Fish Production Correlated with Primary Productivity, not the

Morphoedaphic Index1

John A. Downing, Celine Plante, and Sophie Lalonde

DGpartement de Sciences Biologiques, University de Montreal, C.P. 6128, Succursale 'A', Montreal, Qu4. H3C 317 Canada

Downing, J.A., C. Plante, and S. Lalonde. 1990. Fish production correlated with primary productivity, not themorphoedaphic index. Can. J. Fish. Aquat. Sci. 47: 1929-1936.

Estimates of the biological production of entire lake fish communities were collected from the published literature

on lakes covering a wide range of geographic areas and trophic status. Correlation analysis shows that fish

production is uncorrelated with the morphoedaphic index (p>0.05) but closely correlated with annual phyto-

plankton production (^ = 0.79), mean total phosphorus concentration (^ = 0.67), and annual average fish stand

ing stock (r2 = 0.67). Empirically derived regression equations are presented and compared with previous models

based on catch and yield data. Analysis of these equations suggests that conversion of phytoplankton into fishproduction is 100 times more efficient in oligotrophic lakes than hyper-eutrophic ones, but that a much lower

fraction of fish production can be channeled to sustainable yield in oligotrophic lakes. Sustained yields were

frequently as little as 10% of the annual community fish production.

On a releve", dans les ouvrages publics, des estimations de la production biologique de communaute's entieres

de poissons provenant de lacs couvrant une grande e"tendue ge"ographique et une vaste gamme d'e*tats trophiques.L'analyse de correlation de"montre que la production de poissons n'est pas correMee avec I'incide morphoe"da-

phique (p>0,05), mais qu'elle est etroitement corre'le'e avec la production annuelle de phytoplancton (r2 = 0,79),

la concentration moyenne de phosphore total (r2 = 0,67) et la biomasse annuelle moyenne de poissons (r2 = 0,67).

On pre"sente des Equations de regression obtenues empiriquement et on les compare avec des modeles ante"rieurs

reposant sur des donne"es relatives aux captures et aux rendements. L'analyse de ces Equations revele que la

conversion de phytoplancton en production de poissons est centfois plus efficace dans les lacs oligotrophes que

dans les lacs hypereutrophes, mais qu'une fraction beaucoup plus faible de la production de poissons contribue

a un rendement gquilibre" dans les lacs oligotrophes. Dans de nombreux cas, les rendements equilibre's repre"-sentaient aussi peu que 10% de la production annuelle des communaute's de poissons.

Received October 11, 1989

Accepted April 19,1990

(JA337)

Recu le 11 octobre 1989

Accepte" le 19 avril 1990

At equilibrium, unexploited fish stocks produce exactly

enough biomass to balance natural mortality. The aim

of fisheries management is to replace the greatest pos

sible fraction of natural mortality with fishing mortality without

exceeding the rate of renewal of the stock (e.g. Schaefer 1968).

The renewal of fish stocks is provided by production, which is

the "amount of tissue elaborated per unit time per unit area,

regardless of its fate'1 (Clarke 1946).

Models for the prediction of freshwater fish productivity

abound (review by Leach et al. 1987). These models disagree

about the characteristics of lakes that have the greatest influence

on fish production (Table 1). Early models were based on sim

ple morphometric measures like mean depth (Z) or lake area

(Rounsefell 1946; Rawson 1952, 1955). Later ones, like the

famous "morphoedaphic" model, augmented morphometric

data with indicators of lake fertility like alkalinity (Hayes and

Anthony 1964) and total dissolved solids (Ryder 1965). The

simplicity of the morphoedaphic index (MEI) model has made

it a staple of lake fisheries management (Ryder 1982; Leach

et al. 1987). In the 1970's, models frequently linked fish pro

duction to phytoplankton (Hrbacek 1969; Melack 1976;

McConnell et al. 1977; Oglesby 1977) and benthos (Matuszek

'Publication No. 370 of the Groupe d'Ecologie des Eaux douces,

University de Montreal.

1978) productivity. Most recent studies have used the old

approaches on new fish production data bases (Liang et al.

1981; Jenkins 1982; Jones and Hoyer 1982; Oglesby etal. 1987)

or have reanalyzed the old data (i.e. Ryder 1965) using new

combinations of variables (e.g. Hanson and Leggett 1982;

Youngs and Heimbuch 1982) or methods (Prepas 1983; Schnei

der and Haedrich 1989). Some recent multivariate models pre

dict yields of single species fisheries (e.g. Godbout and Peters

1988). Individual models have advanced alkalinity, algal bio

mass (chlorophyll a), air temperature, area, benthos standing

crop, body size, fishing effort, mean depth, phytoplankton pro

ductivity, total dissolved solids, total nitrogen concentration,

and total phosphorus concentration as the most important pre

dictors of lake fisheries production.

Probably the most important deficiency of current lake fish

ery production models is that few actually predict rates of bio

logical production; instead, many predict variables like "catch"

or "yield" (Table 1). The many different "fish production"

models yield predictions of quite different dependent variables

and thus are difficult to compare and interpret. Short-term sport

or commercial fishing yields probably do not bear a constant

relationship to long-term or sustainable yields. Further, the

relationship between catch and actual fish production is

unknown because both the amount of fishing effort expended

Can. J. Fish. Aquat. Sci., Vol. 47, 1990 1929

Table 1. Number of lakes examined, coefficients of determination (r2), and variables employed in various models for the prediction of fishproduction in lakes. Year is the year of publication, Area indicates lake area, Z is the lake mean depth, Alk. is the alkalinity, TDS is totaldissolved solids, PP is the water column primary production, Chi a is the chlorophyll a concentration in the water column, Benthos indicatesthe standing biomass of benthic invertebrates, TP is the water column total phosphorus concentration, TN is total nitrogen concentration, Air T

is the mean annual air temperature, and Effort is the annual fishing effort.

Year Author Predictor Production estimate

1946

1946

1952

1964

Rounsefell

Rounsefell

Rawson

Hayes and Anthony

12

39

13

138

0.73

0.40

0.80"

0.67b

Area

Area

Z

Area,Z,Alk

1965

1969

Ryder

Hrb&ek

23

13

0.73

0.72

TDS/Z

TN

1969

1976

1977

1977

1977

1978

1981

1982

1982

1982

1982

1982

Hrb&ek

Melack

McConnell et al.

Oglesby

Oglesby

Matuszek

Liang et al.

Hanson and Leggett

Hanson and Leggett

Jenkins

Jones and Hoyer

Schlesinger and Regier

11

24

6

19

19

11

18

20

21

290

25

43

0.72c

0.56

0.93

0.74

0.84

0.80

0.76

0.58

0.87

0.08

0.83c

0.74

PP

PP

PP

PP

Chi a

Benthos

PP

Benthos/Z

TP

TDS/Z

Chlo

Air T

1982 Youngs and Heimbuch 27 0.72 Area

1983 Prepas 23 0.70 Z

1983 Schlesinger and McCombie 92 0.61 Effort

Commercial fishing yields

Sport fishing yields

25 yr average commercial catch

Long or short term sport or commercial

catches weighted by trophic chain length

"Catch records for several years, or from

published estimates based on intensive

fishery surveys1'

Net growth increment in stocked carp

ponds

Net growth increment in carp ponds

Three consecutive years average

commercial yield

Net growth increment in experimental

swimming pools

Fish catch "...in lakes with moderate to

intensive fishing pressure"

Fish catch "...in lakes with moderate to

intensive fishing pressure"

Average annual catch over the 15 yr

period of maximum commercial yield

Net (1-yr) growth increment in stocked

ponds and lakes

"Long term commercial and sport

harvest" (Oglesby and Ryder)

"Long term commercial and sport

harvest" (Oglesby and Ryder)

"Mean of all available annual (catch)

estimates"

Sport fish harvest from creel census

Commercial and sport yields "...assumed

to be a close approximation of the lake's

MSY."

Yield data from Ryder, Oglesby and

Matuszek

"Catch records" of Ryder

Sport fish yields from creel census.

•Computed from Rawson's (1955) table 3.

bLog of yields weighted by trophic chain length.

CFY not as logarithms.

and the capture efficiency have a large influence on the amount

offish actually landed (Schlesinger and McCombie 1983; God-

bout and Peters 1988). The meaning of the variables predicted

by many of these models to either the practical management of

lake fisheries, or the theoretical study of ecology is difficult to

discern. Oglesby (1977) has stated that it would be preferable

to establish fish production models from rigorously defined and

repeatable measurements like fish production rather than rough

indices of average catch, but suggested that reliable, whole

community production data were too rare, during the mid

1970's, to permit this.

All fisheries management models assume that sustainable fish

yield is correlated with fish production. If this correlation exists,

then we would expect fish production to be correlated with the

same lake characteristics as long-term fish yield measures. This

research draws together existing measurements of production

of whole fish communities to find which of the proposed lake

characteristics is most closely correlated with fish productivity.

Methods

Data on the production of entire fish communities were

gleaned from an exhaustive survey of the primary ecological

literature published since 1969. We only retained data where

the production of all significant fish species in each lake had

been measured. Fish communities that had been subject to

recent stocking were not included in the data set because pro

duction by recently stocked fish might not reflect the natural

ecosystem productivity.

Data on biotic and abiotic characteristics of lakes were

derived from published works or in some cases were completed

through direct communication with the authors of the produc

tion studies. Morphometric characteristics collected were lake

area, volume, mean and maximum depth, and area of

watershed. Lake productivity indicators were phytoplankton

productivity, total phosphorus, nitrogen and chlorophyll a con

centrations in the water column, and conductivity (as a measure

1930 Can. J. Fish. Aquat. 5c/., Vol. 47, 1990

Table 2. Whole fish community production data and lake characteristics drawn from the published literature. Abbreviations indicate, FP: fish

production (kg-na^yr"1), FB\ fish biomass (kg-ha"1), PP: primary production (g C-m~2-yr"1), TP: total phosphorus (jxg-L"1), Cond.:conductivity (jiS-cm"1, 20°C), Z: mean depth (m). na indicates that the datum was not available.

Lakes

Alinen Mustajarvi

Batorin

Big Turkey (1985)

Big Turkey (1988)

Botjam

Char

Dalnee

Demenets

Horkkajarvi

La Luisa

Little Turkey (1985)

Little Turkey (1988)

Marion

Myastro

Nakuru

Naroch

0vre Heimdalsvatn

Red deer

Sabanilla

Vitalampa

Washington

Wishart (1985)

Wishart (1988)

FP

30.3

83.0

1.2"

3.8

17.3

2.0

398.0

228.5

15.5

276.9

3.9a

4.7

21.5

64.0

145.8

44.0

14.7

28.9

220.5

9.8

89.4

5.6a

5.7

FB

73.0

163.0

5.2"

7.3

30.3

104.0

197.0

372.2

32.2

325.7

10.9a

11.2

28.9

125.0

59.4

81.0

19.8

11.6

311.7

20.2

93.4

24.4a

24.4

PP

7.0

351.6

0.3

0.3

2.6

4.1

487.0

na

5.6

na

0.7

0.7

8.0

314.8

881.0

97.3

11.5

na

na

4.5

157.0

2.8

2.8

TP

39.3

na

4.2

4.6

13.5

3.8

na

na

114.5

na

6.2

5.0

na

na

9850.0

na

14.7

na

na

16.5

34.8

5.1

5.0

Cond.

24.3

na

41.1

41.1

25.0

196.0

na

na

43.3

na

38.8

38.8

18.2

na

18166.7

na

15.3

35.1

na

18.0

81.1

33.3

33.3

Z

3.0

3.0

12.2

12.2

3.3

10.2

31.5

3.3

7.0

1.4

6.0

6.0

2.4

5.4

2.3

9.0

4.7

3.7

0.6

2.8

33.0

2.2

2.2

Data sources

Rask and Arvola 1985; Arvola (pers.

comm.)

Winbergetal. 1972

Kelso 1985; Lam et al. 1986

Kelso 1985, 1988; Lam et al. 1986

Nyberg 1979; Ramberg 1976

Rigler 1972, 1975

Krogius et al. 1972

Gulin and Rudenko 1973

Rask and Arvola 1985; Arvola 1983;

(pers. comm.)

Hoiafk 1970

Kelso 1985; Lam et al. 1986

Kelso 1985, 1988; Lam et al. 1986

Efford 1972

Winberg et al. 1972

Vareschi and Jacobs 1984, 1985;

Plante 1987; Vareschi (pers. comm.)

Winberg et al. 1972

Kloster 1987; Lien 1978, 1981;

Tangen and Bretum 1978

Chadwick 1976

Hoieik 1970

Nyberg 1979; Ramberg 1976

Edmondson 1977, (pers. comm.);

Eggers et al. 1978

Kelso 1985; Lam et al. 1986

Kelso 1985, 1988;

Lam et al. 1986

"Production and biomass of 1+ fish and older.

of total ion concentration). Phytoplankton production data were

converted to total net annual production per unit lake area.

Chlorophyll, nitrogen, and phosphorus concentrations were

collected as annual averages. The chlorophyll concentrations

of lakes Alinen-Mustajarvi, Botjarn, Vitalampa and 0vre

Heimdalsvatn were estimated using the least-squares regres

sion relationship between log,o chlorophyll and log,0 total

phosphorus found in the data set (r2 = 0.82; n= 10). Total dis

solved solids (TDS) was inferred from conductivity measure

ments using Rodhe's (1949) table 2. This procedure yielded

predicted TDS values that were closely correlated (^ = 0.95;

n = 6) with TDS measurements in an independent set of data

(Oglesby 1977). Annual mean air temperature, which is closely

correlated with surface water temperature (StraSkraba 1980),

was used instead of water temperature (see also Schlesinger and

Regier 1982) because of its greater availability in climatic

reports (Wernstedt 1972). Annual average pH data were also

collected because of the possible influence of pH on the pro

duction of invertebrates (Plante and Downing 1989). Geo

graphic variables, latitude and altitude, were also noted where

possible.

Analysis of relationships between fish community produc

tion and lake characteristics were performed using simple and

multiple regression (Draper and Smith 1981; Gujarati 1978)

after logarithmic transformation, where necessary, which sta

bilized the variance, linearized the responses, and normalized

the residuals. Residuals were examined following the protocol

of Draper and Smith (1981).

Can. J. Fish. Aquat. ScL, Vol. 47, 1990

Results and Discussion

Fish community production data were obtained for 20 lakes

(Table 2) in a wide range of geographic locations (Fig. 1). All

estimates were actual measures of biological production (as

opposed to catch or yield), and were mostly estimated using

cohort, Allen curve, or instantaneous growth methods (Chap

man 1971; Rigler and Downing 1984). Two independent pro

duction estimates were obtained for Turkey, Little Turkey, and

Wishart Lakes (Kelso 1985, 1988) yielding 23 whole com

munity fish production estimates in all. This is a number of

estimates similar to that used in the majority of previous studies

examining average catch data (Table 1). The fish communities

consisted of between one and 22 species of fish, but there was

no correlation between species richness and lake trophic status

(e.g. both the unproductive Char Lake and the tropical Lake

Nakuru had only one important fish species). Lakes varied in

mean annual fish standing stock from 5.2 kg-ha"1 in the oli-

gotrophic Turkey Lake on the Canadian Shield (Kelso 1985) to

372 kg-ha"1 in the small, eutrophic Lake Demenets in thePskov province of the eastern USSR.

Primary Production

Only three of the 16 lake characteristics were significantly

(p<0.05) correlated with fish community production (Table 3),

thus many of the published correlations between fish yield and

various lake characteristics do not hold for fish community pro

duction. Although the top three variables in Table 3 yielded

1931

-60

-180-150-120 -90 -60 -30 6 30 60 90 120 150 180

DEGREES OF LONGITUDEFig. 1. Location of 20 lakes for which fish community production data were obtained from the publishedliterature.

Table 3. Coefficients of determination (r2) for the relationship:

\og,oFP = a + b logloX, for the fish production data in Table 2. n

is the number of observations included in the regression, and p is the

probability that an equal or greater r2 value would be obtained by

chance alone. Values of p are corrected for the number of tests

performed (Kirk 1982).

Variable n r2 o

Primary production 19

Fish biomass 23

Total phosphorus 14

Watershed area 18

Chlorophyll a 9

Latitude" 23

TDS/Z(MEI) 16

pHa 16

Air temperature 22

Conductivity 16

Zmax 23

Z 23

Lake area 23

Total nitrogen 16

Lake volume 23

Altitude 14

0.79

0.67

0.67

0.34

0.57

0.15

0.20

0.18

0.13

0.16

0.10

0.04

0.02

0.01

0.01

0.00

<0.001

<0.001

0.002

0.058

0.149

0.519

0.593

0.727

0.741

0.812

0.871

1.000

1.000

1.000

1.000

1.000

"Independent variable not log,0 transformed.

r2 values that were not significantly different from each other(r-test; Sokal and Rohlf 1981; p<0.05), the correlation coef

ficient for the relationship between annual phytoplankton pro

duction (PP; g C-m^-yr"1) and fish production (FP;kg-ha~l-yr~1) was the highest. The coefficient of determination for the relationship:

(1) log]OFP = 0.600 + 0.575 logloPP

(Fig. 2) was 0.79 (n= 19;/?<0.001). This coefficient of deter

mination is greater than all but six of those found in studies of

correlations between lake characteristics and fish catch or yield

in the published literature (Table 1). The correlation is even

better than the relationship seen between fish yield and primary

production in stocked ponds (Hrbacek 1969; Liang et al. 1981).

The lower residual variation around Eq. (1) is probably because

3-

OO

PRODUCTION DATA (THIS PAPER)ooooe CATCH DATA (OGLESBY 1977)

-1

-0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

LOG PRIMARY PRODUCTION (g C-m-yr )

Fig. 2. Relationship between fish and phytoplankton production for

literature data on whole fish community production (closed squares)

and Oglesby's (1977) fish catch data (open circles). Dashed lines indi

cate approximate 95% confidence bands for predicted values of the

mean logloF/> (Prepas 1984). Oglesby's fish yield values were con

verted to wet mass by multiplying by four (Oglesby 1977) and the

Asian data where gross production of phytoplankton was measured or

where a range of fish catch values was given are not plotted. The solid

line is predicted from Eq. (1). The four lowest points are fisheries

yield data from Lakes Superior, Huron, Michigan and Ontario. The

four highest points are the average of three consecutive years of com

mercial yield from three Indian ponds and the annual yield from an

unnamed carp pond in Israel.

our FP data are uninfluenced by variations in fishing effort.

Confidence intervals for predicted values show that 95% of pre

dicted mean fish production rates will be not less than about

half and not more than about twice the production indicated by

Eq. (1).

1932 Can. J. Fish. Aquat. ScL, Vol. 47, 1990

Equation (1) suggests that the conversion of phytoplankton

into fish production (assuming 1 gC = 10g fresh mass;

Oglesby 1977) is about 1% efficient in very unproductive, oli

gotrophic lakes like Big Turkey Lake, and as little as 0.002%

efficient in hyper-eutrophic lakes like Lake Nakuru. Our results

oppose those of Oglesby (1977). He found that ratios of fish

catch to primary production covered the same range as ours,

but the transfer of primary production to fish yield was most

efficient in hyper-eutrophic carp ponds. Oglesby stresses, how

ever, that his analysis does not necessarily represent true bio

logical efficiency because his yield data were subject to several

factors like fishery management practices, or incomplete catch

records. Our data on biological fish production suggest that the

transfer of energy from algae to fish production is several orders

of magnitude more efficient in oligotrophic ecosystems than in

eutrophic ones.

Oglesby's (1977) fish-catch data are nearly always lower than

our fish production rates (Fig. 2). This may be because the

sport and commercial yields in Oglesby's model do not account

for the production of all fish, only those landed by fishermen.

Another possibility is that the fishing yields summarized by

Oglesby (1977) are sub-maximal, especially considering that

the catches that were lowest, compared to measured production

values, were found in the Great Lakes (Lakes Superior, Huron,

Michigan, and Ontario) where distances are large and effort

might be small relative to the productive area. The latter seems

improbable, however, because Lawrie and Rahrer (1972) and

Wells and McLain (1972) suggest that the Great Lakes were

being overfished during the period when these catch data were

recorded. It is plausible that Rounsefell (1946) was correct, that

extremely large lakes with little perimeter, littoral zone, or allo-

chthonous energy input yield low rates of fish production rel

ative to autochthonous primary production.

The separation between catch and actual production meas

urements seems largest in oligotrophic lakes (Fig. 2). The dif

ference between fish production and sustainable yield should

reflect either the fraction of fish community production that is

unexploited (i.e. rough fish, juveniles) or the amount of fish

production that cannot be channeled into fishing mortality.

Rough fish are generally more abundant in eutrophic lakes,

therefore it is unlikely that the difference between Oglesby's

(1977) yield measurements and our production data is only the

production of unfished populations. If Oglesby's data are an

accurate reflection of real maximum sustainable yields of

exploited populations, then it appears that sources offish mor

tality are different in oligotrophic and eutrophic lakes. Because

nearly all fish production can be harvested on a sustainable

basis when primary production is high, it appears that natural

mortality can be substantially replaced by fishing mortality

through exploitation in eutrophic lakes. This would occur if

most mortality of biomass falls upon harvestable adult fish and

reduction of adult density assuages natural mortality. In oli

gotrophic systems, however, a smaller proportion of the fish

production can be harvested on a sustainable basis, and it

appears that nearly all fish production is required to balance

natural mortality. Natural mortality therefore must fall on a part

of the population that is not harvested, and for which decreased

adult density provides little increase in growth potential or

decreased density-dependent mortality. The data therefore sug

gest that fish production in eutrophic lakes is highly dependent

on adult density, while in oligotrophic lakes a much greater

fraction of the annual community energy budget goes toward

reproduction and juvenile mortality.

Can. J. Fish. Aquat. Sci., Vol. 47, 1990

3.0

v 2.5 -

D.C

(kg-rioNPRODUC"FISHoo

2.0 -

1.5 -

1.0 -

0.5 -

-J o.o0.5 1.0 1.5 2.0 2.5 3.0

LOG FISH STANDING BIOMASS (kg-ha"1)

Fig. 3. The relationship between mean annual fish standing stock and

fish community production in lakes. The solid line represents Eq. (2).

Dashed line indicate approximate 95% confidence bands for predicted

values of the mean log1oFP (Prepas 1984).

This interpretation is speculative, however, because the

degree to which Oglesby's catch data reflect actual sustainable

yields is unknown. For example, ignoring the four lowest and

four highest points in Oglesby's data (Fig. 2), representing the

Great Lakes fisheries and Indian and Israeli pisciculture ponds,

renders the two data-sets approximately parallel, offset by about

one log unit, suggesting that fishing yields are about 10% of

net fish community production.

Standing Biomass

Fish production in lakes was found to covary with fish com

munity standing biomass (FB; annual mean kg-ha"1) as:

(2) log10F/> = -0.42 + 1.084 logloFB

(^ = 0.67; w = 23;/><0.001). These two variables are closelycorrelated (Fig. 3) because, in general, production is propor

tional to the product of growth and standing biomass (Plante

and Downing 1989). Fish production in the arctic Char Lake

falls far beneath the general trend, probably because the aver

age annual air temperature near Char Lake is - 16°C. The slope

for log10Ffi close to 1 in Eq. (2) suggests that PIB does not vary

significantly with FB. The annual production to biomass ratio

{FPIFB) ranged from 0.02 in Char Lake to 2.73 in the equa

torial Lake Nakuru (average FPIFB = 0.76; n = 23; 5£=0.15;

median = 0.52).

Total Phosphorus

Fish community production was also correlated with the total

phosphorus concentration (TP\ jxg-L~l) of the water column:

(3) log,0F/> = 0.332 + 0.531 \o%iQTP

(t2 = 0.67;n= 14;^ = 0.002). The relationship between log, qand \ogl0TP appears to be non-linear when the very rich Lake

Nakuru is included in the data (Fig. 4). The non-linear relation

is approximated by:

1933

,*-N O.U -1

_' 2.5 -i

Df

CT 2.0 -

i—

o

gi.o-

oa:

°" 0.5-

^ 0.0-o

o1

-0.5-

/

/

■■■■■PRODUCTION DATAPREDICTED FISH PRODUCTIONHANSON AND LEGGETT (1982)

, . , ill

0 12 3

LOG ANNUAL MEAN TOTAL P

Fig. 4. The relationship between seasonal average total phosphorus

concentration of the water column and the fish community production

in lakes. The solid curve is predicted from Eq. (4) while the dotted

curve represents predicted yields from Hanson and Leggett's (1982)

Fig. 2. Dashed lines indicate approximate 95% confidence bands for

predicted values of the mean loglo FP (Prepas 1984).

Osi

o

Q

OorQ_

oo

■■■■■ PRODUCTION DATA (THIS PAPER)PREDICTED FROM RYDER (1965)

0.2 0.4 0.6 0.8 1.0

LOG MORPHOEDAPHIC INDEX

1.2

Fig. 5. Relationship between fish community production and the mor-

phoedaphic index of Ryder (1965). Not shown are data for Lake Nak-

uru. Where measurements of TDS were not available, they were

inferred from conductivity using Rodhe's (1949) table 2. TDS =

- 1.85 + 0.81 C + 0.0002 C2 (R2 = 1.00) where C is the conductivityin (iS-cm"1 and TDS is the summed concentration (mg-L~') of allmajor ions. The broken line is Ryder's (1965; Schneider and Haedrich

1989) equation.

(4) log]0FP = -0.319 + 1.441 logI077> - 0.209 (logWTP)2

(^ = 0.79; n=14; ;?<0.001), but this relationship is greatlyinfluenced by the extreme TP concentration in Lake Nakuru.

The close correlation between TP and fish production probably

results from the close correlation of TP with most other com

ponents of lake productivity (Peters 1986) and echos the close

correlation found between long-term fish catch and TP (Hanson

and Leggett 1982). Hanson and Leggett's (1982) model

severely underestimates fish production (Fig. 4), however, but

again it is not known whether this is because these catch data

do not accurately represent sustainable yields or whether sus

tainable yields actually fall an order of magnitude beneath bio

logical fish production (Fig. 4). Equations (3) and (4) may be

of greater practical use in predicting fish production in lakes

than Eq. (1) or (2) because TP is much less expensive to mea

sure than either PP or FB.

Morphoedaphic Index

We found weak correlations between FP and chlorophyll a

concentrations and watershed area, but no other variables or

combinations of variables that we tried yielded significant cor

relations (Table 3). This was especially surprising for the mor

phoedaphic index, which is reputed to be the best predictor of

fish production in lakes. There was no significant (p>0.05)

correlation between morphoedaphic index and FP (Fig. 5). This

was even true when TDS and 2 were entered separately into amultiple regression (see Schneider and Haedrich 1989), either

alone or in combination with other likely variables. Predictions

from Ryder's morphoedaphic equation (Ryder 1965; Schneider

and Haedrich 1989) were significantly (p<0.01) beneath, and

uncorrelated (p>0.05) with actual fish production measure

ments (Fig. 5). Ryder's (1965) analysis was based on a more

geographically restricted range oflakes than ours. Ryder (1965)

has suggested that MEI might not be correlated with fish pro

duction if lakes are not all in the north temperate zone, if lakes

at high altitude (>600 m) are considered, or if lakes are not

all large (>260 ha). Excluding all tropical and high altitude

lakes from our data yields an even lower correlation between

log,0MEI and log]0F? (^ = 0.05; n= 14; p = 0.53). Our data

do not, however, cover as large a range of MEI as did Ryder's,

because we lack data on full community fish production in very

deep lakes. All of our MEI values are >1. When Ryder's data

on "moderately to intensively fished" lakes with MEI >1 (all

but very deep or very dilute lakes) are analyzed, no significant

relationship between fish yield and MEI remains (n=13;

^ = 0.02; /? = 0.82). Although Ryder's (1965) equation provoked a highly influential sequence of research (e.g. Oglesby

1982), it does not make valid predictions of fish community

production within this range of basin morphometry. The anal

ysis presented here shows that fish production is more closely

correlated with primary production, phosphorus concentration,

and fish standing stock in oligotrophic to hyper-eutrophic lakes

of moderate depth.

Acknowledgments

This research was supported by an operating grant to J.A. Downing

from the Natural Science and Engineering Research Council of Canada

(NSERC), a team grant from the Ministry of Education of the Province

of Quebec, a NSERC fellowship to Mario Henri who performed some

of the initial search for data, and a CAFIR grant from the University

de Montr6al. We are grateful to R. T. Oglesby and two anonymous

referees for suggestions on the manuscript.

1934 Can. J. Fish. Aquat. Sci., Vol. 47, 1990

References

Arvola, L. 1983. Primary production and phytoplankton in two small, poly-

humic forest lakes in southern Finland. Hydrobiologia 101: 105-110.

Chadwick, E. M. P. 1976. Ecological fish production in a small Precambrian

shield lake. Env. Biol. Fish. I: 13-60.

Chapman, D. W. 1971. Production, p. 199-214. In W. E. Ricker [ed.] Meth

ods for assessment of fish production in fresh waters. IBP Handbook 3,

Blackwell Scientific Publications, Oxford, UK.

Clarke, G. L. 1946. Dynamics of production in a marine area. Ecol. Monogr.

16: 321-335.

Draper, N. R., and H. Smith. 1981. Applied regression analysis. 2nd ed.

John Wiley and Sons, New York, NY. 709 p.

Edmondson, W. T. 1977. Recovery of Lake Washington from eutrophication,

p. 102-109. In J. Cairns Jr., K. L. Dickson, and E. E. Herricks [ed.]

Recovery and restoration of damaged ecosystems. University Press Vir

ginia, VA.

Efford, 1. E. 1972. An interim review of the Marion Lake project, p. 89-109.

In Z. Kajak and A. Hillbricht-Ilkowska [ed.] Productivity problems of

freshwaters. IBP-UNESCO Symposium on productivity problems of

freshwaters. Warszawa.

Eggers, D. F., N. W. Bartoo, N. A. Rickard, R. E. Nelson, R. C. Wissmar,

R. L. Burgner, and A. H. Devol. 1978. The Lake Washington ecosys

tem: the perspective from the fish community production and forage base.

J. Fish. Res. Board Can. 35: 1553-1571.

Godbout, L., and R. H. Peters. 1988. Potential determinants of stable catch

in brook trout (Salvelinusfontinalis) sport fishery in Quebec. Can. J. Fish.

Aquat. Sci. 45: 1771-1778.

Gujarati, D. 1978. Basic econometrics. McGraw-Hill Book Co., New York,

NY. 462 p.

Gulin, V. V., and G. P. Rudenko. 1973. Procedure for assessment of fish

production in lakes. J. Ichtyol. 13: 813-823.

Hanson, J. M., and W. C. Leggett. 1982. Empirical prediction offish bio-

mass and yield. Can. J. Fish. Aquat. Sci. 39: 257-263.

Hayes, F. R., and E. H. Anthony. 1964. Productive capacity of North Amer

ican lakes as related to the quantity and the trophic level of fish, the lake

dimensions, and the water chemistry. Trans. Am. Fish. Soc. 93: 53-57.

HOLCfK, J. 1970. Standing crap, abundance, production and some ecological

aspects of fish populations in some inland waters of Cuba. Vestn. Cesk.

Spol. Zool. A 33: 184-201.

Hrbacek, J. 1969. Relations between some environmental parameters and the

fish yield as a basis for a predictive model. Int. Ver. Theor. Angew.

Limnol. Verh. 17: 1069-1081.

Jenkins, R. M. 1982. The morphoedaphic index and reservoir fish production.

Trans. Am. Fish. Soc. Ill: 133-140.

Jones, J. R., and M. V. Hoyer. 1982. Sportfish harvest predicted by summer

chlorophyll-a concentration in midwestern lakes and reservoirs. Trans.

Am. Fish. Soc. Ill: 176-179.

Kelso, J. R. M. 1985. Standing stock and production of fish in a cascading

lake system on the Canadian Shield. Can. J. Fish. Aquat. Sci. 42: 1315-

1320.

1988. Fish community structure, biomass, and production in the

Turkey Lakes Watershed, Ontario. Can. J. Fish. Aquat. Sci. 45 (Suppl. 1):

115-120.

Kirk, R. E. 1982. Experimental design: procedures for the behavioral sciences.

2nd ed. Brooks/Cole, Belmont, CA.

Kloster, A. E. 1978. Physical and chemical properties of the waters of 0vre

Heimdalsvatn. Holarct. Ecol. 1: 117-123.

Krogius, F. V., E. M. Krokhin, and V. V. Menshutkin. 1972. The model

ling of the ecosystem Lake Dalnee on an electronic computer, p. 149-

164. In Z. Kajak and A. Hillbricht-Ilkowska [ed.] Productivity problems

of freshwaters. IBP-UNESCO Symposium on productivity problems of

freshwaters. Warszawa.

Lam, D. C. L., A. G. Bobba, D. S. Jeffries, and J. R. M. Kelso. 1986.

Relationships of spatial gradients of primary production, buffering capac

ity, and hydrology in Turkey Lakes Watershed, p. 42-53. In B. G. Isom,

S. D. Dennis and J. M. Bates [ed.] Impact of acid rain and deposition on

aquatic biological systems. ASTM STP928, American Society for Testing

and Materials, Philadelphia, PA.

Lawrie, A. H., and J. F. Rahrer. 1972. Lake Superior: effects of exploitation

and introductions on the salmonid community. J. Fish. Res. Board Can.

29: 765-776.

Leach, J. H., L. M. Dickie, B. J. Shuter, U. Borgmann, J. Hyman, and

W. Lysack. 1987. A review of methods for prediction of potential fish

production with application to the Great Lakes and Lake Winnipeg. Can.

J. Fish. Aquat. Sci. 44 (Suppl. 2): 471^*85.

Liang, Y., J. M. Melack, and J. Wang. 1981. Primary production and fish

yields in Chinese ponds and lakes. Trans. Am. Fish. Soc. 110: 346-350.

Lien , L. 1978. The energy budget of the brown trout population of 0vre Heim

dalsvatn. Holarct. Ecol. 1: 279-300.

1981. Biology of the minnow Phoxinusphoxinus and its interactions

with brown trout Salmo trutia in 0vre Heimdalsvatn, Norway. Holarct.

Ecol. 4: 191-200.

Matuszek, J. E. 1978. Empirical predictions of fish yields of large North

American lakes. Trans. Am. Fish. Soc. 107: 385-394.

McConnell, W. J., S. Lewis, and J. E. Olson. 1977. Gross photosynthesis

as an estimator of potential fish production. Trans. Am. Fish. Soc.

106: 417-423.

Melack, J. M. 1976. Primary productivity and fish yields in tropical lakes.

Trans. Am. Fish. Soc. 105: 575-580.

Nyberg, P. 1979. Production and food consumption of perch, Percafluviatilis

L., in two Swedish forest lakes. Inst. Freshwater Res. Drottningholm.

58: 140-157.

Oglesby, R. T. 1977. Relationship offish yield to lake phytoplankton standing

crop, production and morphoedaphic factors. J. Fish. Res. Board Can.

34: 2271-2279.

1982. The MEI symposium — overview and observations. Trans.

Am. Fish. Soc. Ill: 171-175.

Oglesby. R. T., J. H. Leach, and J. Forney. 1987. Potential Stizostedion

yield as a function of chlorophyll concentration with special reference to

Lake Erie. Can. J. Fish. Aquat. Sci. 44 (Suppl. 2): 166-170.

Peters, R. H. 1986. The role of prediction in limnology. Limnol. Oceanogr.31: 1143-1159.

Plante, C. 1987. Prediction de la production secondaire en milieu aquatique.

M.Sc. thesis, Universit6 de Montreal. 129 p.

Plante, C, and J. A. Downing. 1989. Production of freshwater invertebrate

populations in lakes. Can. J. Fish. Aquat. Sci. 46: 1489-1498.

Prepas, E. E. 1983. Total dissolved solids as a predictor of lake biomass andproductivity. Can. J. Fish. Aquat. Sci. 40: 92-95.

1984. Some statistical methods for the design of experiments and

analysis of samples, p. 266-335. In J. A. Downing and F. H. Rigler [ed.]

A manual on methods for the assessment of secondary productivity in fresh

waters. IBP Handbook 17. Blackwell Scientific Publications, Oxford, UK.

Ramberg, L. 1976. Relations between phytoplankton and environment in two

Swedish forest lakes. Klotenprojektet Rapp. 7. Scr. Limnol. Upsaliensia426. 97 p.

Rask, M., and L. Arvola. 1985. The biomass and production of pike, perch

and whitefish in two small lakes in southern Finland. Ann. Zool. Fenn.22: 129-136.

Rawson, D. S. 1952. Mean depth and the fish production of large lakes. Ecology 33: 513-521.

1955. Morphometry as a dominant factor in the productivity of large

lakes. Int. Ver. Theor. Angew. Limnol. Verh. 12: 164-175.

Rigler, F. H. 1972. The Char Lake Project. A study of energy in a high arctic

lake, p. 287-300. In Z. Kajak and A. Hillbricht-Ilkowska [ed.] Produc

tivity problems of freshwaters, Warszawa.

1975. The Char Lake project, p. 171-198. In T. W. M. Cameron

and L. W. Billingsley [ed.] Energy flow — its biological dimensions. A

summary of the IBP in Canada, 1964-1974. Royal Society of Canada,Ottawa, Ont.

Rigler, F. H., and J. A. Downing. 1984. The calculation of secondary pro

ductivity, p. 19-58. In J. A. Downing and F. H. Rigler [ed.] A manual

on methods for the assessment of secondary productivity in fresh waters.

IBP Handbook 17. Blackwell Scientific Publications, Oxford, UK.

Rodhe, W. 1949. The ionic composition of lake waters. Int. Ver. Theor.

Angew. Limnol. Verh. 10: 377-386.

Rounsefell, G. A. 1946. Fish production in lakes as a guide for estimating

production in proposed reservoirs. Copeia 1: 29-40.

Ryder, R. A. 1965. A method for estimating the potential fish production of

north-temperate lakes. Trans. Am. Fish. Soc. 94: 214-218.

1982. The morphoedaphic index — use, abuse, and fundamental

concepts. Trans. Am. Fish. Soc. Ill: 154-164.

Ryder, R. A., S. R. Kerr, K. H. Loftus, and H. A. Regier. 1974. The

morphoedaphic index, a fish yield estimator — review and evaluation. J.

Fish. Res. Board Can. 31: 663-688.

Schaefer, M. B. 1968. Methods of estimating effects of fishing on fish pop

ulations. Trans. Am. Fish. Soc. 97: 231-241.

Schlesinger, D. A., and A. M. McCombie. 1983. An evaluation of climatic,

morphoedaphic, and effort data as predictors of yields from Ontario sport

fisheries. Ontario Fish. Tech. Rep. 10, Ont. Ministry Nat. Res., Toronto,Ont. 14 p.

Can. J. Fish. Aquat. Sci., Vol. 47, 1990 1935

Schlesvnoer, D. A., and H. A. Regier. 1982. Climatic and morphoedaphic

indices offish yields from natural lakes. Trans. Am. Fish. Soc. Ill: 141-

150.

Schneider, D. C, and R. L. Haedrich. 1989. Prediction limits of allometricequations: a reanalysis of Ryder's morphoedaphic index. Can. J. Fish.

Aquat. Sci. 46: 503-508.

Sokal, R. R., and F. I. Rohlf. 1981. Biometry — the principles and practice

of statistics in biological research. 2nd ed. W. H. Freeman and Co., San

Francisco, CA. 859 p.

StraSkraba, M. 1980. The effects of physical variables on freshwater pro

duction: analysts based on models, p. 13-85. In E. D. LeCren and R. H.

Lowe-McConnell [ed.] The functioning of freshwater ecosystems. Inter.

Biol. Prog. 22, Cambridge University Press, Cambridge, UK.

Tangen, K., and P. Brettum. 1978. Phytoplankton and pelagic primary pro

ductivity in 0vre Heimdalsvatn. Holarct. Ecol. 1: 128-147.

Vareschi, E., and J. Jacobs. 1984. The ecology of Lake Nakuru (Kenya).

V. Production and consumption of consumer organisms. Oecologia

(Berlin) 61: 83-98.

1985. The ecology of Lake Nakuru (Kenya). VI. Synopsis of pro

duction and energy flow. Oecologia (Berlin) 65: 412-424.

Wells, L., and A. L. McLain. 1972. Lake Michigan: effects of exploitation,

introductions and eutrophication on the salmonid community. J. Fish. Res.

Board Can. 29: 889-898.

Wernstedt, F. L. 1972. World climatic data. Climatic Data Press, Lemont,

PA.

WlNBERG, G. G., V. A. BABITSKY, S. I. GRAVILOV, G. V. GLADKY, I. S.

Zakharenkov, R. Z. Zovalevskaya, T. M. Mikheeva, P. S.

Nevyadomskaya, A. P. Ostapenya, P. G. Petrovich, J. S. Potaenko,

and O. F. Yakushko. 1972. Biological productivity of different types of

lakes, p. 383-404. In Z. Kajak and A. Hillbricht-IIkowska [ed.] Produc

tivity problems of freshwaters. IBP-UNESCO Symposium on productivity

problems of freshwaters. Warszawa.

Youngs, W. D., and D. G. Heimbuch. 1982. Another consideration of the

morphoedaphic index. Trans. Am. Fish. Soc. Ill: 151-153.

1936 Can. J. Fish. Aquat. Sci., Vol. 47, 1990