THIS WATERIALI.tAr BePROTecTED BYCOPYRIGHTI.A~

(T;T~:- "1?t 1.8.COD£)Minneapolis 55455

Robert E. Carlsonch2 Center, University of Minnesota,I . al ResearLimno OglC

\

A trophic state index for lakes1

Abstract

A numerical trophic state index for lakes has been developed that incorporates mostlakes in a scale of 0 to 100. Each major division (10, 20, 30, etc.) represents a doublingin algal biomass. The index number can be calculated from any of several parameters,including Secchi disk transparency, chlorophyll, and total phosphorus.

My purpose here is to present a new ap-proach to the trophic classificationof lakes.This new approach was developed becauseof frustration in communicatingto the pub-lic both the current nature or status of lakesand their future condition after restorationwhen the traditional trophic classificationsystem is used. The system presented here,termed a trophic state index (TSI), in-volves new methods both of definingtrophic status and of determining that statusin lakes.

All trophic classification is based on thedivision of the trophic continuum, howeverthis is defined, into a series of classestermed trophic states. Traditional systemsdivide the continuum into three classes:oligotrophic, mesotrophic, and eutrophic.There is often no clear delineation of thesedivisions. Determinations of trophic stateare made from examination of several di-verse criteria, such as shape of the oxygencurve, species composition of the bottomfauna or of the phytoplankton, concentra-tions of nutrients, and various measures ofbiomass or production. Although eachchanges from oligotrophy to eutrophy, thechanges do not occur at sharply definedplaces, nor do they all occur at the sameplace or at the same rate. Some lakes maybe considered oligotrophic by one criterionand eutrophic by another; this problem is

tI

I

1Contribution 141 from the Limnological Re-search Center, University of Minnesota. This workwas supported in part by Environmental Protec-tion Agency training grant U910028.

· Present address: Department of BiologicalSciences, Kent State University, Kent, Ohio44242.

UMNOLOGY AND OCEANOGRAPHY

sometimes circumvented by classifyinglakes that show characteristics of both oli-gotrophy and eutrophy as mesotrophic.

Two or three ill-defined trophic statescannot meet contemporary demands for asensitive, unambiguous classification sys-tem. The addition of other trophic states,such as ultra-oligotrophic, meso-eutrophic,etc., could increase the discrimination of theindex, but at present these additional divi-sions are no better defined than the firstthree and may actually add to the confu-sion by giving a false sense of accuracy andsensitivity.

I acknowledgethe advice and encourage-ment of J. Shapiro in developing this indexand in preparing the manuscript. I alsothank C. Hedman and R. Armstrong for ad-vice and H. Wright for help in manuscriptpreparation.

..

Cu"ent approaches

The large number of criteria that havebeen used to determine trophic status hascontributed to the contention that thetrophic concept is multidimensional, in-volvingaspects of nutrient loading, nutrientconcentration, productivity, faunal andfloral quantity and quality, and even lakemorphometry. As such, trophic statuscould not be evaluated by examining one ortwo parameters. Such reasoning may havefostered multiparameter indices (e.g. Bre-zonik and Shannon 1971; Michalski andConroy 1972). A multiparameter index islimited in its usefulnessbecause of the num-ber of parameters that must be measured.In addition, the linear relationship assumed

361 MARCH 1977, v. 22(2)

362

1

between the parameters in some of theseindices does not hold as evidence presentedbelow indicates.

Alternatives to the multidimensionaltrophic concept have been based on a sin-gle criterion, such as the rate of supply ofeither organic matter (Rodhe 1969) or nu-trients (Beeton and Edmondson 1972) intoa lake. Indices based on a single criterionpotentially could be both unambiguous andsensitive to change. However, there is cur-rently no consensus as to what should bethe single criterion of trophic status, andit is doubtful that an index based on a sin-gle parameter would be widely accepted.

Basis for a new index

The ideal trophic state index (TSI)should incorporate the best of both theabove approaches, retaining the expressionof the diverse aspects of trophic state foundin multiparanleter indices yet still havingthe simplicity of a single parameter index.This can be done if the commonly usedtrophic criteria are interrelated. The evi-dence is that such is the case. Vollenweider(1969, 1976), Kirchner and Dillon (1975),and Larsen and Mercier (unpublished)have developed empirical equations to pre-dict phosphorus concentration in lakesfrom knowledge of phosphorus loading.Sakamoto (1966) and Dillon and Rigler(1974) have shown a relationship betweenvernal phosphorus concentration and algalbiomass,measured as chlorophylla concen-tration. Lasenby (1975) used Secchi disktransparency to predict areal hypolimneticoxygen deficits. If many of the commonlyused trophic criteria could be related by aseries of predictive equations, it would nolonger be necessary to measure all possibletrophic parameters to determine trophicstatus. A single trophic criterion, e.g. algalbiomass,nutrient concentration,or nutrientloading, could be the basis for an indexfrom which other trophic criteria could beestimated or predicted by means of the es-tablished relationships. Alternatively, mea-surements of any of the trophic criteriacould be used to determine trophic status.

.1

i

Carlson

I chose algal biomass as the key descrip_tor for such an index largely because algalbloomsare of concern to the public. An in-dex that is particularly sensitive to suchconcern would facilitate communicationbe-tween the limnologistand the public.

Values for algal biomass itself are diffi-cult to use in the index because biomass isa poorly defined term, usually in turn esti-mated by one or more parameters such asdry or wet weight, cell volume, particulatecarbon, chlorophyll, or Secchi disk trans-parency. I constructed the index using therange of possible values for Secchi disktransparency. In addition to having valueseasily transformed into a convenient scale,Secchi disk transparency is one of the sim-plest and most often made limnologicalmeasurements. Its values are easily under-stood and appreciated.

The relationship between algal biomassand Secchi disk transparency is expressedby the equation for vertical extinction oflight in water

I"= Ioe-(k-+kt>.8, (1)

where Ie= light intensityat the depth atwhich the Secchi disk disappears, 10=in-tensityof light strikingthe water surface,kw= coefficientfor attenuationof light bywater and dissolved substances, kb =coef-

ficient for attenuation of light by particu-late matter, and z =depth at which theSecchi disk disappears. The term k'b canbe rewritten as aC, where a has the dimen-sions of m2 mg-1 and C is the concentrationof particulate matter (mg m-B). Equation1 can then be rewritten as

and rearranged to form the linear equation

(~)(ln ~:) = kw+ aC. (3)

III is about 10% of 10 (Hutchinson 1957;Tyler 1968) and can be considered to be aconstant. Alpha may vary depending onthe size and on the light-absorption and

Co

Trophic state index 363

100

. 100120

100

oo

.o, , , , " .

. " .,o 0

o

oo

.oo ,,

,.'0'.~"''''..

., , , , , , , , ,

o .

0'o

o

0.1 1 10

Secchi Disk Transparency em)

100

... _8.G 8 107 9 11 12 13

Secch i Disk Transparency (meters)

Fig. 1. Relationship between Secchi disk transparency and near-surface concentrations of chloro-phyll a. Insert: log-log transformation of same data. Dashed line indicates slope of 1.0.

light-scattering properties of the particles,but intuitively one would expect the value,which is the reciprocal of the amount ofparticulate matter per square meter in thewater column above the Secchi disk, notto vary as a function of algal concentration,and for this discussion it is considered aconstant

According to Eq. 3, Secchi disk transpar-

ency is inverselyproportional to the sum ofthe absorbance of light by water and dis-solved substances (k",) and the concentra-tion of particulate matter (C). In manylakes studied, k", is apparently small in re-lation to C, and hyperbolic curves are ob-tained when Secchi disk transparency isplotted against parameters related to algalbiomass, such as chlorophyll a (Fig. 1).

;>E

f 10"'5.

isPI 80

E

I- .CI .E-

f"nil 60 0

= ..>.s: . .a. 00 .. .L. .0s: 40

0u 'f ·...

...

20-1. Ie... ..<:. 0.. .. "'.... .. ....-2

364

Constructionof the index

I defined trophic states for the indexusing each doubling of algal biomass as thecriterion for the division between eachstate, i.e. each time the concentration ofalgal biomass doubles from some basevalue, a new trophic state will be recog-nized. Because of the reciprocal relation-ship between biomass concentration andSecchi disk transparency, each doubling inbiomass would result in halving transpar-ency. By transforming Secchi disk valuesto the logarithm to the base 2, each bio-mass doubling would be represented by awhole integer at Secchi disk values of 1m, 2 m, 4 m, 8 m, etc.

I felt that the zero point on the scaleshould be located at a Secchi disk (SD)value greater than any yet reported. Thegreatest value reported by Hutchinson(1957) is 41.6 m in Lake Masyuko,Japan.The next greatest integer on the lo~ scaleis at 64 m. A trophic state index (TSI)value of 0 at 64 m is obtained by subtract-ing the lo~ of 64 from an indexingnumberof 6, giving a final TSI equation of

TSI=10(6-10~SD). (4)

The value is multiplied by 10 to give thescale a range of 0 to 100 rather than 0 to10. Two significant digits are all that canbe reasonably expected. The completedscale begins at 0 at SD=64 m, with 32 mbeing 10; 16 m, 20j 8 m, 30; etc. The theo-retical limit is indefinite, but the practicallimit is 100 or 110 (transparency values of6.4 and 3.2 em).

The scale is intentionally designed to benumerical rather than nomenclatural. Be-cause of the small number of categoriesusually allotted in nomenclatural systems,there is both a loss of information whenlakes are lumped together and a lack ofsensitivity to trophic changes.This problemis alleviated as more and more trophicstate names are added, but the system soonbecomes so encumbered that it loses itsappeal. The scale presented here, a nu-merical classification with more than 100trophic categories, avoids these problems.

Carlson

At the same time it retains the originalmeaning of the nomenclatural trophic sys-tem by using major divisions (10, 20, 30,etc.) that roughly correspond to existingconcepts of trophic groupings.

Adc1ltion of other parameters

The regression of Secchi disk againstchlorophylla and total phosphorus was cal-culated from data readily available. Thenumber of data points and the number ofparameters used can be expanded. Thedata used came from Shapiro and Pfann-kuch (unpublished), Schelskeet al. (1972),Powers et al. (1972), Lawson (1972), Me-gard (unpublished), and Carlson (1975).

Chlorophyll a (Fig. 1) does not give alinear fit to the model predicted in Eq. 2.A nonlinear element in the relationship ne-cessitated a log-log transformlition of thedata. The resulting equation is

In SD=2.04- 0.68InChI (5)(r =0.93, n= 147),

where SD transparency is in meters andChl a concentration is in milligrams percubic meter taken near the sudace. Onepossible explanation for this exponentialrelationship may be that as algal densityincreases, the algae become increasinglylight limited. In response to lower lightper unit cell, more chlorophyll may be pro-duced (Steele 1962).

When all available data were included,the regression of total phosphorus (TP inmg m-3) against the reciprocal of Secchidisk transparency yielded

SD =64.9jTP (6)(r =0.89, n =61).

Total phosphorus should correlate bestwith transparency when phosphorus is themajor factor limiting growth. Correlationsmay be poor during spring and fall over-turn when algal production tends to belimited by temperature or light. Prelimi- "',nary use of the above regression coefficients .

in the index suggested that the equationdoes predict the average seasonal relation-ship between total phosphorus and trans-

l

r

Table 1. CompJetedtrophic state index andits associatedparameters.

ID-o

102030405060708090

100

Secch1dl~

643216

8421

0.50.250.120.062

Surfacephosphorus

~0.751.5

36

1224lie96

192384ill

"S'ilri'acechlorophyll

(mg/m3)0.040.120.340.942.66.4

2056

154427

~

parency, but its predictions for phosphorusare too high in spring and fall and toolow during summer. Because the equationis to be used in an index where agreementof correlated parameters is emphasized, Idecided to use only summer values in theregressionto provide the best agreement oftotal phosphorus with algal parameters dur-ing the season when sampling would nor-mally be done.

There were not enough data availablefor phosphorus and transparency duringJuly and August to produce a meaningfulregression. Instead, chlorophyll was re-gressed against total phosphorus and theresulting equation combined with Eq. 4 toproduce a phosphorus-transparency equa-tion. The chlorophyll-total phosphorus forJuly and August data points yielded

In Chi = 1.449 In TP - 2.442(r = 0.846, n = 43 ).This equation is similar to that derived byDillon and Rigler (1974) for the relation-ship between vernal total phosphorus andsummer chlorophyll:

In ChI = 1.449 In TP - 2.616.

Combining5 with 7 produced

In SD = 3.876 - 0.98 In TP,

or approximately

SD =48(ljTP).

This equation was used in the index.The trophic state index can now be com-

puted from Secchi disk transparency, chlo-

I

.

Trophic state index 365

rophyll, or total phosphorus. The computa-tional forms of the equations are

TSI(SD) =10(6_1~S~). (11)

TSI(ChI) =10(6- 2.04-0~~In Chi), (12)and

(In 48

)TSI(TP) =10 6- InT:. (13)

The completedscaleand its associatedpa-rametersare shownin Table 1.

Using the index

Seasonal TSI values for several Minne-sota lakes are plotted in Fig. 2. The TSIvalues for total phosphorus tended to fluc-tuate less than those for the biological pa-rameters, which approached the valuesbased on phosphorus during July and espe-cially August and September. The extremefluctuations in the TSI based on biologicalparameters in May and June may resultfrom springtime crashes in algal popula-tions. In Lake Harriet, the chlorophyll andSecchi disk TSI values remained below thephosphorus TSI throughout summer. Thisdivergence of the biological TSI valuesfrom the phosphorus TSI cannot yet be ex-plained, but it does emphasize one of the

(7) strongest advantages of individually deter-mined trophic indices. Allparameters whentransformed to the trophic scale shouldhave the same value. Any divergence fromthis value by one or more parameters de-mands investigation. For example, is itreally true that Lake Harriet is P limited,

(8) since it seems to be producing less binmassthan the phosphorus levels would suggest?Thus, the TSI scale not only classifies the

(9) lake but can serve as an internal check onassumptions about the relationships amongvarinus components of the lake ecosystem.

To use the index for classifying lakes re-quires that a single number be generatedthat adequately reflects the trophic statusof the lake. It should be emphasized thatthe number generated is only an index of

"

(10)

HALSTED B4Y

lAK£ NlNttE'0..

LAKE tWlRIET

I A I S I 0 I N

Fig. 2. Seasonal variation in tropbic state in-dices calculated from Seccbi disk transparency( .), cb1orophyll /J (0), and total phospboros( X) for three Minnesota lakes in 1972.

the trophic status of a lake and does notdefine the trophic status. In other words,chlorophyllor total phosphorus are not con-sidered as the basis of a definitionof trophicstate but only as indicators of a morebroadly defined concept. The best indi-cator of trophic status may vary from laketo lake and also seasonally,so the best in-dex to use should be chosen on pragmaticgrounds.

Secchi disk transparency might be ex-pected to give erroneous values in lakescontaining high amounts of nonalgal par-ticulate matter, in highly colored lakes(those with a high kv,), and in extremelyclear lakes where kv,again is an importantfactor in light extinction. The advantageof using the Secchi disk is that it is an ex-tremely simple and cheap measurementandusually provides a TSI value similar to thatobtained for chlorophylI.

Chlorophyll a values are apparently freefrom interference, especially if the valuesare corrected for pheophytin. It may bethat the number derived from chlorophyllis best for estimating algal biomass in mostlakes and that priority should be given forits use as a trophic state indicator.

Accurate index values from total phos-phorus depend on the assumptions thatphosphorus is the major limiting factor for

Carlson

algal growth and that the concentrationsof all forms of phosphorus present are afunction of algal biomass. The former as-sumption does not seem to hold duringspring, fall. or winter (Fig. 2) nor is thelatter the case in lakes having high ortho-phosphorus values (such as some lakes re-ceiving domestic sewage) or in highly col-ored lakes where some forms of phosphorusmay be tied up with humic acids (Moyerand Thomas 1970). The advantage of thephosphorus index is that it is relativelystable throughout the year and, because ofthis, can supply a meaningful value duringseasonswhen algal biomass is far below itspotential maximum.

In instances where the biological TSIvalues diverge from that predicted by thephosphorus index, it is not satisfactory toaverage the different values, as the resultingvalue would neither represent the trophicstate predicted by the nutrient concentra-tion nor the biological condition estimatedfrom biological criteria. I suggest that forpurposes of classificationpriority be givento the biological parameters, especially thechlorophyllindex, during summer, and per-haps to the phosphorus values in spring,fall, and winter, when the algae may belimited by factors other than phosphorus.These priorities would result in about thesame index value during any season of theyear.

Some period should be identified duringwhich samples should be taken for lakeclassification. It might be argued that sam-ples should be taken throughout a calen-dar year to include all variations in trophicstatus, but this would require the switchingof index parameters during the year, asmentioned above, and the large number ofsamples required might limit the usefulnessof the index.

Basing the index on samples taken dur-ing spring or fall turnover would providea phosphorus index based on mean phos-phorus concentration, a measurement usedin most predictive loading models. How-ever, some disadvantages of limiting sam-pling to overturn periods are the brief oreven nonexistent overturn in some lakes

366

80

60

40

" :zo[ojill 70..

50

ui 30

10

&0

40

20. .. I .. I ..

Trophic state index 367

and the possibility of a lessened agreementwith the biologicalparameters. In addition,mean phosphorus concentrations could becalculated at any time if appropriate mor-phometric data were available. An alter-native might be to take epilimnetic sam-ples during summer stratification, whenthere should be the best agreement betweenall of the index parameters. This wouldallow confidence in comparisons betweenstudies that used different parameters. Thisperiod would also coincide, in temperatelakes,with peak recreational usage, increas-ing the utility of the index in communicat-ing the meaning of the classificationto thegeneral public.

A calculation of TSI values from theyearly data for Lake Washington (Fig. 3),supplied by W. T. Edmondson, also illus-trates the utility of the scale. In 1957,thelake was receiving 51% of its phosphorusfrom sewage effluent, and there was a no-ticeable deterioration in water quality.Sewage diversion began in 1963, and by1968,99% of the sewage had been divertedfrom the lake (Edmondson 1910). Al-though the data plotted in Fig. 3 show theincrease and decrease in nutrients and re-sulting biological changes, the rates ofchange differ considerably for each param-eter. Chlorophyll a seems to be only slightlyaffected until 1961, whereas Secchi disktransparency shows a drastic decrease be-tween 1950 and 1955. Conversely, aftersewage diversion in 1963, chlorophyll ashows a dramatic decrease, while there isa 3-year lag before Secchi disk transpar-ency changes noticeably. These discrepan-cies in rates of response are the result ofthe inverse relationshipbetween chlorophyllconcentration and transparency. Secchidisk transparency is very sensitive to bio-mass changes at low concentrations, butis insensitive at high concentrations. Ifonlychlorophyllhad been measured duringthe 19505,it might have seemed that sig-nificant changes in the lake did not occuruntil after 1961. When chlorophyll valuesare converted to TSI, however, they canbe seen to be responding rapidly to enrich-ment, and the changes parallel those in the

t

l

o

70

- .E

Fig. 3. Top: Yearly changes in average sum-mer values of three parameters in Lake Wash-ington, Seattle, Washington. Below: Data trans-formed into trophic state index values.

TSI values based on transparency. The TSIvaluespredicted by total phosphorus closelyparallel the biological TSI values, addingstrength to the argument that phosphoruswas the key nutrient in determining thebiological trophic state of Lake Washing-ton.

Discussion

The trophic index presented here hasseveral advantages over previous attemptsat trophic classification.

The scale is numerical rather than no-menclatural, allowinga large number of in-dividual lake classes rather than three tofive distinct ones. This allows sensitivityin describing trophic changes. The major

6.

60r E.,,_..,.;;

...1

..

.ot- . Muo,rop'uc. I....I....I........I

'033 1950 ,g60 '970

368

trophic divisions have' a logical basis,namely the doubling of concentrations ofsurface algal biomass, which should makethe trophic state classification more ac-ceptable theoretically.

Although the scale itself is constructedfrom a single parameter, the mathematicalcorrelation with other parameters allowssome latitude in selecting the best one fora given situation. The use of correlatedparameters also allows trophic comparisonsbetween lakes where different types of datawere collected in different studies. Calcu-lation of the index for more than one pa-rameter for a given lake also serves as aninternal check both on methodology andon assumptionsas to relationships betweenparameters.

The data used can be minimal or exten-sive,depending on the level of accuracy de-sired and the resources available. Secchidisk values alone can give a trophic stateclassification, information that can be col-lected even by nonscientistsin public-par-ticipation programs at little expense (Sha-piro et al. 1975). If the survey is moreextensive, data on chlorophyll and totalphospborus can provide supplementary oralternative index values. The index numbergives both the public and the limnologista reasonablyaccurate impressionof a lake'swater quality. For the layman, the num-ber may have little meaning at first, but itcan readily be transformed into Secchitransparency, which is easily understood.By analogy, the Richter scale hasmeaning for people other than seismolo-gists. For the limnologist the index canbe an aid in communicationbetween him-self and other scientists. With only theTSI value, a reasonable estimate can bemade of the Secchi disk transparency,chlorophyll, and total phosphorus.

The index can be used as a predictivetool in lake-managementprograms. If themean total phosphorus after nutrient abate-ment can be predicted with loading-rateequations such as those of Vollenweider(1969,1976), then a new TSI can easily becalculated and the biological conditions of

Cal'!$on

the lake estimated. This should prove ofvalue to groups interested in determininghow much nutrient abatement is necessaryto reach a desired trophic condition.

The index can be used for regional classi-fication of all surface waters, includingstreams and rivers. Any body of watercould be classified using the total phos-phorus index, which is essentially a predic-tor of potential algal biomass. Hutchinson(1969) suggested that lakes and theirdrainage basins should be consideredas trophic systems, where a eutrophicsystem is one in which the potentialconcentration of nutrients is high. The in-dex allows the classificationof that poten-tial concentration for a watershed or reogion, at least on the basis of phosphorus.Such a classification could develop anawareness of regional patterns in trophicpotentia~ and could then also allow regionaltrophic standards to be the basis of rationalmanagement.

Finally, a trophic state index is not thesame as a water quality index. The termquality implies a subjective judgment thatis best kept separate from the concept oftrophic state. A major point of confusionwith the existing terminology is that eu-trophic is often equated with poor waterquality. Excellent, or poor, water qualitydepends on the use of that water and thelocal attitudes of the people. The definitionof trophic state and its index should re-main neutral to such subjective judgments,remaining a framework within which vari-ous evaluations of water quality can bemade. The TSI can be a valuable tool forlakemanagement,but it is alsoa valid scien-tific tool for investigationswhere an objec-tive standard of trophic state is necessary.Ihope that the index can serve as a standardof trophic measurementagainst which com-parisons can be made between the manychemical and biological components of thelake system that are related to trophicstatus. The result could be a more com-plete and dynamic picture of how thesecomponents relate to one another and tothe lake ecosystemas a whole.

-

Trophic state index 369

References

BEETON,A. M., ANDW. T. EDMONDSON.1972.The eutrophication problem. J. Fish. Res.Bd. Can. 29: 673-682.

BREZONDC,P. L., AND E. E. SHANNON. 1971.Trophic state of lakes in north centralFlorida. Fla. Water Resour. Res. CenterPubI. 13. 102 p.

c.uu.soN, R. E. 1975. Phosphorus cycling in ashallow eutrophic lake in southwestern Min-nesota. Ph.D. thesis, Univ. Minnesota.

DILLON,P. J., ANDF. H. RxcLER. 1974. Thephosphorus-chlorophyU relationship in lakes.LimnoL Oceanogr. 19: 767-773.

EDMONDSON,W. T. 1970. Phosphorus, nitro-gen, and algae in Lake Washington after di-version of sewage. Science 169: 690-691.

HUl'CHINSON,G. E. 1957. A treatise on lim-nology, v. 1. Wiley.

-. 1969. Eutrophication: Past and pres-ent, p. 17-26. In Eutrophication: Causes,consequences, correctives. NatI. Acad. Sci.Publ. 1700.

KmCHNER,W. B., ANDP. J. DILLON. 1975. Anempirical method of estimating the retentionof phosphorus in lakes. Water Resour. Res.11: 182-183.

LASENBY,D. C. 1975. Development of oxygendeficits in 14 southern Ontario lakes. LimnoJ.Oceanogr. 20: 993-999.

LAWSON,D. W. 1972. Temperature, trans-parency, and phytoplankton productivity inCrater Lake, Oregon. LimnoL Oceanogr. 17:410-417.

MICHALSJC:,M. F., AND N. CoNBOY. 1972.Water quality evaluation-Lake Alert study.Ontario Min. Environ. Rep. 23 p.

MoYER,J. R., ANDR. L. THOMAS. 1970. Or-ganic phosphorus and inositol phosphates inmolecular size fractions of a soil organic mat-ter extract. Soil Sci. Soc. Am. Proc. 34:80-83.

POWERS,C. F., D. W. SCHULTS,K. W. MALUEG,R. M. BRICE,ANDM. D. SCHULDT. 1972.Algal responses to nutrient additions in nat-ural waters. 2. Field experiments. Am. Soc.Limnol. Oceanogr. Spec. Symp. 1: 141-154.

RODDE,W. 1969. Crystallization of eutrophi-cation concepts in northern Europe, p. 50-64. In Eutrophication: Causes, conse-quences, correctives. NatI. Acad. Sci. Publ.1700.

SAKAMOTO,M. 1966. Primary production byphytoplankton community in some Japaneselakes and its dependence on lake depth. Arch.Hydrobiol. 62: 1-28.

SCHELSa, C. L., L. E. F'ELDr,M. A. SANTIACO,ANDE. F. STOERMER.1972. Nutrient en-richment and its effect on phytoplankton pro-duction and species composition in LakeSuperior. Proc. 15th Conf. Great Lakes Res.1972: 149-165.

SHAPIRO,J., J. B. LUNDQ\J1ST,ANDR. E. CARLSON.1975. Involving the public in limnology-an approach to communication. Int. Ver.Thear. Angew. Limnol. v.erb. 19: 866-874.

STEELE,J. H. 1962. Environmental control ofphotosynthesis in the sea. LimnoI. Oceanogr.7: 137-150.

TYLER,J. E. 1968. The Secchi disc. Limnol.Oceanogr. 13: 1-6.

VOLLENWEIDER,R. A. 1969. Possibilities andlimits of elementary models concerning thebudget of substances in lakes. Arch. Hydro-bioI. 66: 1-36.

-. 1976. Advances in defining criticalloading levels for phosphorus in lake eutro-phication. Mem. 1st. Ital. Idrobiol. 33: 53-83.

Submitted: 5 February 1975Accepted: 18 November 1976

Announcement

Pacific Section, American Society of Limnology and Oceanography, Inc.

The Pacific Section will meet 12-16 June 1977 at San Francisco StateUniversity, San Francisco, California. A symposium on "The San Fran-cisco Bay" is scheduled for 13 June. Symposia on "Coastal upwelling"and "Ecology of western streams and rivers" are planned for 15 June.