Alternative Hypotheses in Biogeography: Introduction and Synopsis

PDF document for free

1. PDF document for free

31494_722_2_349.pdf

AMER. ZOOL., 22:349-354 (1982)

Alternativ

e Hypothese s i n Biogeography :Introduction and Synopsis of the Symposium1 JOH N A . ENDLE R

University

of Utah, Salt Lake City, Utah 84112

SYNOPSIS

. Biogeograph y ha s ha d a lon g history , bu t onl y recentl y ha s i t starte d t o b

emore than a descriptive subject. Models have been developed which help to explain pat-terns of geographical distributions of animals and plants. Examples include dispersal,vicariance, and ecological determinism. The dispersal model proposes that present-daydistributions were caused by long distance dispersal among disjunct localities. The vicar-iance model proposes that the distributions resulted from the fragmentation of formerlycontinuous distributions with accompanying speciation. Ecological determinism proposesthat whatever the historical changes in distribution (dispersal or vicariance), present-dayecological factors are the major determinant of distributions. We too often see in theliterature large bodies of data analyzed with respect to only one of these competingmodels. I organized the symposium to bring together what I thought were particularlyinteresting examples of each approach. It is obvious from the papers that there are mul-tiple explanations for most species distributions, and that some effort should be made tointegrate them.

Biogeograph

y ha s ha d a lon g history

,but only recently has it started to be morethan a descriptive subject. Models havebeen developed which help to explain andsometimes predict patterns of geographi-cal distributions of animals and plants. Ex-amples include dispersal, vicariance, andecological determinism. Presumably as aresult of the newness of the quantitativetechniques and ideas in many of themodels, we too often see in the literaturelarge bodies of data analyzed with respectto only one of several competing hypoth-eses. Too often a model is chosen appar-ently because it is familiar, and explana-tions have often been those of plausibilityrather than a comparison of alternativehypotheses. Even those interested in test-ing hypotheses rarely consider alternativehypotheses or even testing other predic-tions of their favorite model if the first testis successful. Biogeography has becomedivided into "schools" which largely ignoreone-another. As a first step towards a newsynthesis, I organized this symposium onAlternative Hypotheses in Biogeography.

1 Fro m th e Symposiu m o n Alternative Hypotheses in

Biogeography

presente d a t th e Annua l Meetin g o f th

eAmerican Society of Zoologists and the Society of Sys-tematic Zoology, 27-30 December 1980, at Seattle,Washington.The contributions represent what I thinkare particularly interesting examples ofeach approach to biogeography.

Biogeograph

y attempt s t o explai n th

edistribution of species. There are threeclasses of phenomena which affect whichspecies and how many are found in a par-ticular area: ecology, dispersal ability, andpatterns of speciation. The species in aparticular place can be there because theirecology permits them to survive, becausethey dispersed there from elsewhere, orbecause they speciated there. These classesof phenomena correspond roughly to thethree main schools of biogeography: eco-logical determinism, dispersal, and vicari-ance biogeography.

Ecolog

y determine s whic h specie s ca

nactually survive in a particular place. Phys-iological limits, presence of competitors, ormerely microhabitat requirements can lim-it the number of species in a given locality(Andrewartha and Birch, 1954; Mac-Arthur, 1972; Krebs, 1972; May, 1976).These factors operate continuously, andmay wholly or partially obliterate historicalfactors. Ecological factors act on a shorttime scale, so their patterns can changewith geological and climatic changes. Eco-logical biogeography tends to assume thatall species can disperse to the location ofinterest, and that speciation does not oc-cur. This is simply a matter of the short

34

9Downloaded from https://academic.oup.com/icb/article/22/2/349/2015960 by guest on 16 August 2023

350JOHN A. ENDLER

tim e scal e o f interes t t o ecologists , a s com -pared to historical biogeographers (Udvar-dy, 1969).

Classica

l historica l biogeograph y as

-sumes that species originated in particularareas and dispersed from those centers oforigin. The distribution of species in a par-ticular place is dependent upon whichspecies happened to disperse there (Ek-man, 1953; Darlington, 1957; Simpson,1965; Briggs, 1974a, b; Pielou, 1979). Eventhough the probability of crossing a bar-rier may be very low, given enough timesome species will cross it. Dispersal bio-geography is primarily concerned withidentifying the centers of origin and is notconcerned with ecology except insofar asit limits the establishment of dispersing or-ganisms (for example, Carlquist, 1965). Itis also not concerned with the process ofspeciation; it assumes that speciation is arare event.

Islan d biogeograph y combine s ecologi

-cal and dispersal ideas, along with somestatistical notions. It seeks to predict thenumbers and kinds of species found in is-lands or isolated patches of habitats on thebasis of probabilities of immigration andextinction. These rates and their relation-ships to species numbers vary in predict-able ways with island size, distance fromthe source of dispersing species, and thedispersabilities of the species concerned(MacArthur and Wilson, 1967; Mac-Arthur, 1972; Diamond and May, 1976).Unlike classical historical biogeography, itis less concerned with unique events andmore concerned with the effects of manyuncommon events. In addition, unlike dis-persal biogeography, it yields specific test-able predictions.

Vicarianc

e biogeograph y ignore s ecolo

-gy, downgrades the importance of dis-persal, and concentrates on the effects ofspeciation on the distributions of organ-isms. It is therefore exclusively concernedwith the effects of history, and in particu-lar the effects of geological changes whichcause species to split, allowing the charac-ter distributions of the split populations todiverge. The basic idea is that: (1) A givenspecies spreads over a large continuousarea or range. (2) The range fragmentsowing to geological or climatic changes. Arange split is called a vicariant event(Udvardy, 1969). (3) The allopatric popu-lations diverge and speciate. (4) The newspecies may spread into each other'sranges. This model has specific predictionsabout the taxonomic relationships of thevicariant species: species resulting from re-cent vicariant events should be sisterspecies (closely related), and species re-sulting from older vicariant events shouldbe more distantly related. Relatedness isestimated by cladistic methods (Nelson,1979; Eldredge and Cracraft, 1980). Inaddition, since many different organismsliving in the same large area experiencethe same sequence of geological changesand vicariance events, there should be ahigh degree of concordance among theestimated phylogenies for each group (Ro-sen 1975, 1978; Platnick and Nelson,1978; Nelson and Platnick, 1978, 1980;Nelson and Rosen, 1981; Wiley, 1981).Ecology is ignored because taxonomiccharacters must be non-adaptive. Dispersalis minimized because most species are as-sumed to have originated in their presentarea, or dispersal took place before the vi-cariant events. This model has the advan-tage of specific testable predictions.

Th e paper s o f th e symposiu m d o no t

fiteasily into the three categories because Iasked the speakers to consider as many al-ternative hypotheses as possible withoutsetting up "straw men." They do, however,vary in their emphasis on ecology, dispers-al, or vicariance.

Stron g an d Re y giv e a goo d definitio n o

fecological biogeography: it is concernedwith how geographical range and local suc-cess is affected by predation, disease, par-asitism, and abiotic factors. Their paper isconcerned with the effects of temporalchanges and randomness in the immigra-tion and extinction curves of island bio-geography. They find that on a very shorttime scale the slopes of these curves maynot be significantly different from zero,but over a long time period, or for manyislands, the classical MacArthur-Wilsonpattern emerges. In addition, over inter-mediate (ecological) time periods theslopes of the immigration and extinctionDownloaded from https://academic.oup.com/icb/article/22/2/349/2015960 by guest on 16 August 2023

INTRODUCTION AND SYNOPSIS OF SYMPOSIUM351

curve s ma y change . Thi s ha s profoun d ef

-fects on the predicted equilibrium numberof species on an island or patch of habitat.The important point is that temporalchanges in the probabilities of immigrationand extinction can make both predictionsand testing of predictions difficult.

Livingstone

, Rowland , an d Baile y ex

-amined fishes of African rivers and foundthat most of the variation in numbers ofspecies could be explained on the basis ofthe total stream discharge. The annualflow through a river is correlated with ter-restrial productivity. This can affect thefish because much of the food of thesetropical rivers comes from the surround-ing land rather than within the streamsthemselves, and space is not limiting. Thisnot only explains many present-day distri-butions, but also some patterns in the past;periods of low discharge (such as a warmdry period) will therefore cause extinctionsof many species in a given drainage. Itwould be interesting to know if the patternapplies to other continents, tropical ortemperate. This is a good example of asimple ecological limitation which canchange in space and time, with testablepredictions.

Edmund

s presente d som e interestin

gdata on relationships, distributions, andecology of new world mayflies. Relation-ships were worked out by cladistic meth-ods. Once Laurasian or Gondwana assign-ments are made, it is possible to examinethe effects of the merger of the two faunaafter the formation of the Central Ameri-can land bridge. Many South Americangenera have penetrated far into NorthAmerica. It turns out that variation in life-history patterns largely explains the dis-persal pattern. Species with tropical af-finities are limited in their northwarddispersal by the length of their larvalperiod. Those with shorter larval periodscan reproduce at higher latitudes thanthose with longer periods; if a larval peri-od is too long the mayfly dies before it cantransform to an adult and reproduce.North American species which hatch inwarmer periods have southern affinitiesmore often than those which hatch in cold-er periods. Thus we have a direct ecolog-ical reason for an historical pattern of thedispersal of species, using cladistic meth-ods and a knowledge of the biology of theanimals. Edmunds also discusses vicariantpatterns among the southern continents.

Davi s examine d th e effect s o f dispersal

,vicariance, and ecology on the large fresh-water snail family Pomatiopsidae. He isable to separate out the effects of all threefactors on species distributions. By com-paring cladistic reconstruction to knowngeological events he was able to follow thedivergence and radiation of the family ona large scale. The breakup of Gondwana-land, passage on India into Asia, and sub-sequent diversification in the newly creat-ed river systems was also reflected in thephylogenetic reconstruction. Vicarianceevents were detectable as a result of thesplitting of Gondwanaland and also theHimalayan orogeny. Dispersal (recon-structed from relationships) was presum-ably responsible for spread into the riversystems (such as the Mekong) which re-sulted from the Himalayan orogeny. Fi-nally the presence of a large number ofnew ecological niches in the newly formedriver systems allowed rapid adaptive radia-tion without vicariance events (allopatricspeciation), merely because these snailshave very low mobility. Speciation andadaptive radiation is thought to have oc-curred by parapatric and even sympatricspeciation in these new rivers. The ecolog-ical factors are sufficiently important insnails that they cause strong convergencein gross morphology and shell morpholo-gy in unrelated species living in the samekinds of habitats. At the same time ecolog-ical factors cause great divergence inclosely related species living in very differ-ent, but adjacent habitats. Davis shows thatmuch more information can be obtainedby considering a variety of approaches,and considering the biology of the organ-isms, than by blindly following the predic-tions and paradigms of a single biogeo-graphic model.

Heye r an d Maxso n d o somethin g simila

rwith South American leptodactylid frogs.They worked out relationships by meansof microcomplement fixation instead ofcladistic techniques and found that this canDownloaded from https://academic.oup.com/icb/article/22/2/349/2015960 by guest on 16 August 2023

352JOHN A. ENDLER

hel p t o reconstruc t th e sequenc e o f vicar

-iant events. Unlike cladistic methods, thismethod estimates the amount of diver-gence as well as the branching sequence(for a discussion of methods, see Hull,1979). In many cases the immunologicalmethod yields similar results to morpho-logical methods, but they found some caseswhere there was a lack of congruence be-tween immunological distances and mor-phological distances among species. Thissuggests that morphology is only an ap-proximation to molecular divergence, orat least that the evolutionary rates of mor-phology and albumins are different. Theythen discuss the problems of using mor-phological and molecular data in biogeog-raphy. If different kinds of charactersevolve at different rates, then they willhave very different effects on attempts toestimate phylogenies. Their results alsoimply that varying rates can affect the cla-distic estimates of branching sequences. Ifsome characters evolve more slowly thanother characters then the slow characterswill not record as many vicariant events asthe rapidly evolving characters. If thecharacter change rates vary in time as wellas between characters, then the recordingof vicariant events will be very sporadic,and an analysis of the final result (the cur-rent species) is very likely to yield very in-accurate branching sequences. Heyer andMaxson also show how ecological con-straints affect species distributions, andtest many different hypotheses about lep-todactylid frog distributions.

Cracraf

t present s a ver y clea r exampl

eof vicariance biogeography of Australianbirds. More important, he shows how vi-cariance biogeography can help to distin-guish between the various modes of spe-ciation. This was also discussed by Wiley(1981). The most important point of thispaper is that the methods of cladistic re-construction and vicariance biogeographyhave never been applied to the study ofspeciation. This is surprising in that cla-distics attempts to estimate genealogicalrelationships, and speciation is the processproducing those relationships. Given thevarious modes of speciation, and the nec-essary assumptions of cladistic reconstruc-tion of phylogeny, Cracraft shows thatsome kinds of speciation yield unique pre-dictions about the relationships among thenewly split species. He then goes on to testthe predictions with Australian bird sys-tematics and biogeography. He does notcover all the modes of speciation, so not allof the conclusions are generalizable. Forexample he does not discuss the mainmode of parapatric speciation, which re-quires ecological differences among thenew sister species (Endler, 1977); insteadhe discusses a very special case of parapat-ric speciation which depends upon geneticdrift and no selection. This leads to hisconclusion that parapatric speciationshould always yield sister species. But thisis a very special case of parapatric specia-tion and a very unlikely one since there isnothing in the system to promote isolatingmechanisms. The regular form of para-patric speciation will rarely if ever result insister species since the system dependsupon very differing environments amongthe species in statu nascendi, and thereforethey would share very few derived char-acters. This reverses his conclusions aboutthe frequency of parapatric speciation inAustralian birds. But in spite of this, thepaper is one of the first attempts to applycladistic reconstruction of speciation eventsto the speciation process, and an excellentexample of the method of vicariance bio-geography.

Chernof

f present s a goo d tes t o f vicari

-ance biogeography by examining how wellmorphological characters work, when usedin a cladistic analysis, to reconstruct knownvicariant events which happened less than50 years ago in Menidia beryllina fish livingin the Mississippi river system. Althoughhe used characters typically used by fishtaxonomists, the recently separated popu-lations were not sister species in the anal-ysis. A multivariate analysis was performedon the morphological and environmentalfactors among all localities. Chernofffound significant canonical correlationsbetween groups of taxonomic charactersand groups of environmental factors, in-dicating that these characters may experi-ence natural selection. This yields the false"vicariance patterns." Environmental fac-Downloaded from https://academic.oup.com/icb/article/22/2/349/2015960 by guest on 16 August 2023

INTRODUCTION AND SYNOPSIS OF SYMPOSIUM353

tor s wer e mor e importan t i n determinin

gthe pattern of shared derived charactersthan history in this data set. It is thereforevery important to eliminate characters sub-ject to natural selection before proceedingwith the methods of vicariance biogeog-raphy. Chernoffs results cast doubt onmuch of work in fish systematics and bio-geography because his results are for char-acters typically used by fish systematists.

M y pape r i s concerne d wit h th e prob

-lems of confounding the effects of histor-ical and ecological factors in biogeography.Unfortunately both classes of factors canyield similar results in distributions and ingenealogies. Two examples are given (1)the Pleistocene forest refuge hypotheses,and (2) vicariance biogeography. ThePleistocene forest refuge hypothesis seemsto work well because only one of three pre-dictions has been tested (Prance, 1981);centers of diversity correspond amongmany different groups of animals andplants. An examination of the other pre-dictions makes the hypothesis seem doubt-ful, and the predictions of two alternativehypotheses - present-day ecology and dif-ferentiation of peripheral isolates - are up-held. The major prediction of vicariancebiogeography, that lineages experiencingthe same sequence of vicariant eventsshould yield highly concordant estimatedphylogenies, is rejected. Concordantcladograms can only result from commonpatterns of shared selection regimes, andthese do not necessarily reflect shared vi-cariant patterns. Sister species in a clado-gram can result either from many sharedderived neutral characters and a recentvicariant event, or independent of history,from many shared derived characters un-der the same selection regime. This is trueeven if not all characters are under selec-tion. Species experiencing different selec-tion regimes will not be assigned to sistergroups by a cladistic analysis unless nocharacters are subject to selection. This isa direct corroboration of Chernoffs re-sults. It is necessary, therefore, to showthat taxonomic characters are not underselection. The problem of distinguishinghistorical from present-day factors is a se-rious one.Brown's paper presents one of the mostextensive sets of biogeographical data andanalysis ever published, as well as one ofthe most balanced in terms of the consid-eration and testing of alternative hypoth-eses. It deals with the distribution of neo-tropical butterflies, and the effects ofhistorical and ecological factors. The ma-jor conclusions are that the distributions ofspecies, as well as zones of high diversity,are caused by: (1) ecologically favorableconditions and resulting regional differ-entiation, (2) unpredictable mild distur-bances over short and longer time scales,differing regionally, and (3) regional dif-ferentiation related to historical changes inclimate and the distribution of the vege-tation, including Pleistocene forest refuges(isolated patches of forest flora and fauna).Brown shows how ecology, dispersal, andvicariance exert varying effects, but all aredetectable as important factors. This paperis a splendid example of considering manyhypotheses, and with a knowledge of thebiology of the organisms, testing the hy-potheses in a reasonable and informativeway.

I t i s clea r fro m th e paper s tha t ther e i

sa great diversity of biogeographical phe-nomena, and a diversity of explanationsfor the phenomena. It is clearly importantto entertain more than one hypothesis ata time, and for each consider all the pre-dictions. A knowledge of the biology of theorganisms concerned helps to eliminate in-correct results which would otherwise ap-pear to be correct. No one method of anal-ysis is the best in biogeography, as eachonly gives a part of the story. There aremultiple explanations for species distribu-tions, and some effort should be made tointegrate them. Biogeography could profitby more attempts to examine the relativeand joint effects of both history and ecol-ogy on distributions. One possible methodwould be to explore the ecological factorsin sufficient detail so that these factorscould be removed, leaving components ofthe distributions which presumably reflecthistorical events. But this is a formidabletask. Brown's paper is a start. It is obviousthat much work needs to be done in at-tempts to understand both historical andDownloaded from https://academic.oup.com/icb/article/22/2/349/2015960 by guest on 16 August 2023

354JOHN A. ENDLER

ecologica l factor s i n specie s distributions , withou t losin g sigh t o f th e alternativ e hy - potheses .

ACKNOWLEDGMENT

S I a m gratefu l t o Stev e Farri s fo r sug -gesting that I organize the symposium,and to NSF-DEB-80-23918 for financialsupport.

REFERENCE

S

Andrewartha

, H . G . an d L . C . Birch . 1954
. The dis-tribution and abundance of animals. University ofChicago Press, Chicago.

Briggs

, J . C . 1974a
. Marine zoogeography. McGraw -Hill, New York.

Briggs

, J . C . 19746
. Operatio n o f zoogeographi c bar -riers. Syst. Zool. 23:248-256.

Carlquist

, S . 1965
. Island life. Natura l Histor y Press

,Garden City, New York.Darlington, P. J., Jr. 1957. Zoogeography: The geo-graphical distribution of animals. John Wiley andSons, New York.

Diamond

, J . an d R . M . May . 1976
. Islan d biogeog

-raphy and the design of natural reserves. In R.M. May (ed.), Theoretical ecology, principles and ap-plications, pp. 163-186. W. B. Saunders and Co.,Philadelphia.

Ekman , S . 1953
. Zoogeography of the sea. Sidgewic kand Jackson, London.

Eldredge

, N . an d J . Cracraft . 1980
. Phylogenetic

pat-terns and the evolutionary process. Columbia Uni-versity Press, New York.Endler, J. A. 1977. Geographic variation, speciation, anddines. Princeton University Press, Princeton, NewJersey.

Hull , D . L . 1979
. Th e limit s o f cladism . Syst . Zool .28:416-440. Krebs , C . J . 1972
. Ecology, the experimental analysis ofdistribution and abundance. Harper and Row, New York

.MacArthur, R. H. 1972. Geographical ecology. Harperand Row, New York.MacArthur, R. H. and E. O. Wilson. 1967. The theory

of island biogeography. Princeto n Universit y Press ,

Princeton

, Ne w Jersey .May, R. M. (ed.). 1976. Theoretical ecology, principles and applications. W . B . Saunder s Co. , Philadel -phia.Nelson, G. 1979. Cladistic analysis and synthesis:

Principle

s an d definitions , wit h a historica l not eon Adanson's Families des Plantes (1763-1764). Syst . Zool . 28.1-21
.Nelson, G. and N. I. Platnick. 1978. The perils of plesiomorphy : Widesprea d taxa , dispersal , an

dphenetic biogeography. Syst. Zool. 27:474-477.Nelson, G. and N. I. Platnick. 1980. A vicariance ap-proach to historical biogeography. Bioscience

30:339-343

.Nelson, G. and D. E. Rosen (eds.). 1981. Vicariancebiogeography: A critique. Columbia University Press , Ne w York

.Pielou, E. C. 1979. Biogeography. Wiley-Interscience,New York.Platnick, N. I. and G. Nelson. 1978. A method of

analysi s fo r historica l biogeography . Syst . Zool .27:1-16.Prance, G. (ed.). 1981. Biological diversification in the tropics. Columbi a Universit y Press , Ne w York .Rosen, D. E. 1975. A vicariance model of Caribbean biogeography . Syst . Zool . 24:431 - 164
.Rosen, D. E. 1978. Vicariant patterns and historical explanatio n i n biogeography . Syst . Zool . 27:159
- 188

.Simpson, G. G. 1965. The geography of evolution. Chil-ton Books, New York.Udvardy, M. 1969. Dynamic zoogeography. Van Nos-

trand-Reinhold , Ne w York .Wiley, E. O. 1981. Phylogenetics, the theory and practice of phylogenetic systemalics. Joh n Wile y an d Sons , Ne w York .Downloaded from https://academic.oup.com/icb/article/22/2/349/2015960 by guest on 16 August 2023

Biogeography Documents PDF, PPT , Doc

[PDF] biogeographic zones

1. Science

2. Earth Science

3. Biogeography

[PDF] biogeographic zones meaning

[PDF] biogeographic zones of india

[PDF] biogeographic zones of india map

[PDF] biogeographic zones of india pdf

[PDF] biogeographic zones of india ppt

[PDF] biogeographic zones of india upsc

[PDF] biogeographic zones of world

[PDF] biogeographical

[PDF] biogeography a study of plants in the ecosphere

12345 Next 200000 acticles