Nordic Society Oikos

Population: A Central Concept for Ecology?

Author(s): Alan A. Berryman
Reviewed work(s):
Source: Oikos, Vol. 97, No. 3 (Jun., 2002), pp. 439-442
Published by: Wiley on behalf of Nordic Society Oikos
Stable URL: http://www.jstor.org/stable/3547665 .
Accessed: 01/01/2013 14:45

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Wiley and Nordic Society Oikos are collaborating with JSTOR to digitize, preserve and extend access to Oikos.

http://www.jstor.org

This content downloaded on Tue, 1 Jan 2013 14:45:17 PM

All use subject to JSTOR Terms and Conditions
FORUM FORUM is intended fornewideashypotheses
providesa chancefor suggesting
or newways
andoffor
interpreting
existing
challenging information.
currentthinking It
on
FORUM ecologicalissues.A lighter
research
reports,
prose,designedto attractreaders,
albeitshort,willnotbe accepted,
willbe permitted.
Formal
shouldbe concise
and all contributions
FOR
va UMVl A summary
shortlistof references.
witha relatively is notrequired.

Population:a centralconceptfor ecology?

Alan A. Berryman, State Univ.,Pullman,WA 99164, USA

Dept of Entomology,Washington
(berrymanqmail.wsu.edu).

"To searchforthebestconceptis no idleconceit,because Is "population"the appropriate central

theexperiments thata scientist
maydeviseand therefore
thefactshe maydiscover, as wellas theexplanations
that concept for ecology?
he offers forthem,dependon howhe conceivesnature."
Andrewartha and Birch(1984) Some ecologistshave argued that the centralfocus in
ecologyshould be on the individualorganism(e.g., see
When mycolleague Mauricio Lima sentme the accom-
Lomnicki 1988, DeAngelis and Gross 1992). Afterall,
panyingpaper (Camus and Lima 2002) it rekindledmy
ecologicallandscapesare made up of individualsgoing
interestin an unfinishedmanuscriptthat attemptsto
about theirdailybusinessand,in so doing,affectingeach
addresssome of the questionsand problemstheypose.
other in a multitudeof ways. The problem is that,
The ideas presentedin thisessayare based on thepremise
althoughan individual-basedapproach may seem logi-
thatecologyneedsa centralunifying conceptupon which
cal, it is impractical.
to build a coherentand practicaldiscipline.I willargue
Firstthereis a computationalproblemfor,ifwe were
that,froma practicalpointof view,the "population" is
to treateveryorganismon a largelandscapeas a separate
themostlogicalchoiceforthatcentralconceptand that,
entity,withits own particularcharacterand behavior,
because of this, the concept requires a precise and
thenkeepingtrackof all the information, motionand
unambiguous meaning. This leads me to propose a
interaction wouldbe an impossibletask,evenformodern
definitionfor "population" that, I think,has these
supercomputers. Anotherproblemis that of measure-
qualities.
ment.In orderto predictthe futurestatesof ecological
systems,we mustfirstmeasuretheircurrentstates.An
individual-basedapproachwould requireus to measure
Does ecologyneeda centralunifying the state and location of everyorganismon the land-
concept? scape, an impossible task in most situations.Thus,
Perhaps I can approach thisquestionby askinganother though an individual-basedapproach may make sense,
- Does biology need a centralunifyingconcept?The it is not practical.This does not mean thatthe studyof
answer,I think,is "yes" becausemostofmodernbiology individual-basedmodels is useless,fortheymay illumi-
is based on thebiologicalconceptofa "species".Without nate importantecologicalprinciples,but ratherthatthe
the speciesconcept,it is difficult to imaginehow order individualapproach is not directlyapplicable to the
and understanding could have been broughtto thevast practicalproblemsof ecologicalestimationand predic-
panorama of life,fortaxonomy,systematics, evolution, tion. In addition,population-levelconceptsand models
and ecologyare all bound togetherby thisunifying idea. should,ideally,be deriveddirectlyfromindividual-based
The question is, if the species conceptis necessaryfor firstprinciples;e.g., Royama's (1992) derivationof the
biology, is it sufficient for ecology and, if not, what logisticmodelfroma spatiallyexplicitmodelof individ-
additionalconceptor conceptsare necessary?My con- ual interactions.
tentionis that ecologyneeds one additionalconcept - One way to make a sciencemorepracticalis to work
the population.I believe that this is necessarybecause at a higherlevel of abstraction.The level of ecological
ecologists need to know more than the presence or abstractionabove the individualis the population - a
absence of species on the landscape, theyalso need to groupof individuals of thesamespecies.Here we concern
know how abundant the species are and how their ourselves,not with individuals,but with the average
abundance changesin space and time. properties ofa groupofsimilar(butnotexactlythesame)

OIKOS 97:3 (2002) 439

This content downloaded on Tue, 1 Jan 2013 14:45:17 PM

All use subject to JSTOR Terms and Conditions
individualsbelongingto the same species. Although idea of a populationis sometimeslost completely.For
variabilityfromthis averagemay be describedstatisti- example,the most recentbook on this subjectdefines
cally,say by thevariancearound themean,theidentity the metapopulationas a group of local populations
and uniquenessof theindividualorganismis lost in the (Hanski and Gilpin 1999). However,Andrewarthaand
abstraction.We have sacrificedinformationabout the Birch (1984) previouslydefinedwhat they called a
individualin orderto develop a practicalscience,as is naturalpopulationin exactlythe same way. A groupof
also truefor the concept of species. From thisline of local populationscannot be, at the same time,both a
reasoning,an ecologicalsciencebased on theconceptof population and a metapopulation.Surelywe need to
populationwill be morerealisticthan one based on the understandwhat we mean by the term"population"
concept of communityor ecosystembecause these are beforewe can comprehendand defineits derivatives,
higherlevels of abstractionand, therefore, furtherre- "local" and "meta" population?
moved fromreality.Thus, forthosewho wishto apply Most definitions of populationincludethe notionof
ecologicalscienceto thesolutionof real problems,there "a group of individuals of the same species living
seems to be littlechoice but to accept the concept of togetherin a particularplace". But thereis considerable
populationas theirmodusoperandi. ambiguityabout what is meant by "place". At one
Building ecological science around the concept of extremeare those who define"place" by the natural
populationgreatlysimplifies its organization.Commu- boundaries to the distributionof the species (An-
nity,landscape and ecosystemare no longer separate drewarthaand Birch 1984), at the other those who
paradigms for ecology but merelydifferentways to leave it to the whimof the observer(Berryman1981,
describeand classifyassembliesof populations(or sub- Ricklefs1990). Althoughthe formerdefinition seemsa
populations).In thesame vein,it no longermakessense bit restrictive,
the latteris muchtoo vague fora funda-
to considerpopulationdynamicsas a particularway to mental concept. Yet ecologists, including this one
view ecology - it's the onlyway. Populationdynamics (Berryman1981),have been veryliberalin theiridea of
becomessynonymous withecologicaldynamics,withthe "place", and this,I believe,is one of the reasons for
variablesofecologicalchangebeingnumbers,biomassor much of the confusion and controversyin ecology
energycontainedwithinthe constituentpopulations. (Camus and Lima 2002). Some authorsattemptto be
Finally,thereis a certainlogic,parsimonyand unity more specific.For example, Huffakeret al. (1999)
in consideringthepopulationas thefundamental unitof define"place" as "a naturalarea of sufficient size that
ecology,forthe conceptis intimately intertwined
with, reproductionand survivalmaintainsthepopulationfor
and interdependent on, the otherconceptualunit,the many generationsand permitsnormal dispersiveand
species.In orderto definea specieswe needto knowwhat migratorybehaviors". This definitionrecognizesthat
a populationis, and to definea populationrequiresthe "place" has somethingto do withthespecialcharacter-
conceptof species(see below). In addition,theprocesses istics of the species, especially the need to disperse
of microevolutionand population dynamicsare inti- and/ormigrate.One of the most thoroughanalysesof
matelyentwinedin the sense thatmicroevolution gives thisproblemmustbe Andrewarthaand Birch'sexplo-
riseto newtraits,whichsetthedemographicparameters rationof "the meaningof population" in chapter8 of
governingpopulationdynamics,whichdrivethe evolu- theirbook TheEcological Web(Andrewarthaand Birch
tion of new traitsthroughthe Malthusian"strugglefor 1984). These authorsconcludedthatthe "naturalpopu-
existence".Each processdrivestheotherin an unending lation" is made up of manyinterbreeding "local popu-
sequenceof adaptation,speciation,and populationdy- lations" (sensu den Boer 1968, Mayr 1970) and that,
namics. The species and the population seem to be although dispersal between local populations is ex-
natural and necessaryunits on which to build the pected,it is "nonexistentor negligiblebetweennatural
disciplineof ecologyor even biologyin general. populations" (because they are isolated by physical
barriersto dispersal).Thus, the naturalpopulation of
Andrewarthaand Birchis what evolutionarybiologists
call "geographicalisolates" (Mayr 1970). Because such
populationsare reproductively isolatedfromothersim-
How do we define
thepopulation? ilar units,theyare thoughtto be a major factorin the
If "population" is to be thecentralconceptforecology, evolutionof new species.In my opinion,however,this
thenit should be definedin a clear and unambiguous definitionis much too broad, even for evolutionary
way. In the past, the termhas been used in a rather biologists,formanybelievethatevolutioncan progress
casual or arbitrary way,sometimes,it seems,to provea at much smallerspatial scales (e.g., Bush 1969).
pointratherthanforscientific rigor.For instance,using Most ecologistswillprobablyagreewithHuffakeret
one definitioncan lead to a non-equilibrium view of al. (1999) and Andrewarthaand Birch (1984) that the
ecologyand anotherto an equilibriumview (Berryman spatial definitionof "population" should at least con-
1987). The problemhas been exacerbatedby the popu- siderthedispersaland/ormigrationrequirements of the
larityof metapopulationtheory,to the extentthat the organismin question. But how do we translatethis

440 OIKOS 97:3 (2002)

This content downloaded on Tue, 1 Jan 2013 14:45:17 PM

All use subject to JSTOR Terms and Conditions
minimal requirement intoa practical rule?Theproblem tions". The dynamicsof local populationswill be
forthefieldecologistis to definean areaoverwhichthe strongly influenced by dispersaland migration, while
populationneedsto be sampledin orderto representthoseof metapopulations willonlybe affected by rare
the"true"population rather thana "local"population. and/orrandomdiffusion betweenneighboring popula-
The tendency has beento definethisratherarbitrarily,tions.As all members ofa population,so defined, have
or forreasonsof practicality; e.g.,"thepopulationof thepotential to reach,through theircharacteristic mo-
aphidsin a wheatfield".Theyshouldask themselves,bility, all othermembers, thisdefinitionis similarto the
however, whethera wheatfieldis an appropriate spatial systematistsconceptof "deme" - a groupof inter-
scaleforstudyinganimalsthatcan movelongdistances breedingor potentiallyinterbreeding conspecifics
and whosenumbers are greatly affectedby conditions (Simpson1961).
in distanthabitats.Studieslimitedto a "wheatfield" The problemfortheappliedecologist, of course,is
havelittlechanceofuncovering themechanisms under- how to identify the populationarea fora particular
lyingthepopulationdynamics of manyaphids. speciesof interest.Remember that,underour defini-
Camus and Lima (2002) argue that populations
tion,populationchangemustbe broughtabout by
shouldhavea spatialidentity thatreflects
themobility
individualbirthsand deaths,whichmeansthatnet
and habitatrequirements of theorganism - whatthey
migration mustbe zero,or closeto it. In otherwords,
call the"populationarea". Considermonarchbutter-
fliesbreeding
in thenorthern temperate zoneand over- thearea must be largeenoughfortheemigration and
wintering in subtropical foresthabitatsthousandsof immigration rates to be negligibleor, at least, roughly
kilometers to the south.The populationarea of the balanced.If you can imaginesuchan areafora partic-
monarch mustencompass a vastterrainand,forpopu- ularspecies,then youcan graspthespatialidentity of
lationstudiesto be meaningful, mortalityfactors must that species'population. Another way to view it is to
be measured overthetotalarea.Thisis notto saythat imaginethe area over whichan averageindividual
studiesof reproductive successin thenorth,or over- wandersduringitsnormallifetime. Thismaybe fairly
wintering survivalin thesouth,are meaningless, only straightforward for relativelysessile organisms, for
thattheinvestigators needto realizethattheyare not those with well-known migratory patterns (directed
studying thetruepopulation. movement), and forpopulationsisolatedby physical
If we accepttheidea thatthepopulation conceptis barriers(e.g., isolatedislands),but it may be more
onlymeaningful if it has a spatialidentity, thenit difficultfor highlyvagilespecieswithunpredictable
followsthatpopulationchangesshouldnotbe depen- dispersal.This is particularly truefororganisms dis-
denton movement but, rather,on the processesof persedby windand oceancurrents. One wayto solve
reproduction and survival alone.In otherwords,popu- thisproblemis to makethearea so largethatpopula-
lationdynamics is thestudyofforces affectingthebirth tionchangeis notlikelyto be affected verymuchby
and deathprocesses (i.e.,changesin numbers, biomass, movements. This is a consequence of two facts:first,
speciescomposition), notredistribution processes(i.e., thatthe perimeter of a circle,fromwhichdiffusion
changesin location).Local populationdynamics, on occurs,increases as a linearfunction ofitsradiuswhile
the otherhand,needs to considerthe processesof the area, withinwhichbirthsand deathsoccur,in-
immigration and emigration. Thislogicleadsto a rule creasesas a squareof theradius;and second,thatthe
ofthumbfordefining thepopulation area;i.e.,thearea largerthe area, the morelikelyis it to includethe
shouldbe largeenoughso thatimmigration and emi- normalmovement patternof the species.What this
grationis rareor,at least,roughly balanced(Berrymanmeansis thatorganisms withfree-drifting pelagicstages
1999).It also leadsto a formal definitionofpopulation (e.g., aphidsand manymarine
specieswithpelagic
as a group of individualsof the same species that live
larvae)mayhaveto be sampledoververylargeareasor
togetherin an area of sufficientsize to permitnormal
verylargedistancesif we intendto understand the
dispersaland/ormigrationbehaviorand in whichnumeri-
processes
determiningtheirpopulation
dynamics.
cal changesare largelydetermined by birthand death
processes.
Noticethatthisdefinition omitstherequire-
mentthatthearea be largeenoughto satisfy all the
requirementsforreproduction and survival. Thisrecog- Whatare theimplications of thepopulation
nizesthatpopulations can becomeextinct ifthehabitat concept forecology?
withinthepopulationarea deteriorates sufficiently.
It
also recognizes
thatthespatialidentity ofa population I believethattheacceptanceofthepopulation concept
dependson the vagilityof the species,not on its as thebasicbuilding blockforecology,
togetherwithits
propertiesof reproduction and/orsurvival. Giventhis precisedefinition
(butperhapsa betterone thanI have
thengroupsof organisms
definition, livingin smaller given)and methods fordeterminingthepropersizeof
areasbecome"localpopulations" (or sub-populations),thepopulation area,willhavea profoundeffecton our
andthoseoccupying largerareasbecome"metapopula- discipline.

OIKOS 97:3 (2002) 441

This content downloaded on Tue, 1 Jan 2013 14:45:17 PM

All use subject to JSTOR Terms and Conditions
 First, and most important,a clear and widely ac- - I would liketo thankMauricio Lima and
Acknowledgements
cepted definitionof the centralconceptwill providea Bradford
Hawkinsforhelpful comments and suggestions
for
improving
earlierversions
of thisessay.
standardfordiscussionand comparison,so thatwe will
at least know whatwe are all talkingabout (something
I'm not quite sure we are doing now).
Second, otherecologicalconceptscan be placed into References
their correct context. For example, the idea of an
H. G. and Birch,L. C. 1984.The ecological
Andrewartha,
ecologicalcommunity as a groupof interactingpopula- web:moreon thedistribution
and abundanceof animals.
tions (Odum 1971) presupposesa spatial scale deter- - Univ. Chicago Press.
minedby its most vagilemembers.In otherwords,the Berryman,
A. A. 1981.Population
systems:
a general
introduc-
spatial dimensionfor community-level studies should tion. - PlenumPress.
Berryman, A. A. 1987.Equilibrium or nonequilibrium:is that
be set by the most mobile organisms(Korpimdkiand thequestion? - Bull.Ecol. Soc. Am.68: 500-502.
Krebs 1996). For example, the vole populations of Berryman, A. A. 1999.Principles ofpopulation dynamics and
western Finland are apparentlystabilized by avian - StanleyThornes.
theirapplication.
Bush,G. L. 1969.Sympatric hostraceformation and specia-
predatorsthatimmigrate rapidlyintoareas of highvole tionin frugiverous fliesof thegenusRagoletis(Diptera:
abundance and emigratefromlow-densityareas (Kor- Tephrididae).- Evolution 23: 237-251.
pimaki and Norrdahl 1991). In this case, the correct Camus,P. A. and Lima,M. 2002.Populations, metapopula-
spatial scale to studythe interactionbetweenvole and tions,and theopen-closed dilemma:theconflict between
operationaland naturalpopulation concepts.- Oikos97:
raptorpopulationsis determinedby the dispersalchar- 433-438.
acteristicsof the predatorsratherthan those of the DeAngelis,D. L. andGross,L. J.(eds)1992.Individual-based
prey. In the case of lions feedingon rovingherds of modelsand approaches in ecology.- Chapman& Hall.
denBoer,P. J. 1968.Spreading theriskand thestabilization
ungulates,it may be the opposite. oftheanimalnumbers. - ActaBiotheor.18: 165-194.
Third,the populationarea conceptprovidesa basic Hanski,I. and Gilpin,M. 1999. Metapopulation biology:
frameworkfor describingand analyzingthe effectsof ecology,geneticsand evolution. - AcademicPress.
Huffaker,C., Berryman, A. and Turchin, P. 1999.Dynamics
spatial heterogeneity on populationdynamics.For ex- andregulation ofinsectpopulations. - In: Huffaker,C. B.
ample,two importantcharacteristics of the population and Gutierrez, A. P. (eds),Ecologicalentomology. Aca-
area may be (1) the proportionof the total area com- demicPress,pp. 269-312.
posed of favorablehabitat,and (2) the degreeof frag- Korpimaki, E. and Norrdahl, K. 1991.Do breeding nomadic
avianpredators dampenpopulation fluctuationsof small
mentationof the favorablehabitat. Population areas mammals? - Oikos62: 195-208.
can thus be describedby average propertiesjust like Korpimiki, E. and Krebs,C. J. 1996.Predation and popula-
populations,and these average propertiescan thenbe tioncyclesof smallmammals. - BioScience46: 754-764.
Lomnicki,A. 1988. Populationecologyof individuals.-
relatedto per-capitabirthand death processes. PrincetonUniv.Press.
Finally (but not exhaustively),the ultimategoal of Mayr,E. 1970.Populations, species,andevolution:an abridg-
ecologyis to elucidatethe generalprinciplesor laws of mentof animalspeciesand evolution.- HarvardUniv.
Press.
interactionbetween populations, and to classifythe Odum,E. P. 1971.Fundamentals ofecology,3rded. - W.B.
emergentpropertiesof these interactions.To achieve Saunders.
these ends we must have a precise and unambiguous R. E. 1990.Ecology,3rded. - Freeman.
Ricklefs,
conceptual frameworkand terminology.If this essay Royama,T. 1992.Analytical population dynamics. - Chap-
man& Hall.
has helpedus towardsthisend, mytimewill have been Simpson,G. G. 1961. Principles of animaltaxonomy.-
well spent. ColumbiaUniv.Press.

442 OIKOS 97:3 (2002)

This content downloaded on Tue, 1 Jan 2013 14:45:17 PM

All use subject to JSTOR Terms and Conditions