sign in sign up
<<
Home , Abiotic component, Niche construction

The University of Chicago

Niche Construction Author(s): F. John Odling-Smee, Kevin N. Laland and Marcus W. Feldman Source: The American Naturalist, Vol. 147, No. 4 (Apr., 1996), pp. 641-648 Published by: The University of Chicago Press for The American Society of Naturalists Stable URL: http://www.jstor.org/stable/2463239 . Accessed: 23/02/2014 15:09

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 and tools to increase and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

.

The University of Chicago Press, The American Society of Naturalists, The University of Chicago are collaborating with JSTOR to digitize, preserve and extend access to The American Naturalist.

http://www.jstor.org

This content downloaded from 129.2.19.102 on Sun, 23 Feb 2014 15:09:33 PM All use subject to JSTOR Terms and Conditions Vol. 147, No. 4 The AmericanNaturalist April 1996

NOTES AND COMMENTS

NICHE CONSTRUCTION

Organisms,through their metabolism, their activities, and their choices, define, partlycreate, and partlydestroy their own niches. We referto thesephenomena as "nicheconstruction." Here we arguethat niche construction regularly mod- ifiesboth biotic and abioticsources of naturalselection in environmentsand, in so doing,generates a formof feedback in evolutionthat is notyet fully appreci- ated by contemporaryevolutionary theory (Lewontin 1978, 1983; Odling-Smee 1988,in press). Adaptationis generallythought of as a process by whichnatural selection shapesorganisms to fitpreestablished environmental "templates." Environments pose "problems,"and thoseorganisms best equipped to deal withthe problems leave themost offspring. Despite the recognition that forces independent of or- ganismsoften change the worlds to whichpopulations adapt (Van Valen 1973), thechanges that bring about in theirown worlds are rarelyconsidered in evolutionaryanalyses. Yet, to varyingdegrees, organisms choose theirown ,mates, and resourcesand constructimportant components of theirlo- cal environmentssuch as nests,holes, burrows, paths, webs, dams, and chemical environments.Many organisms also choose, protect,and provision"nursery" environmentsfor their offspring. Organisms not only adapt to environments,but in partalso constructthem (Lewontin 1983). Hence, manyof the sources of naturalselection to whichorganisms are exposedexist partly as a consequence ofthe niche-constructing activities of past and presentgenerations of organisms. Thereare numerouscases oforganisms modifying their own selective environ- mentsin nontrivialways, by changingtheir surroundings or by constructingarti- facts(Von Frisch1975; Hansell 1984). One earlyexample was describedby Dar- win(1881). , through their burrowing activities, their dragging organic materialinto the , theirmixing it withinorganic material, and theircasting, whichserves as a basis formicrobial activity, change both the structureand chemistryof (Lee 1985). As a resultof the accumulatedeffects of past generationsof earthwormniche construction, present generations inhabit radi- cally alteredenvironments and are exposedto changingsets of selectionpres- sures. Thereis also considerableevidence of evolutionaryresponses to self-induced selectionpressures. For instance,social bees, wasps, ants, and termites construct

Am. Nat. 1996. Vol. 147, pp. 641-648. ? 1996 by The Universityof Chicago. 0003-0147/96/4704-0009$02.00.All rightsreserved.

This content downloaded from 129.2.19.102 on Sun, 23 Feb 2014 15:09:33 PM All use subject to JSTOR Terms and Conditions 642 THE AMERICANNATURALIST elaboratenests that then mediate selection for many nest regulatory,mainte- nance,and defensebehaviors (Rothenbuhler 1964; Spradbery 1973; Von Frisch 1975;Mathews and Mathews 1978; Hansell 1984). Many species of fish, amphibi- ans, reptiles,birds, and mammalsconstruct nests and burrows,which then influ- encefurther selection. For example,comparative evidence suggests that the com- plexburrow systems excavated by moles, rats, badgers, and marmots,with their undergroundpassages, interconnected chambers, and multipleentrances, have servedas thesource of selection for defense, maintenance, and regulationbehav- iorsand componentsof matingrituals (Von Frisch1975; Hansell 1984). Plants,too, changethe chemical , pattern of nutrientcycling, tempera- ture,humidity, and fertilityof thesoils in whichthey live (Ricklefs1990). They mayeven affectlocal climates,the amount of precipitation,and thewater cycle (Shuklaet al. 1990).Many plants also changeboth their own and otherspecies' local environmentsvia allelopathy(Rice 1984),while and chaparraltree speciesfacilitate forest fires by accumulatingoils or litter(Mount 1964; Allen and Starr1982). These specieshave evolveda resistanceto fireand, in some pine speciesthat require a firebefore their seeds willgerminate, a dependencyon it (Allenand Starr1982).

THE EVOLUTIONARY CONSEQUENCES OF NICHE CONSTRUCTION Theseexamples, and others (Lewontin 1982, 1983), suggest that niche construc- tionmay be a generalphenomenon, that it is not restrictedto a few isolated speciesor taxa,and that feedback from phenotypically modified sources of selec- tionin environmentshas evolutionaryas well as ecologicalconsequences. Al- thoughseveral topics in contemporarypopulation are alreadyconcerned withthe evolutionary consequences of the changes that organisms bring about in theirown environments(e.g., ,frequency- and density-dependentselec- tion),so farthese analyses have onlyfocused on geneticloci thatinfluence the productionof the niche-constructingphenotype itself. What is missingis any explorationof thefeedback effects on othergenetic loci. A moregeneral body oftheory is required. In orderto encouragethe developmentof this theory,in whatfollows we discusssome evolutionary consequences of nicheconstruction and detailwhy it is likelyto be of significanceto thebiological sciences.

EXTRAGENETIC INHERITANCE Currently,evolutionary theory rests heavily on theassumption that only genes are transmittedfrom generation to generation,the principal exception being cul- turalinheritance (Feldman and Cavalli-Sforza1976; Boyd and Richerson1985). However,ancestral niche-constructing organisms effectively transmit legacies of modifiednatural selection pressures in theirenvironments to theirdescendants. Thisextragenetic inheritance has previouslybeen calledan "exploitivesystem" (Waddington1959), an "ontogeneticinheritance" (West et al. 1988),and an "eco- logicalinheritance" (Odling-Smee 1988). We willintroduce it in stages.

This content downloaded from 129.2.19.102 on Sun, 23 Feb 2014 15:09:33 PM All use subject to JSTOR Terms and Conditions NOTES AND COMMENTS 643

If,in each generation, each organismonly modifies its environment temporarily or inconsistently,then there will be no cumulativeor consistentmodification of any sourceof naturalselection in its population'senvironment. If, however,in each generation,each organismrepeatedly changes its own ontogenetic environ- mentin thesame way,because each individualinherits the same genescausing it to do so, thenancestral organisms can modifya sourceof for theirdescendants by repetitiveniche construction. The environmentalconse- quencesof such niche construction may be transitoryand may still be restrictedto singlegenerations only, but the same induced environmental change is reimposed sufficientlyoften, for sufficient generations, to serveas a significantsource of selection. For example,individual web spidersrepeatedly make webs in theirenviron- ments,generation after generation, because theyrepeatedly inherit genes in- structingthem to do so. Subsequently,the consistent presence of a web in each spider'senvironment may, over manygenerations, feed back to become the sourceof a newselection pressure for a furtherphenotypic change in thespiders, such as the buildingby Cyclosa of dummyspiders in theirwebs to divertthe attentionof avian predators (Edmunds 1974). Yet, thiskind of feedback does not introducean extrageneticinheritance in ,because no consequenceof nicheconstruction is transmittedthrough an externalenvironment from one gen- erationto thenext. In morecomplex cases, inheritedgenes instruct organisms to modifyrepeatedly the ontogeneticenvironments of theiroffspring as well as, or insteadof, their own. Here theconsequences of niche construction are effectively"transmitted" fromone generationto the nextvia an externalenvironment, in the formof a parentallymodified source of natural selection for their offspring. This transmittal is sufficientto establishan extrageneticinheritance system in evolution. Offspring nowreceive a dualinheritance from their parents, genes relative to theirselective environments,and at leastsome parentally modified sources of selectionin their environmentsrelative to theirgenes. The cuckoois an example.Cuckoo parents repeatedly select nests for their offspring,generation after generation, thereby bequeathing modified selection pressuresas wellas genesto theiroffspring. These modifiedselection pressures have thenapparently selected for changed adaptations in theoffspring, such as a shortincubation period or the behavioralejection by newlyhatched cuckoo chicksof host eggs from the parasitized nests. Also, cuckoos that are raisedin the nestsof a particularhost species may preferentially parasitize that host species in whosenests they were originally raised themselves when they mature, possibly learningthe host characteristics through imprinting (Krebs and Davies 1993).This kindof extrageneticinheritance is currentlymodeled as a non-Mendelianmater- nal inheritance(Cowly and Atchley1992; Schluter and Gustafsson1993). Mater- nal inheritancecan generatesome counterintuitive results, including temporarily reversedevolutionary responses to selection,and timelags thatmay resultin populationscontinuing to evolveafter selection has ceased by an "evolutionary momentum"(Feldman and Cavalli-Sforza1976; Kirkpatrick and Lande 1989). Maternalinheritance is, however,a specialcase, and theeffects of niche con-

This content downloaded from 129.2.19.102 on Sun, 23 Feb 2014 15:09:33 PM All use subject to JSTOR Terms and Conditions 644 THE AMERICAN NATURALIST structiongeneralize to multiplegenerations and to multipleancestors, not just mothers.For example,suppose a niche-constructingbehavior is influencedby geneticvariation at one setof loci, that it spreads through a population over many generations,and thatit progressivelychanges the frequency of someresource in theenvironment as itdoes so. Supposefurther that the frequency of the , now a partof an extrageneticinheritance, feeds back to thepopulation to influ- ence selectionat a secondset of loci. In thesecircumstances, a time lag should developbetween the change in frequencyof alleles at theniche construction loci and the responseto a frequency-dependentmodified selection pressure at the recipientloci, the onlydifference being that here the timelag is likelyto take manymore generations to buildup. Darwin'searthworms are an example.Suppose a firstgenetic locus influences someniche-constructing behavior, such as soilprocessing or burrow lining, which subsequentlyaffects the amount of topsoil or thesoil nutrients in the earthworms' environment.Another locus expresses a phenotypethat is affectedby soil condi- tions,such as the structureof theepidermis or the amountof mucussecreted. In thiscase, ancestralniche construction by manygenerations of earthworms, due to thefirst locus, willeventually feed back to the secondlocus and change its selection,but only after many generations of nicheconstruction. Here again, theeffect of the time lag shouldbe to createan evolutionary"momentum," such thatif the selectionpressures at the firstlocus are relaxedor reversed,the re- sponseat thesecond locus shouldcontinue in theoriginal direction for a number of additionalgenerations. Moreover, assuming many generations are requiredto modifynatural selection on a population,it mightnot be able to evolve fast enoughto preventthe geneticvariation on whichits eventualresponse relies frombeing prematurely lost. This possibilityalso meansthat once a population reachesa stableequilibrium, it maytake a greaterperiod of time,or stronger selection,for the population to moveaway from it.

INDIRECT GENE INTERACTIONS Nicheconstruction provides a wayin whichthe differential phenotypic expres- sion of genotypesat one locus can be influencedby the genotypeat another locus, indirectlyvia theexternal environment. For instance,the pink coloration characteristicof flamingospecies is extractedfrom the carotenoid pigmentation ofthe crustacea they digest (Fox 1979).Here thegenes influencing flamingo prey choiceinteract with those underlying pigment extraction and utilization,via the foodresources in theirenvironment. In severalrespects, the genetic basis ofthis feedbackis reminiscentof epistasis.In contrastto conventionalepistasis, how- ever, niche constructioncan generateinteractions between genes in different populations,even differentspecies. For example,the genes thatunderlie those activitiesof earthworms that modify soil structure, thereby enhancing plant yields (Lee 1985),have influenced the expression of genes in theplant population affect- inggrowth. Thus, the niche-constructing outputs of individualsnot only change selectiveenvironments, which feed back to alterthe fitnesses of alleles at other loci,but may also influencethe phenotypic expression of thosealleles in ontoge-

This content downloaded from 129.2.19.102 on Sun, 23 Feb 2014 15:09:33 PM All use subject to JSTOR Terms and Conditions NOTES AND COMMENTS 645 neticenvironments (West et al. 1988).The effectof these interactions,which influenceboth the nature of the variants subjected to selectionand thepattern of selectionacting on those variants,is to introducethe kindof feedbackloops betweenpopulations and theirenvironments that Robertson (1991) suggests may makea considerabledifference in evolution.

SYNTHESIZING EVOLUTIONARY BIOLOGY AND Because nicheconstruction may applyto interactionsbetween genes in the samepopulation, or in differentpopulations, it providesa mechanismfor driving populationsto coevolvein .Populations may affecteach othernot onlydirectly in theways that are alreadymodeled by coevolutionarytheory, as, forinstance, in host-parasitecoevolution (Futuyma and Slatkin1983), but also indirectlyvia theirimpact on some intermediateabiotic component in a shared ,as in competitionfor a chemicalor waterresource. For example,if nicheconstruction resulting from a gene in a plantpopulation causes the soil chemistryto changein such a way thatthe selectionof a gene in a second population,of plantsor ,is changed, then the first population's nicheconstruction will drive the evolutionof the secondpopulation simply by changingthe physical state of thisabiotic ecosystem component. Since the dy- namicsof the intermediate may be qualitativelyquite different fromeither the frequency changes in thegenes that underlie the niche construc- tionor the numberof niche-constructingorganisms in the firstpopulation, this indirectfeedback between species may generatesome interesting-andas yet unexplored-behaviorin coevolutionarysystems. Accountingfor the evolutionand prevalenceof mutualisticinteractions is a stubbornproblem for theoretical population biologists, who usuallyassume that thereis somecost to thedonor (Roughgarden 1975; Mesterton-Gibbons and Du- gatkin1992). Recent years, however, have seena changein thinkingabout mutu- alism,with increasing emphasis placed on thefact that many mutualisms involve the transferof incidentalby-products, at no cost to the donor(West-Eberhard 1975;Brown 1983; Janzen 1985; Connor 1986; Mesterton-Gibbons and Dugatkin 1992).For example,seed predatorsoften benefit the host plant through the dis- persalof its seeds at no cost to themselves(Janzen 1985). These by-products, whichare clearlya componentof a population'sniche construction, often serve as thesource of selection for interspecific investment in theirproduction (Connor 1986). For instance,fruit represents investment in seed dispersers(Thompson 1982).A thoroughunderstanding of thesecoevolutionary dynamics may involve formalrecognition that the niche-constructing activities of organismscan change selectionpressures and therebyinitiate mutualistic interactions.

A SECOND ROLE FOR PHENOTYPES IN EVOLUTION Whenphenotypes niche construct, they can no longerbe thoughtof as simply "vehiclesfor genes," sincethey are nowalso responsiblefor modifying some of thesources of selectionin theirenvironments that may subsequently feed back

This content downloaded from 129.2.19.102 on Sun, 23 Feb 2014 15:09:33 PM All use subject to JSTOR Terms and Conditions 646 THE AMERICAN NATURALIST to selecttheir genes. Moreover, there is no requirementfor niche construction to resultdirectly from genetic variation before it can influencethe selectionof geneticvariation. For instance,niche construction can dependon learning,as is thecase forthe British tits that may have changedtheir own selectionpressures by openingfoil milk-bottle tops, thereby gaining access to a new resource.The evolutionaryconsequences of this learned innovation are unknown,but it is possi- ble thatthis new resource may now be selectingfor some further change in these tits-forinstance, for different digestive enzymes-or for improved learning abil- ity(Fisher and Hinde1949; Sherry and Galef1984). Similarly, niche construction can also dependon culture,as happenswhen humans increase the frequency of thesickle-cell allele in theirown populations by unwittinglyincreasing the preva- lence of malariain theirenvironments through their agricultural practices (Dur- ham 1991).The consequencesfor a geneat thesecond locus are thesame, pro- videdthe effectof the nicheconstruction on the environmentalresources that constitutethe sourceof selectionis unchanged.The netresult is an additional role forphenotypes in evolution.Phenotypes not only surviveand reproduce differentiallyin the face of naturalselection and chancebut also modifysome sourcesof selectionin theirenvironments by nicheconstruction.

RELATIVIZATION OF EVOLUTIONARY BIOLOGY On the basis of empiricalevidence, Lewontin (1978, 1982, 1983)has argued thatthe "metaphorof adaptation"should be replacedby a "metaphorof con- struction."However, the acceptance of Lewontin's position demands more than just semanticadjustments to evolutionarytheory. Niche constructionchanges thedynamic of the evolutionary process in fundamental ways because it precludes a descriptionof evolutionarychange relative only to autonomousenvironments. Instead,evolution now consistsof endlesscycles of naturalselection and niche construction.Equally, it is no longertenable to assumethat the only way organ- ismscan contributeto evolutionarydescent is by passingon fitor unfitgenes to theirdescendants relative to theirenvironments, because theycan also pass on modificationsinthose environments that are better or worsesuited to theirgenes. Adaptationbecomes a two-waystreet.

ACKNOWLEDGMENTS We are gratefulto R. C. Lewontinfor his inspiration,support, and advice.

LITERATURE CITED

Allen, T. F. H., and T. B. Starr. 1982. Hierarchy:perspectives for ecological complexity.University of Chicago Press, Chicago. Boyd, R., and P. J. Richerson. 1985. Culture and the evolutionaryprocess. Universityof Chicago Press, Chicago. Brown, J. L. 1983. Cooperation: a biologist's dilemma. Advances in the Study of Behavior 13:1-37. Connor, R. C. 1986. Pseudo-reciprocity:investing in .Animal Behaviour 34:1562-1584.

This content downloaded from 129.2.19.102 on Sun, 23 Feb 2014 15:09:33 PM All use subject to JSTOR Terms and Conditions NOTES AND COMMENTS 647

Cowley, D. E., and W. R. Atchley. 1992. Quantitativegenetic models for development,epigenetic selection,and phenotypicevolution. Evolution 46:495-518. Darwin, C. 1881. The formationof vegetable mold throughthe action of works withobservations on theirhabits. J. Murray,London. Durham, W. H. 1991. Coevolution: genes, culture,and humandiversity. Stanford University Press, Stanford,Calif. Edmunds, M. 1974. Defense in animals. Longman, New York. Feldman, M. W., and L. L. Cavalli-Sforza. 1976. Cultural and biological evolutionaryprocesses: selectionfor a traitunder complex transmission.Theoretical Population Biology 9:238-259. Fisher, J., and R. A. Hinde. 1949. The openingof milkbottles by birds. BritishBirds 42:347-357. Fox, D. L. 1979. Biochromy: natural coloration of living things. Universityof California Press, Berkeley. Frisch, K. von. 1975. Animal architecture.Hutchinson, London. Futuyma,D. J., and M. Slatkin. 1983. Coevolution. Sinauer, Sunderland,Mass. Hansell, M. H. 1984. Animal architectureand buildingbehavior. Longman, New York. Janzen,D. H. 1985. The naturalhistory of mutualisms.Pages 40-99 in D. Boucher, ed. The biology of mutualisms.Croom Helm, London. Kirkpatrick,M., and R. Lande. 1989. The evolutionof maternalcharacters. Evolution 43:485-503. Krebs, J. R., and N. B. Davies. 1993. An introductionto behavioural ecology. 3d ed. Blackwell, Oxford. Lee, K. E. 1985. Earthworms:their ecology and relationwith soil and land use. Academic Press, London. Lewontin,R. C. 1978. Adaptation.Scientific American 239:156-169. 1982. Organismand environment.Pages 151-172 in H. C. Plotkin,ed. Learning,development and culture.Wiley, New York. 1983. Gene, ,and environment.Pages 273-285 in D. S. Bendall, ed. Evolution from molecules to men. CambridgeUniversity Press, Cambridge. Mathews, R. W., and J. R. Mathews. 1978. Insect behavior. Wiley, New York. Mesterton-Gibbons,M., and L. A. Dugatkin. 1992. Cooperation among unrelatedindividuals: evolu- tionaryfactors. Quarterly Review of Biology 67:267-280. Mount, A. B. 1964. The interdependenceof the eucalypts and forest firesin southernAustralia. AustralianForestry 28:166-172. Odling-Smee,F. J. 1988. Niche constructingphenotypes. Pages 73-132 in H. C. Plotkin,ed. The role of behavior in evolution. MIT Press, Cambridge,Mass. In press. Niche construction,genetic evolution and culturalchange. Behavioural Processes. Rice, E. L. 1984. Allelopathy.2d ed. Academic Press, Orlando, Fla. Ricklefs,R. E. 1990. Ecology. 3d ed. Freeman, New York. Robertson, D. S. 1991. Feedback theoryand Darwinian evolution. Journalof Theoretical Biology 152:469-484. Rothenbuhler,W. C. 1964. Behavior genetics of nest cleaning in honey bees. IV. Responses of F, and backcross generationsto disease-killedbrood. American Zoologist 4:111-123. Roughgarden,J. 1975. Evolution of marine : a simple cost-benefitmodel. Ecology 56: 1201-1208. Schluter,D., and L. Gustafsson.1993. Maternalinheritance of conditionand clutchsize in the collared flycatcher.Evolution 47:658-667. Sherry,D. F., and B. G. Galef,Jr. 1984. Culturaltransmission without imitation-milk bottle opening by birds. Animal Behaviour 32:937-938. Shukla, J., C. Nobra, and P. Sellars. 1990. Amazon deforestationand climatechange. Science (Wash- ington,D.C.) 247:1322-1325. Spradbery,J. P. 1973. Wasps. Sidgwick & Jackson,London. Thompson, J. N. 1982. Interactionand coevolution. Wiley, New York. Van Valen, L. 1973. A new evolutionarylaw. EvolutionaryTheory 1:1-30. Waddington,C. H. 1959. Evolutionarysystems; animal and human. Nature (London) 183:1634- 1638. West, M. J., A. P. King, and A. A. Arberg. 1988. The inheritanceof niches: the role of ecological

This content downloaded from 129.2.19.102 on Sun, 23 Feb 2014 15:09:33 PM All use subject to JSTOR Terms and Conditions 648 THE AMERICAN NATURALIST

legacies in ontogeny.Pages 41-62 in E. Blass, ed. Handbook of behavioral neurobiology. Plenum, New York. West-Eberhard,M. J. 1975. The evolution of social behavior by kin selection. QuarterlyReview of Biology 50:1-33.

F. JOHN ODLING-SMEE INSTITUTE OF UNIVERSITY OF OXFORD 58 BANBURY ROAD OXFORD OX2 6QS UNITED KINGDOM KEVIN N. LALAND SUB-DEPARTMENT OF ANIMAL BEHAVIOUR UNIVERSITY OF CAMBRIDGE MADINGLEY CAMBRIDGE CB3 8AA UNITED KINGDOM MARCUS W. FELDMAN DEPARTMENT OF BIOLOGICAL SCIENCES STANFORD UNIVERSITY STANFORD, CALIFORNIA 94305 SubmittedFebruary 21, 1995; Revised August 7, 1995; Accepted August 16, 1995

Associate Editor: J. Bruce Walsh

This content downloaded from 129.2.19.102 on Sun, 23 Feb 2014 15:09:33 PM All use subject to JSTOR Terms and Conditions