the origin of new species - wesleyan...

5

THE ORIGIN OF NEW SPECIES

Gillia tenuiflora, a member of the phlox family, and Oligo-

dranes sp. in the family of bee flies, Bombyliidae—the most

numerous and effective pollinator in a population near Cres-

ton, California. (Based on illustrations in Grant and Grant 1995)

5.1 Geographic Specia-

tion, or Allopatric Spe-

ciation

5.2 Sympatric Speciation

5.3 Parapatric Speciation

5.4 Speciation Mecha-

nisms and the Kinds

of Species Involved

5.5 Primary Speciation in

Plants

5.6 Secondary Speciation

in Plants

5.7 Agamic Systems

Summary

74

GEOGRAPHIC SPECIATION, OR ALLOPATRIC SPECIATION 75

arwin's (1872) model of speciation must beD

teased out of his extensive discussion on thesubject, but a careful reading reveals a rather

simple view of the mechanism involved with the evo-

lution of new species. First a species spreads out over

a large area (see Figure 3-4). Then environmental vari-

at ion within this area produces different variations in

t he populations of different environments. Restrictedmovement of individuals between environments re-

sults in independen t natural selection and finally a

new species in each environment.Hybrid zones give rise to maladapted individuals

that survive poorly in both parental environments.Because environments are not completely discrete,variation between species is continuous. The hybridpopulations are small and must compete with paren-

tal populations; they are likely to eventually go ex-

tinct.It will be interesting to keep Darwin's model in

mind and compare it with other speciation modelsthat developed subsequently.

5.1 GEOGRAPHIC SPECIATION, OR

ALLOPATRIC SPECIATION

After Darwin's time, a major building block of theModern synthesis was Mayr's model of geographicspeciation, or allopatric speciation. The biologicalspecies concept serves as an essential component ofgeographic speciation, for it explains how two ormore populations that once shared a common genepool become reproductively isolated. Mayr's viewthat the specjation process is almost exclusively geo-graphic has been largely accepted. As we shall see,alternative mechanisms are also important, and viewson the relative contributions of those mechanismsto the diversity of life on earth are shifting. Eventextbooks on evolution debate the validity of modesof speciation other than geographic speciation.Clearly, no consensus has been reached (cf. Cockburn1991, Futuyma 1986, Minkoff 1983, Ridley 1993,Strickberger 1990, and this chapter).

We may call Mayr's views on geographic specia-tion the classic model of speciation, as they werelargely accepted in the 1940s and are still intact. Theclassic model invokes a geographic barrier as an es-sential part of the scenario (Figure 5-1). A geographicbarrier divides a population, gene flow between thetwo parts stops, each new population evolves inde-pendently of the other, and differences accumulateu ntil the members of one population cannot breedwit h members of the other even when the rangesagain overlap. Thus, two new species have evolved.

detailed (1992) emphasizes the need for a more

aetailed mechanistic understanding of each phase of

allopatric speciation. He notes that the process can bedivided into three stages: the formation of an isolatedpopulation, the persistence of that population, andthe differentiation of the population from the parentalstock. The heuristic value of this subdivision lies inthe increased focus on each phase. For example, inisolate formation, what are the relative roles of extrin-sic mechanisms, such as glaciation effects, and intrin-sic mechanisms, especially dispersal ability or vagil-ity? Does isolate persistence depend on populationsize and stability, ecological amplitude or specializa-tion, environmental rigor, niche space, or adaptation?What are the key processes in isolate differentiation?Keeping Allmon's arguments in mind will help usassess the thoroughness with which scientists havedescribed each process of speciation discussed inthis chapter.

Speciation in Warblers

Mengel (1964) developed an example of geographicspeciation involving northern wood warblers (inNorth America) that divided into two large popula-tions. In the black-throated green warbler group areone eastern species, the black-throated green warbler,Dendroica virens, and three western species with ex-tensive distributions (Figure 5-2): Townsend's, her-mit, and black-throated gray. Mengel argued that theparental species with a southeastern distributionadapted to coniferous forest during the PleistoceneNebraskan glaciation, which compressed vegetationtoward the south. During the interglacial period, thewarblers expanded north and west with the conifers.The next glaciation, the Kansan, reached its southernextremity in central North America and constituteda geographic barrier separating the eastern parentalspecies from a western population. Speciation oc-curred. The process was repeated in the next intergla-cial period, the Yarmouth, and the Illinoian glaciation,cutting off a second western species. A third warblerspecies in the West originated in a similar mannerafter the Sangamon interglacial period and the Wis-consin glaciation. This scenario also fits the distribu-tion of species in the Nashville warbler group, rein-forcing the view that large-scale geographical barriersformed by glaciation caused similar patterns of war-bler distribution and (avoiding complexities not tobe discussed here) explaining why three species ofthe black-throated green warbler group occur in theWest and only one occurs in the East.

Speciation in Darwin's Finches

Alternatively, only one inseminated female or a veryfew individuals may colonize a new locality—per-haps an oceanic island 500 miles off the mainland-

PARENTALSPECIES

BARRIER DIVIDESPARENTAL SPECIES

POPULATIONS EVOLVEINDEPENDENTLY

REPRODUCTIVEISOLATING MECHANISMSEVOLVE GRADUALLYAND SLOWLY

ON SECOND CONTACTTHE NEW SPECIES RETAINDISTINCT GENE POOLS

EH

76 THE ORIGIN OF NEW SPECIES

FIGURE 5-1

Geographic, or allopatric, speciation by division of populations resulting from a geographic barrier.

and found a new population. The founder effect,discussed in Section 5.4, can be defined as the found-ing of a new population by one or a few individuals,which carry only a small proportion of the geneticvariation of the parental population. With the foundereffect, sampling error becomes an important in-fluence in evolutionary change, as Section 5.5 andChapter 15 discuss. The new population evolves inde-

pendently of the parent population and eventuallybecomes reproductively isolated from it.

Because the ocean is so effective as a geographic

barrier, such colonization events are extremely rare.

We find an example in the original colonization ofthe Galapagos Archipelago by a South Americanfinch, the formation of a new species, and subsequentspeciation as new islands were colonized. The speciesnow called Darwin's finches (Figure 5-3) are theresult.

Geographic Barriers

Thus, geographic barriers may take very different

forms. For large mammals and birds, they may belarge expanses of ocean, high mountain ranges, or

Hermit WarblerDendroica occidentalis A fourth Western

species?

Townsend's WarblerDendroica townsendi

,ack-throated Green WarblerDendroica virens

Black-throated Gray WarblerDendroica nigrescens

Wilson's WarblerWilsonia pusilla

GEOGRAPHIC SPECIATION, OR ALLOPATRIC SPECIATION

77

FIGURE 5 - 2

Allopatric speciation of Dendroica warblers, based on Mengel's (1964) argument that major glaciations produced geographicbarriers, resulting in a new species of warbler in western North America after each event. Note that another glaciationevent is likely to compress the distribution of D. virens to the south and cut off another species in the west. The additionalbird illustrated is Wilson's warbler (Wilsonia pusilla), which has a trans-American range, posing the question of how thisspecies escaped the geographic barrier of an ice cap and consequent speciation.

glaciers (e.g., Mengel 1964) (Figure 5-3). For terrestrialanimals in general, they may be deep river valleys,wide rivers, or deserts. For fishes, they may be water-falls, narrows, or salt water. And for smaller orga-nisms, these and many more kinds of barriers arelikely to play roles. The geographic barriers resultingin geographic isolation are clearly environmentalproperties that constrain gene flow between popula-tions.

Reproductive Isolating Mechanisms,or Reproductive Barriers

Reproductive isolating mechanisms are the proper-

ties of individuals, populations, and species that con-strain gene flow between populations and species.

These properties evolve during geographic isolationand may take many different forms. Mayr (1963, 91)defined reproductive isolating mechanisms as "bio-logical properties of individuals that prevent the in-terbreeding of populations that are actually or poten-tially sympatric." His classification scheme illustratesthe great diversity of the mechanisms that are poten-tially involved in reproductive isolation (Table 5-1).Note that in hybrid sterility, so long as the F1 hybridhas a lower fitness than either of the parental types,the two populations remain isolated. Also, there areconsiderable differences between premating andpostmating barriers. In the case of the former, no lossof time, energy, or gametes occurs in seasonal or

habitat isolation or in ethological isolation, and some

loss of time and energy may occur in mechanicalisolation. Postmating mechanisms, however, waste

Dispersal andcolonization ofnew island

Reproductive isolation fromfounding population is acquiredduring geographic isolation

FLOREANA

PINTA

GENOVESAMARCHENA

Secondary contact isestablished and twospecies become sympatric

/Very rare colonizationfrom across a strong geographic barrier

- ..Geospizamagnirostris

FromSouthAmerica

SANTIAGO

BALTRA

FERNANDINA

One or twofoundersestablish apopulation

SANTA FE

78THE ORIGIN OF NEW SPECIES

FIGURE 5-3

Allopatric speciation by the founder effect, based on Grant (1986) relating to speciation of Darwin's finches in the Galapagos.The bird illustrated is the large ground finch, Geospizo magnirostris.

Another colonization event

time, energy, and gametes, and so these mechanismsare particularly inefficient in terms of energetics andreproductive effort. Although natural selection maybe expected to increase efficiency in energy and ga-mete allocation and effect a transition from postmat-ing to premating mechanisms as populations and spe-cies diverge, researchers have had difficulty findinggood evidence for reinforcement of premating isola-tion (Butlin 1989).

One can develop a legitimate argument that iso-lating mechanisms are actually a whole set of charac-teristics that usually facilitate reproduction within a

species, as Chapter 4 discusses (Paterson 1985, Tem-

pleton 1989). The emphasis should therefore be on the

positive mechanisms involved with mate recognition,

acceptance, and interfertility rather than on the resul-tant reproductive isolation between species. That is

why Paterson (1985) emphasized his recognition spe-cies concept. Attention to positive mechanisms maysuggest a term such as reproductive barrier ratherthan reproductive isolating mechanism. However, ageographic obstacle can also act as a reproductivebarrier, and so our preference is to keep the two termsdistinct: a geographic barrier is a property of the envi-ronment, as described earlier, and a reproductive iso-lating mechanism is a property of individuals, as

just explained.

Secondary Contact

After geographic isolation and the divergence of pop-ulations, secondary contact may be established. Theresults of such contact vary depending on the dura-tion of isolation, the effectiveness of the geographic

GEOGRAPHIC SPECIATION, OR ALLOI'ATRIC SPECIATION

79

TABLE 5- i Reproductive isolating mechanisms proposed by Mayr (1963)

Type

Mechanism and Example

2. Zygotic mortality

3. Hybrid inviability

4. Hybrid sterility

Transfer of gametes is prevented.

Potential mates do not meet; e.g., small-mouthed salamander, Arnbystoma texanum,breeds in ponds, but A. barbouri breeds in streams (Kraus and Petranka 1989).

Potential mates meet but do not mate. Some species of Allonemobius crickets (seeChapter 4), leopard frogs (Hillis 1988). In pollination systems, a pollinating insectmay carry pollen but does not deposit it on the stigma.

Copulation is attempted, but no transfer of sperm takes place. Many insect genitaliarequire precise lock-and-key link for transfer of sperm (cf. Eberhard 1985).

Sperm transfer takes place but egg is not fertilized. Pseudogamy in parthenogeneticPoeciliopsis related, for example, to the Sonoran topminnow (e.g., Vrijenhoek 1984),plants, and nematodes.

The egg is fertilized but the zygote dies. Some leopard frogs in the Rana pipienscomplex (Hillis 1988).

Zygote produces an F 1 hybrid of reduced viability, or hybrid recombinations are lessviable—e.g., frogs in the genus Pseudophryne (Woodruff 1979).

F1 hybrid is fully viable but partially or completely sterile, or it produces a deficientF2 . Crosses between horses and donkeys produce mules, which are sterile. Somehybrids in the Rana pipiens complex (Hillis 1988).

A. Premating mechanisms

z. Seasonal and habitat

isolation

2. Ethological isolation

3. Mechanical isolation

B. Postmating mechanisms

Gametic mortality

barrier, and ultimately the degree to which reproduc-tive isolating mechanisms have evolved.

Let us call one scenario case 1. After very effectiveand long-term isolation, on secondary contact popu-lations have diverged into discrete ecological nichesand reproductive isolation is complete. New speciesbecome sympatric without any mutually inducedevolutionary change (Figure 5-4).

Case 2 may occur after less prolonged geographicisolation or because of slower evolutionary rates, suchthat ecological divergence may be less complete, al-though reproductive isolation may be complete. Dur-ing secondary contact, competition between the twonew species is likely to result in mutually induceddivergence in ecologically relevant traits (Figure5-4). This process, which Darwin (1859) called charac-ter divergence and Brown and Wilson (1956) calledcharacter displacement, need not involve morpho-logical characteristics. Rather, it can involve onlycharacteristics of behavior, habitat use, and foodchoice. (Since the 1950s, the term "character displace-ment" has broadened to include character diver-

gence, convergence, and parallel characteristic shifts

in zones of sympatry [cf. Taper and Case 1992].)

In case 3, after shorter periods of geographic isola-tion that result in incomplete reproductive isolation,hybridization may occur between the two incipientspecies during secondary contact. One possibility incase 3, which Figure 5-4 illustrates, is that hybrids haveequal or greater fitness than either of the parental pop-ulations. The result is complete introgression, or theincorporation of genes of one species into the gene poolof another. Any differences between the populationsthat evolved during geographic isolation would beswamped, and something akin to the old parental spe-cies would be reinstated (Figure 5-4). Another possi-bility is that hybrids may have low fitness, and thusintrogression does not occur. Theoretically, natural se-lection might then improve the reproductive isolatingmechanisms so that hybridization would diminish, al-though the reality of this process, called the Wallace

effect, has been questioned (e.g., Butlin 1989).Thus, the model of geographic speciation ac-

counts successfully for many characteristics observedwhen species coexist, from lack of interbreeding andthe presence of distinct ecological niches all the way

to considerable introgression and strongly overlap-

ping ecological niches.

< ,

00 090

R.I.M. completeE.D. incomplete

Evolved divergencein ecological characters,possibly involvingcharacter displacement

00'

R.I.M. incompleteE.D. incomplete or complete

a) Introgession iscomplete

b) Wallace effect.Ecological divergencemay continue

ParentalPopulation

S .C.

S.0

CO 1<

<I c0

So THE ORIGIN OF NEW SPECIES

FIGURE 5- 4

Three possible cases during geographic isolation. Case 1: After prolonged geographic isolation, reproductive isolating

mechanisms (RIMs) are complete and ecological divergence (ED) is complete. Case 2: RIMs are complete but ED is

incomplete. Case 3: RIMs are incomplete. SC indicates time of secondary contact between diverging populations. Timepasses from top to bottom in each case, and the parental species occupies niche A. Case 1 in the left-hand populationillustrates ecological divergence into niche B.

CASE 1 CASE 2 CASE 3

Secondarycontact

Evolvedresponse

R.I.M. completeE.D. complete

No evolutionarychange aftersecondary contact

5.2 SYMPATRIC SPECIATION

Mayr (1963) pointed out that the concept of speciationwithout geographic isolation was much older than hisgeographic speciation model, as Darwin's eventualspeciation model suggests. Mayr gave a brief histori-

cal account of other concepts and the negative criti-cism they had received. One concept seemed moreviable than most others. It concerns the speciationprocess in rather specialized groups of insects andthe evolution of new species within one locality. This

sympatric speciation model can be defined as "theestablishment of new populations of a species in dif-ferent ecological niches within the normal cruising

range of the individuals of the parental population"(Mayr 1963, 449) accompanied by rapid developmentof reproductive isolation between members of thepopulations in different niches.

Two points relating to specialized, or monopha-gous, insect groups are worth noting. First, plant-

feeding monophagous insects frequently belong tolarge genera. Thus, their apparently rapid speciationis unlikely to have occurred during prolonged geo-graphic isolation. For example, Ross (1962) reported

that the Erythroneura group of leafhoppers (Cicadelli-

dae) includes approximately 500 species, and specia-tion has involved about 150 shifts from utilization

of one host plant species to another. Such extensive

Specialists on host plantB are reproductivelyisolated from specialistson plant species A, butare now sympatric

Reinvasion intoparental rangeHerbivore population

specializes on host plantspecies B

Main range of herbivorespecies with species Aas principal host

Herbivore populationbecomes isolatedwhere host B is presentbut A is absent

Peripheral— isolate on1 host species A

SYMPATRIC SPECIATION

speciation is likely where many new niches exist,and with high food specificity, ecological barriers tointrogression are likely to be very effective. Second,it seems unlikely that a geographically isolated mo-nophagous species would repeatedly split into twospecies, each with a different host food plant. In manycases we should expect retention of the ancestral foodplant by both species, but this is not common.

An Allopatric Alternative

In defense of the allopatric speciation model, Mayr(1963) pointed out that peripheral populations arelikely to be exceedingly important in the speciationprocess. We can call this view the marginal-popula-tion allopatric speciation model (Figure 5-5). In amarginal population, a subsidiary host species maybecome the primary host because it is more abundantthan the major host or is favored in other ways. (Thisscenario assumes that the so-called specialized insectspecies is not strictly monophagous but utilizes morethan one host plant species, which is commonly thecase.) Such a shift in hosts results in rapid adaptationto the new conditions and rapid genetic change inthe small isolated population. At the same time, re-productive isolation from the parental populationmay develop. On secondary contact, the new species

on its new host plant remains sympatric with the

parent species, and the history and mechanism of thespeciation process are obliterated. Thus, the mar-ginal-population allopatric model is a special case ofthe geographic speciation model.

Mayr (1963) also noted problems with applyingthe geographic speciation model to animal parasites.For example, a sympatric model for speciation of thehuman louse into the body louse and the head louseneed not be complex (cf. Price 1980). Mayr, however,argues that the current species probably divergedallopatrically, the body louse evolving on the bodiesof heavily clothed people such as Eskimos and thehead louse on people clothed only in breechcloutsand grass skirts or less.

Any discussion on speciation processes will failto prove either the allopatric or the sympatric modelbeyond a doubt unless the actual process of speciationcan be studied in ecological time, over a few decadesor years. This kind of study, of course, is difficult tocarry out, and it has seemed that speciation is sucha gradual and rare event—relative to the 130 or soyears since Darwin—that the issues will remain tanta-lizing but unresolved. Yet increasing evidence forcesus to consider sympatric speciation as a reality. Be-cause this kind of speciation can occur rapidly, wemay indeed observe the process in ecological time.

FIGURE 5 - 5

Mayr's (1963) marginal-population allopatric speciation model.


Others in addition to Mayr (e.g., Futuyma andMayer 1980) have criticized the evidence for sympat-ric speciation and other modes of nonallopatric speci-ation. We discuss these modes here because they offerexplanations of how speciation can result in veryclosely related sibling species, an event that wouldnot be expected to result from allopatric speciation(cf. Bush and Howard 1986, Tauber and Tauber 1989,Cockburn 1991).

Speciation in Rhagoletis Fruit Flies

In the light of tantalizing debates and the skepticismof entomologists and botanists about the universalityof the geographic speciation model, Guy Bush in the1960s undertook a study of host-specific true fruitflies in the genus Rhagoletis and family Tephritidae.He knew that several species in this genus had re-cently established new host races on introduced com-mercial fruits. The apple maggot, R. pomonella, hadmoved from its native host, hawthorn (Crataegus sp.),to introduced apples in the Hudson River Valley,New York, in 1864. Then it had shifted from applesto cherries in Door County, Wisconsin, as recentlyas 1960 (Bush 1974, 1975a). Another host race of R.pomonella occurs on wild and cultivated plums, butless is known of its biology. In western NorthAmerica, the western cherry fruit fly, R. indifferens,in Oregon, Washington, and British Columbia, shiftedfrom the native bitter cherry, Prunus emarginata, tothe domesticated cherries, P. avium and P. cerasus, alittle before 1913, only 89 years after cherries wereintroduced into the region. Where bitter cherry anddomestic cherries grow close together in California,R. indifferens occasionally attacks domestic cherriesas if establishing a new host race. However, the Cali-fornia Department of Agriculture quickly expungesthese populations.

It became evident that much information—be-havioral, ecological, and genetic—was needed to es-tablish the likelihood of sympatric speciation. Bushand his coworkers have provided such information.Figure 5-6 summarizes much of the behavior of Rhago-letis fruit flies in relation to host selection and matingbehaviors. Note that a female fly finally accepts a fruitbased on a chemical cue, which must be received by herchemoreceptor. A mutation in the fly could thereforeaffect the amino acid composition of a receptor proteinand radically change the fly's response to a specifichost-plant chemical. Also, courtship and mating occuron the host fruit, and fruits are essential to fly repro-duction. Thus, host shifts often result in fly popula-tions that are allochronic, or chronologically out ofphase. Once a host shift has occurred, allochronic

isolation can become a factor in limiting gene flow.

For example, the two emergence patterns of R. porno-nella in Door County, Wisconsin, are so different thatthe hawthorn and cherry races of flies show practi-cally no overlap (Figure 5-7). Emergence time is undergenetic control (Smith 1988). T. K. Wood and his asso-ciates have explored in detail the processes involvedin allochronic isolation in the herbivorous membracidtreehopper genus Enchenopa (Wood and Keese 1990,Wood et al. 1990, and references therein).

The shift of R. indifferens from native cherry tothe introduced Prunus avium in California involvesboth temporal and spatial factors. On Mount Shasta,for example, the two tree species overlap in a zonefrom 3,500 to 5,000 feet above sea level, but theirfruiting times are so different that there is only a 2-week period in July and August in which R. indifferenspopulations overlap (Figure 5-8).

At least two genes must be involved in a hostshift by an insect: one for host recognition and selec-tion and another for larval survival once the egg isdeposited. Huettel and Bush (1971) have indeedshown, in a related species of tephritid, that a singlemajor locus controls host selection. In addition,Hatchett and Gallun (1970) found that a gene-for-gene relationship exists between host resistance andparasite survival in the hessian fly, which attackswheat. Genes for plant resistance to the fly arematched by genes for virulence in the fly population,as described in Section 19.2. Apparently, such rela-tionships between plant pathogens and their hostsare common (Day 1974). Thus, a shift from one hostto another of similar chemistry may involve only twogenes, and, because of allochronic isolation and mat-ing on the fruit, a shift causes ecological isolation,strongly reduced gene flow between host races, andessentially instantaneous and sympatric speciation.Isolation of races is followed by changes in other locito adapt the new race to new conditions. Thus, a factorsuch as larval survival may become polygenicallycontrolled once evolution of the new race has pro-

gressed.Bush (1975a, 199) stated,

It is now quite clear that host races of phytophagousparasitic insects have evolved sympatrically. Further-more, these host races are undoubtedly the progeni-tors of the many reproductively isolated sibling spe-cies so frequently found coexisting sympatrically on

different host plants.

Even Mayr in 1970 accepted the real possibility of

incipient sympatric speciation in Rhagoletis.Bush (1975a) observed that sympatric speciation

probably has been common in monophagous andstenophagous parasitic insects, but certain qualitiesof species predispose them to host shifting. (1) Mating

9Flies to tree(visual and

olfactory cues)

c/Looks tor andflies to fruit(visual cues)

Ovipositiona'Attempts to copulate

with any "Rhagoletispptb-rned" fly (cr or 9 )

9 Checks internalcondition of fruit

(chemotactile nv, ․)

Reconnoitersfruit surface (physical

and chemical cues)

Waits on fruit forapproaching femalesemits "aphrodisiac"

c7 Checks for 9 markingpheromone(s)

(chemotactile cue)

SYMPATRIC SPECIATION 83

FIGURE 5-6

The behaviors of fruit flies, Rhagoletis pomonella, involved with host choice and acceptance. Male fruit flies are symbolized

by d and female flies by 9.

I Mating on fruit

ctI 9Arrives on leaf

tactile cueschemical and physical)

9 Drags ovipositor overfruit, lays down

marking pheromone(s)

occurs on or near the host plant. (2) The adult femaleselects the host; the larvae have no choice. (3) Theparasitic species is specialized, attacking groups of

closely related host plant species. (4) Host selectionand larval survival are under genetic control; other

factors reinforce reproductive isolation and acceleratethe speciation process. (5) As Smith (1988) demon-strated, univoltinism (a single generation per year)with genetic control of emergence time increasesthe probability of allochronic isolation. (6) If genetic

84

THE ORIGIN OF NEW SPECIES

FIGURE 5- 7

Availability of host plant fruits (top) and approximate emergence times of apple maggot adults (below) from differentsympatric hosts in Door County, Wisconsin. (Based on Bush 1975a)

Sour cherry

Apple

Hawthorn

7

7 — '.„.. ...---,/ \ I

./ \ / .

/ Apple \ I Hawthorn\I .

/I\/ \

/ \

Sourcherry

JUNE JULY AUGUST SEPTEMBER OCTOBER

control of host-plant induction or conditioning is pre-sent, where an adult insect is induced or conditionedto return to the same host from which it emerged,fidelity to the specific ecological niche increases.

The Genetic Model

Bush (1974, 1975a) based his proposal for a general-ized genetic model of sympatric speciation on theshift of R. pomonella from apples to cherries in DoorCounty, Wisconsin. For a host shift to occur, muta-tions must produce new host selection (H) and sur-vival (S) alleles in a race—e.g., the apple race of flies—that is homozygous at those loci: HIHISISI. H2 and S2can then symbolize the mutations, which enable thebearer to select and survive on a new host—such ascherry. The new alleles might remain in the popula-

tion for a variety of reasons, but certain combinations

would be lethal in the older race. For instance, all

FI,H, and H,H2 would probably oviposit on apple,

but any S2S2 larvae would die on apple. H 2H2 femaleswould oviposit on cherries, but their SI S, progenywould die (Figure 5-9). Heterozygous H,H2 individu-als could oviposit on both plant species. However,Huettel and Bush (1971) found that individuals wereconditioned by larval food and tended to oviposit onthe species from which they had emerged. This factortogether with allochronic isolation would inhibit ran-dom mating, and H,H2 individuals would be mostlikely to remain on apple, where coadapted inductiongenes would operate effectively. Only few H 2H2 flies

that emerged early in the season would mate andoviposit on cherry (cf. Figure 5-7).

Shifts in emergence time of the new race wouldincrease the separation of races in time. An occurrenceof spread to other areas might cause divergence from

a purely sympatric distribution of the new species.

Investigators have documented genetically distinct

host races of R. pomonella on sympatric hawthorn and

apple (Feder et al. 1988, McPheron et al. 1988, Smith

50

Uz(-) 25

0

Zone of overlaySpatialwindow

50%

Temporalwindow

l'11(111.<:

(introduced)

15,000

13,000

11,000

9,000

7,000

5,000

3,000

1,000

0

Fruit availabilityM ! J J I AiS10 1

50%

50% r

- ---------Prunus emarina

PARAPATRIC SPECIATION 85

FIGURE 5 -8

Temporal and spatial availability of the fruits of Prunus emerginata and P. aviurn. Only a very small window, representedby a darkly shaded block, is available for a shift from the original host to the introduced P. avium based on synchronicand sympatric distributions. Fruiting times range from May (M) to October (0). (After Bush 1975a)

1988). The most likely mechanisms maintaining sig-nificant genetic differences are genetic control of allo-chronic emergence times and at least partial premat-ing reproductive isolation because of host-plantfidelity. Diehl and Bush (1989) provide the theoreticalbasis for this divergence: habitat preferences of indi-viduals and mating within habitats (see also Bush1992, 1994).

No doubt, endless natural "experiments" in hostrace and species formation have taken place in wildpopulations, and the successes must constitute a verysmall fraction of the total. However, when one consid-ers the vast number of highly specialized parasiticorganisms on this earth, many of which are liable tosympatric divergence of populations (Price 1980), wemust wonder about the extent to which sympatricspeciation has been more common than allopatricspeciation, especially when we include plant specia-

tion, discussed later in this chapter (see also Bushand Howard 1986). As Barton et al. (1988, 14) stated,

The new acceptability of the view that speciation doesnot always demand a period of geographical isolation

may help to explain the origins of the tens of millionsof species estimated to be alive today. Perhaps thepatron saint of evolution is, after all, sympatric.

5.3 PARAPATRIC SPECIATION

White (1968, 1978) advanced a third mode of specia-tion that he called stasipatric speciation; Bush (1975b)called it parapatric speciation. This model definesthe mechanism that accounts for a pattern of clo_selyrelated species contiguously distributed in space withnarrow zones of overlap (Figure 5-10). Basically, aceitain set of ecorogical 'factors may limit the rangeof a species. A new chromosomal rearrangement orgene combination may enable a few members to colo-nize the unexploited habitat. Heterozygotes betweenthe parental and progeny types are well adapted forneither the original environment nor the colonists'conditions and are rapidly selected out of the popula-tion. Premating isolating mechanisms evolve rapidly,and the new species occupies a range contiguous with

that of the parental species. The new species is likely

Range ofH 2 mutant _

//

IT3 11-1, S,S,H,H,}

H2H2•

Range where H2 andS, mutants overlap

CHERRY

CHERRY

APPLE(H,H,S,S,)

Sympatric shift tocherry by carriers ofH2 and S, mutations

APPLE(H,H,S,S,)

',..Range of

I' — S, mutant

•

(H2H2 die• , on cherry),

(HAHPPsLsEi )

CHERRY

DISTRIBUTION OF APPLE HOST PLANTS

AND CHERRY

H = Host Selection GeneS = Survival Gene— — Distribution at H2 allele— Distribution at S2 allele

Range expansion inparapatric zone

)ERIVEI

PA KINTA ISI'ECIF

-!•••-,1'.'


FIGURE 5 - 9

Bush's (1975a) genetic model ofsympatric speciation, based on ashift by apple maggots in DoorCounty, Wisconsin, from applehosts to cherry hosts. H is thehost selection gene (H, for apple,H2 for cherry), and S is the sur-vival gene (S, for apple, 5 2 forcherry). (After Bush 1975a)

FIGURE 5-10

A general view of parapatric speciation.

Narrow hybrid zone with hybridspoorly adapted to parental andderived species ranges

Limit of parental

Carrier of newspecies A range chromosomal mutation

colonizes new range

to be largely homozygous, with little genic differencesfrom the parental species but with chromosomal re-arrangements that alter major regulatory pathwaysin development.

Speciation in Morabine Grasshoppers

One example of this stasipatric or parapatric modelconcerns the Australian morabine grasshoppers inthe genus Vandiemenella. (The name "morabine"comes from another genus, Morabo.) These are allwingless grasshoppers with low vagility; conse-quently, they move little over a lifetime. They feedmainly on shrubs in the family Asteraceae. All thespecies have parapatric distributions. No two species,or even races within species, are sympatric over muchof their range (Figure 5-11). Zones of hybridizationare only 200 to 300 meters wide in several cases.

About 240 species of morabine grasshoppers oc-cur in Australia. Because most of the species recog-

nized by Key (1974) and White (1968, 1974, 1978) have

not been formally described and assigned specificnames, they are designated P for provisional species.

XP45b (XO)

P24 (XY)(translocation race)

XiYi

I I " " 11 KangarooIsland

P50 X

IIIu1 species

boundary

Adelai e

viatica

MaleVandiemenellapichirichi

0 100

km

race boundary

XV. pichirichi 1111'1'1'1

PARAPATRIC SPECIATIONT 87

FIGURE 5 - 1 1

The distribution of species and racesof Vandiemenella in southern Austra-lia. A male V. pichirichi is illustratedin the left bottom corner. The haploidkaryotype for each provisional specieslisted in Table 5-2 is shown, with thesex chromosome marked with an X orY or both. (Based on White 1978)

In southern Australia they are numbered P24, P25,P45b, and so on (Table 5-2). The distributions of spe-cies and races are almost mutually exclusive (Figure5-11). Each race has a unique karyotype, meaning aunique set of chromosomal morphologies. Figure 5-11 shows the haploid karyotype, its distribution ofmetacentric, acrocentric, and telocentric chromo-somes demonstrating considerable variation in chro-

TABLE 5 - 2 Species and races of some Vandie-menella wingless grasshoppers

Species Races

V. viatica

19-chromosome, 17-chromosomeP24'

XO, XY, translocation

P25

X0, XYP45b

XO, XY

P45c

No races

P50

No racesV. pichirichi = 3 karyotypes

P26/142

See Figure 5-11.

I P signifies an unnamed provisional species.

mosome structure. A metacentric chromosome hasthe centromere in its middle. An acrocentric chromo-some has one arm off the centromere. A telocentricchromosome has two arms extending asymmetricallyoff the centromere. For example, the karyotype ofVandiemenella viatica, race 19, has 9 autosomes in thehaploid state and 18 in the diploid state plus one sexchromosome, the X, adding up to 19 chromosomes.No Y chromosomes exist in this race, and so the raceis designed an XO type of race (cf. Table 5-2). Thenine autosomes comprise one telocentric and eightacrocentric chromosomes.

On Kangaroo Island, the distributions of Vandie-menella species seem very strange until one realizesthat, during the Pleistocene Epoch, solid land filledwhat are now the channels and bays between themainland and the island. This pattern suggests thatthe parapatric distributions have been remarkablystable over the past 10,000 years.

Given the lack of geographic barriers betweenspecies—indeed, the lack of any distinct factor lim-iting distribution—and the very local distribution ofspecies with almost exclusive distributions of specieswithin genera, neither the allopatric nor the sympatricmodel accounts adequately for the distributions of

Vandiemenella. The parapatric model accounts for the

pattern, but more research is needed on what defines

the limits of species ranges.


5 .4 SPECIATION MECHANISMS AND THE

KINDS OF SPECIES INVOLVED

The allopatric, sympatric, and parapatric mechanismsof speciation are so different mechanistically and inthe species distribution patterns they are likely tocause that we should expect rather different kinds oforganisms to engage in each. Bush (1975b) reviewedthe general attributes of the species that are likely tobe generated by each mode of speciation.

Allopatric Speciation

Allopatric speciation should be subdivided into theclassic model, involving speciation by subdivisionof large populations, and speciation by the foundereffect, in which a very small subsample of a popula-tion founds a new population in a geographicallyisolated area. In the classic model, an extrinsic barrierbetween two or more relatively large populations in-terrupts gene flow. In large populations evolutionis likely to be slow, as many genetic changes mustaccumulate during geographic isolation (see Chapter15). Thus, speciation is a long-term, gradual process.Relatively large animal species with high vagility areprobably restricted to this mode of speciation. Forexample, the vertebrate carnivores, many birds, cer-tain reptiles and amphibians, and most marine, lacus-trine and riverine fish probably speciate slowly ingeographic isolation.

Chromosomal evolution, which is a good indica-tor of rapid speciation (see Chapter 14), is minimalin animals with extensive ranges, such as the cats anddogs. Cats (Fells) in the Northern Hemisphere all havea chromosome number 2n = 38. All dogs (Canis) have2n = 78, whereas animals such as foxes (Vulpes andother genera) with smaller home ranges show consid-erable chromosome evolution (2n = 38, 40, 64, and78 in four species). As movement becomes more local,the chance that a chromosome rearrangement willbecome fixed increases, and the second type of allo-patric speciation becomes more likely.

The Founder Effect

Speciation by the founder effect may be more rapidand possibly more frequent than speciation by subdi-vision. A novel environment disrupts coadapted genecomplexes in the parental population, and rapid evo-lution is likely to build to a new adaptive peak, re-sulting in a speciation event (Carson and Templeton1984). Mutations in an almost empty and perhapsnovel environment are more likely to be adaptive and

to generate novel evolutionary steps than those in

the old environment. Advantageous changes in thegene pool become fixed much more rapidly in a small

population than in the original, larger one (see Chap-ter 15). Major adaptive chromosome rearrangementsmay readily become fixed in the homozygous con-dition.

Organisms with moderately low vagility are sus-ceptible to speciation by the founder effect, althoughit can also be effective in some very vagile animals,such as birds colonizing oceanic islands (see Chapter8). Even subsocial or social mammals, such as pri-mates and ungulates that typically form small, cohes-ive bands, can exist in a sufficiently fractionated pop-ulation structure so that founder effects becomeimportant. Cave-dwelling animals with homing be-havior, such as bats, are liable to found new coloniesdistant from parental groups. Carson (1973, 1975)suggested that speciation of some Hawaiian Dro-sophilidae is most likely to be by the founder effect,and many other insects must surely be equallylikely candidates.

The opportunities for chromosomal evolutionduring this mode of speciation are great. Such re-arrangements as fusions, fissions, whole-arm translo-cations, and pericentric inversions may cause rapidreorganization of regulatory mechanisms and thusdrastic changes in the developmental process withouta genic change (see also Chapters 12 and 14). Suchradical changes may permit the exploitation of a setof ecological conditions or resources very differentfrom conditions in the parental environment. Majorshifts in the adaptive mode may occur during specia-tion by the founder effect.

Parapatric Speciation

Parapatric speciation differs from speciation by thefounder effect in several important ways (Bush1975b). No spatial isolation occurs. The involved or-ganisms are exceptionally sedentary. Selection acti-vates reproductive isolating mechanisms at the sametime as individuals with new, highly adaptive genecombinations move into and exploit a new habitat.The organisms are sedentary enough that 100 to 1,000meters may be the limit of movement of an individualor its progeny. Plants, terrestrial snails, pocket go-phers and other fossorial rodents, morabine grass-hoppers, small mammals such as Peromyscus species

and shrews, certain lizards, mole crickets, stick in-sects, and many other flightless insects are likely to

speciate parapatrically.

Sympatric Speciation

Sympa tric speciation differs from parapatric specia-

tion in three ways. (1) Premating reproductive isola-tion develops before the shift to a new niche because

the new variant gene for habitat and host selection

Parental species

PRIMARY SPECIATION IN PLANTS 89

arises within the parental population. (2) Chromo-somal rearrangements do not seem to be involved.(3) Speciation is likely to occur at the center of thespecies range instead of the periphery and thus canlead to the evolution of many sympatric siblingspecies.

Sympatric speciation is common among parasitesof plants and animals, which constitute a very spe-cies-rich and diverse group of taxa.

Additional Speciation Mechanisms

Other common and important modes of speciationinclude interspecific hybridization and polyploidy,which occur frequently in plants. Association withsymbionts commonly adds to the complexity of intra-specific interactions and the development of repro-ductive isolation, as is the case with symbiotic sterilityfactors in Drosophila (e.g. Ehrman 1983) and otherpathogens that induce hybrid inferiority betweenhost populations (Thompson 1989, 1994). Develop-mental pathways may change because of mutationsor environment so that the timing or rates of develop-mental events differ from those of the same eventsin the parental population. The general term for de-velopmental change that results in new morphology,life history, or behavior is heterochrony, meaningliterally "different time" (McKinney and McNamara1991). Clearly, the timing of events such as reproduc-tion of morphological changes may contribute to re-productive isolation, and speciation and higher levelsof macroevolution may result, as McKinney andMcNamara (1991) discuss. (See also Chapter 18,"Rates of Evolution," especially Sections 8.5through 8.7.)

Plants are so different from animals in many waysthat we should examine the extent to which the modesof speciation found in these two kingdoms are similaror disparate. Grant (1971) distinguished between pri-mary and secondary forms of speciation in plants. Byprimary speciation he meant the origination of newspecies from one common ancestral species, as wehave discussed in the three modes of speciation-

FIGURE 5-12

Primary speciation and one example of secondary specia-tion recognized by Grant (1971).

Species C

Polyploidy

Species A Species

Hybridization

Species A Species B

PRIMARY SPECIATION SECONDARY SPECIATION

allopatric, sympatric, and parapatric (Figure 5-12). Insecondary speciation, good biological species pro-duce new species rather rapidly or even instantane-ously—for example, by hybridization or polyploidyor by a shift from an outcrossing to an amictic systemof reproduction. Researchers have studied such sec-ondary speciation extensively in plants because it hasbeen so common. Their accumulated knowledgeabout plant speciation can enrich our appreciation ofthe speciation process as a whole.

5.5 PRIMARY SPECIATION IN PLANTS

Quantum Speciation

Grant classified primary speciation into geographic,quantum, and sympatric modes (Figure 5-13). Geo-graphic speciation is the classic model of speciation.Grant (1971, 114) defined quantum speciation "asthe budding off of a new and very different daughterspecies from a semi-isolated peripheral population ofthe ancestral species in a cross-fertilizing organism."

cal races

POLYMORPHISM

ALLOPATRIC SPECIATION

Geographical races

_ QUANTUM SPECIATION _ _ SPECIES

I

FIGURE 5-13

Modes of primary speciation as classified by

Grant (1971).

- — - — — -SYMPATRIC SPECIATION

• San Jose

NEVADA

Monterey

Gilia tenuiflora

Gilia austrooccidentalis

in relativelyarid habitats

Gilia jacens

MOJAVEDESERT

• Los Angeles

O San Diego


In contrast with geographic speciation, quantum spe-ciation is rapid and results in major shifts in pheno-type and genotype. Two mechanisms seem to playroles in quantum speciation.

In the first mechanism, a shift from an outcrossingmode of reproduction to an inbreeding mode seemsto be the major factor. A large central populationof individuals is adapted for outcrossing by havinggenes that can effectively interact in the genome withnew genes brought in during gene flow throughoutthe population. In this central population, inflow ofnew genes tends to swamp idiosyncratic variation.Once a few individuals become isolated in a periph-eral population, however, inbreeding becomes com-mon among them, with drastic genetic and pheno-typic effects. Most of the change to inbreeding hasstrong negative effects on fitness and may cause ex-tinction of the local population. However, in manyperipheral populations, radical changes in genotypeproduce a few individuals with highly adaptive traits.This outcome occurs especially in peripheral loca-

tions where the parental species has presumably beenan unsuccessful colonizer. Thus, the mechanism ofquantum speciation fits within the umbrella of allo-patric speciation by the founder effect.

Verne and Alva Grant's (Grant and Grant 1960)work on the genus Cilia in California (Figure 5-14)provides an example of the pattern expected fromthis first mode of quantum speciation, in which theshift to inbreeding is important. Gilia tenuiflora is awide-ranging, predominantly outcrossing species.Two other species, G. austrooccidentalis and G. jacens,have smaller, peripheral ranges and are autogamous,or self-fertilizing. Apparently, chromosomal changeshave not been important in speciation. The fact thatthe peripheral species live in more arid habitats thandoes G. tenniflora, the apparently parental species,suggests that a physiological shift in the adaptivemode of the species during speciation enabled a rangeextension beyond the parental species.

A second mechanism that Grant regards as quan-tum speciation fits into the parapatric model. A coin-

FIGURE 5-14

Distributions of the parental species Gilia tenuiflora and two derived species in California. Note the two derived speciesin more arid locations than the parental species. (Based on Grant 1971)

PRIMARY SPECIATION IN PLANTS

parison between Clarkia biloba and C. lingulata sug-gests C. biloba as the parental species and C. lingulataas the derived species (Table 5-3; see also Figure 4-7on left). A chromosomal sterility barrier, resultingapparently from an original reciprocal translocationinvolving two chromosomes, and a subsequent trans-location in the C. lingulata line reproductively isolatethe species. Thus, C. lingulata is an aneuploid of C.biloba, meaning that its chromosome number is not amultiple of the parental haploid number. Because C.biloba is self-compatible, speciation occurred throughchromosomal repatterning rather than through a shiftfrom outcrossing to inbreeding. Lewis (1966) calledthis form of speciation speciation by saltation.

In addition to the factors already mentioned ascontributing to the two mechanisms of quantum evo-lution, populations are likely to be very unstable. Atmarginal sites in severe conditions, mortality is highand population size fluctuates dramatically throughbottlenecks of low numbers and rapid increases dur-ing favorable years. Thus, as Carson (1975) envi-sioned, populations are likely to pass through flush-crash cycles, with a founder effect after each crashstimulating rapid genetic drift and evolutionarychange (Figure 5-15). (Lewis [1962] emphasized thisevolutionary pattern in the genus Clarkia, arguingthat catastrophic selection was an important ingredi-ent in the evolutionary divergence of species.) Geneticdrift is the alteration of gene frequencies throughsampling error (see Chapter 15 for more detail). Whenpopulations become very small, as after a populationcrash, the few remaining individuals usually containonly a small proportion of the genes that were present

in the largest population. Just as two crayons sampled

at random from a 20-color crayon box cannot repre-

sent the color diversity in the box, so a few individuals

cannot usually contain the genetic diversity of a largepopulation. Repeated episodes of the founder effectwith genetic drift result in change in gene frequen-cies—or evolutionary change—even in the absenceof natural selection.

The Wallace Effect

Another potential factor in rapid divergence of popu-lations during primary speciation is the Wallace ef-fect. In his book Darwinism: An Exposition of the Theory

of Natural Selection, Wallace (1889) stated that crossingbetween populations in the same area or adjacent areasthat resulted in inferior-quality individuals would se-lect directly for reproductive isolating mechanisms.As Grant (1971, 136-137) put it, "Those individuals intwo sympatric populations which produce inviable orsterile hybrids will contribute fewer offspring to futuregenerations than sister individuals in the same paren-tal populations which do not hybridize." Therefore,natural selection reinforces reproductive isolatingmechanisms, and speciation by reinforcement occurs(Butlin 1989). Grant (1971) argued that the Wallace ef-fect is most likely to be important in annual plants,since reduced seed production is much more cata-strophic in them than in a long-lived species withmany years for seed production and many chances forcrossing with compatible genotypes.

A nice example of the Wallace effect comes fromGrant's (1966) studies on Gilia. The leafy-stemmed

TABLE 5 - 3 Characteristics of Clarkin biloba and -C. lingulata discerned. by Lewis

and Roberts (1956)

Characteristic Clarkia biloba Clarkia lingulata

Relationship

Geographical range

Presumed parental species

Large area of northernSierra Nevada and adjacent

California

Derived species

Very small area in oneriver canyon on south-

ern periphery of

C. biloba

n = 9

Self-compatible

Drier

Petals tongue-shaped

Chromosome number n = 8

Pollination Self-compatible

Habitat Moist, closer to ancestral

habitats for Clarkin

Morphology Petals heart-shaped

See Figure 4-7.

Flush

Crash

Catastrophic Selection

Population ReducedTo Few Individuals

Founder EffectGenetic Drift

Time

9 2THE ORIGIN OF NEW SPECIES

FIGURE 5-15

The flush—crash cycle with cat-astrophic selection and foundereffects occurring at frequent in-tervals.

Gilias can be divided into two major groups, and allare annuals. The distributions of the five foothill andvalley species in California overlap extensively, withindividuals of two or more species growing andflowering together in many localities. Very strongincompatibility barriers separate these species. Thedegree of separation can be measured conversely bythe number of hybrid seeds per flower produced as aresult of cross-pollination between species. In Grant'sstudy, seed output in this group ranged from 0 to 1.2seeds per flower and averaged 0.2 seeds per flower.Four maritime species are completely allopatric indistribution, occurring on coastal strands in Northand South America. The maritime species cross-fertil-ize easily, yielding 7.7 to 24.8 hybrid seeds per flower,with an average of 18.1 seeds per flower.

Grant noted that in other characteristics the in-land and maritime species had similar amounts ofdivergence between species, suggesting that the evo-lution of fertility barriers has accelerated greatlywhere species are sympatric. Butlin (1989) discussesproblems with interpretation of evidence for the Wal-lace effect.

Sympatric Speciation

Grant (1971) found no evidence for primary sympat-ric speciation. However, one case suggested that de-tailed studies might reveal more frequent instances ofthe process. Gottlieb (1973) argued for the sympatric

origin of a species derived from the parental stock of

an annual plant, Stephanomeria exigua. The parentalspecies was an obligate outcrosser with eight chromo-

somes (n = 8). Reproductive isolation of the newspecies developed sympatrically when an individualcarried a new chromosomal arrangement such thathybrids had reduced pollen viability and seed setand natural selection favored the evolution of a self-pollinating breeding system.

5.6 SECONDARY SPECIATION IN PLANTS

Hybrid Speciation and Amphiploidy

The major forms of secondary speciation in plantsinvolve a sympatric origin of new species. Hybridspeciation, the origination of new species directlyfrom a natural hybrid (Grant 1971), is a secondarymode of speciation because no divergence from acommon parental stock occurs. One of the problemswith hybrid speciation is that two genomes must beintegrated. Commonly, this occurs because segrega-tion stops in the F, generation, although a great diver-sity of mechanisms is involved, resulting in manydifferent kinds of genetic systems. Some methods ofstabilization in hybrid reproduction include vege-tative reproduction, agamospermy (the production ofseeds without fertilization of the ovule), permanentodd polyploidy, and amphiploidy, or allopolyploidy.An amphiploid, or allopolyploid, is "a polyploidthat originated by the doubling of the chromosomes

of a zygote with two unlike chromosome sets, usually

owing to hybridization of two species" (Mayr 1963,663). Polyploidy is the "condition in which the num-

SECONDARY SPECIATION IN PLANTS 93

ber of chromosome sets in the nucleus is a multiple(greater than 2) of the haploid numbers" (Mayr1963, 671).

The most common barrier to hybridization in via-ble plants is that, during meiosis, chromosomes can-not pair up because they are not homologous (Figure5-16). Thus, hybrid species originate where this prob-lem is circumvented by various means. In vegetativereproduction and agamospermy, meiosis is com-pletely absent, and mitosis proceeds normally to pro-duce vegetative individuals or to mitotically produceviable seeds. These kinds of uniparental reproductionare likely to give rise to microspecies. The develop-ment of amphiploidy results in good biological spe-cies that maintain effective sexual reproduction (Fig-ure 5-16). Amphiploidy produces a second version ofeach chromosome, and so homologous chromosomesare available for normal pairing, and sexual reproduc-tion is reestablished.

Clearly, hybrid speciation can result in very dif-ferent kinds of species. Some examples follow. Com-mon hemp-nettle, Galeopsis tetrahit, is a naturally oc-curring biological species (2n = 32), and Waltzing(1932, 1938) reproduced this species by crossing twodiploid species of hemp-nettle (2n = 16), Galeopsispubescens and G. speciosa (Figure 5-17). Despite consid-erable sterility between the two diploid species,Muntzing obtained F I 's, one of which was a triploidindividual. He backcrossed it to the parent G. pubes-cens, which yielded a tetraploid G. tetrahit-like plant.

Karpechenko (1927) produced a totally artificialspecies by crossing the radish, Raphanus sativus, with

FIGURE 5-16

An example of a hybridization event that results in cellswith nonhomologous chromosomes, and the recovery ofsexual reproduction by polyploidy. The identical pairs ofletters for the parental species indicate homologous chro-mosomes.

AmphiploidAABB Sexual

homologous chromosomes reproduction

reestablished

BB

Sexualreproductionprevented

Parental species 1

Parenta species 2

homologous

homologouschromosomes AA chromosomes BB

cabbage, Brassica oleracea, to form a tetraploid, Rapha-nobrassica. Initially a diploid hybrid was formed fromthe parental genera. Then, this largely sterile hybridapparently gave rise to Raphanohrassica after the unionof two diploid gametes. Natural hybridization mayproduce a hybrid complex in which "the morphologi-cal discontinuities between the originally divergentancestral forms" are obscured, giving rise to a speciesgroup in which discrete species are hard to identify(Grant 1971, 296).

Autopolyploidy

Another form of polyploidy in addition to amphi-ploidy is autopolyploidy, in which the same chromo-some set is doubled, tripled, and so on, giving riseto new species. For example, native species of Dahliahave 2n = 32, whereas the cultivated Dahlia variabilis,grown for show in flower gardens, has 2n = 64 andis a tetraploid. Further multiples of chromosomenumber can occur, as in Triticum, the genus to whichcommon wheat belongs, where 2n = 14, 28, and 42,the last being a hexaploid (n = 7, 6 x n = 42). InChrysanthemum, which includes Shasta daisies andmarguerites, 2 X, 4 x, 6 x, 8 x, and 10 x levels ofploidy are represented. Odd numbers of chromosomesets can maintain themselves if asexual reproductionor some degenerate form of sexual reproduction oc-curs and meiosis is avoided. The Rosa canina groupof dog roses in Europe has diploids (2n = 14), tetra-ploids (2n = 28), pentaploids (2n = 35), and hexa-ploids (2n = 42). The Crepis occidentalis group amongthe plants commonly called hawk's-beards has 2 X,3 x, 4 x, 5 X, 7 x, and 8 x levels of ploidy.

Many other forms of polyploidy can occur, sincethe presence of at least a duplicate copy effectivelybuffers change in chromosome number. Hence, chro-mosomes can be lost or added by ones and twos aswell as in whole haploid sets. As a result, an almostinfinite variety of karyotypes is possible in plants,and polyploid complexes are relatively common.

Frequency of Polyploid Plant Species

Polyploidy is so frequent in plants that it is a majorform of speciation. Grant (1963) constructed a fre-quency distribution of chromosome numbers inflowering plants and developed the argument thatmany plant taxa have polyploidization events in theirphylogenetic histories, resulting in chromosomenumbers greater than n = 13. He found a modal

distribution around n = 7, 8, and 9 and suggestedthat the original basic number in the angiosperms lay

in this range. Then, chromosome numbers of n = 14

and above would all be polyploids.

AA

Polyploidy

HybridAB

•-hLologousnon-

chromosomes

PARENTALGENERATION

n =8

GAMETES In =8 2n=16 ln= 8

Large-floweredhemp nettleGaleopsis speciosaThe flowers (up to 30mm) vary in color, butalways have some yellowin them. The plant iswidespread on disturbedsoils (often peaty), and issimilar to common hempnettle, but larger. Theleaves are toothed. 1 m.July-Sept.

Galeopsis Galeopsispuhescens speciosa2n=16 2n=16

2n=nx4=Tetraploid

2n =32Galeopsistetrahit

BACK CROSSTO PARENT

2n =24 2n=rix3=Triploid

GAMETES ln =8 2n =24

Common hemp nettleGaleopsis tetrahitGrows frequently on marshy and alsodisturbed ground, a straggling,coarsely hairy annual, with white, pinkor purple flowers (15-20 mm). Thepetals are often spotted. The leaves aretoothed. 1 m. July-Sept.

FIRST FILIALGENERATION


FIGURE 5-17

The possible events in the evolution of the amphiploid species Galeopsis tetrahit, studied by Muntzing (1932). Galeopsis

speciosa and G. tetrahit are illustrated.

Using Grant's estimate, Goldblatt (1980) sug-gested that many families of monocotyledons com-prise entirely polyploid species—for example, theBromeliaceae (1,700 species), Agavaceae (300 spe-cies), and Lemnaceae (30 species). And some verylarge families have more than 50 percent of speciesthat are polyploids: Cyperaceae, 73% (4,000 species),Poaceae, 60% (10,000 species), and Orchidaceae, 94%(18,000 species). In all, we regard 55 percent of mono-cotyledonous species as polyploids on the basis ofchromosome numbers greater than n = 13. Lewis(1980) called this an estimate of paleopolyploids,meaning that polyploidization has occurred in thephylogenetic background of a species. An estimateof intrageneric polyploid species illuminates morerecent speciation events, because the polyploids havearisen since the genus evolved. For monocotyledons,the paleopolyploid estimate is therefore 55 percent

of species, and the intrageneric polyploid estimate is

36 percent of species (Goldblatt 1980). Both estimates

illustrate the impressive impact of polyploidy in plantspeciation. In the dicotyledons, similar proportionsof species are polyploid (Lewis 1980).

Polyploidy is certainly less common in animalsthan in plants; it occurs in only some 30 species ofreptiles and amphibians (Bogart 1980) and is rare ininsects (Lokki and Saura 1980). Schultz (1980) regardspolyploidy in fishes, however, as a mechanism thathas provided the extra genetic material for majoradaptive breakthroughs. Perhaps all the salmonidsand catostomids are tetraploids, for example, But itis not yet clear how important polyploidy is in ani-mals in terms of speciation and macroevolution.

5.7 AGAMIC SYSTEMS

In addition to hybridization and polyploidy, loss of

gamete fusion has occurred many times. Such agamic

systems of reproduction result in immediate repro-

AGAMIC SYSTEMS 95

ductive isolation from the parental stock, indepen-dent divergence of lineages, and the emergence ofnew species. Sympatric speciation is likely to occur.Because the nomenclature of agamic systems in plantshas developed independently from that in animals,some definitions are in order (see Chapter 14). Apo-mictic reproduction, or apomixis, is asexual repro-duction in plants in general, which may involve de-velopment of a new individual from an unfertilizedegg or from a somatic cell of the parent. Ago-mospermy is seed formation without fertilization bypollen. Vegetative reproduction is the production ofnew individuals from somatic cells, not cells relatedto gametes. A clone consists of all individuals derivedby uniparental reproduction from a single parentalindividual, the genet being the genotype of all indi-viduals and the ramet being one individual in theclone (Harper 1977). Uniparental reproduction is re-production involving one parent by vegetative meansor by a degenerate form of sexual reproduction suchas agamospermy. Pseudogamy is the necessity ofpollination for seed formation in an apomicticspecies. (These genetic systems are discussed morein Chapter 14.)

Agamospermy is very common in plants; per-haps the best-known cases are in the family Astera-ceae, in the genera Hieracium, the hawk weeds, and

Taraxacum, the dandelions. It is common to find apo-mictic and automictic (self-pollinating) or amphimic-tic (outcrossing) forms of the same morphospecies.Taraxacum officinale, the common dandelion, has pop-ulations of apomicts and automicts in the same gen-eral locality. Most agamospermous species are theresults of hybridization, and so speciation by hybridi-zation goes hand in hand with agamospermous spe-cies. Once each lineage is reproductively isolated, theopportunities for development of many microspeciesare extensive, since clones in the same locality candiverge through somatic mutations.

Overall, speciation mechanisms are very diversein detail, although in general they may be distributedamong the allopatric, sympatric, and parapatricforms of primary speciation and the hybridization,polyploidy, and agamic systems of secondary specia-tion (Table 5-4). Each form of speciation results in adiscrete genetic system. Chapter 14 will explore theconsequences of each of these genetic systems.

One of the great accomplishments of the theoryof evolution is that it describes mechanisms to ac-count for the great diversity of species on this planet.Otte and Endler (1989) and references therein discussrecent developments in the fascinating field of specia-tion. The question of how life originated is more dif-ficult, but the next chapter will attempt to answer it.

BLE 5 - 4 Summary of major modes of speciation in animals and plants

Characteristics

Allopatric Parapatric Sympatric

Subdivision Founder Effect Host Shifts Chromosomal

Population size Large Very small Very small Very small Very small

Major processes Small genic dif-ferences accu-

Genetic drift,shift in adaptive

Chromosal mu-tation

Gene mutationsin host selection

Chromosomalmutation hy-

mulate peak and survivalgenes

brids, polyploids

Speciation rate Slow Faster Very rapid Very rapid Very rapid

Speciation location Allopatry Allopatry Edge of parentalspecies

Center of paren-tal range

Anywhere in pa-rental range

Reproductive barrier Any Any Premating • Premating Postmating

Examples Dendroica war- Darwin's finches Morabine grass- Rhagoletis fruit Galeopsis (Muntz-

blers (Mengel (Grant 1986) hoppers (White flies (Bush 1975a) ing 1932, 1938)

1964) 1968, 1978)

Cilia tenuiflora

and relatives(Grant and Grant

Ciarkia biloba

(Lewis andRoberts 1956)

Stephanomeria

(Gottlieb 1973),hybrid species

1960) (Grant 1971)


SUMMARY

Darwin's model of speciation lacked as an ingredientgeographic barriers that isolate populations andcause them to evolve independently. The model ofgeographic, or allopatric, speciation emphasizes theimportance of geographic isolation during whichproperties of populations evolve, ultimately leadingto reproductive isolation and the evolution of newspecies. An illustration of this mechanism involvedDendroica warblers and glaciation acting as a geo-graphic barrier. When one or a few individuals founda new population in isolation, as happened in thecase of Darwin's finches, allopatric speciation by thefounder effect may occur.

Geographic isolation results from environmen-tal characteristics, such as mountain ranges, that pre-vent movement of individuals between populations.Reproductive isolation results from properties ofspecies that constrain gene flow between populations,preventing the production of fully fertile progeny,even when the populations occur in the same locality.After geographic isolation, secondary contact be-tween populations may indicate that the new speciesare fully developed in terms of both reproductiveisolating mechanisms and ecological divergenceinto different ecological niches. If reproductive isola-tion is incomplete, however, secondary contact mayresult in introgression and the blending of the diverg-ing populations.

Sympatric speciation may occur in one locality,especially if parasitic organisms, such as Rhagoletisfruit flies, shift to a new host. Allopatric speciation inmarginal populations may account for the patterns ofoverlap in the distributions of many closely relatedspecies, although it would not lead us to expect thecommon occurrence of sibling species. Host shiftingbetween closely related plant species may requireonly small differences involving the genes for hostselection and survival of larvae in the host. The timingof fruiting and the partial overlap of host speciesin space may result in narrow windows in spaceand time for host shifts to be made, even though

they occur within the cruising range of a parentalpopulation.

In parapatric speciation, a chromosomal muta-tion enables carriers to invade new territory beyondthe parental range, and adjacent distributions of pa-rental and derived species develop. Morabine grass-hoppers in Australia demonstrate patterns of distri-bution and karyology that are best accounted for bythis speciation model.

Each mode of speciation may apply to a particularset of species characteristics. For example, allopatricspeciation involving extensive geographic barriersprobably applies to highly mobile organisms such asbirds and large mammals; sympatric speciation islikely to be common among parasitic species; andhighly sedentary organisms have the potential forspeciating parapatrically.

Among plants, we distinguish between primaryspeciation, involving divergence from a commonstock, and secondary speciation, by hybridization orpolyploidy or both, involving coalescing of genomesor multiplication of genomes. In primary speciation,quantum speciation—rapid divergence from a pa-rental stock in a peripheral population—may includeelements of the founder effect and parapatric specia-tion. The Wallace effect may promote the rapid evo-lution of reproductive isolation.

Secondary speciation in plants results from hy-brid speciation usually coupled with amphiploidy, aform of polyploidy that stabilizes sexual reproductionby providing homologous sets of chromosomes. Auto-polyploids are also reproductively isolated from pa-rental diploid stock and can give rise to new species.Polyploidy has provided the basis for probably morethan half the angiosperm species when the estimateincludes paleopolyploids. Polyploidy, although amechanism of secondary speciation, is a sympatricmode. Agamic systems of reproduction automaticallycause reproductive isolation among lineages and thepotential for evolutionary divergence until stocks canbe recognized as separate species.

QUESTIONS FOR DISCUSSION

In your estimation, do methods exist for objectivelydefining or discovering the mode or modes of specia-tion involved with the development of a set of re-

lated species?

2. This chapter attempted a balanced view on alternative

modes of speciation. Has this been a truly balanced

view, or do you think that preconceived ideas and bi-

ases molded the development of this chapter?

3. Do you think that modes of speciation in plants andanimals are basically the same, and should they be

treated as such?

4. An enigmatic aspect of morabine grasshopper distribu-

tions is the lack of any clear physical feature at the edge

of a range that may be involved in limiting dispersal.

What research would you plan in order to understand

the reasons for parapatric distributions in these species?

REFERENCES 97

5. Do you agree that organisms that habitually live in oron other species, such as parasites, small herbivores,and mutualists, are likely to come under strong influ-ences from the host population during the speciationprocess?

6. Many evolutionary biologists have expressed concernabout the sympatric model of speciation and its realityin nature, because it is relatively difficult to understandhow reproductive isolation could develop withoutsome kind of physical isolation between divergent pop-ulations. What kinds of research would you proposeto clarify this debate?

7. Is it possible that new modes of speciation will be dis-covered when we better understand the nature of thefungal and bacterial species?

8. Do you agree that the kinds of species a taxonomiststudies are likely to influence views on the prevalenceof certain kinds of speciation? For example, would tax-onomists of birds, cloning plants, parasitic worms, andsoil-dwelling fungi all agree that allopatric speciationis the usual or inevitable mode of speciation?

9. What research would you recommend be carried outon a particular group of specialized insect herbivoresto determine whether its species emerged according tothe marginal-population allopatric speciation model oraccording to the sympatric speciation model?

io. How do you think species size plays a role in speciation,given that there are many more small species thanlarge species?

REFERENCESAllmon, W. D. 1992. A causal analysis of stages in allopatric

speciation. Oxford Surveys in Evol. Biol. 8:219-257.Barton, N. H., J. S. Jones, and J. Mallet. 1988. No barriers

to speciation. Nature 336:13-14.Bogart, J. P. 1980. Evolutionary significance of polyploidy

in amphibians and reptiles, pp. 341-378. In W. H. Lewis(ed.). Polyploidy: Biological Relevance. Plenum,New York.

Brown, W. L., and E. 0. Wilson. 1956. Character displace-ment. Systematic Zool. 5:49-64.

Bush, G. L. 1974. The mechanism of sympatric host race for-mation in the true fruit flies (Tephritidae), pp. 3-23.In M. J. D. White (ed.). Genetic Mechanisms of Specia-tion in Insects. Australia and New Zealand BookCo., Sydney. . 1975a. Sympatric speciation in phytophagous para-

sitic insects, pp. 187-206. In P. W. Price (ed.). Evolution-ary Strategies of Parasitic Insects and Mites. Plenum,New York. . 1975b. Modes of animal speciation. Ann. Rev. Ecol.

Syst. 6:339-364. . 1992. Host race formation and sympatric speciation

in Rhagoletis fruit flies (Diptera: Tephritidae). Psyche99:335-358. . 1994. Sympatric speciation in animals: New wine

in old bottles. Trends in Ecol. and Evol. 9:285-288.Bush, G. L., and D. J. Howard. 1986. Allopatric and non-

allopatric speciation: Assumptions and evidence, pp.411-438. In S. Karlin and E. Nevo (eds.). Evolutionary

Processes and Theory. Academic, New York.

Butlin, R. 1989. Reinforcement of premating isolation, pp.158-179. In D. Otte and J. A. Endler (eds.). Speciation

and Its Consequences. Sinauer, Sunderland, Mass.Carson, H. L. 1973. Reorganization of the gene pool during

speciation, pp. 274-280. In N. E. Moreton (ed.). Genetic

Structure of Populations. Population Genetics Mono-

graph 3. Univ. of Hawaii Press, Honolulu.

1975. The genetics of speciation at the diploid level.

Amer. Natur. 109:83-92.

Carson, H. L., and A. R. Templeton. 1984. Genetic revolu-tions in relation to speciation phenomena: The found-ing of new populations. Ann. Rev. Ecol. Syst. 15:97-131.

Cockburn, A. 1991. An Introduction to Evolutionary Ecol-ogy. Blackwell Scientific Publications, Oxford.

Darwin, C. 1859. The Origin of Species. 1st ed. Murray,London.

. 1872. The Origin of Species. 6th ed. Murray,

London.Day, P. R. 1974. Genetics of Host-Parasite Interaction.

W. H. Freeman, San Francisco.Diehl, S. R., and G. L. Bush. 1989. The role of habitat prefer-

ence in adaptation and speciation, pp. 345-365. In

D. Otte and J. Endler (eds.). Speciation and Its Conse-

quences. Sinauer, Sunderland, Mass.Eberhard, W. G. 1985. Sexual Selection and Animal Genita-

lia. Harvard Univ. Press, Cambridge, Mass.Ehrman, L. 1983. Endosymbiosis, pp. 128-136. In D. J. Futu-

yma and M. Slatkin (eds.). Coevolution. Sinauer, Sun-

derland, Mass.Feder, J. L., C. A. Chilcote, and G. L. Bush. 1988. Genetic

differentiation between sympatric host races of the ap-ple maggot fly Rhagoletis pomonella. Nature 336:61-64..

Futuyma, D. J. 1986. Evolutionary Biology. 2d ed. Sinauer,Sunderland, Mass.

Futuyma, D. J., and G. C. Mayer. 1980. Non-allopatric speci-ation in animals. Syst. Zool. 29:254-271.

Goldblatt, P. 1980. Polyploidy in angiosperms: Monocotyle-

dons, pp. 219-239. In W. H. Lewis (ed.). Polyploidy:

Biological Relevance. Plenum, New York.

Gottlieb, L. D. 1973. Genetic differentiation, sympatric spe-ciation, and the origin of a diploid species of Stephanom-

eria. Amer. J. Bot. 60:545-553.Grant, P. R. 1986. Ecology and Evolution of Darwin's

Finches. Princeton Univ. Press, Princeton, N.J.

Grant, V. 1963. The Origin of Adaptations. Columbia Univ.

Press, New York.

.1966. The selective origin of incompatibility barriers

in the plant genus Gilia. Amer. Natur. 100:99-118.


. 1971. Plant Speciation. Columbia Univ. Press,New York.

Grant, V., and A. Grant. 1960. Genetic and taxonomic spe-cies in Gilia, [Part] XI: Fertility relationships of thediploid cobwebby Gilias. Aliso 4:435-481.

Grant, V., and K. A. Grant. 1965. Flower Pollination in thePhlox Family. Columbia Univ. Press, New York.

Harper, J. L. 1977. Population Biology of Plants. Academic,New York.

Hatchett, J. H., and R. L. Gallun. 1970. Genetics of the abilityof the Hessian fly, Mayetiola destructor, to survive onwheat having different genes for resistance. Ann. Ento-mol. Soc. Amer. 63:1400-1407.

Hillis, D. M. 1988. Systematics of the Rana pipiens complex:Puzzle and paradigm. Ann. Rev. Ecol. Syst. 19:39-63.

Huettel, M. D., and G. L. Bush. 1971. The genetics of hostselection and its bearing on sympatric speciation inProcecidochares (Diptera: Tephritidae). Entomol. Exp.Appl. 15:465-480.

Karpechenko, G. D. 1927. Polyploid hybrids of Raphanussativus L. x Brassica oleracea L. Bull. Appl. Bot. Genet.Plant Breeding, Leningrad 17:305-310.

Key, K. H. L. 1974. Speciation in the Australian morabine

grasshoppers: Taxonomy and ecology, pp. 43-46. InM. J. D. White (ed.). Genetic Mechanisms of Speciationin Insects. D. Reidel, Dordrecht, The Netherlands.

Kraus, F., and J. W. Petranka. 1989. A new sibling speciesof Ambystoma from the Ohio River drainage. Copeia1989:94-110.

Lewis, W. H. 1962. Catastrophic selection as a factor inspeciation. Evolution 16:257-271. . 1966. Speciation in flowering plants. Science

152:167-172. .1980. Polyploidy in angiosperms: Dicotyledons, pp.

241-268. In W. H. Lewis (ed.). Polyploidy: BiologicalRelevance. Plenum, New York.

Lewis, W. H. (ed.) 1980. Polyploidy: Biological Relevance.Plenum, New York.

Lewis, W. H., and M. R. Roberts. 1956. The origin of Clarkialingulata. Evolution 10:126-138.

Lokki, J., and A. Saura. 1980. Polyploidy in insect evolution,pp. 277-312. In W. H. Lewis (ed.). Polyploidy: Biologi-cal Relevance. Plenum, New York.

Mayr, E. 1963. Animal Species and Evolution. Belknap Pressof Harvard Univ. Press, Cambridge, Mass. . 1970. Population, species and evolution. Belknap

Press of Harvard Univ. Press, Cambridge, Mass.McKinney, M. L., and K. J. McNamara. 1991. Heterochrony:

The Evolution of Ontogeny. Plenum, New York.McPheron, B. A., D. C. Smith, and S. H. Berlocher. 1988.

Genetic differences between host races of Rhagoletispomonella. Nature 336:64-66.

Mengel, R. M. 1964. The probable history of species forma-tion in some northern wood warblers (Parulidae). Liv-ing Bird 3:9-43.

Minkoff, E. C. 1983. Evolutionary Biology. Addison-Wesley, Reading, Mass.

Mantzing, A. 1932. Cytogenetic investigations on synthetic

Galeopsis tetrahit. Hereditas 16:105-154.

. 1938. Sterility and chromosome pairing in intraspe-cific Galeopsis hybrids. Hereditas 24:117-188.

Otte, D., and J. A. Endler (eds.). 1989. Speciation and ItsConsequences. Sinauer, Sunderland, Mass.

Paterson, H. E. H. 1985. The recognition concept of species,pp. 21-29. In E. S. Vrba (ed.). Species and Speciation.Transvaal Museum Monograph No. 4. Pretoria.

Price, P. W. 1980. Evolutionary Biology of Parasites. Prince-ton Univ. Press, Princeton, N.J.

Ridley, M. 1993. Evolution. Blackwell Scientific Publica-tions, Oxford.

Ross, H. H. 1962. A Synthesis of Evolutionary Theory. Pren-tice-Hall, Englewood Cliffs, N.J.

Schultz, R. J. 1980. Role of polyploidy in the evolution offishes, pp. 313-340. In W. H. Lewis (ed.). Polyploidy:Biological Relevance. Plenum, New York.

Smith, D. C. 1988. Heritable divergence of Rhagoletis porno-nella host races by seasonal asynchrony. Nature336:66-67.

Strickberger, M. W. 1990. Evolution. Jones and Bartlett,Boston.

Taper, M. L., and T. J. Case. 1992. Coevolution among com-

petitors. Oxford Surveys in Evol. Biol. 8:63-109.Tauber, C. A., and M. J. Tauber. 1989. Sympatric speciation

in insects: Perception and perspective, pp. 307-344.In D. Otte and J. A. Endler (eds.). Speciation and ItsConsequences. Sinauer, Sunderland, Mass.

Templeton, A. R. 1989. The meaning of species and specia-tion: A genetic perspective, pp. 3-27. In D. Otte andJ. A. Endler (eds.). Speciation and Its Consequences.Sinauer, Sunderland, Mass.

Thompson, J. N. 1989. Concepts of coevolution. Trends inEcol. and Evol. 4:179-183. . 1994. The Coevolutionary Process. Univ. Chicago

Press.Vrijenhoek, R. C. 1984. The evolution of clonal diversity in

Poeciliopsis, pp. 399-429. In B. J. Turner (ed.). Evolution-ary Genetics of Fishes. Plenum, New York.

Wallace, A. R. 1889. Darwinism: An Exposition of the The-ory of Natural Selection. Macmillan, London.

White, M. J. D. 1968. Models of speciation. Science

159:1065-1070. . 1974. Speciation in the Australian morabine grass-

hoppers-the cytogenetic evidence, pp. 57-68. InM. J. D. White (ed.). Genetic Mechanisms of Speciationin Insects. D. Reidel, Dordrecht, The Netherlands.

. 1978. Modes of Speciation. W. H. Freeman, SanFrancisco.

Wood, T. K., and M. C. Keese. 1990. Host-plant-inducedassortative mating in Enchenopa treehoppers. Evolution

44:619-628.Wood, T. K., K. L. Olmstead, and S. I. Guttman. 1990. Insect

phenology mediated by host-plant water relations.

Evolution 44:629-636.Woodruff, D. S. 1979. Postmating reproductive isolation

in Pseudophryne and the evolutionary significance of

hybrid zones. Science 203:561-563.

the origin of new species - wesleyan...

Documents