Fundamentals

of Ecology
FIFTI{ EDITION

EugeneP.Odum,Ph.D.
oJEcologt
LateoJUniversity
oJGeorglaInstitute

GaryW.Barrett,Ph.D.
of Ecolog4UniversityoJGeor$aInstituteof Ecologt
OdumProJessor

-rHorvlsoN

=+-*

BROOKS/COLE

TheScopeof Ecology
Ecology:Historyand Relevanceto
Humankind
Levels-of.OrganizationHierarchy
The EmergentPropertyPrinciple
TranscendingFunctionsand Control
Pro@sses
EcologicalIntertacing
About Models
DisciplinaryReductionismto Transdisciplinary
Holism

rl

1 Ecology:
HistoryandRelevance
to llumankind
The word ecologiis derived from rhe Greek oihos,meaning "household,"and logos,
meaning "study."Thus, the study of the environmentalhouseincludes all the organisms in it and all the functional processesthat make the house habitable. Literally,
then, ecology is the study o[ "life at home" with emphasison "the totality or pattern of relationsbetween organismsand their environment," to cite a standarddictionary definition of the word (Merriam-Webster's
CollegiateDictionory,1Othedition,
s.v. "ecology").
The word economics
is also derived from the Greek root oikos.As nom[s means
"management,"economicstranslatesas "the managementof the household" and, accordingly, ecology and economicsshould be companion disciplines.Unfortunately,
many people view ecologistsand economistsas advercarieswith antitheticalvisions.
Table 1-1 attempts to illustrate perceiveddifferencesbetween economicsand ecology. l-ater, this book will considerthe confrontation that resuhsbcauseeach discipline takesa narrow view of its subjectand, more important, the rapid development
of a new interfacediscipline, ecoiogical
economics,
that is begnmng to bridge the gap
betweenecologyand economics(Costanza,Cumberland,et al. I997; Barrettand Farina 2000;L. R. Brown 2001).
Ecologlrwas of practical interestearly in human history. In primitive society,all
individuals neededto know their environment-that is, to understandthe forcesof
nature and the plants and animals around them-to survive. The beginning of civilization, in fact, coincided with the use of fire and other tools to modify the environment. Becauseo[ technologicalachievements,humans seem to depend less on the
natural environment for their daily needs;many of us lorget our continuing dependence on nature for air, water, and indirectly, food, not to mention wasteassimilation, rccreation,and many other srvicessupplied by nature.Also, economicsystems,
of whatever political ideology, value things made by human beings that primarily
benefit the individual, but they placelittle monetary value on the goodsand services
of nature that benefit us as a society.Until there is a cdsis, humans tend to uke nat-

Iable 1-1

l"
t-

A summaryof perceiveddifferencesbetweeneconomicsand ecology

Atttibute

Economics

Ecologl

Schoolof thought

Cornucopian

Neo-Malthusian

Currency

Money

Energy

GroMh form

J-shaped

S-shaped

Selectionpressure

r-selected

K-selected

Technological
approach

Hightechnology

Appropriatetechnology

Systemservices
Resourceuse

Servicesprovided
by economiccapital
Linear(disposal)

Servicesprovided
by naturalcapital
Circular(recycling)

Systemregulation

Exponential
expansion

Carryingcapacity

Futuristicgoal

Exploration
and expansion

Sustainability
and stability

SECTION1

Ecology:Historyand Relevance
to Humankind 3

Figute 1.1. Earthscape

as viewedfromApollo17 traveling towardthe Moon.Mew of the ecospherefrom "outside
the boxl'

ps,
1n-

lry,
ictn,

ac:l)',
n5.
!rlcllnt

lap

rlL
of
ti-

,nhe
tnla:ts,
rl)
ies
.1t-

dy

ural goods and servicesfor granted; we assumethey are unlimited or somehow replaceableby technologicalinnovations, even though we know rhar life necessities
such as oxygen and water may be recyclablebut nor replaceable.As long as the lifesupport seruicesare consideredfree, they have no value in current market systems
(seeH. T. Odum and E. P Odum 2000).
Like all phasesof learning, the scienceof ecologyhas had a gradual if spasmodic
development during recorded history. The writings of Hippocrates,Aristotle, and
other philosophersol ancient Greececlearly contain referencesto ecologicaltopics.
However, the Greeksdid not have a word for ecology.The word ecolog,,is o[ recenr
origin, having been frrst proposed by the Grman biologist Ernst Haeckel in 1869.
Haeckel defined ecolosl as "the study of the natural environmem including rhe relations of organismsto one another and to their surroundings" (Haeckel 1869). Before this, during a biological renaissancein the eighteenthand nineteenthcenturies,
many scholarshad contributed to the subject,even though the word ecologrwas not
in use. For example,in the early 1700s,Antoni van Leeuwenhoek,best known as a
premier microscopist,also pioneeredthe study of food chains and population reguIation, and the writings of the English botanist Richard Bradley revealedhis understanding of biological productivity. All three of thesesubjectsare important areasof
modern ecology.
As a recognized,distinct field ofscience,ecologydatesfrom about 1900, but only
in the past few decadeshas the word becomepart of the generalvocabulary.At first,
the field was rather sharply divided along taxonomic lines (such asplant ecologyand
animal ecologl), but the biotic community concept of Frederick E. Clementsand
Victor E. Shelford,the food chain and material cycling conceptsof Ra1'rnondLindeman and G. Evelyn Hutchinson, and the whole iake studles o[ Edward A. Birge and
ChauncyJuday,amongothers,helped establishbasictheory for a unified 6eld ofgeneral ecology.The work of thesepioneerswill be cited often in subsequentchapters.
What can best be describedas a worldwide environmentalawarenessmovement
burst upon the sceneduing two years, 1968 lo 1970, as astronautstook the first
photographsof Earth as seen from outer space.For the first time in human history,
we were able to seeEanh as a whole and to realizehow alone and fragile Earth hovers in space(Fig. I-1). Suddenly,during the 1970s,almosteveryonebecamecon-

C H A P T E R1

The Scope of Ecology

cemed about pollution, natural areas,population growrh, food and energ/ consumption, and biotic diversity, as indicated by the wide coverageof environmental
concernsin the popular press.The 1970swere frequently referred to as the "decade
of the enyironment,"initiated by the first "Earth Day" on 22 April 1970. Then, in the
1980s and 1990s, environmental issueswere pushed into the political background
by concerns for human relations-problems such as crime, the cold war, government budgets,and welfare.As we enter the early stagesof the twenty-first century,
environmental concelns are again coming to rhe forefront becausehuman abuseo[
Earth continuesto escalate.We hope that this time, to usea medical analogy,our emphasiswill be on prevntionrather than on treatment,and ecologyas outlined in this
book, can contibute a great deal to prevention technology and ecosystemhealth
(Barrett2001).
The increasein public attention had a profound effecton academicecology.Before the 1970s, ecologywas viewed largely as a subdiscipline of bioiogy. Ecologisrs
were staffedin biology depanmenrc,and ecologycourseswere generallyfound only
in the biological sciencecurricula. Although ecology remains strongly rooted in biology, it has emergedfrom biologr as an essemiallynew, inregrarivediscipline thar
links physical and biological processesand forms a bridge berweenrhe natural sciencesand the social sciences(E. P Odum 1977). Most collegesnow offer campuswide coursesand haveseparatemajors,departments,schools,centers,or inslitutes of
ecology.While the scope of ecology is expanding, the study of how individual organismsand speciesinterfaceand useresourcesintensifies.The multilevel approach,
as outlined in the next section,brings together "evolutionary" and "systems"thinking, two approachesthat have tended to divide the field in recentyears.

2 levels-of-0rganizationHierarchy
Perhapsthe best way to delimit modern ecologyis to consider the concept o[ levels
of organization, visualizedas an ecologicalspectrum (Fig. 1-2) and as an extended
ecologicalhierarchy (Fig. 1-3). Hierarchy means "an arrangementinto a graded
setes" (Merriam-Webster's
CollegieteDicfionary,10th edition, s.v. "hierarchy"). Interaction with the physical environment (energy and matter) at each level produces
characteristicfunctional systems.A system, accordingto a standarddefrmuon, consists of "regularly interacting and interdependent components forming a unified

BIOTICCOMPONENTS
plus

Genes

t{

ABIOTICCOMPONENTS

Cells

++

t+

Organisms

t+

Matter

equals
BIOSYSTEMS

Organs

Genetic
systems

Cell
syslems

Populations

t+

Communities

r{

Energy
[||
Organ Organismic
systems
systems

Population
systems

Ecosystems

Figure 1-2. Ecological

levels-ot-organization
spectrumemphasizing
the interaction
of living
(biotic)and nonliving(abiotic)components.

r
SECIION2
ntal
de
he
rd

Figure 1-3. Ecologicallevelsof-organization

hierarchy;seven
transcending
processesor lunctions are depicted as vertical
componentsof elevenintegrative levelsof organization
(afier
Barrettet al. 1997).

Energetics
Evolution
Development
Regulation

Levets-of.Organization
Hierarchy 5

Behavior
Diversity
Integration

of
Tl'll5

rh
letli..

ut
;l-

ts..f
:h.
.k-

:ls
ed
ed
es
d

whole" (Meriam-Webster'sCollegiateDictiondry,l0th edirion, s.v .system").Systems

containing living Giotic) and nonliving (abioric) componentsconstirure biosrstems,
ranging from genetic systemsto ecologicalsysrems(Fig. I-2). This specrmm may
be conceivedof or srudied at any level, as illusrrated in Figure 1-2, oi at any inter_
mediateposition convenientor practical for analysis.For example,host-parasiresys_
tems or a two-speciessystemof mutually linked organisms(such as the fungi_algae
partnership that constitutesthe lichen) are intermediatelevels between oooulation
and community.
Ecologyis largely,bur not entirely,concernedwith the systemlevelsbeyond that
of the organism (Figs. 1-3 and 1-4). ln ecology, rhe rerm population, originally
coined to denotea group of people,is broadenedro include groups ofindividuals of
any one kind o[ organism.Likewise,cornmunity, in rhe ecologicalsense(sometimes
deslgnatedas "biotic community"), includes all the populations occupying a given
area. The community and the nonliving environment function together as an eco_
logical systemor ecosystm. Biocoenosis
(lireralt, ..lfe and Earth
andbiogeocoenosis

C HAPTER 1

The Scope of Ecology

Figure 1.4. Comparedwiththe strongset-pointcontrols

at the organismlevelandbelow,organization
andlunctionat
the populationleveland aboveare muchlesstightlyregulated,with morepulsingand chaoticbehavio(but theyare
controllednevertheless
by allernatingpositiveand negative
feedback-inotherwords,they exhibithomeorhesrs
as opposedto homeostasls.
Failureto recognize
thisdifferencein
cyberneticshas resultedin muchconfusionaboutthe balanceof nature.

Ecosphere

Biomes

Lanoscapes

t
t
Communities
t
Populations
t
OFGANISM
Ecosystems

No sefpointcontrols
(+ and -)
feedback
maantaanjng
pulsingstates
withinlimits
HOMEORHESIS

Organsystems

I
I
Tissues

.t

Sel-pointcontrols
(+ and-)
feedback
maintaining
steadystates
withinlimits
HOMEOSTASIS

Cells

Molecules

Atoms

functioning together"),terms frequentlyusedin Europeanand Russianlirerarure,are

roughly equivalent to community and ecosystem,respectively.Refe[ing again to
Figure 1-3, the next level in the ecologicalhierarchy is the landscape,a term originally referring to a painting and defined as "an expanseof sceneryseenby the eyeas
one view" (Mrridm-Webster's
Colleglo.te
Dictionary,lOth edition, s.v."landscape").ln
ecology,landscape is defrnedas a "heterogenousarea composedo[ a cluster of interactingecosystemsthat are repeatedin a similar manner throughout" (Forman and
Godron 1986). A wctershedis a convenientlandscapeJevelunit for large-scalestudy
and managementbecauseit usually has identifiable natural boundaries. Biomeis a
term in wide use for a largeregionalor subcontinentalsystemcharacterizedby a major vegetationt)?e or other identifying landscapeaspect,as, for example,the Temperate Deciduous Forestbiome or the Continental Shelf Oceanbiome. A reglonis a
large geologicalor political area that may contain more than one biome-for example, the regionsof the Midwest, the AppalachianMountains, or the PacificCoast.
The largestand most nearly self-sufficientbiological sysremis ofren designatedasrhe
ecosphere, which includes all the living organismsof Earth interacting with the
physicalenvironmentasa whole to maintain a self-adjusting,looselycontrolledpulsing state(more about the concept of "pulsing state"Iater in this chapter).
Hierarchicaltheory providesa convenientframework for subdividing and examining complex situationsor extensivegradients,but it is more thanjust a useful rankorder classification.lt is a holistic approachto understandingand dealingwith com-

SECTION 3

The Emergent property principte

piex siruations,and is an alternativeto the reductionist approachof seekrnganswers

by reducing problems ro lower-levelanalysis(Ahl and Allen 1996).
More than 50 yearsago,Novikoff (1945) poirued out that rhereis both continuity and discontinuity in the evolurion of the universe.Developmentmay be viewed
as continuous becauseit involvesnever-endingchange,but it is also drsconLtnuous
becauseit passesrhrough a seriesof different levelsof orqanizarion.As we shall dis_
cussin Chapter 3, the organizedstateo[ Ii[e is mainraineJ by a conrinuousbut srepwise flow ofenergy. Thus, dividing a gradedseries,or hierarchy,into componentsis
in many casesarbitrary, but sometimessubdivisionscan be basedon natural discon_
tinuities. Becauseeach level in the levels-of-organization
spectrum is ,.integrated"or
interdependentwith other levels,there can be no sharp lines or breaksin a functional
sense,nor even betweenorganismand population. The individual organism,for ex_
ample,
survive for long without its population, any more than the organ
-cannot
would be able [o survive for long as a self-perpetuaringunit without its organlsm.
Similarly, the community cannot exist wirhour rhe cyclinq of materialsand the flow
of energyin rhe ecosysrem.This argumenris applicablet"orhe previously discussed
mistaken notion that human civilization can exist separatelyfrom the natural world.
It is very rmporrantto emphasizethar hierarchiesin natureare nested-Lharts,
each level is made up of groups of lower-level unirs (populations are composedof
groups of organisms,for example). ln sharp contrast,human-organizedhierarchies
rn Sovernments,cooperations,univercities,or the military are nonnested(sergeants
are not composedofgroups ofprivates, for example).Accordingly,human-organized
hierarchiestend ro be more rigid and more sharply separatedii compared ro natu_
ral levelso[ organization.For more on hierarchrcaltheory, seeT. F. H. Allen and Starr
(1982),O'Neill et al. (1986),and Ahl and Allen (1996).

re
:o
tts
.n
'l-

.d
l;'
a
tl-

a
i-

t.
LE

te
tl-

(l-

3 TheEmergent
Propefi Principte
Al important consequenceof hierarchical organization is that as componems, or
subsets,are combined to produce larger functional wholes, new propenies emerge
that were not preseni at the level below Accordingly, an emergent property of in
ecologicallevel or unit cannot be predicted from the study of the componentsof that
level or unit. Another way to expressrhe sameconcept rs nonreducibie propenythat is, a property ofthe whole not reducibleto the sum ofrhe propeniesof the parts.
Though findings at any one level aid in the study of the next level, they never com_
pletely explain the phenomenaoccurring ar the nexr level,which musr itselfbe studied to complete rhe picrure.
Two examples,one from the physical realm and one from rhe ecologicalrealm,
will sufficeto illusrrate emergentproperries.When hydrogen and oxygen are combined in a certain molecular configuration, water is formed-a liquid wirh properties utterly different from those of its gaseouscomponents.When certain algaeand
coelenterateanimals evolvetogetherto produce a coral, an efficientnutrientiycling
mechanismis createdthat enablesthe comblned system to maintain a high rate of
productivity in watercwith a very low nutrient content. Thus, the fabulousproduc-

CHAPTER 1

The Scope of Ecology

tivity and diversity of coral reefsare emergentpropenies only at the level of the reef
community.
Salt (1979) suggestedthat a distinction be made betweenemergempropeties, as
defined previously,and collective properties, which are summationsof the behavior of components.Both are propertiesof the whole, but the collectivepropertiesdo
not lnvolve new or unique characteristicsresuking from the functioning of the whole
lniL Birth rateis an exampleof a population level collectiveproperty, as it is merely
a sum of the individual births in a designaiedtime period, expressedas a lraction or
percentof the total number of individuals in the population. New propertiesemerge
becausethe componentsinteract, not becausethe basic nature of the componentsis
changed.Partsare not "melted down," as it were, but integratedto produce unique
new properties. lt can be demonstratedmathematicallythat integrativehierarchies
evolvemore rapidly from their constituentsthan nonhierarchicalsystemswith the
same number of elements;they are also more resilient in responseto disturbance.
Theoretically,when hierarchiesare decomposedto theirvarious levelsoIsubsysrems,
the latter can still interac[ and reorganizeto achievea higher level of complexity.
Someattributes,obviously,becomemore complex and variableas one proceeds
to higher levelsof organization,but often other attributesbecomelesscomplex and
lessvariableas one goesfrom the smaller to the larger unit. Becausefeedbackmechanisms(checksand balances,forcesand counterforces)operatethroughout, the amplitude ol oscillations tends to be reduced as smaller units function within larger
units. Statistically,the va ance of the whole system level property is less than the
sum of the varianceof the parts. For example,the rate of photosynthesisof a foresr
community is lessvariablethan that of individual leavesor treeswithin the community, becausewhen one component siows down, another component may speedup
to compensate.When one considersboth the emergentpropenies and the increasing
homeostasisthat develop at each level, not all component parts must be known before the whole can be understood.This is an important point, becausesome contend
that it is uselessto try to work on complex populations and communitieswhen the
smaller units are not yet fully understood. Quite the contrary, one may begin study
at any point in the spectrum, provided that adjacentlevels, as well as the level in
question,are considered,because,as alreadynoted, some attributes are predictable
from parts (collectiveproperties),but others are no[ (emergentproperties).ldeally, a
system-levelstudyis itselfa threefoldhierarchy:system,subsystem(next levelbelow),
and suprasystem(nextlevelabove).Formore on emergenlproperties,seef. F. H. AIIen
and Starr(1982),T. F. H. Allen and Hoekstra(1992),and Ahl and Allen (1996).
Each biosystemlevel has emergentproperties and reduced varianceas well as a
summation of attributes of its subsystemcomponents.The folk wisdom about the
forest being more than just a collection of treesis, indeed, a first working principle
of ecology.Although the philosophy of sciencehas alwaysbeenholistic in seekingto
understandphenomenaasa whole, in recentyearsthe practiceof sciencehasbecome
increasinglyreductionist in seekingto understand phenomenaby detailed study of
smallerand smaller componenls.LaszLoand Margenau(1972) describedwithin the
history o[ sciencean alternation of reductionist and hoiistic thinking (redi{ctiorxism
constructionism
and atomism-holism
are other pairs of words used to contrast these
philosophicalapproaches).The law of diminishing retums may very well be involved
here, as excessiveellort in any one direction eventuallynecessitates
taking the other
(or another) direction.
The reductionistapproachthat hasdominatedscienceand technologysincelsaac

r
SECTION 4

ref
as
lc-t
rle
:\'
Llf

is
ue
e5
he
te.
1S,

rd
'r'the

Transcending Functions and Control Processes 9

Newton hasmade major contdbutions. For example,researchat the cellular and molecular levelshas establisheda firm basis for the future cure and prevention of cancersat the level of the organism.However,celllevel sciencewiil contribute very little
to the well-being or survival of human civilization if we understandthe higher levels
of organization so inadequatelythat we can hnd no solutions to population overgrowth, pollution, and other forms ofsocietal and environmentaldisorders.Both holism and reductionism must be accordedequal value-and simultaneously,not altematively (E. P Odum 1977; Barrett 1994). Ecologyseek slnthesis, not separation.
The revival of the holistic disciplinesmay be due at leastpartly to citizen dissatisfaction with the specializedscientistwho cannot respond to the large-scaleproblems
that need urgent attention. (Histodan L),nn White's 1980 essay"The Ecology of Our
Science"is recommendedreadlng on this viewpoint.) Accordingly, we shall discuss
ecologicalprinciples at the ecosystemlevel, with appropiate attention to organism,
population, and community subsetsand to landscape,biome, and ecospheresuprasets.This is the philosophical basisfor the organizationo[ the chaptersin this book.
Fortunately,in the past I0 years,technologicaladvanceshaveallowed humans to
deal quantitativelywith large,complex systemssuch as ecosystemsand landscapes.
Tracer methodology,masschemistry (spectrometry,colorimetry, chromatography),
remotesensing,automadcmonitoring, mathematicmodeling, geographicalinformation systems(GlS), and computer technologyare providing the tools. Technologyis,
of course,a double-edgedsword: it can be the means of underctandingthe wholenessof humans and nature or of destroyingit.

rp
le-

rd
he
lv

in
rle
.a

..),
an
ia

he
rle
to
ne
oI
ne
n5e
:d
ET

4 Transcending
Functionsand ContrulProeesses
Whereaseachlevel in the ecologicalhierarchy can be expectedto haveunique emergent and collectiveprope ies, there arc basic functions that operateat a]l levels.Examplesof such transcending functions are behavior,development,diversity, energetics,evolution, integration,and regulation (seeFig. l-3 for details).Someo[these
(energetics,for example)operatethe samethroughout the hierarchy,but others differ in modusoperandiat differentlevels.Natural selectionevolulion, for example,involvesmutations and other direct geneticinteractionsat the organismlevel but indlrect coevolutionaryand group selectionprocessesat higher levels.
It is especiallyimpoftant to emphasizethat although positive and negativefeedback controls are universal,from the organismdom, control is sel?oirlf,in that it involvesvery exactinggenetic,hormonal, and neural controls on growth and development, Ieadingto what is often called homeostasis. As noted on the right-hand side
of Figure 1-4, there are no set-pointcontrolsabovethe organismlevel (no chemostats
or thermostatsin nature). Accordingly, feedbackcontrol is much looser,resulting in
pulsing rather than sieady states.The term homeorhesis, from the Greek meaning
"maintaining the floq" has been suggestedfor this pulsing control. ln other words,
there are no equilibriums at the ecosystemand ecospherelevels,but there arepulsing
bclances,such asbetweenproduction and respirationor betweenoxygenand carbon
(the scidioxide in the atmosphere.Failure to recognizethis differencein cybernetics
encedealingwith mechanismsof control or regulation)has resultedin much confusion about the realitiesof the so-called"balanceof nature."

10 CHAPTER1

TheScopeot Ecology

5 Ecologicallnterfacing
Becauseecology is a broad, muLtileveldiscipline, it inrerfaceswell with lraditional
disciplinesthat tend to have more narrow focus. During the past decade,there has
been a rapid rise of interfacefields of study accompaniedby new societies,journals,
symposium volumes, books-and new careers.Ecological economics,one of the
most important, was mentionedin the frrst sectionin this chapter.Others that are receiving a greatdeal oI attention,especiallyin resourcemanagement,are agroecology,
biodiversity,conservationecology,ecologlcalengineering,ecosystemhealth, ecotoxicology,environmentalerhics,and restorationecology.
In the beginning, an interface effort enriches the disciplines being interfaced.
Lines of communication are established,and the experliseo[ narrowly trained "experts" in each field is expanded.However, for an interfacefield ro becomea new discipline, somethingnew hasto emerge,such asa new conceptor technology.The concept oInonmarket goodsand services,for example,was a new concept that emerged
in ecologicaleconomics,but that inidally neither rraditional ecologistsnor economists would put in their textbook (Daily 1997; Mooney and Ehrlich 1997).
Throughout lhis book, we will reler to natural capital and economiccapital. Natural capital is defined as the benefits and servicessupplied to human societiesby
natural ecosystems,or provided "free of cosf'by unmanagednatural systems.These
benehtsand servicesinclude purificadon oI water and air by natural processes,decomposition of wastes,maintenanceof biodiversity, control of insect pests,pollination o[ crops, mitigation of floods, and provision o[ natural beauty and recreation,
amongothers(Daily 1997).
Economic capital is defined as the goods and servicesprovided by humankind,
or the human workforce, tlpically expressedas the gross national product (GNP).
Gross national product is the total monetary value o[ all goods and serl,rcesprovided in a country during one year. Natural capital is typically quantifred and expressedin units of energy,whereaseconomic capital is expressedin monetary units
(Table 1-1). Only in recent years has there been an attempt to value the world's
ecosystemservicesand natural capital in moneury terms. Costanza,d'Arge, et al.
(1997) estimatedthis value to be in the rangeof 16 to 54 trillion U.S.dollarsper year
for rhe enrire biosphere,with an averageof 33 triliion U.S. dollars per year. Thus it is
wise to protect natural ecosystems,both ecologicallyand economically,becauseof
the benehtsand servicesthey provide to human societies,aswill be illustrated in the
chantersthat follow.

6 AboutModels
Ifecology is to be discussedat the ecosystemlevel, for reasonsalreadyindicated,how
can this complex and formidable systemlevel be dealt with? We begn by describing
simplifred versionsthat encompassonly the most important, or basic,propertiesand
functions. Because,in science,simplified versionsof the real world are called models,
it is appropriatenow to introduce this concept.
A model (by definition) is a formulation that mimics a real-world phenomenon

S E C T I O N6

^l
15

le

e-

),
\d.
xt5-

ed
o-

rtbt
:5e

letarn,
rd,
P).
:o:x.rt5

d's
aL.
tar

of
.he

rn8
.nd
ais,

AboutModels 11

and by which predictionscan be made.ln their simplestlorm, modelsmay be verbal

or graphic (infotmal). Ultimately, however, models musr be sraristicaland mathematical Uormdl) i[ their quantitative predictions are ro be reasonablygood. For example, a mathematicalformulation that mimics numerical changesin a population
of insectsand that predicts the numben in the population at some time would b
considereda biologically useful model. lf the insectpopulation in questionis a pest
species,the model could have an economicallyimportant application.
Computer-simulatedmodelspermit one to predict probableoutcomesasparametersin the model are changed,as new parametersare added, or as old ons are removed. Thus, a mathematicalformulation can often be "tuned" or refined by computer operations to improve the "nC'to the real-world phenomenon. Above all,
models summarizewhat is understood about the situation modeled and therebydelimit aspectsneedingnew or better data, or new principles. When a model does not
work-when it poorly mimics the real world-computer operadonscan often provide cluesto the refinemenmor changesneeded.Once a model proves to be a useful
mimic, opponunities for experimentationare unlimited, becauseone can introduce
new factorsor perturbationsand seehow they would affectthe system.Even when a
model inadequatelymimics the real world, which is often the casein its early stages
of development,it remainsan exceedinglyuseful teachingand researchtool if it revealskey componentsand interactionsthat merit specialattention.
Contrary to the feelingof many who are skepticalabout modeling the complexity of nature, information about only a relativelysmall number ofvariables is ofren a
sufficient basisfor effectlvemodels becausekey factors,or emergentand orher imegrative properties,as discussedin Sections2 and 3, often dominate or control a large
percentageofthe action.Watt (1963), for example,stated,"We do not needa rremendous amount o[ information about a great many variablesto build revealingmathe
matical models."Though the mathematicalaspectsof modeling are a subject for advancedtexts,we should review the first stepsin model building.
Modeling usually beginswith the construction of a diagram, or "graphic model,"
which is often a box or companmenLdiagram,as illustrated in Figure 1-5. Shown are
two properties, P1and P2,that interact, I, to produce or alfect a third prope y, P],
when the systemis driven by an energysource,E. Five flow pathways,F, are shown,
with F1 representingthe input and F6 the output for the systemas a whole. Thus, at
a minimum, there are five ingredientsor componentsfor a working model of an ecologrcalsituation, namely,(l) an energ) sourceor other outside forcing function, E;
(2) propertiescalled state variables, P1,Pr, . . . P"; (3) flow pathways, Fr, F2, . . . Fi,
showing where energyflows or material transfersconnectpropertieswith each other
and with forces;(4) interaction functions, I, where forcesand propertiesinteract to
modify, amplify, or control llows or createnew "emergenC'properties;and (5) feedback loops, L.
Figure 1-5 could serveas a model for the production of photochemicalsmog in
the air over Los Angeles.ln this case,P1could representhydrocarbonsand P2nitrogen oxides,two products of automobileexhaustemission.Under the driving force of
sunlight energy,E, rheseinteract to produce photochemicalsmog, P1.In this case,
the interaction function, l, is a slrrergisticor augmentativeone, in that P3is a more
seriouspollutant for humans than is P1or P2acting alone.
Alternatively,Figure 1-5 could depict a grasslandecosystemin which P1representsthe greenplanE that convet the energyof the Sun, E, to food. P2might representa herbivorousanimal that eatsplants, and P:lan omnivorous animal that can eat

1 2 C H A P T E R]

The Scope of Ecology

Figure 1-5. Compartment

diagramshowingthefjvebasjccomponents
of primaryInterestrn
modelingecological
systems.
E : energysource(forcingfunction);pj, p2,p3= statevariables;
F1-F6= flow pathways;
| : interaction
function;L = feedbacklooo.

eitherthe herbivores
or rhe plants.ln rhiscase,the interactionfunction,I, could representseveralpossibilities.
Ir could be a no_preference
swirchi[ observntron
in the
realwotld showecL
that the omnivorep, eatseitherpr or pr, accordingto availability.
Or I could be specifiedto be a constantpercentage
valuei[ it wasfound that rhe dier
o[ P, rvascomposeclof, say, 80 percent plant and 20 percent animal narrer, rrre_
specrivco[ the srateo[ p, or pr. Or I cou]d be a seasonai
switchi[ pr feedson plants
during one part of the yearancLon animalsduring anorherseason.Or I could be
a
thresholdswitch il P, greatlyprelcrsanimalfood and switchesto plantsonly when
P, is rcducedto a low level.
Fccclbcrch
loopsare imporLant learuresoI ecologicalmodels becausethey repre_
sentcontrolmechanisms.
Figure l-6 is a simplilieddiagramoIa systemth fearures
a feedbackloop in which "downstream,,
ourput,or so;e part o[ ii. is fed backor .e
cycleciro affector perhapsconrrol"upstream.,
.o.pon.nrs. For exrmple,the feed_
back loop could representpredalion by 'downsrream',organisms,C, that reduceand
thereby rend to control the grouth o[,,upstream,'herbivoresor plants B and A in the
fooclchain. Often. such a feedbackacruallypromotesthe growth or survivalof
a
downstrcamcol]tponent,such as a grazerenhancingfhe growth of plants (a .,reward
feedback."
asit rvere).

Figure 1-6. Compartmentmodelwith afeedbackorcontrol loop that transformsa linear system into a partiallycyclical one.

Feedback loop

SECTI0N 6

AboutModels 13

Figute 1.7. Interactionof positive

and negativefeedbacksin the relationships of atmosphericCOr, climate
andcarbonse'
warming,soilrespiration,
questration(modifiedafter Luo et al.
2001).

-->

trn

:phe
t\''
let

re'lt5
la
c'n

Negative

Figure 1-6 could also representa desirableeconomicsystemin which resources,

A, are convertedinto useful goods and services,B, with the production ofwastes,C,
that are recycledand used again in the conversionprocess(A --+ B), thus reducing
the waste output of the system.By and large, natural ecosystemshave a circular or
loop design mther than a linear structure. Feedbackand cybernetics,the scienceof
controls, are discussedin detail in Chapter 2.
Figure 1-7 illustrateshow positive and negativefeedbackcan interact in the relationship betweenatmosphericCO2concentrationand climatic warminS. An increase
in CO2 has a positive greenhouseeffect on global warming and on plant growth.
However,the soil systemacclimatesto the warming, so soil resPirationdoesnot continue to increasewith warming. This acclimation results in a negativefeedbackon
carbon sequestrationin the soil, thus reducing emissionof CO2 to the atmosphere,
accordingto a study by Luo et al. (2001).
Compartment models are greatly enhancedby making the shape of the "boxes"
indicate th general function of the unit. tn Figure 1-8, some of the symbols from
the H. T. Odum energylanguage(H. T. Odum and E. P Odum 1982; H. T. Odum
1996) are depicted as used in this book. ln Figure l-9, theses).tnbolsare used in a
model of a pine lorest located in Florida. Also, in this diagram estimatesof the
amount of energ'yflow through the units are shown as indicators of the relativeimportanceof unit functions.
ln summary, good model definition should include three dimensions: (I) the
spaceto be considered(how the system is bounded); (2) the subsystems(components)judged to be important in overall function; and (3) the lime interval to be considered. Once an ecosystem,ecological situation, or problem has been properly
dehnedand bounded, a testablehlpothesis or serieso[ hlpotheses is developedthat
can be rejectedor accepted,at leasttentatively,pending further expedmenLationor
analysis.For more on ecologicalmodeling, seePattenand Jorgensen(1995), H. T.
Odum and E. C. Odum (2000), and Gundersonand Holling (2002).
ln the following chapten, the paragraphsheadedby the word statement are, in
effect,"word" models of the ecologicalprinciple in question.ln many cases,graphic
models are also presented,and in some cases,simplified mathematical formulations are included. Most ofall, this book attemptsto provide the principles, concepts,

14 CHAPTER
1 The Scope of Ecology

Energy
circuit
(A pathwayor
flowof energy)

Energy
source
(Sourceol energyfrom
outsidethesystem)

Consumer
(Usesproducer
energy
for self-maintenance)

-\

t--_r
-Y
----\.

-.L
--:

Heatsink
(Degradedenergy
afterusein work)

Producer
(Converts
and concentrates
sorarener9yl

t-

Slorage
{A compartmentoJ
energy storage)

-v-*

$.-.,.4\--

lnteraclron
(Twoor moreflows
of energyto producea
high-quality
energy)

Capitaltransaction
(Fow of moneyto
payfor flowof energy)

Figule 1-8, TheH. T.Odumenergylanguagesymbolsusedin modeldiagramsin this book.

Figure 1-9. Ecosystemmodel using

energylanguagesymbolsand including
estimatedratesof energyflowfor a Florida pineforest(courtesyof H. T. Odum).

Heatsink(usedenergy)

SECIION 7

DisciplinaryReductionismto Transdisciptinary
Hotism 1O

simplifications,and absrractionsthat one must deducefrom the realworld beforeone

can understand and deal with situations and problems or construct marhematical
models of rhem.

7 llisciplinarylieductionismto Tlansdisciptinary
Holism
a gaper entltled "The Emergenceof Ecologr as a New lmegrarive Discipline,,,
ln
E. P Odum (1977) noted that ecologyhad becomea new holistii discipline, having
roots in the biologicai, physical,and social sciences,rather than jusr a iubdiscipline
ofbiology. Thus, a goal of ecoiogyis to link the natural and sociaisciences.It should
be noted that most disciplinesand disciplinary approachesare basedon increased
specializationin isolation (Fig. l-10). The early evoiution and developmentofecol_

Figure 1.10. Progression

ol relationsamongdisciplines
fromdiscjplinary
reductionism
to transdisciplinary
holism(afterJantsch
1972).
)k.

DISCIPLINARY
specialzingin isolation

tI

MULTIDISCIPLINARY
no cooperation

CROSSDISCIPLINARY
rigid polarizationtoward
specific monodisciplinary
concepl

INTERDISCIPLINARY
coordination
by higher
levelconcept

TRANSDISCIPLINARY
multi-levelcoordinationof
entire education/ innovation
systern

16 CHAPTER 1

The Scope of Ecology

: "many"), espeogli was frcquently basedon multidisciplinary approaches(multi

cially during the 1960s and 1970s.Unfonunately, the multidisciplinary approaches
lacked cooperation or focus. To achievecooperation and define goals,insliutes or
centerswere establishedon campusesthroughout the world' such as the lnstilute of
Ecologylocatedon the campusof lhe University ofGeorgla Thesecrossdisciplinary
uppro".h", (ctott = "traverse":Fig. 1-10) frequently resultedin polarization toward
a specificmonodisciplinary concept,a poorly funded administrativeunit, or a narrow mission. A crossdisciplinaryapproachalso frequently resultedin polarized facuhy reward systems.Institutions of higher leaming, traditiona\ built on disciplinary
structures, have diffrculties in administering programs and addressi.ngenvironmental problemsaswell astaking advantageofopponunities at greatertemporal and spatial scales.
= "among') were
To addressthe dilemma, interdisciPlinary approaches(inter
employed,resultingin cooperationon a higher-levelconcept,problem, or question'
t'oi example, the process and study of natural ecological successionprovided a
higher-level concept resulting in the successof the SavannahRiver Ecological Labotheorized that new systemProperratory (SREL)during its conception.Researchers
and that it is theseproperdevelopment
of
ecosystem
the
course
during
ties emerge
that occur (E P Odum
form
changes
growth
and
for
species
account
ties that largely
approachesare cominterdisciplinary
Today,
for
details).
8
1969,1977, seeChapter
global
levels'
and
landscape,
at
ecosystem,
problems
mon when addressing
problems,
need
to
solve
increased
is
an
There
however.
be
done,
Much remainsto
manin
a
transdisciplinary
resources
manage
literacy,
and
promote environmental
innovation
and
educalion
entire
involves
approach
large-scale
ner. This multilevel,
systems(Fig. I-10). Thil integrativeapproach to the need for unlocking cause-and#ect explanations across and among disciplines (achieving a transdisciplinary understanding) has been termed consilience(E. O Wilson 1998), sustainability scien'e
(Barrett2001) Actually, the continued de(Kateset al. 2O0l), and integrativescietLce
(the
"study
of the household"or''place where we
velopmentofthe scienceofecology
science of the future This
integrative
much-needed
into
that
live';) will likely evolve
and approachesto
principles,
concepts,
knowledge'
book attempts to provide the
process
learning
and
underpin this educationalneed