THE STUDY OF COMMUNITIES
164
productivity, diversity, etc . These fea­
tures of communities will u sually in­
crease through the succession to the
climax, but there is no reason why , in a
given case , the usual direction should
not be reversed. In general, climax
sta tu s should be determined not by
a bstract or generalized conceptions of
what should be ultimate, but by
what populations actually replace other
populations and then maintain them­
selves . .. .
Kormondy, E.J. 1965. Readings in Ecology.
Prentice-Hall, Inc. NJ. 219 pp.
THE CONCEPT OF mE ECOSYSTEM
One of the ends to which science directs itself is the development of
an encompassing theory . In ecology, the principle of the ecosystem
provides such a unifying framework within which specialized study at
the individual, population and community level con be meaningfully
conducted.
The concept of the ecosystem as an ecological unit comprtstng
living and nonliving components interacting to produce a stable system
is not new. Those 1'.'110 have read more or less consecutively in this
anthology have recognized that many ecologists discussed the inter­
relations of a particular biotic assemblage with its environment and
the interrelations of biotic units within the assemblage. However,
the theoretical development of the concept and the implementing of
studies oriented to its clarification are largely products of the period
since 1940, the major impetus occurring in the 1950's.
165
 FRA NCIS C . IplANS 167
ECOSYSrEM AS THE BASI IN ECOLOGY
regarded as an entity in its o wn right, often difficult to det erm ine the limits of
Francis C. E vans- 1956 interacting with its enviro nme nt (which a given ecosys tem. Th is has led some
may include other o rga nisms a s well a s ecologi st s to reject the ecosystem con­
Reprinted by permission of th e author and publisher from Science 123 : physical feature s of the habitat) to form cept a s unreali stic a nd of little use in
1127-1128, 1956. a system of low er rank th at lik ewise de scripti on or an aly sis. One is remind­
involve s the di stribution of matter and ed , howe ver , of th e fact that it is a lso
Although the term ecosystem was first proposed by Tansley ill 1935, the
energy. In turn , each individu al anima l difficult , if not impossible , to delimit a
contex t in which it is used today is much broader in scope. Systematic
o r plant, together with its particular species from its ances tra l or derivative
treatm ent and development of the concep t in its current sense is largely
micr oenvironment, con stitutes a system speci es o r from hoth ; yet this d oes not
associated with Eugene Odum beginning with his 1953 text Fundamentals of
o f still lo wer rank. Or we may wish to destro y th e value o f the concept. The
ecology. However, the f ollowing succinct statement by Professor E vans, of
take a world view o f life and look up on ecosy stem co ncept may indeed be more
the University of Michigan , provides an exce llent introduction to this section.
the biosphere with its total enviro nment useful when it is employed in rel ati on to
as a gigantic eco system . Regardless of the community than to the populati on
the level on which life is examined, the or individual , for its limits may be more
ecosystem concept ca n appropriately be ea sily determined on that level. Never­
The term eco sy stem wa s pr op osed by tati on, er osion a nd dep osition. The ap plied . The ecos ystem thu s sta nds as a theless, its a pplicatio n to all levels
Tansley as a name for th e interac tio n ecologist, then, is prim ar ily concerned b asic unit of ecol og y, a unit th at is as seems fully ju stified.
system comprising living things together with the qu antities of matter and energy imp ortant to thi s field of natural The con cept of the eco system has
with th eir nonliving habitat. T ansley re­ th at pass throu gh a given ecosy stem and science as the speci es is to tax on omy been described under many names,
garded the eco system as including "not with the rates at whi ch they do so . Of and systematics. In any given ca se, the among them th ose of micro cosm ,
o nly the orga nism-co mplex, but also the almost equaL importance, howe ver, are particul ar level on which th e ecos ystem naturkomplex , holocoen and biosyst em .
whole complex of physical fact or s form­ the kinds of orga nisms that are pr esent is being st ud ied can be spec ified with a Tan sley' s term seems mo st success­
ing what we call the en vir onmen t. " He in any particular ecosystem a nd the q ualifying adjective-for example , com­ full y to co nvey its meaning and has in
thus applied the term specifica lly to th at r oles that they occupy in its stru ct ure munity ecosystem , populati on ecos ys­ fact been accepted by a large number of
level of biological organization reprc­ and organization. Thu s, both qu anti­ tem , a nd so forth. present-day eco logists. I hope that it
sented by such units a s the community tative and qualitative aspects need .to be A II ran ks of eco systems are op en sys­ will eventually be ad opted universally
and th e biome. I here su ggest th at it is con sidered in the de scripti on and com­ tem s, not closed o nes . Energy and and that its application will be expand ed
logically appropriate a nd desirable to p ari son of ecosystems . matter continually escape from them in beyond its or igina l usc to include o ther
extend the application of the concept Eco systems are further characterized th e course of the processes of life, and level s of biological organization . R e­
a nd th e term to include orga nizatio n by a multiplicity of regul at ory mechan­ th ey must be replaced if th e system is to cognition of the eco system a s the basic
levels other than that of th e com­ isms , which, in limiting th e numbers of co ntinu e to function. The pathways of unit in ec ology would be helpful in
munity. organism s present a nd in influencing loss and replacement o f matter and focus sing a tte ntion upon the truly
In its fundam ental aspec ts, an ecosys­ their physiology a nd beh avi or , control energy frequ ently connect o ne ceo­ fund am ental aspects o f this rapidl y
tem inv olve s the circulati on , tr ansfor­ the quant ities a nd rates of movement o f system with another, a nd therefore it is devel opin g science .
mation, and accumula tio n of energy both matter a nd energy. Proce sses of
and matter through the medium of growth and reproduction, agencies of
living things and their acti viti es. Ph ot o­ m ortality (ph ysical as well as biological),
synthesis, dec omposition , herbivory , patterns of immigrati on and emigra­
pr ed ation, parasitism, a nd o the r sym­ tion, and habits of a daptive significance
biotic activit ies are am on g the principal are among th e m ore impo rtant gr oups
biol o gical pr oce sses resp on sible for the of regulatory mech anisms. In the ab­
tr an sp ort a nd storage of materi als and sence of such mechanisms, no ecosystem
energy , and the interacti ons of the could continue to persist a nd maintain
orga nisms engaged in th ese activities it s identity.
provide the pathways of distribution . The assemblag e of plants an d a nim als
The foo d-cha in is an example of such a visualized by T an sley as an integ ra
pathway . In the nonliving part of the part of the ec osystem usu ally co nsists
ecosyst em, circulation of energy a nd of numerous species, each represented
matter is completed by such physical by a population of individual o rgan­
proc esses as evaporation a nd precipi­ isms . H owe ver , ea ch p opulation can be
166
 STEPHEN A. FORBES 169
THI~ LAKE AS A f\.fl CRO COSM
productive rate , that is to say, in those is peculiarly unstable an d fluctuating ;
Stephen A . Forbes-1887 that each species swings, pendulum-like
which produce the largest average
number of eggs and young for each but irregularly, betwee n a highest and a
Re printed by pe rmission of the publi sher from Illinois Natural History
Survey Bulletin No . 15. pp . 537-550, 1925, a reprinting of an article which adult. lowest point, and tha t this fluctuation
first appeared ill the Bull etin ofthe Peoria Scientific Association, pp. 77-87, The first to feel this tremendous im­ affects the different ela sse s successively ,
1887. pulse are the protophytes and Protozoa, in the order of their dependence upon
upon which most of the Entornostraca each at her for food . .. .
In tire tradition of Mobius' biocoenosis (see p. 121), Forbes showed considera­ It would be quite irnpossihlc within
and certain minute insect larvse depend
ble insight in 10 the dynamic interrelationships between the living and nonliving reasonable limits, to go into details
for fo od . This sudden development of
comp onents of stable sys tem s. In som e few respects, his micro cosm is but respecting the organic relat ion s of the
th eir food resources cau ses. or course, a
semantically different from today's ecosystem. In addition 10 careful obser­ animals of these waters , a nd 1 will
corresponding increase in the numbers
vations, and perceptive insights, Forbe s was a gifted writer; part of the content my self with two or th ree
of the latter classes, and , through them ,
classic quality of this essay is its literary charm. The breadth of viewpoint illustrations. As one example of the
o f all sorts of fishe s. The first fishe s to
belies Forbes' narrow speciality as an applied entomologist with the Illinois varied and far-reaching relations into
feel the force o f this tidal wave of life
Natural History S ur vey . which the animals of a lake are brought
are the rapidly-breeding, non-preda­
ceous kinds; and the last , the ga me in the general struggle for life, I take the
fishes, which derive from the others comm on black bass. In the dietary of
their principal food supplies. Evideutly this fish I lind, at different ages of the
A lake .. . forms a little world within complicated mechanism of the aquatic each of these classes must act as a individual, fishes of great variety, rcpre­
itself-a m icrocosm with in which life of the locality, both animal and check upon the one pr eceding it. The senting all the important ord ers of that
all the elemental forces are at work vegetabl e, of which his species forms development of animalcules is ar­ class ; ins ects in considerable number,
and the play o f life goes on in full , but a sing le element. .. . rested and soon sent back below its especi ally th e vario us wa te r-b ugs and
but on so small a scale as to bring it The amount and variety of animal highest point by the consequent deve­ larva of day-Hies ; fresh- water sh rimps;
easily within the mental grasp. life contained in them as well as in the lopment of Entomostraea; the latter, and a great multitude of Entomostraca
Nowhere can on e Sec more clearly streams related to them is extremely again, are met, checked, and reduced of many species an d genera. The fish is
illustrated wh at may be called the variable, depending chiefly on. the in number by the innumerable shoals therefore d irectly dependent lipan all
sensibility of such an organic complex, frequency, extent, and duration o f the of fishes with which the water speedily the se classes for its existence. Next ,
expressed by the fact that whatever sprin g and summer overflows. This is, swarms. In this way a general adjust­ lo oking to the food of t he species w hic h
affects any species belonging to it , in fact, the characteristic and peculiar ment of numbers to the new conditions the ba ss ha s eaten, and upon w hich it is
must have its influence of so me sort feature of life in these waters. There is would finally be reached sp onta neo us ly; therefore ind irectl y depend ent , 1 flnd
upon the whole as semblage. He will perhaps no better illustration of the but long before any such settled balance that one kind of t he fis hes taken feeds
thus be m ade to see the impossibility of methods hy which the flexible system o f can be established, often o f course up on mud, algre, a nd Ento mo straca ,
studying completely any form out of organic life adapts itself, without in­ before the full effect o f this upward and another up on nearly every animal
relation to the other forms; the neces­ jury, to widely and rapidly fluctuating influence has been exhibited , a new su bsta nce in the water, including mol­
sity for tak ing a co m p rehensive survey conditions. Whenever the waters of the cause o f disturbance intervenes in the lu sks and decom posin g organic matter.
of th e who le as a condition to a satis­ river remain for a long time far beyond T he inse cts ta ken b y the bass, them­
disappearance of the overflow. As th e
facto ry un d er sta nding of any part. If their banks, the breeding gr ounds of waters retire, the la kcs are again de­ selves take oth er in sects and sm all
one wishes to becom e acquainted with fishe s and other animals are immensely fined; the teeming life which they con­ Crustacea. T he cra wfishes a re nea rly
th e black ba ss, for example, he will extended. and their food supplies in­ omnivorous, a nd of the other cru sta ­
tain is restricted within daily narrower
learn but little if he limits himself to crea sed t o a corre sponding degree. The bounds, and a fearful slaughter follow s; ceans some cat Ent o rnos traca and som e
that species. He must evidently study slow or stagnant hack waters of such an the lower and more defenceless animals algre and P rotoza . At only the seco nd
also the species upon which it depends overflow afford the best situation s a rc penned up more and more closely step, therefo re, we find our hass brought
for its existence, and the various con­ possible for the de velopment of myriads in to dep end ence up on nearl y ever y cla ss
with their predaceous enemies, and
ditions upon which th ese depend . He of Entomostraca, which furni sh, in these thrive for a time to a n extraordi­ of an ima ls in th e wate r.
mu st likewise study th e species with turn, abundant food for young fishes of An d now, if we sea rc h for its com­
nary degree. To trace the further
which it comes in competition, and the all de scriptions. There thus results an p etito rs we Shall find these al so e x­
conseq uences of this oscill a tio n would
entire system of conditions affecting outpouring o f life-an extraordinary trcmcly numer o us . J n the first place, I
take me too far. En ough ha s b een said
their prosperity; and by the time he has multiplication of nearly every species, h ave fo un d t hat all ou r you ng fishes
to illustrate the general idea that the
st ud ied all these sufficiently he will find most prompt and rapid, generally speak­ except the Catostornida: feed a t first
life of waters subject to periodical
that he has run through t he wh ole ing, in such a s have the highest re- almost wholly on E nto m o stru ca , so
expansions of considerable duration,
168
 170
th at th e littl e bass finds himself at th e
very beginn ing of his life engage d in a
scram ble for food with a ll th e o ther
little fishe s in th e lak e. In fact , not on ly
yo ung fishe s but a multitude of o the r
anima ls as well , especially insects and
the lar ger Crustacea, feed up o n th ese
Entomostra ca, so th at th e competit o rs
of th e b ass are not co nfined to member s
of its ow n class. Even mollu sks, while
th ey do not directl y co mpe te with it do
so ind irectly, for th ey ap pr opria te
myriads of th e microscop ic form s up on
which the Ento rnos traca lar gely depen d
for food . But th e ene mies of th e bass d o
n ot all attack it by ap pro pria ting its
food supplies, for man y de vour t he litt le
fish itself. A great variety of predaceou s
fishes, turtles, wat er -snak es, wad ing a nd
diving birds , a nd even bugs of giga ntic
dim ens ions destroy it o n the slightes t
op po rt u nity. It is in fac t hardl y too
mu ch to say th at fishes whi ch reach
maturity are relati vely as rare as ce n­
ten ari ans am on g human kind .
As an illustrati on of the rem ote and
unsu spected rivalries which reve al th em ­
selves on a careful study of such a
situa tion, we may ta ke the relation s of
fish es to the blad derwort I- a flo wering
pl ant wh ich fills many acres of the water
in the shall ow Ja kes o f n orthe rn Illi nois.
Upon the leaves of th is spe cies are
found little bladders- several hundred
to eac h plant- which when cl ose ly
exa mi ned are seen to be ti ny tr ap s fo r
th e ca pture of Entomostrac a and ot he r
minute a nima ls. T he plant usu all y has
no roots, bu t lives entire ly up on th e
ani ma l food obtai ned th rou gh th ese
littl e bladder s. Ten of th ese sacs wh ich I
took at random from a m ature plant
co nt aine d n o less th an ninety-three
a nima ls (mo re th an nine to a bladder).
belon gin g to twenty-ei ght di fferen t
speci es. Seventy-six of th ese wen: Ento­
mostraca , an d eight other s were minute
insect larvse. When we estima te the
myri ad s of sma ll insec ts and C ru stacea
'Utricularia.
THE CONCEPT OF THE ECOSYSTEM
wh ich th ese pl ants must app rop riate
during a yea r to t heir ow n sup po r t, and
co ns ide r th e fac t th at th ese are of th e
kind s most useful as food for yo ung
fishes of nearl y all descripti on s, we
mu st co nclude that t he bladd erworts
co mp ete with fishes fo r foo d , and tend
to keep d own their num ber by dim ini­
shing the food resou rces of the yo ung.
Th e plan ts eve n have a certai n ad­
vant age in this co mpe titio n, since th ey
are n ot strict ly dependent o n Ento mo­
straca, as the fishes are. but so me ti mes
tak e root , develop ing then but very few
leave s and bladders. Th is pr ob abl y
happen s und er co nditions un favorabl e
to th eir suppo rt by th e o the r meth od .
Th ese simple instances will suffice to
illustra te the in tim at e way in whic h the
living forms of a lak e are united .
Perhap s no ph enom en on of life in
such a sit ua tion is more rem arkabl e
th an th c steady hal an ce of organic
nature, which hold s eac h species
within the limit s of a unifo rm average
number, yea r afte r year, althou gh .eac h
o ne is alw ays doing its best to break
ac ro ss bound ari es on every side . The
reproduct ive ra te is usually eno rmo us
and th e str uggle fo r existenc e is cor­
resp ondingly severe . . ..
It is a self-evident proposit ion th at a
speci es ca n no t mainta in itself co n­
tinuo usly. yea r after year , unl ess its
birth-rate a t least eq ua ls it s death-rat e.
If it is pr eyed up on by a no the r species .
it mu st pr od uce regularly a n excess of
ind ividu als for destruction , or else it
must certainly d windle a nd disappear.
O n the ot he r hand . th e dependent spe ­
cie.. evidently mu st not appropr iate, o n
an a verage, a ny m ore than the surplus
and excess of individuals up on whieh it
pr eys, fo r if it does so it will co nt inu­
ou sly dim inish its own food supply, and
thus indirectly but surely exterminat e
itself. The in teres ts of both p art ies will
th erefor e be best served by a n adj ust­
ment of their respective rat es of
multiplication such that t he species
EDG AR NELSON TR ANSEAU 171
devou red sha ll furnis h an excess of wh o lives within his income will fina lly
number s to supp ly th e wants of the d isp ossess his shiftless co mpetito r wh o
devou rer , a nd that the latter sha ll ca n never pay his debts, th e well­
confine its appro priat ion s to th e excess adjus ted aq ua t ic anima l will in tim e
t hus furnished . We thu s see that th ere is crowd o ut its poorly-adj uste d co mpe ti­
rea lly a close community of interes t tors for foo d an d for th e va rio us goo ds
bet ween these two see mingly de adl y of life. Con sequ entl y we may believe
foes . th at in th e lon g run a nd as a general
And next we not e th at thi s comm on rule th ose spec ies w hic h have survived,
interes t is pr om ot ed by t he process of a rc th ose which have reached a fairl y
na tural selec tio n ; for it is th e grea t close adj ustme nt in th is parti cula r.
office of thi s pr ocess to elimina te th e Tw o ideas ar e thu s seen to be suf­
un tit. If tw o species sta ndi ng t o eac h ficient to exp la in th e orde r evolved
o ther in th e relati on o f hunter an d fro m thi s see m ing chaos; the first th at
pr ey a re o r hecorn e badl y adju sted in of a genera l co mm unit y of int ere sts
resp ect to t heir rate s of increase, so a mo ng a ll th e classes of o rga nic bein gs
th a t th e o ne preyed up on is kept very far here assembl ed , a nd th e seco nd th at of
below the norm al number which might th e be ne fice n t po wer of natural selectio n
find food , even if th ey do not pr esently which co m pels such adj ustme nts of the
o bliterat e each o the r th e pair ar e pl aced ra tes of destructi on a nd of multip li­
at a disad vantage in the battle for life , ca tion of the va rio us species as sha ll
a nd must suffer acco rdi ngly. lu st as best p ro mo te t his co m mo n interest.
cer tai nly as the thrifty business man
rUE ACCUMULATION OF ENERGY BY PLANTS
Edgar Ne lson Transeau- 1926
Rep rinted by perm ission of the publisher from the Oh io Journa l of
Science 26 : 1-10, 1926.
If he is to study energy flo w in an ecosystem , the ecologist needs inform ati on
about the energy relations of fire only g roup of organisms that can transfo rm
radiant energy into a f orm useoble by other organisms. Tran seou's study
of tire energy budget of 0 cornfie ld was considerably in ad vance of its tim e
and p rovided the basis fo r what has subsequently becom e a highly active area
of research-primary production,
. .. Le t us no w exa mine the energy sta tio ns a t which so lar rad iati on has
bud get of a hyp oth etical ac re of co rn in been st udie d. T he gro wing seaso n is
th e hea rt of th e co rn belt in north fro m June I · to Septemb er 8, one
centra l lllin ois wh ere corn at ta ins hu nd red days. The best yields have been
yields a s grea t as a nyw here, a nd not far wit h 10,000 plan ts to the acre. O ne
from Ma dison. Wisco nsin, on e of th e hundred bu shels, with a dry weight of
172
2160 k g. p er acr e , is t he yie ld as sume d
a ltho ug h it is n ot a m a ximum crop for
th e corn belt. H ow we ll an acre of co rn
cover s th e area is s hown by t he fac t t ha t
durin g the la tt er hal f o f th e seaso n
ne arly two ac res of lea ves are exposed
to the light.
At m aturity th e average corn p la n t
co nta ins a bo u t 20 per cent of dry
matter a nd a bo u t 80 per ce nt of wat er.
Of the dry m atter, ca rbo n m a ke s up
about 44 .58 pe r cent. T his is th e m ost
im portant figure fo r o u r calculatio ns
d er ived fro m th e chemical a na lyses of
the corn plants . W e mu st al so know th e
total amou nts of mineral elem ents
presen t wh ich is 5.37 p er cen t fr om
whi ch we ca n derive th e fac t th at 94 .63
per cent o f the pla nt' s dry weight is
o rganic m atter. T he dry weight o f a n
av era ge corn pl ant grow ing und er th ese
circ um sta nce s is 600 g., of which 2 16 g.
m ak es up th e grain 200 g. the sta lk ,
140 g. th e lea ves, and 44 g. the ro ots.
The tot al weight of th e 10,000 pla n ts is
6000 kg. Subtracti ng fro m th is th e 322
kg . o f mineral elements in the as h we
have left 5678 kg. o f orga nic matter , of
wh ich 267 5 kg . is ca rbo n.
To est ima te the amo un t of phot os yn­
thesis we must d eterm in e the amo unt o f
car bo n, becau se ca r b on ente rs th e
pl ant only by pho tosyn th eti c r educti on
of CO 2 , T he to tal carbon is 2675
kilogram s and th e gluco se equiva len t o f
this car bo n is 6687 kilograms. This is
the amo un t of prim ary sugar eq uivalen t
to the carbon acc umu la ted in th e
m ature pl a nt.
T ABL£ 3. Glucose eq uivalent 0/ resp iration
Estima te d rare o f C0 3 relea se '- 1 %
of t ho dry wI pe r da y
A ver age d ry weig h t fo r se aso n <I tot al w r.) . • 3000 kg .
A vera ge ra te orc o. release (.0 1 x 3200)
per da y . . . . . . . . . . . . . . . . . . . . . . . )0
T ot al CO l rel ease during season . - ] (lOO
Ca rbon eq u iv a leu t : C ~CO ! - 12 :44 . - 8 18
G lucose eq uiv al en t C, Il " O" C, 180 '72 . . ~ 2045
At m aturi ty how ever , o nly a p art o f
th e car bo n rem ains, for some ha s be en
THE CON CEP T OF TH E ECOSYST EM
lost as CO ~ in re sp iration (Table 3).
The a ver age rate o f CO 2 loss is not far
from one per cent o f t he dry weig ht pe r
day. This would cau se a d a ily loss o f 30
kil og ram s of CO 2 , a nd during th e ent in
sea so n a loss o f 3000 kilog rams . T he
glucos e eq uivalen t to th is am ount o f
carbon dioxide is 2045 kilo gr ams.
Adding th e a m oun t o f th is lo st
glucose to th e glucose eq uivalent of the
ca r b on in the pl ant , gives the total
glu cose m anufactured as 8732 k ilo ­
gra m s. It requires ene rgy eq uiva len t to
3760 ca lo ries to produce o ne ki logra m
o f glucose. H ence it requir ed not far
fro m 33 mill ion ca lorie s to pr o duce the
enti re phot os ynthetic pr oduct. (Tab le 4).
T .AUlF. 4. En ergy cons um ed in pho tosynt hrs is
G lucose eq uiva len t or accu m umted carb on . 6687 kg.
G hJC o~ e qu ivalent of ca rbo n oxidiz ed . . . . . 2o.t 5 k-..
T OI" I glu cose ma nufa c tu red . . . _ . . . . . . . .. 87 J2 k ~ .
~rgy re qu ired 10 p r od uce I k g . gluco'e . . . 3760 C ~ 1.
'T o ta l c ne rrr co nsu m ed in ph o t osyn th esis 33 million Cal.
We are now in a position to estima te
the efficiency o f the co rn plant as a
p ho tosynt het ic agen t : (Tab le 5).
T A ni I: 5. Efficien cy 0/ photo s vnt hc sis
Total ener gy a vailabl e o n acre d u ri n g the
gro ""';n.-:; sea son . .. 204 3 m illio n Cal.
Tot a l CU t tar used in ph ot nayn th esis . , . . 33 n n lh n n COlI.
Pe r cent o f a \o'a tl::..hle cn cf lrlY used by th e co r n plan t
in phoLos)n lhcsi:iJ (c a. c ~nc y o r c orn pl.. m) . 1.6 %
O f the to ta l Jiuh l ap C(,lfU IIi me.u ureaJ. ho wever, o nly
ab out 20 ~.{ is w.ed in ph ()t D ly n th~ s iti he nce the
efficiency oft he photos ynthe tic process is . . . . . . 8 %
T he total energy ava ila b le accordi ng
to th e Smithsoni an figures is 2043
million Cal ories. Th e energy utilized is
33 milli on, or 1.6 pe r ce n t. In photo­
synthesis however, o nly cert a in rays a re
effective, a nd th ese f urn ish a bout 20 p er
ce nt o f th e energy measu red by th e
pyrheliometer. C on sequ entl y th e effi­
cie ncy o f ph ot osynth esis i n 100-bush el
kg . corn is 8 per ce nt.
k g. An other so u rce o f ene rgy loss to th e
k g.
kg . p la nt is tran sp iration . From the wa te r
requirem ent stud ies o f co rn it is pro­
bable th at in Ill in o is not far from 276
kil o gram s o f water a re evapo ra ted
!:'O GAR N£ l.-~ON TRANSEAU
during th e gro wi ng season for eve ry
kil ogram o f its dr y weight. Th e total
weigh t of water lo st in th is way th er efore
is o ne an d a hal f mill ion kil o gr am s.
This is equal to 40 8,000 gall on s o r
s ufficien t water to cov er th e ac re to a
d epth of fifteen inches .
The energy necessary to eva po ra te
o ne kil ogram o f water a t th e a verage
temperature of the growin g seaso n is
593 ca lo ries . Consequ entl y 910 milli on
C al ori es a rc exp end ed in thi s way. Thi s
is equ ivalent to 44 .5 p er ce n t of th e
availa b le energ y.
'I'.... . l.£ 7. Knergv released in respira tion
G lucose c on sumed in re sni nu ion . . . . . . . . . . . . 20 4 5 k g..
One ki lo gr a m n f gluc osl: rc lcu scs . . . . . . . . ... 37(JO C it!.
T o tal cnt r ~y rel ea sed in R esp irat ion
O f lhoa ene rgy mad e poten tia l in pho LPsy n Lhl:sl:i..
R es rtinl il o n releases . _ • ... . . _
A ~ ~ u l r. i n $:. th u t p h u [U s ) J; l l J c~j i 1:fJCS (In J2 hou rs each da y
and rcsph atio n 24 ho urs enc h d uy, We a\'~ r.Ll 8 C dall)' rat e
or ph o rosvnthest s i!\ ab o u t ~ Li m e s (h e ra te C., r n~ l p i r . li ( lI1 .
R espiration again releases a p art of
th e ene rgy rendered pot ent ial in ph oto­
synt hesis. A s we hav e seen 2045 kil o­
gra ms o f g lucose a re thu s oxi d ize d , a nd
in con sequ en ce 7.7 mill ion Cal o ries are
rel ea sed withi n the plant. This ene rgy
ra ises th e temper ature of the pl ant and
esca pes to the env iro nment, or it is used
in sy nth esis o f fats, pr oteins a nd o ther
reduced or ganic subs ta nces.
Th e energy rel ea sed in respirati on
a m oun ts to 23.4 per cent- alm ost o ne­
fourth o f the energy absorbed in
ph ot osyn th esis. This is far more th a n
is needed "fa acco unt for t he en d o­
thermic rea ct ions associa ted with fo od
tran sformati on s wit hin th e p la nt.
A ssuming th at photosynth esis goes
on 12 hours and resp iration 24 h ours
each day , th e a verage rate o f ph ot o­
syn thes is must be a bo ut 8 t im es th e
rate of respi ration.
173
W e m a y no w summarize the energy
budget [T abl e 8] :
T A nLE 8. Summary 0/ budget
T o ta l energy a va ila ble . 204 3 m illi on Cal.
Used in p h o to xy nt h es is .. . . . 33 m illi o n C a l.
U Sl:(l in tra nspira tio n . . . . 9 10 millio n CuI.
To L.:! 1 e nergy c on sumed 943 millio n Ca l.
E nergy no t direct ly used by the plant s . . 1100 million Ca l.
Energy rele a s ed by re sp irati o n . . . . . . 8 mill ion Cal.
O f' tb e a vaila b le ene rgy , 100-bushe l·co rn USe S abou t 46 ~{;
T he cnviro nrnen t takes up a bo rn . 54 %
A s a result we a re in a po siti on to
m ak e a number o f ge ne ra lizat io ns
regardin g the m et aboli sm and gro wt h
o f plants, bot h as t o materi als a nd
en er gy. (1) An a cre o f 100-bush el co rn
uses durin g the gro wing sea son ab out
40 8,000 ga llo ns of wa te r o r 15 acre­
inc he s. (2) T he eva porat io n o f t hi s
7.7 miJltnn Cal. wat er co ns um es ab out 45 p er cent o f
th e ava ila b le light ene rgy. (3) In ph ot o­
_ 2 J .4 ;~
synt hes is th e co rn plant utilizes a bo u t
I. G per cent of the ener gy av a ilable ; its
efficiency is a bou t 8 per cent. (4) An
ac re o f 100 -bu shel corn manufactures
on th e averag e 200 pounds of s uga r a
d ay. (5) Of the ene rgy rend er ed po­
tential in ph o to synthesi s, 23.4 per cent
is again released in respiration . (6) Of
t he s uga r m an ufa ctured nearly one­
fo ur t h is oxi d ized in respiration . (7)
R espi rati on rel eases severa l times as
mu ch energy as is ne eded to a ccount
fo r the redu ct ion s in the synthesis o f
fa ts , prote ins a nd o ther compound s. (8)
A t maturi ty th e grain conta ins about
o ne- fo urt h of the tot al energy utilized
in ph oto synt he sis, or abo ut .5 per cent
o f th e energy avai la ble . (9) The av era ge
rat e o f pho tosy nt hes is is about eight
tim es th e ra te of respiration . ( 10) Sin ce
th e yo ung co rn seed lin g weig hs .3 grams
an d the m ature pl ant weighs 600 gra m s,
o n th e ba sis of th e compound interest
la w o f gro wth the average dail y in ­
cre ment in dry weight is 7.9 per cent. . . .
 C H AN C EV JUDAV
175
AL ENERGY BUDGET OF AN INLAND LAKE

Chancey Juday- 1940 tion that passes int o the water of Lake one gram of animal tissue even in
Mendota is utilized by the aquatic animals tha t feed directly on plants; the
Reprinted by permi ssion of the publisher from Ecolog y 2 l : 438-450, 1940. plants in the process of photosynthesis. pr edaceous animals represent a still
The products of this assimilation. larger quantit y of the original photo­
Transeau 's energy budget calculations (see p. 17/) were limited to one namely, proteins, fats and carbohy­ synthesized material.
growing season of a single crop . Juday's study not only covers an entire y ear drates, thus con stitute the primary Plankton. . . In view of all of these
but considers the totality of energy relation s ofa system . The complex ity of accumul ation or storage of the energy complexities, it may be estimated that
study ing comm unity meta bolism is obvious in this paper which represented a derived from the sub-surfac e illumi­ the average turn over in the organic
major forw ard step in ecosystem analy sis. Its influence on Lindeman's nation. Since this organic material matter of tile mean stand ing crop of
research and formulation of his trophic-dynamics concep t is also apparent manufactured hy the plants serves, pla nkton takes place about every two
(sec p. 179). Juday, with E. H . Birge, pioneered limn ological studies in the either dir ectly or indirectly, as a source weeks throughout the year. It would be
United States and established the University of WiSCVIlSill as the center of of food for all of the non-ehlorophyl­ more frequent than this in spring and
such studies. laceous organisms that inha bit the lake, summer, and less frequent in winter. A
the se latter forms , therefore, con stitute turnover of 26 times per year would
a secondary stage in the storage of the give an annual yield of 6,240 kilograms
ener gy accumulated by the aquatic
T ,\ nf .E TV. Annnal product ton of plunkmn, bot tom fior a,
The variation in the quantity of solar tcrials is so small in most cases that plants. T he original amount of energy b o tto m j ;:w w an d /i sh, irs 'w(~11 a s crude prot ein , ether
radiation delivered to the sur face of an they may be neglected .. .. represented by th ese secondary or gan­ extr act (far ), and carbohy dra t e cons tit uc..nts 0/ the or ganic
matter
inland lake during the course of the . .. the monthly means of solar and isms varies wit h the different forms, Th e re su lts arc stale d in kilog ram s per hect are o n a
year is the principal factor in deter­ sky radiation delivered to the surface of depending upon the number of links in d ry. a sh-Tree busts. The plan kto n yie ld j::. base d on a
mining the ph ysical, chemical and Lake Mendota during the 28 year their respective food chains ; in general turn o ve r ev e-ry two weeks dur in g the year. Th e a ve r age
q ua n tit y or d iss olv ed o r ga nic matter is included also .
biological cycle of chan ges that take period ranged from a minimum of 3,568 they represent a comparatively small
place within the wa ter. Thi s is true calories in Decem bel' to a maximum of D ry E th er C ar bo­
proportion of the primary organic o rganic Crude ex­ h ydrate
especially of lakes which arc situated in 16,392 calories in July . The annual material manufactured by the plants. rua t­ prot ein tract
ter
temperate latitudes where t here arc means for the 28 years varied from a Chemical analy ses of the vari ous -
Total pla n k ro n 6240 2704 431 3105
con siderable differences hetwccn sum­ minimum of 108,597 calories in 1935 to aquatic organisms have now progressed Phyto-
mer and winter temperatures of the a maximum of 129,659 calories in p lankton _ 5850 25 0 1 383 2966
far eno ugh to enable one to compute Zo oplan kton 1 390 203 48 139
air and of the water. ... 1937.. .. the annual mean for the 28 their energy values from the standards B Oll OI1l flo ra. · 1 5 12 64 6 442
45 33 4 8
The annual energy budget of a lake year s is 118,872 calories. ... that have been established by food B ott om fau n a .
Fish . . . . . . . . 5 3.4 I o.
may be regarded as comprisin g the The respective amounts of energy in­ chemists. The standard values are 5,650 Di ssolved or­
gun ic ma tte r ) 52 3 334 68 J 121
ener gy received from sun and sky each clud ed in the four items of the phy sical calories per gram of protein , 9,450 cal o­
. ~l
T ot al o rgan ic
yea r and the expend it ures or uses which energy budget which have been consid­ ries per gram of fat and 4,100 calo ries m a t te r . . . . • 3138.4
I 510 I 467 6 .6
the lake makes of this annual incom e of ered so far are indicated in table III. per gram of carboh ydrate, on a dry
radiation. In general the annual income T ABL6 Ill. Ou orn ity of sola r and sky radiati on used by weight basis . These value s cia not rep­ of dry organic matter per hectare of
and out go substa ntially balance each L ake M endQIQ in ~a-rt QlIs p hysicu! an d bia log rcal p r ocrs.H · ,';
resent the total quantity of energy surface as indicated in table IV. This
other. This is true more particularly of T he resu lts are indi ca led in gram c alo ries per square
utili zed by the aquatic organisms. material would consist of 2,704 kilo­
ce n time ter o f surface.
the ph ysical energy budget. Considera­ Me ltin g o f ice in srrin.g 3, 500 however, hccause a part of the synthe­ grams of pr otein, 431 kilograms of fat
ble biological material produced in one Annual heat bu dge t o [ wa ter .. .. .. . 24 ,:200 sized material is o xid ized in the meta­ and 3,105 kilograms of carbohydrate.
An n ual beu t bud get o r bottom 2.000
energy year lives over into the next , but E n ~II!Y to ~ l by eva pora tion . .. . . . . .. . ... 29,300 bolic pr oce sses of the living organism s. Approximately 94 per cent of the or­
th is overla pping crop of organisms is Ann,,,,l su rface lo ss . . . . . .. . . ... . 28,500
These metabolic oxidation s result in the ganic matter comes from th e phyto­
L o ss by cond uctio n. co n vec tio n and radiation 30. 324
much the same in quantity from yea r to B iological en ergy budget ( ma ximum) . .. .. . 1,048 production of heat which is trans­ plankton and 6 per cent from the
year so that it plays approximately the mitted to the water, but the quantity of zooplank ton .
same annual role . For this reason it The sum of these four items is 7] ,000 Bottom flo ra. . . . the annual crop of
calories, which is approximately 60 per heat derived from this source is extreme­
docs not require an y special consider­ ly small in comparison with that which large aq uatics amounted to 2,000
cent of the mean quantity of energy de­
ation. Th ere is a certain am ount of comes from direct insolation. kilo grams per hectare, dry weight. If
organic material contributed to the livered to the surface of Lake M endota evenly distributed over the entire lake,
annually by sun and sky. ... The amount of organic matter con­
bottom deposits in the deeper water and sumed in the metabolism of plants is this crop would give a yield of 628
to peat formati on in the shall ow water much smaller than that in an imal s kilograms per hectare. A little more
BIOLOG ICAL ENERGY BUDGET
which lasts for long periods of time, but because several gram s of plant material than 18 per cent of this material consist­
the annual energy value of these ma­ A certain amount of the solar radia- may be consumed in the production of ed of ash, so that the organi c matter
174
 17
wa s equivalent to 512 kil ograms per
hectare. .. .
. . . the en ergy va lues .. . a re given in
table IV . The 512 kil ogram s o f dry o r­
ganic matter per hectare consisted of 64
kil ogram s o f prot ein , 6 kilograms o f
e t he r ext ra ct or fat and 442 kilograms o f
carbohydrate.
The bottom dep osits, especia lly in the
de eper wat er , cont ain a rather large
popul ation o f bacteri a . . . . While th ese
o rga n isms a re present in considerable
numbers, they are so sm a ll in size that
th e y add ve ry little to the cr op of
organic m atter in th e lake; so they ha ve
been d isregard ed. Likewi se fungi a re
fairly a bu nda nt in th e bott om d ep osit s,
but no qu antitative st ud y o f them ha s
ye t been m ad e ; it seems proh ahle that
their contribution to the o rga nic co n­
tent o f the lake is negligible from a n
energy sta nd p o int.
Bot/ om fauna . . . The m acroscopic
bott om faun a yield ed 45 kil ograms o f
dry organic matter per bect are ; o f this
amount 33 kil ogr am s con sisted o f
prot ein , 4 kil ogram s o f ether extract o r
fat a nd 8 kil ogram s of carbohydrat e.
While some of th ese organisms live
more th an one yea r, others p ass
th rough tw o or three gen er ations in a
yea r ; the tw o gr oups of o rganisms are
gen er all y considered as bal ancing eac h
o t he r, so th at the a bo ve qu antities m ay
be taken as th e annual crop of thi s m ate­
rial a s sho wn in tabl e IV .
C onsider able numbers o f prot ozo a
and ot he r micro scopic a nima ls ha ve
been found in the bott om dep o sit s, but
n o quantitative study of them has been
mad e. It seem s probable, h ow ever, that
these minute form s would not add a n
appreciable a m o un t o f organic m atter
to the total weight o f the bottom fa un a.
Fish . N o accurate cen su s of the fish
caught by a ng le rs in L ake Mend ota ea ch
year ha s ever been made so that the
a ssessment o f thi s p art of the bi ol ogic al
crop can be estimated o nly rou ghl y. .. .
adding the carp cr op to that of the game
T HE CONCEPT OF THE ECOSYSTEM
a nd pan fish gives a n annu al fish yield
of 22 kil ogr ams per hectare, live weight.
O n a dry, as hfree basis, the tot al yield
am ounts to a little m ore th an 5 kilo­
gra ms per hectare a s indicated in table
IV. By far th e g rea te r p art o f this
m a terial consists of protein a nd fat.
inergy value oj annual crop. T able
IV s ho ws that the total quantit y of
s to red and accumulated ener gy in the
form of dry o rga n ic matter in the a nn ua l
cro p of pl ants and a nim a ls amount s to
6,8 02 kilo grams p er hectare; o f this
qu antity protein constitutes a little
m ore than 2,804 kilogram s, et he r e xtract
o r fat 44 2 kilogram s and carbohydrates
3,556 kil ogram s. On the basis o f the
en ergy equivalents of these three cla sses
of o rga n ic matter, as indicated in a
previ ou s paragraph , the total energy
valu e of th e annual cr op am ounts to 34 6
gram ca lor ies per sq ua re cen tim e ter of
lak e s urface (table V).
TA BlI~ V. EfJl~ ,gy values of {he orga nic mailer in (he
organisms, tas rt hvr wit l: ' he estima ted amoum s 0/ encrs»
n'prl'_'i('nt t!d in their mera bohsm, and In th e di ssol ved organ ic
ma tter
The values are s ta ted in gram ca lories per sq uare
ccn tuu et er of lake: su rface . T he result s for phytopl ankto n
and z oo p lan k ton a re based on a tu r n over eve!')' t wo
week s durrng the YC.:lT.
Phy top lan kt on
Met ab ol ism . .
Zoo p la n k to n ... . .
\1 c tabolism . ... .. . .
Bottom ~ ora .. . .
M e tab oli sm . ... .. . . .
Bouo m fauna and ti sb .. .. . . .
Meta bo lism . ... .. .. ..
D iuol\'cd organic m a u e
T o!a l .. .. . .
In additi on to th e orga nic m aterial in
the plants a n d a nima ls, th e water co n­
tain s a certain am ount of organic m atter
whi ch cannot be recovered with a high
sp eed cen trif uge. It is either in true
so lu tio n o r is in such a finely di vided
st at e that it cann ot be o bta ined with a
cen trifuge; for la ck o f a better term it
has been call ed " diss o lved organ ic
m atter" as co mpare d with the "par ti­
c ula te organic matter" which ca n be
recovered from th e water with a centri-
CHAN CEY JUDAY
fuge . The water of Lake Mendota
contain s 10 to 14 milli grams p er liter,
dry weight, o f this di ssol ved orga n ic
matter; the mean o f so m e 60 deter­
minations is 12 milligrams p er liter.
W hen co m pu te d to an area ha sis, the
a ver age weight of th is materia I is 1,523
kilo grams per he ctare, dry weight , of
which 334 kil o gram s arc protein , 68
kilogram s fat and 1,121 kil ogram s
carboh ydrate. The energy value o f thi s
di ssol ved o rga n ic m atter is about 71
gram cal orie s per square centim eter.
Utilizat ion of'solar energy. The ch lo­
rophyll -bearin g aquatic plan ts are re­
sp o nsib le for the ut ilizat ion o f th e
subs ur fa ce radi ati on ; th at is, the s un
furni shes the [lower and th e chl or o­
phyll a nd as soci at ed p igm ent s of th e
pl ant s se rve as the m achines for the
manu facture o f the fund amental o rga n ic
matter o f the lake. Table IV sho ws that
the ph ytoplankton and the large aquatic
plant s co ns ti t ute the m ajor i te m in the
a n n ua l yield o f b iol ogical material.
Together they contribute 6,3 62 kil o­
grams o f dr y o rga nic m atter per hectare
a s compared with 440 kil ogr am s o f
zoopl ankton , bottom fauna a nd fish ;
that is, th e pl ant contributi on is 93 per
299
100
cent a nd the a ni ma l part is 7 per cent o f
22 the total organic matter.
. ll O
22
T able V give s the en ergy value of th e
7 va rio us constituents of the annual bio­
. 3
IS logical cro p . The two gro ups of pl ants,
71 namel y phyt oplankt on and large a­
649 quatics, h a ve an energy value of 321
gram calories as co mp a red with 25
gram ca lo ri es per squ are centimeter in
the a ni m a ls. The 321 gram calories
represented in the organic matter of th e
pl ants is only 0.27 of o ne per ce n t of the
mean a n nua l radiation deli vered to the
surfa ce of the lake , namely 118,872
ca lo ries. Two co rr ec tio ns need to be
made in this result, however. (I) A s
alread y ind icated so me 28,500 ca lo ries
o f sol ar en er gy a re lo st a t the surface of
the wat er and thu s d o not rea ch th e
aq uat ic vegeta tio n. Ded ucting this
177
amount leav es 90,372 ca lo ries which
pa ss into the water a nd thu s bec ome
available for the pl ants. On this basis
th e percentage o f utilization is increa sed
to a little m ore than 0 .35 of o ne per cen t
o f the a vail able rad iat ion . (2) A certain
am ount of the o rga nic matter synth e­
sized by the plants is u sed in their
metaboli sm a nd thi s d oes n ot a p pea r in
the percentage of utili zation given above .
Experiment s sho w th at some o f the
al gae utili ze in their metab oli c processes
about o ne- t hird o f th e o rga nic m atter
th at they sy nth es ize. No data a re
avail able for the lar ge aqu atics, but
a ssuming th at they al so utilize a sim ila r
proporti on in their metaboli sm , the tw o
groups of pl ant s would represent a
utili zati on of 42R ca lo ries which is
equival ent to 0.47 of o ne per cent of the
annual qu antity o f so la r energy that
actu ally enters the water.
This percent age is based o n an a ver­
age turnover in the phyt oplankton ever y
tw o weeks throughout the yea r, but
there is some evide nc e that the turnover
takes place m ore fr equently , especiall y
fr om April to October. With a n aver­
age turnov er o nce a wee k in the organic
matter of the phyt oplankton during the
year, the en ergy va lue o f this crop
would be 798 calories, includin g me­
taboli sm ; adding to this a mo u nt the 29
cal ories in the annual cro p of bottom
flora gi ves a total of 82 7 cal ories which
is utilized by the pl ants. This is 0 .91 of
one per cent of the 90,372 calories of
en er gy th at penetrate th e water a nd
become ava ilable to th e plants ; in round
numbers thi s ma y be regarded a s a
utili zati on o f o ne per cent.
Thi s sma ll percentage of utilization of
sol ar energy by aquatic pl ants sh ows
that Lake Mend ot a is not a ver y
efficient manufacturer of biol ogical pro­
du ct s in so fa r as utili zing th e annual
suppl y of so la r and sky radiation is
concerned ; o n the other hand it bel ongs
to the gro up of highly productive lakes .
While the a q ua tic pl ant cro p ap pea rs
 L78
to be inefficient in its utilization of solar
energy, it com pares very favorably with
some of the more important land crops
in this respect. Transcau states that
only 1.6 per cent of the total available
energy is used by the corn plant in
photosynthesis during a growing period
of 100 days, or from June I to Septem­
ber 8.... These computations for cul­
tivated crops, however, take into
account only the quantity of solar radia­
tion available during comparatively
brief growing periods and thus do not
cover the entire year as indicated for
the aquatic plants.
Energy value of animals . The organic
content of the animal population of the
lake represents a conversion and further
storage of the material manufactured by
the plants, but no direct utilization of
solar energy is involved in the transfor­
mation. It may be regarded as an ex­
pensive method of prolonging the exist­
ence of a certain portion of the original
plant material. As previously indicated,
it may take five grams of plant food to
produce one gram of animal tissue, so
that the plant equivalent of the animal
crop may be reckoned as five times as
large as the organic content of the ani­
mals; in the predatory animals, however
it would be much larger.
THE CONCEPT OF THE ECOSYSTEM
·'-a.ble V shows that the energy value
of the bottom and fish population is 25
gram calories per square centimeter; on
the fivefold basis, this would represent
the conversion of at least 125 gram calo­
ries of original plant organic matter.
This utilization is approximately 40 per
cent of the potential energy stored in the
annual plant crop of 321 gram calories
which is based on a turnover in the
phytoplankton every two weeks during
the year. A turnover in the phyto­
plankton every week would give a plant
crop of 620 calories and an animal
utilization of a little more than 20 per
cent.
Dissolved organic matter. The energy
value of the dissolved organic matter is
indicated as 71 gram calories per square
centimeter in table V. This material is
constantly being supplied to the water
by the various organisms and the stand­
ing crop of it remains fairly uniform in
quantity during the different seasons of
the year as well as in different years.
While there is a regular turnover in this
organic matter, it needs to be taken into
account only once in computing the
organic crop of the lake because it has
its source in the plants and animals for
which an annual yield has already been
computed. .. .
THE TROPHIC-DYNAMIC ASPECT OF ECOLO GY
Raymond L. Lindeman-1942
Reprinted by permission of the pu blisher from Ecology 23: 399-418. 1942.
This paper incorporates the most significant formulation in the development
of modern ecology, It provided not only a conceptual framework within
which to work but also stimulated a great profusion of effort because of the
basic questions it posed. Hutchinson stated the case weft in a postscript to
the paper, " . . . here for the first time, we have the interrelated dynamics of
a biocoenosis presented in a form that is amenable to a productive abstract
analysis." Lindeman's potentially productive career which began at the
Univer sity of WiSCOIISill and ended at Yale was foreshortened at age 27
while the following paper was in pre ss.
The trophic-dynamic viewpoint, as even such a unit as a lake primarily as
adopted in this paper, emphasizes the a biotic community appears to force a
relationship of trophic or "energy-avail­ "biological" emphasis upon a more
ing" relationships within the com­ basic functional organization.
munity-unit to the process of succession. This concept. .. is inherent in the
From this viewpoint, which is closely term ecosystem, proposed by Tansley
allied to Verriadsky's "biogeochemical" for the fundamental ecological unit. . . .
approach and to the "oekologische The ecosystem may be formally defined
Sicht" of Friederichs, a lake is con­ as the system composed of physical­
sidered as a primary ecological unit in chemical-biological processes active
its own right, since all the lesser "com­ within a space-time unit of any magni­
munities" are dependent upon other tude, i.e., the biotic community plus its
components of the lacustrine food cycle abiotic en vironment. ...
fo r their very existence. Upon further
consideration of the trophic cycle, the TROPIllC DYNAMICS

discrimination between living organ­ Qualitative food-cycle relationships.

isms as parts of the "biotic community"
and dead organisms and inorganic
nutritives as parts of the "environment"
seems arbitrary and unnatural. The
difficulty of drawing clear-cut lines be­
tween the living community and the non­
living environment is illustrated by the
difficulty of determining the status of a
slowly dying pondweed covered with
periphytes, some of which are also con­
tinually dying... . much of the non­
living nascent ooze is rapidly rein­
corporated through "dissolved nutri­
ents" back into the living "biotic com­
munity." This constant organic-inor­
ganic cycle of nutritive substance is so
completely integrated that to consider
179
Although certain aspects of food rela­
tions have been known for centuries,
many processes within ecosystems are
still very incompletely understood. The
basic process in trophic dynamics is the
transfer of energy from one part of the
ecosystem to another. All function, and
indeed all life, within an ecosystem
depends upon the utilization of an exter­
nal source of energy, solar radiation. A
portion of this incident energy is trans­
formed by the process of photosynthesis
into the structure of living organisms.
In the language of community econo­
mics introduced by Thienernann, auto­
trophic plants are produ cer organisms,
employing the energy obtained by
180 THE CONCEPT OF THE ECOSYSTEM RAYMOND L. LINDEMAN 181

photosynthesis to synthesize complex the following productivities : Ao (rate of mechanically constituted , some "can­ solutes reavailable to photosynthetic
organic substances from simple inorgan­ incident sola r rad iation), )..,1 (rate of nibalism" within such an arbitrary level producers. These saprophages may also
ic substances. Although plants again re­ photosynthetic production), A 2 (rate of can be expected, so that the actual serve as energy sources for successive
lease a portion of this potential energy primary or herbivorous consumption), value for predation loss from A ~ proba­ levels of consumers, often considerably
in catabolic processes, a great surplus of )..,3 (rate of secondary consumption or bly will be somewhat above zero. The supplementing the normal diet of herbi­
organic substance is accumulated. Ani­ primary predation), and A, (rate of predation loss from level "\"- 1 will vores.. . .
mals and heterotrophic plants, as tertiary consumption). The total a­ represent the total amount of assimi­ APPLICATION. The value of these the­
consumer organisms, feed upon this mount of orga nic structure formed per lable energy passed on into the higher oretical energy relationships can be
surplus of potential energy, oxidizing a year for any level An, which is com­ level (i.e., the true productivity, All)' illustrated by analyzing data of the
considerable portion of the consumed monly expressed as the annual "yield," plus a quantity representing the average three ecosystems for which relatively
substance to release kinetic energy for actually represents a value uncorrected content of substa nce killed but not comprehensive productivity values have
metabolism, but transforming the re­ for dissipation of energy by (I) respi­ assimilated by the predator. . . . The been puhlished (table 1). . .. The calo­
mainder into the complex chemical ration, (2) predation, and (3) post­ predation loss from level :\ ,,_, will rific values in tahle 1, representing
substances of their own bodies. Fol­ mortem decomposition. Let us now likewise represent the total amount of annual production of organic matter,
lowing death, every organism is a poten­ consider the quantitative aspects of assimilable energy passed on to the next are uncorrected for energy losse s.
tial source of energy for saprophagous these losses . level (i.e., A" ,), plus a similar factor for T/\l)U~ r. Prodn ct ivrties off ood-groups in thr e e aqu atlc
organisms (feeding directly on dead RESPIRATORY CORR ECTIONS. The a­ unassirnilated material, as illustrated by ("( 'o," y yu'm\' , as g-ctJl/,:m ' /j,'('ar. uncorrected [or IO S Sl:X dueto
rvsnt ration, pre dation ami decomposit ion. Dot a [ rom
tissues), which again may act as energy mount of energy lost from food levels the data of tables II and Ill. The various Brujewic t ('39). J uday ("40) and Lindeman ("4Jb)
sources for successive categories of by catabolic processes (respiration) categories of parasites are som ewhat - -
Ot<pian 1.a ke Cede
consumers. Heterotrophic bacteria and varies considerably for the different comparable to those of predators, but Sea Men- Bog
Lak e
fungi, representing the most important stages in the life histories of individuals, the details of their energy relationships d ota

saprophagous consumption of energy, for different levels in the food cycle and have not yet been clarified . Phytoplan ers : Al . . . . . . 59.5 299 25.8
Phytobent ss : AI ... .... 0.3 22 44.6
may be conveniently differentiated from for different seasonal temperatures. In DECOMPOSITION CORRECTIONS. In con­ Z oopl :.m k 'S : A:a . . . . . . .. 20 .0 22 6.1
animal consumers as specialized decom­ terms of annual production, however,
posers of organic substance. Waksman individual deviates cancel out and
formity with the principle of Le Chate­
lier, the energy of no food level can be
Ben thic b r ","'Sl' TS: All •• . .•
Bent h ic pr tai ors : AJ ....
Pla nkton eda io rs : :\ ~ . . .
J 20.6
I
1.8"
0.9"
0.8
0.2
0.8
has suggested that certain of these respiratory differences between food completely extracted by the organ isms " For a ge" he" A,( ·, A, ?).
Ca rp : .~ ,, ( · A ~ ·!) . . ... . . .
0.6
0.0
'!
0.2
0.3
0.0
bacteria be further differentiated as groups may be observed... . which feed upon it. In addition to the "G . mc" fi lea: ,1\ .1 (- ' - \ 31). 0.6 0.1 0.0
Considering that predators are usual­ Se a k : A5- . . . . . . . . . . . . 0.01 0.0 0.0
transformers of organic and inorganic energy represented by organisms which -
compounds. The combined action of ly more active than their herbivorous survive to be included in the "annual " R oughly assumin g th:::t , of th e bonom fauna 1S
h erbiv o rous.
animal consumers and bacterial decom­ prey, which are in turn more active yield," much energy is contained in
posers tends to dissipate the potential than the plants upon which they feed, it "killed " tissues which the predators are Correcting for the energy losses due
energy of organic substances, again is not surprising to find that respiration unable to digest and assimilate.. . . J\l­ to respiration , predation and decorn­
transforming them to the inorganic with respect to growth in producers (33 though the data are insufficient to war­ position , as discussed in the preceding
state. From this inorganic state the per cent), in primary consumers (62 rant a generalization, these values sug­ sections, casts a very different light on
autotrophic plants may utilize the dis­ per cent) and in secondary consumers gest increasing digestibility of the higher the relative productivities of food levels.
solved nutrients once more in resynthe­ ( > 100 per cent) increases progressively. food levels , particularly for the benthic The calculation of corrections for the
sizing complex organic substance, thus These differences probably reflect a components of aquatic cycles. Cedar Bog Lake values for producers,
completing the food cycle.. . . trophic principle of wide application: The loss of energy due to premature primary consumers and secondary con­
Produ ctivity. DEFINITIONS. The quan­ the percentage loss of energy due to death from non-predatory causes usual­ sumers are given in table II. The appli­
titative aspects of trophic ecology have respiration is progressively greater for ly must be neglected, since such losses cation of similar corrections to the
been commonly expressed in terms of higher levels in the food cycle. are exceedingly difficult to evaluate and energy values for the food levels of the
the productivity of the food groups PREDATION CORRECTIONS. In consid­ under normal conditions probably rep­ Lake Mendota food cycle given by
concerned. Productivity has been rather ering the predation losses from each resent relatively small components of Juday, as shown in table III, indicates
broadly defined as the general rate of level, it is most convenient to begin the annual production . ... that Lake Mendota is much more
production , a term which may be ap­ with the highest level, An. In a mechani­ Following non-predated death, every productive of producers and primary
plied to any or every food group in a cally perfect food cycle composed of organism is a potential source of energy consumers than - is Cedar Bog Lake,
given ecosystem.... organically discrete levels, thi s loss by for myriads of bacterial and fungal while the production of second ar y con­
In the following pages we shall con­ predation obviously would be zero. saprophages, whose metabolic prod ucts sumers is of the same order of magni­
sider the quantitative relationships of Since no natural food cycle is so provide simple inorganic and organic tude in the two lakes.
182 THE CONCEPT OF T HE ECOSYSTE M RAYMON D L . l.I NOf' \I AN 83

At first sight, thi s generalization o f success ion, productivity in cre a ses much
T ABl R 11. Producti vity values f o r the C edar Bog Lak e degr ee of ut ilizati on o f its potentia l food
fo od ('y d ,', iff N-ca l /i n, t ,.,I'r a r. ,'U corr tvc tcd by usin): (h(! incr ea sin g efficien cy in higher co ns ume r mor e slo wly. . . .
('ort11k it·nt J derived in flu p r~ n: r!,'m: sect ions s upp ly o r energy so u rce . All efficiencies
gro ups would a p pea r t o co nt ra d ict th e
discussed in the follo wing pages a re SUCCES SIONAL PR ODU CTIVIT Y Cl; RYES
Troph ic
level
IuncorrC'( Ie<J1 R
proc.tuC:li", lry ~ :ra ~
.< :-It~
Pre-
·
~~~. r~~r~~Jo-
posi-
rion li ~i~~
progressive efficiencies, ex pressed in
terms of relat ive product ivities
previ ou s generali zati on th at th e loss of
ene rgy due to respirat ion is progrcs­ J n recapitulat ing th e pr oba ble ph ot o­
on Io n sively grea ter for high er levels in th c sy nthetic producti vity relation sh ip s in
P-r-o-d-u-cc-r-s-,-,,-,1·7-0-.4- t- IO
- .-1-4 ; 23.4JI4.~ 2.M Ill :;
((>"n/>"n-I ) 100). It is import ant to re­
member th at e fficiency and producti vity
food cycle. These ca n be reco nc iled by hydr ar ch success io n, we sha ll vent ure . ..
Primary rememberin g th at increa sed ac tivity of a hypothetical hydroserc , dev elo ping
consum cr s : A, 7.0 -= 1.07 4.4 3. 1 I 0.3 14.8 are not syn onym ou s. Productiv ity is a
pred at ors con sid era bly increa ses the Irorn a mod erately deep lake in a
Sec on dar y
co ns ume rs : A' I I~ 0.43 " J.8 0.0 0.0 L J~ rate (i.e. , in t he units here used , ea lfcm ~/
yea r) , while effi ciency , bein g a rat io , is
cha nces of en counterin g suit a ble pr cy. fertil e co ld t emper ate re gion under
is va lue includes the product ivit y of the small Th e ultimate effect o f such a ntago nistic rela tively co nst a nt clim atic co ndi tio ns.
cypr.ic o id fishes found In Ih e lak e. a dim ensi onl ess number. T he point s of
princip les would present a pict ure of a The initial peri od of o Iigo t ro p hy is
reference for a ny efficiency value sho ud
pred ator completel y wca ring it self o ut believed to be relat ively sho rt. . . . wit h
III. Produc tivity values f or thr Lake Mv ndotu f ood a lway s be clearly sta te d .
TAR LE
in th e process o f co m plete ly ex ter mi­ product ivity rapidly increasing until
cyc le , in g -callt m ' /y rat', as co rrecte d by using co efficients The p rog ress ive efficien cies (11. ,.1>..,.. ,)
derived in Th t' prec ed mg sections. a nd as given by Juday natin g its p rey. a very improb abl e sit ua ­ eutr op hic stage-eq uilibrium is a tta ined.
(' 40) 100) for the tr ophi c levels of Ceda r Bo g
tion. H o wever , Elton point ed o ut th at T he durati on o f hi gh eutroph ic pro­
Lake and Lake M end ot a, as o btained
ncoedr- !Res - Pre- D e--
ureel I Co r• co r- IJUd,Y'S from the prod ucti vities derived in
food- cycle s rar ely hav e mor e th an f ve
tr ophic level s. Am ong th e severa l Ia ct or s
du ctiv it y dep end s upon the mean depth
o f the basin a nd up on th e rate o f
Trophic Le ve l] pro- p,ir3- ~3 · ~~: ~~~~~. ree led tables II and Il l, a re presented in tab le
~ll.l c. li on Iu o n rio n I l 1'v I' tY r
uvuy Pr.(}Vll
o ~uc
y - inv olved, increa sin g resp ira t ion o r su c­ sed im enta tio n, a nd producti vity fluctu­
Pr odu ccr s : 1\ I .12 " 107" '42 W 480 - 42S - ­
cessive levels of pred at or s co ntrastcd a tes a bo ut a high eutrophic mean until
l AOLE IV . Producrtvit irs and pro f r, .u i'·4! ~tJicit!n cie s il l lilt'
Prima ry co n­ C eda r B t.1g La k e ami Lu/.,(· M endota food cy r lt' f, us with th eir su ccessively increasin g c r­ the lake bec omes too sha llo w for ma xi­
sume rs : 1\2 24 15 2.3 0.3 41.6 144
g - c4/" f1'J1 1' i~o r ficicncy o f pred ati on ap pea rs to be mum g rowth of phytoplankton o r
Seco ndary
con sum ers : A3
Ter tia ry con­
It I 0.3 0 .0 2. 3 6
Ced llr R"~ l -,, ~e 1 l. a ke M ~nd o t a important in restrictin g the n umber o f rege ne ra t ion of n utrients from the ooze.
sumcrs : A" 0. 12 0. 2 0.0 0 .0 0.3 0.7 Pr odnc-
t ivit y
I Em- I
ci ency
Pr od uc-
tivity
EIIi­
c icncy
trop hic le vels in a fo od cycle. . . . A s the lak e becom es sha llower and mor e
senescent, pr oducti vity is increasingly
TROPHI C-DYK A\ll CS IN

• Hutchinson g ives e viden ce th at thi s va lue is pro ba bly

to o high and m ay ac tually be: as 10 ...... us 250.
]b d illlicm I~I 1l. ~ 7 . j
II I.J O. I O ~{
j lI M.Il72
41l0'
.
I 0.4 0 ~{; SU CCESSION

influe nced by clima t ic fluctu a tions a nd

j A ppu r cn t fy such o rgnrusms as sma ll " Ioruge'' fas.h ~
P rod ucers : ,\ 1' . gra d ua lly decl ines to a m inimum as th e
Primmry
a re n OI, inclu de d In any part or JUd'1Y·S ba la nce shee t, cens u mers : 1\'.1. 14.8 U . 3 ~{ 4 1.6 ' 8.7 % Dynamic processes within a n ecosys­ lak e is co mpletely filled with sed i­
The in cl us ion o f these Co rms might be expec ted to inc re ase Seco nd a ry
con siderabl y the pr c d ucuvi ty o f seco nd ary co nsarmptio n. co nsurner s.: A ~ 3 . 1 1 22. 3 ~;, 2. 3Il 5.5% tem , over a peri od o f tim e. tend to ment s. ...
Ter tiary
com umers I 0.3 , 1 3 . 0 :.~
produce certa in o b vio us ch anges in its Efficie ncy relationships in su ccession,
_: _J'
,\ ~ !I - _
species-co mpos itio n, soil characte ristics Th e succes sio na l cha nges of phorosyn­
Bi ological effic iency. T he qu antit ative "P ro b ably to o hi8h ; "" fo o m o te o f lab lc 111.

theric efficiency in natural a rea s (with

t Pr ob ab ly 10 0 I" ", ; see f oo tno te or t~ b l c III.
and productivity. Ch an ge, accord ing to
rel ation ships o f a ny food -cycle level ma y Co oper , is th e essenti al criterio n o f respect to so lar rad iat ion , i.e. , (>.. J >..o)
be ex p ressed in term s of its efficiency 100 ) ha ve not been inten sively studied.
1V. In view of the uncert ainties co n­ successio n. Fro m th e tr ophic-d ynamic
wit h respect to low er levels. Qu oting
ce rn ing so me of th e La ke M endo ta viewpoin t, success io n is th e pr ocess o f In lak e successio n, ph ot osynth etic ef­
Hutchinson 's defin ition , "the efficiency
pr oduct ivitie s, no de fin ite conclusio ns development in an ecosyste m , br ou ght ficiency would hc expec ted to foll ow th e
of th e productivit y of a ny level (A.,,)
can be dr a wn from th eir relat ive ef­ abo ut primaril y by the effec ts of th e sa me co urse deduced for productivity,
relati ve to th e pr od uetivit y of a ny
ficiencies. Th e Ced ar Bog Lak e rati os, organ isms o n the environ me nt a nd up on rising [ 0 a mo re o r less con stant va lue
previous level (A m) is defined a s (11."'/>"14) each o t he r, to ward s a relat ively sta ble du rin g eutro phic sta ge-eq uilibr ium, a nd
100. If th e rat e of sola r energy entering ho wever, indicat e th at th e pr ogressive
efficien cies increase from a bo ut 0.10 per co nd ition o f eq uilibr ium. d eclining d ur ing senesc ence, as sug­
the ec osystem is den oted as 11. 0, the It is well kn own th at in th e initial
cent for production , t o 13.3 per cent for gested by a ph ot osynth etic efficiency of
efficiencies o f all levels ma y be referred ph ases of hydrarch successio n (o ligo ­
primary cons umptio n, a nd t o 22. 3 per at least 0.2 7 p er cen t for e utro phic
back t o th is qu antity 11.0," In ge nera l,
tr ophy -s-eutrophy) producti vity in­ La ke Mendot a a nd of 0.10 per cent for
however , th e most interesting ef­ cent for second ar y co nsumpt io n . . ..
These progre ssivel y in cre a sin g efficien­ creases rapidl y ; it is equall y a ppa rent se nesce nt Ceda r Bog Lake . Fo r th e
ficicncies a re th ose referred t o th e
tha t th e colonization of a bar e tcrrcstri a I terrestri al hydroser e, efficien cy wo uld
previou s level ' s productivit y (>"n- '), o r cies may well represent a fund amental
trophic principle, nam ely, th at th e a rea represents a sim ila r accelera tio n in likewise foll ow a curve sim ilar to that
th ose expressed as (>"n/>"n-I) 100. These
co ns ume rs at pr ogressively higher levels pr oduct ivity. In th e lat er pha ses of p ostulat ed for productivity. .. .
latter may be term ed the progressiv e
effi ciencies of th e va rio us food-cycle in the food cycle a re progressivel y more
levels , indicatin g for ea ch level the efficient in the use of th eir food sup ply.
 G EORG E I.. C1.AH K E 185
DYNA!\HCS OF PRODUCTION IN A MARINE AREA
SI ZE OF ST ANDING CORP the a rea but which are co ns umed, or die
George L. Clarke- 1 946 EOUILI8R IUIl STANDING CROP in othe r ways, and dec ompose so as to
enter the ecological cycle again within
Reprinted by permi ssion of the author and publi sher from Ecol ogical
Monograp hs 16 : 321- 335, 1946. the sa me a rea a rc not considered to be a
yield in the sens e recommended here .
'"
w
However , dead orga nic matter whi ch
Probfem s and conf usion in the description and measure ment 0./ the cncrgy '"
relation s within ecosys tems , and concom itantly in the attemp t to compa re 'z"
::>
has become permanently ina cce ssible ,
the energetics of ecosystems, were considerably clarified by Clarke . I II this as for example, if buried under bottom
clear-cut discussion of concepts of production, he indicates th e lim itat ions deposits , is a n irrevocable los s to the
of tradition ally used m easurem ents and sugges ts a basis f or meaningfu l system , and hence forms an other subdi­
compa rison which is fl OW widely fo llowed. (See p. 27 f or another paper by vision of the material rem oved .
Clarke .) FIG. 2. Cha nges of populatlon size with tim e in These different subdivisio ns have in
three hypothetical sit uations (fo r fu }! explanation comm on the fact that in ea ch instance
see text] , the particular materi al whi ch is remo ved
from the a rea does not return again to
to th e rat e of net increa se. If, however , that a rea . .. . Because of the inevitabl e
be ade q ua te for the given sit ua tion. the sta ndi ng crop is measured before th e loss of mat eri al s (a nd energy) at each
CON CEPTS OF P RODUCTION
Eva lua tion o n th e basis of energy con­ asy mp to tic level is approached , th e level in t he food cha in, the yield will
All the ideas and me asurements of tent ha s the ad vantage that the efficiency sit ua tio n with th e more rapid incre ase ordinarily be much sma ller th an the
p rod uctivity which ha ve an eco logica l of the utili zation of th e incident so lar will have th e lar ge r standing crop. T he su pply. Measurem ent of the rati o of
applica tio n, may be gro up ed under th e r adiation may bc ca lcula ted directl y. size of the sta nding cro p at a ny time is yield to supply, and h ence th e efficiency
foll owing three fund am ental co nc ep ts: Standing cro p. M eas uremen ts of the the resul t o f the summ ati on of the of the formation of the yield, is th ere­
Standing cr op-the a mo u nt of or gan­ sta nding cro p , an d hence of the con­ excess of prod uction over destru ction fo re of gre at importance to . . . ascertain
isms existing in th e area at the time of centrati on, of th e vario us species inhab ­ from the beginning of th e gro wth o f the whether the actua l yield represents a
observation. itin g the area a rc essential in judging popul ation to the mom ent of o bser­ needlessly low utili zation o r an over­
Mat erial rem o ved- ·thc a mo unt of the harmful o r benefi cial effec t of expl oit ation of th e ar ea .
vat ion.
orga nis ms rem oved from the area per cr owding within the species, and the In the third hypothetical sit ua tio n, Product ion rate. The co nce p t o f the
unit time by man , or in other ways . effectivene ss with which dep end ent the sa me le ngt h of time is required for production rate as the am ount of
Production rate-the amo u nt of or­ spe cies can feed up on fora ge species . the p op ula tion to reach a max imum or ganism s formed per unit tim e (per
ga nism s j ormed within th e ar ea per unit When applie d to the exploitat io n of a size as in th e seco nd sit uat io n, but th e unit area o r vo lume) is complicat ed by
tim e. natural po p ulatio n by ma n, the m agni­ maximum level reached is lower. In the fact that in most natural areas
All three of th ese major con cepts of tude of the sta ndi ng crop simila rly both sit ua tio ns, th e po pulat ion may o rga nic matter is being formed o r tr an s­
produ ctivity arc imp ortant , and are infl ue nces the size of the catch per unit fluctuate a bov e a nd belo w the ave ra ge form ed at several trophic levels sim ul­
essential for a co mplete underst anding effort. Ho weve r, a kn owledge o f th e limiting value (b ro ken line s at right of taneously, i.e., by the plants and
of the operati on of th e are a as an sta nd ing crop docs not give any infor­ cur ves), a nd since eq uilibri um is bein g animal s of the food cha in which depend
ecol ogic al complex. In addition, th e mation as to the time whi ch has been m aintained, at least temporarily, the upon o ne a no ther. In order for meas­
qu antities inv olve d are to a certa in required to p roduce the cro p , or its aver age rate o f pr oducti o n m ust be ur ements of production rat e t o be useful
extent mutuall y dep end ent. To avo id rcpl aceability, exac tly ba la nced by the average rate of it is necessar y to k eep separate the
a m biguity in discu ssin g th e ecological T o illus tra te the se points, th e chang es dest ructio n.. . . values for the vario us tr ophic levels a nd
relati on ships of th e ar ea , it is sugges ted of population size with time for three M at erial removed. T he mat eri al re­ in each case to di stinguish between
that th e terms "productivity" and hyp othetical situations a re rep resented m ove d fr om a n area may fall int o gros s and net ra tes of form ation.
" pro d uctio n" be not used in referring to dia grammati call y in Figu re 2. In the severa l categor ies. In the first place , the These trophic rel ati on s a mong the
the sta nding cro p o r to th e materi al first situa tio n the growth rate o f the a mo unt of o rga nis ms ha rvested by man components of the pr oduction p yramid
rem oved unl ess a phrase is added to popul at ion is much more rap id th an in during a certain period ma y be me as­ may be ... [considered] first , for a
make the meaning clear. Measurements the seco nd, but co ndit io ns are suc h that ur ed a nd desi gnated as t he yield. In situa tion in which the co n str uctive and
made und er an y of the three categories the sam e size of sta nd ing cro p is attained ad di tio n, o ther or gan ism s m ay be de struct ive pr ocesses are equal so tha t
ma y be sta ted in terms o f nurn ber of pro vided th at the seaso n is suffic iently perma nently removed fro m the area by there is no ga in o r Joss in the amo u nts of
indi vid ua ls, weight or " b iomass," ener­ lon g for th e ma ximum valu e to be wind or cur rents, o r by em ig ra tio n organisms present at th e end of the
gy content, or a ny o ther characteristic reached . 1n thi s case, th e final size of the accomplishe d by the o rga nisms' own period ove r that a t the beginning.. ..
(such as chl or ophyll co ntent) which ma y sta nd ing crop gives no info rmat io n as In tr acing the energy and material
locomotion. Organisms which gr ow in
184
186
thro ug h the pr oducti on pyramid we ma y
start wit h th e incide nt light whic h fa lls
up on th e area. A porti on o f thi s light
(sma ll, in the aqu ati c en viron ment)
reac hes th e plant cells and is absor bed
by them . A small portion o f the a b­
sorbed energy appea rs as th e ca rbo ­
hydrate which ha s been fo r med by
p hot osynthesis (P j). Th e a mo unt o f
mat erial p ro d uced by thi s a na bo lic
process (or of the energy represent ed) is
term ed th e gross plant p roduc tion. It has
a lso been ca lled the primary production
becau se the an ima l substa nces which
res ult fro m co ns uming th e plants, a nd
each other, a re transformati on s of th e
origina l plant mat erial , a nd thu s rep­
rese nt altern a tive forms of the sa me
materi al (a nd energy). .. . A lar ge part
of the gross pr oducti on is lost as th e
res ult of ca ta bo lic proc esses. Thi s loss
ma y be measured b y the a mo unt of
resp iration . Th e rem ain ing fracti on of
the gross producti on accou nts for th e
new plant growth (P ,) a nd thi s is
termed th e net plant p roduction, o r
simp ly, th e plant producti on, for unit
a rea and time.
Since in th e case con sider ed th er e is
no p ermanent increase o r decr ease in
pl ant material , all of thi s new plant
gro wt h is d estroyed a nd th e energy
tr an sfo rm ed before the end of th e peri od .
P art of th e plant materi al p roduced is
co ns umed by herbivores and th e re­
maind er dies in so me other mann er and
d ecomposes. In many situa tio ns a lar ge
pr oporti on of the plant material co n­
sumed is not assimila ted by the her­
bivo res, but is pa ssed through t he gut
undigest ed . This fr action ( U) may be
added t o th at wh ich has decomposed
foll o wing de structi on in o the r ways .
A simila r a na lysis ma y be mad e of t he
mat eri al which enters into each of th e
succes sive t rophic levels represented by
th e a nima ls and colo rless plants. . ..
At t he e nd of the peri od for thi s
hyp oth eti cal case of co mplete eq uili­
br ium th e sta nding cro p is of exactly th e
THE CONCE PT OF THE ECOSYST EM GEO RGE L. CL ARKE 187
same magnitud e as at the beginnin g a nd In the discu ssion o f produ ct ion rate wh ole year in a n a tte mpt t o reac h a
there has o bv io us ly been no yield. thu s fa r nothing has been said a bo ut the "to ta l" an n ua l va lue. If the gr owth of
Produ cti on has been go ing on, ho wever , len gth o f the p eri od co nsidered . The the plants is measured as energy, it is
at definite ra tes at th e va rio us levels, a nd tim e required for eac h t ype of o rga nis m permi ssibl e to summa te the am ount of
in this case destru ction of the material s to co mp lete its grow t h, to die a nd energy wh ich has been tran sformed du r­
has exac tly bala nced th eir formation deco mp ose, a nd to start tile cycle ove r in g the yea r since th e energy ca n be used
duri ng th e peri od . . . . aga in is kn own as its peri od of turn over. by the plant s o nly o nc e. A co mpa riso n
Let us conside r no w a seco n d case, in T he len gth of th e turn over period ma y thu s be mad e bet ween the a nnua l
which sma ller a mo unts of o rgan isms us ua lly differs wid ely fo r th e o rga nisms incident radiation a nd the energy con­
a re destroyed th an are fo rm ed during a t th e d ifferent tr ophic levels o f th e tent o f th e plants produced.
the period . Un de r th ese circ umsta nces pr odu cti on pyramid , a nd may differ for In th e case o f th e o rga nisms at succes­
the grow t h (P 2) of a nyo ne of th e th e sa me level in different situa t ions and sively highcr t ro phic levels, both th e
ca tego ries of o rga nisms would be a t differ ent seaso ns of th e yea r. The mat er ial s a nd th e energy are used over
acco unted for a t th e end o f the period gree n plants o f a terrestrial a rea may again one or m ore times. A su mma tion
in three ways : on e porti on cons umed . have esse ntia lly o n ly o ne turn ove r per o f amo unts of pr oducti on a t all t rophic
o ne p ort ion de composed , and o ne yea r, wherea s the ph ytopl ankt on of an levels fo r a lon g peri od , suc h as a year ,
portion rem ain ing in ex istence a nd aq ua tic a rea may turno ver within sho rt therefor e sim ilarly ha s little meanin g. . .
rep resen ting a n incr em ent whi ch may be peri od s var ying wide ly fro m a few days . .. Product ion is there for e best
term ed th e net increas e (P 3)' The to severa l week s. In the latter eas e the mea sured as a set o f rate s a pp lying to
sta nd ing cr op a t the e nd o f the period is sa me material may be used ove r again the gross p rodu cti on. net production,
cons eq uently lar ger th an at th e begin­ severa l times during th e year , a nd it and net incr ease fo r eac h ca teg o ry of
nin g of the pe riod by th e a mo un t o f t he wo uld thu s hav e littl e meaning to ad d organisms o r each tr ophi c level. . . .
net in cr ea se. In a situatio n in which th e up the increments of growth for th e
a mo unt of destruct ion d uring the peri od
was g reate r th an th e amount of pr o­
du cti on , th e net incr ease (p.l ) wo uld be
negati ve. and the sta nding cr op wo uld
be reduced .
If th is surplus mat eri al is perma nently
removed from the syste m by man, it
will co nstit ute a yield. Obv iou sly these
org a nisms co uld n ot co n tinue t o be
rem oved fro m th e system in succeed ing
period s unl ess a n eq ua l a mo unt o f
materi al, as nutrients or in so me other
form , were added eac h period . If the
o rga nisms representing the net increase
are not rem oved fro m the sys tem, and if
circ umsta nces are s uch th at they ca n
co nt inue in existence in th e a rea, there
will be a perm an ent increase in thc size
of the sta ndi ng crop . Again , this ca n
take place o nly to th e extent to which a n
eq uiva lent supp ly of nutrients or ot her
mat eri als are adde d to th e area . T he
sta ndi ng cro ps of la kes a re freq uently
see n to incre ase in thi s way with the
result th at th e lak es become more
eutrop hic a nd the p ro cess is kno wn as
eutrop hication ,
 HOWARD T . ODUM 189

TROPHIC STRU CTURE AND PRODUCTIVITY OF SILVER

KILO' CALORIES
SP RINGS, FLORIDA . , :.:', PH OIOSY~THET I C PE~ YEA~
. ,
~ . t .. BlOMASS 101"

Howard T. Odum-1957
Reprint ed by permission of the author and publisher from Ecological
Mo nographs 27: 55-112,1 957.

The signifi cance of this paper lies in its extensive application of Lindeman's
trophic-dynamics concept . It has been a model for subsequent studies on
different ecosys tems. It also intr oduced an energy flow model which has had
successful instructional value in incorporating the various energy relations
ofan ecosystem into a readily comprehensible pattern. Following a productive
career as Director of the Institute of Marine Science, of the University of .'
'{.
CO....UNITY

~~ " .'~~
Texas, Odum recently assumed the directorship of th e terrestrial ecology '8 ,796 RESP1RATION

program of the Puerto Rico Nuclear Center. I'

The study of Silver Springs reported

here has been made with the purpose of
determining the basic structure and
workings of flowing water ecosystems
ficiency maximum power hypothesis... ,
wherever an energy transformation
occurs mo st of the energy is dispersed
into heat that is unavailable for further FIG. 7. Energy flow diagram with estimates of
r

systems. These are defined as follows

by the careful study of one stream under
some unusually favorable conditions
provided. In recent years such holistic
con sideration of the energy flux and
biomass have provided a fruitful ap­
proach to the understanding of many
types of ecological communities. . . .
use by the organisms in the com­
munity,
Since the overall metabolism ha s now
been estimated. , . it is now possible to
make some rough estimates of the
component flow rates . Where estimates
are complete, the diagram may be used
energy flows in kilo-colones per square meter per using symbols from the flow diagram:
year in the Silver Springs community . The small
diagram contains symbols usedin text discussion P = production rat e (rate of net organic
synthesis in th e form of the species
( P, production; I, energy intake; A, consumer
of the trophic level or in storage
assimilation; R, respirations.
produc ts)
A ~ rat e of ass imila tion
I = rate of ingestion (consumption) or
energy intake

Diagrammatically, the object of the as a check since the parts must agree R = rate of resp irati on

energy approa ch may be said to be with the whole. Where estimates are directly using varied means depending Ell = utili zati on efficiency

the compl ete quantitative determination incomplete the diagram may be used to on the organisms . As described in ~_ h ~ .!J.. ..!J.. _ I ;
of the states and flow in Figure 7 as fill in the part from the total. The figure sections on animal growth rate above, PI 1'2 1'.1
must be regarded as extremely tentative thi s has been done in some cases and E. = assimilation efficiency
well as the control mechanisms by which
such a picture is sustained. The story in because of the many possible errors and reported as turnover. The best esti­ 'c. A~ = .l'I.3 = A j = A"

mates from this table are used in the J, I, 1I I ,

this paper concerns the details and incomplete data. E, = tissue growth efficiency

workings previewed in Figure 7. .. . As discus sed in the animal growth energy flow diagram in Figure 7. The _ 1'2 _ p ! _ P, _ 1'5

Energy flow diagram . In Figure 7 the
flow of energy through the community
in Silver Springs was shown according
to trophic levels. The use of the diagram
is applicable in general form to any
system . It shows distinctly the workings
of the first two laws of thermodynamics.
rate section , there are several procedures input of energy into a trophic level is - A2 - A; - Aj - A;
required to work out the flow rates of mainly the sum of the respir ation and E.. = ecological growth efficiency
production at the various trophic levels. net production if one neglects egested = 1'I ,, = ( Ei) (En)
To obtain the respiration of a trophic unass imilated matter. The flow picture
level one determines the standing crops, in Figure 7 shows the relationship of E, = Lind em an efficiency
(ratio of intakes of trophic levels)
estimates the respiratory metabolism growth, utilization, assimilation, and
per gram, and multiplies to obtain the heat loss for each trophic level. Fr om ~ .!L .,., (E.)
(E u )
1I •

According to the first law the total total respiratory metabolism in cach these concepts one may also derive the p trophic level production ratios

influx of energy equals the total flux out.
According to the second law, especially
as interpreted with the optimum ef­
trophic level. The net production of the several types of efficiencies (ratios X I'
trophic level follows from the deter­ 100) useful in describing ecological = 1':
= (E u ) (E e) = (Eu.) (E,) (E 3 )
mination of the growth rates (output)

188
190
The dat a in F igure 7 permit th e calcula­
tion of the ecol ogi cal gro wth efficiencies
(E.) and th e tr ophic level producti on
ratios (E ,,, syn on om ou s with Lindem an's
usage of tr ophic level efficiencies o nly
Eo
Plant N et 8,833 = 42 %
Prod uctio n 20,810 0 20,8JO
Herbi vores 1'±~~ = 44 %
3,368 0 0
Carn ivores 3~; = 17%
T op
Carnivores Z¥ = 29 %
Great co n fus ion o fte n res ults fr om
m isunderst anding as to which efficienc y
is meant. The energy flow d iagr am is a
help in thi s clar ificatio n. I n an oth er
com mu nica tio n som e current u ses of
the term produ cti on as syno no mo us
with assimi la tion a re qu esti on ed.
Th e general principle from Lindeman
a nd Dineen th at tr ophic level efficiency
(E I or E j ,) inc reas ed a long the fo od
cha in wa s not entirely co nfir me d for
Silver Springs a lt ho ug h the estimates o f
rates at th e high er tr ophic le vels ar e
pr ob abl y not acc ura te enough to test
th e rel ati on ship defini tely.
Turnover, climax and steady state.
With a gr oss prim ar y pro duct ion during
a yea r of 6390 gms/rn' an d with a
standing cro p biom a ss o f a ll compon­
ent s of 819 grns/rn' it is clear fro m the
rati o th at th e co mm unity turnov er is 8
times/ year. A s di scu ssed elsewhere, th e
smaller th e co mpo ne n t orga nis m, th e
more rapid th e turnov er. Th us the
overa ll co mm unity turn over ha s m ean­
ing only in th at it expresses th e relati on­
ship o f th e size o f th e co mmu nity an d
th e total productiv ity.
Wh ere a ll of th e co mpo ne nts o f a
com mu nity turn ov er seve ral tim es a
yea r th er e would be ampl e op port u nity
for c ha nges to occ ur if the re wer e n o
self regulative mech an ism s. In Silv er
T HE CONCEPT OF T HE ECOSYSTEM
in th e first tr ophic leve l). The eco lo gica l
gr owth efficienci es E" a nd th e trophic
level production rati os a re g iven as
foll ow s:
Er E,
8,833 = 42 % 20,RJO = 5 %
Q 4 10,000 0
1,478 ~ = 17 %-
8,8 3 3 ~
3,368- = 16 %*
20,810 Q
67 _ 5 %-
1,478. - 0
383 _ IJ 'Yo_
3,368- ­ 0
~ =9 % 21 6 %
383 = 0
Springs there is ap paren tly a fairl y
high d egre e of sta bility. Shelford &
Eddy pr esented eviden ce that the cli­
max concep t wa s ap plic able to strea ms .
The Silver Springs co m m uni ty is stro ng
evid ence that where th e hyd rographic
climate is co nsta n t th e co m m u nity m ay
de velop a stea dy sta te . Th e po ssibl e
relationship to th e balan ced a qua ri um
idea is di scu ssed elsew he re .
It is sug gested th at th e sta bility of a
system be m ea sured by the num ber o f
times it tu rn s over witho u t change.
Thus communities o f sma ll org a nism
which per sist for a yea r with a da ily
turn over may be co nsi dere d as sta ble in
thi s sense as a co mm un ity of larg e
organism s which last s 300 years wh ile
turning over o nce p er year.
Th e s um mer pul se of p rim ary pro­
duction due to th e gr eat er influx o f
light mainl y lead s to a n incr ea se in
reproduction in th e co nsume rs ra the r
th an to pul ses, blooms, a nd cha nges in
populati on s. T he increa ses in rates a t
an y on e trophic lev el are accompan ied
apparently by increa ses in utili zatio n a t
o the r tr ophic levels so th at the sta nd ing
crops do n ot change markedl y.. . .
-These figures ha ve been correc ted from the
o rigin al publ ica tion a t th e req uest o f th e
a utho r [Ed .],
NI TR ATE IN TH E SEA
H. W. Har vey-1926
Reprinted by pe rm ission of t he p ubl ishe r fro m Journal of th e M a rine
Biol o gical Asso cia tio n of t he United K ingd o m t4 : 71- 88, 1926. Publi shed
by Cambridge U niversi ty Press .
H. W. Harv ey , a pionee r ill ma rine biological chemi stry , describes here 1I0t
onl y th e inverse relationship betwee n the size of diat om populations and
nitrog en availability but also the regulation of en vironm ent by organisms,
In the nitrog en cycle, utili zable nit rogen is returned to the sys tem by organ­
ism s acting on the hound nit rogen in the excreta and ronains of other
organism s; the maio" fun ction 0/ the en vironme nt ill the p rocess is in physi­
cally distributing the requisi te materials .
T h us we have in the sea a closed cycle.
T Il E NIl ROGEN CYCLE
Denit rify ing bacteria have been foun d PROT EI NS. N IT RATE S
in wa ter near the sho re a nd in mud o f etc.. of Plant s- - - Dead By 1 fiy ni tr ate
eat en by and plants ba cterial forming
the Baltic, but as pointed o ut by G ra n forming rhe action bacteria
th ey are not likel y to pl ay a part in th e pro teins o f ·_ · A MMONI U M
econ omy o f the ope n ocea ns, since th e An imal s - - Dead SAn'S
wh ich give animals
wa ter a lmost inva ria bly co n tai ns a rise to -- -- Anima l
sufficiency o f oxy gen for th ese bacteria , excreta
with out their having recourse to attac k
the sma ll quantities of nitrates p resent. Phospha te is regenerated in a sim ilar
Azo tobacter, fixin g d issolved nitr o gen ma nner, a nd th e evid ence so far
has been found in th e slime of Baltic o b ta ined sho ws th at in both cases th e
a lga; a nd in bottom deposit s fro m near cycle is practicall y fI closed o ne, th e
th e land. There is no evid ence as ye t incr ement due to land wa shings not
that they are gen er al in o ccurre nce o r bein g ver y grea t compared with th e
that th ey add more th an a minute fr ac­ a mo unt co n tin uall y be ing regenerat ed
tion to the combined nit rogen in th e fro m dead marin e organism s.
sea . A d eficiency o f nutr ien t sa lts limits
As far as we k no w th e ino rga n ic sal ts both th e rat e o f m ulti plica tio n of
necessary for pl ant life a re a lways vegetab le pl ankt on . a nd in all proh a­
present in sea -wat er in a m ple a mo un t bil ity th e rat e o r car bo n ass imilatio n as
exce pt p hosph a tes a nd nitr at es. I ron is a well. A ll the ene rgy expend ed in the vit al
possibl e excepti on. T he nit rates are processes of plants an d a ni ma ls in the
convened by the al ga: int o prot ein s, sea is d er ived fr om t he en ergy o f light
etc. Some of th ese alga: die , a nd from a bso rbed during th e co u rse o f ca r bo n
th e de cay of th eir corpses a m mo n iu m ass imila tio n by ph ytopl ankton -the
salts a n: produced . O the rs nouri sh fixed a lga: bein g but a sma ll pr oporti on
marine a n imals whi ch in turn are fed o r th e pl ant life.
up on by o ther a nima ls. Th ese exc rete Hence a qu ant itative kn owledge o f
the products of their metabo lism a nd in th ose facto rs, which co n tro l and lim it
du e co urs e d ie ; amm o nium sa lts a re th e mu ltipl icati on o f vegetable pl ank ton
produced fro m th eir corpses a nd ex­ a nd which a ffect its e ffic iency as an
cr et a by ba cterial actio n. ene rgy a bso rber, is the first step in a ny
9l
192
fundamental inquiry concerning the
fertility of the sea.
The considera tion of this cycle indi­
cates that the fertility of a ny area of the
open oceans, no t su bjec t to considerable
in flows of wa ter from ot her areas,
depends up on three main fa ctors. (a)
The len gth of tim e protein formed by
ph ytoplan k ton rem ains pa rt of the
plant or nouris hed animal's body. (b)
Th e time wh ich ela pses during the decay
and formation o f ammonium salts and
p hosphate from corpses and excreta. To
this must be added the time ta ken for
nitrate forming bacter ia to convert the
ammonium into nitrat es. (c) T he time
whieh elapses before the reformed
nitra te and ph o sphate aga in reaches the
up pe r lay ers where th ere is sufficient
light for pho tosy nth esis.
Dea ling with th ese three factors in
turn, tbe first is cont rolled by t he
natural length of life of the plants and
animals, and by the proportion of the
plants which a re ea ten. This decides the
proportion of to ut! living matter t o the
rate of loss by natural death. The
proportion of total living ma tter to its
rate of los s by respiration and exc retion
is controlled by temp erature , In a warm
sea the los s will be in excess of that
taking pl ace under colder conditions,
since the ra te of metabo lism rou ghly
doubles for a rise of 10QC. , and m ore
food will be required by animals to
provide for their irreducible meta bolism
necessa ry to mainta in life.
Of the second factor (h) there is litt l
informati on. Presum ably the bre a k­
do wn prod ucts o f proteins and ph os­
pho p rotcins prod uced in the course of
metabo lism arc excreted mos tly as urea
and pho sph a tes, a nd the deca y of
corpses is largely brought about by
bacterial agency . F ro m the analyses of
water co llect ed in the A t la ntic d uring
the exp ed ilion of the Planet it a p pea rs
that then: is rar ely less th an 25 mil­
ligrams per cu bic me tre of a m mo ni um
nitr ogen in the wa ters of the open
ocean , T he d ist ributio n of nitrate form -
THE CO NCEPT OF TH E ECOSYSTEM
ing bacteria, possibly identical with
the nitroba cter of the land in the open
ocean is quite un known.
T he third factor (c), being the length
of time that the reformed nitrate lies
unu sahie below the illumin ated upper
layer s, is of grea t magnitude . . . . [there
is) a n eno rmo us store of nitrate belo w
the upper 100-metre layer. A rough
calcu lution indicates an amount some­
where in the order of 250 thou sand
million metric to ns of nitrate-nitrogen
in the dee per layers of the great oceans,
lying dorm ant until such time as
currents bring the deep water into the
upper sunlit layers. There is every reason
to suppose that tbi s third factor regu­
lates the speed at which nitrogen and
phosph orus pass through the complete
cycle in the sea as a whole, being the
slowest in th e series of changes.
(n particular shallow areas , however,
such as when: tidal stre ams and suf­
ficient surface coo ling to set up con­
vection currents reaching to the bottom
cause vertical circulation, during the
whole or part of the year , the regener­
ated nitrate is subject to no dela y
before again becoming available.
T hese areas, usually coastal , are very
fertile . The North Sea , English Channel,
and th e Shallow She lf between the
G ra nd Banks a nd Cape Cod arc su ch.
H ere the condition s are somewhat com­
plicated by a certain amount of inflow
of o ceanic water, and by the fact that
the ra te of regeneration of nitrate, and
ph osphate, from dead or ganisms over­
takes the rate at whi ch plant life utilises
it during th e autumn a nd short days of
winte r. A store of a vailable nitrat e and
pho sphate is thus formed which is
rapidly used up as so on as some three
hours of suns hine per day occur in the
spring in th e English Channel. Insuf­
ficient illumination becomes the limiting
factor during th e winter months, delay­
in g for a period the rate at which com­
bined nitrogen and phosphorus pa sses
th ro ugh the co m plete cycle .
In the deep open ocea ns the winter
O. ARRHENIUS 193
cooling of th e surface layers in temper­ Gulf Stream, etc .. to be re placed by
ated and arctic regions will set up wa ter up wellin g from below . He nce in
convection current s. By lessening the these re gions richer plankton may be
den sity gradient, th e convecti on cur­ expected than in th e subtropical. These
rents will be a ssisted by wa ve m otion in expectation s ar e born out by the
bringing about more effectual mixing results of He nsen's Plankton Expedi­
with the layers below. Hence in such tion .
latitudes , as oppo sed to the subtropical IT. is a remarkable fact that plan t
regions, a small store of nutrient sa lt s gr owth sho uld be abl e to strip seawater
ma y be expected in the sp ring, and in of both nitrates and phosphates. and
general a richer plankton pa rti cularly that in the Engli sh Channel the store of
in the early summer. In the t ro pical these nutrient salts formed during
regi on s of the Atlantic th e heated autumn and winter should be used up at
sur face water streams away to form the about the same time . . . .
HYDRO GEN IO N
AN D GRO WT H 0
O. Arl'henius-J')22
Reprinted by per m issio n of the publisher from A rk iv fUr Bot anik til:
1-54 ,1922.
There are two major ecological prin ciples developed i ll this paper: (l) the
prin ciple of interaction offa ctors, nam ely ofsubst ra te, clim at e and organism s
in soil formation and react ion, and (2) the princspl« of regu tation. namely
the regulat ory role of soil ill determining plant associations and the mo dif y­
ing influence 0/ plants VII the nature and properties 0/ soil.
We may therefore say , that in every selective ad sorption , but there rem a ined
so il there exists a certain concentrati on still so me very esse ntia l po ints un ex­
of hydrogen ion s, and hy measuring plained .
this, we also measure the actual acidity A lso with t he conceptions of the
of the soil. co lloid c hem ists it ha s been q uite
The question of the ori gin of this impossible to find a ny sa tisfactory
acidity has caused a vivid d isc ussio n . sol utio n of the qu est io n . But there arc
It bas been shown , that n ot only humu s man y facts spe a king fo r, t hat we must
but also sterilized cotto n a nd ot her ta ke a new po int of view in the so il­
substances quite free from a cids arc science. . . .
able to "absorb" alk al i. Th ere arc a lso A lso seve ra l a ut hor s ha ve shown tha t
pretty many acid min eral soil s kn o wn, so ils arc acti ng as bu ffers (a nd it is no t
where it is impossible t o ident ify a n y th e ino rga nic sa lts in the soils which
particular acid , but still th ey are giving ar t: the ma in facto rs) so that humus
an acid react ion. so ils a rc very str o ng buffers, clay s
T his has been explained by so ca lled medi um a nd sa nd wea k, but even the
194 THE CONCEPT OF T HE ECOSYSTEM O. ARRH ENIU S 195

buffer actio n of pure sa nd is greater out mu st then have go ne o n with a the so il react ion with aid of geolog ical fo rm cd. Tn o ther vegetatio n types th ere
than that of e. g. hydrochlo ric acid o r mu ch high er ra te th an no wad a ys, a nd map s, neither u sing th e rocks nor the was found less, down to 5 cm . Also o n
SORENSEN'S ph osph ate solutions. as thi s is not small it is easy to und er­ min er al so ils as indi cat or s. O ne o nly volcan oes wh ich rece ntly have been in
On the o the r hand clay beh a ves as stand ho w imm en se areas have gained very genera lly ca n state , that o ne rna v acti o n there were o bserved ra the r (hick
gelatine a nd thi s beh a viour is indepe n­ in sa lt co ntent a nd t h us a lka lini ty expect alk aline , ne utra l o r slightly acid humu s dep osits, up to 0.5 m ., whi ch
dent of orgin or origina l reacti on. wherea s ot hers a re so well washed out soils, where th ere is mu ch lim e in the m ust have been for med durin g th e last
It is very pro ba ble, tha t all sub stances th at th e soils mor e o r less ha ve co me to subsoil, whereas in siliceou s region s th c 20 - 30 year s.
behave in th e same manner, i.e., t ha t their isoe lectric point. soi ls may be more or less aci d. . . . As soo n as this humu s la yer is for med
th ey are real sa lt fo rme rs. We can thu s Throu gh thi s washi ng out pr ocess Th e mo st impo rt ant fact or in t he th e pla nt s a n: more o r less ind ep end ent
say that th e same laws hold for par ticles and lat er clima te cha nge s new soil co n­ for ma tio n of th e su iI react io n, how eve r, of unde rlyin g stra ta a nd th e proper ties
not visi ble in the ultr am icr oscop e up to d itio ns have been develop ed which may a re th e plants a nd the ir hum us fo rm ing of the humus play th e grea t role.
sto nes big as the fist or mor e. be loo ked at as pr imar y co ndi tio ns, act io n. Th e humus deposi ts also influence t he
These a mph olyte s give a certain when we onl y ta ke a shor t peri od in to F ro m lon g a go it has hcen wel l furt her dcvelo pern cnt of the soil profil e
hyd rogen io n concentra tion to the co nsidera tio n. kn own th at und er ce rta in assoc iatio ns throu gh their mo re or less acid ifying
solutio n du e to th eir very low di s­ T he process is goi ng o n so th at the o ne find s a specia l soil. T he best a nd influen ce on the percolati ng wa ter. . . .
socia tio n a nd thi s is the chief so urce of rain wat er load ed with car bo n dio xid e most close st udy o f these th ings is th e sa ps of differ en t plants ha ve widely
the hydrogen ion co nce nt ration o f the and acid ified by the humu s decomposes delivered by P. E. M UI. LI.:R in his book d ifferent aci d ity a nd als o plant debris of
soil. th e alkali- an d al ka li ea rt h silica tes and on the humus format ion in th e Da nish different plan ts have a reac tio n which
A fter what is said a bove o ne may sa y carries away more readi ly soluble salts forests, wh ere h e shows that th ere is a see ms LO lie ver y near Lo t ha t of the
th a t t he soi l is a system of a rnp holytes , of th ese eleme nts. In th e low la nd s correla tio n bet ween th e two types: pre-sa p from the livin g plant.
par tly dissociated , partly undi ssocia ted , where the wat er is sp rea ding o ver wide O xali s- Fa gu s-Wood a nd Trien ta lis­ Now it seems . . . as if th ose plan ts
water, air a nd small a mou nts of salts . .. a reas , e. g. as seepa ge in the soil o r by Fagus-Wood a nd the mild a nd raw for ming a certa in stab ilized ass ocia tio n,
Th e first loo se earth crus t was p rob­ irr igat ion , a nd th e eva po ra tio n is humus. But he and man y follo wing a nd th us also fo rm ing the humu s and
a bly nearl y neu tr al. Th is assu mptio n is st ro ng, th e so lutio n heeom es mor e a nd a ut ho rs sa w in t he soi l the pr imary, and the rea ct io n, als o ha ve the ir o ptimal
founded o n the fac t, tha t on one hand mor e co nce ntra ted and a lso the car bon the vegetation was looke d at as a gro wth a t th at reacti on , but in all
most volca nic ashes inves tiga ted a re di oxide of double ca rbo na tes is give n seco nda ry fac tor. unsta hilized assoc iatio ns th e reactio n
neu tral or nea rly neutral , o n th e o the r off by stillsta nding [sic], the salts th ere­ Som e yea rs ago the a ut ho r sho wed, fo rming powe r plays a great ro le in the
hand th at when digging into the soil fore ar e mor e or less dep osit ed and the that th e soil typ e, th e soi l reacti o n a nd co mpetition.
a nd pene trati ng deep er th an th e up­ soil mad e alka line . . .. the vegeta tio n type ar e depe nden t of Bu t also in sta bili zed associa tio ns
permo st region of great fluctua tions o ne It is o nly as lon g as th e rocks ne arly each o the r, so th a t the soil reacti on is there will occur cha nges. F or insta nce
mos tly reaches stra ta, which a re pr acti­ rea ch th e so il surface th at the y have a th e primar y fact or b ut in man y and the trees fall or fi re devasta tes t he fore st
cally neu tr al. gre a t influ en ce. A s soo n as co vered with esse ntia l ways influe nced by th e vege­ a nd then one gets a na t ura l dearing. It
It a lso seem s very pr ob abl e th at th e debris they lo ose th eir direct influ ence. tat ion . . . . has bee n shown by OLSEN th at o n such
ea rth ca me o ut of its sme lting stage in Als o th e min er al -soil looses a grea t deal T he process may be rega rded as wind clea rings thc reacti on may cha nge
a rather neu tr al a nd ho mogeneou s co n­ of its imp ortan ce as fo rme r of th e follo ws : The pion eer-plan ts mo ve in o n ra th er m uch, for instan ce fro m 5.2-6.6.
di tion. surface soil reacti on , whe n the pl ants the min eral soil, a nd with in a sho r t tim e A very grea t influenc e is ca used by
T he fac to rs influ encing th e d evelop­ have tak en th e gro un d . But indirectl y a rather thi ck humus layer is fo rmed. In man wh o so metimes incidentall y so me­
ment of a lka line and acid soils a re t he mineral so il a nd the rock s are northern co un tries as so uthern Swede n tim es co nsciously c hanges the soi l
chiefly the rela tion bet ween precipitat e always influencing the soil reacti on ther e ca n be forme d a layer of up to re acti on .
a nd evapora tio n, the vegeta tio n a nd the throu gh the soi l wa ter. .. . ha lf a met er or mo re in 100 yea rs . The By defor esta tion , thu s pr eventing a
rocks and their debris. Wit hin a very sho r t tim e, 50 yea rs or ra te of humus fo rm at io n is very different good deal of hum us form a tion , th e
If we assume tha t the surface of th e less, the gro und is co vered with a la yer in different p laces hut is at least o f the acidifying action o f th e so il is dec reased ,
soil fro m the beginn ing had a neut ral o f decomposing p la nt subs ta nces , mo uld sa me o rde r of degree in te mpera te and Th e cattle, whe n g razing, are also taki ng
reaction and th a t it is ac ti ng as a n or pea t, and then th is cove ring layer tr opi c parts of the worl d whe n th e away a n eno rm o us a mou nt of or ganic
a mpho lyte it is ea sily underst ood how will play th e grea test role fo r the soi l cli ma te is h umid . T he a uthor o bserved ma tt er wh ich o t herwise wou ld have
the differ ences, no w so g rea t, ha ve aci di ty as influenci ng t he vegeta tio n, by visi ting Verla ten Eila nd , one of th e tak en a part in the hu mu s forma tion . . .
developed. Dun ng ea rlier geo log ical beca use th e plan ts a lways first co me in island s in the Krakat oa-archipelago More co nscio us {sic] ha s a cha nge in
pe riods the climate was much mo re co ntact a nd gro w a grea ter p art o f their de str o yed by er up tio n in 1883, ho w in th e react io n been made fo r agri cultural
mo ist a nd rainy and the at mosphe re life in th is soil. th e Ca sua rina vegetati on a layer o f p urposes. T he first used meth od was
ric her in car bo nic dio xide T he washing ierefore it is impossible to pred ict humus o f ab out 10 em th ic kness wa s pro ba bly the dcns hiring of woods th us
 196 THE CON CEPT O F THE ECOSYSTEM ,.\ LFRED C. REDFI ELD 197

forming alkaline ash and clearing the been using lime, marie and clay to a ocean , Below thi s zone, life depends on
wood, both actions which change the great extent, trying to improve the soil organic matter carried down by organ­
soil reaction. Later on , when agriculture condition. . . . isms sinking from above or by the
was more developed the farmers have vertical migrations of animals back and
forth between the depths. Although the
greater part of the nutrient chemicals
absorbed in the autotrophic zone com­
plete the cycle: in this layer, the portion
which sinks as organic matter tends to
ICAL FACfORS IN deplete the surface layers of these
chemicals and, with the decomposition
of the organic matter in the depths, to
Alfred C. Redfield-1958
enrich this heterotrophic zone with the
Reprinted by perrrussion of the author and publisher from American products of decomposition.
Scientist 46: 205-221, 1958. The existence of the vast reservoir of
deep water in which organic matter may
Redfield's essay has as its central theme that tile nitrate in the sea and accumulate and decay out of reach of
the oxygen ill the atmosphere are controlled by tile biochemical cy cle. This autotrophic resynthesis is a distinctive
is environment regulated by organism . However, the biochemical cycle is feature of the oceanic environment
ultimately determined by the solubility of phosphate, This is organism which enables one to separate, in ob­
regulated by environment. Operationally, both types of regulatory mecha­ servation and thinking, the constructive
nisms function to promote stability . Dr. Redfield has been a staff member of and destructive phases of the biochemi­
Harvard University and the Woods Hole Oceanographic Inst itution for cal cycle....
over thirty five years.
CORRESPONDENCE BETWEEN
For the second alternative, it may be
said that the phytoplankton do have
some ability to vary their elementary
composition when one element or
another is deficient in the medium in
which they grow. Such physiology might
account for the coincidence in the
nitrogen-phosphorus ratios . However,
it is not evident how adaptation could
determine the oxygen relation since this
depends more on the quantity than the
quality of the organic matter formed,
and the oxygen requirement is fclt only
after the death of the living plant.
For these reasons the third alternative
deserves serious consideration . Mecha­
nisms should be examined by which
organic processes may have tended to
control the proportions of phosphorus,
nitrogen, and oxygen available for life
in the sea....
THE PHOSPHORUS-OXYGEN RATIO
. .. It is widely held among geoche­

mists that the primitive atmosphere was

REQUIREMEKT AND A VAILAIlJl.ITY

devoid of oxygen, or at least contained

OF PHOSPHORUS, NITROGEN,

very much less oxygen than at present.

It is a recognized principle of ecology activity of microscopic floating plants, AND OXYGEN

During the course of geological history

that the interactions of organisms and
environment arc reciprocal. The en­
vironment not only determines the
conditions under which life exists, but
the organisms influence the conditions
prevailing in their environment. . . .
The purpose of this essay is to discu ss
the relations between the statistical
proportions in which certain elements
enter into the biochemical cycle in the
sea, and their relative availability in the
water. These relations suggest not only
that the nitrate present in sea water and
the oxygen of the atmosphere ha ve been
produced in large part by organic
activity, but also that their quantities
are determined by the requirements of
the biochemical cycle ....
THE BIOCH EMICAL CYC I.E
The production of organic matter in
the sea is due to the photosynthetic
the phytoplankton. and is limited to the
surface layers where sufficient light is
available. The formation of organic
matter in the autotrophic zone require s
all the elements in protoplasm, of which
carbon, nitrogen, and phosphorus are
of particular concern. These are drawn
from the carbonate, nitrate, and pho­
sphate of the water. Following th e dea th
of the plants the or ganic matter is
destroyed, cit her by the metabolism of
animals 01' the action of microorgan­
isms . No rmally, decomposition is com­
pleted by oxida tion so that carbon,
nitr ogen , and phospho rus arc returned
to the sea water as car bo nate, nitrate,
and phosphate, while requisite quanti­
tie s of free oxygen are wit hdra wn from
the water.
The autotrophic zone has a d ept h of
200 meters at most and incl udes less
than five PC I' cent of the volume of t he
The stoichiometric relations . . . in­
dicate that phosphorus, nitrogen, and
oxygen arc available in ocean water in
very nearly the same proportion s as
those in which they enter the bio­
chemical cycle. In discussing the remark­
able coincidence in the supply and
demand for nitrogen and phosphorus it
has been pointed out that it might arise
from: (1) a coincidence dependent on
the accidents of geochemical history;
(2) adaptation on the part of the organ­
isms; or (3) organic processes which
tend in some way to control the pro­
portions of these elements in the water.
Of the first alternative not much can
be sai d excep t that the probability that
th e ratio in the sea be what it is
rather than any other is obviously
small. That the coincidence applies to
the oxygen as well as to the nutrient
elements compounds the improbability.
atmospheric oxygen is thought to have
been prod uced by the photochemical
dissociation of water in the upper
atmosphere and by the photosynthetic
reduction of carbon dioxide, previously
present in much greater quantities. .. .
Estimates of the quantity of red uced
carbo n present in the earth's crust as
coal and petroleum indicate that
photosynthetic processes have been
much more than adequate to produce
the present oxygen content of the
atmosphere . It has not been suggested ,
to my knowledge, why this process has
proceeded just so far as it has; that is,
why there is 21 per cent of oxygen in
the atmosphere at present, no more or
no less. It is, however, on this fact that
the quantity of oxyge n dissolved in the
sea depends.
My supposition is that the actual
quantities of oxygen present in the sea
1911
may have bee n reg ulated by the act i­
vities of sulfat e-red ucing bacteria. This
gro up of bacteria a re know n to ha ve t he
a bility to use sulfates as a so urce o f
oxyge n wh en free oxygen is absent a nd
organ ic co m po u nds arc presen t to
supp ly a so urce of energy . Th e o ver-all
reacti on is SO; - S" + 2 O~'
The pr ocess sho uld be br ok en do wn
into two steps, each of which takes
place in th e sea un der different e nviron ­
mental co nditions, (I) Sul fa te Re duc­
tion , SO ;" + 2 C ' 2 CO 2 + S· which
occ urs at dep th und er a nae ro bic co n­
dition s. a nd (2) P hotosy nt hesis 2 CO 2
- .2C -+ O 2 whic h occ urs near the
surface in th e pr esence of light.
In th ese eq ua tions C represents th e
reduced ca r bo n pr esent in o rga nic
mat t er. T he decomposition of this
materi al by sulfate -red uci ng bact er ia
accordi ng to t he first step also liber at es
a co rres po nd ing qu antity of nit ro gen
an d ph osph orus, which permit th e CO 2
fo rm ed to re-enter t he biological cycle
when the seco nd step co mes int o pla y.
T he C O 2 p ro d uced in t his way ca n thu s
co nt rib ute to th e prod uctio n of oxyge n
in a way in which t he excess ca r bo nate
normally p resent in sea wat er ca nnot.
Th e first ste p, whic h de pend s up on
th e presen ce of o rga nic matt er in excess
of the free oxygen req uired to co mp lete
its decom posit ion , will init iate a mech a­
nism which will ten d to increase th e
oxyge n whe n, a nd only whe n. th e
qu antity of ava ila ble free oxygen is
deficient. If the total mechan ism has
opera ted on a lar ge e no ug h sca le in the
co urse of geoche mical hist or y, it m ay
have k ept the supp ly of oxygen av a ila ble
in th e sea adj ust ed to th e requirements
o f the b iochemi cal cycle .
T here is very good evide nce th at
sulfa te red uct ion does opera te on a
lar ge sca le in th e sea whe rever a na ero bic
co nd itio ns exist. . . .
GEOC HEMICAL CONS IDERATIONS
As a final check on th ese specu la tio ns,
we ca n look at the relative ava ila bility
THE CONCEPT OF r un ECOSYSTE M
of the princip al mat er ial s o f th e bio­
chem ica l cycle o n t he earth' s su rface,
to sec if they confo rm t o t he postulates ...
. . . the number s of a to ms relati ve to
the ato ms of ph osph oru s in t he ocea n. ..
are indicat ed in F ig ure 3 which presen ts
in dia gr amm at ic form th e bioc he mica l
cycle as de scribed .
The dia gram sho ws ph osph at e,
nitrate, a nd ca rbo na te ente ri ng th e
o rga nic ph ase o f th e bioc he mica l cycle
ncar the sea su rface, th rou gh th e pr o­
cess of ph ot osynth esis. Ph osph orus,
nitr ogen , and carbon a re selec ted by t he
sy nthe tic pr ocess in th e proporti on s of
1 :15 : I05. T his is the ste p whieh coordi­
nate s the cycles of th e seve ra l ele men ts
in a unique way a nd gives meaning to
th e compari son s. Th e elements a re
carried in these prop orti on s t o th e
p oint of de composit ion where th ey a re
oxidized t o th eir o rigina l sta te as
ph osphate, nitrat e, a nd ca rbo nate. T he
o xygen requir ed is just tha t set free by
ph ot osynthesis. Such a cycle co u ld run
ind efinit ely in a n ot he rwise closed
system so lon g as light is supp lied .
To acco unt for th e co r respo nde nce in
the rati os of ph osphoru s and nitr ogen
in the orga nic ph ase o f th e cyc le a nd in
th e inorgani c en vironmen t , bacteria l
processes of nitrogen fixati on and
de nitrification a re indica ted at th e upper
r ight a nd, similarly, t he sulfa te redu c­
ti on pr ocess is show n a t the lo wer left.
Thi s latter is ass umed t o o pe ra te
effective ly only whe n th e envi ro nme nt
beco mes a naero bic. Fina lly, th e ex­
cha nges with th e a tmosp here a nd th e
sed imen ts of the sea bottom a re show n .
If the se processes a rc operative it is
necessar y that supp lies a re adeq uate a nd
th at the ir prod ucts exist in su ita ble
qu anti ties.
Con siderin g first nitrogen , th ere exist
in sea wate r for each a to m of ph ospho­
ru s 15 ato ms of nitrogen avai la ble as
NO , a nd a reser ve of 510 atoms of
n itrog en as dissol ved N, which 111 ay be
drawn on by nit ro gen- fixin g bact eria . In
a ddition, there is a rese rve of nit ro gen
A LFRED c . nr orun D
o x YGErJ
2 30 0 0
~J
PHOTOSYNTHESISX
SUL FATE REDUCT I O ~
S0 4- S +? O2
\
~ SU LFUR
10 00 0
- ~-~----~-----
PHOSPHOR US
40000
r io . 3. The Biochemi cal Cycle. NUll/has repre­
sent quant ities 0/ respective elem ents p resent in
the a tmosphere, the ocea n, and the sedimentary
rocks, rela tive 10 the num ber ofato ms 0/ phos­
phorus in the ocean .
in the at mosphe re equ ivale nt to R6.000
ato ms of phosph oru s, whic h is ava ila ble
t o replace that dissolved in th e sea
were it to be dra wn on. T he nitrogen of
th e sedime nta ry ro cks is a bout o ne­
sixt h t ha t in th e atmos phere a nd
t went y t im es t hat in t he ocea n. M ore
th an four-fifth s of thi s is fossil nitr ogen
whic h ma y be ass umed to be derived
fro m o rga nic matter. Con sequent ly,
larg e qu antities of nitrogen ha ve passed
through the bio chemical cycle in its
passage from the a tmos phe ric reser ves
t o be deposited in sed iments at th e sea
bottom . The qu antity with dr awn in
thi s way is sma ll, ho wever , ill co mpari­
so n 10 the reserv e in th e at mos phere.
Clearl y, th e nitrog en supp ly is a deq ua te .
Sulfate is o ne of th e mo st a bunda nt
io ns in sea water. Tn th is form there is
pr esen t s uTfur eq uivale n t to 10.000
at om s of ph osph oru s. It wo uld be
199
CARBON DIOXI DE Nl r ROGEN
16 620 00
J!~~~
NH
] ~ N2
5 10
\ '\ NITROG EN FIXATION
p0 4 NO] CO ]
I 15 10 0 0
CA RBON
40 0 0 0 0
NI TROGE N COAL a OIL 160 000
10 0 0 0
ca pab le o f supplying oxygen equ ivalent
to 40 .000 ato ms of ph osph oru s o n
redu cti on . Clearl y, the su lfate reduction
mech an ism co uld co ntinue to ope ra te
for a lo ng tim e. If it ba s o pera ted as
p ostu lat ed in the past m uch su lfide ma y
ha ve been rem o ved from th e sea.
Sedime ntary rock s a re estim at ed to
co nta in sulfu r equivalent to 10,000
a to ms o f oceani c ph osph oru s. If thi s
were a ll th e p rod uct o f sulfa te redu ct ion ,
t his wo uld ha ve produced oxyge n eq ui­
va lent t o 40,000 atoms of ocea nic
phosphoru s, which is a lmos t t wice th at
present in th e at mos phe re. It is not
clea r h o w mu ch of the sulfur in sedi­
mentar y rocks is present as sulfides, bu t
m uch of it is. Clearl y, m uch oxygen ca n
ha ve been produced in th e pa st by
sulfa te redu ct ion a n d p ossibl y thi s
pr ocess has co nt ribu t ed to an imp ort ant
degree in pr oducin g th e oxyg en of th e
at mos phe re.
Ca rbon is pre sent in th e sea, chiefly
as cu r ho na tc ion s, in ahout ten tim es th e
qua ntity requi red for the bio chemi cal
cyc le. Mu ch or th e large deposit s of
carbo n in th e sedimenta ry rock s is
present as ca r bo na tes and ca nnot ha ve
200 THE CONCEPT OF THE ECOSYSTEM ALDO LEOPOLD 201

contributed to the production of free may be concluded that the quantity of All land rep resents a downhill flow of internal economy of each , balances the
oxygen. The estimated carbon present nitrate in the sea, and t he partial nutrients from the hills to the sea. one against the other.
as coal and petroleum, equivalent to pressure of oxygen in the at mosp here This flow has a rolling motion. Plants The food balance between soils and
160,000 atoms of oceanic phosphorus , is are determined through the require­ and a nimals suck nutrients out of the waters is accomplished not only by
sufficient to yield oxygen on reduction ments of the biochemical cycle, by the soil an d ai r and pump them upward circuitous routes of flow, but also by
equivalent to 320,000 atoms of oceanic solubility of phosphate in the ocean . thr o ugh the food cha ins; the gravity of eddies and back-currents. That is to say,
phosphorus, which is more than ten This is a physical pr operty of a unique death spills them back into th e soil and some animals pump food back uphill.
times the present content of the atmos­ chemical compound and as such is not air. Mi nera l nutrients. between their These local reversals of the downhill
phere. su bject to change except in so far as successive trips through this circuit, flow helve not, to my knowledge, been
The known facts of geochemistry do alterations in conditions may influence tend to be washed downhill. Lakes described or measured . They may be
not appear to contrad ict the sup­ the activity coefficients of the ions retard this downh ill wash, and so do important to science, and to land health ,
positions pre sented on the mechanism involved. It follows then that the soils . Without the impo unding action or conservation.
which may have controlled the relative nutrient supplies in sea water, and the of soils and lakes, plants and animals
availability of phosphate, nitrate, and oxygen content of the atmosphere have MOVEMENTS FROM WATER TO
would have to follow their salts to the
oxygen in the sea . Sources of nitrogen been about as at present for a long time coast line . LAND A"D FROM LAND TO
in the past and will remain at much the WATER
and sulfate are available in great excess The rate of retardation depends, for
and the by-products o f the reactions same level into the future. This argu­ one thing, on the length and the ter­ Bacr-currenrs are likely to be clearly
can be adequately accounted for. Ac­ ment may then be added to those mini of the food chains. A nutrient salt visible in areas inhabited by some
cording to these suppositions, phospho­ reviewed by Rubey that the composition impounded in an oak may take a animal requiring a larger supply of a
HIS is the master element which con­ of sea water and atmosphere has varied century to pass through an acorn, a particular nutrient than the soil sup­
trols the availability of the ot hers .. .. surp risingly little at least since early sq uirrel, a redtail, and parasite before plies . The red deer on the Scottish
If the argument presented is sound it geologic time .. .. it re-enters the soil for another upward highlands is a case in point. Here
roll. The same particle may take only nutrients are scarce because the soil is
a year to pass through a corn plant and derived from sterile rocks. The red
a fieldmouse to the soil. Again it may deer's yearly production of new antlers
pass through a grass, a cow , a pig, and a calls for more calcium and phosphorus
member of the Townsend Club, emerg­ than his highland range can supply.
LAKES IN RELATION TO TERRESTRIAL LIFE P AlTERN S ing not into the soil, but into a sewer Where and how does he get them '!
and thence into a lake. Civilization Fraser Darling records the facts as a
Aldo Leopold-1941 shortens food chains, and routes them calcium-phosphorus food chain. The
into lakes and rivers instead of fields deer gets a little, but not enough, of
Reprinted by perrmss ion of the copyright owners, the Regents of the and pastures . these horn-building salts from the
University of Wisconsin, from A symposium on hydrobiology, Ma d iso n , The rate of retardation depends also native herbs and grasses. His supply
The University of Wisconsin Press, pp. 17-22,19-11.
on the fertility of soils . Fertile soils increases when fires concentrate in
This essay by an outstanding conservationist not only reflects his attitude wash slowly . They support long chains ashes the dilute supplies stored in the
regarding man's intervention in "our biotic constitution,' but also focuses if we let them do so. Food circuits are heather. Hence gamekeepers practice
on another aspect of organism-environment regulation . The issue here is intricately adjusted to maintain normal rotation burning on the moors. To
the role of organisms ill the dynamics of energy excltange between two rates of retardation. A normal soil make good his deficiency in horn­
major environments as one of many interrelations between them, balances its intake from the decompo­ building salts, the deer taps the aquatic
sition of rocks against its loss from food chain of the lakes and tarns, where
downhill wash. We now know, to our rich supplies are concentrated in a­
cost, the disturbing effects of too Iowa quatic animals. He eats frogs immobili­
acknowledge this interaction between rate : erosion. A normal water balances zed by frost. As his relative, the
FOOD CIRCUITS IN SOIL AND
water and land after erosion or pol­ its intake from the soil against its reindeer, is reported to do, he may eat
WATER
lution makes them sick , but we lack a outwash to the sea. Pollution is an duck eggs and dead fish, By acquiring
Soil and water are not two organic "language" for describing their normal excess of intake arising from erosion, or such unusual or "depraved" food habits,
systems, bu t one. Both are organs of a interactions . Such a language must deal , from routing land wastes to water. the deer requisitions from lakes what
single landscape; a derangement in for one thing, with their exchanges of Underfed soils thus mean overfed his terrestrial range fails to provide.
either affects the health of both. We nutrient materials. waters. Healthy land, by balancing the That the stored salts are what the red

J ~~.7 .
de er is a fte r is sho wn by ce rta in o t he r
extrao rdi na ry food habits which he lp
bal ance his ca lci um- phos p ho rus eco no ­
m y. H e eats t he ve lvet fr om hi s o wn
h orns ; the bones of d ea d deer left by
hunters ; hi s o wn ca st a n tle rs or th o se o f
o t he r deer ; th e rabbit or vo le which ha s
ex tracte d sa lts fro m these sa m e ma­
terial s a nd the n cha nce d to die. Direct
ingesti on of lemmin gs a nd mi ce b y wild
reindeer has been recorded a nd m ay
a lso be pr acti c ed by the red deer.
T he ing est io n of aq uatic a nim a ls by
deer is a n uph ill mo vem ent o f nu trient s ;
a back -curren t o f the d ow n hi ll strea m.
F o od which ha s a lready " p ass ed" the
te rres t ria l deer bu t lies temporarily
impounded in lak es is puJled ha ck into
the terrestrial circ uit. O n the o t her hand ,
the in gesti on by deer of velv et, horns,
b on es, a nd d ead rod en ts is not a bac k­
current, but ra th er a sho rt circ uit in the
u su al r oll of t he fo od cha in. T he sa lts
contained in t hese bod y pa n s would
normall y re-enter the so il a nd bec ome
(i n part) ava ila ble to th e d eer as p la nts ,
but by sho rt ing thi s no rm al circ uit he
re co ver s t hem in less l im e a nd wit h le ss
wa ste.
The qu antity of min eral s in vol ved in
t hese m ovem ents is sm a ll, but eve n
small qu an tities may , o n p o or so ils, be
of critica l im po r ta nc e . Ran ge m anag ers
n ow rea lize that t he co n tin ue d " de po r­
tation " of ph osphoru s a nd ca lci um in
th e bones of ca tt le a nd shee p m ay
eve n t ua lly impoveri sh gra zi ng ran ges.
D arlin g hin ts t ha t dep ortat ion in sheep
ma y have help ed to impoveris h t he
Sc o tti sh mo ors.
M any a nimals o the r th an red de er ta p
a q ua tic foo d c ha ins and restore f oo d to
terrestrial circ uits . M any a lso m ov e
foo d in the o p posite di recti on . T he net
re tard ation, o r prep onderanc e of uphill
tra nspo rt , varies fro m zero up ward.
Th e length of uphill tr an sp o rt a lso
varies fro m sho r t to long di stan ces.
T h us river du ck s, geese, g ulls, terns,
rails, bitterns, fr ogs, snakes, and mu sk-
THE CONC EPT Of THE ECOSYSTEM
ra ts eat in o r a t t he edge of wate r a nd
d ie o r de feca te inla nd , b ut t hey lik e wise
eat inl and a nd di e or d efeca te in wa te r.
T here is no cle ar prepo nderance o f
uph ill transport. Th e first t hree ra nge
far inla nd , th e o t hers not fa r. Eagles,
cro ws, swallows, bears, d eer, ca ri bo u ,
a nd mo ose ca rry food bo th to a nd from
water, but th ey probabl y mov e m ore
fo od uph ill t ha n downhill, a nd to a
co nsi de ra ble d ist an ce inla nd . R iver­
spa w ning sa lmo n whic h di e in la nd
perfo rm a large a nd long uphill t ra ns­
p ort. Gu ano birds, penguin s, he ron s,
otte rs , min ks, sk un ks, ba ts, a nd cer tai n
wa ter- ha tchi ng , la nd -dy ing i nsects per­
f orm a p rep ondcra ne e of uphill t ra ns­
p ort, but o nly to a sh o rt di st an ce
inl and .
Probably no o the r fo od cha in con ­
centrates so much foo d o n so sma ll a n
area a s th at e ndi ng in gu ano birds. T he
wh ole a q uatic ga rd e n o f t he so uth
Paci fic s hip s it s prod uce, via th e up wel­
lin g Humbo ldt cu rre nt , to th e co asta l
g ua nays, whi ch dep osit it o n . th eir
ra inl ess island rooke ries as g ua no .
Here then is a bo ttl en eck wh ere t he
o cea nic fo od circui t a c hieves a " vo ltage"
o f e xt ra o rd ina ry int en sit y. T he g ua no
deposits, how ev er , lie so ncar t he sho re
and in so dry a clim a te t hat until they
a re moved fur th er inl and by m an th ey
ha ve little effe ct o n terrestrial circ ui ts.
Antarctic p en gu in s lik ewise ca rry
ocea nic food s inl and , b ut thei r dep osit s
a re refrigerated a nd evc nt un lIy slid e
bac k in to the se a . AS again st t he lon g
list of higher a nimals whic h tr an sport
food in a preval ently u phill di rect ion ,
I can think o f o nly two , m an a nd th e
beaver, which get most o f th eir fo od o n
land and deposi t m ost o f it in t he wa te r.
Marsh-roostin g bl ackbirds a lso d o thi s,
bu t only in autumn .
M ost a nima ls merel y circ ul a te food
wit hin the te rrestrial o r a q ua tic ci rc uit
whic h is the ir h abita t. Thu s th e d iving
ducks, except whe n ca ug ht by so me land
preda tor, feed fro m a nd d ie in to the
ALIJO LEOPOLD 203
a q ua tic circ uits. G allin aceou s bird s, d o wnhill flo w, reduce th is defi ci t. It is
furt he r redu ced by sto rage in so ils an d
exc e p t wh en sho t by a hun ter wit h
la kes. T he co ntin uity a nd sta bility o r
modern co nve nie nce s, feed from a nd
inl and co m m unit ies probably dep end
d ie int o the terrestrial circ uits.
on th is reta rd ation a nd sto rage .
T hese m o vement s of foo d su bsta nces
I . O ~; G-Di S rA N C E
see m to co nst it ute , co llectivel y, the
T RA NS I' I.A NTA l IONS
nu tri ti on al syste m of the b io tic o rga n­
M igra tor y birds a nd fish mo ve food
ism. It ma y be sur m ised, b y analo gy
to a di stance fr om its p oint o f o rigin . wi th indivi dua l pl ant s a nd anim al s,
U ntil man began to s hip foo d s an d
tha t it ha s qu a lita tive a s well a s quanti­
fertilizer, th e o nly long-di st an ce m o ver s
ta tive a spects. The recent hist ory of
we re water, air, a nd mi gr ato ry a n im a ls.
hi ol o gy is largel y a di scl osure of the
M igra tor y b ird s must mo ve a cou sid era ­
im port an ce of q ualitati ve nu tri ti on
hie vo lu me of fo od , with more th an
w it hin pl ant s a nd a n ima ls, a nd with in
tr an sit ory eITect s. Thu s H a wkin s point s
lan d and wat er co m m u n it ies . I s it a lso
o ut th at th e pl ant co mmu nity und er
im portant a s bet ween land and wat er ?
pa ssen ger pigeon roo sts was di stin­
D oes th e wild goose, reconnoiterin g the
guisha ble for d ecad es aft er th e p ige o ns
farm er's co rnfield, brin g so me t hi ng
were gon e. more th a n wild mu si c fro m the la ke,
T ra nsp la ntatio ns by migr a tory a ni­
ta ke so met hing more than wa ste corn
m al s have no dear o rie nta tion uphill
from his field '!
o r d o wnhill. Su ch qu esti on s are , fo r the m om ent,
beyon d th e boundari es o f precise k now l­
SUMMA R Y AND DIS C USSION edge, but not beyond th e boundari es o f
in tell igent sp ecu la tio n. We ca n at least
Soil health and water health a re no t
fo resee th a t th e prevalent mutilation s of
t wo p ro ble m s, but o ne. T he re is a
so il a nd wate r system s, and wh ole sal e
circ ulato ry syste m of food su bs ta nces
si mp lifica tio n of native faun a s a nd
co m m o n to both, as well as a c irc ula­
flo ra s , m ay ha ve unpredi ct abl e rep e r­
tory sys te m wit hin ea ch . T he d ownhill
c uss io ns. Ne it he r agric ult uris ts nor
now is carried by grav ity , t he u phill
a q uicult urists have so far s ho wn a ny
now by a nim a ls.
co nsc io usness o f this possibility. A
There is a d eficit in uphill tr ans port ,
prud ent tech nol ogy s ho uld alter th e
which is met by th e de compo siti on o f
natural o rder as littl e a s p o ssible .
rock s. L ong food c ha ins , by retard in g
 G . EVELYN HUTCHIN SON 205
HOMAGE TO SANTA ROSALIA OR '''H Y ARE Tl mRE SO
ANY KINOS OF ANI1\IA I....s r
munity. If we assume t hat in ge nera l 20 of th e m arine fa una, is cle a rly due
per cent o f the energy p a ssing t hro ug h la rge ly to th e diversit y p ro vided by
G. Eve ~v1/ Hutchinson- 1959 o ne lin k ca n enter the ne xt link in the te rres tr ial pl ants. Th is diver sity is
Re nrinted by perm ission o f the a uthor and pub lishe r from T he Am erica n chain , which is o vergen erou s (Sl obod­ a ctu ally two -Co ld . fir st ly, si nce ter­
Natu ralist 93 : 145-159, 1959. kin in an unpubl ished study fin ds 13 restrial p la nts co m pe te for ligh t. th e y
per ce nt a s a re a son a ble upper limit) a nd hav e te nd ed to evo lve in to str uc t u res
A lthough he may be best recognized f or his contributions to biogeochem istry , if we sup p o se th at each predat or has g row ing into a gas eo us m edium of
Hut chinson's public ations in limnulogy and population ecology have been no tw ice the ma ss, (or 1.26 the linear negli gibl e bu o yan cy. Thi s ha s led to th e
less significa nt. In this essay , not withstanding its chance homage to a saint dimen sion s) o f its prey , whi ch is a very form ati on o f spec ia lize d su pport ing ,
of unknown history p resumed f or purposes of the essay to be the patroness of low estim a te of the s ize d ifference ph ot osynt h et ic, and reproducti ve st r uc­
evolutionary studies, Hutchinson exp lores the diversity of species as a between link s, the fifth a n im a l link will tures w h ic h inevi ia bl y d iller in c he m ica l
fun ction of the complex ities of trophic organi zation and niche diversification have a p opulation of o nc tcn th ou­ a nd ph ysical pro p erti es.. . . A m ajor
in ach ie ving stability . Pr ofessor Hntchinson's successio n of produ ctive sa nd t h (10- I) o f the f rst , and th e so urc e of terrestrial di ver si ty was thu s
students from his " itinerant ivory tower" at Yale and his significan t contri­ fiftieth animal li nk . if t he re wa s o ne , a introduce d hy the e vo luti on o f a lmos t
but ions to ecology led to his being recognized as Eminent Ecologist by the p opulati on of 10-l u the size of the first . 200,000 s pe cies o f flowering pl ants, an d
E cologi cal So ciety of A merica in 1962. Five animal links are certainly p ossibl e , the three qu arters of a m illio n in sect s
a few fa irl y clear c ut ca ses ha vi ng been su pposed ly kn own toda y a re in p art a
in fact re corded. If. however , we produ ct of th at diversit y. But o f it sel f
wanted 50 link s, sta rting with a proto­ me rel y pr o vidin g five or ten kind s of
zoa n or rotifer feedin g o n al gae with a food of di ff e re nt co ns iste nc ies a n d
There are at the presen t time sup­ FOOD CHA INS d en sity o f 101' ce lls per ml , we sho uld co mpositio ns d o es n ot ge t LIS much
p osed to be a bout o ne milli on desc ribed ne ed a vo l ume o f 10' " cu bic ki lo me te rs furthe r th an th e live or ten link s of a n
species o f a nimals . Of th ese a bo u t A n ima l ecol o gi sts frequ entl y thin k in to ac comm od at e on a n av era ge one E lto n ia n pyra m id . O n th e w ho le th e
three-qua rters are i nse c ts, o f whi ch a terms of food c ha ins , o f t he form spec im en of th e ultimat e pred at or, and pro ble m sti ll re ma in s. but in the new
quite d isproporti on at ely large num ber individuals of species S , are eaten by thi s is vas t ly g rea te r than th e volume fo rm : wh y are there so m an y kind s o f
a re m embers of a si ng le o rder, th e those 0/ S 2< 0/ S 2 by S 3, 0/ S :j by S " etc . o f the wo rld ocean . C lea rly th e Eltonia n pl a nt s ? A s a zoo logist I do n o t want to
Col eoptera .' T he m arin e fa una a lt ho ug h In such a foo d chai n S, will o rd ina ri ly foo d -c ha in o f itse lf ca n no t give an y at tac k th a t qu esti on d i rec tly, I wa nt to
it has at it s d isp o sal a much grea ter be so me hol ophyli c o rga n is m o r m a­ grea t di versit y, a nd th e sa me is a lmos t sti ck wit h a nimals, b ut a lso to get th e
a rea th an ha s th e terrestri al , la ck s th is terial deri ved fro m s uc h orga nis ms . The c erta in ly tru e o f th e o t he r t ypes o f fo od a nswe r. S ince, ho wever, th e pl ants a re
a st oni sh ing di versity . If the insec ts are si m ples t case is t ha t in w hic h we have a chai n, ba sed o n detr itu s feedin g o r o n p art o f the ge ne ra l syste m o r co m ­
excl uded , it wo uld see m to be more tru e predat or chain in Odurri's CO n­ para siti sm .... munitics, a ny s ufficie ntly a b st ract p ro­
di verse. The proper a nswe r to my veni ent termin ol og y, in which th e lo west Effect oj" size . A sec ond important p erti es o f suc h co m m u nitie s a re likel y
in iti al questi on [why th ere are s uch a n link is a gree n p lant , th e next a her­ limitat ion o f th e len gth of a rood c hain to be rel evant to p la n ts as we ll as to
en orm ou s number of anim al specie s] b ivorou s a ni ma l, th e next a primary is du e to th e fact th at ord inaril y animal s h er bi vores a nd ca r nivo res. . . .
wo uld be to d evelop a th e ory a t lea st ca rnivo re , t he next a seco nda ry ca rn i­ cha nge th eir size dur ing free life. If th e
predictin g a n o rde r o f m a gnitude for vo re, etc. A specia lly imp ortant typ e of terminal member of a cha in were a fish JC':lER R EL AT!O NS OF FOO D CH AINS

th e number of speeics o f 10 6 rathe r th an
1O ~ o r 10 4 • T his r ce rta in ly cannot d o .
At m ost it is m er el y p ossible to p oin t
o ut so me o f th e fac to rs whi ch wo uld
h ave to b e co ns ide re d if suc h a th eory
was ev er to be co nstruc te d .. . .
"Ther e is a sto ry, po ssibly a pocry pha l, of the
distin guish ed British biologist, J. B. S. Haldane,
wh o fou nd himself in the company o f a grou p
o f th eologians . On bei ng aske d wha t o ne cou ld
co nc lude as to th e nat ure o f the Cre a to r from
a study of h is crea tio n, H aldane is said to ha ve
a nswe red , "An inordinate fo ndness for beetles ."
204
p red at or c hain m a y be design at ed
Elto nia n , bec ause in recent yea rs C. S.
E lto n has emphasized its wid esp read
sign ifica nce, in wh ich th e p re da tor a t
e ach level is larger a nd rare r th an its
prey. T his p he no me no n wa s re cognized
much ea rlie r, not ably by A. R. W all a ce
in his co n tributiori to the 1858 co m ­
mu nication to th e L inne an Societ y of
L ond on.
In s uc h a sys te m we ca n m ake a
th eoret ical gu ess o f th e o rde r o f
ma gnitude of th e di ver sit y th at a single
food ch a in ca n in trodu ce into a com­
th at g re w fr om sa y o ne em to 150 ems
in the co urse or a n o rd ina ry life . this
siz e cha nge wo uld se t a limit by co rnpc ­
tit io n t o th e p ossibl e number or o t her­
wi se co nce iva b le link s in th e 1- 150 ern
range. At lea st i n fish es thi s t ype o f
process (rne ta p hoe tes is) ma y involve
the smalle r spec ime ns bel on gin g to
link s bel ow th e larger a nd th e c ha in
length is thu s len gth ened , th ou gh under
st ro ng limitati on s, by ca nniba lis m . .. .
Effects 0/ terr es tr ial plant s. The ex­
tra ordin ar y di versit y of the terrestrial
fauna, whi ch is much g rea te r tha n t hat
Biol o g ical co mm unities d o n o t co n­
sist o f in depend ent fo od chai n s, but of
tood webs , of s uc h a kin d th at an indi ­
vid ual at a ny leve l (co r respo nd ing to a
link in a s ingle chai n) ca n u se so me h ut
not al l o f th e foo d p rov ide d by spec ies
in t he level s bel ow it. . . .
M acA rthur co ncl udes th a t in th e
evo lut io n o f a natural co m m u n ity two
p artl y a ntago nistic p ro cesses a rc occu r­
rin g. M ore efficient species wi ll repl a ce
less e fficient species, b ut mo re sta ble
co m m un it ies will o u tlast less sta ble
com m unities. I n th e process 0 f co m­
206
m unity fo rma tion, the ent ry of a new
species may invo lve o ne of t hree p o s­
sib ilities. It ma y co mp letely d isp lace
a n old sp ecies . T his o f itse lf does not
necessaril y c ha nge t he sta bility, th ou gh
it ma y do so if the new species inh erentl y
has a more sta ble p opula tio n t ha n the
o ld . Secon dly, it ma y occu py a n u nfilled
ni ch e, wh ich m ay, by p ro viding new
pa rtially in dep en den t link s, inc rease
stability . Thirdl y, it may partiti on a
nich e wit h a pr e-exist ing spec ies. E lto n
in a fasc inating wo rk la rgely devoted to
t he fa te of species accidenta lly or
purposefull y introd uce d by man , co n­
cl udes t hat in very d iverse co mm unit ies
suc h in tr oduction s a re difficult . Ea rly in
th e h ist ory of a co m m unity we may
sup pose m an y nich es will be empty a nd
in vasion will proceed easily; as the
co m m u nity becomes m or e d iversified ,
th e p rocess will be pro gr essively m ore
d ifficult. Som etimes an ex tre mely suc ­
ces sful invad er ma y o ust a species but
a d d littl e o r n othing to sta bility, at
o t her times the in vader b y so rnc specia l­
izati on will be a ble to co m pete suc­
cess fu lly for th e margina l p art s of a
nich e. In a ll cases it is proba ble tha t
inva sio n is mo st likely wh en one o r m or e
species happen t o be fluctu ating an d are
underrepresented at a g iven momen t.
A s th e co m m unities build up , th ese
op po rt u nities will ge t p ro gressive ly
rarer. In thi s way a co mp lex co m m unity
co nt aining so me highl y specialize d
spec ies is co nstructed asy mptot ica lly .
M od ern eco log ica l th eory t he re fore
a ppears t o a n swe r ou r initi al qu estion
at least pa rt ially by say ing th a t th ere is
a great diver sity of o rga nisms beca use
communiti es of man y d iver sifie d o r­
gan ism s a re bett er a ble to per sist th an
a re co m m uni ties o f fewer less diver sified
o rga nisms . Eve n th ough th e ent ry of a n
in vader whic h ta kes over part o f a
n ich e will lea d to the redu ctio n in th e
average po p ulat ion of the species o rigi­
n all y pr esent , it will a lso lead to a n
increa se in sta bility redu cin g th e risk
THE CONC EPT OF TH E ECOSYSTEM
of th e or igina l popul a tion bein g at times
u nderre presented to a da ngero us degree
In th is way lo ss of so me nich e spa ce
may be co m pe nsa ted b y redu ct ion in
t he a mp lit ude o f fluctu ati on s in a wa y
tha t ca n be ad va ntageo us to bo th
species. T he process h ow ever ap pears
lik ely to be asy m ptotic a nd we ha ve now
to consid er wh at se ts the asy mp to te, or
in simp ler words w hy a re t here not m or e
diffe rent kind s o f a ni ma ls ?
LIMITATI ON OF D IVERSITY
It is first o bvio us th at t he p ro cesses of
evo lu tio n of co mm uni ties must be
und er vario us sorts of ex te rna l co n tro l,
a nd th a t in so me cas es s uch contro l
limi ts th e possible di ver sity. Severa l
inv estigators, n o ta bly Od um and M ac­
Arth u r, ha ve po int ed o ut t hat t he mor e
or less cycl ica l oscillatio ns o bserved in
a rctic a nd boreal fa un a may be d ue in
part to th e com m u nities no t bein g
s ufficie ntly comp lex to damp o ut osc il­
lati ons. It is certa in th a t th e fau na of
any suc h region is qu a litat ively po orer
t han that o f wa rm tem pe rate a n d
tr opical areas of co mp ara ble effec tive
p recipitation. . . . It is reasonable to
sup pose th a t th e t otal biomass may be
invo lve d. If t he fund amen t al pro d uct i­
vity o f a n area is lim ited by a sho rt
gro wing sea son to such a d eg ree that
th e total biom a ss is less tha n un der
m ore favorable cond it io ns, then the
rarer species in a com mu n ity m ay be so
ra re th at th ey do not exi st. It is a lso
p robabl e th at cer tai n absolute limita­
tions o n growt h-forms of p lants, s uc h
as th o se th a t m ake t he development of
for est imp oss ib le ab ove a cer tain lati ­
tude, m ay in so acti ng, severely lim it
th e number of nic he s. . . .
NICHE REQUIR EMENTS
The vario us evolut io nary te n dencies,
notably rnet a ph o etesis, wh ich operate
o n single fo od chai ns mu st operate
equally o n th e foo d -web, but we a lso
h ave a new, if co m para ble, p ro bl em as
O . EVELYN Ill ;T CHJ ~ S () ~
to h o w mu ch diffe ren ce between tw o
species at th e sa me leve l is need ed to
p revent them from occupy ing the sa me
nic he . Wh er e metr ic character s are
invo lved we ca n ga in so me insi ght int o
t his ex tre mely important probl em by th e
stu dy o f wh at Bro wn a nd Wilson have
ca lle d chara cter displa cem ent or t he
divergenc e show n wh en tw o par tl y al­
lopatric sp ecies o f co mpara b le niche
requirem en t s becom e syrnpa tr ic in pa rt
of t he ir ran ge.
.. . In th e case o f t he aq ua tic in sects
wit h w hic h I began my a dd res s, we have
over mo st o f E uro pe t h ree very clo sely
all ied species of Corixa , th e largest
punctata, being abo ut 116 per ce nt
lon ger th an th e mi ddl e size d s pecies
m ucrocepha la, and 146 per ce nt lon ger
th an the sma ll spec ies affinis. In no rth­
we ste rn E uro pe there is a fo urt h
specie s, C. dentipes , as large a s C.
p unc ta/a a n d ve ry sim ila r in a p peara nce .
A single o bservatio n (Brown) suggests
that t his is what I have el sewh ere
te rme d a fugiti ve species, ma intainin g
it se lf in t he face of co m pe titio n m a inly
o n acco un t of grea te r mobility. Acc o rd ­
ing to M acan w hile both affin is a nd
macroceph ala may occu r with puuctata
they never are found with eac h o t her ,
so tha t a ll three spec ies never occ ur
toget her. In th e eastern part o f t he
ran ge , macro ceph ala d rops out. a n d
punctata ap pe ars to hav e a d isco n­
tinu ou s d istributi on, bein g record ed as
far ea st as Siml a , but not in so ut he rn
Persia o r K a shm ir, wh er e affinis occ urs.
In th ese ea ste rn local ities, where it
occu rs by itself, affinis is lar ger a nd
darker tha n in the west, a nd su pe r­
ficia lly lo oks lik e ma cro ceph ala .
T his ca se is very int eres t ing becau se it
lo o k s a s th ou gh cha rac te r d isp lace me nt
is occu rri ng, bu t th at th e size differen ces
between th e t hree species arc j ust not
grea t eno ug h t o a llow th em all to co ­
o ccur. O the r characte rs th an size are in
fact clearly in vol ved in t he se par a tio n ,
m acrocephala preferring d eep er wate r
207
th an affinis and th e la tter bein g more
tolera nt o f hracki sh co ndi ti o ns . It is
a lso interest ing beca use it ca lls a t­
tentio n to a mark ed di ffer ence th at m ust
occ ur betw een he mimc taho lo us insect s
wit h a n nua l life cycles invol vin g rela­
t ive ly lo ng growt h peri od s, a n d birds
or mamm als in whi ch the pe rio d of
growt h in lengt h is s ho rt a nd o f a very
sp ecia l na t ure co m pa red wit h the tot al
life span . In the latter, niche se pa ra t io n
may be po ssibl e merely throu gh genetic
size d ifferen ces, whil e in a pair o f
a n ima ls like C. puncta ta a nd C. affinis
we need not o nly a size d iffer en ce but a
seas o nal o ne in reprodu ction ; th is is
likely to he a rath er co m plica te d ma t­
ter. Fo r t he larger of two species a lways
to be larger , it must nev er br eed lat er
th a n th e smalle r o ne . I cia not d ou ht
that t his is wh at was happenin g in th e
pon d on M o nte Pellegrin o , but ha ve no
id ea h ow the di ffer en ce is a chi eved . . . .
.\10SA IC NAT URE OF T il E
E?\ VIRO N MEN T
A fina l as pec t of th e limit ati on of
po ssible d iversity , a nd o ne th at per haps
is o f greatest impo rta nce, co ncern s
what may he ca lled the mosai c na ture of
th e en viro n me nt. E xce p t perhaps in
ope n water whe n o nly uniform qu a si­
horizontal surfaces a rc co nside re d ,
every a rea co lo nize d hy orga nis ms has
so me local d iversit y. T he sign ifica nce
o f s uch lo cal di ver sity depend s very
largely o n t he size of th e orga nisms
un der co ns id erat io n . In a no ther paper
M a cA rthu r a nd I have . . . p ointed out
th at eve n if we co ns ide r o nly t he
he r bivo ro us level or o nly o ne o f th e
ca rn ivo ro us level s, ther e are likely,
a bove a ce rta in lo wer limit o f size , 10 be
mo re sp ceies o f sma ll o r mediu m sized
o rgan isms tha n o f large o rga nis ms. It is
difficu lt to go mu ch beyon d cr ude
qua litative im p ressions in testin g this
hyp oth esis, h ut we find t ha t for ma mm a l
fa una s, which co ntai n such d iver se
orga nis ms tha i th ey may we ll be
208 THE CO~ CEPT OF THE ECOSYSTEM ROBERT H. MACARTHUR AND JOH N W. MACARTHUR 209
re garded a s models of whole faunas , is indeed quite possible th at the re sults keys with frequencie s P" Thus, for to the birds. (T he rea so ns for this will
1her e is a de finite hint of the kind of of st ud ies o n these wond erful animal s instance , a o ne species community be discu ssed later. ) Looked at fro m this
th eo retical distributi on that W I.: deduce. would in long-range th ough purely alwa ys ha s zero di versity ; 2 spe cies , point of view, we can see the troubl e
1n qualitative terms the phenomenon practical terms pay for the esta blish­ one with 99 individuals and o ne with I with the other su hdivi sions. F or d e­
can be exem p lilied by any of the larger ment of grea ter reservation s and Na­ individual , will ha ve diversit y of -. 99 finiteness, co nsider 0-15',15 '-30'. :> 30' .
sp ec ies of ungulates whi ch may require tional P ark s than at present exist. . . . lo g• . 99 -- .0 1 log• .01 - - .046 " .0 10 Addin g a 0-1 5' layer to a habitat with out
a n u mber of different kinds of terrain We may , therefore, conclude th at th e = .056 (close to zero), while 2 spe cies it ca uses a mu ch gr eater increas e in
within their home ranges, anyone of reason why there are so many sp ecies of ea ch with 50 individual s will have hi rd spe cie s di ver sit y than the addition
which types o f terrain m ight be the animals is at least partly because a diversity of .347 ;- .347 -" .694. This of the layer :> 30' . There is nothing
habitat of some sma ll species. Most of complex trophic o rganiza tio n of a illu strates why diversity is a better biological about the number of la yer s
the genera o r even s ubfami lies of very community is more stable th an a simple mea sure than actual number of sp ecies, c hos en. F our or 5 layers in a roughly
large terrestrial anima Is contain on ly one, but th at limits are set by the ten­ for the community with 99 of on e and 1 similar subdivision would be more
one or tw o syrnpatric species. In thi s dency o f food chains to sh orten o r of the other seems closer to the com­ cumbersome to analyse but would
connection I cannot refrain from point­ become blurred, by unfa vorable physi­ munity with one species. Margalcf has presumably be even more accurat e. In
ing out the immense scientific im­ cal factors, by space, by the fmc ne ss of frequ ently used a similar measure in particular, the layers o-y , 1'-6', 6'- 15'
p ort ance of obtaining a re ally full p o ssible subdivision of niches, and by hi s plankton st ud ies. In terms 01' this, and > IS' s uggeste d by Elton and
in si ght into the ec ology of the large th ose characters o f the environmental the question becomes : "What is it about Miller allow a rather go od prediction
mammals of A frica while they can still mo saic which permit a greater diversity the environment which controls the o f the bird species diver sity.
be studied under natural condition s. It of sm all than of large allied sp ecies . . . . bird sp ecies di versity ?" The next questi on is : H ow much o f
The pr ocedure of the research de­ the remainin g scatter, i.e., how mu ch of
scribed here wa s to censu s a wide th e variability in bird species diversit y
variety of habitats, differing in (I) plant n ot accounted for by the variation in
species co mpo sitio n, (2) foliage height foliage hei gh t diver sit y, can be ac­
profiles, and (3) latitude , and to deter­ counted for in vari ati on s of plant sp e­
ON BIRD SPECIES DIVE RSIT Y mine how much each o f th ese fa ctors cie s diversity and latitude '! Remarkably
influ enced the bird diversity.. . . eno ugh, the a ns wer is " N o ne" . . . . Thus,
Robert H. Ma cA rthur and John W. MacArthur-1961 RESULTS
although plant species diversity al one is
a go od predictor o f bird sp ec ies d iver sit y
Reprinted by permission of the authors a nd publisher from Ecol ogy . .. [When] the corresp onding foli age it is because plant specie s diversity is
42 : 594-598, 1961. height diversity and bird spec ies di­ high when folia ge height diversit y is
vers ities . . . a re plotted as a gr aph . . . hi gh, and , when thi s is taken account
The recognition of the stabilizing role 0/ diversity in ecosystems prompts [they show] a clo se fit to the line : bird of, plant spe cies di versity ca n con­
investigation into the description and regulation 0/ diversity . This paper is species di versity = 2.01 foliage height di­ tribute nothing furth er. In o th er word s,
all excellent example ofsuch all investigation and its relevance 10 the discus­ versity-]-.46, ca lcu lated by least sq uares. habitats o f the sa me profile have the
sions by Hutchin son (see page 204), Eugen e Ot/UIII (see pag e 21 f) and Various other su bdivisions of the profile same bird sp ecies diversity wh ether
Murgoteft see page 215) will be readily apparent, into horizontal layers were tried, and the co mpos ed of few or many plant sp e­
layers 0-2 ',2'-25' and > 25' were chosen cies. ...
as th ose lay ers whic h made the col­
lection of points o n the graph m ost D1SC USSION
It is common exp erience that more The actual number of species is better orderly. ... The linearity of the cl uster These results are rather statis tical in
species of birds breed in a mixed wood replaced by a number called th e "bird of p oints indicates that the addition o f nature. What is their meaning in terms
th an in a field of comparable size. It is species diversity," calculated as fol­ a new la yer of a given amount of foliage o f indi vid ual birds o r species ? The
also well known that tropical forests lows : Let P, be the proportion of all of re sults in the same inc reas e in bird simplest explanati on whi ch seems to
seem to support more species than their the bird individuals which belong to the spe cies di versity , (not however the sa me ac count for the o bservatio ns, de scribes
temperate counterparts . These facts it" species . T h en the bird species div er­ increase in number of bird species ) no the "shape" of a bird 's ni che. Let us
a re o ft en explained in terms of th e matter which layer (0-2',2-25' or > 25') return to the pi cture of many territories
sity, is - L P, log. P" This is a formula
number of "n iches " or " way s of life' , is added, a n d no matter which other distributed over an area and consider
which the habitat provides. In thi s used by communication engineers to la yer s are present to begin with . Thu s, the follo wing evolutionary argument. A
p aper, a somewhat more precise an al ysis calculate the informati on generated, we can say that the la yers 0-2', 2'-25' large number of spe cie s can be a ccom­
is attempted, e.g., by a typist who uses the different and > 25 ' are roughly eq ually im po rtan t m odated in an environment in a variety
210
of wa ys of which th er e a re 2 ext re mes.
Ea ch spe cies may h ave di fferent habit at
prefer ence and feed throughout this
h abitat o n all kind s o f food , or, a ll
sp ec ies m a y sha re th e enti re habitat,
each species feeding o n a different
va r iety of food or in a differ ent sit ua t io n
wit hin the h abitat. T he first extreme
violates wh at might be ca lle d the
"j ack of all trad es-m a ster of n one"
pr incipl e th at natu ral selec tio n fav ors
th e in crea sed efficienc y resulting fr om
a ce rtai n a mo un t of spec ia liza tio n. In
th e oth er extrem e, spec ia liza t io n ha s
pr o ceed ed so fa r th at t ime a nd ene rgy
are wasted in tr avellin g between spo ts
fo r whic h the specia liza tio ns a re
a da p te d. It is hard t o say just where
th e balance o f th ese op posing req uir e­
rnents would be reach ed, but it is cle ar
th at greater sp ec ia liza t io n res ulting in
in cr eased efficiency would a lways be
fa vored as long a s no tim e or ene rgy
are wa st ed . And no t ime o r ene rgy will
be wa st ed it' nich es a re " con vex" in th e
se nse that b et ween a ny 2 fairl y di st ant
feeding pl aces th er e will be a fairly
natural route also co ns ist ing of feeding
place s. A specia lizatio n to a single tree
sp eci es in a mix ed for est would cle arl y
violate thi s since, in pa ssing fro m on e
suita ble tree t o a no t he r, the bird would
go throu gh m any un suitable o nes. Thus,
n atural selec t io n would t end to eli­
minat e a sit uatio n in whi ch bird spec ies
diversity dep ended up on tr ee species
di ver sit y, unl ess, as in so m e fru it ea ti ng
spe cies, a very rem arkabl e improvem ent
in effici en cy is ac hieve d a lo n g with the
restricti on in feedi ng p osit ion . Thus,
one p rinc ipal res ult of these ce n suses
ca n be pr edi cted on ass um in g that
nich es a re co nve x.
Next , we m a y as k " why a re th e layer s
0-2', 2- 25' , :> 25' eq ua lly im po rtant? I s
it b ecau se bird s res po nd to d iffer ent
heights, or is it becau se th ey res po nd to
differ ent co nfig u ra t io ns of ve geta tio n in
different la ycr s ?" In t he latter case,
herbs, bu shes a n d trees presum ably
TH E CONCEPT OF THE ECOSYSTf. M
co rresp o nd to the layer s 0- 2', 2-25'
a nd :> 25' re sp ect ively, a lt ho ugh sma ll
trees count as bu sh es, et c. Th ere is go o d
eviden ce for this latter ex pla na t io n. F or,
althou gh dec idu ous forest s va ry prin­
cipally with heig ht a bo ve t he gro und
a n d hence h ave a bird d iver sity pr e­
dictable from the height pr ofile , co n ifers
(especially sp ruce) have a m arked
"i ns ide" a nd " outside" fo r which
species are specialized . H en ce b ird
species d iver sity would be h igh in a
mature sp r uce forest even if fe w layer s
were present. This is pr ecisel y w ha t
happen s in th e M ain e whi te sp ruce
wood mentioned ea rlier , wit h bird
sp ecies d iversity of 1.71 2 a nd fo liage
height d ivers ity o f .28 7 whic h is serio us ­
ly off the gr ap h of d eciduou s forest s.
A differ ent wa y of lo oking at th e dat a
gives addi tio na l ins ig ht. Wall has
p ointed o ut th at plants arc distributed
in patch es . Hutchin son a nd M a cArthur
attempted to ex p la in th e sizes of coex­
isting organ ism s in terms o f a n envi ro n­
ment composed o f a m osaic of kinds o f
patches. Differ ent co m bi natio ns o f
patches formed the habitats selec te d by
different species . T he present resea rch
ca n be eas ily interpret ed in terms of t h is
pi cture of th e enviro n me n t. In fac t, o ur
results su ggest th at t he p at ch es forming
the birds' en vir onm ental mo sai c are
e
sectio ns of can opy (o ver 25'), p atches
of bu sh es B fr om 2- 25', a n d th e her­
baceous a nd o t he r cove r H less t ha n 2'
fr om the ground. And th e se q ue nce o f
p atches enco u ntered in mo ving t hro ug h
t he habitat (o r in takin g eve r larger
sa mp les) is th en rep resented by a
seq uenc e of letter s, e.g., c, B, H , H , B,
C, . .. with ce rta i n random proper ti es
but a lso su bj ect to the con dit io n th at
the long term frequ en cy o f C's, B's, a nd
H 's sho uld co nform to th ei r resp ecti ve
d en siti es (Pi) in th e p arti cul ar habi tat.
If the se que nce is ergo dic, which defines
wh at we ca ll a homogen eou s habitat ,
then it is well kn own t hat t he un cer ­
tainty of th e next letter s in the se q uen ce
EUG ENE P. OD UM 211
is a p p ro p r iately me asured by the for­ height d ivers ity, whi eh is essent ially th e
mul a - ~ P, lo g, Pi wh ich we used . If, pr ed icted valu e o f the bird speeics
i
di ver sit y. Thu s we ca n say th at bird
inst ead o f co nside ri ng th e un cert ainty
o f futur e single letters in th e se q ue nce, spe cies di versit y is d etermined a s if th e
we a sk for th e un certaint y of f ut ure bird s recogni zed su ita b le habitats by
p air s of letter s, th e formula be com es pairs o f foli a ge types ( > 25', 2-2 5',0-2').
Th e species area curve co u ld then be
-2 ~ Pi lo ge P, which is 2 Yo foli a ge
\ predict ed fro m thi s.
RELATIONSHI P S BETW EEi'l STR UCT URE AND F UI\'CTION
IN ECO SYSTEMS
Eugene P. Odum-1962
Rep rinted by pe rm issio n of th e au tho r a nd pub lisher fro m t he J ap an ese
Journal of Eco logy 12 : 108 118, 1962.
In defining ecology as the study of the structure andfunct ion of ecosyst ems,
Odum brings Into mu ch clos er alliance these t wo m ajo r traditional approaches
In biology . Th ere is intimation of causal relat ions between the t vvO in line
with current thinkin g on the molecular and subce llular levels of biological
organiz atio n. By discussing aquatic and terrest rial sys tems in p arallel,
Odum strengthens his strongly espoused contention regarding the univer­
sality of applying the ecosystem approach, This view is the them e of bot h
editions of his tex t , Fundam ent als of ecology (1953, 1959 . Philadelphia,
W. B. S aunders, Inc.) which enjoys wide use,
. . . A s yo u kno w ecol ogy is ofte n material s suc h as nutrients, water, etc. ;
d efined as : The st udy of int errel ati on­ (3) th e ra nge, or grad ie nt , of co nd itio ns
ships between o rga nisms a nd e nv iro n­ o f ex iste nce s uch as temperature, light,
ment. I feel that this co nve nt io nal etc. Di viding eco logical str ucture into
definiti on is n ot s uita ble; it is to o va gu e th ese three di visions is, of co u rse , a rbi­
and too br oad . Pe rs on all y, 1 prefer to tr a ry but 1 b elieve co nve n ient for actual
de fine eco logy a s: The st udy o f th e st udy of both aq ua tic and terrestrial
str uct ure a nd fun cti on of ecosyste ms. sit ua tio ns .
Or we might say in a less technical By fun cti on we me an: (I ) The rate of
way : The study of struct ure and Iunc­ biol ogical ene rgy flow through th e
ti on of nature . ecosystem, th at is, the rates o f pro­
By st ruc t u re we mean : ( I) Th e co m­ du cti on a n d th e rat es o f respiration of
position o f the biological co m m u ni ty t he popul ati on s and the co m m u nity;
incl ud ing sp ecies , numbers, b ioma ss , (2) th e rate o f - mate ria l o r nutrient
life history and dis tribution in sp ace of cycl ing , th at is, the bi ogeoch emical
p opul ati ons ; (2) t he qu anti ty a nd cycles; (3) biol ogical o r ecol ogical
di stribution of th e a b io tic ( no n -livin g) reg ula tio n includi ng both regulation of
211
organisms b y environ ment (a s, for
exam p le, in photoperiodi sm) and re­
gulat io n of en vironment b y orga nism s
(a s, for example, in nitrogen fixation b y
microorganism s). Again , dividing eco­
logical function into these three divi­
sions is arbitrary but conven ient for
study... .
Both aquatic and terrestrial com­
munity types ha ve severa l structural
features in comm on. Bo th must have
th e sam e three necessary biological
compon ents : (1) Producers or green
plants capable of fixin g light ene rgy
(i. e. , autotrophs) ; (2) animals or m acro­
co ns um ers which consume part iculate
organic matter (i .e., ph a gotrophs) ; and
(3) microorgani sm d eco m p osers which
di ssolve organ ic m atter releasing nut ri­
e nts (i.e., osrno tr op hs) , Both eco­
sys tems must be supp lied with the sa me
vital material s such as nitrogen, pho s­
phorus, trace mineral s, etc. Both
eco syst ems are regulated a nd limited by
th e sam e co nd itio ns o f existence such as
light a nd temperature. F ina lly, the
arrangement o f biol ogi ca l units in
vert ica l space is ha sicall y the same in
the tw o co n trasting types of ecosys­
t em s. Both have two st rata, a n autotro­
phic stratum above and a he ter ot rophic
st ratum below. The ph otosynthetic
machinery is concentrated in the upper
stratum or ph otic zo ne where light is
available, while the con sumer-nutrient
regenerating machinery is co nc ent ra ted
largel y bel ow the photic zo ne. It is
important to em p has iz e th at while th e
ver tica l extent or thickness of com­
munities va ries greatly (esp ecially in
water), light energy co m es int o th e
eco system o n a hori zontal su rface
ba sis which is every where the same.
Thus , different eco systems sh ould b e
compared o n a squa re meter basis, n ot
on a cubic or volume b asis.
On the o t her hand, aquatic and
terrestrial ec o systems differ in st ruc tu re
in several important wa ys , Species
composition is, of course, compl etel y
different; the roles of producers , con-
THE COl'CEPT OF 'I1 IE ECOSYS11:M
su mers a nd de com posers a rc ca rried o ut
by ta xo no mica lly different organisms
which ha ve bec om e a da p ted through
e vo lution , T rop h ic str uctu re a lso di ffers
in th at land p lant s t en d t o be la rge in
size but few in n u mber whi le the a uto­
trophs of open water eco syst ems (i.e.,
phytopla nk to n) a re small in size but
very n u me ro us. In general, au to tr o ph ic
bio m ass is m uc h gr eater th an hetero­
trophic bi om a ss on la nd, wh ile the
reverse is often true in th e se a . Perha ps
t he most impo rta nt difference is the
fol lowin g T he matrix, or s up p or. ing
fra mewo rk, o f t he com m un ity is la rgely
physical in a q ua tic ec osyste m s, but m or e
stro ng ly biologica l on land. T hat is to
say, th e community itself is im po rtant
as a ha b itat on land, but not so im­
p ortant in water.
N ow , we ma y a sk : How do th ese
simila rities and ditTerences in structure
a ffect ecolo gical fun cti on ?
One i mpo rt a nt aspect of functi on is
. . . th e energy flow tb rough the eco­
sys tems begi nni ng with the inco ming
so lar en er gy and p a ssing t hrou gh the
success ive trophic level s. A t each
transfer a large part o r the en ergy is
di ssip at ed in re sp ira t io n a n d passes o ut
of th e system as heat. The amo nnt o f
energy rem a ini ng aft er th ree steps is so
sm all th at it ca n be ignored i n so far as
th e ener getics of the community a re
concern ed . H ow ever, t ertia ry consumers
("top carn ivo res ") can be important as
regulators ; th at is, predati on may h ave
an important effect o n ener gy flow at
the herbivore level. . ..
The autot rophic-h eterotrophic stra ti­
fication, which we emph a sized as a
uni versal feature o f community st ruc­
ture, res u lt s in tw o ba sic fo od chains.
. .. Th e co n su mp tio n o f livin g pl ant s by
herbivores whi ch live in the a ut o­
trophic stra t um to gether with their
p redators m ay be consid ered as the
grazing fo od chain. Th is is the classi cal
fo od chain of ecology, a s, for exa mp le,
th e phytoplan kt on-zo opl ankton-fIsh
sequ ence or the gras s-ra bb it-fox se-
EUG EI'E P . onusr
qu ence , However, a large p ro portio n o f
t he net pro d uct io n may no t be co n­
sum ed until d ea d, thus beco ming t he
sta rt o f a ra ther differe nt energy flow
w hich we ma y co nv enien t ly designa le
as th e detritus f ood chain. Th is en ergy
flow ta kes pl ace la rgel y in th c hc tcro ­
trophic strat um . . .. th e detritus ene rgy
flow takes place ch iefly in t he sedi me nts
o r wa ter sys tems, a nd in th e litter a nd
so il o f la nd system s.
: co logis ts have to o oft en over loo ked
th e fa ct that the det ritus fo o d chain is
th e mo re irnpor tn nt energy pathwa y in
ma ny eco sys tems . ... a la rger porti on
of net pr o d ucti o n is estima ted to be
cons ume d by grazers in the marine ba y
t ha n in the fo rest ; n in e-t enth s o f the
net p rod uct io n o f t he for est is est ima ted
to be co ns u med a s detrit us (de ad leave s.
wo od , etc .). It is n ot clea r wheth er th is
difference is a di rect or indirect result of
t he d iffe ren ce in co m mu nity struct u re.
O ne ten t ativ e genernlizat io n might be
pr oposed as fo llows : com m u ni ties o f
small, rap id ly growin g prod uce rs s uc h
as ph yt opl ankt on o r gr ass can t ol erate
hea vier grazing pre ss ure t ha n co rn ­
munit ies o f la rge , slo w-growing pla nt s
such a s tr ees o r large seaweed s. .. .
Despite th e large d ifference in relati ve
size o f sta nd ing cro ps in th e tw o ex­
trem e types o f ec osy ste ms, th e ac tua l
en ergy flow may be o f the same o rd er o f
magnitude if ligh t and avai lahl e n utri­
ent s are sim ilar. ... Th us . XO KCa is o f
ph yt oplankt on m ay hav e a net pro­
du ct ion a lmos t a s large as 5000 KCal s
o r trees (or 500 KC a ls o f green leaves).
T herefo re , producti vit y is not pro­
p ortion al to the size of the st a nd ing cro p
except in sp ecia l cases invo lving annual
p lants (as in some ag riculture) . U n­
fortun at el y, ma ny eco log ists confu se
p roductivit y a nd sta nd ing c rop. Th e
relat ion between st ruc ture an d functi on
in thi s ca se depends o n th e size a n d ratc
of met ab ol ism (an d rat e of t urn over ) o f
the o rga nisms .
T o s um ma r ize, we see t hat bi ol o gical
struc tu re influen ces th e pattern o f
213
en ergy flow, part icul arly t he fate o f net
produ cti on an d the rel ati ve im po rta nce
of grazers a nd detrit us con sumers.
H owever, total en er gy Ilo w is less
a ffected by str uct ur e, and is thu s less
varia b le th an sta nd ing crop. A func ­
t ion al homeo sta sis ha s b een evo lved in
nature desp ite the wide ran ge in species
st ruct u re a nd in bi omass st ructure. , . •
. . . No w let us turn to st r uc t ure and
functi on a t th e p opulati on level and
consid er a seco nd ma j or a spect of
functi on . namely, the cycl ing of nutri­
en ts, . . .
First, we sh all take a look at the sal t
m arsh ecosystem a nd th e d ist r ib utio n o f
t he speci es in th e marsh . The mu ssels
live partl y buried in the sed imen ts and
attach ed to the ste ms and rh izomes of
the mar sh gras s. Sp artiua alternifiora.
Individual s a re g ro up ed in to co lo nies
(clumped di stribut ion ), b ut th e co lo nies
are wid ely scattered ove r th e marsh .
N u m bers av er a ge 81M 2 for the entire
marsh and 321M' in th e most favo rable
parts of the marsh . Hi ornass in terms of
ash-free d ry weight ave rages I J.5 gins]
M ~. W hen th e tide co ve rs the col on ies
th e va lves partl y ope n and the an imals
begin to p ump Jarge quantit ies o f
wa ter.
... Each da y the p opulati on re­
mo w s a la rge p art of the ph o sph oru s
from th e wa ter, espe cially the part iculate
frac ti on. M o st of th is d oes not ac t ua lly
pa ss throu gh the b od y but is sedi­
merited in the for m o f pse ud ofeccs
w hich fall o n th e sed i me n ts. Thus, th e
mussel s mak e la rge quanti ties or phos­
ph oru s availab le to microorgani sm s
a nd t o the a u to t ro p hs (be nthic a lga e a nd
marsh grass} .. . the ene rgy flo w was
esti mate d t o be abou t 0.15 KCal~/M ' 1
day.
Til e mo st important finding o f th e
study is .. . the rati o between flux a nd
a mo unt. N ote that over o ne third o f the
14 mgrns of part iculate pho sph orus is
rem oved from the wat er eaeh da y b y
th e p opu lati on , a nd thereb y ret ain ed in
the m arsh. In co n tras t , less th an o ne pe r
21
T"'[ - A"O VT THHt! WL[ 'S
FlO. 3. Th e Ma rgelef model of ecolog ical sue­
cession showing a simple typ e of succession.
which can be demonstrat ed in laboratory cultur es.
The fla sk s show changes ill species composition
occurring ......hell succession is set in motion by
the intr oduction of lIew nutrient media into all
old "climax" culture. The graph shows resultant
changes in two aspects of diversity uud in the
relation betw een pr oduction and biom ass (Pi B).
S ee text f or det ails of the experime nt.
cent of the 20 KCals o f potential ene rgy
(net production estimate) available is
actually utili zed by the mu ssel popu­
lation . In other words, the mu ssel
population h as a much more imp ortant
effect on the community ph o sphoru s
cycle than it ha s on community energy
flow .. ..
To summarize , the mu ssel stud y
brings out two important points: (1)
It is nece ssary to study both en ergy flow
and biogeochemi cal cycles to determine
th e role of a particular species in its
eco system , (2) anima ls may be impor­
tant in the eco syst em not only in terms
of food ene rgy, but a s agents which
make basic nutrient more available to
a uto trophs. . . .
N ow let us co nsid er the third im ­
portant a spect o f ecological function ,
th at is, co m m unity regul ati on. Eco lo­
gical su ccession is on e of the mo st
important pr o cesses which result from
the community mod ifying the en vir on-
TIl E CONCEPT OF THE ECOSYSTE.\ 1
D _ 'Q. QlII.Tr... ~
" OflltI:NAL.G.AE
8
A
l:II«')I'LAGC:l &.ATE S
I'OT'W(
ment. Fi g. 3 illu strates a ver y simple
typ e of eco lo gical s uccessio n wh ich can
be d em on str at ed in (I laboratory ex­
periment. Yet the ba sic pattern sho wn
here is the sa me as Occur s in more
complex s uccess io n of natural com­
munities. Th e di agram (Fi g. 3) was
suggested to me by Dr. Ram on
Mi\RGELFF, hence we may call it the
M i\RG E L EI ~ mo del of succession ,
At th e to p of th e diagram (Fig. 3) are
a series of culture fla sk s co n ta ining
pl ankton communities in different sta ges
of suc cession. The graph sho ws c ha nges
in two as pe cts of str ucture and in one
a spect of functi on. The first flask o n the
left contain s a n old a nd relatively
sta ble community; thi s fla sk represent s
the climax . Diversity of spec ies is high
in the clima x; species of diat om s, green
flagellates, d in oflagellates a nd r otifers
arc shown in th e diagram to illu str ate
the vari et y of plants and a nima ls
present. Biochemical diversity is a lso
high as indicated by the ratio of yellow
plant pigments (optical density at
430mj1.) to chlorophyll-a (optical den sity
a t 665mf.L) . On th e other hand the ratio
of production to bi oma ss (P jB in rig. 3)
is low in the old or climax culture. a nd
gross production tends to eq ua l com­
muni ty respirati on . If we a d d fresh
culture medium to the old culture, a s
sh own in Fig. 3, ec ological s uccess io n is
set in motion . An early stage in sue-
RAM ON M AR GALEF 215
ce ssion is shown in the seco nd flask . The cha nges which we have just des­
Speci es div ersity is low , with one o r tw o cribed arc apparentl y typical of all suc­
spec ies of ph ytoplankton d omin ant. cession regardless of enviro n men t o rtypc
Chlorophyll s pred ominate so that the of ecos ystem . Althou gh much m ore
yellow / gr een rati o (0 .D.430 jO .D .665) study is needed , it appears that differ­
is low , indicatin g low biochcmica I cn ccs in community structure mainly
di versit y. On the other hand . production a ffect th e tim e required , that is, whether
now exceeds respiration so that the ratio the hori zontal scale (X- axis in Fig . 3) is
o f production to bioma ss becomes measured in week s, months or years. . . .
higher. In o t he r word s, autotroph y T o s um ma rize , I. a m suggesting that
grea tly exceed s heter otrophy in the the ba sic pattern o r functi onal chan ge
pi on eer or early success ion stage. Tile in eco log ical succession is the same in
tw o flask s o n the right side of th e a ll eco systems, but that the species
dia gram (Fig. 3) sho w the gradual co mpositio n, rate of c ha nge and du­
return to the climax or stead y state ra tiou o f succession is determined hy
whe re a uto tro p hy tend s to balance th e physical environment and th e
heterotrophy. resultant community structu re.. ..
ON CER TAIN UNIFYING P RINC IPLES IN ECOLOGY
Ramon Margal ef-1963
Reprinted by permission of the author a nd publisher from the American
N aturalist 97 : 357- 374, 1963.
A dmo nishing ecologis ts (iJ I' haviut; been reluctant to orient their eff orts in
terms of a general theory, M arguk-f proposes SO If/(' UI/ify illf,' principles with
major emphasis all the ntat urity of the system as measured by diversit y
and ill term s of energetics. Maturity is a qttantitative measure of tlu: pattern
In which the components of rite ecosys tem are arranged and related. In
spite of its theoretical orientation, lite easy readabilit y of the essay is credit
to the auth or' s clear p erception and ills(I{ht. The essay is largely a speculative
interpretation enco mpass ing both orieinat and undeveloped ideas . Because
o] its spe culative nature ami theoretical ideas it should he read not as all
instan ce of accomplished fa ct but as a possible indicator of future lines of
inquiry.
T he main point is th at the "real"
STRUCT URE or TIlE ECOSYSTBI
struct ure o f a n eco syst em is a property
Ec o systems have II struct ure , in the that rema in s o ut of reach , but thi s
se nse that they are composed of dif­ complete structure is reflected in man y
fe ren t parts or elem ents, a nd these a re a spect s of the ecosy ste m that can be
ar ra nge d in a definite pattern. The su bjected to ohse rva tio n : in the d is­
interrelations between th e con stitucn t tribution of indivi du als into spe cies, in
elemen ts a re the basis o f the structure. . . the pattern of the food net, in the
216
distribution of total ass imila tory pig­
ments in kinds of pigme nts, an d so on.
St ruct ure, in general, becomes more
com plex, mor e rich. as time passes ;
structu re is linked to histor y. Fo r a
qu a ntitative mea sure of structu re it
seems con venient to select a na me that
suggests this histo rical character, for
instance , maturity. In gene ra l, we may
spea k of a mor e complex ecos ystem as
a more mature ecosystem. .. . Th e term
maturity suggests a trend , a nd moreover
maint ains a co ntact with the trad ition a l
dyn ami c a pproach in the st udy of
natural comm unities, whic h has always
been a so urce of inspira tion.
Maturity, then, is a quality that
in creases with ti me in an y undis tu r bed
ecosystem . Field ecologists usc many
criteria to es tima te th e maturit y of an
ecosystem, wit hou t the need of assess­
ing it s pr ecise pla ce in a n ac tual
success ion. Empi rica l kn owled ge of
succe ssio n lead s o ne to co nsider as
mor e m ature the ecosystems that arc
mor e complex ; tha t is, com posed of a
grea t number of elements, wit h lon g
food chain s, a nd with rela tio ns be­
tween species well defined or more spe ­
cialized . . . .
TH E ECOSYSTE.M IN ReLATI O N
TO EJ','F.RG Y AND MASS
The ecosys tem has di fferent com­
plementary as pects : 11' we co nsider the
eleme nts a nd the rela tions bet ween the
elements, we have t he struct ure, whereas
in co nsideri ng m atter and energy, we
ha ve to deal wirh metric properties
which are perhaps ea sier to express.
Th e ecosys tem is formed by a cert ain
a mo unt of matt er (biomass) and there
is a budge t of matter and energy.
For th e moment, let us co nside r a n
ecosystem in a steady sta te, wit h a
material o utp ut equ al to th e materi al
input. Here we need to con sider onl y
two qu antit ies : th e matter present , or
biom ass, in the ecosys tem, always to be
ex pressed in the sa me form (to ta l
T HE COr-;CEI'l' or 1 11E I:COSYSTI.M
weight. d ry weight) ; a nd the p ote nt ia]
en ergy nCCCSS.lry fo r maintena nce in
the ecosystem, amo unting to tot al
respiration a nd o t her losses. Bot h
q uan tities can be considered in every
cosysrern and simply equa ted to prima­
ry production (P) and biomass ( 8);
both co ncepts are of co mmon usage in
eco logy. Their relatio n (Pi B) ca n be
sta ted as flow o f energy per unit bio­
mass : it is the I urnover rat e or
C ushing,
H umphrey, Banse and Lacvastu a nd
the productivity ind ex under nat ural
light condi tion s o f St rickland .. . .
What is imp ort ant is the em pirical
relati on between st r uctu re and energy
flo w per unit biomass. More mature
ecosystems, with a richer struc ture, have
a lo wer primary production pe r unit
biomass. . . . T he rutin PI B is ta ken as
the rat io exp ressed by primary pro duc­
tionltotal biomass, includ ing all cle­
ment s of the ecosystem , such a s the
consumers, etc. In ecosystems of higher
maturity there is a mo re co mplete usc of
foo d, there is a grea ter prop orti on of
anima ls, and energy cascades th rou gh a
mo re con sid erable n umbe r of steps.
T his is true in aq uatic ecosy stems, but
in terrestria l ecosys tems a so mew hat
par ado xica l situat io n ar ises owing to a
ce rta in exaggera ted do mina nce of
vegetatio n. O n the other hand , the
grea t number or possible kind s or
relations in a matu re ecosystem a llows
a higher efficiency in every rela tion. If
these re lations are considered as com­
mun icat ion cha nnels, less noise co mes
in to them.. . .
T he ideas d eveloped so far can be
sum mar ized as follows. A n ecosystem
t hat has a co mp lex stru cture, rich ill
informa tion, need s a lower a mou nt of
ene rgy fo r mainta ining such structure .
If we co nsider the interrelati on s be­
tween the eleme nts o f an ecosystem as
co mm unicatio n channels , we ca n sta te
that such cha nnel s funct ion on the
average mor e effectively, with a lower
no ise level, if they are multiple and
R AMO N MA R G AL EF
diverse, linking elements not subjected
to great cha nges. T hen. loss of energy
is lower , a nd the energy necessar y for
pre venti ng decay of the whole eco­
system amo unts relat ively to less. T his
seems to be one of the basic prin ciples
o f eco logy, p ro ba bly recognized tacitl y
by most wr iters, altho ugh rar ely put in
an explicit way.
SUC CESSIO N AN D I' I.l IC TLJATION S
Any ecosys tem not subjected to
st ro ng disturbances co ming fro m o ut­
side, changes in a progressive and
direction al way. We say tha t the eco ­
syste m beco mes mor e ma tu re. The two
most noti ceab le ch anges accom pa nying
th is process arc the increase o f co m­
plexity of struct ure and th e decrease of
the energy flow pe r un it bio mass . T his
theo retica l backg round leads us to
accept a so rt of na tural selection in the
possi ble rea rr ange ments of th e eco­
system: Lin ks between the elements of
an eco system ca ll be substituted by
o the r links tha t work with a highe r ef­
ficiency, req uiring II chan ge in th e
elements a nd o ften a n increa se in th e
num ber o f elements and co nnectio ns.
T he new si tua tion no w has an excess o f
potentia l energy. T his can be used in
develo ping the ecosystem furt her. for in­
sta nce. by add ing biomass after d riving
mo re matter into the system. A mo re
co mplex Stale. with a red uced waste
of energy. allows main te na nce of the
sa me biom ass with a lower supply of
energy-or a higher b io mass wit h th e
sa me supp ly of energy-s-aud replaces
au to maticall y a ny previous state.
he on ly limit set to l itis p rogressive
change is interference from the ph ysica
enviro nmen t. Succession can build
istory only when the environmen t is
sta ble. In t he case of .1 cha nging en­
viro nment , t he selected ecosysrern will
be co mposed of species with a high
re pro du ctive ra te and lower special
req uireme nts . Suc h a n eco system is
less diverse an d less co mplex ; the surface o f equa l ma t urity. Remember
217
energy flow per unit biomass rem ain s
relati vely hig h.. . .
. .. Th e co nclus ion is that in any
estimat e or maturit y, not o nly d iver sity,
b ut al so predict abilit y o f cha nge wit h
time has 10 be co nside red. Ord inaril y
bot h cha rac ter s ar e co rrela ted . Less
mature ecosys tems n ot o nly ha ve a
lower diversity, but in them tran siti on
between s uccess ive sta tes includ es a
higher amo unt of u nce rta inty. And
mort; diverse ecosyste ms have, in gener­
al, mor e pred ict able fut ure states . In
ot her words. in more mature ecosystems
the fut ure sit uation is mor e d ependent
on the pre sent than it is on inputs
co ming fro m o utside. Homeo sta sis is
high er. On t hc OTher hand , futur e sta tes
in less mature ecosy ste ms a rt; hea vily
influenced by exter na l input s, by
changes in t he ph ysical e nviro nme nt. . .
In genera l, th e expec ted differen ces in
the character of fluctu at ion s in less
mat ure and mor e mat ure co mm un ities
would be as fo llows. In less matu re
co m munities, enviro nme nta l fluctu­
at ion s a rc st ro ng and a ble to stop the
tren d to incr ease mat uri ty at a ce rtai n
level. Ma t urity d ocs not incre ase be­
ca use abiotic fluct uati ons arc too stro ng,
and hom eostasis is difficult to a tta in in
a poo rly orga nized , ofte n a p ion eer
commun ity. In a mor e sta ble enviro n­
me nt, successio n proceed s a nd ma turit y
incr eases: now W I: ha ve to expect
rh ythm s tha t arc mo re regular , mo re
inde pend ent or env iro nme nt a nd often
endo genous. A nticipatory power has
survival value and is the expressio n of a
co mplex system. ab le to produce very
effi cie nt ho meo sta tic mech ani sms. Up
to a ce r ta in level, t hese hom eosta tic
mech anisms can protect the system
fro m disru ptio n du e to ex ternal a gents.
Mat uriry is self.. preser ving. . . .
L\: TE:-': SIVE SY S TF.~ I S \VITII LOeA'I. D II'­
FEREN C E,<; 1;-1 '111 1. V A L GE or MATURIIY
Let us exp lo re wha t ha ppens alo ng a
218
t hat at one side we have a su bsystem of
lower mat urity , with a high product ion
per unit bioma ss, wit h less st ro ng links
bet ween species, su bject to wider fluctu­
a tion s and to a n easy dispersal of t he
ele ments. A t t he ot her side we find a
subs ystem with a g reate r biomass for
th e sa me e nergy flo w, wit h well o r­
ga nizc d relation s over elements mo re
st ro ngly locali zed .
If ma turi ty increases in the less
mature syste m, especially at the prox­
im it y of th e boundary (which is to be
expec ted from successio n) the sur fac e
of eq ua l mat ur ity moves to ward s th e
less mature s ubsys tem . T his is pr ob­
a bly acc o mpanied by a flo w of energy
going th e co nverse way. Th is mean s
th a t matter (biomass a nd non-living
matter) goes in both directi on s, since
both co up led suhsyste rns are act ua lly
open , b ut the co nte nt of potential
e nergy of s uch matt er is, On th e average,
h igher in th e matter goi ng the way of
increas ing maturity th an in t he matter
goi ng tile way o f decrea sing maturity.
T he subsyst em with a lower matu ri ty
ma intain s a high er rat io bet ween p ri­
mary pr odu ct ion a nd tot al present
biom ass, beca use it actu all y loses
bioma ss, in go ing across th e bord er to
th e mor e mature co upled subsy ste ms.
Le t us rc me m bel' t ha t successio n is
simply t he exc hange of an excess
ava ilable energy in the present, for a
fut ure increase of biomass. A n eco ­
system in its present state is less mature
and has a n exce ss prod ucti on th at go es
to th e futu re and h elp s reorga nize th e
eco system in a mo re mature form . If
t her e is no a va ila ble excess producti on
or it is dr ain ed ou t of th e syste m,
successio n p roceeds no furt her. . . .
UT ILITY OF A SYNTlJl,T1C AP PROACH
M ost of what has been d iscu ssed ca n
be summa rized in two very simple
p rincip les :
( I) T he relat ive a mo unt of energy
necessa ry fo r ma int ainin g a n ecosy ste m
TH E CONCEPT Of TH E ECOSYSTEM RA MON MAR GALEF 219
is relat ed to th e degree o f st ructu re or bu t with no effect o n pot enti al ene rgy from less mature (ru ra l) ar eas to mor e
o rga nizatio n o f th is ecosyste m . Less flo w; ra dia tio n, then , mu st redu ce the mat ure (urba n) areas . The urb an ce n­
energy is necessar y for a mor e co mplex maturity o f ecosystems, in part by ters re prese nt loca lized elem ents that
ecosy ste m, a nd th e natural tr en d in selective destructi on of the mor e have accu m ulated h igh a mo unts of
successio n is to wa rds a decreasing flow mature element s of the ecosyste m. in formati on , fed on th e p roduction of
of energy per unit of biom ass a nd Thu s, a great increase in rad iat ion ma y neighborin g subsys tems, a nd ha ve ex­
tow ards incre asin g orga nizatio n. Brief­ mean a new pu sh given to an a lrea dy erte d a directiv e ac t io n. Ver y old sys­
ly st ated th e trend is to ward s incr ea sing lag ging evolution. tem s ca n survive wit h a small /low of
maturity. Most of th e sa me p rin cipl es ca n be energy, a nd lik e th eir ecolog ical
(2) When tw o syste ms of d iffere nt applied to human organiza tio ns. T ak­ co unterpa rts ca n bre ak d own as a
maturity meet a lo ng a bound ary th at ing as criteria the dive rsificati on of co nseq ue nce of a min o r en vir onm en tal
all ows an excha nge, energy (p rod uc tio n) skills a nd j ob s (dive rsity), or th e rela­ cha nge. It is possible to deal o bjec tively
flow s tow ard s th e mor e ma ture sub­ tive flow of p ot enti al energy , it is a nd qu antit ati vely with big and co mplex
system, a nd th e boundary or surface of p ossibl e to map th e "ma t ur ity " of str uctures , if o ne never for gets th e
eq ua l maturity shows a t re nd to move stat es and co nti ne nts in the ceo logical co m plementa ry as pec ts of ener gy as
in an op posite direct io n to such energy sense o f o rga nizat ion. Energy Row goes relat ed to matter , a nd struct ure.
flo w.
Th ese gener al principl es clari fy man y
ecol ogical inte rac tions and pr ocesses
and a llo w q ua ntitati ve fo r mulatio n.
Th ey ca n be use d or tested in pr edict ing
ch a nges ind uced by hum an ac tio n.
Ex p loita tio n is lik e inflicti ng a wo und
up on a het erogen eou s o rganic st ructure :
so me ti ssu es o r su bsys te ms (more
mature) d o not rege nera te ; o the rs (less
mature) do a nd these supply th e basis
for a further even tua l inc rea se of
maturity. M aintain ed ex ploita tion keeps
th e m aturity of t he exp lo ited system
con stantly low. Ex p loited natural co m­
muniti es co me to ha ve a higher pr imary
production per unit biom ass, a lower
species d iversit y a nd, p resuma bly, a
lo wer rati o D o o/D (!05' Mo re ene rgy
goes into fluctu a tion s suc h as th ose
represented by exploi ted popul ati on s
o r by popul ati o ns th at are integra ted
into exp loit ed eco syste ms. Fo r exa m ple,
pests ha vc fluctu at ions with a wider
range a nd shorter peri odicit y th an
simila r population s t hat arc integ ra ted
into more mature, event ua lly uncx­
pl oited, ecosystems . Extre mely matu re
ecosystems, suc h as tro p ica l fore sts,
are unable to go back a nd a rc totally
disrupted by h um an exp loita tio n. . . .
. . _ Ra dia tio n increase ca n be ex­
pected to ac t dest ruct ively to accu mu
lated info rma tio n (th at is, t o biomass)