VESIENTUTKIMUSLAITOKSEN JULKAISUJAPUBLICATIONS OF THE WATER RESEARCH INSTITUTE

Jorma Niemi: Simulation of water quality in lakesTiivistelmä: Veden laadun simulointi järvissä 3

Jorma Niemi: Mathematical modeling of phytoplankton biomassTiivistelmä: Kasviplanktonmallien laadinta 16

Urpo Myllymaa & Sakari Murtoniemi: Metals and nutrients in the sediments of smalllakes in Kuusamo, North-eastern FinlandTiivistelmä: Metallien ja ravinteiden esiintyminen Kuusamon pienten järvien sedimenteissä 33

Urpo Myllymaa: Quality of lake water and sediments and factors affecting these in theKuusamo uplands, North-east FinlandTiivistelmä: Järvien veden ja sedimenttien laatu ja niihin vaikuttavat tekijät Kuusamonylänköalueella 49

Lea Kauppi, Kaarle Kenttämies & Eeva-Riitta Puomio: Effect of nitrogen and phosphorusremoval from sewage on eutrophication of lakesTiivistelmä: Asumajätevesien fosforin ja typen poiston vaikutus järvien rehevöitymiseen 70

VESI- JA YMPÄRISTOHALLITUS—NATIONAL BOARD OF WATERS AND ENVIRONMENT, FINLANDHelsinki 1986

Tekijät ovat vastuussa julkaisun sisällöstä, eikä siihen voidavedota vesi- ja ympäristöhallituksen virallisena kannanottona.

The authors are responsible for the contents of the publication.It may not be referred to as the official view or policy

of the National Board of Waters and Environment.

ISBN 951-47-0864-4ISSN 0355-0982

Helsinki 17. Valtion painatuskeskus

70

EFFECT OF NITROGEN AND PHOSPHORUSREMOVAL FROM SEWAGE ON EUTROPHICATION

OF LAKES

Lea Kauppi1),Kaarle Kenttämies2)& Eeva-Riitta Puomio3)

Kauppi, L., Kenttämies, K. & Puomio, E-R. 1986. Effect of nitrogen andphosphorus removal from sewage on eutrophication of lakes. Publications ofthe Water Research Institute, National Board of Waters and Environment,Finland, No. 69.

Experimental enclosures were used to study the effects of different sewagetreatment methods on primary production, biomass and species compositionof plankton as well as on nitrogen fixation by blue-green algae in three lakesof different trophic status. The limiting nutrient was determined by algal testusing Selenastrum capricornutum as test organism. The sewage treatmentmethods studied were: 1) biological + P removal and 2) bioiogicai + Premoval + N removal by nitrification — denitrification. A preliminaryexperiment was performed using inorganic fertilizers. Three kinds ofexperimentai enclosures were tested and a fairly large (50 m3) one with anatural bottom and pianktivorous fish population was found to be superior tosmali (1—2 m3) ones.

In the chemical enrichment experiment, only nitrogen and phosphorustogether were able to increase primary production to a significant extent inthe eutrophic, previous recipient, lake Vesijärvi. Planktonic nitrogen fixationwas enhanced by P-enrichment, while nitrogen alone inhibited it. Sewageresulted in a faster growth rate for periphyton than chemicai P-enrichment,while only slight differences were observed in planktonic production. Jo spiteof the different chemical composition the bioom of heterocystous blue-greenaigae occurred simultaneously in the lake and in ali the enclosures.

Jo the ohgotrophic, pristine forest lake, sewage enrichment in a benthicenciosure resulted in a rapid, short-term increase in aigal growth foilowed by adecline in production. Grazing by zooplankton, together with the nutrientcontent, reguiated primary production. Jntroducing fish fingerlings effectiveiyfavoured primary production by decreasing the zooplankton biomass. No Nfixation could be observed, the biomass of biue-green algae reaching only1.4 % of the total biomass at its maximum.

Benthic enciosures were aiso used in the third sewage enrichrnentexperiment in a heaviiy loaded recipient. Enrichment clearly increased thephytoplankton biomass, aithough the internal phosphorus load fromsediments — varying from 7.25 to 107 mg m2 d’ had a decisive impact onthe P budget of the enclosures. A very dense zooplankton popuiation causedovergrazing of phytopiankton towards the end of the experiment. Again,despite the differences in the quality of the sewage used for enrichment thephytopiankton succession during the summer was similar in ali enclosures,Pyrrophyta and Chlorophyta dominating. The Cyanophyta biomass remainedinsignificant.

The results showed that decisions concerning nitrogen removal fromsewage should be made case by case. Carrying out simple enrichmentexperiments in the actual recipient are recommended before decision-making.

Jndex words: Sewage, lakes, P removal, N removal, eutrophication, planktonicnitrogen fixation, enclosures

1) Water and Environment Research Institute, National 2) Ministry of the Environment, P0. Box 306, SF-00531Board of Waters and Environment, P0. Box 250, SF- Helsinki, Finland.00101 Helsinki, Finland. 3) Helsinki Water and Environment District, P.0. Box

58, SF-00421 Helsinki, Finland.

71

1. INTRODUCTION

Efficient removal of phosphorus from sewage hasbeen accomphished in most treatment plants inFinland. On an average over 90 per cent of thephosphorus is removed, mainly by chemicalprecipitation. The organic matter is effectivelyremoved at the same time. In order to furtherimprove sewage treatment nitrogen removal shouldalso be considered.

Of the known methods for removing nitrogenfrom sewage, physical-chemical processes are notregarded as suitable owing to the high costs andtechnical complexity. The biological denitrificationprocess, in which nitrification occurs in thepreceding phase, has been considered the mostfeasible in the municipal treatment plants. Theeffective removal of phosphorus is usually aprerequisite of the official permit if the plant isallowed to use a lake as a recipient. The chemicalremoval of phosphorus using ferrous sulphate asthe precipitant is the most common method.

In Finland the dominant position of ferroussulphate as precipitant is well explained by itsavailability as a cheap waste product from thetitaniumdioxide industry. In biological sewagetreatment plants the precipitant is mixed withsewage in the aeration basin in order to achievesimultanous decomposition of organic matter andthe precipitation of phosphorus. Unintentionaltemporary nitrification of ammonium compoundsis not a rare event in many plants during periods oflow load. The plant can be made to nitrify effectively by making relatively small changes to thepurification process. The main difficulty is the iowalkalinity of sewage in Finland, which may causetoo low pH-value in the process. The inhibitory effect of ferrous iron on nitrification is not a majorproblem as the usual concentration levels (Valve1985).

The effects of nitrogen removal may, however,be conflicting: on the one hand the highconcentrations of nitrogen compounds — which assuch may be harmful — are reduced and algalgrowth decreases if the nitrogen supply becomeslimited, while on the other nitrogen removal mayenhance the growth of certain organisms that areable to fix molecular nitrogen. The most important group is cyanobacteria, which can cause odourand taste problems in water and fish whenoccurring in high biomasses (Persson 1982). Somecyanobacteria also contain endotoxins which arereleased after the death of these organisms. Therelative significance of beneficial and detrimentaleffects has to he evaluated before making thedecision whether to implement nitrogen removal.

If the original problem is the harmfully highconcentration of nitrate, nitrogen removal wouldhave only beneficial effects, because it is veryprobable that phosphorus will in any case remainthe limiting nutrient. If, however, eutrophication isthe main probiem then the probability andsignificance of the occurrence of nitrogen-fixingcyanobacteria has to be weighed against thedecrease in primary production by algae.

This article is a summary report of thelaboratory and field studies carried out at theWater Research Institute in 1981—1983, the aimof which was to estimate the effects of nitrogenremoval on lake ecosystems. The emphasis was onphytoplankton composition, primary productionand nitrogen fixation by cyanobacteria.

Experimental studies, involving the constructionof enclosures in the lakes, were performed in aheavily loaded lake, in a lake where the earlierheavy Ioading has ceased and in an unpollutedforest lake.

2. BACKGROUND2.1 Nitrogen: phosphorus ratios in

wastewaters

In some industrial effluents the ratio of total N tototal P is quite low (Table 1). The N:P of theeffluents of the pulp and paper industry is, on anaverage, clearly less than 10. The food processingindustry also produces effluents with a very lowN:P. In treated sewage, on the other hand, the N:Pratio is at present quite high due to the efficientremoval of phosphorus. At the beginning of the1970’s, however, the situation was quite different,the N:P value in treated sewage being less than 10.

Table 1. Nitrogen: phosphorus ratios in the industrialand domestic waste waters at the beginning of the 1980’s(Yrjänä and Kauppi 1984).

Source Tot.N:Tot.P

Pulp and paper industry 6.6Mechanical wood processing 3.3lii and petrochemical industry 34Fertilizer industry 15Other chemical industry 11Metal industry 190Textile industry 12Food processing industry 8.9Domestic waste waters 16Non-point sources 19Fish farming 4.8

72

15000 or industrial effluents. The low N:P ratio in thesetn a lakes is due to the low N:P value of the effluents

(eg. pulp and paper industry) and/or high internal10000 phosphorus loading. The second distinctive group

was formed by man-made lakes with a high organic‘ load frorn former terrestrial soils, and the third one

by lakes located in areas of phosphorus-rich5000 mineral soils. In conclusion, it is ciear that judging

by the N:P values nitrogen is only rarely theprimarily limiting nutrient in Finnish lakes. It may,however, limit decomposition and thus indirectlyalso production (Alexander 1977).

During the last fifteen years the phosphorus ioadhas rapidly decreased, while that of the nitrogenhas steadily increased (Fig. 1). However, the mainfactors balancing the N:P ratio in Finnish lakes arenatural input and agricultural runoff, which arestrongly nitrogen-weighted (Kauppi 1979).

2.2 Nitrogen: phosphorus ratiosin Finnish lakes

The ratio of total nitrogen to total phosphorus canhe used as an initial estimate when determiningthe iimiting nutrient. If the ratio is small, nitrogencan be assumed to limit algal growth. A range ofvalues have been cited in the literature for thecritical ratio, commonly varying between 8 and 17(Voiienweider 1968, Seppänen 1970, Forsberg et al.1978, Rhee and Gotham 1980).

The water quality data bank of the NationalBoard of Waters and Environment was used in order to get an overview of the ratio of total N andtotal P in Finnish lakes (Yrjänä and Kauppi 1984).Ali the iake surface water (0.5—1.5 m) observationsrecorded between January and August from 1978on were included in the study. The data wasgrouped into four groups according to the totai Pconcentration and the season of observation.

In eutrophic (total P > 30 g i1) lakes the N:Pratio was, on an average, much lower than that inoligotrophic (tot. P < 30 g i) lakes. Thesummer mean value of eutrophic iakes was only 16,whereas in oligotrophic lakes it was 34. The lakeswhere iow N:P values were observed could heroughly identified on the basis of the type ofioading. The biggest homogeneous group was thatformed by the iakes heavily polluted by domestic

2.3 Occurrence of nitrogen fixationin some Finnish lakes

Kanninen (1980) studied nitrogen limitation andthe potential significance of nitrogen fixation forthe N budget in three eutrophic lakes — Vesijärvi,Hiidenvesi and Tuusuianjärvi — in 1979. Thestudy was based on phytopiankton biomasscomposition and aigal assays. In lake Vesijärviwhere mineral N, especiaily nitrate, was aimostexhausted in June and July the growth of the testalga, Selenastrum capricornutum Printz, was hmitedby nitrogen from May to the middie of August.The highest biomasses were observed in July andAugust, when cyanobacteria dominated. 23—24per cent of the totai biomass consisted ofheterocystous cyanobacteria (A nabaena circinalisRabenhorst, A. solitaria f. planctonica (Brunnthaler) Komarek and Aphanizomenon flos-aquae(Linn) Ralfs f. flos-aquae.

In the two other lakes, concentrations of bothmineral nitrogen and mineral phosphorus remainedat quite a high level throughout the summer.According to their ratio, phosphorus would havebeen the primarily limiting nutrient in L. Tuusulanjärvi. In May this was confirmed by algal tests, butduring the summer only both nutrients togetherincreased the growth, the effect of N becomingstronger towards the end of the summer. Thebiomass maximum in L. Tuusulanjärvi occurred inearly June, and after this point cyanobacteriadominated; a bioom of a heterocystous cyanobacteria, Anabaena circinalis was observed in July(Kanninen 1980). In Lake Hiidenvesi the algai testsconfirmed nitrogen iimitation, but no significantamounts of heterocystous cyanobacteria were observed (Kanninen 1980).

In 1980 the study was concentrated in L.Vesijärvi because the results from 1979 impliedthat nitrogen fixation by biue-greens could significantiy contribute to the N budget of the lake. In

D

0-c

-c0•.

n a•

2000

1000

197 1976 1978 1980 1982 198

Fig. 1. Loads of phosphorus and organic matter fromdornestic effluents in Finland in 1974—1984.

73

addition to the cyanobacteria biomasses, nitrogenfixation was measured by the acetylene reductionmethod (Yrjänä 1982). In the main basin of thelake planktonic nitrogen fixation was responsiblefor 31 per cent of the total annual N input, and for52 per cent of the N input during the growingseason (Yrjänä 1982). These results encouraged usto carry out experiments on the influence ofremoving nitrogen from sewage on the recipients.

3. MATERIAL AND METHODS3.1 EnclosuresThe experiments involving the construction ofenclosures in lakes (limnocorrals) were planned inorder to gain a better understanding of theecological consequences of removing nitrogen fromsewage. Since the enclosures had originally beendesigned for pelagial plankton studies, a bagiike,closed type enclosure with a size of 2.5 m3 (Fig. 2a) was introduced in 1981. During the course of thestudy the internal load of the recipients proved todemand much more attention. The chemical andbiological effects of sediments subsequently weretaken into account in the two types of benthicenclosures — a cylindrical one, 1 m3 and a cubicalone, 56 m3 (Fig. 2 b and c). In one case fishfingerlings were introduced into the enclosure.

The wall material of the enclosures was atransparent, 0.20 mm thick polythene fiim. Thetransparent “collar” in the pelagial enclosures wasmade of PVC plastic. The enclosures were assumedto he absolutely waterproof. The benthic enclosures were made of twofold polythene fiim.

The water sampies were taken from enclosureswith a normal Ruttner-sampler. In the homothermal, relatively shallow enclosures chemical andplankton analyses were made on a compositesample obtained by mixing the vertical subsamples.

The chemical analyses were made by thestandard methods of the Finnish Water and En—vironment Research Institute (National Board ofWaters 1981). The algal test using Selenastrum cap—ricornutum has been described by Kanninen et al.(1982) and the acetylene reduction method used inestimating nitrogen fixation by Yrjänä (1982).

Primary productivity “in vitro” and nitrogenfixation “in vitro” were preferred to “in situ”analyses from the practical point of view. Comparable results for the enclosures (treatments) werethe most important goal.

bStyrox—f(oaf

PLastic wut[Wooden frQme

Anchor ropeSond bcigs

Fig. 2. Types of enclosures used in the enrichment experiments a) a bag-like, closed enclosure (2.5 m3), b) a cubical,benthic enclosure (56 m3) and c) a cylindrical, benthic enclosure (1 m3).

3.2 Sewage from treatment plantsIn this study, treated sewage from both pilot andfuli scale plants were used in the enclosure tests.

The Kariniemi sewage plant in the city of Lahtiis a conventional active sludge plant with simultaneous phosphorus precipitation. The precipitant isferrous sulphate. Phosphorus removal is usually ofa high degree of efficiency. In the summer duringhigher temperature and low load phases especially,the plant tends to nitrify ammonium compounds.During the enclosure tests in Lake Vesijärvi thesewage plant was even too effective! The phos

ci

col[ar

bond

c

Ptostic bond

6 471400R

74

phorus concentration in the treated sewage was onoccasions as low as 0.3—0.4 mg 1—1 P and thus toofar from the “normal” Finnish mean of 1 mg 11 P.Preclarified sewage was therefore added to thesewage used for enrichment of the enclosures inorder to produce the typical, higher phosphoruslevel.

The pilot plant of the city of Hyvinkää was abiological plant that had simuitaneous precipitation of phosphorus with ferrous suiphate asprecipitant. The nitrification and denitrificationprocesses were accomplished in the same reactorusing the so-called intermittent method, in whichoxic and anoxic phases were interchanged.

After passing through these two processes thewater was reaerated and settled. The key probiemsin the process are the inhibitory effect of ferroussuiphate on nitrification, and ineffective phosphorus removal due to the presence of unsettied,fine sludge. Treated sewage from the pilot plant ofHyvinkää was used in the enclosure tests of lakeValkea Mustajärvi, an oligotrophic clearwaterforest lake. Before addition to the test enciosure,the treated sewage was diiuted with lake water(1:1 ratio) and aged in a separate enciosure for twoweeks. This procedure was thought to bettersimulate the prevaiiing case where the oligotrophic recipient is relatively far from the loadingpoint. Both phosphorus and nitrogen removalcontinued throughout the aging process before thetreated sewage was added to the enciosureecosystem proper.

The treatment plant of the city of Mikkelicarried out denitrification in a separate small unitfor the enciosure tests in the heavily loaded lakeRohmavesi. However, the test has to be classifiedas being done on a whole plant scale. In theprocess, denitrification took place in the firstreactor where carbon-rich raw sewage and nitrogen- rich water from the second effectively aeratednitrification unit were mixed (so cailed dn-process).

After the third sedimentation phase the treatedsewage was pumped into the enciosures. Thephosphorus was precipitated simultaneously withferrous sulphate. The conventionally treated, Pprecipitated sewage was obtained from the mainunit of the plant.

The chemical characteristics of ali the types ofsewage used in the study are presented in Tabie 2.Ali additions to the enclosures were done using arelativeiy high degree of dilution in order toachieve better simulation of the real conditions inthe recipients. The continuous flow of sewage intoenciosures was not reaiizabie in pratice, even if itwould have been theoretically the best way ofcarrying out the simuiation.

4. RESULTS4.1 Chemical enrichment experimentA prehminary experiment with chemicai enrichment was made in iake Vesijärvi. The bag-iikeenclosures (2.5 m3) used in this experiment wereciosed at the bottom thus ehminating the effect ofthe lake sediment. They were enriched with purechemicals, i.e. with polyphosphate (Na5P3O10),ammonium nitrate (NH4NO3) and ammoniumchioride (NH4G1) (Table 3).

Both the nitrogen and phosphorus concentrations decreased in the enclosures during the courseof the experiment. The added phosphorus inparticuiar disappeared rapidiy from the waterphase, although the concentrations stiil remainedat a higher ievei than that in the control enciosure.Since nitrogen was much more stable, thetot.N:tot.P ratio increased in ali the enclosuresduring the experiment. In minerai nutrients the

Table 2. Chemical characteristics of the treated sewage used in the enclosure experiments.

Sewage plant Date tot. P tot. N N02-N N03-N NH -Nmg 1—’ mg 11 mg h’ mg h’ mg 1—1

Lahti city 22.7.1981 0.34 8.5 0.540 0.52 0.5312.8.1981 1.0 18 0.800 9.0 3.59.9.1981 1.4 24 0.600 13 5.7

Hyvinkää pilot plant 6.7.1982 0.48 2.0 0.002 0 0.0317.8.1982 0.68 7.9 0.008 0.68 1.5

Mikkeiicity,, 29.6.1983 0.57 9.5 0.200 8.5 0.80denitnfying line

Mikkeli ity,, 29.6,1983 0.53 14 0.490 7.2 6.0conventional line

Mikkeli ity27.7.1983 0.34 12 0.440 7.2 3.4denitrifying line

75

Table 3. Nutrient additions in the enrichment experiment with inorganic fertiiizers in Lake Vesijärvi 1981.

Enclosure Nutrient addition Limiting nutrientbased on aigal

assayP0 N0

II 0.21 g P(0.84 g Na5P3O10) 120III 7.56 g N (10.8 g NH4NO3+ 14.4 g NH4C1) 4700IV 0.21 g P + 7.56 g N (0.84 g Na5P3O10 120 2900

+ 10.8 g NH4NO3+ 14.4 g NH4CI)

development was not that straightforward (Table4) perhaps due to uptake by aigae.

Only the N+P enrichment significantiy increased the primary productivity and chiorophyii aconcentration (Figs. 3 and 4). These variabiesremained at the same level as the control in ali theother enclosures. The chiorophyli a levei reached amaximum in the N+P enciosure one week after thepeak of primary productivity.

Nitrogen fixation by biue-green algae wasenhanced by P enrichment, while nitrogen additionaione inhibited it (Fig. 5). The effect was evenmore pronounced in the “in situ” measurementsmade two weeks after the enrichment (Fig. 6). Themaximum biomass of biue-greens occurred in theenclosure enriched with both N and P (Fig. 7). Thedominant species in ali the enclosures wereOscillatoria agardii Gomont and Microcystis sp.

Table 4. Nitrogen: phosphorus ratios and hmitingnutrient according to algai assay in the enrichmentexperiment with inorganic fertilizers in Lake Vesijärvi in1981.

Enclosure Date Tot.N: N:Pp04 Limiting nutrientTot.P based on aigal

assay

Control 24.6. 12 4.7 N1.7. 24 1.6 P,N7.7 20 5.3 P, N

15.7. 90 4.2 N, P

P- 24.6. 5.0 0,55 Nenriched 1.7. 9.7 0,27 N

7.7. 8.8 0,29 N15.7. 13 0,48 N

N- 24.6. 130 Penriched 1.7. 120 490 P

7.7. 150 440 P15.7. 520 540 P

N+P 24.6. 24 37enriched 1.7. 27 220 P

7.7. 34 120 P15.7. 44 93 P

Fig. 3. Primary productivity in the chemical enrichmentexperiment in 1981. — D control; 0 — 0 P-enriched;

— N-enriched; L N+P -enriched.

0

0

0

0

>-

ci0

-cLJ

100

Big. 4. Chlorophyll a concentration in the chemical enrichment experiment in 1981. Symbois as in Fig. 3.

76

7 13 Days 21

Fig. 5. Nitrogen fixation “in vitro” (measured by theacetylene reduction method) in the chemical enrichrnentexperiment in 1981. Symbols as in Fig. 3.

The heterocystous blue-green algae increasedduring the experiment in the control enclosure andespecially in the P-enriched enclosure, while Nenrichment resulted in a decrease in their biomass.

According to the aigal tests, nitrogen limited thegrowth of Selenastrum capricornutum in the Penriched enclosure, while phosphorus was thelimiting nutrient in the N-enriched and N+Penriched enclosures. In the control enclosurenitrogen limited the growth of Selenastrum in thebeginning, but during the course of the experimentonly both nutrients together resulted in increasedgrowth (Table 4). The results are in agreementwith the nutrient ratios measured in the enclosures.

4.2 Sewage enrichment experiments

The results of chemical enrichment experimentscannot be directly applied to the estimation of theeffects of sewage on recipients. In addition tophosphorus and nitrogen, sewage contains numerous compounds which influence the response ofthe lake. Therefore we continued the enclosureexperiments with sewage enrichment.

4.2.1 Previous recipient

The first experiment with sewage enrichment wasmade in Lake Vesijärvi using pelagial bag-likeenclosures (2.5 m3). “Normal” biologically andchemically (P removal) treated sewage was obtainedfrom the treatment plant of the city of Lahti. Thetreatment plant removed nutrients very efficiently,especially during the experiment in July. In orderto get more “normal” sewage for the experiment

Fig. 6. Nitrogen fixation “in situ” (measured by theacetylene reduction method) in the chemical enrichmentexperiment in 1981 two weeks after the enrichment.

7B

N m 2h1

.2 50

><

c0)

0

25

)Jg

Control P— N— NP —

enc[osure enriched enriched enrichedenc[osure encLosure enc[osure

mg

15

10

060

5 [mi1,7. 77.15.7 1.7 77.15.7 1.7 77.15.7Contro[ P—enriched N—enrichedencLosure enetosure enc[osure

Fig. 7. Phytoplankton biomasses in therichment experiment in 1981.

1.7 77lS7DcifeN + P—enrichedenclosure

chemical en-

77

15 per cent preclarified sewage was mixed with it.

Despite this, the nutrient concentrations of theenrichment were quite low (Table 2).

The first enrichment with phosphate andsewage, both aiming at the same total P concentration, was made on July, 22, and repeated twice, onAugust, 12 and September, 9. The highest total Pconcentration was found during the whole period

in the chemicaliy P-enriched enclosure. The samewas true for phosphate. In the sewage-enrichedenclosure the total P and P04-P concentrationsdecreased to the control level in one week, exceptafter the Iast enrichment when they stayed abovethe control for the whoie observation period (Fig.8). Initial nitrogen concentrations were naturailyhighest in the sewage-enriched enclosure, i.e. theonly one where nitrogen was also added. Thedecline was, however, quite rapid (Fig. 9). Themineral nitrogen concentrations were almost zeroduring the whole experiment in the controi and Penriched enclosures and in the sewage-enrichedenclosure it was almost exhausted in three weeksafter each enrichment. According to the ratiobetween mineral nutrients, nitrogen was limitingaigal growth in the control and P-enrichedenclosures. In the sewage-enriched enclosure, theN:P ratio was close to the optimum.

Sewage strongly increased primary productivityafter the second and third enrichment, whilephosphorus alone had only a slight effect (Fig. 10).The chlophyll a values followed a slightly differenttrend: the sewage-enriched enclosure had thelowest chlorophyll values, except after the lastenrichment (Fig. 11). The smaller aigal biomass(measured as chlorophyll a) in the sewage-enrichedenclosure was at least partly compensated byperiphyton growth: almost no periphyton grew inthe control and P-enriched enclosure, while sewagecaused the development of a strong periphytongrowth:

Chl. a (sg m2)ControlP-enrichmentSewage enrichment 1 000

The biomass of blue-green algae in the enclosures remained smaller than that in thesurrounding lake. At the time of the secondenrichment the biomass of Cyanophyta was clearlyhigher in the sewage enciosure than in the others.The proportion of heterocystous blue-greens wasnormaily iess than 10 per cent of the total biomass,but immediately after the third enrichment theybecame dominating in ali the enclosures (Fig. 12).Maximum nitrogen fixation was also observed atthe same time (Fig. 13). The dominant heterocystous aigae were Aphanizomenon flos-aquae anddifferent Anabaena-species. For the rest of thetime Oscillatoria agardhii was dominating. According to the algai tests, as weli as the nutrient ratios,the P-enriched enclosure was continuousiy nitrogen limited, while in the two other enclosures thehighest growth increase was normally obtained by

0

0

c0)

00

:3

0•

0.00•

Fig. 8. Total phosphorus and phosphate phosphorus

(shaded area) concentrations in the sewage enrichrnentexperiment in 1981.

1500

1-

00

11000

500

2638

Control P—en ri ched Sewage—enrichedenelosure enetosure enelosure

Fig. 9. Total nitrogen and mineral nitrogen (shaded area)concentrations in the sewage enrichment experiment in1981.

78

Fig. 10. Primary productivity in the sewage enrichmentexperiment in 1981. EI — EI control; 0 0 P-enriched;

— i sewage-enriched.

Fig. 13. Nitrogen fixation “in vitro” (measured by theacetylene reduction method) in the sewage enrichmentexperiment in 1981.

adding both nutrients (Table 5). This was not,however, reflected in nitrogen fixation: maximumnitrogen fixation was observed at the same time inali the enciosures, as well as in the surroundingwater. No ciear differences were observed in therate of nitrogen fixation between the enclosures.Thus the growth of nitrogen-fixing biue-greenalgae in Lake Vesijärvi was obviously stronglyregulated by environmental factors other thanactual nutrient concentrations. Radiation, temperature and probably some micronutrients mightbe the factors which determine the succession ofthese algae during the growing season.

1,00

mgC m3d -I

300

0

>.

0

0•

200

100

22.7. 29.7 128.19.8. 2.9. 9.9. 16.9. 30.9. Ekste

Fig. 12. Phytoplankton biomasses in the sewage enrichment experiment in 1981.

Contro[ P—enriched Sewage—enrichedenciosure encLosure enc[osure

00

0

0

00

0

>,.00.0

0

30

1ig

20

10

022.7. 297 12.8.19.8. 2.9. 9.9. 16.9. 30.9.Date

Fig. 11. Ghlorophyll a concentrations in the sewage enrichment experiment in 1981. Synibois as in Fig. 10.

79

Table 5. Nitrogen: phosphorus ratios and limitingnutrient according to aigal assay in the sewage enrichment experiment in Lake Vesijärvi in 1981.

Enclosure Date Tot.N: Nm:Pp4 Limiting nutrientTot.P based on aigal

assay

Control 22.7. 29 1.329.7. 12 2.5 N, P12.8. 10 1.1 P, N19.8. 17 2.5 N, P2.9. 25 1.6 N, P9.9. 16 7.0

16.9. 20 5.7 P, N30.9. 12 0.86 N, P

P- 22.7. 7.0 0.71enriched 29.7. 3.0 0.20 N

12.8. 3.3 0.1 N19.8. 1.7 0.1 N2.9. 3.8 0.04 N9.9. 1.7 0.06

16.9. 2.7 0.05 N30.9. 2.9 0.06 N

Sewage- 22.7. 19 11enriched 29.7. 25 21 P

12.8. 12 5.8 N, P19.8. 6.3 12 N, P2.9. 28 7.5 P, N9.9. 16 13

16.9. 26 25 P, N30.9. 13 19 P, N

c0

0

00J

00

00)

0

011.9 28. Dote

00

0

0

00

0

0

0.

.00•

Fig. 15. Chlorophyll a concentration and primary productivity in the sewage enrichment experiment in 1982.

ber Pyrrophyta (Cryptomonas sp.) and Chrysophyta(Uroglena americana Caikins) became dominant.Cyanophyta were almost absent throughout thewhoie summer. They formed only 1.4 per cent ofthe total phytoplankton biomass at the maximum.No nitrogen fixation was observed, although someAnabaena and Aphanizomenon species were foundin ali the sampies.

The zooplankton biomasses foliowed the vanation in phytopiankton with a short time iag.Their highest biomass vaiues were usuaily observed2—3 weeks after the phytoplankton maximum(Fig. 17). Cladocera were dominant, except in thebeginning of Juiy, when zooplankton consisted

70

4.2.2 Unpoliuted forest lake

Two types of benthic enclosures were used in LakeValkea Mustajärvi, a iarge cubical, 4 m deep one(56 m3) and a small cylindnical enclosure (1 m3).

Lake Valkea Mustajärvi is an oligotrophic lake.After the sewage ennichments, the nitrogen andphosphorus concentrations doubled in the largeenciosure (Fig. 14). Later on they rapidly decreased, however, to almost the same level as thatin the iake. The sewage additions resuited in arapid increase in pnimary production as well as inthe chlorophyll a vaiues (Fig. 15). The mostdramatic increase was, however, short-term, andthe vaiues returned near to the oniginai levei.

The phytopiankton biomass also increased as aresult of sewage addition, but decreased againalready within one week simultaneously with anincrease in the zoopiank :on biomass (Fig. 16). Thedominant algal groups varied with time, startingwith Chrysophyta (dom. species Synura uvella Steinemend. Korshikov), foilowed by Pyrrophyta(Rhodomonas lacustris Pascher & Ruttner). Afterthe second addition Chlorophyta (Kirclrneriellaobesa (W. West) Schmidle) dominated. In Septem

Fig. 14. Total nitrogen and total phosphorus concentrations in the sewage enrichment experiment in 1982.

30 1500—1 ‘.3 -1

1 mgC0m d

0

.2

20 10000 00) -n0 23 0.

>.

o

10 500

00

137.207.277. 24,5 1.9.

80

Fig. 16. Phytoplankton biomass in the sewage enrichmentexperiment in 1982.

Fig. 17. Density of zooplankton in the sewage enrichmentexperiment in 1982.

mainly of Copepoda. After introducing the whitefish fingerlings (Coregonus tavaretus (L.)) thezooplankton density decreased rapidly, from555 11 to 8 1 in two weeks. The phytoplanktonbiomass, chlorophyll a and primary productivityincreased at the same time. This implies thesignificance of fish in regulating the biomass andproduction of different trophic leveis of theecosystem.

The phytoplankton biomasses and species composition in the small enclosures were almostidentical despite the different sewage additions (6% and 20 %). The maximum biomass values wereobserved one week after the addition, and duringthe next two weeks they decreased to one tenth. Inaddition to smaller biomasses, the small enclosuresdiffered from the large enclosure as regards thespecies composition. The zooplankton bioniasseswere also smaller in the small enclosures.

The size of the enclosure seemed to affect theplankton biomasses. Despite the higher nutrientlevel in the small enclosures, the biomass valueswere clearly higher in the large enclosure. The mainreason is probably the lack of turbulence (and thusefficient sedimentation) in small enclosures.

4.2.3 The heavily Ioaded recipient

The enclosures used in this experiment were of alarge, cubical shaped benthic type with a natural,undisturbed sediment bottom. Sewage for theexperiments was pumped directly from the Mikkelimunicipal sewage treatrnent plant, which issituated about 100 m frorn the experimental area.It was possible during the experiment to operatethe parallel treatment lines of the plant in differentways. Phosphorus removal in the treatment plantwas very efficient and nitrogen removal was ofconsiderable magnitude, too (Table 2). Thus theeffect of the 2 per cent sewage addition on theenclosure recipient was not vety drastic, especiallybecause the recipient was aiready eutrophic without any enrichment.

The nitrogen concentration in the controlenclosure remained at a relatively stable level(Table 6). However, because the phosphorusconcentration increased markedly, the N:P ratiodiminished in the course of the experiment. Theoverail decrease in the tot.N:tot.P ratio could beseen in the sewage-enriched enclosures, too. As wasthe case in the control enclosure, the internal loadfrom the sediment makes it difficult to evaluate thesignificance of the nutrient ratios.

The chlorophyll a content in the enrichedenclosures remained larger than that in the controlthroughout the time (Table 6). However, towardsthe end of the experiment the phytoplanktonbiomass in enclosure III (P removed sewage)decreased drastically, even below that of thecontrol. The zooplankton densities in the controland enclosure III were vety high, and wereobviously the main cause of the decrease in thephytoplankton biomasses.

The phytoplankton in the control enclosure was

0E0

13

rng

11

10

9

8

7

6

5

4

3

2

06.7. 13.7. 27.7. 17.8. 24.8 1.9. 14.9. 28.9.

900

kpl l

700c0. 600

1 500

400

300

200

100

06.7. 13.7 27.7. 17.8. 24.8. 1.9. 14.9. 28.9.

81

Table 6. Characteristics of the enclosure ecosystems after sewage enrichments in the eutrophic Lake Rohmavesiin 1983.

Variable Control enclosure Enclosure 11*) Enclosure III’)

0 7days l4days 0 7days l4days 0 7days l4days

Tot.Nsg1’ 1700 2200 1400 2900 2700 2200 4100 4500 3500N03—N jig 1’ 540 290 100 1200 880 630 2000 2000 1700NH4—N sg M 270 270 300 440 230 260 990 440 680Tot. P sg l1 190 350 400 190 240 190 190 290 360P04—P jsg 1 99 220 300 71 67 56 59 110 250Tot. N tot.P 8.8 6.4 3.6 15 12 11 21 16 9.8N : P04—P 8.9 3.0 1.6 24 18 17 55 26 11Cia sg 1 59 45 27 76 91 86 83 45 5.3Zooplankton (Cladocera+ Copepoda) md. h’ 1300 1600 5000 820 460 710 1500 4600 1900

) Enriched with P and N -removed sewage (“denitrifying line”)Enriched with P-removed sewage (“conventional line”)

dominated by Pyrrophyta and Chlorophyta (Fig.18). The dominant species was Chlamydomonas sp.The enriched enclosures were first Chlorophytadominated, then became dominated by Pyrrophyta.It is remarkabie that the phytopiankton composition developed in the same way in ali the

________________________________________

enclosures after the enrichments. Thus factorsDays 11+ other than the nitrogen and phosphorus concentra

tions (which did not level off very much towardsthe end of the experiments) seem to have a verystrong leveliing effect on phytoplankton composition.

Cyanophyta are seidom dominant in this type ofbrown water recipient. In the control enclosure

______

Cyanophyta did not exceed 0.5 % at its maximum.Their highest proportion was observed in enciosure II, but even then the percentage was only2.8 %. Nitrogen fixing blue-green algae werenaturaliy of even less importance in the enciosures.They were only occasionaily present and theircontribution to the nitrogen budget of theenclosures remained insignificant.

4.3 Internal phosphorus loading fromthe sediments

The enciosures with genuine lake sedimentbottoms made it possible to estimate the release ofphosphorus from sediment into the overlyingwater in the heaviiy loaded recipient. In June one

Control enctosureGJ

d

D

ci

5,>

ci5,

7 DaysllEncLosure 11 (P÷ N removot)

0

7Enriosure III (P—removat)

Pyrrop[iyto

Chrysophyto

j Euglenophyto

Ellil ChLorophyto

Cyanophytci

Fig. 18. Relative abundance of different phytoplanktongroups in the sewage enrichment experiment in 1983.

82

enciosure was fiiled with lake water for controipurposes. In late summer the same enclosure wasemptied, refilled with lake water and enriched withsewage. The properties of the sediment remainedthe same as earlier.

The source of phosphorus, the surface iayer ofthe sediment, had a P concentration of 6.6 mg Pg1 (dry weight). The chemical characteristics ofthe water in the enclosure in July (a) and August(b), after adding treated sewage revealed big differences in the weekly net phosphorus budgets(Table 7). The net load remained positive. Thus theminimum weekly load was +7.25 P mg rn2 d’,and the maximum +107.25 P mg m2d1.

It would appear that the sediment load was nottemperature dependent. The oxygen concentration1 m above the sediment surface also remained atthe same relatively high level. The pH-vaiue didnot change much during the test. The COD, BODand total N values increased during both of thehigh peaks in P release from the sediment. Especially prominent was the weekly BOD increase of3.7 mg O2 in August, which occurred at thesame time as the highest release of phosphorus,+107 mg m2 d2 was measured. On the otherhand, the nitrate concentration decreased in boththe (a) and (b) parts of the experiment uniformlyfrom the beginning to the end. In test (a) thechlorophyll a concentration decreased ali the time,while in test (b) it increased strongly. Thus thephytoplankton biomasses and production correlated positively with phosphorus in the latter partof the experiment oniy. Instead, the crustacean

zooplankton biomasses correlated fairly weil withtotal phosphorus in the first part of the experiment. It is obvious that very high zooplanktonbiomasses are temporariiy abie to maintain phosphorus concentrations in the water phase of theaquatic ecosystem. The net phosphorus loads fromthe sediment, calculated from the concentrationchanges in the enciosure, do not reveal whetherthese changes are due to changes in the sedimentation rate or to changes in release from thesediment.

5. DISCUSSION5.1 The use of enclosures in enrichment

experimentsLarge enclosures are a useful alternative foriaboratory tests and for wholelake experimentswhen studying the effects of sewage on lakerecipients. Many naturai phenomena are moreeasily quantified in enclosures than in lakes. It isalso possible to make parallel experiments and toexamine the effects of different factors in acontrolied manner. Nevertheless, enclosures arepartially closed systems and it is therefore notpossible to make direct comparisons between theenclosure and the iake ecosystem. Moreover,observations made in the enclosures are not

Table 7. Changes in the chemical and biological characteristics of an enclosure with a polluted sediment in LakeRohmavesi.

Variable Untreated phase After enrichment22.6. 29.6. 6.7. 13.7. 27.7. 3.8. 10.8.

t°C 13.8 16.5 17.8 21.2 21.3 20.6 21.3) 02, mg h 13.5 10.5 9.3 7.9 9.6 8.4 7.8pH 9.2 8.2 7.6 8.3 7.9 7.7 7.3COD, rng °2 1—1 16.5 13.9 15.4 15.0 13.5 12.9 14.5BOD7,mg 02 1 11.0 7.3 9.3 7.0 5.8 5.5 9.2Tot.N, j.g 2300 1700 2200 1400 3100 1800 2200NO,—N, i’g 45 67 94 90 65 44 33N03—N, ig 11 730 540 290 100 1600 440 300NH4—N, g i1 250 270 270 300 200 200 74Tot.P, ig 1 170 190 350 400 500 560 860zlTot.P, g 1 d1 +2.9 +22.9 +7.1 +8.6 +42.9P04—P, g l’ 50 99 220 300 440 470 590zXTot.P, mg m2d1 +7.2 +57.2 +17.8 +21.5 +107.2Chl.a, pg 1 100 59 45 27 12 19 100Zooplankton (Cladocera±Copepoda),ind.1’ 410 1300 1600 5000 3600 1800

) 1 m above the bottom

83

necessary valid in a natural water body.The physical, chemical and biological conditions

in enclosures differ to some extent from those inthe surrounding water. However, the larger theenclosure, the smaller the difference, and it maywell be said that short-term experiments in fairlylarge enclosures reflect well enough the naturalconditions in a lake.

These experiments were performed in threedifferent plastic enclosures placed in lakes: (1)baglike, the lower end closed containing 2.5 m3 ofwater, (2) cylindrical, each end open containing 1m3 of water and (3) cubical, each end opencontaining 56 m3 of water.

The production of periphytic algae was significant in each type of enclosure. Detached wallgrowth caused problems in sampling and in theanalyses. Periphyton caused the least samplingproblems in large enclosures. A loss of illuminationinside the cylinder was detected only in the bagiikeenclosures during the 9 .weeks experiments. Noshading-effects due to wall growth, measured witha submerged photocell, were detected in the largeenclosures during the 3 month experiments.

The disturbing effects of enclosures increaseduring long-term experiments. Usually a period of2—5 weeks is suitable in experiments with small(less than 10 m3) enclosures. The larger theenclosures, the longer the experimental times thatcan be used. The phytoplankton communities canremain similar to those of the lake even after 2.5years of isolation (Lack and Lund 1974).

A lack of turbulence caused a rapid loss ofnutrients and a decline in the phytoplanktonbiomass in the smalier types of enclosures.However, the phenomenon was less pronounced inthe larger ones. After sewage enrichment in theenclosures phytoplankton was more abundant inthe large enclosures than in the lake. In contrast,plankton production in the smaller enclosures wasless, or at its maximum as high as in the open lakein spite of nutrient enrichment. The ratio of watervolume to wall-area has to be large enough tomaintain the natural phyto- and zooplanktoncommunities (Smyly 1976, Kenttämies 1981). Inlarge enclosures with a small volume to wall-arearatio, the plankton communities are similar topelagic ones. In studies with enclosures of less than10 m3, the phytoplankton has changed to the typeof community found in a pond or in the littoralzone (Lack and Lund 1974, Smyly 1976).

The effects of sewage addition on primaryproduction were more pronounced after introducing fish in a large enclosure. Fish predationprevents zooplankton communities from becomingtoo large. Fish excretion and mechanical stirring

stimulate the nutrient cycles and retard theoligotrophication of an enclosure.

The largest (56 m3) experimental enclosure witha natural bottom and planktivorous fish population appeared to he the most suitable forlimnological studies. The most notable difficultywas the significant wall growth of periphytic algae.This cannot be avoided in any kind of enclosure.However, analysis of the chlorophyil content ofthe periphytic aigae gives a fairly good insight intothe importance of periphytic production.

5.2 Chemical vs. sewage enrichmentPure chemical enrichment, instead of sewage, ispreferred in many empirical studies on the effectson lakes of municipal waste water. Besides beingmore practical, chemical addition has been regarded as a more scientific way of study. Ourresults from chemical enrichment were quitesimilar to those obtained in corresponding studieselsewhere (Lundgren 1978a, b, Flett et al. 1980,Schindler 1980). Phosphorus enrich ment enhancednitrogen fixation, but only phosphorus andnitrogen enrichment together increased primaryproductivity. However, this study was planned toprovide answers concerning the total effects of anew type of sewage treatment method on therecipients. The total effects of oider, widely usedtreatments can he seen in thousands of eutrophiedlakes and rivers! According to this study, the totalphosphorus and nitrogen concentrations did notvery well explain the effects of sewage even onprimary production. Many other factors, such asinhihition, heterotrophic activity and species succession, affect the results of such experiments. Theeffects of Ioading on the whole ecosystem may hethe most prominent, in the hypertrophic level ofproduction especiaHy. The separate testing of alithe components of sewage (macro- and micronutrients, toxic compounds organics etc.) in awhole ecosystem is not a practical possibility. Itwas hoped that testing and comparing the totaleffects of a sewage type, a real treatment method,on lake ecosystems would give straight answers toits usefulness.

5.3 Effects of nitrogen removalon the aquatic ecosystem

The primary hypothesis that cyanophyta would hefavoured by a relatively small lowering of thenitrogen load from sewage, had to he abandoned at

84

least in the circumstances prevailing in the heavilyloaded, brown water lake Rohmavesi, the recipientof the city of Mikkeli. Nutrient loads from thesediments were high and dominated primaryproduction. The duration of the internal phosphorus load from sediments might be fairly longafter the impiementation of phosphorus removalfrom sewage, and thus the total benefits of bettertreatment will only be seen in the future, maybenot until decades have passed. The problemsassociated with restoring polluted lakes to theirformer condition are well known in many countrieswhere the chemical removal of phosphorus fromsewage has been implemented (Björk 1982).However, the internal load of phosphorus can becontrolled to some degree by external measuressuch as regulation of the oxygen consuming load orartificial oxygenation of the hypolimnion. Thechemical oxygenation of water with nitrates fromnitrifying sewage treatment plants has beenintroduced as a new mitigation method, too (Ripi1976).

6. CONCLUSIONS

The role of nitrogen as a factor limiting primaryproduction in lakes is not yet well understood. Inaddition to direct effects, nitrogen aiso has indirecteffects on primary production by regulatingheterotrophic activity. Therefore the question ofnitrogen removal from sewage also has no simpieanswer. Our resuits imply that the decision onnitrogen removal should be made case by case. Insome lakes, like lake Vesijärvi, biooms of nitrogenfixing blue-green aigae may at least partly compensate for nitrogen removal and, as such, causeproblems for recreation and other forms of wateruse. In typical Finnish humic lakes, however, bluegreens hardly ever occur in high biomasses. Thismight reflect the iack of some natural prerequisitesfor their growth. In such cases nitrogen removalwouid have only beneficiai effects.

A simple test procedure, like aigal assays andenrichment experiments in enclosures, would forma more soiid basis for decision-making. Taking intoaccount the high costs of the impiementation ofnitrogen removal in the treatment plants, the costsof this type of testing are only marginal.

7. SUMMARY

Eutrophication is the main probiem resuiting fromthe discharge of municipai waste waters towatercourses. In Finland an average of about 90 percent of sewage phosphorus is removed in treatmentplants. During the last fifteen years the phosphorusload has steadily gone down, while the nitrogenload has increased. Today the N:P ratio of treatedsewage is cleariy over 10, whereas in the early1970’s it was stili less than 10. However, since inmany cases the effects of purification on lakes haveremained smalier than expected, nitrogen removalhas been considered as the next step in improvingsewage treatment. If the probiem to be solved isthe toxic level of nitrogen compounds in therecipient, the removal of nitrogen from sewagewould have only beneficiai effects. However, it hasbeen proposed that nitrogen-fixing cyanobacteriaincrease in eutrophic lakes, thus compensating forthe benefits of nitrogen removal in the treatmentpiant and as such being harmful for recreationaland other water use.

The aim of this study was to increase ourunderstanding of the eutrophication process, andespecialiy knowledge concerning the role ofnitrogen fixation in heavily loaded recipients.

According to a preliminary study carried out inlake Vesijärvi, southern Finland, planktonic nitrogen fixation was responsible for 31 per cent of theannual nitrogen input to a eutrophic lake wheresewage loading had been ceased entirely a coupie ofyears earlier.

The experimental part of the study wasperformed using enclosures (limnocorrals). A fairlylarge (56 m3), benthic type enclosure with a naturalsediment bottom proved to be suitable for theecosystem studies in which the effects of bothtreated sewage and internal load from old sediments were investigated. Ali three treatment plantsused in the study as sewage producers employedthe active sludge method and simultaneous phosphorus precipitation with ferrous sulphate. Nitrogen removai with a nitrification-denitrificationprocess was employed in two plants. The effects onthe ecosystem were tested a) in a heavily loaded, b)in a previousiy loaded and c) in a pristine forestlake.

The experiment with chemicai N and Penrichment proved that only nitrogen and phosphorus together could increase primary productionsignificantly in the eutrophic lake Vesijärvi.Nitrogen fixation by blue-green aigae was enhanced by P enrichment, while nitrogen additionaione inhibited it. The dominant species both in

85

the control and in ali three manipuiated enciosureswere Oscillatoria agardhii and Microcystis sp.

The results of algai tests on the hmiting nutrientcarried out using a green aiga Selenastrumcapricornuturn were in good agreement with thechemical nutrient ratios. The different response ofperiphyton growth to the same total P concentration in the chemicaHy enriched enclosure and thatin the sewage enriched enciosure was prominent. Sewage resulted in much better growth ofperiphyton than chemical P enrichment, while oniyslight differences were observed in pianktonicproduction. In spite of the different chemicalcomposition, heterocystous blue-green algae appeared to dominate simultaneousiy in the lake andin ali the enclosures. Different nutrient ratios didnot appear to regulate nitrogen fixation verystrongiy if N or N+P are limiting factors.

The response in the benthic enclosure in anohgotrophic, pristine forest lake to sewage enrichment was a peak in growth followed by a decreasein production. Grazing by zoopiankton, togetherwith the nutrients concents, reguiated primaryproduction. Thinning of zoopiankton by fishfingerhngs effectiveiy favoured primary production. The addition of efficientiy treated (N+Premoval) sewage did not favour blue-green algae,which reached a maximurn of only 1.4 per cent ofthe totai phytoplankton biomass. No nitrogenfixation could be observed, either.

In the third, heavily loaded recipient only large,benthic type enclosures were used. Treated sewagecleariy increased the phytopiankton biomass in thiscase, too. The internal ioad made the ratio of totalN to total P decrease in ali the enclosures,inciuding the controi one. A very dense zoopiankton popuiation caused overgrazing of phytoplankton towards the end of the experiment.Despite differences in the sewage used forenrichment, the phytopiankton succession duringthe summer was similar in ali enclosures, changingfrom Chlorophyta to Pyrrophyta. Cyanophytaremained rare thus implying that the removai ofnitrogen from sewage was not able to enhance theirgrowth in this recipient.

The internal net load of phosphorus fromsediments of the heavily ioaded recipient variedfrom 7.25 mg m2 d’ to 107 mg m2 d’.Since the net ioad was caicuiated on the basis ofconcentration changes in the water phase, changesin sedimentation rate and the actuai reiease fromsediments couid not be separated from each other.In any case, the internai ioad from sediments wasof decisive importance in the phosphorus budget ofthis heaviiy polluted lake area.

AKNOWLEDGEMENTS

The study was financed by the National Board ofWaters and Environment and the Foundation forResearch on Natural Resources in Finland. However, without the coliaboration with many authorities and iaboratories this study would not havebeen possibie. We would like to thank Helsinkiand Mikkeli Water and Environment Districts foranaiyticai work, as wefl as Lahti, Hyvinkää andMikkeli municipal water and sewage works forproviding the sewage for the experiments. TheGame and Fisheries Research Institute, Evo fieldstation, helped in the field work in VaikeaMustajärvi. Numerous persons contributed to thestudy. In particuiar we wish to express ourgratitude to Juha Keto, Jarmo Kivinen and OlaviSandman for collaboration. Matti Valve andMarkku Mäkelä gave valuable comments duringthe study. Liisa Lepistö, Pirkko Kokkonen, ReijaJokipii and Maija Niemelä did the microscopy onphyto- and zooplankton, for which our thanks aredue. Finaily we would like to thank John Deromefor revising the Enghsh ianguage of the text.

TIIVISTELMÄ

Järvien rehevöityminen on asumajätevesien huomattavin vesistöhaitta. Nykyään n. 90 prosenttiaasumajätevesien fosforista poistetaan Suomessapuhdistamoilla. Viimeisten viidentoista vuoden aikana asumaj ätevesista johtuva fosforikuorma onkinjatkuvasti laskenut, kun taas typpikuorma on lisääntynyt. Puhdistetun jäteveden N:P on nykyäänselvästi yli kymmenen kun se 1970-luvun alkupuolella oli selvästi sen alle. Kuitenkin monissa vesistöissä puhdistuksen vaikutus vesistön tilaan onjäänyt odotettua vähäisemmäksi. Typen tehokkaampaa poistoa on harkittu seuraavaksi, jätevesienpuhdistusta parantavaksi askeleeksi. Mikäli typenliian korkea, myrkyllinen pitoisuustaso on typpiyhdisteiden pääongelmana vastaanottavassa vesistössä, on typenpoiston tehostamisesta yksiselitteisesti hyötyä. Rehevöitymishaittojen torjuntaan typenpoiston on odotettu soveltuvan huonommin,mikäli tällöin stimuloidaan typpeä sitovien, vedenkäytölle huomattavan haitallisten sinilevien massaesiintymistä.

86

Tämän tutkimuksen tarkoituksena oli lisätä jätevesien aiheuttaman rehevöitymisilmiön tuntemusta ja erityisesti tietoa typensidonnasta raskaastikuormitetuissa vesistöissä.

Esitutkimuksessa todettiin että rehevöityneessäHollolan Vesijärvessä, jota Lahden kaupunki olilakannut kuormittamasta pari vuotta aikaisemmin,planktinen typensidonta oli 31 prosenttia typenvuosikuormasta.

Kokeelliset vesistötutkimukset tehtiin muovikelmusta rakennetuissa, vesistöön sijoitetuissa koealtaissa. Vertailtaessa eri allastyyppien sopivuuttaosoittautui kohtalaisen tilava (56 m3), n. 3,5 m syvyiseen veteen sijoitettu kuutionmuotoinen allasparhaaksi silloin, kun haluttiin tutkia myös pohjankuormitusvaikutusta. Jätevesiä saatiin tutkimuksiin kolmesta eri aktiivilietepuhdistamosta, jotkakaikki käyttivät fosforin rinnakkaissaostusmenetelmää ferrosulfaatti saostuskemikaalina. Kahdessalaitoksessa toteutettiin typenpoistoa biologisellanitrifikaatio-denitrifikaatio-menetelmällä. Koevesistöinä, joissa jätevesien vaikutuksia testattiin olia) aikaisemmin kuormitettu vesistö (Vesijärvi) b)luonnontilainen metsäjärvi (Valkea Mustajärvi,Evo), c) raskaan kuormituksen vaikutuksen alainenvesistö (Rohmavesi).

Tutkittaessa Vesijärvellä epäorgaanisen typen jafosfaatin vaikutusta perustuotantoon todettiin, että vain nämä ravinteet yhdessä aikaansaivat merkittävän tuotannon lisäyksen. Sinilevien typensidontavoimistui fosfaatin lisäyksestä, kun taas yksinomainen typpiravinteiden lisäys ehkäisi typensidontaa.Kokeiden aikana olivat valtalajeina kaikissa, niinkontrolli- kuin käsittelyaltaissakin sinilevälajit Oscillatoria agardhii ja Microcystis sp. Koealtaiden vedestä tehdyt levätestit, joissa käytettiin testilevänäviherlevä Selenastrum capricornutumia, antoivatkemiallisten ravinnesuhteiden kanssa yhtäläisiä tuloksia minimiravinteesta. Perifytontuotanto reagoijätevesilisäykseen ja sen fosforipitoisuutta vastaavaan fosfaattilisäykseen täysin eri tavoin siten, ettäjätevesilisäys aikaansai kertaluokkaa suuremmankasvun. Planktinen perustuotanto erosi altaissa sensijaan vain vähäisessä määrin. Huolimatta vedenerilaisesta kemiallisesta koostumuksesta heterokystillisistä sinilevistä kehittyi dominoiva ryhmä samanaikaisesti järvessä ja eri koealtaissa. Jossainmäärin erilaiset N:P-suhteet eivät näyttäneet säätelevän typensidontaa, jos minimiravinteena oli N taiN±P.

Pienessä, kirkasvetisessä metsäjärvessä koealtaaseen tehty jätevesilisäys aikaansai tuotantohuipunjota seurasi nopea eläinplanktonin “laiduntamisesta” johtuva lasku. Eläinplankton sääteli näin yhdessä ravinnepitoisuuksien kanssa perustuotantoa.Runsaan kalanpoikaspopulaation siirto koealtaa

seen nosti tehokkaasti perustuotantoa, kun eläinplanktontiheys laski romahdusmaisesti kalojenpredaation seurauksena. Jätevesi, jonka fosfori- jatyppipitoisuutta oli huomattavasti alennettu puhdistuksessa, ei kuitenkaan suosinut sinilevien kasvua. Sinileväbiomassa oli suurimmillaan vain 1,4 %kasviplanktonin kokonaisbiomassasta. Typensidontaa ei myöskään havaittu.

Mikkelin jätevesien kuormittaman Rohmavedentutkimuksessa käytettiin Valkea Mustajärven tapaan suurta koeallastyyppiä, jossa on luonnonpohja. Puhdistetun jäteveden lisäys (2 %) altaaseen nosti myös täällä jonkin verran perustuotantoa. Sedimentistä lähtevä altaan sisäinen kuorma oli kuitenkin vanhalla jätevesien purkualueella huomattava jase sai ravinnesuhteen tot.N:tot.P pienenemään kaikissa koealtaissa kontrolli mukaan lukien. Altaisiinkasvoi ylitiheä eläinplanktonkanta, joka ilmeisestirajoitti perustuotantoa. Kasviplanktonsukkessio olikoealtaissa samantapainen vaikka ravinnelisäyksetaltaisiin erosivatkin eri jätevesityypeissä. Alkutilanteen Pyrrophyta-Chlorophyta dominanssi muuttuiensin Chlorophyta- ja sitten takaisin Pyrrophytadominanssiksi. Sinilevät (Cyanophyta) jäivät kaikissa kokeiden vaiheissa melko vähämerkityksellisiksi, eikä fosforinpoistoon yhdistettyä typenpoistoa voida pitää Rohmaveden olosuhteissa sinilevienkasvua suosivana.

Rohmavedellä oli fosforin nettokuorma sedimentistä 7,25—107 mg m2 d—’. Nettokuormamääritettiin veden konsentraatiomuutosten perusteella suljetussa systeemissä, eikä siitä voida näinerottaa todellisia liukenemis- ja sedimentoitumisnopeuksia. Huomattavan suurella sedimentistä lähtevällä fosforikuormituksella oli ratkaiseva merkitys raskaasti kuormitetun vesistönosan fosforitaseessa.

REFERENCES

Alexander, M. 1977. Introduction to soil microbiology.2nd ed. 467 p. New York.

Björk, S. 1982. Goals, methods and possibilities for directing development of limnic ecosystems. Hydrobiologia86:177—183.

Burris, R.M. 1972. Measurements of biological N2fixation with 15 N2 and acetylene. pp. 3—14 inSorokin, 3. and Kadota, M. (eds.): Techniques for theassessment of microbial production and decomposition in fresh waters. IBP Handbook No. 23.

87

Flett, R.J., Schindler, D.W., Hamilton, R.D. &Campbell, N.E.R. 1980. Nitrogen fixation in Canadian precambrian shield lakes. Can. J. Fish. Aquat.Sci. 37:494—505.

Forsberg, C., Ryding, S-O., Claesson, A. & Forsberg, A.1978. Water chemical analyses and/or aigal assay? —

Sewage effluent and polluted lake water studies. Mitt.Int. Ver. Limnol. 21:352—363.

Kanninen, J. 1980. Typen merkitys perustuotannonrajoittajana kolmessa eutrofisessa Etelä-Suomen järvessä. [Role of nitrogen as a growth limiting factor inthree eutrophic lakes in southern Finland.]. M. Sc.Thesis. Univ. of Helsinki. Dept. of Limnology. 91 p.

Kauppi, L. 1979. Phosphorus and nitrogen input fromrural population, agriculture and forest fertilizationto watercourses. Publications of the Water ResearchInstitute, 34:35—46.

Kenttämies, K. 1981. Asumajätevesien kasvinravinteiden,erityisesti typen vaikutus vesien perustuotantoon;allaskokeita oligotrofisissa järvissä. [Effect of wastewater nutrients, especially nitrogen, on aquaticprimary production.] pp. 61—72 in Sipponen, M.(ed.), Tarvitaanko typen poistoa jätevesistä. Vesi- jaKalatalousmiehet ry. Helsinki.

Lack, T.J. & Lund, J.W.G. 1974. Observations andexperiments on the phytoplankton of Blelhem Tarn,English Lake District. 1. The experimental tubes.Freshwat. BioI. 4:399—415.

Lundgren, A. 1978a. Nitrogen fixation induced byphosphorus fertilization of a subarctic lake. EcoI.Buli. (Stockholm) 26:52—59.

Lundgren, A. 1978b. Experimental lake fertilization inthe Kuokkel area, northern Sweden: Changes insestonic carbon and the role of phytoplankton. Mitt.Int. Ver. Limnol. 20: 863—868.

National Board of Waters 1981. Vesihallinnon analyysimenetelmät. Methods of water analyses used by waterauthorities. National Board of Waters, Finland,Report 213. 136 p.

Persson, P-E. 1982. Muddy odour: a problem associated

with extreme eutrophication. Hydrobiologia 86:161—164.

Rhee, G.Y. & Gotham, I.J. 1980. Optimum N:P ratiosand coexistence of planktonic algae. J. Phycol.16:486—489.

RipI, W. 1976. Biochemical oxidation of polluted lakesediment with nitrate — A new lake restorationmethod. Ambio 5:132—135.

Schindler, D.W., Ruszczynski, T. & Fee, E.J. 1980.Hypolimnion injection of nutrient effluents asmethod for reducing eutrophication. J. Fish. Res. Bd.Can. 37:320—327.

Seppänen, P. 1970. Typpikysymys vesiensuojelussa. [Theproblem of nitrogen in water protection.] Vesitalous11:12—15.

Smyly, W.J.P. 1976. Some effects of enclosure on thezooplankton in a small lake. Freshwat. Biol. 6:241—251.

Valve, M. 1984. Combined nitrogen and phosphorusremoval from wastewaters. Publications of the V7aterResearch Institute, 58:28—54.

Vollenweider, R.A. 1970. Scientific fundamentals of theeutrophication of lakes and flowing waters, withparticural reference to nitrogen and phosphorus asfactors in eutrophication. OECD. 159 p.

Vuorio, H. 1977. Fixation of molecular nitrogen byplanktonic cyanobacteria in the Helsinki sea area in1974. English summary. Rep. Water ConservationLab., City of Helsinki 9(1):1—58.

Yrjänä, E-R. 1982. Planktisten sinilevien typensidontaHollolan Vesijärvessä. [Nitrogen fixation by planktonic blue-green algae in lake Vesijärvi, Hollola].M.Sc. Thesis. Univ. of Helsinki. Dept. of Limnology.89 p.

Yrjänä, E-R. & Kauppi, L. 1984. Kasvinravinteetrehevyystason säätelijöinä. [Plant nutrients and regulation of the trophic status of lakes.] pp. 103—117 in:Kylä-Harakka T. & Heinonen P. (eds.) Vesiensuojeluo tavoitteiden asettaminen. Vesi- ja kalatalousalanammattijärjestö VKA ry., Helsinki.