter 1) and each part is opened with an specific Introduction (Chapters 2, 6 and 10). Part A is focussed on phytoplankton, especially its use in asses sing water'quality in an ecological way (Chapter 3), while in Chapter 4 a comparison is made between historical and recent phytoplankton and chemical data with the purpose to look for ecological objectives for combatting eutrophication. In Chapter 5 a description is presented of the phosphate-phytoplankton relationships in the Reeuwijk lakes. Part B deals with sediments, especially the availability of sedi ment phosphates for algal growth. This availability is determined in four lakes with a rather laborious bioassay technique and compared with two chemical extraction techniques (Chapter 7), while in Chapter 8 a new and quicker technique is proposed, which is applied to the sediments of eight lakes in the Rijnland Waterboard area in Chapter 9. -205-

Part C treats bioassays, which are used to assess the algal growth potential and to determine the limiting nutrient(s) for algal growth in lakes and canals. This is both done with the aid of the na tural phytoplankton association (Chapter 11) and with testalgae like Scenedesmus quadricauda (Chapter 12). In Chapter 13 bioassays with two different testalgae, i.e. Stigeoclonium tenue and Scenedesmus quadricauda, and with different procedures are compared, while in Chapter 14 a comparison is made between two bioassay techniques and several other methods to assess growth limiting factors in the Reeuwijk lakes.

Part A: Phytoplankton

The phytoplankton composition in a surface water reflects the physical and chemical conditions and can therefore be used very well to assess the water quality in an ecological sense. So the phytoplank ton association of a water responds quite specifically to pollution, like discharges of organic material, causing saprobity, or discharges of inorganic material, causing eutrophication. About twenty years ago two german limnologists, Caspers and Karbe, presented a system for water quality assessment based on physiological and ecological aspects, that tried to combine the leading concepts on saprobity and trophism. They showed that both are running parallelly. In Chapter 3 an elabo ration is presented of Caspers and Karbe's water quality system for the larger waters (canals, lakes, peat lakes and deep sandpits and pools) in the western part of the Netherlands. Significant correlations have been established between several physical and chemical water quality parameters and phytoplankton community structures such as diversity and saprobity. Based on this a proposal is presented to quantify the water quality classes in Caspers and Karbe's scheme by measuring the bióactivity, oxygen regime and phytoplankton commu nity structure. The phytoplankton composition can also be used to show the con sequences of eutrophication in time. By comparing the phytoplankton and chemical data from 1941 and 1942 with recent data it has been established in Chapter 4 that the inorganic nitrogen and orthophosphate concentrations in canals and lakes in the Rijnland area have been increased strongly in the last 45 years. The inorganic N/P ratio has been dropped, especially in the Gouwe canal, where water is taken in from the Hollandsche Ussel, a branch of the river Rhine. This indicates that the limiting nutrient factor for the algal growth in the intake water of the Rijnland Waterboard area has probably been changed from phosphorus to nitrogen over the last 45 years. Moreover, it has been shown that the average seston volume, measured by filtrating 100 liter water through a planktonnet, has generally been doubled. Although in 1941/1942 the blue-green alga Microcystis aeruginosa Kiitz. was blooming regularly too, the phyto plankton composition was shown to be impoverished in the last decen nia, since several taxa have disappeared and some others are strongly reduced in numbers. The saprobic degree, measured with Pantle and Buck's index, has not been changed significantly. By using the relationships between chlorophyll-a and BOD and between transparency and seston volume, BOD, dry weight and ash free dry weight, which were established on data from 1986 and 1987, -206- an attempt has been made to assess the chlorophyll-a concentrations and transparency values in 1941 and 1942. It is made plausible that the average chlorophyll-a content in the lakes is doubled or tripled in the last 45 years and that the mean transparency in the Gouwe canal is reduced from 75 to 50 cm. This can be used to develop ecological objectives for combatting eutrophication in canals and lakes and possi bly even the lower course of the river Rhine. In two parts of the Reeuwijk lakes area, lake Elf hoeven and lake Nieuwenbroek, dose-effect relationships have been investigated be tween the phosphorus concentration and the phytoplankton biomass (Chapter 5). It is demonstrated that the two parts are rather diffe rent. Lake Elfhoeven has a relatively high P-load and high phospho rus concentration, while lake Nieuwenbroek has a lower P-load and P-concentration. However, lake Elfhoeven is showing the lowest algal biomass, consisting of greenalgae and diatoms, while lake Nieuwen broek appears to have the highest algal biomass consisting almost enti rely of filamentous blue-green algae. The carbon/chlorophyll-a ratios also proved to be different in both lakes. Another striking difference is that in lake Elfhoeven a relatively high zooplankton biomass occurs, while in lake Nieuwenbroek zooplankton is almost absent. This proves that the relationship between phosphorus concentration and phyto plankton biomass is not so simple as is sometimes supposed and that other factors, like species composition of the phytoplankton and gra zing by zooplankton, can also be determining for the phytoplankton biomass in lakes.

Part B: Sediments

The sediments of lakes can be substantial sources of phosphorus and are for that reason important in eutrophication studies. Before expensive management measures are taken for the reduction of the external P-load, the contribution of the internal nutrient source from the sediments should be considered. Several ways of estimating the phosphorus release from sediments are applied in limnology, such as chemical extraction techniques and bioassays with sediment. In Chapter 7 the phosphorus available for algal growth in four lakes in the Rijnland area (Braassem lake, Westeinder lakes, Nieuw koop lakes and Reeuwijk lakes) has been assessed by means of two extraction techniques, viz. a NTA column method and a stepwise NH4Cl-NaOH-HCl shaking method, and one bioassay technique, using Scenedesmus quadricauda (Turp.) Bréb. as testalga and sediment as the sole source of P. The mean phosphorus concentration of the in vestigated lake sediments varied from 0.8-3.6 mg P/g dry weight, of which 0.4-36% proved to be available for algal growth in the bioas says. The lowest availability was found in the Nieuwkoop lakes. The chemical extraction techniques produced on average 36-69% (NTA co lumn) and 27-62% (NaOH extraction) of the total phosphorus, although it is likely that NaOH extracted from peaty soils also large amounts of non-available organic phosphorus. Anyhow it was demonstrated, that the sediments of the investigated lakes showed great variability and that the chemical extraction techniques cannot replace the bioassays to assess the amount of phosphorus available to algae. The bioassay technique with algae and sediment used normally has the disadvantage, that it is rather time-consuming. The following -207-

of the algal growth by means of cell counting is very laborious, so an other technique was looked for. This was found in Chapter 8 by ap plying derivative spectroscopy, a technique which can differentiate between turbidity in samples caused by algae and caused by sediment. This much quicker method has been tested and found to be very ac curate and sensitive, even at low algal densities and high sediment turbidities With the aid of this derivative spectroscopy technique the sedi ments of eight lakes in the Rijnland Waterboard area have been inves tigated in order to estimate the potential release of phoshorus from sediments (Chapter 9). The average total-P concentration of the top 5 cm sediment layer varied from 1.1-5.7 mg/g dry weight. Based on the bioassays with sediment, available phosphorus varied from 10-45% of total P. The uptake by algae was proven to increase strongly with rising pH. Among the P-binding components in the sediment (Fe, CaC03, clay and organic matter), only the organic matter content was shown to diminish the availability of the sediment phosphates. The amount of available P was highly positively correlated with the ortho- P concentration in the interstitial water and in the water immediately above the sediment. This provided a quick estimate of the available sediment P.

Part C: Bioassays

In studies related to the eutrophication problem it is important to know the possible algal growth (the so-called algal growth potential "AGP") in surface waters and to establish the growth limiting factor(s) for phytoplankton in these waters. This applies especially for lake restoration programs, where the effects of phosphorus load re duction have to be predicted with some precision. One of the most popular techniques to measure algal growth potential and to establish the limiting nutrients is the bioassay technique. The principle of the bioassay method is based on von Liebig's Law, which states that the yield (biomass) of a species (or population) is determined by the nutrient, which is present in the least available quantity in relation to the demand. Addition of this nutrient will raise the yield. With such bioassays the algal growth potential and the limi ting nutrient(s) can be established. This is mostly done under standarized conditions in the laboratory. Bioassays can be carried out both with the indigenous phytoplankton population as well as with special testalgae. In Chapter 11 the first type of bioassays are described carried out from 1980-1982 with water from four lakes (Braassem lake, Westeinder lakes, Nieuwkoop lakes and Reeuwijk lakes) in the Rijnland Waterboard area. It was shown that no significant lowering was detec ted. of the phosphate concentration and the algal growth potential in the investigated lakes as a result of phosphorus removal at sewage treatment plants. All lakes proved to be principally nitrogen limited except the Reeuwijk lakes, which showed clearly, after a primary ni trogen limitation, a secondary phosphorus limitation. It was concluded that the main attention with respect to phosphorus reduction should be concentrated on the Reeuwijk lakes in the first place. For the other lakes in the investigated area it was inferred that phosphorus removal will, when it is the only measure taken, presumably not lead at short notice to a decrease of the algal biomass. i -208-

In Chapter 12 the results are presented of a large number of bioassays with Scenedesmus quadricauda (Turp.) Bréb. as testorganism, carried out with several surface waters in the Rijnland area from 1983-1986. The bioassays are used to assess the algal growth potential and to determine the limiting nutrients. Special attention has been paid to the effects of deep-freezing and autoclaving as pretreatment on pH and nutrient concentrations. It was shown that pH and particulate P increased, while N-Kjeldahl, NH4-N, particulate N, total-N, ortho-P and total-P decreased significantly by the pretreatment. The algal growth potential of the various sampled waters was proven to range from very low yields in the relatively isolated polder lakes to very high yields in the canals and lakes, which form part of the basin system of Rijnland. From the bioassays with nitrogen and phosphorus enrichments it is concluded that the lowest yields are ge nerally observed in nitrogen and phosphorus co-limited waters, while the highest yields are found in waters limited by nitrogen alone. Very high correlation coefficients have been demonstrated between the algal growth potential and inorganic and total nitrogen concentrations, from which could be concluded that the algal growth potential is primarily determined by the amount of nitrogen in the samples and only secon darily by the amount of phosphorus. From a graphical presentation of the yields against the N/P ratios in the samples critical ranges for nitrogen and/or phosphorus limitation could be derived. The ranges indicating phosphorus limitation turned out to be considerably higher than the ranges reported in the literature so far, which implicates that phosphorus limitation in the Rijnland surface waters can only be achieved at relatively high N/P ratios and low phosphorus concentra tions. It was shown that, once the relation between AGP and nutrient concentrations are established, as in this study, AGP tests do not have to be carried out to monitor algal growth potential of surface waters on a routine basis. They still can be very useful in special studies, e.g. in lake restoration projects and to detect possible toxi cants in effluents and wastewater discharges. In Chapter 13 thr,ee different bioassay techniques, two culture tube test methods with respectively Stigeoclonium tenue Kütz, or Sce nedesmus quadricauda (Turp.) Bréb. and one bottle test with S. qua dricauda, are compared. The yields obtained in the various tests proved to be linearily correlated. The same primary limiting nutrient was indicated by the bioassays in most cases. However the algal growth in the tube test using Stigeoclonium was more often P-limited. In the case of S. quadricauda both test methods (tube and bottle) were nearly equally effective. The yields of N-limited samples were significantly correlated with the inorganic-N as well as total-N con centration of the water samples. A significant correlation of the or tho-P as well as total-P concentration with the yield of the P-limited assays was only found for Stigeoclonium tenue. The critical N/P ratio (by weight) for N of P limitation proved to be approximately 17:1 for Stigeoclonium tenue and 22:1 for Scenedesmus quadricauda. This should be kept in mind, while applying one of these test algae in field studies to evaluate the effects of P-reduction measures. Finally, in Chapter 14 five different methods to asses.s growth limiting factors for algae are compared and applied to data from two parts of the Reeuwijk lakes area (lake Elf hoeven and lake Nieuwen broek) over the years 1983-1985: (1) bioassays with the natural phytoplankton population, (2) bioassays with Scenedesmus quadricauda

/ -209-

(Turp.) Bréb., (3) ammonium and phosphate uptake experiments, (4) nutrient ratios and (5) statistical relationships. It was shown that the different methods did not lead to the same results. Both types of bioassays and the N/C ratios predicted in general a primary N-limitation for the Reeuwijk lakes, while all other methods predicted generally a P-limitation. The reason for this is not quite clear. Several possibili ties are discussed and the various methods are evaluated. Based on this it is suggested to apply in lake restoration studies aimed at re ducing the P-load several methods simultaneously. Furthermore it is concluded that the bioassay method with the natural phytoplankton population is probably the most direct and closest to the field situa tion with respect to assessing the growth limiting factors. Despite the variable results of the different methods it is evident that lake Nieuwenbroek with the lowest P-concentrations but with the highest algal biomass is probably closer to a P-limitation than lake Elfhoeven with the highest P-concentrations and the lowest algal biomass. There fore P-reduction measures will probably show more and quicker re sults in the first lake.

-211-

CHAPTER 16:

SAMENVATTING EN CONCLUSIES

"Een samenvatting bevat de inhoud van het hele rapport in zeer be knopte vorm. Deze 'bijkomstige' tekst is bedoeld voor de lezer die geen tijd of geduld heeft om het hele rapport grondig te bestuderen, of die na het lezen het geheel nog eens in grote lijnen wil overzien."

H. de Boer et al., 1972. Schriftelijk rapporteren. Aula-boeken nr. 54, p. 90. Uitgeverij Het Spectrum N. V., Utrecht. -212-

SAMENVATTING EN CONCLUSIES

Dit proefschrift beschrijft verschillende onderzoekingen naar de eutrofiëring van meren, plassen en kanalen in het westen van Neder land. Eutrofiëring is de overbemesting van het oppervlaktewater met voedingsstoffen, met name fosfaten en stikstofverbindingen, met als gevolg een toename van de groei van de algen en waterplanten. Het is in essentie een natuurlijk proces, maar door diverse menselijke activi teiten, zoals de aanleg van waterleiding en riolering en het gebruik van kunstmest en fosfaathoudende wasmiddelen, is de eutrofiëring in deze eeuw in een stroomversnelling geraakt. Dit heeft geleid tot aller lei ongewenste verschijnselen in het oppervlaktewater, zoals het op treden van algenbloeien, een vermindering van het doorzicht in meren en plassen en soms zelfs zuurstofloosheid gevolgd door vissterfte. Evenals in andere delen van de wereld, zoals in de Great Lakes in Noord-Amerika, zijn ook vele Europese meren, zoals de Zwitserse en Italiaanse Alpenmeren en vele Oost-Europese en Scandinavische meren, door de eutrofiëring sterk in kwaliteit achteruitgegaan. In de jaren zestig vestigde met name Dr. Golterman de aandacht op de snel toenemende eutrofiëring van de Nederlandse oppervlakte wateren. Hij was de initiatiefnemer van een Stuurgroep Fosfaten bin nen de Koninklijke Nederlandse Chemische Vereniging, die een rapport samenstelde over de oorzaken en gevolgen van eutrofiëring in Neder land en die maatregelen suggereerde die genomen zouden moeten wor den om de fosfaatbelasting van onze binnenwateren te verminderen. Dit rapport was mede aanleiding tot de zogenoemde Fosfatennota van de ministeries van Volksgezondheid & Milieuhygiëne en Verkeer & Waterstaat in 1979, waarin het regeringsbeleid inzake de eutrofiërings- bestrijding werd uitgelegd. Een reductie van de fosfaatniveaus zou bereikt kunnen worden door fosfaatverwijdering op zuiveringsinstalla ties en door vervanging van fosfaten in wasmiddelen. In het gebied van het Hoogheemraadschap van Rijnland werd de eutrofiëring eerder (van 1973-1976) bestudeerd door mevr. Schmidt- van Dorp, die aantoonde dat dankzij de hoge niveaus aan fosfaat in de meeste meren in het gebied, fosfaat niet meer de beperkende fac tor voor de algengroei was, maar dat deze rol door stikstof was over genomen. Zij stelde dat in dit gebied meer aandacht aan stikstof re ductie gegeven zou moeten worden. Algemeen werd echter aangenomen dat de eutrofiëring toch beter via fosfaatbeperking aangepakt zou kunnen worden, omdat ten eerste stikstof veel meer verspreid (o.a. door de landbouw) wordt geloosd dan fosfaat, waardoor het moeilijker te beheersen is en ten tweede omdat in de meeste wateren stikstof bindende blauwwieren voorkomen, die stikstof uit de lucht kunnen aanwenden voor hun groei. Bovendien moet niet uit het oog worden verloren dat stikstof een beperkende factor voor de algengroei is geworden juist doordat in de laatste tientallen jaren zoveel fosfaat is geloosd. Om die reden heeft het hoogheemraadschap van Rijnland van 1979 tot 1982 een grootschalig defosfateringsexperiment uitgevoerd op drie rioolwaterzuiveringsinstallaties, namelijk te Gouda, Bodegraven en Nieuwveen, met het doel om de eutrofiëring in het zuidoosten van het gebied een halt toe te roepen. Vanwege het proefkarakter van de fos faatverwijdering en vanwege de ingewikkeldheid van het eutrofiërings- probleem, werd dit experiment begeleid door uitgebreid hydrobiolo- -213- gisch onderzoek met het doel de gevolgen van de fosfaatverwijdering nauwkeurig te beschrijven. Delen van dit onderzoek zijn opgenomen in dit proefschrift (Hoofdstukken 7, 8, 9 en 11). Aangezien op grond van bovengenoemd experiment geconcludeerd moest worden dat alleen fosfaatverwijdering op zuiveringsinstallaties niet zal leiden tot een vermindering van de algengroei in de meren, die onderdeel uitmaken van Rijnlands boezemstelsel, werd de aandacht verschoven naar de waterstaatkundig meer geïsoleerd liggende polder- plassen, waarin de uitvoering van integrale maatregelen ter bestrij ding van de eutrofiëring naar verwachting wel effectief zijn. Daarom zijn integrale eutrofiëringsbestrijdingsprojecten ontwikkeld voor de Reeuwijkse, de Nieuwkoopse en de Langeraarsé plassen (Geerpias) met het doel om alle mogelijke fosfaatbronnen gelijktijdig aan te pakken. Delen van het hieraan ten grondslag liggende onderzoek zijn eveneens opgenomen in dit proefschrift (Hoofdstukken 5, 12 en 14).

Dit proefschrift handelt over het hydrobiologische onderzoek, dat de laatste tien jaar is uitgevoerd bij het Hoogheemraadschap van Rijn land in het kader van de eutrofiëringsproblematiek. Het is verdeeld in drie delen: Deel A is gericht op fytoplankton, de in het water zwe vende microscopisch kleine plantaardige organismen (meestal algen), dat als gevolg van eutrofiëring sterk in aantal kan toenemen en qua soortensamenstelling kan veranderen. Deel B handelt over het uitge voerde sedimentonderzoek, met name over de in de bodems van meren voorkomende hoeveelheden fosfaat en de beschikbaarheid daarvan voor de algengroei. Deel C gaat over eutrofiëringstoetsen, zogenaamde bioassays, die gebruikt zijn om de groeikracht van diverse wateren te bepalen en om de voor dat water beperkende factor(en) voor de al gengroei vast te stellen. De drie delen worden voorafgegaan door een algemene inleiding (Hoofdstuk 1), terwijl elk deel begint met een spe cifieke inleiding (Hoofdstukken 2, 6 en 10).

Deel A: Fytoplankton

De in het water aanwezige fytoplanktonorganismen kunnen ge bruikt worden om iets te zeggen over de kwaliteit van het water. Ze reageren namelijk zeer specifiek op bepaalde verontreinigingen, zoals saprobiëring (verontreiniging met organische stoffen) en eutrofiëring (verontreiniging met anorganische stoffen), twee typen waterveront reiniging die nog vaak voorkomen in de praktijk van het waterbeheer. Twee duitse hydrobiologen, Caspers en Karbe, hebben twintig jaar geleden laten zien dat beide verschijnselen nauw met elkaar verbonden zijn: afbraak van organische stof levert anorganische stoffen op (o.a. fosfaten en stikstofverbindingen), die door algen weer gebruikt kun nen worden voor de opbouw van organische stof (= groei). Op grond hiervan is door Caspers en Karbe een waterkwaliteitssysteem opgezet, dat echter door hen zelf nooit is uitgewerkt tot een in de praktijk bruikbaar systeem voor het beoordelen van de waterkwaliteit. In dit proefschrift (Hoofdstuk 3) is een uitwerking gepresenteerd van het waterkwaliteitssysteem van Caspers en Karbe voor de grote wateren in West-Nederland (meren, kanalen, plassen, diepe putten). Duidelijke verbanden zijn hierbij aangetoond tussen verschillende fysische en chemische eigenschappen van het water en de samenstelling van de in dat water aanwezige fytoplanktonlevensgemeenschappen. Op grond -214- hiervan is een voorstel gedaan voor een concrete invulling van het systeem van Caspers en Karbe met grootheden die de opbouw en de afbraak, de zuurstofhuishouding en de samenstelling van de plankton levensgemeenschap in het water beschrijven. De fytoplanktonsamenstelling kan ook gebruikt worden om het verloop van de eutrofiëring in de afgelopen decennia aan te geven. Door een vergelijking te maken tussen chemische analyses en plank tonanalyses uit 1941 en 1942 met recente gegevens is in Hoofdstuk 4 aangetoond dat de (anorganische) stikstof- en fosfaatgehalten in kana len en meren in Rijnlands boezemstelsel in de afgelopen 45 jaar sterk zijn gestegen. De stikstof/f osfo rverhou ding is gedaald, met name in de Gouwe bij Gouda, alwaar Rijnland water inlaat uit de Hollandsche IJssel. Dit wijst er op dat de beperkende factor voor de algengroei in het ingelaten water in de afgelopen 45 jaar verschoven is van fosfaat naar stikstof. Omdat uit 1941/1942 geen gegevens bekend zijn over de algenhoeveelheden en de helderheid van de Rijnlandse wateren is via een indirecte methode, door gebruik te maken van verbanden tussen het chlorofyl-a-gehalte (een maat voor de hoeveelheid algen) en het doorzicht enerzijds en verschillende andere grootheden die iets zeggen over de algenbiomassa anderzijds, een poging gedaan om te berekenen hoeveel algen 45 jaar geleden aanwezig waren en hoe helder het water toen was. Hieruit is gebleken dat de algenhoeveelheden in de boezem meren verdubbeld tot verdrievoudigd is en dat het doorzicht in de Gouwe is gedaald van 75 naar 50 cm. Deze resultaten kunnen worden gebruikt om ecologische normen te ontwikkelen voor kanalen en meren en mogelijk zelfs voor de benedenloop van de rivier de Rijn. In twee delen van het Reeuwijkse plassengebied (Elfhoeven en Nieuwenbroek) is bijzondere aandacht gegeven aan het verband tussen de fosfaatconcentratie en de hoeveelheid fytoplankton (Hoofdstuk 5). De twee deelplassen blijken nogal verschillend te zijn in hydrobiologisch opzicht. Elfhoeven heeft een hoge fosfaatbelasting en een rela tief hoge fosfaatconcentratie, terwijl Nieuwenbroek een lage fosfaat belasting en een lage fosfaatconcentratie heeft. Wat de hoeveelheid algen betreft is het juist andersom: Elfhoeven vertoont de laagste algengroei, bestaande uit een redelijk gevarieerde planktonsamenstel ling van o.a. groenalgen en kiezelwieren, en Nieuwenbroek de hoogste, vrijwel uitsluitend bestaand uit draadvormige blauwwieren. Waarschijn lijk heeft dit verschil in planktonsamenstelling te maken met een ander verschil tussen deze twee plassen: In plas Elfhoeven wordt relatief veel dierlijk plankton aangetroffen, terwijl dierlijk plankton in plas Nieuwenbroek over het algemeen vrijwel afwezig is. Hieruit blijkt dat het verband tussen fosfaat en planktonhoeveelheid niet zo eenvoudig is als soms wordt verondersteld en dat andere factoren, zoals de soor tensamenstelling van het fytoplankton en de graas door zooplankton eveneens bepalend kunnen zijn voor de algengroei in meren.

Deel B: Sediment

De bodems van meren en plassen kunnen aanzienlijke bronnen van fosfaat zijn en spelen daarom een belangrijke rol in de eutrofië- ringsproblematiek. Daarom moet, voordat kostbare maatregelen geno men worden om de eutrofiëring terug te dringen, de bijdrage van de bodem aan het in stand houden van de algengroei kritisch worden be keken. Hiervoor bestaan verschillende technieken, zoals chemische -215-

extractiemethoden en laboratoriumproeven met algen (bioassays) waar bij sediment als enige fosfaatbron wordt toegevoegd. In Hoofdstuk 7 zijn de resultaten beschreven van experimenten waarmee het voor algen beschikbaar fosfaat is bepaald in vier Rijn landse meren (Braassemermeer, Westeinderplassen, Nieuwkoopse plas sen en Reeuwijkse plassen). Dit is gedaan met behulp van twee che-. mische extractiemethoden, namelijk een NTA kolom techniek en een stapsgewijze NH4Cl-NaOH-HCl techniek, en met een bioassay methode, waarbij gebruik gemaakt is van de testalg Scenedesmus quadricauda (Turp.) Bréb.. De gemiddelde fosfaatconcentratie van de meerbodems varieerde van 0,8 tot 3,6 mg P per gram droge grond. De beschik baarheid van deze fosfaten voor algen varieerde in de bioassayproeven gemiddeld van 0,4 tot 36%, terwijl met de NTA-methode 36-69% fosfaat werd. geëxtraheerd, dus veel meer dan in de bioassays. Met NaOH werd gemiddeld 27-62% fosfaat geëxtraheerd, hoewel deze methode, met name in veenachtige bodems, ook grote hoeveelheden niet-beschikbaar organisch fosfaat zal hebben meebepaald. In ieder geval is aan getoond dat de bodems van de onderzochte meren grote verschillen vertonen en dat de chemische extractie-methoden niet de bioassays kunnen vervangen als het er om gaat de hoeveelheid voor algen be schikbaar fosfaat te bepalen. Bovengenoemde bioassays met algen en sediment hebben als na deel, dat ze vrij bewerkelijk zijn. Met name het volgen van de algengroei door middel van algentellingen met behulp van de microscoop is zeer tijdrovend, zodat werd omgezien naar een andere minder bewer kelijke techniek. Deze werd gevonden door gebruik te maken van af geleide spectroscopie, een techniek waarmee de hoeveelheid algen ge meten kan worden in monsters die vertroebeld zijn door bijvoorbeeld modder. Deze methode is uitgetest in Hoofdstuk 8 en zeer gevoelig en nauwkeurig gebleken, zelfs bij lage algendichtheden in sterk door humus verkleurde monsters. Tevens is met deze techniek aangetoond dat een verhoging van de zuurgraad in de bioassays met sediment leidt tot een aanzienlijke verhoging van de hoeveelheid beschikbaar fosfaat. Met behulp van de bovengenoemde afgeleide spectroscopie-techniek zijn de bodems van acht Rijnlandse meren onderzocht om de po tentiële afgifte van fosfaat te kunnen schatten (Hoofdstuk 9). De ge middelde fosfaatconcentratie in de bovenste 5 cm van de meerbodems bedroeg 1,1 tot 5,7 mg P per gram droge grond, terwijl 10 tot 45% hiervan beschikbaar bleek voor algen. Een hoog gehalte aan organi sche stof bleek samen te hangen met een lage beschikbaarheid van de sedimentfosfaten. De laagste gehaltes totaal en beschikbaar fosfaat werden aangetroffen in de Nieuwkoopse en de Reeuwijkse plassen. Verder is een duidelijk verband aangetoond tussen de hoeveelheid voor algen beschikbaar fosfaat en de (ortho-)fosfaat-concentratie in het poriënwater van de bodem en in het water vlak boven het sedi ment. Dit laatste kan een snelle schatting opleveren van het voor al gen beschikbaar fosfaat in de bodem.

Deel C: Bioassays

In het kader van eutrofiëringsonderzoek is het belangrijk om de groeikracht van oppervlaktewateren te kennen en om de groeibeperkende factoren voor algengroei in deze wateren vast te stellen. Dit -216-

geldt des te meer in eutrofiëringsbestrijdingsprojecten, waar de effec ten van fosfaatverlaging met enige nauwkeurigheid moeten worden voorspeld. Eén van de meest gebruikte methoden om de groeikracht van water te bepalen en om de groeibeperkende factoren voor algen vast te stellen is de bioassay-techniek. Het principe van deze techniek is gebaseerd op de wet van het minimum van de landbouwkundige von Liebig uit de vorige eeuw, die stelde dat de opbrengst van een gewas bepaald wordt door dié voe dingsstof, die verhoudingsgewijs het minst aanwezig is. Toevoeging van die voedingsstof zal de opbrengst vergroten. Op deze wijze kan de beperkende factor(en) worden opgespoord. Dit geldt niet alleen voor landbouwgewassen, maar ook voor algen in het water. Bioassays of eutrofiëringstoetsen kunnen in het laboratorium zowel met de in het water van nature aanwezige algenpopulatie worden uitgevoerd als met speciaal gekweekte testalgen. In Hoofdstuk 11 zijn bioassays beschreven, waarbij gebruik is gemaakt van de in het water aanwezige algenpopulatie. Hiermee zijn de gevolgen van de in de periode 1979-1982 uitgevoerde fosfaatver wijdering op zuiveringsinstallaties nagegaan. Hierbij bleek dat geen duidelijke verlaging aangetoond kon worden van de fosfaatconcentratie en de groeikracht in het water van vier meren in Rijnlands gebied (Braassemermeer, Westeinderplassen, Nieuwkoopse plassen, Reeuwijkse plassen). De maximale algengroei in deze meren bleek voornamelijk door stikstof beperkt te zijn, behalve in de Reeuwijkse plassen, waar naast een primaire stikstofbeperkirig ook een secundaire fosfaatbeper king werd aangetoond. Geconcludeerd werd daarom, wat fosfaatver mindering betreft, de aandacht vooral te richten op de Reeuwijkse plassen. In Hoofdstuk 12 zijn een groot aantal bioassays beschreven, uit gevoerd met behulp van de testalg Scenedesmus quadricauda (Turp.) Bréb., in water afkomstig van 24 locaties in het gebied van Rijnland. Bij deze experimenten is speciale aandacht gegeven aan de effecten van de voorbehandeling van het water, namelijk diepvriezen en steri liseren. Een aantal chemische eigenschappen bleek door de voorbehan deling duidelijk te veranderen. Gemiddeld verdween ongeveer 4% stik stof en 16% fosfaat tijdens de behandeling. In hoeverre dit de resulta ten van de bioassays beïnvloedt is de vraag. De groeikracht van de verschillende bemonsterde wateren bleek te variëren van zeer laag in de meeste polderpiassen tot zeer hoog in sommige kanalen en meren, die deel uitmaken van Rijnlands boezem systeem. Uit verrijkingsexperimenten met fosfaat en stikstof is gecon cludeerd, dat de laagste groeikracht te vinden is in wateren, waarin de algengroei zowel door stikstof als door fosfaat wordt beperkt, ter wijl de hoogste groeikracht wordt aangetroffen in wateren, die uit sluitend door stikstof worden gelimiteerd. Zeer goede verbanden zijn aangetoond tussen de groeikracht van het water en de in het water aanwezige stikstofverbindingen. Hieruit mag worden geconcludeerd dat de groeikracht in de betreffende wateren vrijwel uitsluitend wordt bepaald door de hoeveelheid stikstof en slechts in veel mindere mate door de hoeveelheid fosfaat. Verder is vastgesteld bij welke stikstof/ fosfor-verhoudingen de algengroei in de Rijnlandse wateren wordt be perkt door voornamelijk stikstof, stikstof en fosfaat samen of vooral fosfaat. Hieruit is gebleken dat fosfaatbeperking in deze wateren al leen bereikt wordt bij relatief hoge stikstof/fosfor-verhoudingen en -217- lage fosfaatconcentraties. Verder is geconcludeerd dat, als de relaties tussen de groeikracht en de voedingsstof concentraties eenmaal bekend zijn, zoals na dit onderzoek, eutrofiëringstoetsen niet meer routine matig hoeven te worden uitgevoerd. Voor speciale studies, zoals in eutrofiëringsbestrijdingsprojecten en om mogelijk giftige stoffen in effluenten en afvalwaterlozingen op te sporen, kunnen ze wél nuttig zijn. In Hoofdstuk 13 is een vergelijking gemaakt tussen bioassays met twee verschillende technieken en twee soorten testalgen, namelijk Sti- geoclonium tenue Kütz. en Scenedesmus quadricauda (Turp.) Bréb.. De eerste soort wordt representatief geacht voor het slootmilieu in Nederland, terwijl de tweede zeer algemeen voorkomt in het plankton van grotere wateren. Uit de vergelijking bleek dat beide methoden en soorten vrij goed vergelijkbaar zijn, maar dat Stigeoclonium tenue, iets vaker een fosfaatbeperking aangeeft. Ook in deze experimenten bleek dat de onderzochte wateren eerder door stikstof dan door fosfaat ge limiteerd zijn. Tenslotte is in Hoofdstuk 14 een uitgebreide vergelijking gemaakt tussen vijf verschillende methoden om groeibeperkende factoren voor algen vast te stellen in twee delen van de Reeuwijkse plassen (Elf- hoeven en Nieuwenbroek), namelijk (1) bioassays met de natuurlijke populatie, (2) bioassays met testalgen, (3) ammonium en fosfaat-op name experimenten, (4) verhoudingen tussen voedingsstoffen onder ling en (5) relaties tussen chlorofyl-a en mogelijk beperkende facto ren. Uit de vergelijking blijkt dat de verschillende methoden niet al lemaal dezelfde resultaten opleveren. De bioassays en de stikstof/ koolstof-verhoudingen geven over het algemeen een stikstofbeperking aan voor de Reeuwijkse plassen, terwijl alle andere methoden eerder wijzen op een fosfaatbeperking. De redenen hiervoor zijn niet geheel duidelijk. Verschillende mogelijkheden zijn nader beschouwd. Aange raden wordt om in projecten, die tot doel hebben de fosfaatbelasting van meren te verminderen, verschillende technieken naast elkaar toe te passen. Uit dit vergelijkende onderzoek komt echter wel de bioassay-methode met de natuurlijke algenpopulatie als de meest direkte en waarschijnlijk ook de makkelijkst naar de veldsituatie te vertalen methode voor het bepalen van groeibeperkende factoren naar voren. Verder bleek uit dit onderzoek dat de plas Nieuwenbroek, die nu de hoogste algenbiomassa vertoont, dichter tegen een fosfaatbeperking aanzit dan plas Elfhoeven en daarom waarschijnlijk het duidelijkst en het snelst zal reageren op fosfaatbeperkende maatregelen. -218-

Foto pag. 219: Het groenwier Pediastrum biradiatum Meijen, indicator voor schoon water, in de Nieuwkoopse plassen (Noord- einderplas d.d. 11.11.87).

-220-

CURRICULUM VITAE De auteur van dit proefschrift werd geboren op 31 oktober 1946 te Sappemeer. Vanaf 1958 bezocht tiij het Willem Lodewijk Gymnasium te Groningen, waar hij in 1966 het eindexamen gymnasium p aflegde. Militaire dienstplicht werd vervuld van 1966 tot 1967, waarna in ok tober 1967 een aanvang werd gemaakt met de studie biologie aan de Vrije Universiteit te Amsterdam. Het kandidaatsexamen werd afgelegd in 1970 en het doktoraalexamen in 1973 (beide cum laude). Als hoofd vak voor de doktoraalstudie koos hij Hydrobiologie en als bijvakken Algemene dierkunde met statistiek en (te Kisangani, Zaïre) Tropische visserijbiologie. Daarnaast werkte hij als studentassistent van 1970 tot 1973 bij de Vakgroep Plantensystematiek en Oecologie van lagere plan- te van de subfaculteit Biologie aan de Vrije Universiteit. Na het voltooien van zijn biologiestudie was hij tot 1976 als weten schappelijk medewerker verbonden aan het Instituut voor Milieuvraag stukken van de Vrije Universiteit, waar hij in opdracht van het Mini sterie van Volksgezondheid & Milieuhygiëne onderzoek verrichtte naar de effecten van verontreiniging op benthische wieren in sloten in het Groene Hart van Holland. Sinds 1976 is de auteur werkzaam bij de Technische Dienst van het Hoogheemraadschap van Rijnland, aanvankelijk als hydrobioloog, sinds 1987 als plaatsvervangend hoofd van de afdeling Waterhuishou ding. Hij is lid van diverse landelijke commissies en werkgroepen op het terrein van het waterkwaliteitsbeheer. Hij publiceerde enkele tien tallen populair-wetenschappelijke en wetenschappelijke artikelen over effecten van waterverontreiniging, biologische waterbeoordeling, eutro fiëring en normstelling van oppervlaktewater.

PUBLIKATIES

Klapwijk, S.P., 1975. Literatuuroverzicht betreffende effecten van verontreinigingen op benthische wieren. Vrije Universtiteit, Instituut voor Milieuvraag stukken, werknota no. 48.

Hülebrand, H. & S.P. Klapwijk, 1976. Distribution of multicellular benthic algae in an eutrophic ditch. Hydrobiol. BuU. 10: 48-58.

Klapwijk, S.P., 1976. Effecten van verschillende verontreinigingstypen op de benthi sche wierflora in polderwater van het Groene Hart van Holland. Vrije Universiteit, Instituut voor Milieuvraagstukken, werknota no. 58. -221-

Klapwijk, S.P., 1977. De fosfatennota van de interdepartementale coördinatiecommissie voor de milieuhygiëne. Waterschapsbelangen 62: 432-437.

Klapwijk, S.P., 1977. Experimentele fosfaatverwijdering op praktijkschaal in Rijnland. Waterschapsbelangen 62: 284-289.

Klapwijk, S.P., 1978. Saprobiological evaluation of Dutch ditches using benthic algae on artificial substrates. Verh. internat. Verein. Limnol. 20: 1811-1815.

Klapwijk, S.P., 1979. Eutrofiëring en fosfaatverwijdering in Rijnland. De Klaarmeester 14: 7-14.

Klapwijk, S.P., 1979. Graskarpers, ook goed voor de waterkwaliteit? Gemeentewerken 8: 309-311.

Klapwijk, S.P., 1979. Graskarpers, ook goed voor de waterkwaliteit? Waterschapsbelangen 64: 484-489.

Klapwijk, S.P., 1979. Ingezonden: Fosfaatbelasting Friesland. H20 12: 133-134. Klapwijk, S.P., 1980. Defosfatering: het belang van hydrobiologische studies. Syllabus Herfstinformatiedag, Belgische Commissie IAWPR.

Klapwijk, S.P., 1980. Effect of laundry wastewater on benthic algae in ditches in The Netherlands. Hydrobiol. Bull. 14: 142-152.

Klapwijk, S.P., 1981. Le déphosphatage: 1'importance des études régionales hydrobio- logiques. Trib. Cebedeau 34: 513-518.

Klapwijk, S.P., 1981. Limnologisch onderzoek naar effecten van defosfatering in Rijn land. H20 14: 472-483. Klapwijk, S.P., 1981. Onderzoek in Nederland naar het effect van het uitzetten van graskarpers op de waterkwaliteit. In: WERKGROEP GRASKARPER (Coördinatiecommissie Onkruidonderzoek van de Nationale Raad voor Landbouwkundig Onderzoek): Verslag graskarpercontactdag 1981, gehouden op 2 september 1981 te Noordwijkerhout.

Heinsdijk, M.A. & S.P! Klapwijk, 1982. Waterkwaliteitsmodellen bezien vanuit het beheer. Symposium Wa terkwaliteit en Modellering, Waterloopkundig Laboratorium, Delft, pp. 91-98. -222-

Hovenkamp, I.R.M., S.P. Klapwijk & J.E.F. Landman, 1982. Biologische beoordeling van de waterkwaliteit in Noord- en Zuid- Holland. H20 15: 406-412. Hovenkamp-Obbema, I.R.M., S.P. Klapwijk & J.E.F. Landman, 1982. Nawoord op het commentaar van H.W. Kroes op het artikel "Bio logische beoordeling van de waterkwaliteit in Noord- en Zuid- Holland" (HzO 15: 406-412, 5 augustus 1982). H20 15: 675-676. Klapwijk, S.P. , 1982. Hydrobiologisch onderzoek naar de uitwerking van het waterkwa liteitsklassensysteem van Caspers en Karbe voor grotere wateren in Zuid-Holland. Rapport Hoogheemraadschap van Rijnland, Tech nische Dienst, Leiden.

Klapwijk, S.P., J.M.W. Kroon & M-L. Meijer, 1982. Available phosphorus in lake sediments in The Netherlands. Hydrobiologia 92: 491-500.

Klapwijk S.P., T.F. de Boer & M.J. Rijs, 1983. Effects of agricultural wastewater on benthic algae in ditches in The Netherlands. In: R.G. Wetzel (ed.): Periphyton of Fresh water Ecosystems, pp. 311-319. Junk, The Hague.

Bruning, C. & S.P. Klapwijk, 1984. Application of derivative spectroscopy in bioassays estimating al gal available phosphate in lake sediments. Verh. internat. Verein. Limnol. 22: 172-178.

Does, J. van der & S.P. Klapwijk, 1985. Effecten van fosfaatverwijdering op de waterkwaliteit in Zuid- Oost Rijnland. HzO 18: 381-387.

Klapwijk, S.P. & C. Bruning, 1986. Available phosphorus in the sediments of eight lakes in The Netherlands. In: P. G. Sly (ed.): Sediments and water interac tions, pp. 391-398. Springer Verlag, New York.

Tieman, P.CM. & S.P. Klapwijk, 1986. Waterkwaliteitsbeheer en automatisering. Waterschapsbelangen 71: 514-517.

Vlugt, J.C. van der, S.P. Klapwijk & J.A.A.M. van Eijk, 1986. Waterkwaliteitsonderzoek Reeuwijkse plassen WOR 1983-1985; ver slag van drie jaar veldonderzoek. Rapport Rijksinstituut voor Volksgezondheid & Milieuhygiëne nr. 840156001, Büthoven.

Boulan, R.P., M. Donze & S.P. Klapwijk, 1987. Fosfaatbalans van de polder Reeuwijk en een aantal deelgebieden. Mededelingen nr. 10 van de Vakgroep Gezondheidstechniek en Waterbeheersing, T.U. Delft. -223-

Bruning., C. & S.P. Klapwijk, 1987. Toepassing voor afgeleide spectroscopie bij bio-assays die het voor algen beschikbaar fosfaat in meersedimenten schatten. Labstract nr. 6: 12-15.

Does, J. van der & S.P. Klapwijk, 1987. Effects of phosphorus removal on the maximal algal growth in bioassay experiments with water from four Dutch lakes. Int. Revue gés. Hydrobiol. 72: 27-39.

Klapwijk, S.P., 1987. Ecologische normstelling voor het zoete water. Landschap 4: 28-40.

Klapwijk, S.P. & C. Bruning, 1987. Beschikbaar fosfaat in de sedimenten van acht meren in Rijnland. Labstract nr. 7: 20-26.

Vlugt, J.C. van der & S.P. Klapwijk, 1987. Dose-effect relationships between phosphorus concentration and phytoplankton biomass in the Reeuwijk lakes (The Netherlands). Verh. internat. Verein. Limnol. 23: 482-488.

Vlugt, J.C. van der & S.P. Klapwijk, 1987. Waterkwaliteitsonderzoek in de Reeuwijkse plassen 1983-1985. H20 20: 86-91. Vries, P.J.R. de & S.P. Klapwijk, 1987. Bioassays using Stigeoclonium tenue Kütz. and Scenedesmus quadricauda (Turp.) Bréb. as testorganisms; a comparative stu dy. Hydrobiologia 153: 149-157.

Hove, L. van den, U.G. Dijkstra-Stam & S.P. Klapwijk, 1988. Biologische waterbeoordeling bij het Hoogheemraadschap van Rijn land. Proceedings Symposium Biologische Waterbeoordeling, April 1987, Wageningen.

Hovenkamp, I.R.M., S.P. Klapwijk & Y. Scheffer-Ligtermoet, 1988. Toepassing van. het raamwerk van Caspers & Karbe voor de wa terbeoordeling in Noord- en Zuid-Holland. Proceedings Symposium Biologische Waterbeoordeling, April 1987, Wageningen.

Klapwijk, S.P., L. van den Hove & P. Nieuwpoort, 1988. Een vergelijking tussen historische en recente gegevens van hy- drochemie en plankton in het gebied van Rijnland. In: R.M.M. Roijackers (ed.): Hydrobiologisch onderzoek in Nederland. Fun damentele en toegepaste aspecten. Publ. nr. 6 Hydrobiol. Ver., Amsterdam: 93-103.

Klapwijk, S.P. & C.J. Smit, 1988. Gouda en de waterkwaliteit van Rijnland. In: Ludy Giebels (ed.): Waterbeweging rond Gouda van ca. 1100 tot heden: geschiedenis van Rijnlands waterstaat tussen IJssel en Gouwe. Uitgave Hoog heemraadschap van Rijnland, Leiden. -224-

Ruiter, M.A. de, T.H.L. Claassen & S.P. Klapwijk, 1988. Fosfatennota vertelt halve waarheid. Volkskrant 24 maart 1988, p. 19.

Vlugt, J.C. van der & S.P. Klapwijk, 1988. Reeuwijk lakes 1983-1987. Internat. Symp. Lake restoration by reduction of phosphorus loading, April 1988, Noordwijkerhout, The Netherlands. -225-

DANKWOORD Het schrijven van een proefschrift naast een volledige dagtaak is niet eenvoudig. Zoiets kan alleen tot een goed einde worden gebracht dank zij de ondersteuning en de hulp van velen, die ik hierbij van harte wil bedanken. In de eerste plaats wil ik mijn promotor, Prof. Dr. M. Donze, bedanken voor zijn stimulans tot en zijn belangstelling tijdens het schrijven van dit proefschrift. Zonder zijn vaak ironische en spits vondige correcties zou dit boekje een stuk dikker en onleesbaarder zijn geworden. Ook de overige leden van de promotiecommissie ben ik erkente lijk voor de tijd die ze gestoken hebben in het doorlezen en becom mentariëren van het concept-proefschrift. Het college van Dijkgraaf en Hoogheemraden van Rijnland en de beide diensthoofden ben ik zeer dankbaar voor zowel de morele als de materiële ondersteuning, die ik ervaren heb tijdens het maken van dit proefschrift. Onderzoek is vaak een gezamenlijke inspanning, zoals ook blijkt uit het feit dat een groot deel van de in dit proefschirft opgenomen artikelen geschreven is in samenwerking met collega's, zowel van bin nen Rijnland als van verwante instituten en universiteiten. Vooral de discussies met mijn collega's en vakgenoten Gerda Bolier, Cees Bruning, Joop van der Does, Jo van der Vlugt en Pierre de Vries hebben direct invloed gehad op de inhoud van dit proefschrift. Daar naast hebben talloze collega's en vakgenoten, met wie ik de laatste jaren in verschillende werkgroepen en commissies heb gezeten, bijge dragen aan mijn meningsvorming over biologische waterbeoordeling en de eutrofiëringsproblematiek. Mijn chefs en naaste collega's hebben veel belangstelling en be grip getoond toen ik in de afgelopen maanden mijn tijd moest verdelen tussen het afronden van dit proefschrift en mijn normale werkzaam heden. Tientallen studenten van diverse universiteiten, hogescholen en laboratoriumopleidingen, die de afgelopen jaren een stage hebben ver richt bij Rijnland, hebben gedeelten van het praktische werk uitge voerd. Ik hoop dat ze nog iets van hun noeste arbeid in dit boekje kunnen terugvinden. De heren Guijt en van Duijvenbode hebben in weer en wind alle monsters op tijd naar het laboratorium gebracht. Het is goed om te weten dat de basis van dit onderzoek, de monsterneming, in zulke vertrouwde handen was. Medewerkers van Rijnlands centrale labora torium hebben duizenden chemische analyses uitgevoerd. Uit de resul taten blijkt hoe stabiel de kwaliteit van hun werk is. Het grootste gedeelte van het veld- én laboratoriumwerk voor dit onderzoek is uitgevoerd door de analisten Uka Dijkstra, Linda van den Hove, Peter Nieuwpoort en Marion Pach. Dankzij hun praktische hulp en die van mijn naaste collega's Kees Bender, Joop van der Does en Piet Kuijt kan blijkbaar een a-technisch iemand toch promoveren aan een Technische Universiteit. Verder wil ik de medewerking van de tekenkamer noemen, waar verschillende van de figuren zijn vervaardigd en de zeer grote inzet van de typekamer, met name van Carla van Dijk, die steeds opnieuw bereid bleek om nauwgezet het typewerk voor de zoveelste keer te verbeteren. -226-

Gerard Zwarts heeft enthousiast als altijd het drukwerk uitge voerd op een wijze en in een tijd, die geen professionele drukkerij hem nadoet. In eerste instantie heeft Paul Gutteridge en later vooral Anya Honnef mij met het Engels geholpen. Anne Post toe Slooten heeft voor het boekje een fraaie omslag, een titelblad en enkele sprekende foto's gemaakt.

De bijdrage van mijn kinderen, Jelle en Maartje, aan de totstand koming van dit proefschrift is groter geweest dan ze zelf denken. Dank zij hun overredingskracht heb ik me vorig jaar laten verleiden om een personal computer aan te schaffen, waar ik veel gemak van heb gehad bij het schrijven van dit proefschrift. De meeste steun en stimulans om dit proefschrift af te ronden heb ik ontvangen van mijn vriendin Floor. Zij was, als ik zelf wel eens twijfelde, zeker van het resultaat. Samen met onze vriendinnen Loeke en Sjak is ze dan ook voornemens om van een plechtige aange legenheid als een promotie vooral ook een feestelijke gebeurtenis te maken. Alleen al om die reden is dit boekje in de eerste plaats voor haar.

Ik draag dit proefschrift in dankbaarheid op aan mijn overleden vader, die mij als kind zowel de liefde voor de natuur als de belang stelling voor haar wetmatigheden heeft bijgebracht. Beide eigenschap pen hebben in belangrijke mate bijgedragen aan mijn studie- en be roepskeuze en vormen nog steeds een grote bron van inspiratie voor mijn werk. -227-

COLOFON

Uitgave Hoogheemraadschap van Rijnland Breestraat 59, 2311 CJ Leiden Tel.: 071-259125

Correctie Engelse tekst Anya Honnef, Leiderdorp

Typewerk Carla van Dijk, typekamer Rijnland

Drukwerk Gerard Zwarts, huisdrukkerij Rijnland Ontwerp omslag, titelblad en foto's Anne Post toe Slooten, Groningen

Kleurendruk omslag Grafisch Bedrijf Codex, Schevingen

Brochage Boekbinderij P.M. Jansen b.v., Leiden