Addicott, Aho Et Al (OIKOS) 1987

7/30/2019 Addicott, Aho Et Al (OIKOS) 1987

http://slidepdf.com/reader/full/addicott-aho-et-al-oikos-1987 1/7

OTKOS 49: 3;l_1346. Copenhagen 1987

Ecological neighborhoods: scaling environmental patterns

John F. Addicott, John M. Aho, Michael F. Antolin, Dianna K. Padilla, John S. Richardson and Daniel A. Soluk

Addiott, J. F., Aho, J. M., Antolin, M. F., Padilla, D. K., Richardson, J. S. andSoluk, D. A. f987. Ecological neighborhoods: scaling environmental patterns. - Oi-kos 49: 3&-%.

In this paper we review, develop, and differentiate among concepts associated withenvironmental patterning (patch,-division, and heterogeneity), spatial and temporalscales of ecological proaesses ecological neighborhoods), and responses of organismsto environmental patterning (relative patch size, relative patch duration, ielativepatgh isolation" and grain lespoqse), We gencralizc the concept of ecological neigh-borhoods to represent regions-of activity or influence during periods of t'ime appro-pri.ale- o- particular ecological processes. Therefore, there is no single ecol6gicalneighborhood for any given organism, but rather a number of neighborhoods, eachappropriate to different processes. Neighborhood sizes can be estimated by examin-ing the cumulative distribution of activity or influence of an organism as a function ofincreasingly. large spatial units. The spatial and temporal dimensions of neighbor-hoods provide the scales necessary or assessing nvironmental patterning relaiive toparticular ecological processes for a given species. Consistent application of theneighborhood-concept will assist n the choice of appropriate study-units, compari-sons among different studies, and comparisons between empirical studies and-the-oretical postulates.

J. E Addicon and D- K. Padilla, Dept of zoology, univ. of Albena, Edmonton, AI-berta, Canada T6G 289. t. M. Aho, Savannah River fuology lrb., Aiken, SC 29801,USA. M. F. Antolin, Dept of Biol- Sci., Florida State Univ., Tallahassee, FL 32306,USA. I. S. Richardson, Inst. of Animal Resource Ecology, Univ. of British Columbia,Vancouver, B. C., Canada V6T IWs. D. A. Soluk, Dept of Zoology, Erindale Col-lege, Mbsbsauga, Ontario, Canada L5L lC6.

Introduction

Environmental patterning refers to the non-uniform,spatial and temporal distribution of resources andabiotic conditions that influence species or species nter-actions. Such patterning is pervasive in nature and isknown or hypothesized to affect many ecological,pro-cesses and phenomena, including population dynamics,life histories, dispersal, foraging behavior, patterns ofnatural selection, coexistence of speqies, predation, and

species diversity (e.g" Huffaker 1958, Southwood L962,lrvin and Paine 1974, May 1974, Roff 1974, Wilbur etal. 1974, rvin 1976, Chesson 1978, Pleasants andZim-merman 1979, Shorrocks et al. 1979, Denno et al. 1980,Hassell 1980, Wilson 1980, den Boer 1981, Paine andlrvin 1981, Kareiva 1982, Spence 1983). Therefore, the

study of environmental patterning and how organismsrespond to it has beome a central focus of current eco-logical research.

However, there are no general procedures and crite-ria for determining how organisms respond to or are af-fected by environmental patterning. C.onvenient but ar-bitrary spatial and temporal study units may be inappro-priate for the processes being studied (see discussion by:Brown and Kodric-Brown tW, Connell and Sousa1983, C-onnor et al. 1983, Cunningham 1986), and con-

clusions appropriate to one scale of environmental orpopulation patterning may be inappropriately trans-ferred to another scale e.g. Price 1980). Without a rea-sonable means of scaling, it is difficult to compare re-sults from the same species n different environments,from different species n the same environment, or be-

Accepted 12 February 987o otKos

340 OIKOS {9: -1(1987)



a

&raFr{ds}

Fig. l. Examples fcombinations of divided andheterogeneousenvironments. Differenrpalch types are denored bydifferent patterns ofshading. Notice the variationin patch size, an d th eexistence f patches fpatches.

D IV ID E DHOMOG NEOUS

tween theoretical postulates and empirical results.Therefore, a precise and unified approach to the studyof patterning and responses o patterning is needed.

In this paper we first discuss oncepts associated withenvironmental patterning: patch and environmental di-vision/heterogeneity (Wiens 1976, Shorrocks et al.1979, Gould and Stinner 19&4).Seoond, we develop ageneralized concept of ecological neighborhoods(Wright 1943, 1946, Southwood

ln7, Antonovics andkvin 1980). Ecological neighborhoods hen provide amechanism for scaling both spatial and temporal pat-terns of the environment relative to organisms. Next,we develop explicit procedures and criteria for imple-menting the concept of ecological neighborhoods. Fi-nally, we discuss why scaling s necessary n ecologicalstudies. Our intent is to encourage he use of more ex-plicit criteria for defining and rgporting on environ-mental patterning and the responsis of qrganisms o it .

Environrirental patterning

The study of environmental patterning involves descrip-tions of the spatial distributions of both resources andabiotic conditions. when environmental variabres arediscontinuous among arbitrarily chosen sampling units,environments are said to consist of patches. A patch isdefined as a discontinuity in environmentar characterstates where the discontinuity matters to the organism(Wiens 1976), as a *bounded, connected disonlinuityin a homogeneous reference background. (Levin andPaine 1974), or as any place in the environment wherethe abundance of either resour@s or organisms s highor low relative to its surroundings (Roughgarden lg7i).These

definitions can be implemenged by using any of avariety of statistical procedures see pielou 1969, nipteyr98l).

We distinguish tw o qualitatively different kinds ofpatterning: division and heterogeneity shorrocks et al .1979, Gould and Stinner 1984). Division involves theseparation f patches by regions of relatively unsuitableenviionmental onditions Fig. la,c), where suitabilityis defined n terms of the fitness or components of fit-

OIKOS {9:-1 tgtt?)

r1^ess)hat an organism would experience within a patch(southwood rg77). Heterogeneity involves the Lxist-ence of two or more qualitatively different patch types(Fig- lb,c) that may or may no t differ in suitability. En-vironments could be divided and homogenrou, 1rig.1a), heterogeneous ur no t divided (Fig. tb), or both di -vided and heterogeneous Fig. lc). tn a diviaed envi-ronment, the interpatch region could represent environ-

ments where the fitness of an organism would be zero(southwood L977), such as .".ine environments forterrestrial insects or all plants except flowering yuccasfor yucca moths. Alternatively, thi interpatcf, regioncould represent environments n which fiiness s rela-tively low, but non-zero. Each type of patterning existsas a continuum, with variation from continuouJ to oi-vided environments, and from homogeneous o hetero-geneous environments.

Patchy environments ncrease he difficulty of makingappropriate choices of temporal and spatial scale foistudy units. The distribution of organisms and patternsof resolrrces may not reflect an organism's ur. oi the en-vironment. Simply identifying the existence of patchi-ness does not mean that the patchiness s importint fora particular process. Next, environments can simulta-neously exhibit patchiness at a number of different spa-tial scales, rom millimeters to kilometers (Krebs l9z-g).However, a simple description of patch structure ofeither an organism or its environment would not indi-cate which spatial scale s appropriate for the study of agiven prooess or a given organism (Connor et al. i9gl,Heads and l-awto-n 9g3). Firially, itifferent species maytreat patch sizes differently. A simple description of en-yirgnnleltal or population patch structure provides nobiologically meaningful way of comparing patchesamong different organisms (Southwood tgil)'. Com-parisons must be based on the responses f organisms.

A general concept of ecological neighborhoodsTherefore, we need a mechanism or determining th escale at which experiments or observations should be

U N D I V I E DH E T E R O G E N E O U S

D V I D E DH E T E R O G E N E U S

341

c

@k

7&

ts=-_=\ary/At A

&P

@



made, and this scale will in turn measure Patch size'

patch duration, patch isolation and use of heterogen-Lou, put"hes relative to organisms. The concept of eco-

togical neighborhoods s appropriate for this task' Here

*J"*p"nd

and generalize neighborhood concepts de-

veloped by Wright (1943, 1946), Southwood (1977) and

Anronovics and lrvin (1980) or scaling genetic or eco-

logical processes. our generalized version of the con-

cept of ecological neighborhoods refines earlier uses o

rnut" it appliiable to any ecological process and to bothmobile and sessile rganisms.

We define ecological neighborhoods by three prop-

erties: an ecological process, a time scale appropriate to

that process, and an organism's activity or influenceduring that time period. The ecological neighborhoodof an organism for a given ecotogical process is the re-

gion wit[in which that organism s active or has some n-

flu"n* during the appropriate Period of time' This con-

cept of neighborhood could aPPty o any ecological pro-

.eis, such as population growth, predator-preyinteractions, @mpetition, foraging behavior, or territo-rial defense. The choice of an ecological proccss will de-

fine an appropriate time scale over which to measureneighborhood-size. As with genetic neighborhoods, it

*"! b" convenient to refer to either the physical size of

theneighborhood or the number of individuals within

the neighborhood.For relatively mobile organisms, he movement of in-

dividuals will usually define the neighborhood (Dobz-

hansky and Wright 1943, Lloyd t967, Koeppl et al'

Lg7s,Ju*".s tg76,Ford and Krurpme 1979)' For sessileorganisms, neighborhoods will freqqently be restrictedtoih" immediate vicinity of a single ndividual, depend-ing upon the precise mechanism of resource utilizationoiini".f"rence with other individuals (Connell 1961'

Mack and Harpet 1977, Antonovics and l-evin 1980)'

However, in some cases egions of influence can extendwell beyond individuals whether they are mobile or ses-sile. This ivill occur when 1) consumable resources aremobile with respect to an organism such as for sit-and-wait predators or filter feeding invertebrates (e'g' Buss

and Jackson 1981), 2) an organism produces some Po-tentially deleterious by-product that is carried in the air

or watir (e.g. allelopathy) (Pratt 1966, -ewis 1985), and3) an organit* attracts its consumers or its mutualisticforagers-from a wide region (e.g. Manasse and Howele83).

Responses o environmental patterning

Relative atch size, solation, nd duretion

Ecological neighborhoods provide appropriate scalesfor measuring he relative size. solation' an d duration

342

of patches n a patterned environment- Relative patchsize is simply the ratio of patch size to neighborhoodsize. Patches of a given size are relatively small if aneighborhood encomPasses number of them, and rela-tively large f a neighborhood encompasses nly part ofone patch (see Southwood 1977). n a divided environ-ment, patches are separated rom each other. The sizeof the region separating atches may be just as mport-ant as he size of patches hemselves Southwood 1977).

Therefore, it is also appropriate o scale nterpatch dis-tances by neighborhood izes, eading o th e concept ofrelative patch isolation. For example, Southwood's(1977) definition of an isolated habitat ncludes he ideathat interpatch distance must be large relative to an or-ganism's migratory range, "the area over which it canmove when it is not reProducing".

Temporal patterning of environments is also import-ant, because patch duration (: temPoral persistence)may range from seconds o millennia (Lrvin and Paine1974, Paine and l-evin 1981). The basis for comparingpatch duration is the time period appropriat€ to a given

neighborhood (see C.onnell and Sousa 1983)- Relative

patchduration provides an assessment of whether

patches are Permanent or ephemeral with respect to aginen process and organism (Southwood t977, Connellind Sousa 1983). As with patch size, any given patchcould be relatively permanent or ephemeral simulta-neously, depending uPon the neighborhoods and organ-isms being considered.

The interaction between relative patch size and rela-tive patch duration may be complex- Temporal environ-mental patterning may affect the mobility or life historyof organisms, and these will in turn affect neighborhoodsize. If resources n a given patch are insufficient for thecompletion of an entire life history stage, then individ-ual organisms must disperse rom one patch to another

as patches are depleted (see Chew 1977). An alternativeresponse to ephemeral resources is the evolutionarymodification of either dispersal behavior or duration oflife history stages o match the temporal availability ofresources e.g. ThYlor l980a,b).

Grein: respotulcs o cnvironmentel heterogeneity

For heterogeneous patches, utilization of the differentpatch types can be random, in proportion to their avail-ibitity, or non-random, with some patch types usedmore frequently than their relative availability and oth-

ers less frequently. These represent fine andcoarse

grain resPonses, esPectively. Although the term grainhas been used n other contexts (MacArthur and Levins1964, -evins 1968), t is most useful as a concept for theresponse of an organism o environmental heterogen-eity. As with relative patch size and duration, grain isrelative o a given process nd period of time- With thisdefinition, grain is not a property of an environment:there are not fine and coarse grained environments'

OIKOS J9: -1 e$7)



N 61

o\

z9Ff

IG,Fo

oUI

=

fI5c)

N t p

S PAT I A L U N I T(o t suxcE,AR E A,voLu t{ E)

Fig. 2. Hlpothetical examples of the rclationship btweer\cumulative distribution of movement or influence against spa-tial unit. Horizontal dotted lines at arbitrary points on the cum-ulative distribution indicate different neighborhood sizes fordifferent decision criteria (Fig. 2a). Each curve in Fig. 2b rep-resents a cumulative distribution with an associated neigh-borhood size for a different ecological process, using a decisioncriterion of.95%.

only fine and coarse grained responses by organisms toenvironmental heterogeneity.

Neighborhoods, relative patch sizes, and grain arealso potentially interrelated. If patches are large relativeto a given neighborhood, grain response to those

patches must be coarse. If patches are relatively small,however, grain response could be either fine or coarse.More complex situations will arise when patches of dif-ferent quality also differ in size or durat ion.

Estimating neigbborhood size

There are two problems involved in measuring neigh-borhoods. The first is to develop procedures for meas-uring the distribution of activity or influence amongsampling units. The second s to develop criteria for de-termining the boundaries of a neighborhood. These cri-teria will be arbitrary, but they should be explicit. Theprocedures needed o measure he distribution of activ-it y or influence will depend on the kinds of neigh-borhoods and organisms being studied.

For neighborhoods uch as ocal reproductive popula-tion neighborhoods ha t involve relatively ong time pe -

OIKOS .t9:3 1987)

riods, the most appropriate ndicator of activity may bea measure of net movement of individuals (o r their ga-metes or progeny) among sampling units during theperiod of time appropriate o the neighborhood. Thereare two general sets of procedures or measuring ne tmovements. One is the direct measurement f dispersaldistances, sing a mark-release-recapture echnique at anumber of sampling sites o determine he cumulativefrequency distribution of dispersal distances Dobzhan-sky and Wright 1943). The other is to measure he dis-tribution of locations where an organism occursthroughout an appropriate period of time (Koeppl et al.1975, Ford and Krumme 1979). For some kinds of or-ganisms and neighborhoods, assessment of influencewill not be a function of the movement of that organism.For example, for interaction neighborhoods in sessileorganisms, the regions from which resources are ob-tained or depleted, or the regions from which predatorsor mutualistic foragers are attracted by neighboring in-dividuals must be measured (e.g. Buss and Jackson1981, Manasse and Howe 1983).

Given a distribution of activity or influence, neigh-borhood size is then defined by the distance, area, orvolume representing some arbitrary but explicit pointon the distribution (Fig. 2a). lt. the concepr of neigh-borhood is to be useful, the decision criterion must beconsistent, explicit and reasonable. For population dy-namics we want to identi$ a neighborhood where mosrof the population growth is due to in situ reproductionrather than due to immigration (Connor et al. 1983).For interaction groups we would want to identify theneighborhood from which most potentially limiting re-sources are obtained (Antonovics and l-evin 1980). It isnot clear a priori whether a criterion of 50% , 67"/o or95% is reasonable Koeppl et al. 1975). Moreoyer, thearbitrary nature of the decision criterion is clearly

shown by comparing the importance of dispersal n a ge-netic and ecological context. The movement of a verysmall number of individuals can maintain genetic uni-formity among populations (Roughgarden 1979:203),but the same amount of dispersal may have little effecton population dynamics. Until there is an accumulationof studies of neighborhoods, it is not so importanr tharthe criterion be common to all workers, but it is imper-ative that the criterion be explicit. Fig. 2b provides anexample of how a decision criterion of.95"/" would beapplied to define three different neighborhoods for asingle organism

Why scating is necessary

Three major problems arise f the size of study units isarbitrary rather than based upon a scale appropriate toorganisms and questions. First, it will be difficult or im-possible o make comparisons f experimental or obser-vational results among different species n either thesame or different systems, because arbitrarily chosen

l l 11

L O C A L

? o ? u L A r l o x

343



rir :

istudy units may represent different scales (: neigh-borhood types) for each species. For example, in a studyof the establishment, xtinction, and dynamics of aphidson fireweed, Addicott (1978a) assumed ha t aphids onindividual ramets represented ocal populations for eachof four species of aphids. This arbitrary assumptionseemed easonable based upon the mobility of apterousaphids, population densities f aphids on individual ra -mets,'and the work of others (Hughes L963, Sanders

and Knight 1968). However, we now know that this as -sumption was appropriate for Aphb varians but not forMacrosiphum valerianae; there is little or no movementof apterous A. variaru from ramet to ramet within aclone of fireweed, whereas M. valerianae are relativelymobile (Antolin and Addicott, unpubl.). Therefore,analysis of data from individual ramets of fireweed re-lates to different levels of population structure in thetwo species. Comparisons of species with distinctly dif-ferent life histories will be even more difficult. For ex-ample, it will be difficult to omPare the responses topatterned environments for organisms as different asaphids (e.g. Addicott 1979) and birds (e.g. Rotenberryand Wiens 1980) unless we can tell whether or not eachhas been studied at the same relative scale.

The second problem arising from an arbitrary choiceof spatial units is that a given patch size need not necess-arily correspond to the particular ecological neighbor-hood appropriate for examining a given ecological mo-del (see C-onnor et al. 1983, Heads and l-awton 1983).Therefore, it will be difficult to relate field experimentsor observations to particular theoretical models or con-ceptual postulates, and conclusions\about those modelscould be flawed. This would clearly bi the case in thefireweed system, where individual ramets do not rePre-sent local population neighborhoods for M. valerianae.Data on M. valerianae on individual ramets should not

be related to models of the dynamics of patch occuPa-tion that are based upon local populations as the unit ofobservation. This does not meian hat it is necessarily n-appropriate to study s'6tems at arbitrary and conveni-ent spatial scales. However, it may be inappropriate tocompare data sets among different systems or with con-ceptual models. Since such comparisons are an esscntialpart of our science (C,onnell 1983, Schoener 1983), it isimportant that study units be based upon appropriateand clearly identified scales.

The third problem is that different Processes in thesame system may occur at differeirt scales, and there-fore it may not be sufficient to examine a particular sys-tem at only one spatial or temporal scale. For example,in the aphid-fireweed system there are a number of pro-cesses hat affect the dynamics of local populations ofaphids, but these processes occur at a variety of scales.Competition neighborhoods can be as small as the im-mediate vicinity of an individual aphid for behavioral in-teractions (e.g. Whitham lng\, individual ramets forcompetition based upon plant resour@s, or clones offireweed for competition for the services of ants (Addi-

344

cott 1978b). Predation neighborhoods an be as small assingle ramets or the predaceous mites of aphids, one ormore clones of fireweed for insect predators or parasit-oids, or all fireweed ramets n whole drainages or avianpredators of aphids. The variety of neighborhood sizescomplicates the study of organisms n patterned envi-ronments, because t is extremely difficult to examineprocesses imultaneously that may be occurring at twoor more spatial or temporal scales. However, an ade-

quate understandingof the dynamics or behavior

of asystem may demand us t that.

Discussion

ln this paper we have made three main points. First, webelieve that it is essential to distinguish betrveen threeconceptually distinct but biologically interrelated con-cepts: t) the distribution of environmental variables asrepresented by temporal and spatial patterns; 2) thearea of activity or influence of organisms during appro-priate periods of time that define what we call ecologicalneighborhoods; 3) the use of ecologicat neighborhoodsto scale environmental patterning, thereby allowing as-sessment of relative patch size, relative patch isolation,relative patch duration, and relative utilization of het-erogeneous patches (grain response).

Second, the concept of ecological neighborhoodsneeds o be general; there should not be ust one kind ofprocess or a limited number of time or dispersal dis-tances that should be used to scale ecological processes.lnstead, each process and interaction should be exam-ined to determine an appropriate neighborhood, de-fined by movement or influence during an appropriateperiod of time. Thus, a single, general'concept of eco-logical neighborhoods can be applied at a variety of lev-els, rather than having a series of apparently unrelated

concepts and terms such as ambit (Lloyd L967, Jumars1976), home range (Koeppl et al. tnl), territory, localpopulation (Wright 1946), etc. Therefore, we have gen-eralized the concepts of scaling and emlogical neigh-borhoods, as developed by Southwood Qyn) and An-tonovics and kvin (1980).

Third, scaling of eological observations is extrernelyimportant. Ecological neighborhoods provide a basisfor scaling the patterning of environments relative to or-ganisms, which in turn provides the basis for oompar=ingspecies n the same system, comparing processes:ainongorganisms in different sJrutems, making appropriatecomparisons between field studies and theoretical mod-els, and for designing field studies of population inter-actions at one or more appropriate scales.

Implementing the neighborhood concept by meas-uring neighborhood sizes and using them to scale envi-ronmental patterning will be hindered by at least fivemajor problems. First, there will continue to be conflictbetween the ease of choosing study units to correspondto patch sizes and the fact that ecological neighborhoodsdo not necessarily orrespond o patches. Similarly, th e

OIKOS a9:3 1987)



scale at which a study can be conveniently conductedmay not correspond very closely o a neighborhood thatshould be ultimately of interest to t he investigator.Third, as has been he case with genetic neighborhoods,it will be difficulr to actually obtain the data required toassess eighborhood izes. Fourth, the criteria fo r estab-lishing neighborhood size are necessarily rbitrary, andthere is as ye t no consensus n rhtionale for choosingone decision criterion over another. Finally, organismsare affected by a variety of processes, ach of whichmay have a different neighborhood size associated withit .

Despite hese practicar difficulties, t is imperative orecologists to begin to reporr the scale(s) at which theirobservations and experiments are made. This could in-volve a detailed analysis of neighborhood sizes relativeto the sizes of study units and patches, or it could simplyinvolve reporting qualitative impressions of the naturalhistory of lhe study organism to justify the use of a par-ticular study unit. However, the futuie development ofecology will be hindered not by rhe absence of ippropri-ate concepts and techniques for scaling environmental

patterning, but rather by the failure to recognize theoverriding importance of scaling for facilit"ting com-parisons between studies and comparisons betweln the-oretical postulates and empirical work

*:l::y!dg:^errrs- Thispaper was conceived n 19g2 uring a

ropulatro.n,Frology eminar n the role of heterogeneity-inecolo.g5r. c would like to thank the followingpirsons'for59a$ing nd_commenting n rhe manuscript: . H]drown, f. C.!Io!no, P. Kareiva, . e Marino, M. A.'Mort, p- W. price,J.R. Spence, nd J. A. Wiens.

Refercnces

Addicott, J. E, 1978a. he population ynamics f aphids onfireweed: a complllon oi local ani metapopulations.C^an. . Z&t. 56.2ss+2s64.

- t:?!b. Co_mpetitionor mutualists: phids and anrs. Can.I.7nl. 56:2@T2CFI6.

- lylg- A multispecies phid-ant ssociation: ensity epend-enoc and species-specific ffects. Can. J. Znl:57: 55g_569.

Antonovics, . and [rvin, D. A. 1980.The ecological nd ge-n:tt! @nsequences f densitydependent rigulation-inplants. Ann. Rev. Ecol. Syst. l: 4ll4SZ.

Brown, J. H. and Kodric-Brown, n. lg7. Tirrnover ates n in-:ylar biogeography: ffect of immigration on extinction.Ecolqy 58:44H49.

den Boer, P- J. 1981. n the survival f poputations n a het-glogj^neous nd variable environment. Oecologia Berl. )50:3L53.

Buss, . W. and Jackson, . B. C. 1981. lanktonicfood avail-ability and suspension-feeder bundance: vidence or insitu depletion. J. Exp. Mar. Biot. Ecol. 49 : 15l_161.

c-herson, . L- 1978.Predato_r-prgyheory and variability.Ann. Rev. Ecol. Sysr. 9: 32T347.

chew' F. s- 1977. oevolution f pierid butterflies nd heircruciferous oodplantr Il- The distribution f eggs n po-tential oodplants. Evolution l: 56g-579.

connell, J- H. 196t.The nfluence f interspecificompetitionand other actors n the distribution f the barnacle htha-malus stellatus. Ecology 42: 133-146.- 1983.On the prevalence nd etative mportance f inter-

2l OIKOS 49:3 t987)

- and Sousa, W. P. 1983. On the evidence needed o judgeecological stability or persistence. Am. Nat. r2r: 7g9_824.

C-onnor, E. F., Faeth, S. H. an d Simberloff. D. 19g3. Lraf_miners on oak: the rore of.immigration and in situ'repro-ductive recruitment. - Ecology 6a: llgl_204.

cunninghar, M- A. 1996. Disfrsal in white-crowned spar-rows: A computer simulation of the effect of study-areasize on estimates f locar recruitment. - Auk 103: 7g-g5.

Denno, R. F. , Raupp, M. J. , Thllamy, D. W."n d

R.i.h"l-derfer, c- F. 1980. .Migration in heterogeneous enuiron-ments: Differences in habitat selection 6.t*e.n the wingforms of the {mgrptrig.planthoppe r, p ro kerisa mirginai|

^ .(Homoptera: Delplracidie). - Eiology 6l: 8j9_S6t.Dobzhansky, T._?n.<l wright,-S. 1943."ben.tio oi naturar

populatio_$- X- Dispe-rsion ates in Drosophira pseudoob-scura. - Genetics ?3:30f340

Ford, R. G. and Krumme, D. W- 1979. The analysis of spaceuse patterns. - J. Theor. Biol. 76: 125__155.

Gould, F- and stinner, R. E. l9&4. Insects n heteroqeneoushabitats. In: Huffaker, C. B. and Rabb, R. L. (eOii, Eco-logical-entqgglogy. Wiley, New york, pp. aZTaacj.

Hasscll, M. P. 1980. &T" cohsequences or trauitat heteiogen-elty for population dynamics. - Oikos 35: l5Gl60.

Heads, P. A. and Lawton, J. H. 19g3. Studies on the naturalenemy gomplgx of the

holly leaf-miner: the effects of scaleon the detection of aggregitive respon_ses nd the implica-tions for biological control. - Oikoi 40:267_276.

Huffaker, c. B. 1958. Experimentar studies on predition: dis-pgrsigl factors and piedator-prey oscilrations. HirtaraiaT7:34T383.

Huqhg_s,_R. . 1963; population dynamics of the cabbage ap_hid Brevicoryne brassicae L.).- - J. Anim. Ecor,1i7 $-424.

Jumars, P. A- 1976. pgep-1ea species diversity: does it have acharacteristic scale? - J. Mir. Res. 34: Zli_246.

Kareiva, P. 1982. Experimental and mathematical analyses ofherbivore movem€nt: qu-antifying he influence oi ptunt

if;ffi'lf,Tf, #fllb:"-foraging-discrimination.Ecor.

KoeppL . W., Slade,N. A. and Hoffmann,R. S. 1975. bi -variate ome ange Sode_r ith possibte ppticaiioni;;rd-

ological ata analysis. J. Mammal. SO:'gi_90.

Krebs, . J. 1978- cology: he experimental narysis f distri-bution and abundance. nd EO. - Harper d Ror",-N"r"York.

levin, S. A. 1976.Population ynamic models n heterogen_eous environments. Ann. Rev. Ecol. Syst. 7: Zgljrc.- and Paine, R. T. 1974.Disturbance, atih formition andpT-m_ulily structure. proc. Nat. Alad. Sci. U.S.A.7l:27&T147.

Irvins, R. l%8. Frvolulion n changing nvironments: omethgore{gl explorations. prinlt6n Univ. pr"*s,-prin-

ceton, NJ.I-ewjs-,w. M., Jr. 198d. Evolutionary nterpretations f al-lelochemical nteractions n phytoitanttdn

"rgae.---nrn.at . 127: 184-194.!-!oy9,\,t. 1967. Mean crowding.- J. Anim. Ecol. 36: _30.MacArthur, R- H. and rvins,

-R.t96+. c;p.iitio",-trauitat

selection, nd cha.racter i_sprac_ementn ui"rciry in"iron-ment. Proc. Nat. Acad. Sci. U.S.A. 5l:'tZOl.intO.Mack,R- N. and Harper, . L. 1927. nterference n June

"n -lyal-si_srylal arternand neighborhood ffects. i-Ecot.65:31-363.Manasse, . S. and Howe, H. F. 19g3. ompetition or disper-sal agents lToqg. tropicll trees: nfluenies

"i".il-tUorr. _

Oecotogia Bert.) 59: tg5_190.May' R. M- 1974- cosystem atterns n randomry luctuatingenvironmenrs- [n : Rose,'R. nd Snet. e. ..irj, Fi"i."r.

specific competition: evidenceAm. Nat. 122:661-696.

from field experiments. _

345



.:.1

l , : .' _ !

in theoretical biology. Academic Press, New York, pp.l-50.

Paine, R. T. and lrvin, S. A. 1981. ntertidal landscapes: is -turbance and the dynamics of pattern. - Ecol. Monogr- 5l:r45-178.

Pielou, E. C. 1969. An introdu ction to mathematical ecology-Wiley, New York.

Pleasants, . M. and Zimmerrnan, M. 1979. Patchiness n thedispersion of nectar resources: Evidence for hot and coldspots. Oecologia Berl.) 4L:733-?8.

Pratt, D. M. 1966. Competition between Skeletonemacostatum

and Olisthodbcrc luteus in Narragansett Bay and in cul-ture. - Limnol. Oceanogr. ll 447455.Price, P. W. 1980. Evolutionbry biology of parasites. - Prin-

oeton Univ. Press, Princeton, NJ.Ripley, B. D. 1981. Spatial statistics. Wiley, New York.Roff, D. A. 1974. Spatial heterogeneity and the persistence of

populations. - Oecologia (Berl.) 15:245-258.Rotinberry, J. T. and Wiens, J. A. 1980. Temporal variation in

habitat structure and shrub steppe bird dynamics. - Oco-logia (Berl.) 47: L-9.

Roughgarden, J. D. Sn. Patchiness n the spatid distributionof a population caused by stochastic fluctuations in re-sources. Oikos 29:52-59.

- lng. ft9ory of populatio! genetics and evolutionary eq!-ogy: an introduaion. - Macmillan, New York.

Sanders, C. J. and Knight, F. B. f%8. Naturd regulation of thqaphid Pterocomma populifoliae on bigtooth aspcnin north-ern L,ower Michigan. - Ecology 49:23+244.

Schoener, T. W. 1983. Field experiments on interspecific com-petition. - Am. Nat. 122: 24U285.

Shorrocks, 8. , Atkinson, W. D. arid Charlesworth, P. 1979.Competition on a divided and ephemeral resource. - J.Anim. Ecol. 48: 899-908.

Southwood, T- R. E. L962. Migration of terrestrial arthropodsin relation to habitat. - Biol. Rev. 37 : L7l-2L4.

- lW. Habitat, the templet for ecological strategies? J.Anim. Ecol. 46: 337-165.

Spence, J. R. 1983. Pattern and process n co-existence of wa-ter-striders Heteroptera: Gerridae). - J. Anim. Ecol. 52 :

497-5r2.Thylor, F. 1980a. Optimal switching to diapause n relation tothe onset of winter. - Theor. Pop. Biol. 18: 125-133.

- 1980b. Timing in the life histories of insects. Theor. Pop.Biol. 18: lL2-124.

Whitham, T. G. 1979. Territorial behaviour of.Pemphigas gallaphids. - Nature (tond.) 279: t7a875.

Wiens, J. A. 1976. Population respooses to patchy environ-ments. - Ann. Rev. Ecol. Syst. 7: 8l-120.

Wilbur, H. M., Trnkle, D. W. and C-ollins,J. B. 194. Envi-ronmental certainty, trophic level, and resource availabil-ity in life history evolution. - Am. Nat. 108: 80H17.

Wilson, D. S. 1980. The natural selection of populations andcommunities. - The Benjamir/Cummings Publishing Com-pmy, Inc., Menlo Park, California.

Wright, S. 1943. solation by distance. Genetics 28: 114--138.- 1946. solation by distance under diverse systems f mating.

- Genetics 3l: 39-59.

346 OIKOS a9:l (t987)

Addicott, Aho Et Al (OIKOS) 1987

Documents

Transcript of Addicott, Aho Et Al (OIKOS) 1987