-
7/28/2019 Bush - 1994 - Sympatric speciation in animals new wine
in old bottles.pdf
1/4
REVIEWS
1718
19202122
Wiley, R.H. and Rabenold, K.N. (1984) Euolution 38,609-621Waser,
P.M. (1988) in The Ecology ofSock/ Behaoior (Slobodchikoff,C.B.,
ed.), pp. 109-130, Academic PressZack, S. and Rabenold, K.N. (1989)
Anim. Behau. 38,235-247Komdeur, J. (1992) Nature 358,493-495Emlen,
S.T. and Wrege, P.H. (1994) Nature 367, 129-132Heinsohn, R.G.,
Cockburn, A. and M ulder, R.A. (1990) Trends Ecol.
252627282930
Euol. 5,403-40723 Heinsohn, R.C. (1991) Am. Nat. 137,864-881
3124 Rabenold, K.N. (1984) Ecology 65,871-875 32
A though much has beenwritten and debated onthe origin of
species,,details of the actual pro-
Sympatricspeciation n animals:new wine in old bottles
cesses involved still elude us.Increasing concern for the
main-tenance of species diversity hasstimulated renewed interest in
howspecies arise in nature. The evol-ution of sympatric sister
species(those whose distributions par-tially or comp letely
overlap) hasbeen a particularly contentiousproblem. Leading
proponents ofthe neodarwinian orthodoxylJstill maintain that all
sexuallyreproducing, spatially overlapping,sister species result
from second-ary contact between spec ies thatevolved in geographic
isolation.Recent theoretical studies and re-search on the evolution
of natural
Austad, S.N.and Rabenold, K.N. 1985)Behau. Ecol. Sociobiol 17,
18-27Emlen, S.T. and W rege, P.H. (1991)5. Anim. Ecol.
60,309-326Malcolm, J.R. and Marten, K. (1982) Behar;. Ecol.
Sociobiol. 10, 1-13Goldizen, A.W. (1987) Behao. Ecol. Sociobiol.
20,99-109Ligon, J.D. and Ligon, S.H. (1978) Living Bird 17,
159-197Packer, C. et al. (1988) in Reproductive Success: Studies of
IndividualVariation in Contrasting Breeding ystemsClutton-Brock.
T.H., ed.),pp. 363-38 3, University of Chicago PressAlexander, R.D.
(1974) Annu. Rev. Ecol. Syst. 5,325-383Creel, S.R. and Creel, N.M .
(1991) Behao. Eco[. Sociobiol. 28. 263-270
Guy L. BushRecent research on natural host races
and symp atric sister species, c omp arativephylogenetic
analyses, laboratory
experiments and theoretical models hasgreatly strengthened the
case for
symp atric speciatlon. Traits evolving inresponse to divergent
selection
experienced by subpopulations adaptingto different habitats
provide sufflclent
intrinsic prema ting isolatlon for sympa tricspeclatlon to
occur. The initiationof speciatlon through a habitat shift in
animals which m ate within a preferredhabitat (such as many
phytophagousand parasltlc invertebrates and somevertebrates,
including birds) requires
few genetic changes.
Guy Bush is at the Dept of Zoology, Michigan StateUniversity,
East Lansing, MI 48824- 1115, USA.
populations challenge this widely held view. There is mount-ing
evidence that in some groups of habitat specialistssister species
may arise by sympatric speciation3 (Box 1)- a process described by
Kondrashov and Mina4 n geneticterms as one which in its course the
probability of matingbetween tw o individuals depends on their
genotypes alonerather than on physical barriers.
late between sympatric sisterpopulations as they adapt to
dif-ferent habitats are sufficient toinitiate and eventually
completethe speciation process in the ab-sence of physical
isolation. Evi-dence reviewed here indicatesthat adaptive traits,
such as habi-tat or host preference, selectionand fitness,
intimately involved inthe shift to a new niche are oftenthe same
traits that result in habi-tat specific assortative mating andthe
evolution of partial or com-plete reproductive isolation be-tween
sympatric populations(Box 3). Once identified, traitsresponsible
for habitat subdiv-ision among habitat specialistswhich mate only
within preferredhabitats are amenable to geneticscrutiny,
experimentation andtheoretical treatment. The time, place and other
circum-stances of the origin of such pop ulations are sometimes
known or can be deduced.
Species are natural popu lations sufficiently
geneticallydistinct from one another for each to follow
independentevolutionary paths (Box 2). To establish the mode an d
pr@cess of speciation in nature, it is necessary to determinehow a
reduction in gene flow occurs between sister popu-lations
sufficient to allow each to become irrevocablycomm itted to
different evolutionary paths. Upon reachingthis threshold of
genetic differentiation, which representsthe transitional stage
between race and species, gene flowmay affect the rate and pattern
of divergence, but not theoutcome. Unfortunately, we lack specific
genetic and bio-logical details about this boundary of irreversible
evol-utionary comm itment for any sexually reproducing animal.
Several broad and somewhat overlapping approachesare employed to
study and test hypotheses of sympatrichabitat race formation and
speciation in sexually repro-ducing organisms. These include: (1)
characterization ofgenetic polymorphisms responsible for habitat
choice anduse; (2) direct ecological, ethological and genetic
studiesof naturally established sympatric habitat races in the
prclcess of divergence; (3) determination of historical factorsand
traits responsible for the origin and maintenanceof sympatric
sister species; (4) analysis of cospeciationand habitat or host
shifting as revealed by patterns ofphylogenetic relationships; (5)
laboratory experiments thatattempt to simulate aspec ts of
allopatric a nd sympatricspeciation; a nd (6) development of
theoretical models ofthe speciation process.
The debate between those who believe sympatricspeciation occurs
in nature and those wh o do not boilsdown to whether or not genetic
differences which accumu-
Polymorp hisms responsible for habitat choiceand useA key
element in the initiation of sympatric speciationby habitat shift
is the acquisition of genetically based
TR E E 001. 9, no X Au gust 1.994 0 1994, Elsevier Science Ltd
28 5
-
7/28/2019 Bush - 1994 - Sympatric speciation in animals new wine
in old bottles.pdf
2/4
REVIEWS
Box 1. SpeciationThe p rocess o f c l adogenes i s ( sp l i t t
i ng o f li neages) w hereby gene f l ow i s reducedsu f f i c i en
t l y be tween s i s te r popu la t i on s to a l l ow each to
becom e i r revocab l ycom m i t ted to d i f f e ren t evo lu t i
onary pa ths . The fo l l ow ing m odes o f spec ia t i on canb e d
i s t i n g u i s h ed :A l l opa t r l c (w i th geograph i c i so
la t i on )D ichopa t r l c ( v i ca r l an t ) spec ia t i on :
The acqu i s i t i on o f spec ies s ta tus by phys i ca l l yi so
la ted s ubpop u la t i ons , each su f f i c i en t l y l a rge to
exc lude i nb reed ing as a fac to ri n t he spec ia t i on p
rocess .Per l pa t r ic ( f ounder e f fec t ) spec ia t l on : The
rap id acqu i s i t i on o f spec ies s ta tus asa by -p roduc t o
f i nb reed ing , se lec t i on , d r i f t , genom e reorgan i za
t i on and subs tan t i a lm orph o log i ca l and eco log i ca l
sh i f t s i n a geograph i ca l l y i so la ted po pu la t i on es
tabl i shed by a sm a l l num ber o f f oun d ing i nd i v i dua l
s .A l kp rapa t r i c ( re i n fo rcem ent ) spec la t l on : I n
i t i a l s tages o f spec ia t i on occur i ngeograph i ca l l y i
so la ted p opu la t i ons and a re then com p le ted i n sym pat
ry as a resu l to f se lec t i on aga ins t i n te r - rac ia l
hybr i ds .Nona l l opa t r l c (w i thou t geograph i c i so la t
i on )Parapa t rl c spec ia t l on : S i s te r spec ies evo l ve
wh i l e adap t i ng to con t i gu ous ,spa t i a l l y segrega ted
hab i ta t s ac ross a nar row con tac t zone .Sym pat r i c spec
ia t i on : New s i s te r spec ies evo l ve w i th i n t he d i
spersa l range o f t heo f f sp r i ng f rom a s ing le dem e.
Parapa tr i c and sym pat r i c s pec ia t i on represen t ex t
rem es o f a con t i nu um in t hepa t te rn and ex ten t o f hab i
ta t - im posed spa t i a l segrega t i on and gene f l ow reduc t
i ontha t occurs dur i ng non-a l l opa t r i c d i ve rgence .
Even i n cases o f sym pat r i c s i s te rpopu la t i on s , i n t
r i ns i c d i f f e rences w i l l genera te so m e spa t i a l
and /o r t em pora l s eg-rega t i on as they seek ou t and adap t
t o d i f f eren t f ragm ented and pa tchy hab i ta t s .
B o x 2. The species problemTry ing to exp la in spec ia t i on
w i th i n t he con tex t o f a p reconce i ved spec ies concep tp
laces the ca r t be fo re the hors e . I t i s an unders tand ing o
f t he fac to rs t ha t resu l ti n t he reduc t i on and even tua
l e l im ina t i on o f gene f l ow be tween s i s te r popu la t i
on s- t he ve ry p rocess o f spec ia t i on i t se l f - t ha t i
s necessary be fo re a c l ea r spec iesde f i n i t i on i s poss
ib le . To es tab l i sh the p roper t i es tha t cons t i t u te t
he boundarybe tween s i s te r spec ies , we m us t f i r s t i den
t i f y key t ra i t s respons ib le f o r m aterecogn i t i on , s
uch as those respons ib le f o r hab i ta t cho i ce , non-hab i ta
t assor ta t i vem at i ng , f it ness and genom ic i ncom pat i b
i l i t y . Spec ia t i on research s hou ld focus onthe gene t i
cs o f spec ia t i on ra ther t han on the gene t ics o f spec ies
d i f f e rencesse .
Spec ies c l ea r l y ex i s t i n na tu re , ye t i t i s doub
t fu l whe ther a s i ng le , un equ i voca lde f i n i t i on o f
t he spec ies c a tegory c an sa t is f y a l l t he b io log i ca
l and ph i l osoph i ca lc r i t e r i a deem ed essen t i a l by
au thor i t i es o f d i f f e ren t t axas . Therefo re , I l eave
i t t othe reader t o se lec t any o f t he de f i n i t i ons p
resen ted i n Ereshe fsky ssg an tho log yo f spec ies conc ep ts
and the i r a t t endan t ph i l oso ph i ca l p resum pt ion s .
The p rocessand ou tcom e o f spec ia t i on ( sym pat r i c o r o
therw ise ) does no t depend on thea pr i o r i i nvoca t i on o f
any par t i cu la r sp ec ies de f i n i t i on .
differences in habitat preference and fitness between theparent
and daughter habitat races. When new habitat prefer-ence alleles
arise in a population they ca n initiate habitatshifts. Whe n
mating is restricted to preferred habitats,these new alleles can
lead to the establishment of geneti-cally distinct sympatric
habitat race s. Over time such racesmay evolve into distinct
species.Genetic polymorphisms which govern habitat prefer-ence and
habitat-related fitness appear to be common insympatric natural
populations. Some adaptive feedingpolymorphisms are controlled by
single genes, and pro-duce rather spectacular phenotypic
differences such ashandedness of mouth opening in Perissodus
microlepis, aLake Tanganyika scale-eating cichlids, and bill size
poly-morphism in the African finch, Pyrenestes o.
ostrinus6.Moresubtle, but equally important, habitat- or
resource-prefer-ence polymorphisms are also well documen ted3JJ.
Resultsfrom studies on insects and other invertebrates suggestthat
habitat preference may sometimes involve only a fewalternative
alleles at a single locus, a major locus withmodifierssJ19J0, or
require the action of many locir. Even
286
when the genetic ba sis for habitat preference and fitnessis
polygenic a s in recently established habitat races of anestuarian
gammarid amphipod, Eogummarus conferuicolus,divergence can occur
rapidly and contribute significantlyto sympatric population
subdivision*r.The study of naturallyestablished habitat racesThe
study of sister populations in various stages ofgeographic and
habitat race form ation and speciation is themost reliable way to
distinguish key genetic changes re-sponsible for speciation from
other differences accum u-lated after reaching the speciation
threshold. A habitat orhost race is a population of a species that
is partiallyreproductively isolated from other conspecific
sympatricsister populations as a direct consequence of adaptationto
a specific host or habitatlzJ3. Postmating hybrid incom-patibility
betw een habitat races is absent, and the racesshow
habitat-associated tradeoffs in fitness. Although sym-patric
habitat races offer the best opportunity to studyspecies in statu
nascendi, evidence for their existence w asinitially regarded as
unconvincing and the possibility oftheir sympatric origin
questioned on theoretical gro undslJ4.Contributing to this view is
the fact that con ditions thatpromote the formation of habitat
races, such a s the avail-ability of a new habitat or resource,
occur infrequently.Furthermore, once established, new habitat and
host racesappear to rapidly acquire the attributes of
species.Criteria for the recognition a nd characterization of
suchraces are now available3J3. It is not surprising that
mostexamples of habitat races involve insects that have shiftedonto
introduced hosts and have become major pests. Forthis reason, their
origin and subsequent history are fairlywell documented.Of the
several well-documented examples of sympatrichabitat races, the
most thoroughly studied is the nativehawthorn (Crutaegus spp.)
infesting haw fly (Rhugoletispomonella). This species established a
new host race onintroduced Old World apples (M alus pumukz) well
withinthe range of Crataegus along the Hudson River, USA, inabout
186013. he apple population diverged in several keytraits from the
native haw flies and now represents a dis-tinct race1 5 or
incipient species. Not only do the two sym-patric race s differ
genetically in host preference, eclosiontime and host-associated
fitness trade-offs, bu t they alsoconsistently maintain distinct
frequency differences a tseveral allozyme loci. A considerable
degree of allochronicand spatial isolation occurs between these
host races be-cause they mate only on their host plants which fruit
atdifferent times. Also, a genetically ba sed difference betweenthe
races in their response to chemical cues emitted by thehost fruit
plays a key role in host choicel3. This restrictionof movement
between host species by genetic differencesalone is sufficient to
allow strong habitat-directed selectionto overcome the effects of
gene flow, and thus maintainthe distinct racial difference from
year to year. Sim ilarcases of habitat race formation by recent
habitat shiftsare documented in several invertebrates11J2J6-18.The
study of sympatric sister species
The mate recognition systems, life history traits andlack of
convincing evidence of prior geographic isolation ofseveral
sympatric sister species suggests that they evolvedsympatrically
rather than as a result of geographic isolation(consult Ref. 3 for
specific citations). They are h abitatspecialists which use their
habitat or host as a rendezvousfor courtship and m ating and
generally show little evidenceof postmating incompatibilities.
Represented are DipteraTR EE uol. 9, no. 8 Augu st 1994
-
7/28/2019 Bush - 1994 - Sympatric speciation in animals new wine
in old bottles.pdf
3/4
REVIEWS
(Rhagoletis an d Drosophifa3), Neuroptera (lacewings3),Homoptera
(treehoppers and aphidss), Hymenoptera(socially parasitic ants and
parasitic waspss1i9),Lepidoptera(Spodoptera moths and Heliconius
butterfliess), Acarina(humm ingbird flower mitesa), Crustacea
(parasiticcopepodsQ Mollusca (pulmonate snaik=G), eleostei
(rain-bow smelt*0 and sticklebacks*) and Aves (crossbill+).Although
each example combines a unique set of biologi-cal conditions, they
share one important characteristic:mate choice occurs only on or
within a preferred habitator host. Once it appears in such h abitat
specialists, geneticvariation for habitat preference promotes
habitat assort-ative mating resulting in population subdivision,
adaptationand divergences. Subdivision and the reduction in gene
flowunder these conditions result from genetic rather
thangeographic barriers.Comparative phylogenetics of sister
speciesPhylogenetic reconstruction offers an indirect approachfor
establishing the mode of speciation within a taxon bytesting for
concordance between patterns of relationshipsand distribution among
sister species*s. Lynch24, usingphylogenetic trees, compared
species ranges of represen-tative vertebrate groups (freshwater
fish, frogs and birds)with ancestral ranges estimated from the sum
of all de-scendents ranges. He found that classical dichopatric(Box
1) events explain the origin of most species (71%).However,
post-speciation dispersal could not account forthe origin of some
sister species (6%) which appe ar to havespeciated sympatrically.
Although there was evidence ofperipatric speciation (Box 1) (15%),
Lynch suggests that,in vertebrates, this mode of speciation may be
even rarerthan sympatric spe ciation. Microdichopatric or
parapatricdivergence is a more parsim onious argumen t for their
evol-ution because dispersal does not explain the distributionof
species in the groups studied. A general w eakness ofthis approach
is its inability to distinguish between micredichopatric,
allo-parapatric, peripatric and parapatric di-vergence (Box 1).
Without ecological data, it is difficult toevaluate the incidence
of non-allopatric speciation by anecotone or habitat shift.The
contribution of habitat shifts to the speciationprocess becomes
more apparent when it is possible tofactor ecological information
into a phylogenetic analysis.For example, there is a tight
correlation between color pat-terns use d in interspecific and
intraspecific comm unicationby warblers of the genus Phyfloscopus
and light intensityof the habitatzs. Species with the brightest
colors inha bitdark, forest habitats while those with duller
plumage livein open habitats. Sister species w ith contrasting
plumagepatterns occupy adjacent habitats. This parapatric patternof
association between sister species, coup led w ith the factthat
conspicuousness is habitat related and mating occurswithin
preferred habitats, sug gests that these specieshave diverged
parapatrically through a combination ofhabitat shifts and sexual
selection acting on courtshipand territorial displays.Phylogenetic
methods are useful in evaluating the inci-dence of cospeciation
(speciating with the host) versushost shifts in the evolution of
insect grou ps. Mitter et ~1.26found that clades of phytophagous
insects are consistentlymore species rich than their primitively
non-phytophagoussister groups. This suggests that herbivory
promotesdiversity. In a survey of examples from 14
phytophagousinsect families and genera26, 10 showed little or no
phylo-genetic concordance between the insect and host
plantphylogenies, three showed partial concordance, and oneTR EE u
ol 9. no N Auqu st 19.94
-J3ox 3. Sympatric speciation and reinforcement
An obs tac le to reach ing a consensu s on the l i ke l i hood o
f sym pat r i c spec ia t i on i sthe w ide l y he ld percep t i on
tha t i t r equ i res so m e fo rm o f rei n fo rcem ent40 .
However ,i n an im a ls t ha t m ate w i th i n a p re fe r red hab
i ta t a f t e r d i spersa l , t he reduc t i on o fgene f l ow
and i so la t i on be tween subpo pu la t i ons evo l ves wh i l e
t hey adap t t o d i f -f e ren t hab i ta t s , no t as an ou tcom
e o f se lec t i on to reduce the p rodu c t i on o f i n fe r i o
rhybr i ds32 . D i rec ted se lec t i on w i th i n each hab i ta t
p rom o tes a runaway p rocessz8 .Th i s com b ina t i on acce lera
tes i ncorpo ra t i on o f va r i an ts t ha t im prove f i t ness
andhab i ta t cho i ce , t hus reducingrrors i n m ate cho i ce . I
t a l so resu l t s i n t he reduc t i ono f com pet i t i on fo r
m ates and resources , and reduces o r e l im ina tes se lec t i on
tores t r i c t hybr i d i za t i on be tween subpopu la t i ons
.
showed extensive concordance and thus evidence of co-speciation
where speciation of parasite and host apparentlyoccurred
simultaneously. The majority of speciation eventsin these
phytophagous insect specialists are accompaniedby shifts to new
hosts, an d cospeciation events are m uchless frequent. If this
pattern holds for the m ajority of otherhost or habitat spe
cialists, such as insect parasitoids,nematodes and mites with
similar parasitic life histories,then sympatric divergence via host
shifting may be verycomm on. Although it is argued that allopatric
speciationin parasites occurs via host shifts in peripheral
isolatesfollowed by the extinction of the original hosts23.27,
vi-dence from contemporary host and habitat shifts, such
asRhagoletis and other examples cited above, does not sup-port this
view. These shifts occurred well within the nor-mal dispersal range
of the parent population and werefollowed by rapid genetic and
phenotypic divergence.Laboratory experiments of allopatric and
sympatricspeciationRice and Hostertza review m any of the
experiments ofthe past half century that attempted to duplicate all
or partof the processes of allopatric and non-allopatric
speciation.Their survey found little experimental evidence that sam
-pling drift, genetic bottlenecks and reinforcement aredirectly
responsible for the development of pre- or post-mating reproductive
isolation, although these factors maycontribute to divergence which
in turn m ay pleiotropicallycause isolation. This view is supported
by a recent detailedstudy involving replicated large-scale
experimental tests.Caliana et ~1.~~ found that a change in
assortative matingleading to speciation is not likely to occur
under conditionsof the founder-flush-crash model of speciation. M
ountingexperimental** and theoretical3O evidence indicates
thatperipatric speciation (Box 1) by founder effects, viewedas the
major source of new specie9, occurs infrequentlyin nature. In
contrast, there is substantial experimentaldocumentation for the
evolution of reproductive isolationas a result of strong selection,
pleiotropy and/or hitchhikingeffects28, with or without physical
isolation.Theoretical models and computer simulationsThe
maintenance of habitat-specific polymorphismsoccurs most easily
when individuals of different genotypesare able to select the
habitat in which they are most fit.Jaenike and H olt7 modeled a
life cycle in which there ishabitat selection and global random
mating as follows:(start generation t) zygote formation * dispersal
andhabitat choice * habitat residence and selection (differ-ential
mortality) j random m ating across entire population(start of
generation t + 1). They confirmed previous studiesthat habitat
selection coupled with some frequency and/ordensity dependent
selection (soft selection) broadens con-ditions for the maintenance
of niche polymorphisms.
28 7
-
7/28/2019 Bush - 1994 - Sympatric speciation in animals new wine
in old bottles.pdf
4/4
REVIEWS
However, as long as habitat preference is random withrespect to
mate choice, as in the above life cycle, habitatrace formation and
speciation will not occur. Gene flowprevents the formation of
strong linkage disequilibriumbetween habitat related preference and
fitness alleles, a pro-cess require d for the evolution of
different mate-re cognitionsystems between races and species. This
reduction in geneflow occurs when m ating takes p lace at random
within ahabitat following dispersal to a preferred habitat, such
asin the follow ing life cycle: (start gener ation t) zygote
for-mation a habitat residence and selection (differential
mor-tality) * dispersal an d habitat choice * random matingamong
individuals w ithin a preferred habitat (start of gen-eration t +
1). Com puter simulations of single- or multiple-variation models
examine the p roperties of this life cyclebased on a correlated
response between habitat selec-tion and mate choice and their
effects on sympatricspeciationsJlJ2. Assum ing independent
regulation of popu-lation size, such habitat-specific assortative
mating pro-motes the fixation of mutations that increase
habitatfidelity and survivorship. These m odels show that
sym-patric race formation and speciation may occur whenindividuals
with strong habitat-dependent positive assort-ative mating
experience superior fitness over genotypesof intermediate fitness.
Conditions for sympatric specia-tion are further relaxed if habitat
preference is epistaticto fitnes9.
AcknowledgementsI thank Don Hall, Dan Howard, Richard Lenski,
Jim Smithand Don Straney for helpful comm ents on earlier drafts
ofthis article, and for commen ts of the three
reviewers.References1
2
3
9
This contrasts with the results of Felsensteinss, whoexplored a
model w ith variation in non-habitat-associatedpositive assortative
mating and habitat fitness genes basedon the following life cycle:
(start gener ation t) zygoteformation 3 habitat-independent
assortative ma ting 3random dispersal across habitats 3 habitat
residence andselection (start of genera tion t + 1). In this model,
habitatchoice is random and adults mate assortatively indepen-dent
of the habitat. W hen there is no correlated responsebetween
habitat choice and mate choice, conditions ofselection, gene flow,
linkage and gene penetrance necessaryto surmount the antagonism
between recombination andselection are so restrictive that
non-allopatric speciationis unlikely to occur.
1011121314151617181920212223
24
Obviously, life-cycle and mating-system componentssignificantly
influence the speciation process. To broadenour understanding of
sympatric speciation, it is essential toexplore models
incorporating other factors that promotedivergence, such as,
runaway sexual selection34, populationsize4, temporal isolation35,
resource availabilitya, environ-mental com plexity and
variabilitys, small body size andshort generation time36, and
non-genetic environmentaleffectssr. These models should also
examine the effects ofcombining both non-habitat- and habitat-based
assortativemating genes.
2526272829303132
3334353637
3639
40
Mayr, E. (1988) Towards a New Philosophy ofBiology:
Obseruationsofan Euolutionist, Harvard University PressCarson, H.L.
(1989) in Genetics, Speciation, and the Founder Principle(Giddings,
L.V., Kaneshiro, K.Y.and Anderson, W.W., eds), pp.345-362, Oxford
University PressBush, G .L. (1993) in Euolutionary fottems and
frocesses (Lees, D.R.and Edwards, D., eds), pp. 22 9-249, Academic
PressKondrashov, AS. and Mina, M.V. (1986) Biol. J. Linn. Sot.
27,201-223Hori, M. (1993) Science 260,216-219Smith, T.B. (1993)
Nature 363,618-620Jaenike, J. and Holt, R.D. (1991) Am. Nat.
137,S67-S90de M eeus, T., Michalakis, Y., Renaud, F. and Olivieri,
1. 1993) Euot.Ecol. 7, 175-198Carson, H.L. and Ohta, A .T. (1981)
in Euotution Today (froc. 2ndInter. Gong. Syst. and Euot. Biot.)
(Scudder, G.G.E. nd Reveal, J.L.,eds), pp. 3 65-370, Hunt
Institute, Carnegie Mellon Universityde Bell, J.S. and Sokolowski,
M.B. (1989) J. Insect Behau. 2,291-299Stanhope, M.J., Hartwick, B.
and Baillie, D. (1993) Mol. Ecol. 2,99-l 12Diehl, S.R. and Bush,
G.L. (1984) Annu. Reu. Entomot. 29,471-504Bush, G .L. (1993) Psyche
99,335-357Futuyma, D.J. and Mayer, G.C. (1980) Syst. Zool.
29,254-271Feder, J.L., Chilcote, CA. and Bush, G.L. 1990) Euolution
44,570-594Carroll, S.P. and Boyd, C. (1992) Euolution 46,
1052-1069Yoshihisa, A. (1991) Entomot. Exp. Appt.
58,15-20Vouidibio, J. et al. (1989) froc. Nat/Acad. Sci. USA
86,8442-8446Whitfield, J.B. (1990) Euol. Theory9,211-213Taylor,
E.B. and Bentzen, P. (1993 ) Evolution 47,813-832Schluter, D. and M
cPhail, J.D. (1992) Am. Nat. 140,85-108Benkman, C.W . (1993) Ecol.
Monogr. 63,305-325Brooks, D .R. and McLennan, D.A. (1993)
Pamscript: Parasites and theLanguage ofEuolution, Smithsonian
Institution PressLynch, J.D. (1989) in Speciation and Its
Consequences (Otte, D. andEndler, J., eds), pp. 527-5 53,
SinauerMarchetti, K . (1993) Nature 362, 149-152Mitter,
C.,Farrell,B. and Futuy ma,D.J. 1991)Trends Ecol. Euol.
6,290-293Mayr, E. (1963)Animal Species and Euolution, Harvard U
niversity PressRice, W.R. and Hostert, E.E. Evolution (in
press)Galiana, A., Moya, A. and A yala, F.J. (1993) Evolution
47,432-444Coyne, J.A. (1992) Nature 355,511-515Rice, W.R. (1984)
Evolution 38, 1251-1260Diehl, S.R. and Bush, G.L. (1989) in
Speciation and Its Consequences(Otte, D. and Endler, J., eds), pp.
345-365, SinauerFelsenstein, J. (1981) Euotution 35, 124-138Wu,
C-l. (1985) Evolution 39,66-82Butlin, R.K. (1990) Heredity
65,47-50Marzluff, J.M. and Dial, K.P. (1991) Ecology
72,428-439Wood, T.K. (1993) in Euotutionaryfatterns and Processes
(Lees, D.R.and Edw ards, D., eds), pp .299-317, Academic
PressTempleton, A.R. (1981) Annu. Rev. Ecot. Syst.
12,23-48Ereshefsky, M. (1992) The Units ofEuotution: Essays on fhe
Nature ofSpecies, MIT Press
Conclusions Paterson, H.E.H. 198 2) S. Afr. J. Sci. 78,53-57It
is no longer possible to dismiss sympatric divergenceas a mode of
speciation in anima ls, particularly in thosetaxa in which m ate
choice is directly dependent upon hab i-tat or resource choice. Sym
patric speciation is likely to befar more comm on than has been
assumed in some sex-ually reproducing anima l groups such a s
insects, m itesand nematodes which represent the majority of
animalspecies. We need in-depth biological and evolutionaryresearch
on allopatric and sympatric habitat races andsister species of
known ag e and place of origin in thesegroups. Only then can we
gain a true perspective onthe frequency of allopatric and
non-allopatric m odes ofspeciation.
288 TR EE uol. 9. no. 8 August 1994
