MÉMOIRES DU MUSÉUM NATIONAL D’HISTOIRE NATURELLE

Directeur de la publication : Philippe Bouchet

Rédacteurs (Editors) : P. Bouchet, A. Dubois, C. Erard

Secrétariat : Bernadette Charles

Conception graphique : Alain Defilippi

Rédaction : 57, rue Cuvier

75005 Paris

Les Mémoires du Muséum national d’Histoire natu¬

relle publient des travaux originaux majeurs (100

pages et plus) dans les domaines suivants : Zoologie

(série A), Botanique (série B), Sciences de la Terre

(série C). Les auteurs sont invités, pour toutes les

questions éditoriales, à prendre contact avec le direc¬

teur de la publication.

Vente en France

(uniquement)

Éditions du Muséum

Lionel Gauthier

38, rue Geoffroy St-Hilaire

75005 Paris

Tél. : (1) 43-31-71-24

Telex MUSNAHN 202641 F

CCP : 9062-62 Y Paris

Parution et prix irréguliers. Les ordres permanents

d’achat et les commandes de volumes séparés sont

reçus par le service de vente. Catalogue sur demande.

Une liste des derniers titres parus figure en page 3 de

couverture.

Mémoires du Muséum national d’Histoire naturelle

publishes major original contributions (100 pages and

over) in three different sériés : Zoology (série A),

Botany (série B), Earth Sciences (série C). Prospective

authors should contact the Editor. Manuscripts in

French and English will be considered.

Sales Office

(France excluded)

E. J. Brill

P.O. Box 9000

2300 PA Leiden

The Netherlands

Tel. : (071) 312624

Telex 39296 brill ni

Fax. (071) 317532

Volumes are published at irregular intervals, and at

irregular prices. Standing orders and orders for single

volumes should be directed to the Sales Offices. Free

price list and catalogue on request. Recently published

memoirs are listed on page 3 of the cover.

The Muséum national d’Histoire naturelle also

publishes a Bulletin.

Source : MNHN, Paris

The genus in ^ oology:

a contribution to the theory of evolutionary systematics

English version.

This text also exists in French version :

Le genre en zoologie : essai de systématique théorique.

Mèm. Mus. natn. Hist. nat., (A), 139 : 1-132 — Paris

ISBN 2-85653-152-0.

Source : MNHN, Paris

ISBN : 2-85653-151-2

ISSN : 0078-9747

Source : MNHN, Paris

-Z6o

MÉMOIRES DU MUSÉUM NATIONAL D’HISTOIRE NATURELLE

SÉRIE A

ZOOLOGIE

TOME 140

Alain DUBOIS

Muséum national d’Histoire naturelle

Laboratoire de Zoologie, Reptiles et Amphibiens

25, rue Cuvier

75005 Paris

The genus in %oology:

a contribution to the theory

of evolutionary systematics

ÉDITIONS

DU MUSÉUM

PARIS

I988

Source : MNHN, Paris

CONTENTS

Pages

Abstract. 11

Résumé. 13

INTRODUCTION

Brief historical survey. 15

PRESENTATION OF THE PROBLEM. 16

THE CONCEPTS OF THE GENUS

EMPIRICAL CONCEPT. 19

Phenetic concept. 21

Cladistic concept . 21

Synthetic concept. 22

THE GENUS as a genetic unit. 23

THE GENUS as a phylogenetic unit. 24

THE GENUS AS AN ECOLOGICAL UNIT. 26

THE PROBLEM OF GENETIC SIMILARITY

Genetic “ distances ” and “ similarities ”. 29

Structural genes and regulatory genes . 32

Hybridization and genetic similarity. 35

A FEW GENERAL FACTS ABOUT ANIMAL HYBRIDIZATION

The mechanisms of interspecific isolation. 37

Exogeneous factors of isolation. 37

Geographical barriers. 37

Temporal barriers. 37

Source : MNHN, Paris

BlOLOGICAL MECHANISMS OF ISOLATION. 38

Pre-ejaculatory mechanisms. 38

Ecological mechanisms . 38

Behavioural mechanisms . 38

Mechanical mechanisms . 38

Post-ejaculatory mechanisms. 38

Prezygotic mechanisms . 38

Postzygotic mechanisms . 39

SOME GENERAL RULES DRAWN FROM THE STUDY OF HYBRIDS. 40

VARIABILITY OF RESULTS WITHIN A GIVEN TYPE OF CROSS. 40

STUDY OF RECIPROCAL CROSSES. 41

The major stages of failure of hybridization. 41

Arrest of development at the end of the blastula stage. 41

Arrest of development at an embryonic stage subséquent to the blastula stage. 42

Infertility of hybrids. 42

Genic expression in hybrids. 43

CONSEQUENCES AS TO THE USE OF HYBRIDIZATION IN SYSTEMATICS. 44

Elimination of “ parasitic ” factors interfering with hybridization. 44

DETECTION OF TRUE DIPLOID HYBRIDS. 45

INTERSPECIFIC HYBRIDIZATION AND SUPRASPECIFIC CLASSIFICATION. 46

VARIABILITY OF THE RESULTS WITHIN A TAXINOMIC GROUP. 46

Hybridization and molecular divergence between species. 46

Hybridization, phenetic similarity and cladistic kinship between species .... 48

The different types of “distances” between species. 48

Phenetic distance. 49

“ Genetic ” distance . 50

Cladistic distance . 50

Karyological distance. 50

Ecological or eco-behavioural distance. 50

Hybrid distance. 51

INTERSPECIFIC HYBRIDIZATION

AND THE CONCEPT OF GENUS IN ZOOLOGY

HYBRIDIZABILITY as a CRITERION FOR THE DEFINITION OF GENERA. 53

Brief historical survey. 53

PRECISE FORMULATION OF THE CRITERION AND OF ITS CONDITIONS OF USE. 54

TAXINOMIC CHARACTERS AND RELATIONAL TAXINOMIC CRITERIA. 56

THE CRITERION OF HYBRIDIZABILITY AND THE PROBLEM OF THE EQUIVALENCE OF HIGHER TAXA 59

Introduction. 59

THE CRITERIA OF EQUIVALENCE BETWEEN TAXA. 60

Phenetic criteria . 61

Source : MNHN, Paris

“ Genetic ” or molecular criteria. 61

Ecological criteria. 61

Absolute âge of taxa. 62

VAN Valen’s metataxinomic criterion. 64

Hybridizability criterion. 65

Choice of the developmental stage . 65

Choice of the taxinomie rank . 66

The hybridizability criterion and the classification of the Vertebrata Gnathosto-

MATA. 67

Introduction. 67

Amphibians and Reptiles. 67

Bony fishes. 68

Mammals. 68

Birds. 69

CRITICAL STUDY OF THE USE OF THE HYBRIDIZABILITY CRITERION TO DEFINE GENERA. 72

SOME ARGUMENTS AGAINST THE USE OF THIS CRITERION. 72

SOME PRACTICAL ARGUMENTS IN FAVOR OF THE USE OF THIS CRITERION. 75

The criteria of the genus. 76

Conclusion. 78

GENETIC REVOLUTION AND GENIATION:

THE GENUS AS AN EVOLUTIONARY UNIT

Phylogeny and ontogeny. 79

PHYLETIC GRADUALISM AND QUANTUM EVOLUTION: ARE GENERA DISCONTINUOUS?. 83

TRANSILIENCE, GENETIC REVOLUTION AND GENIATION. 86

GENIATION. 86

MAYR’S MODEL OF GENETIC REVOLUTION . 86

OTHER MODELS OF GENETIC REVOLUTION. 88

Genetic révolution and chromosomal rearrangements. 91

Genetic révolution as a mode of spéciation among others. 92

Questions of terminology. 94

Genetic révolution and geniation. 96

Conclusion. 99

THE TAXINOMIC CATEGORIES

BETWEEN THE GENUS AND THE SPECIES

SuPERSPECIES, ULTRASPECIES AND SUPRASPECIES. 101

SPECIES COMPLEX AND SPECIES GROUP. 102

Source : MNHN, Paris

Synklepton. 103

The subgenus. 104

Introduction . 104

The criteria of the subgenus. 105

Distinction between subgenus and genus. 105

Hybridizability . 105

Evolutionary reversibility of characters . 106

Absence of discontinuities between subgenera . 107

Distinction between subgenus and species group. 107

NOMENCLATURAL INTEREST OF THE SUBGENUS. 108

Conclusion. 109

Acknowledgements. 109

REFERENCES . 111

Source : MNHN, Paris

Abstract

DUBOIS, A., 1988.II.18. THE GENUS IN ZOO-

LOGY: A contribution to the lheory of evolutionary

systematics. Mém. Mus. natn. Hist. nat. (A), 140 :

1-124. Paris ISBN 2-85653-151-2.

(1) Despite its importance, both theoretical and

practical, in animal systematics, the concept of the

genus has until now been largely neglected by the

theoreticians of classification. The présent work offers

a reflection on this concept and on related ones, and a

detailed study of a new criterion proposed to define

généra, that of hybridizability.

(2) The analysis proceeds within the framework of

an “ evolutionary ” or “ synthetic ” conception of

classification. It is suggested that généra should be

defined as genetic, phylogenetic and ecological units,

three concepts here made explicit. Thus defined,

généra are discontinuous evolutionary units which

exist really in nature, and not créations of the human

mind.

(3) The problem of the genetic similarity between

two organisms is studied in detail. The analysis

presented insists upon the importance of the rôle of

regulatory genes in the morphological évolution of

organisms and in the phenomena of spéciation, as well

as on the independent évolution of regulatory genes

from that of structural ones. It follows that criteria

like “ genetic distance ”, which measure the divergence

between organisms at the level of structural genes, are

of little use for the construction of an evolutionary

classification. Rather, classification must rely upon

synthetic criteria, such as those derivable from the

analysis of the morphology, or also from the study of

interspecific hybridization.

(4) It is to be hoped that in the future evolutionists

and systematists will grant more importance than thus

far to the study of interspecific hybridization, of its

mechanisms and conséquences, in a double perspec¬

tive: analysis of evolutionary phenomena in zoology,

and applications at the level of supraspecific classifica¬

tion. Most attention should be devoted to the positive

results of interspecific hybridization, which hâve a

clear meaning (criterion of functional genetic similar¬

ity, and proof of a common phylogenetic origin of the

hybridized species); in contrast négative results are of

little interest to systematists. The concept of “ hybrid

distance " deserves review in the light of the présent

suggestions and of the works already achieved in this

field by Gregory S. Whitt and his co-workers; the

results obtained with this index should be compared

with those generated by other comparative techniques.

It is likely that such comparisons with other types of

interspecies “ distances ” (phenetic, “ genetic ” or

molecular, cladistic, karyological, eco-behavioural dis¬

tances) will provide interesting lessons about the

modalities of animal évolution.

(5) Review of the major results drawn from the

study of both natural and experimental hybrids in the

animal kingdom, and of the relations which exist

between these results and the other available data

concerning animal species, leads to the proposai of the

new criterion of hybridizability to identify généra in

zoology: whenever two species can give viable adult

hybrids, they should be included in the same genus; if

other valid criteria had led them previously to be

placed into different généra, these must be merged.

(6) The criterion of hybridizability is a relational

taxinomie criterion. Such criteria rely on the charac-

teristics of the relations between the organisms that are

compared. They differ from traditional taxinomie

characters , which are gathered on the organisms taken

separately and later compared “ from the outside ”, in

the mind of the observer. It is suggested that such

relational criteria may play an important rôle in

“ evolutionary ” or “ synthetic " systematics, although

they hâve thus far been neglected in favor of the

" analysis of characters ”, and that they deserve a

more thorough theoretical and practical investigation.

(7) The new criterion gives the genus category a

deep biological and evolutionary meaning and makes

possible a standardization of supraspecific systematics

in the whole animal kingdom. Better than other

possible criteria, it provides a partial solution to the

problem of the équivalence of higher taxa among

different groups.

(8) The conséquences of the application of this

criterion to the current classifications of the five major

classes of gnathostome vertebrates are examined, in

the light of Van Valen’s (1973) metataxinomic crite¬

rion. This study suggests that application of this

criterion would hâve much more important consé¬

quences in some groups (like birds) than in others. It

would be bénéficiai, as permitting the suppression of

some biases of the current classification, due in

particular to the overestimation of the importance of

certain characters. Other arguments in favor and in

disfavor of the use of this criterion are studied.

(9) The mechanisms responsible for the birth of a

new genus (geniation) are discussed. Généra appear

discontinuous in nature, in morphological, genetic and

ecological terms. It is proposed that most geniation

phenomena involve spéciations by genetic révolution.

Source : MNHN, Paris

ALAIN DUBOIS

within small isolated founder populations. The notions

of genetic révolution and of transilience are discussed.

The importance of regulatory genes in the processes of

geniation by genetic révolution is emphasized. These

phenomena occur on the level of populations and do

not involve the sudden emergence of individual “hope-

ful monsters The study of the mechanisms of

geniation may permit an objective estimate of the

respective importance of genetic révolution and other

mechanisms in these events.

(10) Finally, the taxinomie categories between the

genus and the species (superspecies, ultraspecies, spe¬

cies complex, species group, synklepton, subgenus) are

discussed, and examples of the use of these various

categories are offered in the class of Amphibia.

Source : MNHN, Paris

Résumé

DUBOIS, A.. 1988.11.18. THE GENUS IN ZOO-

LOGY: A contribution to the theory of evolutionary

systematics. Mèm. Mus. natn. Hist. nat. (A), 140 :

1-124. Paris ISBN 2-85653-151-2.

(1) Malgré son importance, théorique et pratique,

en systématique animale, le concept du genre a été

jusqu’à nos jours largement délaissé par les théoriciens

de la classification. Le présent travail est consacré à

une réflexion sur ce concept et les concepts voisins, et

à l’étude détaillée d’un nouveau critère proposé pour

définir les genres, le critère d’hybridabilité.

(2) Nous plaçant dans l’optique de la conception

« évolutionniste » ou « synthétique » de la classifica¬

tion, nous préconisons de définir les genres comme des

unités génétiques, phylogénétiques et écologiques. Ces

trois concepts sont explicités. Ainsi définis, les genres

constituent des unités évolutives discontinues qui

existent réellement dans la nature, et non pas des

créations de l’esprit humain.

(3) Le problème de la similitude génétique entre

deux organismes est étudié en détail. L’analyse pré¬

sentée insiste sur l’importance du rôle des gènes de

régulation dans l’évolution morphologique des orga¬

nismes et dans les phénomènes de spéciation, ainsi que

sur l’indépendance de l’évolution des gènes de régula¬

tion par rapport à celle des gènes de structure. Il

résulte de cette analyse que les critères tels que la

« distance génétique », qui mesurent la divergence

entre organismes au niveau des gènes de structure,

sont de peu d’utilité pour la construction d’une

classification évolutionniste. Celle-ci en revanche doit

reposer sur des critères synthétiques, comme ceux que

permet de dégager l’analyse de la morphologie, ou

encore l’étude de l’hybridation interspécifique.

(4) Il est à espérer que dans l’avenir les évolution¬

nistes et systématiciens accorderont plus d’importance

qu’ils ne l’ont fait jusqu’à présent à l’étude de

l’hybridation interspécifique, de ses mécanismes, de ses

conséquences, dans une perspective double : analyse

des phénomènes évolutifs en zoologie, applications au

niveau de la classification supraspécifique. À cet égard,

il sera fondamental d’accorder la plus grande attention

aux résultats positifs de l’hybridation interspécifique,

qui ont une signification claire (critère de similitude

génétique fonctionnelle, et preuve d’une origine phy¬

logénétique commune des espèces hybridées), alors

que les résultats négatifs sont de peu d’intérêt pour les

systématiciens. Il sera indiqué d’explorer le concept de

« distance hybride », à la lumière de nos suggestions et

des travaux déjà effectués dans ce domaine par

Gregory S. Whitt et ses collaborateurs, et de confron¬

ter les résultats obtenus au moyen de cet indice avec

ceux fournis par d’autres techniques de comparaison

des organismes. Il est probable que la confrontation

de cette distance avec les divers autres types de

« distances » susceptibles d’être mesurées entre espèces

(distances phénétique, « génétique » ou moléculaire,

cladistique, caryologique, éco-éthologique) sera riche

en enseignements sur les modalités de l’évolution

animale.

(5) Après un rappel des principaux résultats tirés de

l’étude des hybrides, naturels et expérimentaux, dans

le règne animal, et des relations qui existent entre ces

résultats et les autres données dont on dispose sur les

espèces animales, un nouveau critère, le critère d’hybri¬

dabilité, est proposé pour reconnaître les genres en

zoologie. Il est suggéré que lorsque deux espèces

peuvent donner naissance entre elles à des hybrides

adultes viables, ces deux espèces doivent être incluses

dans le même genre ; si ces deux espèces étaient

auparavant classées, sur la foi d’autres critères vala¬

bles, dans deux genres distincts, ceux-ci doivent être

réunis.

(6) Le critère d’hybridabilité est un critère taxino¬

mique relationnel. De tels critères s’appuient sur les

caractéristiques des relations entre organismes com¬

parés. Ils s’opposent en cela aux caractères taxinomiques

traditionnels, qui sont recueillis sur les organismes pris

séparément et comparés ensuite « de l’extérieur », dans

l’esprit de l’observateur. Il est suggéré que de tels

critères relationnels peuvent jouer un rôle important

en systématique « évolutionniste » ou « synthétique »,

où ils ont été jusqu’à présent négligés au profit de

P « analyse des caractères », et qu’ils devraient faire

l’objet d'un examen, théorique et pratique, plus appro¬

fondi.

(7) Le nouveau critère donne à la catégorie de genre

une profonde signification biologique et évolutive et

rend possible une homogénéisation de la systématique

supraspécifique dans l’ensemble du règne animal,

permettant ainsi, mieux que les autres critères envisa¬

geables pour une telle opération, de résoudre partielle¬

ment le problème de l’équivalence des taxons supé¬

rieurs entre groupes différents.

(8) Les conséquences de l’application de ce critère

aux classifications actuelles des cinq principales classes

de Vertébrés Gnathostomes sont examinés, notam¬

ment à la lumière du critère métataxinomique de Van

Valen (1973). Il est conclu que cette application, dont

les conséquences seraient bien plus importantes dans

certains groupes (comme les Oiseaux) que d’autres,

Source : MNHN, Paris

ALAIN DUBOIS

serait bénéfique, car elle permettrait de supprimer

certains biais de la classification actuelle, dus notam¬

ment à la surestimation de l’importance de certains

caractères. D’autres arguments en faveur et en défa¬

veur de l’emploi de ce critère sont étudiés.

(9) Pour finir, les mécanismes responsables de la

naissance d’un nouveau genre (géniation) font l’objet

d’une discussion. Il est constaté que les genres sont

discontinus dans la nature, en termes morphologiques,

génétiques et écologiques. Il est proposé que la

majeure partie des phénomènes de géniation se pro¬

duisent à l’occasion de spéciations par révolution

génétique, au sein de petites populations fondatrices

isolées. Les notions de révolution génétique et de

transilience sont discutées. Le rôle important des gènes

de régulation dans les processus de spéciation par

révolution génétique est souligné, ainsi que le fait qu’il

s’agit d’évènements populationnels, et non de l'émer¬

gence brusque de « monstres prometteurs » indivi¬

duels. Il est à espérer que dans l’avenir des travaux

seront consacrés aux mécanismes de la géniation et

permettront d’estimer de manière objective l’impor¬

tance respective des phénomènes de révolution géné¬

tique et d’autres mécanismes éventuels dans ces évène¬

ments.

(10) Finalement, les catégories taxinomiques entre

le genre et l’espèce font l’objet d’une discussion, et des

exemples d’emploi de ces diverses catégories sont

donnés dans la classe des Amphibiens.

Source : MNHN, Paris

INTRODUCTION

Brief historical survey

This work has its origin in a strange observa¬

tion. Having demonstrated, during a study of the

amphibians of the Himalayan région (Dubois,

1974 a, 1975, 1976), the existence of a well

defined group of closely related species of Ranidae,

characterized by a peculiar ecology, I felt it

necessary to name this group, and I wondered

about the taxinomie 1 rank which should be

given to it: “ Genus, subgenus or species group? ”

(Dubois, 1976: 27). When I looked at the existing

scientific literature, I realized with surprise how

few publications had been devoted to a study of

the genus concept (and of related concepts) in

zoology. At this date and after a long bibliogra-

phical search, I know of only 42 publications

bearing the word “ genus ” (or “ subgenus ”) in

their title, and dealing with this concept: Cope,

1868; Clark, 1911; Alphéraky, 1912; Metcalf,

1915; Pia, 1920; Schenck, 1937; Bartlett,

1940; Camp, 1940; Greenman, 1940; Sherff,

1940; Hubbs, 1943; Mayr, 1943, 1965; Simpson,

1943; Williams, 1951; Edwards, 1953; James,

1953; Caïn, 1954, 1956; Mandelbrot, 1956;

Paclt, 1957; Inger, 1958; Tortonese, 1962;

Voous, 1964; Beck & Beck, 1968; Illies, 1970;

Rowell, 1970; Clayton, 1972; Moore, 1976;

Duellman, 1977; Van Gelder, 1977; Alberti,

1978; Bock & Farrand, 1980; Dubois, 1981 a,

1981 c, 1982 a, 1988; Plateaux, 1981; Bernardi,

1983; Daget, 1983; Stoyan, Stoyan & Fiksel,

1983; Lemen & Freeman, 1984. In addition, a

few interesting discussions concerning this pro-

blem appeared occasionally in some general

books (e.g.: Mayr, Linsley & Usinger, 1953;

Simpson, 1961; Mayr, 1969, 1982 a; Crowson,

1970; Ross, 1975) or in papers on taxinomy

(révisions, faunas, etc.) or on general zoology,

where they are difficult to trace (e.g.: Ghigi,

1936; Montalenti, 1938; Ripley, 1945; Simp¬

son, 1945; Inger, 1954; Laurent, 1956, 1964,

1972, 1973; Michener, 1957; Sibley, 1957; Caïn,

1958; Savage, 1958; Johnsgard, 1960; Parkes,

1961; Rosen & Bailey, 1963; Pasteur, 1964,

1982; Poynton, 1964, 1976; Kluge, 1966; Short,

1969; Lynch, 1970, 1971; Pépin et al., 1970;

Martin & Watson, 1971; Dubois, 1975, 1976,

1980b, 1981b, 1983a, 1983c, 1984a, 1984c,

1987 b; Gorham, 1977; McAllister & Coad,

1978; Avise & Aquadro, 1982; Sibley & Ahl-

quist, 1982).

The above list of references is certainly still

incomplète (I would be grateful to anyone who

could help me to complété it!), but the very fact

that it could be built up and that it only counts a

few dozen titles contrasts with the very high

number of papers and books devoted to a

discussion of the species concept (there certainly

exist several hundred, or even more, scientific

publications including the word “ species ” in

their title); this seemingly anecdotal observation

underlines the fact that the genus concept has

attracted the attention of the theoreticians of

zoological classification much less than did the

species concept. However, the fact that, in the

Linnaean System of nomenclature, the generic

name is part of the Latin binomial attributed to

ail species, and therefore appears in every scien¬

tific paper dealing with living beings, gives this

1. I use the correct spelling "taxinomy" instead of " taxonomy following Pasteur (1976) and Fischer & Rey

(1983).

Source : MNHN, Paris

ALAIN DUBOIS

name an important rôle in systematics, certainly

much more important than that of the names of

higher taxa.

Having realized this deficiency, I then devoted

several years to reflection on the genus concept

in zoology and to a discussion of this question

with various colleagues, either personally, or by

letters, or during conférences, symposia, meet¬

ings, etc. I also read several lectures on this

subject, and this gave me the opportunity for

fruitful discussions. Several papers summarizing

my ideas on this question hâve already been

published (Dubois, 1981a, 1981c, 1982 a, 1983 a,

1988). At the same time, I hâve started applying

the principles and criteria proposed on the basis

of general theoretical arguments, in works dea-

ling with the systematics of the Amphibia, my

major field of research (Dubois, 1975, 1976,

1977 c, 1979 a, 1980 b, 1981b, 1983 c, 1984 a,

1984 b, 1984 c, 1984 e, 1987 b). To the best of my

knowledge, until now only one author has

discussed my proposais in a publication, and

briefly studied the conséquences of the latter on

the classification of a given animal group: this

author is Daget (1983), and the group concer-

ned is that of bony fishes.

In August 1981, I discovered the papers of

Van Gelder (1977, 1978) where this author

proposed to use a criterion of hybridization to

identify généra in zoology. A similar suggestion

had been made by myself (Dubois, 1981 a,

1981 c, 1982 a) and by Plateaux (1981) on 14

March 1978 at a round table of the French

zoological Society in Paris (see Dubois, 1981 a).

The convergence between the three proposais is

most interesting: on the practical level, they ail

lead to the same resuit (combination in a same

genus of ail species liable to give birth to viable

hybrids), although the theoretical reasons given

by the three authors are appreciably different.

In this paper, I présent my ideas on this

question at the end of 1985, and I try to combine

in a single reflection ail the data, hypothèses and

discussions which I was led to meet during this

long search. This is certainly not a final word on

this question, on which I hope a rich discussion

will now develop.

PRESENTATION OF THE PROBLEM

Zoological classification in the Linnaean

System is based on a number of categories, from

subspecies to super-kingdom. A critical survey of

this System, of the concepts and methods, has

already been in progress for a long time. How-

ever the different categories hâve not ail been

subjected to an équivalent study. The most

abundant literature has certainly been produced

on the “ species problem ” (see e.g.: Mayr, 1963,

1970, 1982 a; Bocquet, Génermont & Lamotte,

1976, 1977, 1980). The question of the “higher

categories ” (family and above) is currently the

matter of a much debated discussion among

systematists (references to this question may be

found e.g. in Dupuis, 1979 and Mayr, 1981,

1982 a). However the genus, a category which

occupies an intermediate position between the

species and the higher categories, has been

largely neglected so far in these discussions.

As a matter of fact, for many zoologists the

only taxinomie category which corresponds to a

reaiity in nature is the species, and ail the other

categories are artificial. If this was true, there

would be no point in attempting a theoretical

approach of the genus concept. However, if one

follows Mayr (1969, 1981, 1982 a) and others in

recognizing that zoological classification does

not hâve for its only function to be a method of

identification, but should also provide a System

of storage for information, a genuine biological

theory which may be used as a basis for ail kinds

of comparative works, it would seem justified to

devote some attention to this category, which

plays an important rôle in supraspecific system¬

atics.

The practical agreement which exists between

specialists as to the délimitation of généra varies

from one zoological group to another. In many

groups, this agreement is poor, and this results in

a great instability of the generic classification

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

and nomenclature. It is true that a similar

phenomenon also exists for the higher categories.

However, a fundamental différence exists be-

tween these and the genus: in the Linnaean

System of nomenclature, the generic name is part

of the Latin binomial given to every species,

which makes the need for its stability particularly

impérative.

In Linnaeus’ mind, an important function of

the generic name was to relieve the memory

(Caïn, 1958), by collectively designating a group

of “ related ” or “ similar ” species. Nowadays,

ail systematists would probably agréé on a

définition of the genus as loose as this: “ a genus

groups together species doser to each other than

to species of other généra ”. There remains to

define what is meant by “ close ”, and, according

to the définition which will be given of this term,

radically different conceptions of the genus will

ensue. In other words, the whole “ genus prob-

lem ” boils down to deciding which information

must be carried by the generic name.

Many systematists hâve dreamed of a classifi¬

cation of the animal kingdom in which the

different taxa of a same category would be

équivalent in the various groups of animais (i.e. a

genus of butterflies would be équivalent to a

genus of birds or of molluscs, etc.). The search

for this équivalence has led some systematists to

adopt simple, or even simplistic, criteria to define

généra, which will be discussed below. This

problem of the équivalence of taxa is a difficult

one, because of the absence of common charac-

ters between different groups (see Schaefer,

1976 ), but it may not be insolvable, as we shall

see.

Another problem related to the preceding one

is that of the reality of higher taxa (Ball, 1983 ).

The question may be put in the following way:

are the taxa which zoologists recognize artificial

groupings of individuals, i.e. entities made up

entirely by biologists — or entities which really

exist in nature, independent from the conscious-

ness systematists may hâve of them? According

to the answer which will be given to this

question, different methodological impératives

will resuit for systematists. If biological taxa are

créations of the human mind, it will be impor¬

tant to fix rules to establish them. Several types

of criteria may then be chosen to reach this aim,

but the choice of the best criteria will finally be

determined by pragmatical considérations: if

taxa hâve no proper existence in nature, the best

classification will be the one which will make

easiest the work of systematists, and possibly of

other biologists (the most “ practical ” one, in

the various meanings of this term). On the other

hand, if taxa do exist in nature, independently

from the idea we may hâve of them, the task of

systematists will then be to find them, to rec¬

ognize them, even if this is not easy, and if this

does not necessarily facilitate the work of biolo¬

gists later on (e.g. insofar as particular or heavy

techniques must be called upon to recognize

them).

Mayr (1982 a: 207-208) rightly emphasized

the fact that this problem is partly semantic, and

cornes partially from the confusion which has

long existed between the concepts of category

and of taxon. A taxon is a group of organisms of

any rank which is distinct enough to be worth

naming and assigning to a given category. In

terms of logics, a taxon is an individual, and the

animal or végétal organisms which constitute it

are the parts of this taxon. On the other hand, a

category , in the contemporaneous sense of this

term, indicates a rank or level in a hierarchical

classification. It is a class, the members of which

are ail the taxa which are ascribed a given rank.

Relying on this distinction, Mayr (1982 a:

208) writes:

“ The question. Are the higher categories real ? must thus be dissolved

into two separate questions: (1) Are (most of) the groups (taxa) which we

rank in the higher categories well delimited ? and (2) Is it possible to give an

objective (nonarbitrary) définition of such higher categories as genus, family,

or order ? The answer to the first question is clearly yes, but to the second

one it is clearly no ”.

According to this conception, which has often

been defended and illustrated in the scientific

literature, the classificatory process would con-

sist in two steps: first the récognition of taxa

(whatever the method used to do it), then the

establishment of the rank of each taxon. Only

the first of these two steps would really be

“ nonarbitrary ”. The allocation of given ranks

Source : MNHN, Paris

ALAIN DUBOIS

to taxa would be made in particular on the basis

of the size of the divergence between them. Such

a conception implies that ail taxa are fundamen-

tally of the same nature, but fit into each other

like in a nest of dolls: thus eventually a subgenus

would be a “ small genus ” or a “ nascent

genus ", a genus would be a “ nascent family ”,

etc.

At the level of the species, the key-category of

the Linnaean hierarchy, application of this con¬

ception would be wrong: the subspecies, as it is

now understood. is not a “ small species it is

not even, or not necessarily, a “ nascent spe¬

cies Most contemporaneous systematists agréé

to say that the species category is not an

invention of the human mind, but that it corres¬

ponds to an objective reality in nature. In other

words, independently from the conception biolo-

gists may hâve of them, entities exist in nature

which correspond to the species concept as

biologists now define it, i.e. a “ closed, or

protected, gene pool ” (Bocquet, Génermont &

Lamotte. 1976, 1977, 1980): to use again Mayr’s

(1982 a) words cited above, this définition is

therefore “ objective ” and “ nonarbitrary ”. The

task of systematists is then to recognize the

species in nature and not any more to “ create ”

species. In the scale of Linnaean hierarchy, the

species would thus be a fixed point, the position

of which would be given in an objective way; on

the other hand, the position of the higher

categories would be arbitrary, and there would

be no point in trying to fix it in an objective

manner.

Yet, the species is not the only systematic

category fiable to be defined in a rigorous and

objective way. Bernardi (1980) recently pro-

vided a study of several categories designated by

this author as “ the taxinomie categories of

evolutionary systematics ”. Ail these do not

show the same interest or importance, but some

of them, like the superspecies (“ monophyletic

group of entirely vicariant species ”, Bernardi,

1980: 385) and the prospecies (the vicariant

species which together make up a superspecies),

indisputably correspond to real entities in

nature. In the case of categories like the species

or the prospecies, the distinction made above

between récognition (or délimitation) of the

taxon and establishment of its rank is not

warranted any more. The criteria which allow

the récognition of the taxinomie unit and its

attribution to a given category are the same

ones.

Is such a criterion proper to the species

category and to the categories just above and

below the species, studied by Bernardi (1980),

or is it possible to recognize also natural units at

a higher level in the Linnaean hierarchy? This

would only be the case if it was possible to find

objective, nonarbitrary criteria to define these

taxa. Such criteria would allow, as in the case of

the species or of the prospecies, to recognize

concomitantly both the existence of the natural

taxon and its taxinomie rank. The thesis which

will be defended here is that such criteria exist

and may be found: as concerns the genus

category, on which the following discussion is

centered, I propose the use of a new criterion,

the success of interspecific hybridization.

Before discussing this point, however, it is

useful to make a rapid survey of the four major

types of concepts of the genus category which

may be found in the literature, so as to be able to

place the new proposai within this general frame.

Source : MNHN, Paris

THE CONCEPTS OF THE GEN US

Empirical concept

An empirical concept of the genus underlies

the practice of many systematists, who consider

that there is no need for a theory of the genus.

For these authors, the genus is only a practical

convention, généra are pigeonholes which make

the identification of species easier. Therefore the

genus does not correspond to a real unit in

nature, it is a création of the mind.

These authors insist upon the fact that the

genus must be useful, “ pratical This notion of

“ practical ”, however, is not clear. Does it mean

“ easy to recognize ”? or “ easy to identify ”?

“ not too large ”? “ bringing such or such type of

information ”?

For many systematists, the criterion of size is

given pre-eminence, which may be expressed by

saying that “ a genus must contain neither too

many nor too few species These authors tend

then to group together the isolated species in

artificial généra, and to break up large généra, in

order to obtain finally a mean number of

“ pigeonholes ” of similar “ volumes

A few authors are in favor of généra of a

rather large size:

“ I personally feel that one should use rather large généra, such a solution

being préférable in general biology, where scientific names of animais must

be familiar to the largest number. ” (Bernardi, 1983: 136; translation mine).

Other systematists, probably more numerous,

recommend on the contrary to reduce the size of

généra as much as possible:

“ When a genus contains a large number of species and that it is possible

to recognize within it natural groups by whatever means, it is désirable to

split it in several généra. ” (Laurent, 1956: 230; translation mine).

“ In entomology there is sentiment in some quarters for setting an upper

limit (perhaps 40) to the number of species allowed in a single genus. "

(Ross, 1975, cited by Van Gelder, 1977: 2).

Rosen & Bailey ( 1963 ) hâve stressed the fact

that, as the systematics of a group develops, one

may often observe the following phenomena:

first discovery, description and counting of the

species; then tendency to “ put order ” in this

mass of species and to group these together by

affinities. The authors then often tend to create a

genus for any group of species that may be

shown to be closely related, and to break up

généra as soon as new heterogeneities are dis-

closed in them. Eventually, they tend to rec¬

ognize smaller and smaller, often monospecific,

généra.

Such a practice dénotés a misunderstanding of

the fundamental meaning of the Linnaean bino¬

mial, where both names hâve different functions,

the spécifie name expressing the singularity, and

the generic name the existence of a group of

Source : MI'JHN, Paris

ALAIN DUBOIS

units which are “ close ” (or “ similar ”, or différence. The generic name as it is conceived by

" related ”, or both). The genus, contrary to the the splitters does not carry information any

species, is a collective unit, and the first function more, or hardly so:

of the generic name is to express similarity, not

“ This différence in the functions of species and genus names is completely

ignored by many recent taxonomists, particularly the so-called generic

splitters. It is their aim to express différence not only in the spécifie, but also

in the generic name. This tendency, if carried to its logical extreme, leads to

uninomialism, and some of the leading generic splitters hâve openly or in a

veiled form endorsed this principle of nomenclature. To me it seems to

indicate a complété misunderstanding of the principle of binomial nomen¬

clature, if somebody uses the generic name primarily to express différence.

This is the function of the species name. ” (Mayr, 1943: 138).

It is useful in this respect to reproduce the (1963) on poeciliids, where similar ideas are

following extract of the work of Rosen & Bailey expressed very clearly:

“ It is évident that the phylogenetic relationships between different species

or between distinct groups of species are refiected in a host of features, some

anatomical and behavioral, some physiological and biochemical, and some

genetic and developmental — in short, in some features of ail the major

Systems and processes that characterize organisms. Hence, investigators in

the fields of comparative anatomy and comparative development, as well as

many experimental biologists, may contribute directly and indirectly to

systematic knowledge. For the non-systematist, however, a classification

consisting of too many small généra présents a major obstacle to his efforts

at recognizing différences and similarities between related organisms that are

worthy of study. In an earlier classification of the poeciliids, for example,

two species now shown to be intimately related were placed in different

généra because one of them possesses an asymmetrical extemal genitalium.

Under this taxonomie arrangement, a developmental biologist interested in

problems of asymmetry and hence in the asymmetric species would hâve

difficulty identifying the symmetrical relative, the comparative study of

which might be expected to yield important dues as to the origin of the

asymmetric condition. The use of separate généra is usually, and justifiably,

taken as a mark of the gross genetic incompatibility of the species thus

separated taxonomically, and in the foregoing example may be expected to

draw attention away from important biological properties common to both.

It may even, at times, tend to prevent experimental workers from realizing

that the comparative study of both species is appropriate. When a single

genus is used for these species, the comparative materials are collected

together, and the likelihood is increased that studies in other fields will be

performed by investigators whose results are important to systematics. In

general, the masking effects of an oversplit classification may be remedied by

the use of comprehensive généra that assemble, and thereby underscore,

some contrasting features as well as the many unifying characters to be

found among intimately related organisms. The function of broad and co-

ordinate généra, when data on complex and little-known groups are made

available to non-systematists, is often overlooked by the taxonomist. ”

(Rosen & Bailey, 1963: 6).

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

Phenetic concept

If one admits that the genus must group

together species which are “ doser ” to each

other than they are to other such groups, what

does the word “ close ” mean?

For the pheneticists, classification is based on

the estimation or measurement of affinities,

understood here in the sense of phenotypic

similarities, which are in general supposed to

represent genotypic similarities.

The old method, which can be traced to the

early stages of systematics, consists in comparing

the morphology of species, and in using these

data for estimating, more or less subjectively,

their resemblance. The latter is sometimes, but

not always, supposed to express their genetic

likeliness and their affinity. The modem aspects

of this method rely on biometry (quantification

of characters) and more recently (in the last 20

years) on numerical taxinomy, which takes into

account a high number of characters (Sneath &

Sokal, 1973). These characters may be taken

from the morphology, but also from the ecology,

the behaviour, the chromosomes, etc., i.e. the

analysis bears on the totality of the holomorph

(Hennig, 1950). This leads to the estimation of a

“ distance ” between species, and ail the dis¬

tances between several species may be presented

graphically, e.g. on a phenogram. Pheneticists

hâve sometimes ascribed a given systematic level

to a given level of morphological “ divergence

thus, two groups of species separated by such a

distance will by définition be considered as two

généra, by such another distance as two families,

etc.

This leads to the grouping together of similar

species. In general this resemblance is due to the

common presence, in these species, of characters

retained from a recent common ancestor. But

this is not always the case. The resemblance may

be due to the existence of a remote common

ancestor, a parallel évolution having taken place,

as a resuit of the presence at the start of genetic

factors common to two long separated stocks.

The resemblance may also be due to a conver¬

gence between different lineages, when these tend

to adapt to similar modes of life.

The numerical methods of measurement of

phenotypic similarities recently developed are

valuable because they allow an objective, or

almost so, estimate of the resemblance between

two types of organisms, but they alone do not

allow for the construction of a classification of

living species. Such methods would be sufficient

to classify objects, but not living beings which

are the resuit of a history and which live in an

environment.

Cladistic concept

As has been stressed by several authors, and

singularly Mayr (1974), it is important clearly to

distinguish between two fundamental aspects in

the works of Hennig and his disciples. The aim

of cladistic analysis is to reconstitute as accura-

tely as possible the phylogeny of a given group

(establishment of a cladogram). Remarkable

progress has been achieved in this domain by the

cladists through the élaboration and formaliza-

tion of principles and methods of work, some of

which had been applied already long before

Hennig but in a much less systematic and

rigorous manner. Ail zoologists who are inter-

ested in the study of phylogeny must become

acquainted with the works of Hennig and his

disciples in this domain, and it is surprising that

a few works are still published on these questions

(Clark, 1977; Blandin, 1978), where the con-

Source : MNHN, Paris

ALAIN DUBOIS

cepts and methods of the cladist school are not

even mentioned. The works by Dupuis (1979,

1984), which offer an almost complété list of the

significant references in this field, may be con-

sulted fruitfully in this respect.

The principles of cladistic classification, on the

other hand, do not at ail ensue directly from the

preceding analysis. They constitute in reality a

set of arbitrary rules and conventions aiming at

the automatic transcription of the phylogeny

into a classification.

As a matter of fact, for the cladists classifica¬

tion must only be a transcription, as exact as

possible, of the phylogenetic tree or cladogram

on another level. The rules adopted for this

transcription are relatively simple. First, any new

cladogenesis automatically gives birth to two

new taxa. Then, only a taxon which contains ail

the descendants of a given ancestral species and

this species itself is considered monophyletic.

Finally the cladists hâve looked for a simple

criterion making ail taxa équivalent in different

groups: it has thus been proposed that the rank

of a taxon be automatically given by its âge, or

by the number of cladogeneses having taken

place since its appearance. While classical methods

of détermination of the âge of taxa posed

important problems (see e.g. the discussion in

Dupuis, 1979: 47-50), Sibley & Ahlquist (1982)

recently suggested that DNA hybridization would

allow the dating of cladogeneses in a relatively

précisé way. The use of such a criterion would

lead for example to place ail higher Primates in a

single genus, or on the contrary to upgrade the

frogs genus Rana to the rank of an order.

As far as they are concerned, systematists of

the “synthetic” school (e.g. Gisin, 1964, 1966;

Mayr, 1969, 1974, 1981), think that the cladist

conception of classification, based on phylogeny

alone, or rather on a restricted conception of

phylogeny (considering only one of its aspects,

cladogenesis), is singularly poor. As a matter of

fact, cladist classification does not take into

account the more or less important divergence

between lineages which results from the existence

of different rates of anagenesis between different

lineages and at different epochs. Genealogical

kinship and genetic similarity are not équivalent.

A purely genealogical classification does not give

any measure of the morphological, ecological

and genetic resemblances between species. It

does not take into account the transitions into

novel adaptive zones, and the speeding up of

evolutionary rates during these ecological shifts.

It may not be superfluous to point out that the

criticism of some aspects of the “ monophyletic ”

classification, i.e. based on genealogy alone, had

been done in its broad fines very clearly and even

before its birth by Bigelow (1956: 146) in a

forgotten passage which is worth quoting:

“ Without overlooking the fact that resemblance reflecls phylogeny, it is

well to bear in mind that différence refiects évolution, and that the nature

and extent of these similarities and différences, not the time during which

they hâve been retained or effected, is the primary concern of evolutionary

classification. Organisms whose ancestors evolved very little relative to one

another should not be separated merely because évolution has been slow, or

grouped with organisms with whom they share a more recent common

ancestry despite extensive overall différences that hâve evolved between

them. Evolution is change, not time. If classification is to correspond with

évolution, it must be based on the extent of overall différence, not on time.

Monophyletic classification is based on recency of common ancestry (i.e. on

time), and therefore should not be regarded as even a ‘ theoretical ' idéal. "

Synthetic concept

For the generic name, which is part of the

Latin binomial attributed to each species, to be

useful, this name must contain the greatest

amount of information possible, and an informa¬

tion distinct from that carried by the spécifie

name. The three concepts of the genus that we

hâve briefly reviewed above hâve in common the

fact that the generic name carries in every case

Source : AANHN, Paris

THE GENUS IN ZOOLOGY

little information or no information at ail: no

defined information for the empirists, informa¬

tion concerning the resemblance alone for the

pheneticists, and the degree of kinship alone for

the cladists.

The supporters of the “ evolutionary ” (Mayr,

1969, 1974, 1981), “ synthetic ” (Gisin, 1964;

Dubois, 1981c, 1982 a) or “ quantic ” (Gisin,

1966), school of classification, on the other hand,

do not forget that the scientific name of species is

meant to be used also by nonsystematists and

must give them a synthesis of our knowledge on

the évolution and the mutual relationships of

groups. Although these three aspects are closely

connected, it may be useful to consider separa-

tely three types of information which may be

carried by the generic name: généra must be

evolutionary units, i.e. genetic, phylogenetic and

ecological units. We will now examine these

three aspects in more detail.

The GENUS AS A GENETIC UNIT

In his remarkable paper “ Biological classifica¬

tion: toward a synthesis of opposing méthodolo¬

gies ”, Mayr (1981) quite rightly explains that

the classificatory process, according to the sup¬

porters of the “ synthetic ” school, is necessarily

composed of several stages, and always begins by

a phase of grouping “ by inspection ” the species

considered “ doser ” to each other than they are

to species belonging to other groups. For this

work, empirical methods hâve long been used,

but it is now possible to call upon the more

elaborate methods of numerical phenetics alluded

to above. As we hâve seen, these methods permit

to group together “ similar ” species. In many

cases, this resemblance stems from the existence

of a strong genetic similarity between the species

which are being compared. As far as the artificial

groups due to evolutionary parallelism or to

convergence are eliminated (see below), the units

defined by such criteria may be interpreted as

genetic units: at any rate it is only this hypothesis

which justifies, in an evolutionary perspective,

the grouping of species according to their mor-

phological similarity.

Other criteria than morphological resemblance

can be imagined for recognizing genetic units.

One of these is the comparison of the proteins of

the species studied, which leads to what is

commonly called “ genetic distances ”. One of

the unexpected results of the research in this field

during the later years has been the discovery that

morphological évolution and spéciation on one

hand, and protein évolution on the other, are

largely independent, and that the study of the

two categories of phenomena may sometimes

lead to contradictory conclusions. The following

question may then be posed: which of the two

methods of estimation of the genetic similarity of

two organisms is the most reliable, the most

significant, one? Is it the measure given by what

is commonly called “ genetic distance ”, which is

based upon the characteristics of certain struc¬

tural genes of the species studied, or the estimate

given by the “ phenetic distance ” between these

species, which is based on a more synthetic

criterion, the global resemblance between the

two phenotypes compared? We shall address this

question in the next chapter, where we shall also

examine another possible method of comparison

of the genetic characteristics of two species, i.e.

interspecific hybridization. At the moment we

shall retain the traditional methods of study of

genetic resemblance of animal species: in this

respect it is clear that the oldest method, the

overall comparison of the phenotypes of the

species studied, remains by far the most generally

used one by systematists. As we shall see below,

this is not due only to the “ laziness ” or to the

“ lack of modernism ” of systematists, but also

to deeper causes: although it may seem strange

to hâve to précisé it, it is important to stress that,

in many cases, the fact that two organisms hâve

similar phenotypes is simply due to the fact that

they hâve similar génotypes, because they share a

common ancestor! We shall corne back to this

question.

Source : MNHN, Paris

ALAIN DUBOIS

The genus as a phylogenetic unit

The last sentences must not obscure the fact

that the resemblance between two organisms

may be due not to their having similar génotypes

retained from a common ancestor, but to phe-

nomena of convergence or of evolutionary paral-

lelism. For ail evolutionary systematists, it is

clear and indisputable that a taxon can only be

considered “ natural ” if it corresponds to a

monophyletic group. This means that it is very

important, in the construction of a classification,

to try to eliminate as completely as possible

artificial groups based on resemblances between

species due to convergence and, less often, to

evolutionary parallelism. This is the second of

the stages described by Mayr (1981) in the

building up of a classification. The methods to

use in this respect were first described by Hennig

(1950, 1966), then by his disciples, who hâve

considerably refined them (see e.g.: Dupuis,

1979; Farris, 1979; Wiley, 1981). They are now

part of the essential methods of ail work of

taxinomie révision, especially at higher levels.

With the help of these methods, the task of

systematists is to try as much as possible to

reconstruct the phylogeny, to break up ail poly-

phyletic groups and keep only monophyletic

groups.

A clarification is necessary here about the

définition to give to the term “ monophyletic ”.

For a long time, no précisé définition has been

elaborated for this word, which was simply used

to designate ail groups composed of species

descended from a same ancestor, i.e. ail non-

polyphyletic groups: this was consistent with the

first proposai of this term by Haeckel (1868),

who created “ monophyletic ” as opposed to

“ polyphyletic ”, and this was also consistent

with the etymology of these words. Simpson

(1961: 124) proposed the following définition:

“ Monophyly is the dérivation of a taxon through one or more lineages

(temporal successions of ancestral-descendant populations) from one imme-

diately ancestral taxon of the same or lower rank. ”

This définition is unacceptable for it is not On the other hand, Hennig (1950, 1966)

rigorous enough and it is based on the ranks of proposed a new définition of monophyly, which

taxa, i.e. on a criterion external to the phylogeny was reformulated by Wiley (1981: 76) as follows:

itself.

“ A monophyletic group is a group of species that includes an ancestral

species (known or hypothesized) and ail of its descendants. ”

This définition is rigorous and may be used to

build up a classification according to cladistic

principles. However by proposing it Hennig has

stripped the word “ monophyletic ” of its initial

sense (monophyletic being defined as opposed to

polyphyletic), to give it a completely new sense

(monophyletic being defined as opposed both to

polyphyletic and to paraphyletic). This approach,

which breaks with ail anterior taxinomie tradi¬

tion, has important conséquences in classifica¬

tion, and it has been severely criticized by several

synthetist authors (e.g. Mayr, 1974; Ashlock,

1980).

Because of the terminological confusion intro-

duced by the cladists in this field, it has been

necessary to coin new terms. Ashlock (1971,

1972, 1980), who devoted several excellent papers

to the clarification of this question, proposed the

new term of holophyletic to designate the concept

called “ monophyletic ” by Hennig and his

disciples. Even though, the cladists hâve con-

tinued to use the word monophyletic for this

concept, while on the other hand synthetist

authors use this word in its traditional sense. A

considérable confusion ensues in the contem-

poraneous taxinomie literature, which has led

Dubois (1986) to suggest abandoning completely

the term monophyletic, to use holophyletic for

“ monophyletic sensu Hennig ”, and the

new term homophyletic for “ monophyletic sensu

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

Ashlock Four different terms, designating

four different types of taxa, may thus be recogni-

zed. The following définitions are derived from

those of Ashlock (1971, 1972, 1980), taking

advantage of the remarks by Farris (1974),

Platnick (1977) and Wiley (1979, 1981) (see

fig. 1):

( 1 ) A homophyletic group (monophyletic sensu

Ashlock) is a group which contains the species

which it its most recent common ancestor.

(a) A holophyletic group (monophyletic sensu

Fig. 1. Phylogram illustrating the terms proposed by

Dubois (1986). The vertical axis represents the time, the

horizontal axis the divergence (in genetic, phenetic, ecolo-

gical. etc., terms).

(1) Groups ABCDE, AC, BDE and DE are holophyle¬

tic (monophyletic sensu Hennig).

(2) Groups AB, ABC and ABCD are paraphyletic.

(3) Group CDE is polyphyletic (heterophyletic).

(4) Groups of categories (1) and (2) are homophyletic

(monophyletic sensu Ahslock).

(5) Groups of categories (2) and (3) are merophyletic.

Acceptation of the above terms and défini¬

tions, which would not necessarily imply agree-

ment with one taxinomie school or another,

would greatly help in the clarification of debates

between the different schools, and therefore

appear most justified. In such a perspective,

cladists should admit that they try to recognize

only holophyletic groups and that they reject as

unnatural both polyphyletic and paraphyletic

groups (merophyletic groups; Dubois, 1986). On

the other hand, synthetists consider that ail

homophyletic groups may be natural, and that

only the polyphyletic groups must always be

rejected as unnatural.

As a matter of fact, how would the fact that a

group has arisen from another group change the

nature of the initial group? To use a famous

example, if birds had never appeared, reptiles,

including crocodiles, would be holophyletic. The

appearance of birds makes them paraphyletic,

but this does not in the least deprive reptiles

from any reality: the latter remain a homophyle¬

tic group which corresponds to a defined

“ grade Quite irrelevant in this respect is the

fact that birds, which correspond to a new grade

Hennig) is a homophyletic group which contains

ail the descendants of the species which is its

most recent common ancestor.

(b) A paraphyletic group is a homophyletic

group which contains only a part of the descen¬

dants of the species which is its most recent

common ancestor.

(2) A polyphyletic (or heterophyletic) group is

a group which does not contain the species which

is its most recent common ancestor.

AB CDE

and hâve conquerred a different adaptive zone,

are issued from the same stock as crocodiles. As

was well shown by Mayr (1974), the concept of

paraphyletic groups is devoid, for the synthetists,

of ail interest in classification. This divergence is

certainly, and by far, the most important one

which exists between the cladist and synthetist

conceptions of classification, which has not

always been well perceived: many of the authors

who hâve discussed cladism and compared it to

the synthetic systematics hâve only touched

lightly upon this problem and hâve even some-

times entirely omitted it. In the opinion of

synthetists, the automatic rejection of paraphyletic

groups stems from a methodological error, just

like the generalization of formai notions like

“ sister groups ”, “ sister species ”, etc., when in

many cases it would be much more justified to

speak of “ child groups ” or “ child species ”, etc.

For the synthetists, classification is based on

the phylogenetic tree but is not a simple and

automatic transcription of this tree: it takes

additional information into account. Thus a

same genus will be used to group species consid-

ered very close, according to their various char-

Source : MNHN, Paris

ALAIN DUBOIS

acteristics, and derived ail from a common

ancestor belonging to the genus. But this does

not imply that ail the descendants of this

common ancestor be included in the taxon: as a

matter of fact some of these descendants may

hâve considerably diverged, and may not be

“ very close ”, by their characteristics, to the

cladistically related species; it will then be justi-

fied to create for them a different genus, or

several.

The phylogenetic criterion allows therefore the

séparation, into distinct généra, of species which

show important morphological resemblances, due

for example to convergences, but which hâve

different phylogenetic origins. On the other hand

this criterion is insufficient in itself for deciding

whether various species of the same phylogenetic

origin must be grouped together in a same genus

or not. For this purpose, it is also necessary to

take into account other criteria mentioned above

and below (genetic and ecological unit).

The genus as an ecological unit

Every species may be characterized by its

ecological niche, a concept which refers to the

various interactions between the species and its

biotic and abiotic environment. Similarly, the

higher taxa may also be characterized by their

" niche ”, which is “ wider ” than that of any of

their included species.

Simpson (1944, 1953) proposed the phrase

adaptive zone to designate the various relation-

ships between groups of organisms and their

environment. Huxley (1958) popularized the use

of the term grade to designate the levels of

organization corresponding to given adaptive

zones. One may thus speak of the tetrapod

grade, of the homeotherm grade, of the mamma-

lian grade, etc. A grade is a group of organisms

which possess in common a number of adaptive

characteristics. A grade may be polyphyletic.

Inger (1958) proposed to define the genus as a

group of closely related species occupying the

same adaptive zone.

It is important to note that this criterion

applies only after the preceding ones:

— within a homophyletic group (therefore within

a single family);

— in agreement with the data concerning the

genetic similarity of the species.

This criterion applies then within a group of

species which are “ obviously close ” according

to ail other criteria, and among which (morpho¬

logical) subgroups may be recognized. in order

to try to establish whether these subgroups hâve

or not attained a generic grade.

Such a conception of the genus implies neces-

sarily that généra be separated by morphological

gaps. It was already an old conception of the

genus to admit that good généra must be

separated by a discontinuity, i.e. that no interme¬

dia te species exist between them. It was often

argued that the absence of these intermediate

forms was due to their extinction. However

Simpson (1961), while recognizing this fact, insisted

that extinctions, by producing these gaps, were

giving us a nonarbitrary criterion to define taxa.

It may be added that, at least in many cases,

extinctions are not random. The nonadaptive

zone which séparâtes two adaptive zones may

only be crossed by transitory, short lived popula¬

tions, which generally leave no fossils, and which

exhibit a high rate of anagenesis, thus allowing

genuine evolutionary jumps.

From a practical point of view, the problem is

to find a way of recognizing that different groups

of species occupy disjunct adaptive zones. The

idéal would be to hâve précisé data on the

ecology of the species, and in this respect the

works of ecologists may be most helpful to

systematists. To tell the truth, until now ecolo¬

gists hâve mainly devoted themselves to the

concept of ecological niche (see Blondel &

Bourlière, 1979) and to the comparison of

écologies of closely related species. One may

however hope that, in the future, comparative

works at higher levels, like those of Heyer (1973,

1974, 1976, 1979), Crump (1974), Barbault

(1974 a, 1974 b, 1980, 1984), Inger & Colwell

(1977), Duellman (1978) or Scott (1982), to

take only examples in the fields of batrachology

and herpetology, will develop. Such works could

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

allow a better knowledge of the characteristics

of adaptive zones, and a more objective estima¬

tion of ecological resemblances and divergences

between species of a same ecosystem or of

various ecosystems.

Admittedly, this is yet still largely impossible,

and, furthermore, for many groups of animais,

field data are rare or even completely lacking. In

such cases it will be necessary to infer from the

sole morphology the adaptive function of char-

acters. It goes without saying that such works

may be carried out only by specialists of the

group, having in particular a knowledge of the

ecological characteristics of at least a part of its

species. Clearly, in groups where ecological stu-

dies are difficult, as well as in paleontology (at

least in the groups which are only known as

fossils and for which comparisons with living

species cannot be done), works of this type are

difficult, if not impossible, to carry out.

In the absence of data, it may be useful to rely

on the study of convergences : thus a character, or

better, a set of characters, liable to appear

independently in several lineages, in animais

facing similar conditions of environment, is

likely to be adaptive.

This criterion leads us to give more weight, in

classification, to characters having a clear adaptive

meaning than to those which do not hâve such a

clear meaning. In the absence of any indication

on its function, one should avoid recognizing a

genus for animais which exhibit a somewhat

spécial morphological character.

The works of Inger (1954, 1958), where the

use of ecological criteria was suggested for better

identification of généra, were not approved unan-

imously by specialists of amphibians, despite

the most convincing examples given by this

author. Some authors hâve accepted these pro¬

posais favorably and hâve sometimes applied

them in their own works (Poynton, 1964, 1976;

Lynch, 1970, 1971; Martin & Watson, 1971;

Dubois, 1975, 1976, 1980 b, 1981b, 1983 c,

1987 b; etc.), while others hâve criticized them.

sometimes severely (Laurent, 1964: 145-146;

1972: 5-7, 26-28; 1973; Kluge, 1966; Duellman,

1977). What seems to hâve escaped some of these

contradictors is the fact that Inger, contrary e.g.

to Illies (1970) (see below), has not in the least

proposed an “ ecological concept of the genus ”

(Laurent, 1972: 26), but has simply proposed to

take into account, when this is possible, the

additional information that ecology may give. It

is quite obvious that no classification can be

based on the ecology alone, because of the vast

body of convergences which punctuate biological

évolution! As for the fact, which is sometimes

invoked, that this criterion cannot always be

used, because it is not always possible to know

the ecology of the species, nothing is more true,

but, as for other criteria discussed above or for

the hybridizability criterion which will be discussed

below, this does not preclude one from using it

when it is possible! At any rate the existence of an

ecological gap between généra generally implies

that of a morphological gap, and the morpholo¬

gical characters involved are generally complex

(Inger, 1958).

As we shall see below, some criteria, and in

particular the new hybridizability criterion, can

be used only to group together species within a

same genus, never to break up a genus. The

ecological criterion, on the other hand, may be

particularly useful, precisely when the data on

hybridization do not contradict it, for separating

into distinct généra species which are morpholo-

gically close: it is the case for example for the

amphibians généra Rana and Amolops (Inger,

1954, 1966), or Bufo and Ansonia (Inger, 1954,

J 958). This criterion is less useful for grouping

species together, since it may be as easy to

demonstrate the existence of important ecologi¬

cal différences between species as it may be

difficult, unless their ecology is particularly well

known, to be certain that there exists no signifi-

cant différences between two species in such or

such domain of their ecology.

Source : MNHN, Paris

THE PROBLEM OF GENETIC SIMILARITY

The history of life on earth is the history of

organisms, of their cells, and of of the appear-

ance, the transmission and the évolution of the

DNA molécule. It is therefore a unitary process,

but not for ail that a continuum: DNA can

usually be exchanged, recombined, only between

organisms which are “ close ” enough. Discontin¬

uités do exist, which allow the subdivision of the

world of living beings.

In this respect, the species is a well-defined

genetic unit: it is a closed gene pool, more or less

protected from other similar pools (see e.g.

Bocquet, Génermont & Lamotte, 1976, 1977,

1980). The genus, which groups together such

units which are disjunct in nature, cannot be

defined in the same way. Is it possible however to

combine in a genus species having “ close ”

génotypes, because they dérivé from a same

ancestor? This will be possible if we hâve a way

of estimating the genetic similarity between

species. Several methods may be thought of for

obtaining such an estimate: we shall examine

them successively.

Genetic “ distances ” and “ similarities ”

To estimate the genetic resemblance between

species, one might first think of making use

of the various measurements of genetic “dis¬

tances ” and “ similarities ” which hâve recently

developed. Most of these methods are based on

direct or indirect comparisons of the molécule of

a given protein such as it exists in several species:

indirect methods are based on electrophoretic or

immunological techniques, while direct methods,

heavier and more costly and therefore still much

less used, are based on the reconstruction of the

primary structure of proteins (for a présentation

of these methods and of the results they permit,

see e.g. Ayala, 1977).

The methods of measurement of genetic dis¬

tance based on proteins are of great interest

because they give us valuable information for the

reconstruction of the phylogeny of a given

group. As a matter of fact recent works indicate

that, at least for many proteins, the appearance

of différences between homologous molécules of

different species deriving from common ances-

tors happens at a relatively constant rate for a

given protein and within a given group. This rate

is a function of the time elapsed since the

séparation of the two lineages, and is indepen-

dent from the rate of morphological évolution,

as well as from the spéciation rate, in these

lineages. These phenomena are still the subject of

a lively discussion among biochemists, but it

seems well established that we now dispose of a

molecular clock of évolution (Zuckerkandl &

Pauling, 1962 ; Wilson, Carlson & White,

1977): provided some methodological précau¬

tions are taken, it is possible, within a given

group, to hâve at least an approximate idea of

the time elapsed since the séparation of two

lineages which hâve led to two living species,

simply by measuring, by one of the methods

evoked above (and in particular those, of more

generalized use, which are based on protein

electrophoreses or on immunological techniques),

Source : MNHN, Paris

ALAIN DUBOIS

the genetic distance between these species. This

genetic distance is of great interest in studies of

phylogeny, inasmuch as it allows the construc¬

tion of hypothèses on the dates of cladogeneses,

and also the détection of certain morphological

convergences (see e.g.: Maxson & Wilson, 1974;

Maxson, 1977).

Such methods allow us therefore to hâve an

idea of the âge of the lineages of which we

nowadays observe the descendants. To classify

the animais according to the similarity of their

proteins would therefore largely corne down to

classifying them according to the greater or

smaller âge of their common ancestor: such a

criterion would be acceptable from a cladistic

point of view, but it is not so for the synthetists,

who aim at expressing in the classification, not

only the properly cladogenetic aspects of évolu¬

tion, but also its anagenetic aspects, and there¬

fore at taking into account the greater or lower

holomorphological divergence which may hâve

occurred in the various lineages after the clado¬

geneses which hâve separated them.

But would it not be possible to obtain a more

trustworthy measure of the genetic similarity of

organisms by directly comparing their DNAs?

Such direct comparisons, on a large scale, raise

of course important technical problems, but it is

now possible to tackle this problem through the

study of DNA hybridization: the quantitative

measurement of the success of this hybridization

indicates the degree of similarity of the chains

compared (see e.g.: Hoyer, McCarthy & Bol-

ton, 1964; Ayala, 1977; Sibley & Ahlquist,

1982; Diamond, 1983). Furthermore it seems

that, in a rather close future, the direct compari-

son of the structure of portions of DNA chains

will be possible, thanks to recent methods which

allow drawing up gene sequences (Abelson,

1980). The data currently available on compari¬

sons of DNA in different species are still not

numerous, but they seem to indicate that the

measurements thus obtained are much better

correlated with those derived from the compari¬

sons of proteins, therefore with the time elapsed

since the séparation of lineages, than with the

holomorphological divergence which has occurred

during this period between the lineages in ques¬

tion (Wilson, Maxson & Sarich, 1974; Sibley

& Ahlquist, 1982). The DNAs of two given

species are however appreciably more different

than their proteins, probably because of the

redundancy of the genetic code or of the existence

of différences in the non transcribed régions of

the DNA chain (King & Wilson, 1975). Despite

these différences, we are more or less brought

back to the preceding problem.

It is interesting to note that even before the

existence of methods allowing the comparison of

the DNAs of different species, some authors had

had the perceptiveness to foresee that the knowl¬

edge, even complété, of the structure of the DNA

of species would be, although certainly very

useful, insufficient in itself to build up a classifi¬

cation of animais, contrary to what other authors

believed then (e.g. Sibley, 1962) or still believe

(e.g. Sibley & Ahlquist, 1982). Thus Simpson

(1962: 502) wrote:

“Sibley (1962), agreeing with some others, has suggested that the

ultimate (i.e., the touchstone?) for classification would be the complété

DNA code. Certainly I can think of nothing more désirable as an addition to

our criteria for classification, but I strongly doubt whether even that most

désirable of criteria would be sufficient in itself. At high taxonomie levels,

particularly, I suspect that interprétation of DNA resemblances and

différences would be as difficult as interprétation of anatomical resemblances

and différences and that the two would hâve to be combined, with each

other and with ail other classes of data, for the soundest resuit. ”

Sibley (1965: 120) then implicitly admitted the

validity of this criticism by writing:

“ Finally, let it be clearly understood that the application of the methods

of molecular biology to systematics does not insure the solution of ail of our

problems. The new techniques provide new kinds of calipers which can

measure previously unavailable characters but the interprétation of the data

still requires a systematist who knows, appréciâtes and understands the

other available information about the group of organisms he is studying.

The molecular data are enormously exciting, and hold great promise for

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

future discoveries, but they must be viewed as additions to, not substitutes

for, what is already known about the genetic relationships and evolutionary

history of plants and animais. ”

In the same spirit, Moore (1967: 466-467) between two species was insufficient to measure

underlined that the number of different alleles the genetic divergence between them:

“ One has to know the conséquences of the génie action. A few allelic

différences might resuit in two profoundly different populations; a greater

number of allelic différences might resuit in only trivial différences.

There is no way at the moment to compare the complété génotypes of

different individuals, but this may soon become possible with the improve-

ment of techniques for the hybridization of DNÂ’s of diverse origins. This

method will be subject to the important restriction mentioned above, namely

that ail genes are not equal in their efifects. ”

For his part, Mayr (1970: 321-322) wrote:

“ Indeed, it is becoming increasingly évident that an approach that merely

counts the number of gene différences is meaningless, if not misleading.

(...) Nor can species différence be expressed in terms of the genetic bits of

information, the nucléotide pairs of the DNA. That would be quite as

absurd as trying to express the différence between the Bible and Dante’s

Divina Commedia in terms of the différence in the frequency of the letters of

the alphabet used in the two works. The meaningful level of intégration is

well above that of the basic code of information, the nucléotide pairs. ”

Lewontin (1974: 20) expressed similar ideas in

different words:

“ To concentrate only on genetic change, without attempting to relate it

to the kinds of physiological, morphogenetic, and behavioral évolution that

are manifest in the fossil record and in the diversity of extant organisms and

communities, is to forget entirely what it is we are trying to explain in the

first place

Finally Stanley (1979: 56) also defended this

viewpoint:

" Genomic components hâve significance only in terms of phenotypic

expression. A bear probably has been transformed into a panda by a few

genetic alterations, but the resuit is an enormous amount of adaptative

change, not a little. The notion that rates of évolution ideally should be

measured by genomic rather than by morphological parameters (...) excludes

from considération the phenotype, upon which sélection opérâtes. We desire

to understand the genetic mechanism of major evolutionary transformations

of the sort that occurred in the origin of the giant panda, but the kinds of

genetic information to be sought can be gleaned only through study of

phenotypic change. ”

If I deemed necessary to produce these various

quotes, it is because, despite these few stands,

many authors are still not conscious of these

problems, and one still much too often finds

publications where the “ genetic distance ” based

on proteins is considered a good measure of the

overall genetic resemblance between the species

compared, which is obviously wrong.

The large discrepancy which exists between

the morphological and molecular resemblances

between species (Wilson, Carlson & White,

1977; Cherry, Case & Wilson, 1978) has been

largely realized only in the last years. It poses

interesting problems which we shall discuss again

below, but let us note from now on that there

would be no question of attributing automat-

Source : MNHN, Paris

ALAIN DUBOIS

ically a given systematic rank to groups of

species having between them a given divergence

at the molecular level, as it has been contem-

plated and even put into practice by some

authors (e.g.: Wallace, King & Wilson, 1973;

Maxson & Wilson, 1975; Lanza, Cei & Crespo,

1976; Maxson, 1976; Sibley & Ahlquist, 1982;

see Diamond, 1983, for a sériés of référencés on

“ taxinomy by nucléotides ”): this would elimi-

nate a whole aspect of évolution, morphological,

ecological and other divergences, which may be

more or less important and more or less rapid

between genealogically closely related species.

The fact that human polypeptidic chains are

more than 99 % identical to those of chimpanzee

(King & Wilson, 1975), which corresponds to

the différence which exists, in other groups of

organisms, between dualspecies (see Bernardi,

1980, for the use of this term rather than that of

“ sibling species ”), is of great interest for it

expresses the fact that hominid évolution has

been particularly rapid. It does not imply at ail,

however, the need for abandoning classifying

man and chimpanzee in two distinct families.

Structural genes and regulatory genes

Despite their high technical foundations, the

measures of genetic similarity that we discussed

above give us only a static and distorted idea of

the resemblances between two génotypes. As a

matter of fact, they inform us about the purely

structural resemblances between these génotypes,

but scarcely on their functional resemblances.

The génotype is not a sum of genes simply placed

side by side. It is an integrated whole of genes

which interact together (Mayr, 1975, 1982 b).

The fundamental biological properties of an

organism are the resuit of these interactions, and

not an addition of isolated génie activities.

Therefore if we want to estimate, not only the

phylogenetic kinship of the species as allowed by

the molecular methods mentioned above, but the

whole similarity of génotypes, as functional units,

we will need a weighted method of measurement,

which takes into account the fact that ail genes

do not play the same rôle, do not hâve the same

importance, in the building up and the function-

ing of an organism, in other words a method

based on properly biological criteria, and not

only biochemical or molecular ones.

The first method which cornes to mind in this

respect is that of phenetic comparisons. As a

matter of fact it is certain that, up to a certain

point, morphological resemblance expresses in a

synthetic way the similarity of génotypes, and

that a classification based on phenetic compari¬

sons alone is generally a very good starting point

for any synthetic classification (Mayr, 1981).

However, as we hâve seen, this resemblance may

be misleading (convergence, parallelism), and on

the other hand it may not be at ail correlated

with molecular data, i.e. with a measure which is

apparently more précisé, finer, of the genetic

characteristics of the forms compared. How can

we solve this contradiction?

We must here turn to recent developments of

genetics, which point to the long ignored impor¬

tance of regulatory genes, both during the onto-

genesis of a given individual and during the

évolution of living beings. Without going into

the details, for which I refer the reader to the

reviews by Zuckerkandl (1976 a, 1976 b), Wil¬

son, Carlson & White (1977) and Raff &

Kaufman (1983), I shall mention some results of

this research which are particularly important as

regards the problem that concerns us here.

Let us first recall the fundamental distinction

between two types of genes:

— structural genes code for messenger RNAs

which are themselves translated into polypeptids;

a mutation in one of these genes entails often

(but not always, because of the degenerated

nature of the genetic code) a modification of the

primary structure of the polypeptid synthesized;

such modifications are detected by the methods

of comparison of proteins mentioned above;

— the rôle of regulatory genes is the régula¬

tion of the expression of structural genes: accord-

ing to the cell and to the moment, they induce

certain genes to be active, they inhibit others,

etc.; a mutation in one of these genes may induce

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

a modification of the activity of many other

genes.

The distinction between structural and regula-

tory genes was first established experimentally by

Jacob & Monod (1961) in Bacteria. Later

on, the existence of regulatory genes was also

demonstrated in Eucaryotes: these genes are

situated in the “ noncoding ” parts of the DNA,

which are of several types and occupy an

important proportion of the DNA of Eucaryotes

(see e.g.: Walker, 1979; Raff & Kaufman,

1983). The définition of regulatory genes given

above may appear vague, but it is difficult at the

moment to be more précisé:

“ Considering the complexity of régulation and how little we know of its

details, it may be foolish to even attempt to define a regulatory gene.

Nevertheless, it is important to hâve at least a working définition if we are to

study the évolution of regulatory genes and their rôle in adaptation. Thus,

let us define a regulatory gene as any gene that directly affects the amount,

the tissue distribution, or the developmental profile of another gene product.

This working définition tells us, if nothing else, what kinds of phenotypes

might resuit from genetic changes at regulatory loci, namely (a) différences

in the levels of a structural gene product in some or ail of the tissues of an

organism (quantity variants), (b) différences in the presence or absence of

the structural gene product in different tissues of the organism (tissue

variants), and (c) différences in the time of appearance during development

of the structural gene product (temporal variants). It should be clear that

these need not be mutually exclusive categories of regulatory gene variants.

In fact, in most cases (see below) there is substantial overlap in the

phenotypes, e.g., a particular strain may hâve an elevated level of a

structural gene product because that gene becomes active earlier in

development. ” (Macintyre, 1982: 265-266).

The processes of genetic régulation in Euca¬

ryotes are obviously very complex, and if models

hâve already been proposed in an attempt to

describe them (Zuckerkandl, 1964, 1976 a; Brit-

ten & Davidson, 1969; Davidson & Britten,

1973; Whitt, Philipp & Childers , 1977; Raff

& Kaufman, 1983; etc.), those remain probably

still oversimplifications of the reality. Despite the

still very preliminary stage of research on genetic

régulation in Eucaryotes, it is interesting to look

into the results already obtained.

The first important resuit of recent research is

the discovery that the évolution of regulatory

genes and that of structural genes are largely

independent from one another. As we hâve seen,

the évolution of structural genes is relatively

regular: for a given gene and within a given

systematic group, it seems to be simply propor-

tional to time or almost so. On the other hand,

the évolution of regulatory genes does not at ail

seem to exhibit the relative regularity of that of

structural genes. It seems that in certain groups

(the “ living fossils ”) these genes hâve not

evolved during hundreds of millions of years —

or hardly at ail — while in others (e.g. hominids)

they evolved very quickly.

The disparity in the évolution rates of both

types of genes was well put in evidence in

vertebrates by Allan C. Wilson’s team in Berke¬

ley. Thus the évolution rate of albumin seems to

hâve been roughly identical in anuran amphib-

ians and placental mammals, while morpholo-

gical évolution has been slow and weak in the

first group and rapid and important in the

second (Wallace, Maxson & Wilson, 1971;

Wallace, King & Wilson, 1973; Wilson, Sarich

& Maxson, 1974; Maxson & Wilson, 1975;

Cherry, Case & Wilson, 1978; Cherry et al.,

1979, 1982). According to these authors, the

évolution of phenotypes would be correlated

more to the évolution of the Systems of genetic

régulation than to that of structural genes. This

would explain the fact that morphological diver¬

gence is largely independent from molecular

divergence measured at the level of structural

genes of proteins. Both phenomena express

different aspects of the génotype and, from the

viewpoint of évolution, morphological diver¬

gence is much more important a synthetic indica-

tor than divergence at the level of protein

molécules. The latter may be a simple, more or

less exact, function of time, while morphological

Source : MNHN, Paris

ALAIN DUBOIS

modifications take place at variable speeds and

are linked to the history of the group, and in

particular to the history of its adaptations.

It is therefore not unreasonable to think that it

is the évolution of the Systems of genetic régula¬

tion which accounts for the major part of the

morphological évolution of organisms. New types

of structures, of organs or of organisms may

appear following modifications in the Systems of

genetic régulation, but often without the appear-

ance of noticeable différences at the level of

structural genes.

As we hâve seen, polypeptidic chains and

DNAs of man and chimpanzee are more than

99 % identical (King & Wilson, 1975), although

these two species are extremely different in their

morphology, development, behavioral capacities,

etc. The différences between these two species

would be accounted for by différences in their

Systems of genetic régulation.

Thus, the results of recent research on the

molecular aspects of évolution throw new light

on the already ancient conceptions, which may

be found e.g. in the works by J. S. Huxley,

G. G. Simpson or E. Mayr, on the existence of

different rates of anagenesis at different epochs

and in different lineages: these different rates of

anagenesis could correspond to different rates of

évolution of regulatory genes.

The construction of a weighed, synthetic index

of genetic similarity is therefore not an easy task.

Quite rigorously, such an index should take into

account the five following distinct types of

similarity (Zuckerkandl, 1980): (1) structural

similarity of the genes (functional units of the

DNA) and of the DNA as a whole; (2) functional

similarity of the direct (RNA) and indirect

(proteins) products of the genes; (3) similarity of

the types of interactions between genes, in differ¬

ent tissues and at different moments; (4) simi¬

larity of the quantitative aspects of these interac¬

tions; (5) similarity of the results, at the various

supramolecular levels, of these interactions.

These factors are still far from having ail been

analyzed and, furthermore, we still do not hâve a

method which would allow an intégration of the

results of the analyses made at these different

levels so as to obtain a single global index of

similarity. A rigorous synthetic measure of the

genetic similarity of two organisms is therefore

impossible at the moment, but some éléments do

exist for such a measure.

First of ail, as we hâve seen, biochemical

techniques allow us now to obtain a direct or

indirect measure of the structural similarity of

génotypes, (1) above: percentage of genes, or

even of triplets, that two organisms hâve in

common (which are identical in both). This

similarity is usually indirectly estimated by the

“ genetic distance

On the other hand, the similarity of results,

(5) above, is largely measured by phenetic

methods. The analysis must bear on the holo-

morph, and not be limited to the morphology

alone.

But these data are still insufficient: they do not

allow one to measure the functional similarity of

génotypes (similarity of the Systems of genetic

régulation which govern the expression of the

structural genes, (2) to (4) above).

In particular it must be underlined that ail

genes are not active in the adult and that a

measure of genetic similarity between species

which would only be based on the characters of

adults would be largely biased. During ontogeny

(embryogenesis and growth), many genes, which

were inactive in the egg and which will be so

again in the adult, become successively active. In

other words, during development, modifications

occur in the types of interactions between genes

and the quantitative aspects of these interactions,

(3) and (4) above, and these modifications are

important to take into account in any analysis of

genetic similarity.

The species with castes, the animais with

métamorphosés and even more so the parasites

with cycles, give a good illustration of the

various potentialities of a génotype which are

revealed only successively or alternatively. Thus

for example, in the Digeneous Trematoda, mira¬

cidium, sporocyst, redia, cercaria, metacercaria,

adult, are successive stages which make évident

the importance of the rôle of the Systems of

régulation in the expression or the inhibition, at

various stages of the life of an animal, of the

various morphogenetic potentialities présent in

its génotype.

Thus in order to be able to take into account,

in the classification of animais, the genetic

resemblance between species, one must not con¬

tent oneself with the “ classical ” methods of

measurement of “ genetic ” distances, but one

must also possess a method of estimation of the

functional similarity of génotypes. Until now, a

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

single, indirect, method, has largely been used by

systematists in this aim: the measure of overall

morphological resemblance between species. I

suggest here that a second synthetic method of

measurement of the functional genetic resem¬

blance, which until now has been only very little

used by systematists, could be taken advantage

of: it is the study of a natural or artificial

experiment, that of interspecific hybridization.

Hybridization and genetic similarity

The hybridization of individuals (and not of

their cells, of their DNAs or other molécules) is a

synthetic indicator which has often been neglected

but which may be very useful to apprécia te the

degree of biological (and not biochemical) compat-

ibility of two genomes, in an organism and to

build an organism, and to measure their degree of

functional (and not only structural) resemblance.

The fact that two genomes may “ agréé together ”

and succeed, together, in “ building up ” an

organism, indicates that, not only their structural

genes are similar or at least compatible, but also

that their Systems of genetic régulation are

compatible (Wilson, Maxson & Sarich, 1974;

Oliver, 1979). This is an indication of a very

high functional genetic similarity.

What is particularly significant in this criterion

is that it appeals to the ontogeny of the orga¬

nism, and therefore that it takes into account,

among others, ail the genes which become active

only during a period, sometimes very short, of

development, and are inactive in the adult.

The différence between the hybridization of

individuals and that of cells must be underlined.

Cell hybridization consists in the putting together,

in a cell culture, within a single cell, chromo¬

somes of two different species. Such cell hybrids

may be obtained between extremely distant

species, such for example as man and the

mosquito Aedes aegypti (Zepp et al., 1971). These

hybrid cells live and divide during several généra¬

tions, for they are not obliged to develop and

build up an organism. It is likely that few genes

are active in these cells as compared with the

number of genes which take part in the whole

ontogeny of an organism.

On the other hand the hybridization of indi¬

viduals generally succeeds only between species

which are considered, after ail other biological

criteria, as relatively close or very close to each

other. From the zygote until the adult, these

hybrid organisms are able to activate, one after

another and in a coordinate way, their various

genes and Systems of genetic régulation without

provoking a lethality. This is a very strong

indication that both species are closely related

and hâve a high overall genetic similarity.

Classical hybridization (Crossing of two indi¬

viduals and obtaining of one or several hybrid

zygotes) is not the only method allowing a

measurement of this genetic compatibility be¬

tween species. Other more recent methods should

produce interesting data in this field. The most

important of these methods, discussed in detail

elsewhere (Baltzer, 1952; Fankhauser, 1955;

Moore, 1955; Brachet, 1957; Briggs & King,

1959; Chen, 1967; C. L. Gallien, 1970; L.

Gallien, 1972; Subtelny, 1974; Danielli &

Diberardino, 1979; Diberardino, 1980), are the

following ones:

— production of haploid hybrids: after fertii-

ization of the ovum of species A by a spermato-

zoon of species B, and before amphimixy, the

female pronucleus is retired from the egg; the egg

develops then with a cytoplasm A and a single

set of chromosomes B; in amphibians, Moore

(1967) has shown that this System is more

sensitive than normal hybridization to detect

genetic incompatibilities between species;

— doubling of the paternal stock of chromo¬

somes in the egg of the previous experiment: this

egg then develops with a double stock of chro¬

mosomes B in a cytoplam A;

— production of polyploid hybrids, e.g. by

fertilizing a diploid ovum A with a spermato-

zoon B; various other combinations are also

possible, which allow a fine study of the develop-

mental conséquences of the presence of different

doses of chromosomes of the two parental

species in the egg;

— nuclear transplantation: a diploid nucléus

Source : MNHN, Paris

ALAIN DUBOIS

of a species B is introduced in an enucleated egg

of species A;

— injection of cytoplasm of a species B into

the egg of a species A or into a hybrid egg: it is

then possible to study the effect on the develop¬

ment of different cytoplasmic, and not chromo¬

somal, doses (Ansevin & Williams, 1974; Aimar

& Delarue, 1976; Aimar, 1977; Delarue, 1977 a,

1977 b; Aimar, Delarue & Vilain, 1981);

— graft of léthal hybrid tissues on viable

embryos: this allows one to détermine if the

lethality of these hybrids is due to factors présent

in ail the tissue of the hybrid, or on the contrary

présent in certain tissues only, from which for

example toxic substances may diffuse in the

whole embryo and provoke its death;

— in vitro culture of hybrid tissues or cells.

Ail these techniques are extremely interesting.

Unfortunately they hâve only been used until

now on a small scale in experimental works of

developmental biology bearing on nucleo-cyto-

plasmic relationships and realized in limited

groups of animais, mainly amphibians. They

cannot therefore be the subject of generalized use

in zoology as yet, and in what follows we shall

only consider the results of classical hybridiza¬

tion. However, in order to interpret these results

correctly and to be able to use them in system-

atics, we must first recall the different types of

phenomena which hâve been observed during

animal interspecific hybridization; this question

is the subject of the next chapter.

Source : MNHN, Paris

A FEW GENERAL FACTS ABOUT ANIMAL HYBRIDIZATION

The mechanisms of interspecific isolation

Species are protected genetic pools, which

keep their genetic identity thanks to the existence

of factors or mechanisms of interspecific isola¬

tion. As a matter of fact it is important to

distinguish between biological mechanisms of

isolation proper, and exogeneous factors of isola¬

tion, for the latter are not determined by the

génotype of the species but by external con¬

ditions, extraneous to the species themselves.

As for the mechanisms of isolation, they are of

two types: pre-ejaculatory and post-ejaculatory

mechanisms.

Other names hâve been given to the last two

categories, but they seem to me inadéquate for

the following reasons: the words pre-copulatory

and post-copulatory do not apply to animais in

which there exists no copulation in the strict

sense of the term (Dubois, 1977 b); the words

prereproductive and postreproductive, which are

too imprécise, and the words prezygotic and

postzygotic, which are more précisé, are not

adéquate to designate these categories, because

gametes of the individuals of both sexes may be

emitted without this being followed by the

formation of zygote(s). What is particularly

important, in natural populations, is that male

or female gametes be emitted or not, because if

hybridization fails these gametes will hâve been

wasted in vain; natural sélection will thus act in

the sense of reinforcing the mechanisms of

isolation which intervene before the émission of

gametes (Mecham, 1961; Watson & Martin,

1968; Littlejohn, 1969; Dubois, 1983 a). I pro¬

pose the use of the term “ éjaculation ” as a

general term designating the émission of gametes

of both sexes, not only of male gametes: this is

conform to the etymology, the Latin verb ejacu-

lare meaning “ to project with strength a liquid

secreted by the organism ”, not necessarily sperm.

Furthermore, in many animal species, both types

of gametes are emitted at the same time, at a

given moment of the mating act, which may be

designated as “ ejaculatory ”. In the classifica¬

tion of isolation mechanisms which follows, the

major dichotomy is placed between pre- and

post-ejaculatory mechanisms, rather than between

pre- and post-zygotic mechanisms, which also

exist but as subdivisions of the post-ejaculatory

category.

Exogeneous factors of isolation

Geographical barriers

The individuals of two allopatric populations

never hâve an occasion to meet, and hence there

exists no material possibility of hybridization

between them.

Temporal barriers

Similarly, individuals belonging to two allo-

chronic populations hâve no possibility to hybrid-

ize: it is the case of fossil species from different

periods of time, or in relation to contempora-

neous species, but the scale of time considered

may also be very short.

Source : MNHN, Paris

ALAIN DUBOIS

BlOLOGICAL MECHANISMS OF ISOLATION

Pre-ejaculatory mechanisms

They prevent mating and/or the émission of

gametes of individuals of both species.

Ecological mechanisms

They may be spatial (both species having

different habitats or reproduction sites), temporal

(both species having different periods of activity

or of reproduction, be it different seasons or

different periods in the nycthemerous), or both.

In this case, animais of both species do not meet,

at least not during their reproduction.

Behavioural mechanisms

The animais of both species may meet, but not

breed together, for one of the following reasons:

— total absence of attractiveness of the ani¬

mais of both species one for another;

— the animais of one sex (generally the males)

of one species are attracted by those of the other

sex of the other species, but the later repuise

them and refuse to mate;

— the animais of both sexes are attracted, the

preliminaries to breeding start, but do not go to

their end because of the existence of too impor¬

tant différences between the nuptial parades or

other characteristics (morphology, colors, etc.)

of both species.

Among the behavioural mechanisms of isola¬

tion, the following ones may be cited, according

to the physical sense on which they call:

— those which involve hearing: specificity of

the mating calls;

— those which involve sight: specificity of the

colors, of the shape, of the size and of other

visible morphological characters; specificity of

the behaviours in sexual parade;

— those which involve touch: specificity of

the shape, of the size, of the texture, of the

movement;

— those which involve Chemical senses: speci¬

ficity of the sexual pheromones or of other

Chemical stimuli.

Mechanical mechanisms

In this case mating begins but does not lead to

the émission of gametes, for the copulatory

organs of both species are not compatible and do

not allow a complété copulation. Size in itself

may be a factor of mechanical isolation: a male

and a female with too dissimilar sizes do not

always succeed in mating, even in the absence of

behavioral mechanisms of isolation between them.

Post-ejaculatory mechanisms

These mechanisms act after the émission of

gametes. The classification which follows is

based on the results of the numerous works

devoted to interspecific hybridization in various

groups of animais, and in particular of echino-

derms, insects, teleosts and amphibians (Monta-

lenti, 1938; Moore, 1955; Stebbins, 1958; Cou¬

sin, 1967; Blair, 1972 b; etc.).

Prezygotic mechanisms

In this case the émission of gametes of both

sexes, or at least of male gametes, takes place,

but these die before coming into contact, or

corne into contact but without leading to a true

fertilization. Several distinct situations may be

met with in this category:

— Absence of attraction of sperm for ova:

this case may occur in aquatic animais in which

gametes are directly emitted in water without

copulation.

— Destruction or immobilization of sperm in

the female génital tract, in particular following

an antigénie reaction of female towards them.

— Inability of sperm to go through the jelly

of the ovum, either because they are stopped or

killed by the jelly, or because in the later some

factors are lacking which are essential for the

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

pénétration of the spermatozoon in the ovum

itself (see Elinson, 1974).

— The spermatozoon does not enter the

ovum but leads to its activation, and sometimes

to its development by gynogenesis; this may give

birth to false hybrids which are usually haploid

and sometimes diploid (in the case where the

ovum was diploid or when a doubling of the

maternai stock of chromosomes has occurred

after activation and before the first division of

the egg).

— The spermatozoon enters the egg but amphi-

mixy does not occur. Paternal chromatin does

not differentiate in chromosomes, it degenerates

and is later destroyed or expulsed from the egg,

which may however sometimes develop and give

birth to a false hybrid.

Poslzygotic mechanisms

These mechanisms only act when there exists a

true zygote, i.e when fertilization has been

complété and when amphimixy (karyogamy) has

taken place between the male and female pronu-

clei. In the hybridization experiments which hâve

been realized in the whole animal kingdom, ail

intermediates hâve been observed between amphi¬

mixy followed by no development and obtention

of adult hybrids, fertile and normal in ail:

— Amphimixy followed by no development.

— Amphimixy followed by the subséquent,

early or late, total or partial, élimination of

paternal chromatin. Here again, development

may take place with the maternai stock of

chromosomes alone (which gives birth to an-

other type of false-hybrid), or with the maternai

stock and a part of the paternal stock (which

gives birth to an aneuploid hybrid or “ partial

hybrid ”).

— Amphimixy followed by the development

of the diploid hybrid zygote. In amphibians, the

following different cases can further be distin-

guished:

• Development till the end of the blastula

stage, the embryo proving unable to realize gas¬

trulation.

• Arrest of development during a subséquent

embryonic stage: gastrulation, neurulation, tail

bud stage.

• Hatching takes place, but the larvae are

abnormal (oedemas, microcephaly, etc.), do not

feed and die.

• The larvae are apparently normal, they feed

and grow but die after a certain time.

• Inability to get over the hurdle of metamor-

phosis.

• Metamorphosis takes place but the young

amphibians are abnormal and die very soon.

• In the other cases, the animais which hâve

gone through metamorphosis generally reach the

adult stage. Several cases are still possible:

* The adults are viable but présent various

somatic anomalies which may interfère with their

survival or their reproduction.

* The adults of both sexes are unequally

represented (inbalance of the sex-ratio), and

sometimes even one sex may be completely

absent.

* The adults of both sexes are not fertile.

* The adults of one sex are not fertile.

* The adults are fertile but their F2 or

backcross progeny exhibits some of the anoma¬

lies described above for the Fl: arrest of develop¬

ment at a given stage, anomalies, infertility.

* The adults are viable and fertile but

show a repression of their génie activity at

certain loci.

* Finally, the adults are viable, fertile and

fully normal, and their progeny itself is normal.

A few other types of post-ejaculatory postzy-

gotic mechanisms of isolation also exist, which

do not involve mechanisms internai to the zygote

or to the embryo, but factors external to it. Two

of them may be mentioned:

— The hybrid embryo may die at the stage of

hatching, e.g. because it is unable to get out of

the jelly of the egg (Elinson, 1974).

— In viviparous animais, in particular mam-

mals, the mother may develop an antigenic

reaction against its embryos and synthesize anti-

bodies against them, which cause abortion (see

e.g. Medawar, 1953; Billington, James &

Kirby, 1968; Kerr, 1968; Clarke & Hethe-

rington, 1972; Maxson, Sarich & Wilson,

1973; Wilson, Maxson & Sarich, 1974; Gutt-

man, 1985).

Source : MNHN, Paris

ALAIN DUBOIS

SOME GENERAL RULES DRAWN FROM THE STUDY OF HYBRIDS

The study of artificial hybridization in ani¬

mais, in particular in insects, echinoderms and

amphibians, has allowed a certain number of

general rules to be drawn (Montalenti, 1938;

Moore, 1955; Stebbins, 1958; Cousin, 1967;

Blair, 1972 b; etc.). We will only mention here a

few of them, those which hâve a particular

interest for the systematist who desires to take

advantage of hybridization facts for the estab¬

lishment of a supraspecific classification of ani¬

mais.

Variability of results within a given type of cross

In a same type of cross (e.g. between the male

of a species A and the female of a species B), it is

frequent that the results of different experimental

sériés show between them significant différences.

Depending on the cases, these différences may be

attributed to geographical variations in the gene-

tic characteristics of the species crossed (animais

of a single species but coming from distant

régions of the distribution area of the species, or

even from different populations in a same région),

or simply to individual genetic différences (differ¬

ent animais from a same population). Such a

variability can also be found in the vast majority

of biological phenomena and has therefore nothing

to surprise us, but it must be taken into account

in the interprétation of results: it is very impor¬

tant, in this domain like in ail others in biology,

not to content oneself with a unique observation,

but to multiply as much as possible the observa¬

tions, therefore here the crosses, of the same

type.

In natural and artificial crosses there exist

numerous possibilities of an exogenous disrup-

tion of fertilization and of the development of

hybrids. In many cases the failure of develop¬

ment of hybrids is due to such artefacts. It is one

of the reasons why négative results of hybridiza¬

tion must always be accepted with some réserva¬

tion.

Blair (1972 b) and his coworkers carried out a

very wide survey of interspecific hybridization

within a cosmopolitan genus of anuran amphib¬

ians, the genus Bufo (true toads). They made

hundreds of crosses and reared thousands of

hybrids. They were thus able to observe this

variability of results for a single type of cross.

Blair (1972 b) considers that an important part

of the failures of development may be attributed

to the difficultés of rearing, to diseases and other

artefacts. For this reason, for a given type of

cross the resuit which Blair (1972 b) takes into

account for subséquent analyses is not the mean

of the observed results, nor even the extremes,

but the best resuit observed.

In reality, the variability of results is not only

due to the artefacts introduced by man, but also

to the genetic variability proper to species,

populations and individuals; it is however diffi-

cult, not to say impossible, to take both kinds of

variability into account. Furthermore, as we

shall see, the failure of development may be

caused by very simple and little significant

genetic factors (a single gene may be enough),

and has much less genetic, phylogenetic and

taxinomie significance than the success of devel¬

opment of a hybrid. It is therefore justified to

generalize Blair’s (1972 b) method and to con-

sider in what follows, for a given cross, only the

best resuit observed.

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

Study of reciprocal crosses

One might think that the major if not unique

factor which is responsible for the failure of

development of hybrids is the fact that the

patemal and maternai genomes do not succeed

in working correctly together and to induce a

normal development of the embryo. To be sure,

the phenomenon exists, but another phenome-

non plays an important part, from the start of

development, in many cases of léthal hybridiza¬

tion: it is the incompatibility between the pater-

nal chromosomes (which are brought by the

nucléus of the spermatozoon) and the cytoplasm

of the ovum.

The cytoplasm of the ovum has been synthe-

sized during oogenesis, i.e. under the control of

maternai genes alone. It contains in particular

messenger RNAs which play an important rôle

in the first stages of development. The patemal

genes, or some of them, often prove incompat¬

ible with this cytoplam, which leads to a failure

of development. This phenomenon has been well

demonstrated by various methods, and it has

been possible to show that the incompatibility

between the ovum’s cytoplasm and the patemal

genes could be of several types (see in particular

Stebbins, 1958). Without going into details, let

us emphasize that the importance of this phe¬

nomenon appears very clearly in particular in the

case of reciprocal crosses.

As a matter of fact, when two species A and B

are crossed, the resuit is frequently different

according to whether the cross has been made in

the sense female A by male B or female B by

male A. Sometimes only one of the two crosses

gives birth to viable hybrids, sometimes only one

gives birth to fertile animais, sometimes both

give birth to animais the development of which

stops at different stages, etc.

In both types of crosses the genetic material

présent in the zygote is the same: one set of

chromosomes (and genes) A and one set of

chromosomes (and genes) B. What differs is only

the cytoplasm containing these chromosomes.

The experiments show that the same hybrid

génotype AB may be able to give a normal

development in cytoplasm A, but unable to do so

in a cytoplasm B (for the detailed analysis of a

case of this type, see Elinson, 1981).

As we hâve seen, for the systematist the

positive results of hybridization are more mean-

ingful than the négative ones. In the cases where

reciprocal crosses give different results, it will

therefore be indicated again only to retain the

“ best resuit ”, the resuit of the most successful

of the two types of crosses.

The major stages of failure of hybridization

Despite the vast diversity of the stages of

failure of development of hybrids that we men-

tioned above, some of these stages are more

significant and more important than others to

consider, for the developmental arrest occurs

preferentially there. We shall insist here only on

three of them, which are most frequent and

general in the whole animal kingdom.

Arrest of development

at the end of the blastula stage

In a great number of interspecific hybridiza-

tions, in particular in amphibians, development

proceeds normally until the end of the blastula

stage. When one looks closely into it, one may

observe that it very much resembles the develop¬

ment of normal non hybrid eggs of the maternai

species (if it shows any différences as compared

with that of the patemal species). It has been

possible to demonstrate by various methods that

the segmentation of the amphibian egg until the

end of the blastula stage is occurring without

intervention of the egg nucléus, or with a very

limited intervention of it. What is crucial is the

egg cytoplasm: the segmentation is principally or

entirely determined by the génie products tran-

scribed during oogenesis from maternai genes

and présent as messenger RNAs in the cytoplasm

of the ovum.

Source : MNHN, Paris

ALAIN DUBOIS

It is only at the beginning of gastrulation that

the massive activation of the genes of the embryo

nuclei begins, and therefore, in addition to the

maternai genes, of the paternal genes. It is often

at this moment that the incompatibility between

both paternal and maternai genomes, or between

the paternal genome and the maternai cyto-

plasm. expresses itself, and this is why an arrest

of development is observed at this stage in many

cases of animal hybridization, especially in am-

phibians.

The fact that two species may be able to give

viable hybrids until the end of the blastula stage

is therefore of little genetic or phylogenetic

meaning and is of little interest to the system-

atist.

Arrest of development

at an embryonic stage subséquent

to the blastula stage

This arrest may take place during gastrulation,

neurulation, or later. This phenomenon clearly

expresses an incompatibility, either between the

paternal and maternai genomes, or between the

paternal genome and the hybrid cytoplasm, or

still both phenomena together. However, in

detail, the causes and modalities of the inviability

of hybrids may be numerous (see in particular

Stebbins, 1958), and the failure of development

at these stages is therefore difficult to interpret.

This failure may in some cases be due to a

single léthal gene; this is very strikingly shown by

the recent description of the gene Lhr (“ Léthal

hybrid rescue ”) of Drosophila simulons (Wata-

nabe, 1979; Takamura & Watanabe, 1980).

Since the discovery of this latter species (Sturte-

vant, 1919, 1920), it has been known that the

crosses between Drosophila melanogaster females

and Drosophila simulons males only give birth to

female hybrids, while the reciprocal crosses give

only male hybrids, the development of the other

sex being blocked during larval stages. However

the hybrids of both sexes carrying the gene Lhr

are viable in both types of crosses. This gene

constitutes therefore in itself an efficient postzy-

gotic mechanism of isolation between the two

species of Drosophila , but, “ apart from this

gene ”, these two species remain genetically very

close, their genomes being compatible and able

to induce together a normal development:

“ In the évolution of melanogaster and simulons into separate species there

must hâve a stage in which the hybrids of both sexes were viable. Then, at a

later stage the unisexual inviability that now characterizes the hybrids

somehow arose. It is very likely, in view of the results reported in this paper,

that this was a mutation from Lhr to Lhr ' and, if so, then the Lhr gene

represents an evolutionary step backward. This encourages the search for

other mutants which reverse the evolutionary process of reproductive

isolation.” (Watanabe, 1979: 330-331).

If the inviability of hybrids is of unclear

meaning, on the other hand the success of the

development of hybrids expresses without ambi-

guity the absence of major incompatibility between

the two genomes brought face to face and the

cytoplasm, and has therefore a clear meaning,

which we shall discuss again later.

In amphibians, expérience has shown that in

many cases, when the development of the hybrid

has taken place in a harmonious or almost

harmonious manner, until the stage of tailbud,

then it goes on normally: developmental arrest at

the stage of hatching, of larval development, of

metamorphosis and of post-metamorphosis growth

do exist, but they are rarer and probably less

significant.

Infertility of hybrids

In many crosses. Fl hybrids become adult but

prove then infertile, or give birth to non viable,

abnormal or infertile progeny (F2 hybrids or

backcrosses with one of the two parental spe¬

cies).

Here again, we won’t go into details, but let us

mention that the causes of this infertility, as

numerous as they may be, ali fundamentally

express the same phenomenon as the non-viabi-

lity of Fl hybrids: a disequilibrium, an incompat¬

ibility between the two parental genomes and/or

between the paternal genome and the cytoplasm

of the hybrid. Let us remember, however, that

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

two major types of hybrid sterility hâve been

traditionally distinguished, génie (or develop-

mental) sterility, due to an incompatibility between

the genes of the two hybridized species, and

chromosomal sterility, due to the existence of

structural différences between the chromosomes

of the two parents. Let us also remember that in

many cases where the hybrids of a single sex

prove non viable or stérile, they often, but not

always, happen to belong to the heterogametic

sex (Haldane’s 1922 rule). An excellent and

detailed discussion of the problems related to the

sterility of Fl or F2 hybrids and to the non-

viability of F2 hybrids will be found in the work

of Stebbins (1958).

Therefore, the fact that an adult hybrid is

infertile may be due to a number of causes. It is

demonstrated that certain of these causes do not

call upon numerous and complex genetic factors,

and that sometimes, as for the non-viability of

Fl hybrids, a single gene may be involved. In

other cases, this infertility “ simply ” proceeds

from the fact that the species crossed do not hâve

the same degree of ploidy (e.g. in the cross

between a diploid species and a tetraploid species

derived from the first one): despite the very great

similarity of the two species at the génie level,

their hybrid, although perfectly viable until adult

stage, is not fertile.

The infertility of hybrids having such an

unclear meaning, it seems préférable not to take

it into account in a work having taxinomie

préoccupations.

Genic expression in hybrids

Rather recently, various works hâve been

devoted to the study, by protein electrophoreses,

of genic expression in insect and vertebrate

hybrids (see the référencés given by Dubois,

1983 a: 51, and also: Dickinson, 1980 a, 1980 b;

Philipp, Parker & Whitt, 1983; Dickinson,

Rowan & Brennan, 1984; Pasdar, Philipp &

Whitt, 1984; Pasdar et al., 1984; Parker,

Philipp & Whitt, 1985 a, 1985b).

When two parental species hâve different and

electrophoretically détectable alleles, it is pos¬

sible to study in hybrids the expression of alleles

coming from both parents. Several types of

results may be observed:

— synchronous expression of both parental

alleles: this is the most general resuit, which is

obtained, at least for some genes, even in hybrids

made between species deemed relatively distantly

related by systematists, e.g. placed in different

tribes (Whitt, Childers & Cho, 1973);

— delayed expression of one of the alleles

(generally the paternal one), which only becomes

active at a later stage of development than the

other allele;

— absence of expression of one of the alleles,

which is said to be repressed : this is more often

the paternal allele, more rarely the maternai

allele and finally sometimes both alleles of a

given gene.

A rather weak relation exists between the

degree of allelic repression and taxinomie rela-

tionships between the species crossed. Thus in

teleosts, Whitt, Childers & Cho (1973) hâve

observed that there exists usually no allelic

repression between species of a same genus, and

that the number of repressed alleles tends to

increase with the systematic distance between

species (belonging to different généra, different

tribes, etc.). But exceptions to this “ rule ” also

exist (see e.g. Lucotte & Dubouch, 1980).

Source : MNHN, Paris

ALAIN DUBOIS

CONSEQUENCES AS TO THE USE OF HYBRIDIZATION IN SYSTEMATICS

What interests the systematist, in the study of

hybridization, is the information that it may

bring to him as to the resemblance of the genomes

of both hybridized species. The measure of this

genetic resemblance will be applicable to esti-

mate the phylogenetic kinship of these species

and their greater or smaller genetical divergence

since their séparation, and for the construction

of a supraspecific classification.

Given this objective, it will be necessary, on

one hand to try to eliminate as many as possible

of the “ parasitic ” factors which interfère with

hybridization, so as to leave only phenomena

which can actually be explained by the structural

and functional resemblances between the geno¬

mes of the two compared species, and on the

other hand to dispose of methods allowing one

to verify that what has been obtained are really

true diploid hybrids between the two species.

Elimination of “parasitic” factors interfering with hybridization

Various types of methods can be called upon

to try to eliminate these “ parasitic ” factors,

according to the nature of these factors. Several

cases may be distinguished:

— the case where the pre-ejaculatory factors

and mechanisms of isolation between both spe¬

cies in nature are non-existing or imperfect: it

will then be possible to study the eggs, larvae and

adults in the zone where natural hybridization

occurs;

— the case where the species are separated in

nature by geographical barriers: the mere putting

in contact, in captivity or in semi-captivity, of

individuals of the two species separated in nature

will sometimes be enough to obtain hybrids;

the case where the species are separated in

nature by ecological or behavioural mechanisms

of isolation: their placing in captivity may be

enough in certain cases to obtain hybrids (certain

species which do not hybridize, even in sym-

patry, in nature, may do it in a cage or in an

aquarium); in certain cases, an intervention of

man is necessary to erase the behavioural barriers

between both species, and this intervention may

go as far as artificial insémination or fertiliza-

tion;

— the case where mechanical or physiological

barriers prevent fertilization, in particular in

species with internai fertilization: one must then

hâve recourse to artificial insémination or fertil¬

ization, sometimes after taking the ovum from

the female génital tract, in vitro artificial fertiliza¬

tion and reimplantation of the egg in the female;

— the case of other more spécifie barriers; we

will give two examples of these:

. In amphibians, some species never produce

hybrids between themselves although they appear

very close from ail other viewpoints, or they may

produce viable hybrids, which may sometimes

become fertile adults, when the cross is made in

one sense, but in the reciprocal cross the egg is

never fertilized. In some of these cases, it has

been possible to show that the failure of hybrid¬

ization was to be ascribed to the jelly of the egg

of one of the species, which stops the spermato-

zoon of the other one (Elinson, 1974, 1975 a,

1975 b; Brun & Kobel, 1977). When virgin ova

of the species A, taken from the general cavity of

a female of this species before they go through

the oviduct, are placed in the general cavity of a

female of species B which is laying its eggs, these

ova go through the oviduct where they are

covered with the jelly of this species. (This

technique was invented and first applied by

Rostand (1933), and for this reason I proposée!

(Dubois, 1982 a, 1983 a) to call it “ Rostand’s

technique ”). The ova are then liable to be

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

fertilized by sperm of the species B, and in some

of these cases the development may then occur in

totality and give rise to normal adults. Accord-

ing to Elinson (1975 a), the fact that the jelly of

the ova of Rana clamitans prevent the fertiliza-

tion of the latter by Rana catesbeiana sperm

could dépend on two genes only, and the same

probably applies to other similar cases. In ail

these cases, the genomes of the two species

concemed may be very little different, and the

few genes implied in the phenomenon of the

block to fertilization in the jelly are, as concerns

the problem of overall genetic compatibility of

both genomes, artifacts, parasitic factors which

must be disposed of: in a study of overall

similarity between species, it will therefore be

indicated in these cases to call upon Rostand’s

technique.

. In viviparous animais, the mother may develop

an antigenic reaction against the hybrid foetus,

which leads to abortion. This process is still

imperfectly understood but here also such a

reaction could well dépend upon a small number

of genes. It could be possible to avoid this

artifact by making an in vitro culture of the

hybrid embryo in order to follow its develop¬

ment in the absence of antibodies produced by

the mother (see in this respect: Maxson, Sarich

& Wilson, 1973; Wilson, Maxson & Sarich,

1974).

Despite their diversity, ail the interventions

that we hâve just mentioned are, with respect to

the problem which concerns us here (resem-

blance of two genomes and their functional

compatibility), of the same nature: they tend to

suppress the factors which oppose the meeting of

gametes of the two species considered, as well as

ail exogenous factors susceptible to oppose the

development of the hybrid zygote.

DETECTION OF TRUE DIPLOID HYBRIDS

This being done, an additional précaution

must be taken: it is necessary to ascertain that

the animais produced by a cross are true diploid

hybrids and not haploid or diploid “ false-

hybrids ” (eggs developed by gynogenesis or

having expulsed the paternal chromatin), aneu-

ploid “ partial hybrids ” (a part of the paternal

stock of chromosomes having been eliminated at

the beginning of development), or even triploid

hybrids (having e.g. two maternai and one

paternal chromosomes sets). Such anomalies

hâve been observed by Bogart (1972) among the

numerous products of the crosses realized by

Blair (1972 b) and his coworkers in the genus

Bufo: in this case, the major cause seems to be

that Bufo females often produce a low percent-

age of diploid ova, which may develop by

gynogenesis or be fertilized and give rise to

triploid embryos, but other mechanisms may be

responsible for similar anomalies in other cases.

As was stressed by Bogart (1972), there is

good reason for being particularly vigilant when

the number of viable hybrids obtained is low, for

example when, in a cross of amphibians implying

hundreds or thousands of eggs, the number of

eggs giving birth to larvae is very reduced as

compared to the number of fertilized eggs: these

larvae often prove to be gynogenetic or triploid.

However they exhibit a normal morphology and

cannot be detected as such by the sole examina¬

tion of their phenotype.

Because of these problems, vérification would

be indicated in ail cases where the progeny

obtained is composed of real diploid hybrids by

doing the following analyses:

— karyotype of the hybrid, allowing one to

ascertain that it is diploid (or, more precisely,

that it has a number of chromosomes equal to

the sum of the haploid chromosomal numbers of

the two hybridized species, which may be differ¬

ent);

— examination of the external phenotype

(morphology, colors), permitting the discovery in

some cases of the existence of a mixture of

paternal and maternai characters;

— when this examination gives doubtful results,

recourse to electrophoretic methods to see if an

expression of both paternal and maternai alleles

can be observed, at least at some loci.

Source : MNHN, Paris

ALAIN DUBOIS

Interspecific hybridization and supraspecific classification

This chapter shall be devoted to a study of the and the other types of data available concerning

relationships between the data of hybridization the species.

VARIAB1LITY OF THE RESULTS WITHIN A TAXINOMIC GROUP

This variability has long been emphasized

by students of hybridization. If one considers

a traditional taxinomie group, e.g. a genus

of amphibians, the rule is that the results

of interspecific hybridization are most varied,

according to the species of this genus crossed two

by two, from total failure or fertilization to

complété development of normal hybrids. Fur-

thermore there often exists no transitivity in the

results: e.g. the species A and B give between

them viable hybrids, B and C also, but the

hybridization between A and C leads to a failure.

These results caused the authors to be pru¬

dent, if not suspicious, as to the use of the results

of hybridization to estimate taxinomie relation¬

ships between species. In reality this suspicion is

only partly justified. In my opinion, it is based

on a methodological error: the one which con-

sists in giving as much importance and meaning,

in the genetic and phylogenetic interprétation of

the results of hybridization, to the négative

results as to the positive ones. For the reasons

expressed above, positive results alone hâve a

clear meaning in this field, and furthermore

among these results it is préférable to take into

account only the best results obtained. If this

important distinction is made, the results of

hybridization prove much less “ anarchical ”

from a taxinomie point of view and may give us

very useful indications.

Hybridization and molecular divergence between species

The functional genetic similarity measured by

the synthetic criterion of hybridization does not

coincide with structural genetic similarity mea¬

sured by “ genetic ” distance, as was shown e.g.

by Avise & Smith (1974) in Centrarchidae, or by

Allan C. Wilson and his coworkers in various

groups of vertebrates. With the help of immu-

nological methods, these later workers hâve

estimated the molecular divergence, at the level

of molécules of albumine and of transferrine,

between species of anuran amphibians, of birds

and of placental mammals (see in particular:

Maxson, Sarich & Wilson, 1973; Wilson,

Maxson & Sarich, 1974; Prager & Wilson,

1975). Relying on the data concerning the “evo-

lutionary molecular clock ” (see above), these

authors hâve deduced the presumed dates of the

cladogenesis which has separated both lineages

which hâve led to the two species now being

compared. Furthermore they hâve taken into

account the aptitude of these species to give

viable hybrids between them. The results of these

works are presented in table I, which also gives

the mean number of species per genus in the

concemed groups.

If we consider, following the arguments pre¬

sented above, that the aptitude to give hybrids is

the due to a strong functional genetic similarity,

one must admit that the genetic évolution of

birds and of amphibians has been similar and

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

very slow: 20 to 23 millions years after the

cladogenesis which separated their ancestor, two

species of birds or of amphibians are still able to

give viable hybrids. On the other hand the

genetic évolution of mammals was much more

rapid: after 2 or 3 millions years, this aptitude to

hybridize is lost.

Table I. — Data on interspecific hybridization and generic classification in three groups of vertebrates (after

Wilson, Maxson & Sarich, 1974, Prager & Wilson, 1975, and the data of Table II).

Group

Mean âge of the divergence

between hybridizable species

(in millions of years)

Percentage of the

“ intergeneric ” hybridizations

among the successful

interspecific hybridizations

Mean number

of species

per genus

Placental mammals

Birds

Anuran amphibians

2-3

20-23

4.04

4.41

If the mean rate of spéciation has been similar

in these different classes (which is not demon-

strated but is not impossible), one expects there-

fore mammal généra to hâve a mean number of

species lower than amphibian or bird généra. It

is indeed what is observed for amphibians, but

not for birds. Otherwise, if we consider the rate

of “ intergeneric ", or so called, hybridizations, it

is almost null in amphibians, higher in mammals

and much higher in birds. These data indicate

that supraspecific taxa are probably not équiva¬

lent in genetical terms in the various classes of

vertebrates. We shall go back again in detail to

this problem in the next chapter.

From a morphological point of view, mamma-

lian évolution has been very rapid and diversi-

fying; on the other hand, amphibian évolution

has been much slower and less important. Now,

the mammals hâve also lost the ability to

hybridize much quicker than the amphibians.

According to Wilson and his coworkers, both

phenomena would ensue from the same cause:

mammals would hâve undergone more rapid

modifications of their Systems of genetic régula¬

tion. However, the recent results of Wyles,

Kunkel & Wilson (1983) on anatomical évolu¬

tion in birds indicate that it was as rapid as in

mammals, which is not consistent with the just

mentioned data concerning hybridization: this

indicates that there may exist several types of

Systems of genetic régulation, which may evolve

in a relatively independent way one from another:

one would be responsible for the évolution of

morphology, and another for the loss of the

ability to hybridize. To the best of my knowl¬

edge, this latter hypothesis has not yet been

proposed or discussed in the literature.

Wilson, Sarich & Maxson (1974) and Wil¬

son et al. (1975) hâve also underlined the fact

that mammals hâve shown a much higher rate of

chromosomal repatterning than amphibians: a

rapid évolution of the Systems of genetic régula¬

tion could therefore be associated with a rapid

évolution of karyotypes. The validity of this

hypothesis is not demonstrated, and it seems

more probable that the évolution of Systems of

genetic régulation obeys several very distinct

modalities, which do not ail imply chromosomal

repatterning: we shall address this problem again

below. However these results point to the interest

that the measure of karyological distance between

species would hâve for ail works dealing with the

évolution and classification of a group. This

distance would certainly be correlated with the

genetic divergence measured by the synthetic

criterion of hybridization.

Source : MNHN, Paris

ALAIN DUBOIS

Hybridization, phenetic similarity and cladistic kinship between species

McAllister & Coad (1978) recently devoted

an interesting work to the Cyprinidae of north-

ern America. These authors wanted to compare

the classifications of this group such as obtained

by the phenetic principles and by the cladist

principles. In order to achieve this, they did not

compare these classifications directly between

themselves, but they compared both of them

with a third set of data, those concerning

hybridization. Généra were compared two by

two. For every one of the 304 pairs of généra on

which their analysis bore, the authors calculated

a coefficient of phenetic similarity and a coeffi¬

cient of cladistic kinship between both généra,

and they ascertained whether or not natural

hybrids between these généra were known. The

authors observe that the cladistic classification of

this group coincides better with the data of

hybridization than does the phenetic classifica¬

tion.

Several aspects of the analysis of these authors

are open to criticism. Thus these authors con-

sider that the terms “ close taxinomie affinities ”

on one hand, and “ close phylogenetic relation-

ships ” on the other, are équivalent: such a

postulate is only valid for systematists who

adopt the cladist conception of classification; for

those who adopt the synthetic conception, both

phrases are not synonymous, since genetic and

ecological factors must also be taken in consid¬

ération, in addition to phylogenetic factors, for

the construction of a classification.

Furthermore, McAllister & Coad (1978)

consider that the production of hybrids is a

measure of close phylogenetic kinship. Now, if it

is true that hybridizability expresses an unde-

niable phylogenetic kinship, this kinship may be

more or less recent since, as we hâve seen, the

loss of the ability to hybridize seems to occur at

very different speeds from one animal group to

another. What the ability to give hybrids clearly

expresses, is a great genetic similarity of the two

concerned species, this similarity having of course

its origin in the fact that these species descend

from a common ancestor, but this ancestor being

more or less distant.

Other aspects of McAllister & Coad’s (1978)

analysis may be discussed, in particular the fact

that these authors hâve taken into account only

natural hybrids, while, as we hâve seen, artificial

hybrids should also hâve been considered: in a

certain way, these authors hâve measured by

their method the existence of pre-ejaculatory

mechanisms of isolation more than the genetic

potentiality of the species of both généra to give

viable hybrids. Finally their analysis is biased

because of the fact that they compared généra,

i.e. taxa the nature of which is already determined

by the conception of classification chosen (in

their case the cladist conception): to avoid any

bias of this kind, the analysis should not bear on

pairs of généra, but on pairs of species.

Be that as it may, the work of these authors is

very interesting from a methodological view-

point, for it leads to quantify various types of

relationships between species and to compare

these various measures. Such a work falls com-

pletely within the concerns of synthetist system¬

atists, who are neither pheneticist nor cladist, but

make use of information of ail kinds, phenetic,

cladistic, and other, to make a synthesis of them

and base their classification on this synthesis.

The different types of “ distances « between species

Similarly, and pushing further McAllister &

Coad’s (1978) method, it could be interesting to

make for several distinct and even very different

groups of animais overall analyses bearing on

different types od “ distances ” (or “similar¬

ités ”) between species taken two by two: phe¬

netic distance, “ genetic ” or molecular distance,

cladistic distance, karyological distance, ecologi¬

cal or eco-behavioral distance, and distance

measured by the criterion of hybridization.

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

The comparison of ail these data, or at least of

the available part of them, would certainly teach

us many things, both from the viewpoint of the

study of animal évolution and from that of

systematics. In many groups the data are still

insufficient for such an analysis, but a work of

this type could certainly be tackled for the

different classes of vertebrates and for several

groups of insects.

Several techniques are already available for

such an analysis: we shall briefly review them.

Phenetic distance

To the numerous, and now already “classi-

cal ”, methods of numerical taxinomy (Sneath &

Sokal, 1973), one must now add more recent

methods, which are based on a current reflection

on the notion of “ biological shape ” and on the

distinction which must be made between the

factors “ size ” and “ shape ” in the analysis of

morphology (Jolicoeur & Mosimann, 1960).

Some authors advocate the use of qualitative

characters to measure the distance between spe-

cies or higher taxa (e.g. Findley, 1979), while

others, more convincingly, argue that quantita¬

tive characters alone allow a non biased analysis

(e.g. Cherry et al., 1979, 1982). Some authors

insist upon the fact that the morphology of an

organism is the resuit of its growth, during

which, in particular, phenomena of allometry

take place, and they try to take these factors into

account in the analysis of shape (e.g.: Gould,

1966; Lande, 1979; Lemen & Freeman, 1984).

On the contrary, others consider that both

problems are independent and that the question

of the origin (in its genetic, ontogenetic and

phylogenetic senses) of the morphology of an

organism must not be confused with that of the

description of this morphology and of the compar¬

ison of the shapes of different species:

“ one must avoid confusing the need for a quantitative description of the

degrees of organismal différence with the need for explanations of those

différences.” (Wilson, Kunkel & Wyles, 1984: 1158).

These latter authors hâve recently studied in

detail the problem of the establishment of a

reliable index of measurements of morphological

or phenetic distance between species taken two

by two, in a sériés of works which are of a great

theoretical and practical interest (Cherry, Case

& Wilson, 1978; Cherry et al., 1979, 1982;

Wyles, Kunkel & Wilson, 1983; Larson, Pra-

ger & Wilson, 1984; Wilson, Kunkel & Wyles,

1984). Independently from these authors, other

biologists hâve recently addressed this problem

of the phenetic distance between species starting

from different viewpoints (see e.g.: Laurent,

1953, 1967, 1981; Dubois, 1976).

It is interesting to note that it is only recently

that the reflection has really developed on these

methods of measurement of phenetic distances.

The major reason for that is probably the one

emphasized by Wilson, Kunkel & Wyles (1984)

in their answer to a criticism of their previous

work (Wyles, Kunkel & Wlson, 1983) by

Hafner, Remsen & Lanyon (1984):

“ Two Perspectives in Evolutionary Biology — It appears to us that the

Hafner et al. (1984) criticism is a manifestation of the ‘ populationist '

perspective, which has dominated systematic and evolutionary biology since

the 1940’s. It focuses on the tips of the evolutionary tree and on the

uniqueness of every trait, individual, population and species (...). In

contrast, we hâve been influenced by what might be termed the ‘ distance ’

perspective, which entered evolutionary biology more than 20 years ago as

biochemists began to compare proteins from species belonging to different

branches of the tree (...).

The ‘ populationist ’ perspective’s emphasis on uniqueness engenders

respect for the généralisation referred to by Hafner et al. (1984): The set of

characters that best discriminâtes among members of one taxinomie group is

unique to that group. Such perspective makes one wary of comparing the

degree of différence between a pair of species in one taxonomie group with

that in another taxinomie group. Molecular evolutionists, by contrast, hâve

long been comfortable with the pratice of using the same yardstick (i.e.,

number of substitutions) to examine and compare évolution in vastly

different taxonomie groups. The criticism of our work by Hafner et al.

Source : MNHN, Paris

ALAIN DUBOIS

(1984) has made us realize what a deep gulf there is between these two

perspectives and how important it is to explore that gulf on another

occasion." (Wilson, Kunkel & Wyles, 1984: 1158-1159).

This problem meets that of the “ gap ” which

exists between the “ populational ” approach to

the study of évolution and the study of macro-

evolutionary phénoménal we shall corne back to

this problem in more detail below.

“ Genetic ” distance

This expression classically désignâtes distances

like Nei's (1972) or Rogers’s (1972), which are

based on the results of protein electrophoreses.

This distance is very badly named, since it only

measures in the fact the divergence at the level of

a few structural genes, without taking at ail into

account the divergence at the level of Systems of

genetic régulation, and a phrase like “ molecular

distance ” would certainly suit it much better.

Pasteur (1985) recently gave a discussion of the

different types of “ genetic ” distances based on

the results of protein electrophoreses which are

currently in use. Other types of “ genetic ”

distances must be added to these, like “ immunol-

ogical distance ” based on the micro-complement

fixation method (Wilson, Carlson & White,

1977; Pasteur & Pasteur, 1980), or the distances

based on DNA hybridization (Sibley & Ahlquist,

1982; Diamond, 1983).

Cladistic distance

To measure such a distance, one could make

use of the coefficient proposed by McAllister

& Coad (1978).

Karyological distance

Some authors, like Wilson, Sarich & Max-

son (1974) or Cothran & Smith (1983), calcu-

lated a karyological or chromosomal distance on

the basis of the number of chromosomes and the

number of chromosomes arms of the compared

species. In the future, it would be necessary to

hâve a more précisé, finer measurement of

karyological distance, for the same chromosome

number and the same fundamental number may

be obtained in a totally independent manner in

different species. A finer comparison could for

example take into account the total quantity of

nuclear DNA as well as the place of constrictions

and of bands revealed by the techniques of

banding (Dubois, 1983 a: 56-57). This will only

be possible in the case of groups which hâve

already been the subject of a rather advanced

cytogenetical study. Thus Martin & Hayman

(1965) proposed to compare karyotypes of closely

related species, a method which takes into

account the relative lengths of the arms of

chromosomes and the relative quantities of

DNA by genome; as far as they are concerned,

Prevosti, Ocana & Alonso (1975) proposed an

index based on the différences of frequencies of

chromosomal arrangements to measure a dis¬

tance between populations of the genus Droso-

phila.

Ecological or eco-behavioural distance

Works where ecological or eco-behavioural

distances between species hâve been estimated

are still rare, although such distances would be

very interesting. At the moment the interspecific

distances which hâve been measured in this field

concern only certain aspects of the ecology and

behaviour of the species: e.g. distribution of the

species in microhabitats (Ortega, Maury &

Barbault, 1982), techniques and sites of alimen¬

tation (Landres & MacMahon, 1980), trophic

spectra of the species (Barbault, 1981: 119),

characteristics of the mating calls of males

(Duellman & Pyles, 1983). In his interesting

1978 work on an équatorial community of

amphibians and reptiles, Duellman calculated

somewhat more complex distances, which take

into account various types of parameters (habi¬

tat, vertical distribution, periods and types of

activities, feeding, size, mode of reproduction,

characteristics of males mating calls). It would

certainly be interesting to develop research in

this direction, and to construct overall indexes of

eco-behavioural similarity between species, which

could take into account the physico-chemical

characteristics of the niche occupied by every

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

species (climatic and microclimatic parameters,

pedological parameters, etc.), as well as certain

biological parameters allowing a characterization

of this niche in démographie terms, in terms of

biomass and of energy flux, and of parameters

drawn from behaviour (e.g. bio-acoustic charac-

teristics of Sound émissions, etc.). It would

certainly be most instructive to compare such

distances with the other distances mentioned

here. For example, it is not unreasonable to

think that the overall morphology of a species

(its “ shape ”) expresses in a certain synthetic

way the ecological niche of this species, and it is

likely that the divergence between two species in

one or another of these fields is correlated with

their divergence in the other. It is to be wished

that ecologists got interested in these problems

and proposed methods allowing such estimâtes.

In the meanwhile, one must recognize that the

notions of ecological niche and of adaptive zone,

although fundamental in the interprétation of

evolutionary phenomena in the light of the

synthetic theory, still remain very little function-

al. This is in my opinion the field in which the

most important progress remains to be made for

a truly synthetic appraisal of evolutionary facts.

Hybrid distance

What has been said above shows that it would

be very interesting to hâve a distance index

between species measured by the criterion of

hybridization. For more simplicity, I propose to

designate such a measurement by the name of

“ hybrid distance ”. As was already emphasized

(Dubois, 1983 a: 58), the following stages of

development of hybrids seem to be generally

valid for the whole animal kingdom, and could

constitute the framework for a unique scale of

measurement of this distance:

(1) failure at fertilization;

(2) failure at the beginning of the gastrula

stage;

(3) failure during the postgastrulean embry-

onal, larval, or young stages;

(4) infertile adult animais, or adult animais

having a disturbed progeny;

(5) fertile adult animais with a normal prog¬

eny.

To this rough scale, it should soon be possible

to add a finer scale for the crosses which lead to

the development of a hybrid, at least in the first

stages (levels (3) to (5) in the scale above). In this

respect, the way was opened by Parker, Philipp

& Whitt (1985 a, 1985 b), who proposed to use

various indexes to estimate what they call the

“ regulatory distance ” between two species able

to hybridize: percentages of fertilization and or

hatching of hybrid eggs, extent of the disruptions

in the temporal expression of various enzymes

during embryonic development (in relation to

the normal temporal expression in one of the two

species crossed), extent of the disruptions in the

rates of activity of these various enzymes (in

relation to the normal rates of activity in one of

the two species crossed). At the moment these

are only several distinct indexes which give

sometimes somewhat different results, but it is

not forbidden to think that it will be possible in

the future to combine these various data into a

single “ hybrid distance ” index between two

given species. To be able to do this, however, it

will be necessary to study the relationships which

exist between the different criteria mentioned

here: some of them give similar results, others

very divergent ones. It is probable that these

criteria are not independent one from another,

and it is therefore not possible, at the moment,

to calculate a single overall index of “ hybrid

distance ” simply by adding the values of the

different indexes (Whitt, 1985). Given the vari-

ability of results, at ail levels, of hybrid crosses

(according to the direction of the cross, to the

populations and individuals used, etc.), it is at

any rate probable that such an index should not

be based on a mean of the results observed, but

rather on the best results obtained, as we hâve

seen above.

It is clear, as emphasized by Parker, Philipp

& Whitt (1985 a, 1985 b), that the relative

success of development of the various types of

hybrids expresses in a synthetic way the impor¬

tance of the divergence which has occurred

between the two compared species at the level of

their Systems of genetic régulation. It would

therefore be most interesting in the future, as

well from the viewpoint of the study of evolu¬

tionary mechanisms as from that of supraspecific

systematics, to develop methods of measurement

of “ hybrid distance ” between species.

Source : MNHN, Paris

INTERSPECIFIC HYBRIDIZATION

AND THE CONCEPT OF GENUS IN ZOOLOGY

HYBRIDIZABILITY AS A CRITERION FOR THE DEFINITION OF GENERA

Brief historical survey

The criterion of hybridizability has until now that as early as in the 19th. century some authors

been only very little used by systematists to

recognize taxa above the species level. It is true

“ The species is characterized by the

limited fertility” (translation mine),

and he considered that two species like the

donkey and the horse or like the jackal and the

dog, able to produce hybrids between them,

should be placed in the same genus (see also

Flourens, 1845 a: 298-301; 1845 b: 119-128).

More recently, Ghigi (1936) proposed to use

the hybridization criterion in macrotaxinomy: he

“ The capacity of two groups to

their categorical rank. ”

Similar ideas were more recently expressed by

Hubbs & Drewry (1960). Other authors hâve

mentioned in passing the theoretical possibility

of admitting that species liable to give viable

hybrids between them should be placed in the

same genus (see e.g.: Simpson, 1961: 90, note 10;

Short, 1969: 87; Hubbs, 1970; Pépin et al.,

1970), but these proposais had no claim to a

general value. In the recent years, finally, three

different authors (Van Gelder, 1977, 1978;

Plateaux, 1981; Dubois, 1981 a, 1981 c, 1982 a,

1983 a) independently made a similar proposai.

had tried to do it (see in this respect Fischer,

1981). Thus Flourens (1856: 6) wrote:

continuous fertility; the genus by the

suggested that according to the degree of success

of the hybridization between two species (success

measured by the degree of fertility of the Fl

hybrids of both sexes), these species be referred

to the same genus, to different généra or to

different families. In the same spirit, Kinsey

(1936, in Simpson, 1937: 265) writes:

hybridize is inversely correlated with

giving it a general value for the whole animal

classification and justifying it by rather different,

albeit not contradictory, arguments. This conver¬

gence is interesting: it shows in my opinion that

times are ripe for the use of such a criterion in

zoology.

The main argument presented by Van Gelder

(1977) to justify this proposai is the need of a

certain internai cohérence of the classificatory

System between the notion of species, defined by

a mixiological criterion, and that of genus:

Source : MNHN, Paris

ALAIN DUBOIS

“ Basically, the logic of a reproductively isolated and self-contained genus

seems inescapable if one accepts the concept of the reproductively isolated

species. The genus neither can nor should be of less dimension than the

species, and if the parameters of the species are ultimately established by its

reproductive capabilities, then the genus, too, must be so proscribed. The

greatest extent of reproductive compatibility allowed between species is

generally the production of stérile offspring. The production of fertile

offspring in nature is usually sufficient grounds for merging the parental

stocks into a single species with their récognition only as subspecies.

Similarly, for allopatric species, captive hybridization with fertile offspring

may be used to consider the parental stocks conspecific. If these are the

reproductive limits of species it would seem to follow that the genus must be

reproductively at least, if not more, separable, and that crosses between

généra be wholly incapable of producing a live offspring. ” (Van Gelder,

1977: 18).

“ (...) the upper limit for the species (reproductive incompatibility) should

also be contained in the définition of the genus, and at least represent its

lower limit. This would imply that intergeneric hybrids should not be

possible by so defining the genus. The arbitrariness of the définition of the

genus exists in its width and upper reaches, not at its interface with the

species, where its définition is the same as that of a species. ” (Van Gelder,

1977: 4).

As far as he is concemed. Plateaux (1981) meon ” (in the sense of Cuénot & Tétry, 1951)

justifies his proposai mainly by an argument of to designate a group of species liable to hybridize,

“ common sense Using the term “ synga- he writes:

“ If the syngameon can corne to include several généra, the genus does not

mean much any more. One could think of replacing it by the syngameon.

but the latter is usually not yet delimited. It is better to consider something

wider. But, at least, it should be a group the lower limits of which may be

traced before they join those of the species!

It seems to me that one could take the strict rule not to place in different

généra species able to produce together fertile hybrids, even if this fertility is

only very partial. In most cases, it would even be wiser to place in a same

genus ail the species able to produce between them hybrids of any kind. "

(Plateaux, 1981: 518; translation mine).

As far as I am concerned, I made very précisé preceding analysis (Dubois, 1981 a, 1981 c, 1982 a,

propositions for the use of a criterion ofhybridi- 1983 a, 1985 b). Let us now examine these

zability in animal systematics, on the basis of thp propositions in detail.

PRECISE FORMULATION OF THE CRITERION AND OF ITS CONDITIONS OF USE

The first point to insist upon here is the fact

that the success or the failure of hybridization

does not at ail hâve the same meaning or

importance. A single gene brought by one of the

parents may be enough to prevent the develop¬

ment of a hybrid zygote, even though ail other

genes are compatible (example of the Lhr gene of

Drosophila simulons, discussed above). The fail¬

ure of hybridization indicates that the two

populations of which the individuals are inter-

sterile do not belong to the same species. It does

not tell us anything more about the genetic and

phylogenetic relationships between both species.

The same is true for the infertility of some adult

hybrids: it can be due to several types of causes,

some of which involve only a few genes or even a

single gene, and it is therefore of rather unclear

meaning. The factors of lethality and of infertil-

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

ity of hybrids being very diverse, it is impossible

to take into account the négative results of

hybridization (lethality during development, infer-

tility of adults or failure in F2) for a phylogenetic

and genetic analysis of the relationships between

species. For not having realized that, some

authors hâve believed that hybridization could

be of no use in supraspecific systematics, while

only its négative results would be.

On the other hand, positive results give very

interesting information: if it is rather easy to

prevent the development of a zygote issued from

two very closely related species, it is impossible

to do the contrary, i.e. to obtain a normal

development starting from a diploid egg issued

from the hybridization of two distant species.

Given the complexity of the genome of eucaryotes,

it is quite out of the question that two genomes

could be functionally compatible by convergence

or by chance. The compatibility of two genomes

proves that the two species which bear them

descend from a relatively recent common ances-

tor, from which they hâve conserved the homolo-

gous parts of their genomes. The criterion of

hybridizability has therefore both a genetic and a

phylogenetic meaning, although its phylogenetic

meaning is less clear and more ambiguous than

its genetic meaning, since the loss of the ability

to hybridize occurs at a different speed in various

animal groups.

My proposition is therefore to consider that

when two species are liable to give rise between

them to viable adult hybrids, these two species

must be included in a same genus. Let us

remember that these are true diploid hybrids,

possibly obtained in experimental conditions,

which may be fertile or not, and finally that we

only take into account the best resuit observed in

various crosses between two species, possibly in

some only of the types of crosses which may be

realized between them (e.g. male of one species

with female of the other, but not the reverse, or

animais coming from certain populations only).

The criterion of hybridizability must therefore

be used only in one direction, to group together

species in a same genus, but not to separate

généra: when viable adult hybrids may be obtained

between the species A and B, these species

belong to the same genus; on the other hand if

hybridization does not occur or if the hybrids are

not viable, no information is given and the

criterion must never be used to place two species

in two distinct généra.

We hâve here, according to Simpson’s (1951,

1961) terminology, a nonarbitrary criterion as to

inclusiôn, but which must never be used for

exclusion:

“ A group is nonarbitrary as to inclusion if ail its members are continuous

by an appropriate criterion, and nonarbitrary as to exclusion if it is

discontinuous from any other group by the same criterion. It is arbitrary as

to inclusion if it has internai discontinuities and as to exclusion if it has an

external continuity. ” (Simpson, 1961: 115).

A second very important point is the fact that

this criterion takes into account the genetic

potentialities which exist to build an organism,

and not at ail the fact that hybrids do exist or

not in nature, various “ parasitic ” factors (in

particular eco-behavioural and geographical ones)

being liable to be responsible for their absence.

The criterion is obviously ail the more valid

when natural hybrids do exist, but the use of the

criterion to group together species in one genus

only implies the ability to obtain adult hybrids

between both species, even if for this it has been

necessary to call upon particular techniques, like

artificial insémination and fertilization, or even

more elaborate techniques aiming at solving

certain spécifie problems (e.g., in amphibians,

“ Rostand’s technique ”, described above).

Obviously, this criterion is not used alone: it

intervenes as a new and additional piece of

information but it does not nullify ail the other

pièces of information which had already been

gathered, by the other more “ classical ” methods,

on the species considered. Thus, before the

hybridization between species A and B was

observed, these two species may hâve been

placed in two distinct généra I and II, on the

basis of other criteria (morphological, molecular

and ecological resemblance, data on the phylo-

geny of the group, etc.). The fact that these

species prove able to give adult hybrids must first

Source : MNHN, Paris

ALAIN DUBOIS

prompt the systematist to a critical reappraisal of

the validity of taxa I and II: it might well be that

these taxa, or one of them, constitute artificial

groupings, e.g. placing together species of differ¬

ent phylogenetic origins and resembling each

other by convergence, or that one of both species

A and B has been placed by mistake in the genus

I or II but belongs indeed to the other one. The

results of hybridization may thus suggest the

realization of works of systematic révision at the

generic or familial level and lead to rectify

certain mistakes. However it frequently happens

that the revisional work leads to a confirmation

of the validity of groups I and II and the

respective membership in these two groups of the

species A and B. The fact that these two species

are hybridizable implies then not only that they

should be grouped together in a same genus, but

also that ail the other species which by other

criteria were classed in the same genus as A and

in the same genus as B be placed in this genus —

in other words to group the former généra I and

II together in a single genus. If both groups are

separated by a certain morphological, ecological,

or other, discontinuity, it may be well to retain

for them the status of distinct subgenera within

the new genus. It is important to emphasize that

the fact that a single pair of species belonging to

the former généra I and II gives adult hybrids is

enough to group both généra together, even if no

other pair of species of the two généra is known

to give viable hybrids. As a matter of fact to

require that ail species of both généra be hybrid¬

izable two by two and to refuse to join the

généra if they are not would corne down to use

the négative results of hybridization for the

construction of the classification, and we hâve

seen on the contrary that only positive results

may be used for this aim.

To sum up, the new criterion may be formulated

as follows: when two species are able to give birth

to viable adult hybrids, be these fertile or not, both

species must be included in the same genus;

furthermore, if these two species had previously

been attributed, on the basis of valid criteria, to

two distinct généra, the latter should be merged

together.

Taxinomic characters and relational taxinomic criteria

The word “ classification ” is used in two designate classificatory activity itself. Mayr

distinct senses (Mayr, 1969: 4): (1) to designate (1982 a: 185) proposed the following définition

the product of the activity of taxinomists; (2) to of classification as an activity:

" Classification is the ordering of organisms into taxa on the basis of their

similarity and relationship as determined by or inferred from their

taxonomie characters. ”

As for the notion of taxinomic character,

Mayr (1969: 121) defines it as follows:

" A taxonomie character is any attribute of a member of a taxon by which

it differs or may differ from a member of a different taxon. ”

These définitions are those of a synthetist characters (Dupuis, 1979, 1984), proposed slightly

systematist. As far as they are concemed, cia- different définitions. For example Wiley’s (1981:

dists, who hâve devoted an important reflection 116) définition reads as follows:

to the concept of character and to the analysis of

“ A character is a feature of an organism which is the product of an

ontogenetic or cytogenetic sequence of previously existing features, or a

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

feature of a previously existing parental organism(s). Such features arise in

évolution by modification of a previously existing ontogenetic or cytogenetic

or molecular sequence. ”

According to these définitions, or to équiva¬

lent définitions which could be found in many

other works, (1) classification is based on taxino¬

mie characters, and (2) taxinomie characters are

attributes of the organisms that one wants to

classify.

The construction of a classification requires

therefore a two-step procedure: (1) analysis of

taxinomie characters of the to-be-classified orga¬

nisms; (2) comparison of these organisms on the

basis of the results of this analysis.

This process is summarized for example by

Sibley (1965: 114), who writes:

“ There is, in systematics, only one basic technique, that of comparison.

Because comparisons between whole organisms présent insuperable difficul¬

tés it is customary, in fact necessary, to compare characters. ”

From this viewpoint, there exists no difTerence

between the various conceptions of systematics

currently in existence: they ail construct classifi¬

cations on the basis of characters, which are

recorded on individuals; on the other hand what

distinguishes these conceptions are the methods

of comparison used, some of which (cladists,

synthetists) rely on an analysis of the évolution

of characters in a lineage (plesiomorphous to

apomorphous characters), while others (empir-

ists, pheneticists) do not.

The importance of the analysis of characters

(be these morphological, molecular, ecological,

etc.) in systematics is considérable, and it is not

my intention to negate it. However I think that it

is not only on this basis that classifications can

be built. This fact is particularly obvious at the

level of the key-category of Linnaean taxinomy,

that of species, as I hâve already emphasized:

“ There exist no ‘ morphological ' species, no * ecological 4 genetical

etc., species, not even ' biological ’ species: ail species of living beings are

• biological '! There does not exist either ‘ criteria ’ for the species, or, rather,

there exists only one, which is the coincidence between natural reality and

the ‘ theoretical ’ concept of species. Several ‘ criteria ’ do exist which allow

one to differentiate individuals (or groups of individuals) within popula¬

tions, or populations, and to quantitatively appreciate the importance of

divergences, but none of these criteria by itself tells us if the observed

différences are of a spécifie ‘ nature '. The importance of divergences

between populations will be liable in certain cases to give us dues for

example on the duration of the séparation which may hâve existed between

them, but it will not allow us to know if these remain able or not, e.g. on the

occasion of a new geographical contact between them, to merge together

and constitute again a single génie pool. Except for the karyological and

mixiological criteria, and, even there, (...), in certain cases only, no criterion

allows one to assert that the step of spéciation has been crossed between two

populations or groups of populations. ” (Dubois, 1977 b: 205, translated).

As a matter of fact, the only true “ criterion ”

of the species is the conformity with the défini¬

tion of “ protected gene pool ”, whatever the

method used to demonstrate it. It is true that

generally the decisions of systematists at this

level are dictated by the analysis of characters,

but they may sometimes dérivé from the use of

other criteria, such as that of the existence or

non-existence of a natural hybridization between

two sympatric or parapatric groups of animais.

This existence or non-existence may sometimes

itself be demonstrated by the analysis of charac¬

ters, but sometimes by other methods, like the

observation of the behaviour of the animais of

both groups when they are in contact. The

criterion then used bears on the type of relation

which exists between both groups of organisms

compared. It is a relational taxinomie criterion.

Source : MNHN, Paris

ALAIN DUBOIS

which is interested in the particularities of the

relation between these organisms, but not in the

attributes of these organisms taken one by one.

The criterion of hybridizability, which I pro¬

pose to use in supraspecific systematics, is pre-

cisely a criterion of this type. Its use demands the

study of the interaction which takes place, not in

nature between two groups of individuals, but,

this time, between two species when their genomes

must collaborate to build an organism together.

In a sense, it is the species themselves which

compare themselves, not an outside observer. In

this sense, the criterion of hybridizability is an

objective, nonarbitrary criterion, which makes

the genus a “ natural ’’ taxon, in the same sense

as the species.

Classification can therefore not rely upon the

single analysis of characters. It demands a

synthetic approach, and the use, in addition to

the taxinomie characters (proper to any of the

compared organisms taken separately), of rela-

tional taxinomie criteria which are based on the

properties of the relation which exists, in nature

or in experimental conditions, between the orga¬

nisms compared.

For this reason, I think that Mayr’s (1982 a)

définition of classification given above cannot be

retained. I advocate rather the use of a définition

which is not based on the concept of taxinomie

character, like for example those proposed ear-

lier by Mayr (1969):

“ Classification. The délimitation, ordering, and ranking of taxa. ”

(Mayr, 1969: 400).

“ Biological classification. The arranging of organisms into taxa on the

basis of inferences concerning their genetic relationship. " (Mayr, 1969:

399).

“ Zoological classification is the ordering of animais into groups on the

basis of their similarity and relationship. ” (Mayr, 1969: 55).

In what précédés, I mentioned two relational

taxinomie criteria, one of which is useful at the

level of the species, and the other one at the level

of the genus. Both criteria are based on facts of

hybridization, and one could think that it is only

around these facts that such criteria could be

proposed. This is not true, just as it would not be

correct to consider that both criteria mentioned

above are of the same type. As a matter of fact

the relational criterion used at the level of the

species (existence or not of hybrids in nature and

of a génie flux between both groups of sympatric

or parapatric animais) is highly synthetic, since it

takes into account both eco-behavioural or mor-

phological phenomena (existence or not of pre-

ejaculatory mechanisms of isolation between

species), and genetic and developmental phe¬

nomena (existence or not of post-ejaculatory

mechanisms of isolation). On the other hand, the

criterion of hybridizability used at the level of

the genus takes only into account genetic and

developmental phenomena, since it is only inter¬

ested in the existence or not of postzygotic

mechanisms of isolation between species.

Conversely, one can perfectly imagine, although

at the moment their use is virtually non existent

in systematics, relational taxinomie criteria which

would take into account other phenomena than

genetic or developmental ones, e.g. ecological

phenomena. A good example in this field is that

of the criterion proposed by Illies (1970) to

define généra — a criterion which, to my knowl¬

edge, has been received with complété indiffér¬

ence by zoologists until now. Basing himself on

Monard’s (or Gause’s) principle, according to

which two species having very similar or identi-

cal ecological niches cannot live in sympatry, this

author proposed to consider the ecological coexis¬

tence of two species as a criterion of membership

of the latter in two distinct généra. Whatever

may be thought of the validity of this criterion

(see below), it is undeniably a relational taxino¬

mie criterion.

The use of such criteria is justified in a

“ synthetic ” perspective of zoological classifica¬

tion, but would hâve little meaning for systema-

tists adopting the empirical, phenetic or cladist

conceptions of classification. These criteria are in

fact totally incompatible with any typological

conception of classification (Mayr, 1969, 1982 a).

They allow one to put once and for ail an end to

the notion of “ typical " character , to the concep¬

tions according to which an organism would

hâve two kinds of characters, some “ spécifie ”,

some “ generic ”, “ familial ”, etc. One must

class organisms, not characters, which was already

expressed by Linnaeus (1751: 119) when he

wrote:

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

“ Characterem non constituere Genus, sed Genus Characterem.

Characterem fluere e Genere, non Genus e Characlere.

Characterem non esse, ut Genus fiat, sed ut Genus noscatur".

In a synthetic conception of classification then,

the genus could not “ rest ” on a single charac-

ter, on a presence/absence dichotomy in a déter¬

mination key. On the contrary généra can be

polythetic (Sneath, 1962; Mayr, 1969; Sneath &

Sokal, 1973), i.e it can well be that certain

species do not possess some particularities “char-

acteristic ” of the genus, that no “ diagnostic ”

character be common to ail species of the genus.

This is contradictory to an attitude still rather

frequent among some systematists, and which is

well expressed for example by Alphéraky (1912:

36, translation mine):

“ Every Species, or member of a Genus, must absolutely possess ail the

characters proper to the Genus, and if one of them possesses, be it only one

single additional character, or if it lacks one, it must be excluded from this

Genus and placed in a distinct Genus

Needless to say, from a purely empirical and

pragmatical viewpoint, the opinion expressed in

this citation is perfectly justified. It is not so

insofar as one considers that systematists must

try to recognize only taxa which correspond to

natural evolutionary units, and not “pigeon-

holes ” aiming at facilitating the identification of

specimens.

The criterion of hybridizability

AND THE PROBLEM OF THE EQUIVALENCE OF HIGHER TAXA

Introduction

One of the main interests of the new criterion

of hybridizability is that it permits a standardiza-

tion of systematics in the whole animal kingdom,

and that it allows one to solve in part the

problem, which has preoccupied many systemat¬

ists (e.g.: Hennig, 1950, 1966; Crowson, 1970;

Van Valen, 1973; Schaefer, 1976; Sibley &

Ahlquist, 1982), of the équivalence of higher

taxa in different groups. By making use of this

criterion, in a certain way a genus of ants would

be équivalent to a genus of mammals. It would

of course be so only partly, in particular because

the criterion is not symmetrical and cannot

always be used, problems which we shall discuss

again below, but it would nevertheless make for

important progress in this direction.

The comparative study of classifications is the

field of “ comparative systematics ”, in the sense

of Mayr & Short (1970) and of Bock &

Farrand (1980). As was emphasized by the

latter authors, this domain is still little explored:

“ Comparative systematics is a new area of inquiry within taxonomy, so

recent that it has not been discussed in general texts on systematics and its

major goals hâve not yet been clearly formulated. A preliminary statement

of the goals of comparative systematics may be — the analysis of the

structure and composition of taxa (i.e., the number of component subgroups

in each taxon and their nature) and of their evolutionary history. These

goals may change as more is learned about the comparative systematics of

diverse groups of organisms. ” (Bock & Farrand, 1980: 22).

Source : MNHN, Paris

ALAIN DUBOIS

However, the need to hâve criteria of stan-

dardization and of comparison, permitting the

récognition of équivalent taxa in different groups,

is not due to a whim or simply to a taste for

“ inteilectual elegance As was emphasized in

particular by Mayr (1969, 1974, 1981, 1982 a),

zoological classification must not only be an

identification System, but also a true theory, in

the light of which ail biological facts find their

true meaning.

The works of the last years on various aspects

of biological évolution fully confirm this view-

point. More and more frequently indeed, classifi¬

cation is taken as a System of référencé to

interpret the results of comparative works car-

ried out in the most varied fields (molécules,

morphology, behavior, ecology, etc.). Let me

give a few examples. It is on the basis of the

current classifications of these groups that Pra-

ger & Wilson (1975), Cherry, Case & Wilson

(1978), Cherry et al (1982) and Wyles, Kunkel

& Wilson (1983) hâve pointed to the disparities

of the rates of morphological, karyological and

molecular évolution in the different classes of

vertebrates. Similarly, it is on the basis of these

classifications that Avise & Aquadro (1982),

Aquadro & Avise (1982) and others hâve

estimated that évolution at the level of structural

genes has been slower in birds than in other

vertebrates, results which are disputed by Sibley

& Ahlquist (1982). Finally, it is on the basis of

the current classification of teleostS that Whitt,

Childers & Cho (1973), Champion & Whitt

(1976), Philipp, Childers & Whitt (1979),

Philipp, Parker & Whitt (1983) and Parker,

Philipp & Whitt (1985 a, 1985 b) hâve asserted

that the disruption of allelic expression in certain

hybrids (inhibition of certain alleles, modifica¬

tion of the rates of expression of certain other

ones, etc.) is a function of the “ systematic

distance ” between the hybridized species. Such

“ systematic distances ” are more and more used,

and compared with the other types of distances

(“ genetic ”, phenetic, karyological ones, etc.)

discussed above. It is évident that such a practice

is meaningful only if classification is based, at

least in part, on objective, nonarbitrary criteria,

and is not completely empirical. It becomes thus

more and more urgent, as synthetic works of this

type multiply, to find criteria having a general

value for ail animal groups.

One could believe that the only category

allowing such a hope is that of the species, since

it is only at this level that genetic material is

exchanged and that real genetic units do exist in

nature, independently from the interprétation

which we can make of it. Once spéciation has

taken place, there does not usually (with a few

exceptions) exist genetic exchanges between the

individuals of two different taxa, and it would be

necessary to abandon the hope of recognizing

“ naturel ” or “ équivalent ” groups. However

we hâve seen above that, at the level of the genus

at least, it is possible, by taking into account the

results of artificial hybridization (then not only

phenomena which occur spontaneously in nature),

to recognize “ naturel ” units, on the basis of a

criterion which relies only upon the properties of

the cross realized between two species. This

criterion allows one to recognize taxa which are

équivalent from one group to another. It is not

the only one in this case, and we shall now

devote a deeper study to the various criteria of

this type.

The criteria of Equivalence between taxa

Schaefer (1976) has addressed this problem of considers it insolvable, because of the absence of

the équivalence of taxa in different groups. He common criteria between different groups:

“ In short, it seems not likely that higher categories can be made

équivalent even in related groups. I do not think the reasons are obscure.

For there to be bases for establishing équivalence, there need to be some

common criteria: either common adaptations to the same or different

environments, or perhaps different adaptations to the same environment.

Family-groups with the same adaptations to the same environment are ipso

facto not different family-groups, if by ‘ same adaptations ’ we mean

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

genetically the same. It is improbable that two groups would arrive at or

achieve, the same adaptations to different environments, since the environment

after ail culls from the genetic variety, and different environments will not

cull the same adaptations from that variety.

The possibility of different adaptations to the same environment is more

interesting. Fish and cetaceans are adapted to roughly the same environ¬

ment, as are kangaroos, bison and African antelope: Can the family-groups

here be made équivalent? I much doubt it. Equivalence is a taxonomie

judgment, and such judgments are based on assessments of genetic

similarity. Where there is no genetic similarity, such judgments as équiva¬

lence cannot be made. However similar the adaptations of different groups

to the same environment may appear, these adaptations will not bear close

scrutiny; they are only superficially similar, having been attained by different

genetic routes; they therefore cannot be compared except superficially. "

(Schaefer, 1976: 2).

As far as he is concerned, Van Valen (1973)

tried and built a list of the criteria which could

be used to compare taxa from one group to

another, and which prove more numerous than

those considered by Schaefer (1976). The fol-

lowing list is inspired by that of Van Valen

(1973), to which however appréciable modifica¬

tions hâve been brought.

Phenetic criteria

A first criterion could be phenotypic diversity

(Van Valen, 1973: 334). This could be estimated

on the basis of the phenetic distances discussed

above. Such a measure would be interesting for

comparing taxa (and possibly to deduct from this

certain modalities of their évolution), but, as

remarked for example by Lemen & Freeman

(1984: 1236), they would not at ail allow the

définition of supraspecific taxa, just like species

cannot be defined by their intraspecific variabi-

lity.

“ Genetic ” or molecular criteria

A second type of criterion contemplated by

Van Valen (1973: 334) is that of genotypic (or

genetic) diversity. As was shown by ail the

preceding discussion, such a measurement poses

many problems. In Van Valen’s (1973) mind,

such a diversity could be estimated by the

diversity of proteins (indirect method) or of

DNA (direct method). In reality, as we hâve

seen, an index of this type would inform us

about the structural divergence between the

genomes compared, but not about their functional

différences. However, because the évolution of

structural genes is largely proportional to time,

an index ot this type could possibly allow one to

estimate the âge of taxa. We will corne back to

this aspect below.

Ecological criteria

Van Valen (1973: 333-334) considers the

possibility of using an ecological criterion to

compare taxa from one group to another only in

a relatively restricted way: he proposes to esti¬

mate the number of individuals, or the biomass,

or the energetic value, represented at a given

moment by the group. In a certain way, such a

measure would give an idea of the “ evolutionary

success ” of a group. However the groups which

hâve the highest number of individuals are

probably not the same as those which hâve the

highest biomass or energetic value. Furthermore,

the “ ecological success ” of a group may not

necessarily be measured in quantitative terms:

some species produce relatively few descendants

at each génération, but these hâve a high survival

rate, while others produce a large number of

descendants but which undergo a high mortality

at each génération. Eventually, the only real

measure of the “ success ” of a group is its

survival and perpétuation, and only the relative

extinction of species in different groups could

give us a négative estimate of it (see also on this

question Wake, Roth & Wake, 1983).

Other ecological criteria could be used, for

example ecological diversity, estimated on the

basis of the “ ecological distances ” mentioned

above. Independently from the practical prob¬

lems raised by such a mesure, it would raise the

Source : MNHN, Paris

ALAIN DUBOIS

same theoretical problem as the phenetic dis¬

tance discussed above: it would allow the compa-

rison of taxa, but not their définition.

Illies (1970) recently proposed an interesting

criterion to define généra: the ecological coexis¬

tence of two species would be considered as

meaning that these must be referred to two

distinct généra. This is a nonarbitrary criterion

as to exclusion, in the sense of Simpson (1951,

1961). As we hâve seen, this is a relational

taxinomie criterion, like that of hybridizability.

However, despite its interest, this criterion does

not seem to be utilizable for recognizing généra.

As a matter of fact, if it is true that the ecological

niches of two species cannot be identical, there

often exists a wide overlap between the niches of

the species which occupy a same adaptive zone,

the latter being wider than any of the niches

which compose it. The application of the ecologi¬

cal exclusion criterion proposed by Illies (1970)

would lead in practice to multiplying the names

of généra considerably, and to empty the notion

of genus of almost ail phylogenetic meaning. It

could however be interesting to explore this type

of criterion in more detail, by taking into

account not only the spatial dimension of the

niches of species (coexistence) but also other

dimensions of the latter (compétition at the level

of resources, of the acoustic niche, etc.). Criteria

based on such analyses could prove useful to

define certain supraspecific and infrageneric taxa

(subgenus, species group, etc.).

Absolute âge of taxa

An attractive criterion to make taxa équivalent

from one group to another is that of the absolute

âge of taxa. This criterion, first proposed by

Hennig (1936, 1950, 1966) and adopted by

several authors (Kiriakoff, 1954, 1965; Crowson,

1970; etc.), raises practical problems of applica¬

tion which appeared insuperable a short time

ago (see e.g. the discussion of this question in

Dupuis, 1979: 47-50). Recently, Sibley & Ahl-

quist (1982) asserted that the methods of hybrid¬

ization of the DNA would allow one to reliably

date the cladogeneses which hâve separated

lineages leading to contemporaneous species,

and suggested the use of this criterion to ascer-

tain the ranks of taxa. Actually, as was empha-

sized e.g. by Simpson (1962) or Mayr (1969: 72,

230; 1974), and as was recalled above, the use of

such a criterion has a meaning only within the

frame of the cladist conception of classification

(the latter being supposed to be a direct transla¬

tion of the phylogenetic tree), but not within that

of the synthetic or evolutionary conception of

classification: the attribution of the same rank to

taxa of the same âge, independently from the

Table II. Number of living taxa in the four major categories of the Linnaean taxinomie hierarchy (species,

genus, family, order) in the six classes of Vertebrata Gnathostomata, and positions of these categories in

the distance species-class, calculated according to Van Valen's (1973) method of analysis, slightly

modified: here the categories species and class hâve been chosen for the extremities of the axis (with the

respective values 0 and 1), and the position P of any other category is given by the relation P = (1 — R)

x 100, where R is the ratio of the logarithm of the number of taxa of this category to the logarithm of

the number of species in the class (this position may also be determined graphically, as shown by Van

Valen, 1973).

N = number of taxa of the category in the class.

P = position of the category.

Sources for the numerical data on taxa: (1) Nelson, 1984; (2) Dubois, 1985 a; (3) Frost, 1985;

i‘o-7? UELLMAN ’ 1979; (5 * BoCK & Farrand - 198 °: (6) Anderson & Jones, 1967; (7) Van Gelder, 1977,

Class

References

Species (S)

Généra

(G)

Families (F)

Chondrichthyes (CH)

793

151

24.8

51.8

Osteichthyes (OS)

20857

3881

16.9

418

39.3

Amphibia (AM)

2,3

4015

395

27.9

56.8

Reptilia (RE)

5954

885

21.9

55.9

Aves (AV)

9021

2045

16.3

160

44.3

Mammalia (MA)

4060

1004

16.8

122

42.2

—

960

17.4

—

Orders (O)

N P

6 73.2

42 62.4

3 86.8

6 79.4

28 63.4

20 64.0

Source : MNHN, Paris

THE GENIUS IN ZOOLOGY

S G F O C

0 0.25 0.50 0.75 1.00

Fig. 2. — Positions of the major categories of the Linnaean taxinomie hierarchy in the six classes of gnathostome vertebrates

(living species only; from the values of Table II).

In a strictly “ balanced " or “ symmetric ” classification, the positions of the intermediate categories would be

regularly spaced between the two extremities (0.25 for genus, 0.50 for family, 0.75 for order). When the observed values

are lower than these “ expected " values (i.e. when the observed point is to the left of the vertical line corresponding to

the " expected ” value), the classification of the class may be described as “ oversplit " according to Van Valen's

metataxinomic criterion; on the contrary, when these values are higher than the “ expected ” ones (or the point to the

right of the vertical line), the classification may be described as “ overlumped " with respect to this criterion.

For the meaning of the abbreviations, see Table II. For the class of mammals, the position of the category genus

is shown according to two different generic classifications (see text): (1) Anderson & Jones, 1967; (2) Van Gelder,

1977, 1978.

Source : MNHN, Paris

ALAIN DUBOIS

fact that they may or may not hâve experienced

an important divergence or diversification since

their appearance, greatly reduces the information

contents of the classification, which does not give

any more indications on the different evolutionary

rates from one group to another, on the shifts to

different adaptive zones, etc. For this reason, and

although it has the advantage over many other

criteria of being objective and nonarbitrary (at

least in the form advocated by Sibley & Ahlquist,

1982), it is my opinion that this criterion should

not be used to détermine the ranks of taxa. On

the other hand, the “ genetic ” distance between

species measured by the very sensitive criterion

of DNA hybridization is obviously of a very

great interest to détermine the phylogenetic

relations between species, for the study of the

rates of évolution, etc.

Van Valens’s meta taxinomie criterion

A simple criterion of comparison of classifica¬

tions is the number of species and of higher taxa

of every category in the group under study. Such

a criterion would allow one to uncover dispari-

ties between groups: thus, if one compares the

“ mean ” family of insects to the “ mean ” family

of mammals, one notices that the first one

contains many more species than the second

one ; the différence is less important for the

number of généra (Van Valen, 1973: 333). Like

some ecological criteria mentioned above, a

criterion of this type would roughly measure the

“ ecological success ” of a group. However the

meaning of the number of taxa taken by itself is

not clear, since various factors interfère with this

number: the size of species, the “ width ” of the

adaptive zone occupied by the group, the prés¬

ence or absence of other animal groups in this

zone, etc. The criterion cannot therefore be used

to standardize the classification of different

groups, although this has been contemplated and

even put into practice by some systematists

having an empirical conception of classification.

Although such criteria cannot be used to

construct a classification, they allow one on the

other hand to compare classifications between

themselves. Several authors hâve already addressed

this question and produced quantitative analyses

of zoological classifications (Williams, 1951;

Mandelbrot, 1956; Mayr & Short, 1970;

Clayton, 1972; Van Valen, 1973; Gorham,

1977; Bock & Farrand, 1980; Stoyan, Stoyan

& Fiksel, 1983). In this respect, the most

interesting analysis seems to be Van Valen’s

(1973). This author proposed a new criterion,

which he called “ metataxinomic criterion ”, to

analyse biological classifications and ascertain

whether categories occupy similar “ positions ”

in the classification from one group to another.

The “ position ” of every category is determined,

for every group, by the number of taxa belong-

ing to the category, in relation to the total

number of species of the group (for more clarity

and details, see table II, fig. 2 and their legends).

As shown in fig. 2 and as will be discussed in

more detail below, according to this criterion

some classifications may be described as “bal-

anced ”, others as “ oversplit ” and others as

“ overlumped ”. The meaning of these différ¬

ences is not clear, however Van Valen (1973:

341) notes:

“ Although there is no apparent reason other than symmetry why

categories should tend to be equally spaced, it is interesting that the two

most studied groups of animais, the Chordata and Insecta, approach equal

spacing more closely than any other major taxa except perhaps the

Platyhelminthes and Protozoa.”

One may suppose that, when this criterion is

applied to groups which are important enough

(phyla containing thousands of species and of

higher taxa), the overall resuit “ erases ” the

disparities which may exist at a lower scale as to

the “ quality ” of the current classification (groups

having been the subject of more or less recent

révisions, by authors with a tendency toward

splitting or lumping, in favor of such theory of

classification, etc.), and expresses in a synthetic

way the particularities of the classification stud¬

ied, allowing therefore comparisons. It would

thus be a good criterion of “ comparative sys-

tematics ”.

Van Valen’s metataxinomic criterion is there¬

fore interesting for it allows one to compare

classifications and possibly to draw conclusions

as to either the validity of a classification, or the

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

evolutionary modalities within a group (see

below), but it does not in itself make définition

of taxa possible, nor does it make them équiva¬

lent from one group to another. It could permit

one to do it within the frame of a totally

empirical conception of classification, in which

the only objective would be to build the most

“ practical ” possible System of identification: in

such a perspective, it would be necessary to

modify the existing classifications at any price in

order to make them as “ balanced ” or “symmet-

rical ” as possible, a classification wherein ail

categories would be équidistant being the richest

in information contents. On the contrary, within

the frame of an evolutionary conception of

systematics, it is interesting to observe disparities

in the classification from one group to another,

for they may indicate the existence of different

evolutionary phenomena: we shall corne back to

this problem below.

Hybridizability criterion

As finally concerns the criterion of hybridiza¬

tion, Van Valen (1973: 334) believes that it

cannot be retained as a criterion permitting one

to make taxa équivalent from one group to

another:

“ Because it is affected by sympatry, is possible only for low categories,

and can be oligogenic or even monogenic, it does not seem to be a good

estimator. ”

Furthermore he thinks that this criterion is

probably équivalent to the criterion of “geno-

typic diversity ” discussed above: by so doing he

ignores the différence between structural and

regulatory genes on which the présent work

insists.

Other authors hâve also considered that inter-

specific hybridization could not give useful infor¬

mation for constructing supraspecific classifica¬

tion, because of the variability of the results of

the hybridization between species considered

very close along other criteria, of the disparities

observed between reciprocal crosses, and gener-

ally of the poor corrélation between the results of

hybridization and the current classification (see

e.g.: Montalenti, 1938; Moore, 1955; Cousin,

1967). These authors hâve not realized that the

négative results of hybridization do not hâve the

same value as the positive ones.

On the other hand, the argumentation here

presented relies upon the very particular mean-

ing that is attributed to the success of hybridiza¬

tion. According to this interprétation, the crite¬

rion of hybridizability has a deep synthetic

biological meaning, which largely exceeds the

meaning of every morphological, molecular, eco-

logical, or other, criterion taken individually. In

the light of this criterion, the genus stops being

an artificial category to become, in the same way

as the species, an evolutionary systematics cate¬

gory, expressing the existence of real evolution¬

ary phenomena in nature. The groups thus

defined are équivalent, in functional genetic

terms, to one another. Within each of these

“ genetic units ”, the variance of the Systems of

genetic régulation remains moderate enough to

allow the préservation of a possibility of hybrid

development between two éléments of this unit.

Therefore the hybridizability criterion allows

one to recognize real taxa (defined by a rela-

tional, objective and nonarbitrary criterion) which

are équivalent between them from one group to

another. Such criteria being rare, it is important

to use this one well, and to define its use in order

to make it the most general possible at the scale

of the whole animal kingdom. In this respect,

two aspects of my proposition may be discussed

in a more detailed way: (1) the choice of the

developmental stage retained for considering

that the hybridization between two species has

succeeded; (2) the decision to assign the rank of

genus to the taxon defined by this criterion of

hybridizability.

Choice of the developmental stage

In order to décidé that the hybridization

between two species has “ succeeded ”, one must

dispose of a stage of référencé, the minimum

stage to be attained by the hybrid product. In

this respect , Van Gelder’s (1977) proposition

slightly di fiers from mine, since this author

suggested the criterion of “ birth of a living

offspring ”, while I suggested that of “ obtaining

of a viable adult hybrid ” (Dubois, 1981 a). This

last criterion seems to me to merit rétention, for

Source : MNHN, Paris

ALAIN DUBOIS

it has a larger generality in the whole animal

kingdom than that of birth. Virtually ail animal

species (except maybe some Protozoa) possess an

adult stage, characterized by sexual maturity and

the ability to reproduce, and differing in that

from the stages which précédé it (embryo, larva,

young, subadult, etc.). The latter hâve no gener¬

ality, not any more than the stages of birth, of

hatching, of metamorphosis, etc., which can

occur at very different moments of development

from one group to another. The notion of

“ birth ", as used by Van Gelder (1977), is valid

only for mammals and other viviparous animais.

In most animais, if the adult stage is defined by

the acquisition of sexual maturity and by the

ability to reproduce, this stage may be recognized

by many other characters (size, morphology,

behavior). This allows the use of this criterion

even in the case of infertile hybrids, which are

admittedly unable to reproduce but which hâve

nevertheless reached the adult stage of develop¬

ment, which, for the reasons given above, is

enough in my opinion to consider a hybridiza¬

tion as “ successful

As concerns mammals, expérience shows that

a number of hybrids which reach the stage of

birth alive later hâve a more or less normal

growth and usually live until the adult stage:

from a purely practical point of view, the

application of the criterion of the viable adult

would entail virtually no différence as compared

with Van Gelder’s (1977) criterion of the living

newborn offspring.

Choice of the taxinomie rank

Turesson’s (1922) concept of coenospecies and

its synonyms (see Bernardi, 1980), among which

is the syngameon in the sense of Cuénot &

Tétry (1951) but not of Lotsy (1918), applies to

a group of species liable to give viable hybrids in

the laboratory, be they able or not to do so in

nature.

The few authors who hâve until now made use

of the word coenospecies (or of its synonyms)

used to designate by this term a group of species

devoid of taxinomie meaning: for them, the

coenospecies sometimes included several généra,

while in other cases several distinct coenospecies

were maintained in the same genus.

On the other hand, the above propositions

amount to saying that it would be good, because

of the great biological meaning of the nontaxi-

nomic category of coenospecies, to make the

latter coincide with the taxinomie category of

genus. The choice of this latter category could be

discussed. One could contemplate the possibility

of making the coenospecies coincide either with a

lower (subgenus) or with a higher (tribe, sub-

family, or even family) category than the genus.

In a few cases, such a proposai would hâve the

advantage of entailing less taxinomie disrup-

tions: thus in birds, where, as we shall see, the

application of this criterion would considerably

modify the generic (and, by way of conséquence,

familial and ordinal) classification, it would

appear justified to make the coenospecies coin¬

cide with the family, which would much less

modify the current classification. But, the objec¬

tive of my proposition being to standardize

systematics in different groups, what would be

“ gained ” on the side of birds, would be “ lost ”

in ail other groups, where it would be necessary

to rise généra to the rank of families: eventually,

the disruption would be the same or even

greater, but it would concern other groups than

birds.

The choice of the category genus for the

coenospecies was imposed on me, so to speak, by

a set of reasons. First of ail, an intuitive one. The

genus is the first important higher category, and

it seems logical to place at this level the first

important break above the species: species are

genetic pools protected from each other, généra

genetic units definitively isolated from each

other, but within which, at least in artificial

conditions, exchanges and relations may exist. It

is also what has been felt by ail the other authors

who hâve proposed the use of a hybridizability

criterion in supraspecific systematics (Simpson,

1961; Van Gelder, 1977; Plateaux, 1981): they

ail suggested the use of the genus, not the family

or another category, for grouping together hybrid-

izable species.

Furthermore, it so happens that the choice of

this category modifies relatively little, except in

exceptional cases like birds, the generic classifica¬

tion of many groups, as if this criterion had

already been more or less unconsciously used by

systematists since long ago. In fact the genus so

defined generally coincides well with the genus

that the other “ synthetic ” criteria mentioned

above, in particular the “ ecological ” criterion

(Inger, 1958), recognize. Actually, it is this

Source : AANHN, Paris

THE GENUS IN ZOOLOGY

agreement which first drew my attention and led

me to formulate the concept of geniation (see

below).

Let me finally note that the use of the

hybridizability criterion at the leve! of the genus

category will give rather balanced results in the

light of Van Valen’s metataxinomic criterion, as

dicussed below, while if the coenospecies was to

coincide with a higher category like that of

family, this would lead to an important imbal¬

ance in classifications, according to this criterion.

The hybridizability criterion

AND THE CLASSIFICATION OF THE VERTEBRATA GNATHOSTOMATA

Introduction

Van Valen’s (1973) metataxinomic criterion

was presented above. The application of this

criterion to the current classifications of the six

classes of Vertebrata Gnathostomata (Table II,

fig. 2) allows one to disclose the existence of

three types of classifications (Dubois, 1988):

(1) a “ balanced ” or “ symmetric ” pattern, in

which the major categories of the Linnaean

hierarchy (species, genus, family, order, class) are

roughly équidistant: only the smallest class of

Gnathostomata, that of Chondrichthyes, cur-

rently has a classification of this type;

(2) an “ overlumped ” pattern, in which the

taxa of the intermediate categories (genus, family,

order) are “ not numerous enough ”, at least

according to the scale of the metataxinomic

criterion: such a classification is observed for

amphibians and partially (only for the higher

categories, but not at the level of the genus) for

reptiles;

(3) an “ oversplit ” pattern, in which the taxa

of the intermediate categories are “ too numer¬

ous ” according to this criterion: the three classes

Osteichthyes, Aves and Mammalia présent clas¬

sifications of this type.

As was remarked above, the meaning of the

“ balanced ” or “ unbalanced ” pattern, accord¬

ing to this criterion, of the classification of a

group is far from being clear, but one may at

first contemplate two factors which may be

responsible for an unbalanced classification: (1)

mistakes in the building of the classification

(wrong weighing of characters, lack of informa¬

tion, etc.); (2) particularities proper to the mode

of évolution of the group studied.

The confrontation of Van Valen’s metataxi¬

nomic criterion with the hybridizability criterion

in the whole group of Vertebrata Gnathostomata

may bring us additional information in this field.

It is thus striking to observe that the three

classes which appear “ oversplit ” according to

Van Valen’s metataxinomic criterion (Osteich¬

thyes, Aves, Mammalia) are precisely those

which hâve the highest rate of “ intergeneric

hybrids ”, and therefore in which the application

of the hybridizability criterion would reduce the

most the number of généra and, by way of

conséquence, of other higher taxa.

Amphibians and reptiles

In amphibians almost ali hybridizations fiable occur between species which are traditionally

to give viable adults known until now (Monta- classed in the same genus. Only two examples of

lenti, 1938; Moore, 1955; Blair, 1972b; etc.) “intergeneric” hybridizations in this class hâve

Source : MNHN, Paris

ALAIN DUBOIS

been known until now: between the “ généra ”

Hyla and Pseudacris (Ralin, 1970) and between

the “ généra ” Pleurodeles and Tylototriton (Fer-

rier, Beetschen & Jaylet, 1971). In both cases

the fact of merging both généra, while conserving

the name of the second one as subgenus of the

first one (Dubois, 1982 a, 1984 b), does not raise

any particular problem, and even throws a new

light on the phylogenetic relations within these

groups. At the level of the classification of the

whole class of amphibians, these modifications

hâve virtually no effect.

The same is essentially true for reptiles, where

until now no adult hybrid is known which

appears as “ intergeneric ” according to the

current classification (Mertens, 1950, 1956, 1964,

1968, 1972; Arnold, 1973; etc.).

Now, according to Van Valen’s metataxi-

nomic criterion, the classifications of amphibians

and reptiles appear to be not very far from a

“ balanced " or “ symmetric ” type. At the level

of the genus, reptiles appear a little oversplit, and

at the other levels, a little overlumped; amphib¬

ians appear a little overlumped at the levels of

genus and of family, and much overlumped at

the level of the order. It is therefore “ logical ”, if

both criteria are congruent, that the hybridizabi-

lity criterion could be applied without leading to

an appréciable réduction in the number of

généra, or of higher taxa.

Bony fishes

The situation is very different for the classes of

Osteichthyes, Aves and Mammalia: ‘‘interge¬

neric ” hybrids are numerous in these groups,

and the grouping together of généra which the

use of the hybridizability criterion would require,

would probably hâve to be followed by a

grouping together of families and other higher

taxa, for otherwise many suprageneric taxa

would become monogeneric or almost so.

In bony fishes, the potential “ intergeneric ”

hybrids are numerous (Moenkhaus, 1910; Hubbs,

1955; McAllister & Coad, 1978; Daget, 1983).

A finer analysis shows that these hybrids are

much more abundant in freshwater fishes than in

marine fishes (Hubbs, 1955; Daget, 1983), which

can be partially accounted for by the fundamen-

tal disparities between both types of environ-

ments, in particular in terms of diversity and

stability (Hubbs, 1955), and by the fact that both

groups certainly show important différences in

their mechanisms of spéciation.

“ which resuit mainly, in continental waters, from positional isolation and

from the splitting up of ecological niches, while, in marine waters, they are

mainly based on reproductive isolation." (Daget, 1983: 401; translation

mine).

Mammals

In mammals also, potential “ intergeneric ”

hybrids are numerous (Gray, 1972). Van Gel-

der (1977, 1978) has undertaken a révision of the

generic classification of this group basing himself

on the criterion described above (an hybridiza¬

tion is considered “ successful ” when it gives at

least one viable newborn offspring). At the

moment, these works hâve led him to downgrade

44 names of généra of mammals to the rank of

subgenera or even of synonyms of other generic

names. With this operation, the number of

généra of mammals cornes down from 1004

(according to Anderson & Jones, 1976) to 960.

The conséquence of this réduction within the

framework of Van Valen’s metataxinomic crite¬

rion is shown in fig. 2: although it tends to lower

the différence between the value observed for the

position of the category genus and the “ expected ”

value in the case of a balanced classification, this

réduction is slight and the mammals remain

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

appreciably oversplit even after this action.

Three explanations of this phenomenon may be

suggested, which are probably ail partially valid:

(1) There remains certainly other cases of

potential “ intergeneric hybridization ” in mam-

mals, which were not yet known at the time of

Van Gelder’s (1977, 1978) reviews. In particular

it is likely that artificial hybridizations, in condi¬

tions allowing one to avoid the pre-ejaculatory

and post-ejaculatory mechanisms of isolation

(artificial fertilization and in vitro culture of the

embryo in particular) will allow the discovery of

many other potential “ intergeneric ” hybridiza¬

tions in mammals. When these are taken into

account, the number of généra of this class will

keep on decreasing.

(2) As we hâve seen above, the criterion of

hybridizability is only one of the criteria which

may be used to recognize généra as genetic,

phylogenetic and ecological units. In the light of

these various criteria, it is likely that the number

of généra of mammals would decrease even

more.

(3) As we hâve also seen, there is at the

moment no serious reason to believe that

a “ balanced ” or “ symmetric ” classification

according to Van Valen’s metataxinomic crite¬

rion would be “ better ” or “ more natural ”

than another one. On the contrary, a departure

from this “ balanced ” pattern may correspond

to a reality, and be the indication of the existence

of certain particularities proper to the group

studied. It may in particular express the fact that

the characteristics of the évolution of this group

are atypical as compared to those of the related

groups. As concerns mammals, evidence exists

that the group has experienced a particularly

rapid évolution of the Systems of genetic régula¬

tion, of morphology, of karyology, and of the

loss of the ability to hybridize between related

species (Maxson, Sarich & Wilson, 1973; Wil¬

son, Maxson & Sarich, 1974; Wilson, Sarich

& Maxson, 1974; King & Wilson, 1975; Wil¬

son, 1975; Wilson et al., 1977; Wilson, Carl-

son & White, 1977; Cherry, Case & Wilson,

1978; Cherry et al., 1979, 1982; Bengtsson,

1980; Larson, Prager & Wilson, 1984; etc.). It

is therefore likely that the “ oversplit ” pattern of

the supraspecific classification of this group

corresponds at least in part to the reality.

This example allows one to sense in concrète

terms the interest of the use of standardization

criteria, like the hybridization criterion: the fact

that, even after the use of this criterion (and of

other synthetic ones), the classification of a

group remains “ atypical ” as compared to the

“ mean ” or “ balanced ” classification, or to the

classifications of neighboring groups, will be

fiable to draw the attention on particularities

proper to the évolution of this group, and to

stimulate research on the evolutionary mecha¬

nisms responsible for these disparities. This will

not be possible if one does not possess any

criterion allowing one to refer ail classificâtions

to a common yardstick.

BlRDS

The classification of birds will give us a

négative example confirming this interprétation.

The number of “ intergeneric hybrids ” in this

class is extremely high (Gray, 1958; Prager &

Wilson, 1975; Milstein, 1979; etc.). The use of

the hybridizability criterion would entail a radi¬

cal change in the systematics of this class, and

particularly of some of its families, like that of

Anatidae (Johnsgard, 1960), where the number

of “ intergeneric ” hybrids is very high. Would

such a modification be disastrous, as certain

ornithologists seem to believe, or would it corre¬

spond to a real need?

The fact that, as compared to other groups,

the classification of birds is much oversplit, has

already been emphasized on several occasions

(see e.g.: Sibley, 1957; Crowson, 1970; Prager

& Wilson, 1975; Bock & Farrand, 1980;

Dubois, 1982 a ; Sibley & Ahlquist, 1982; Pas¬

teur, 1985). A noticeable effort of réduction in

the number of supraspecific taxa of this class,

which was extremely high at the beginning of the

Source : MNHN, Paris

ALAIN DUBOIS

century, has already been made by ornitholo- number of généra is much too high, as was

gists. The fact remains that the current classifica- specially emphasized by Crowson (1970):

tion is much oversplit, and in particular that the

“ We cannot help feeling that students of birds and of moths would be

better zoologists and better systematists if, despite the difficultés, they

seriously tried to observe and appreciate the generic characters in their

animais. ” (Crowson, 1970: 51).

Il is interesting to observe that numerous are

the ornithologists who, reporting upon the dis-

covery of natural hybrids or the obtention of

artificial hybrids between species of birds classed

in different généra, expressed some doubts as to

the validity of the séparation of these généra.

However, with the help of the strength of

tradition, they generally merely formulated these

doubts in the Discussion of their work, without

going so far as to group together the species of

both généra in a single one, as may be illustrated,

without any concern for exhaustivity, by the

following citations drawn from papers dealing

with “ intergeneric ” hybrid birds:

“That these two species should be considered members of different

généra, in the light of the présent evidence, seems open to question. ”

(Williamson, 1957: 122).

“ a serious study of the generic limits in the Trochilidae is in order. ”

(Banks & Johnson, 1961: 26).

" The discovery of this new intergeneric North American hybrid hum-

mingbird combination (...) lends additional support to the oft-expressed

view (...) that the time is ripe for a thorough study of the generic limits

within the Trochilidae. ” (Lynch & Ames, 1970: 212).

“ The existence of the hybrid, and its mating with T. verticalis, emphasize

the close relationship between T. verticalis and M. forficata and support the

proposai advanced by Smith (...) that M. forficata be placed in the genus

Tyrannus. " (Davis & Webster, 1970: 42).

“ it is clearly apparent that serious considération should be given to

merging the généra Lophortyx and Callipepla with Colinus.” (Johnsgard,

1970: 87).

“ Because of the general morphological similarity of swallows, Mayr and

Bond (...) questioned the reality of generic limits in this family and

suggested that grounds for separating Petrochelidon from Hirundo were

particularly weak. The occurrence of hybridization between Hirundo and

Petrochelidon and the biochemical evidence of close génie similarity between

H. rustica and P. fulva strongly support this view-point. ” (Martin &

Selander, 1975: 364).

One may wonder why the classification of

birds is so oversplit. One reason is certainly the

fact that this class has been the subject of a very

high number of works: as a matter of fact it is

very frequent that very well known groups are

excessively divided as compared to the less

studied neighbouring groups (see in this respect

Crowson, 1970: 48-49). On the other hand many

généra of birds, in particular among diurnal

species, “ rest ” on characters of the plumage,

and often of the plumage of the males only. The

importance attributed by systematists to these

characters is certainly in relation with the fact

that these are very visible, sometimes spectacu-

lar, characters, and that man, a species in which

sight is more developed than the other senses,

tends to give greater importance to characters

accessible to this sense than to others.

During the round table of the French Zoologi-

cal Society on “ Genus, subgenus and species-

group ” (Paris, 14 March 1978), Philippe Dreux

insisted upon the fact that the systematics of

birds would certainly be much less divided if

abstraction had been made of the feathers to

build it. Concerning pheasants, among which the

known “ intergeneric ” hybrids are numerous

(see Gray, 1958), he humorously summarized

this observation: “ Pluck them, and no one will

recognize them, even by their taste! ” (Dubois,

1982 a: 32).

The evolutionary meaning of the important

différences in the plumage of males which is

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

often observed between species of diurnal birds,

which are in other respects very close, is very

clear: these are pre-ejaculatory mechanisms of

isolation allowing the avoidance of the hybrid¬

ization of these species in sympatry. These are

therefore characters related to spéciation, not

characters expressing a more important diver¬

gence. The rôle played by the différences of

plumage (and also of mating call, of nuptial

parade, etc.), in diurnal birds is played by other

characters in other groups of animais. Thus in

anuran amphibians the mating call of males

plays a fundamental rôle in the pre-ejaculatory

isolation between species: in these animais it is

frequent to encounter species morphologically

identical or very similar but having very different

mating calls. A classification of anurans which

would give mating calls an exaggerated impor¬

tance, comparable to that sometimes given plum¬

age characters by ornithologists, could lead to

classifying these species in different généra. The

same would be true with a classification of

micromammals which would give a great impor¬

tance to olfactory criteria.

We are indebted to Sibley (1957) for an

interesting paper where ideas close to the pre-

ceding ones are expressed in a more detailed

way, and where this author most justly writes:

“ The high incidence of monotypic généra in groups of sexually dimorphic

visual animais is due to erroneous human évaluation of the taxonomie value

of signal characters. Morphological structures evolved under the sélection

pressure of deleterious hybridization and/or sexual sélection seem highly

‘ specialized ’ to the intelligent discrimination of the human taxonomist who

therefore accords them generic rank on a ‘ degree of différence ' basis. This

is a coincidental resuit of the fact that we too are visual animais and hence

can and do utilize visible characters in taxonomy. It is significant that

‘ intergeneric ‘ hybrids are found almost exclusively in visual animais,

principally birds and, to some extent, fish. It is apparent that généra in such

groups should not be based only upon secondary sexual characters nor upon

characters which hâve been reinforced by sélection against hybrids since

these, inevitably, are species characters. ” (Sibley, 1957: 187).

It seems therefore that the réduction of the

number of généra of birds which would be

entailed by the use of the hybridization criterion

proposed above would be a salutory operation:

généra thus defined would hâve much more

biological meaning than the numerous monospe-

cific généra which are currently based on plum¬

age characters or on other characters expressing

a simple divergence between sympatric related

species. It is likely that the réduction in the

number of généra of birds, if it was accepted by

ornithologists, would be followed by an impor¬

tant réduction in the number of families and

orders of this class. Moreover, the whole current

classification of birds seems still susceptible of

important modifications, despite the numerous

works which hâve already been devoted to it. It

is in particular possible that such modifications

become necessary as a resuit of the reassessment

which seems to be necessary of some aspects of

the phylogeny of this group (see Cracraft,

1972).

In recent years very interesting works hâve

been devoted to studies of the molecular évolu¬

tion of birds, and hâve led to the rather sur-

prising conclusion that divergence, at the level of

the structure of proteins, between lower taxa of

birds is extremely weak as compared to the

divergence which exists between numerous other

vertebrates of similar taxinomie levels (see e.g.:

Prager et al., 1974; Avise & Aquadro, 1982;

Aquadro & Avise, 1982; Avise, 1983; Pasteur,

1985; Viot, 1985); it is similarly so for the

divergence at the level of the sequence of mito¬

chondrial DNA (Kessler & Avise, 1985). Sev-

eral hypothèses hâve been put forward to account

for these observations, among which the most

often mentioned and discussed (see e.g.: Zink,

1982; Avise, 1983; Kessler & Avise, 1985) are

the two following ones: (1) the taxa of birds

studied would hâve a more recent origin than the

taxa of the other groups; (2) molecular évolution

would be slowed down in birds as compared to

other vertebrates:

“ One possibility is that protein évolution is decelerated in birds: the

protein ‘ clock ' may tick at a slower pace.” (Avise, Patton & Aquadro,

1980: 303).

Source : MNHN, Paris

ALAIN DUBOIS

A third hypothesis, which also deserves con¬

sidération (Avise & Aquadro, 1982; Sibley &

Ahlquist, 1982; Viot, 1985), is precisely that

according to which the supraspecific classifica¬

tion of birds is oversplit.

Once again, we here face the practical interest

of having a criterion of standardization like the

hybridizability criterion: in the absence of such a

criterion, it remains rather gratuitous to discuss

the possible accélération or décélération of the

molecular évolution rate in an animal group.

This was well emphasized for example by Sibley

& Ahlquist (1982), who strongly feel the neces-

sity of such a criterion of standardization of the

different classifications. Unfortunately, the crite¬

rion proposed by these authors (the âge of taxa,

as it may be estimated by DNA hybridization) is

not able to play this rôle well, for the reasons

detailed above.

As we hâve seen, according to Van Valen’s

metataxinomic criterion, the current classifica¬

tion of birds appears to be “ oversplit ”, which

supports the preceding remarks. It would be

most interesting to construct, at least for infor¬

mation only, a new supraspecific classification of

birds where would be grouped together two by

two ail the généra with two species at least being

liable to hybridize, then where the number of

families and orders would be reduced according

to these groupings. Upon examination of the

lists of hybrids of birds (Gray, 1958) and of

mammals (Gray, 1972) currently known, it is

clear that the réduction in the number of généra

entailed by the use of the criterion of hybridiza¬

bility would be much more drastic in the former

than in the latter. The classification of birds

would become, after such an operation, certainly

much doser to a “ balanced ” classification

according to Van Valen’s metataxinomic crite¬

rion than that of mammals after the lumpings of

généra realized by Van Gelder (1977, 1978) (see

Table II and fig. 2). Then, it would be possible to

validly test the hypothèses mentioned above on

the différences between the evolutionary rates of

the Systems of genetic régulation between differ¬

ent groups.

CRITICAL STUDY OF THE USE

OF THE HYBRIDIZABILITY CRITERION TO DEFINE GENERA

Several arguments can be put forward for or these were already discussed above. Some others

against the use of the above defined criterion to remain, which we shall now examine,

group together species in a same genus. Some of

Some arguments against the use of this criterion

(1) A first argument consists in saying that

this criterion cannot always be used. In certain

cases, it cannot be used because of intrinsic

properties of the compared species: thus the

criterion cannot be used in paleontology, nor for

living species with uniparental reproduction (spe¬

cies with a true asexual reproduction; species with

uniparental reproduction derived from bipar-

ental sexual reproduction: autofertilization, par-

thenogenesis, etc.). In other cases, the criterion

cannot be used for purely material reasons: in

many groups of animais, breeding is difficult,

artificial insémination cannot be achieved as

easily as in amphibians or echinoderms, and it is

therefore very difficult or impossible to study

hybridization in the Iaboratory.

In reality an argument of this type could be

used against most of the methods used in

systematics. In modem systematics data are used

which corne from morphology, anatomy, bio-

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

chemistry, karyology, ecology, behaviour, bio-

acoustics, parasitology, etc. It is quite rare that

ail these data should be available for a given

group and, in paleontology, the only available

data are those derived from the study of the

fossilized parts of animais. However no author

has ever suggested, at least let us hope so, to use

in systematics only the method which is the

smallest common denominator to ail possible

studied, and ever proposed to base ail animal

systematics on the study of the sole fossilizable

parts! Systematics make use of the highest

possible quantity of information concerning

living beings. In certain cases the information is

rich, in others it is less so, but it is always

désirable to hâve as much information as pos¬

sible. The criterion of hybridizability can cer-

tainly not be used in ail cases, but this does not

forbid its use when it is possible.

Furthermore, one may hope that in some of

the groups where this criterion cannot be used at

the moment for simple material reasons, the

progress in our knowledge of the biology of these

animais (including in particular the achievement

of breeding and of artificial insémination) will in

the future allow us to use it.

As concerns the species with uniparental repro¬

duction, everything dépends on their more or less

isolated or excepti mal nature. When only a few

species are concerned, e.g. with parthenogenetic

reproduction, within a vast group most of the

species of which still use sexual reproduction,

analogies with the latter will sometimes help in

building up the classification, including at the

generic level. On the other hand, in groups where

the rule is uniparental reproduction, as e.g.

bdelloid rotifers (De Beauchamp, 1965), such a

resort to analogy is hardly possible and one must

admit not to be able to define généra by using,

even in an indirect way, the criterion here

proposed.

(2) This criterion may be blamed for its

asymmetry: it only takes into account the posi¬

tive results of hybridization and can therefore be

used to group species together within a genus,

not to sépara te généra.

Such an asymmetry is the fact of many other

criteria of current use in systematics. To give

only one example, one of the criteria which may

be used to ascertain that two different popula¬

tions belong to two distinct species is based on

the fact that hybridization between individuals of

these two populations is impossible or always

leads to a failure of development. In this case an

absolute genetic isolation exists between both

populations, and by définition these cannot

belong to the same species. On the other hand

the reverse resuit does not at ail allow one to

draw the reverse conclusion. The ability of two

populations to give birth to hybrids between

them, even sometimes in nature, does not at ail

imply that they belong to the same species.

Hybrids may occur in nature sometimes in the

zone of hybridization between two subspecies of

a same species, sometimes in the zone of over-

laping and hybridization between two prospecies

of a same superspecies, and finally sometimes as

isolated hybrid individuals, in a zone of wide

sympatry between two good species. In ail these

cases, what will allow one to choose between

these different possibilities are arguments other

than the simple presence of hybrids (see e.g.

Dubois, 1977 b; Bernardi, 1980). This is here

also an asymmetrical criterion, which does not

prevent it from being very useful where it can be

used.

(3) Another objection to the use of this crite¬

rion is that its adoption would entail important

modifications in the systematics of certain groups.

The importance of these changes would be

extremely variable according to the group con-

sidered, as was shown above by the examples

taken in the vertebrates. I discussed in a rela-

tively detailed manner the problem of the birds,

because this class is probably the one where is

posed with the highest acuteness the problem of

the disruption of the classification consecutive to

the application of the new criterion. The argu¬

ments developed above, or other similar ones,

are also applicable to other groups, where

“ intergeneric ” hybrids are numerous.

Be that as it may, it is not exceptionally that

the introduction of new arguments entails modifi¬

cations in the systematics of a group, and these

arguments cannot be rejected under the sole

pretext of “ preserving the stability of nomencla¬

ture The stability of nomenclature and of

classification is certainly désirable in general, as

long as new information does not contradict the

tradition, but it should not be a brake on the

improvement of systematics which is sometimes

Source : MNHN, Paris

ALAIN DUBOIS

demanded by the progress of our knowledge of

the living beings.

As concerns the groups, like birds, where the

application of the criterion of hybridizability

would lead to important changes at the generic

level, it might be advisable, at least as a provisional

measure, to conserve the very well known generic

names as subgeneric names.

While the above discussions are mainly based

on purely formai arguments and for this reason

seem to me of little importance, the last two

objections which I will consider touch on real

biological problems and are more interesting. In

the current State of our knowledge, they do not

seem to prove justified, but we must nevertheless

examine them.

(4) The first objection bears on the interpréta¬

tion which I hâve adopted here of the genetic

meaning of the success of the hybridization

between two species. Following other authors

(Whitt, Childers & Cho, 1973; Wilson, Max-

son & Sarich, 1974; Whitt, Philipp & Chil¬

ders, 1977; Wilson, Carlson & White, 1977;

Oliver, 1979; Philipp, Parker & Whitt, 1983;

Parker, Philipp & Whitt, 1985 a, 1985 b; etc.),

I hâve here admitted that the success of the

development of a hybrid until the adult stage

expresses a strong similarity and a compatibility

of the Systems of genetic régulation of the two

hybridized species. Another interprétation could

be considered: that according to which only one

of the two Systems of genetic régulation présent

in the hybrid would in fact be active. If it so

happened that the genome of one of the two

species was totally inactivated (repressed) in the

hybrid, the latter would correspond from the

viewpoint of its active genetic material to a

haploid or parthenogenetic individual, and the

criterion of hybridizability would lose the funda-

mental biological meaning which was attributed

to it here.

The known facts do not seem at ail to support

this hypothesis. In some hybridizations between

relatively distant species, evidence exists that

certain structural genes of one or the other of

both parental stocks are inactive, because of

phenomena of repression , but the repression on

one hand only touches a limited proportion of

genes, and on the other hand concerns some-

times the maternai, and sometimes the paternal

alleles (see e.g. Whitt, Childers & Cho, 1973),

which indicates that both genomes take part, at

least partially, in the ontogenesis. In the case of

the inactivation of the genes situated on one of

the two X chromosomes of mammals, the study

of certain hybrids, some Canidae and some

Equidae, shows that it is sometimes the maternai

X, and sometimes the paternal X which is

inactivated (Serov, Zakijan & Kulichkov,

1978 a, 1978 b). Discussing the results of a study

bearing on hybrids of teleosts, Whitt, Childers

& Cho (1973: 59) Write:

“ These results and those of previously published studies support the

postulate that there is a positive corrélation between the evolutionary

distance of the parental genomes and the extent of allelic repression in the

Fl hybrid. ”

Thus, in the hybrids between very close spe¬

cies, there may exist no allelic repression at ail

(see e.g. Champion & Whitt, 1976). On the

other hand, it seems that when the divergence

between the two genomes becomes too great,

rather than a complété repression of one of the

two and a “ normal ” development due to a

single genome, what occurs is a failure of

development. It will be important in this respect

to follow the future works on génie expression in

hybrids, but in the current State of knowledge

this objection does not seem to be relevant.

(5) The last objection is the following one:

might not the use of this criterion lead to the

grouping together in a same genus of organisms

liable to hybridize step by step and constituting a

“ chain ”, so to speak, the extreme links of which

would be extremely dissimilar? Such a situation

would occur if hybridization was a success

between A and B, then between B and C,

between C and D, and so forth without interrup¬

tion. If this was the case, the whole classification

might corne apart like knitting, to lead to the

maintenance of only a few généra within each

great group!

In front of this theoretical hypothesis, only

expérience can answer. Now, the examination of

lists of species liable to give between them adult

viable hybrids (thus, in the vertebrates: Suche-

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

tet, 1897; Montalenti, 1938; Mertens, 1950,

1956, 1964, 1968, 1972; Hubbs, 1955; Moore,

1955; Gray, 1958, 1972; Blair, 1972 b) shows

that successful hybridizations allow one in reality

to define relatively small hermetic groups, sepa-

rated from other similar groups by discontinui-

ties, and not open chains. Therefore, by calling

upon this criterion, it would appear that généra

are closed communities, natural units, just like

species, although in a different manner.

SOME PRACTICAL ARGUMENTS IN FAVOR OF THE USE OF THIS CRITERION

The use of the hybridizability criterion to

group species together in a same genus is of great

theoretical and practical interest.

I already discussed at length the theoretical

aspects of this question. A few purely practical

arguments in favor of the use of this criterion

must also be mentioned.

(1) Généra recognized according to this crite¬

rion will probably be a little larger on the

average than they are currently, i.e. they will

include a higher average number of species. In

many groups, where the excessive number of

généra has already been emphasized by many

authors (e.g. Mayr, 1943; Rosen & Bailey,

1963; Crowson, 1970), such a change would be

most welcome:

“ The désirable trend now would be to reduce large numbers of currently

accepted généra to the level of subgenera or even species-groups (...), and at

least the idealists among us may hope that a change so clearly in the

interests of the scientific majority is almost bound to corne about. ”

(Crowson, 1970: 298).

(2) Although this criterion has already been

mentioned by some authors and used in a few

cases, no systematic attempt to use it to redefine

généra within a given group has so far been

made, except for that of Van Gelder (1977,

1978) in mammals. As we hâve seen, the changes

that such an operation would bring would be of

a very variable scope from one group to another,

e.g. very limited in amphibians and very great in

birds, which is certainly not liable, despite the

arguments in favor of this proposai presented

above, to lead specialists of groups like birds to

be enthusiastic about it! However it must be

insisted upon that the taxinomie disruption

would occur once and for ail and that, once it has

occurred, the generic nomenclature of the group

would be very much stabilized. The hybridizabi¬

lity criterion, if it is used appropriately, avoiding

the few pitfalls pointed out above, is a “defini¬

tive ” criterion, which will never hâve to be

reconsidered later: two species liable to give

viable adult hybrids will remain in the same

genus, independently of ail other arguments

concerning their morphology, their biology, etc.

For many cases in ail the groups where generic

status is currently a matter of discussion but

where viable adult hybrids do exist, such a

stabilization will be welcome: it will stop nomen-

clatural comings and goings between several

generic names for a given species. Despite an

important initial disruption in some groups, the

use of this criterion would in the long run hâve a

strong stabilizing effect on generic classification

and nomenclature in zoology.

(3) This criterion is of a relatively easy and

“ economical ” use, since the discovery of a

single hybridizable pair may lead to the merging

of two généra even if these contain a much

higher number of species.

(4) Finally, while in some groups the use of

this criterion is difficult for material reasons, in

other ones it is easier than long morphological,

molecular, ecological analyses. In some groups

where the studies of these last types are progress-

ing slowly, the use of this criterion should

contribute to a rapid stabilization of the generic

nomenclature, while allowing of course the contin¬

uation of more detailed studies on the other

aspects.

Source : MNHN, Paris

ALAIN DUBOIS

The criteria of the genus

As was emphasized by Crowson (1970: 48-

49), the specialists of a particular group, who

know it well and appreciate ail its subtleties,

often tend to give it a greater importance and to

subdivide it to the maximum, to recognize in it

many hierarchized subgroups, and often later to

elevate the ranks of the latter (as compared to

the neighbouring groups which hâve been the

subject of less detailed works). It is important

always to try to put “ one’s ” group back in the

general context, to allow as much as possible

classification to play its universal information

rôle. The principles and criteria discussed above

may be somewhat useful in this respect.

1 insisted particularly on the hybridizability

criterion, because it is new and its application

would be followed by appréciable modifications

in the current classification of many groups.

However, it is clear that this criterion cannot,

and must not, be used alone to identify généra

and build up classifications. It must be used

within the framework of the “ synthetic con¬

cept ” of the genus as it has been characterized

above, and in conjunction with the other criteria

available within this framework. By the way,

several of these criteria hâve already been used

for a long time by many systematists.

As has been shown elsewhere (Dubois, 1977 b),

there exists a certain hierarchy among the criteria

which allow one to décidé if two sets of popula¬

tions are or not distinct species, some criteria

being more important, more conclusive than

others:

“ The species concept (protected gene pool) is a synthetic concept and in

this field the use of a single criterion is often not enough to reach definite

conclusions. However, the joint considération of several characters (...) often

allows one to remove difficulties. It is particularly important to dispose of

data on several independent characters, and to ascertain whether they reach

similar conclusions or not. In practice it is this joint use of several

independent characters which permits, in many cases, decision. It is thus in

general useless, from a practical point of view, to dispose of data on a high

number of characters. The combinations of characters which may be used

are very diverse and it is not useful here to give examples. However there

exists a certain hierarchy among criteria, which may be briefly summarized

as follows.

The criterion of genetic compatibility (which may be demonstrated, or

deduced from other considérations, e.g. from karyology) is indeniably the

surest criterion of the existence of two species (apart from the réservations

made above on this question). In the cases of genetic compatibility, it will

first be necessary to study the spacio-temporal relationships between both

forms, and to ascertain whether they are sym-, para- or allopatric (or

-chronic). In sympatry, ail the criteria showing a marked discontinuity

between both groups can be used to indicate that two separate gene pools do

exist; it will be valuable in this case to use independent criteria (morphology

of adults and larvae, mating calls, biochemistry, ecology, etc.). In parapatry,

the study of the detailed distribution of both forms, of the hybridization and

introgression in the contact zones take a particular importance. In allopatry,

the joint examination of various independent characters will again be most

useful. The higher the number of independent characters for which a

divergence between both stocks will hâve been demonstrated, the clearer it

will be that the genetic divergence between these stocks is high, and

therefore that the process of différentiation or of spéciation is advanced.

However, even so, in many cases it will be impossible to conclude: only a

field expérimentation, putting in contact populations which hâve been

separated by natural obstacles and which hâve diverged, would allow us to

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

know how they would behave then, and to décidé if they are subspecies or

species. No laboratory work will ever furnish an answer with certainty to

this type of question, and such experiments are difficult to do with anurans,

which are a little too big to be bred in demometers! The experiments of this

kind realized by Twitty (1961, 1964, 1966) on urodelans of the genus

Taricha are, to the best of my knowledge, the only ones of this type to hâve

been made in amphibians. ” (Dubois, 1977 b: 234, translated).

Similarly, the adoption of the genus concept

here advocated implies the acknowledgement of

a hierarchy in the use of the criteria presented

above.

First, the fact that two species A and B are

liable to produce viable adult hybrids is an

absolute and definitive proof that both species

possess very close functional genetic characteris-

tics and must therefore be grouped together in a

same genus (nonarbitrary criterion for inclusion).

Given the complexity of Eucaryote genome, one

may without hésitation exclude as completely

impossible that such a genetic similarity could be

obtained by convergence between two species of

two phylogenetically distinct groups, and this

criterion of hybridizability can therefore be also

considered as a criterion of homophyly. But to

entirely satisfy this criterion, the species A and B

must belong to the same genus as their most

recent common ancestor, which leads to group-

ing in this genus ail the other species which, by

other criteria (homophyly, morphological and

ecological resemblance, etc.) were previously

classed in the same genus as A and in the same

genus as B.

In the absence of successful hybridization, this

criterion cannot be used in a négative way. It will

then be useful to compare the holomorph of the

studied species, to ascertain whether discontinui-

ties exist or not within the group in question.

The presence of such marked holomorphological

discontinuities, whatever their “ size ”, provided

they correspond to characters for which the

supposed genetic determinism is complex and

irréversible in the strict sense of the term, is a

good argument for considering that several

généra do exist (nonarbitrary criterion for exclu¬

sion). The groups which remain must finally be

submitted to a cladistic analysis. If this analysis

demonstrates the existence of phenomena of

parallelism or of convergence, the existing poly-

phyletic groups in their turn must be broken up

(nonarbitrary criterion for exclusion ), to leave

only homophyletic (i.e. holophyletic or paraphy-

letic) groups.

The criteria of morphological and ecological

resemblance and of homophyly must always be

used with caution for inclusion , because real

morphological or ecological différences, as well

as real convergences, may always escape analysis

when the available information is insufficient

(see e.g. in amphibians: Maxson & Wilson,

1974; Maxson, 1977; Fouquette & Delahous-

saye, 1977). The criteria of inclusion, except that

of hybridizability, are less reliable in general than

those of exclusion and here the expérience that a

systematist has of the group he studies takes ail

its importance.

Defined by this set of criteria, généra may be

of very variable “ sizes ”, some being monotypic

while others containing very numerous species. It

is therefore very useful to recognize taxinomie

subunits below the genus. We shall examine

them in more detail below, but a few words may

be said here already.

The récognition of taxinomie subunits within

the genus is mainly based on the type of

divergences which exist between the different

natural groups which phenetic analysis allows

one to recognize. When these groups show

between them appréciable ecological différences,

without for ail that being separated by discontin¬

uities, they should be given the status of subgenera,

while groups which do not show between them a

marked ecological différentiation will be considered

as species groups (and possibly, more finely, as

species complexes, synkleptons, superspecies or

ultraspecies). The subgenus category may also be

used in some cases to conserve at least provisio-

nally old well-known generic names when the

older généra hâve been merged to satisfy the

above criteria.

Source : MNHN, Paris

ALAIN DUBOIS

Conclusion

In his interesting work on cladism, Dupuis

(1979: 52, translation mine) writes:

“ As a matter of fact, it is obvious that the current dispute of so many

classical ideas in biology cannot be reduced to a simple affair of opinion and

could never hâve been the fact of taxinomists atone, be them hennigians or

others. It results, before everything, from the considérable contemporaneous

progresses of experimental biology, of paleogeography, of paleontology. For

this reason, 1 am persuaded that the convincing light in phylogenetic

taxinomy will corne from new experimental facts. Not long ago, to speak of

expérimentation concerning phylogeny might hâve been regarded as impos¬

sible. Today, immunotaxinomy, enzymotaxinomy, molecular hybridization,

genes-structures relationships, ontogenetic régulations and epigenetic

amplification hâve become experimentally accessible (it is roughly the

'experimental systematics ' of Crowson, 1970: 296). More numerous data

in these fields will further modify our views of évolution. There is only to

await the taxinomie constructions which they will impose on us. ”

Although this author fails to mention inter-

specific hybridization among modem and inter¬

esting methods, the hybridizability criterion here

advocated to recognize généra is typically a

criterion of this “ experimental systematics ”

which belongs according to Crowson (1970:

292) to the “ future of systematics ”. Obviously

the application of criteria of this type will not be

possible without appreciably modifying existing

classifications. Let us hope that, despite this

difficulty, the new criterion will be taken into

considération by taxinomists, and that it will

escape the pessimistic prédiction of Sibley &

Ahlquist (1982: 14):

“ it may take a génération or two of systematists to win acceptance. 1

Source : MNHN, Paris

GENETIC REVOLUTION AND GENIATION:

THE GENUS AS AN EVOLUTIONARY UNIT

Phylogeny and ontogeny

Biology is not a unified science as yet. There

runs through it the fundamental divorce between

what Jacob (1970: 14-15) calls the “ integrist

or evolutionary ” attitude and the “ tomist or

reductionnist ” attitude, which has played a great

rôle, not only in the recent history of biology,

but also in that of modem society (see for

example Commoner, 1969, in particular chapter

III). Today’s reality is that the second attitude is

the prevailing one, and biology suffers from a

radical division into varied “ disciplines ”, which

are often completely separated from each other,

or nearly so, and which use different concepts, so

that the “ specialists ” can hardly share their

expériences, their knowledge and their problem-

atics. As a matter of fact, the différentiation into

a certain number of disciplines was historically

necessary to let the “ science of life ” blossom

forth: one had to clarify the concepts, to refine

the methods for the study and understanding of

biological reality at its various levels of intégra¬

tion (molécule, cell, tissue, organ, individual,

population, ecosystem). However, for a great

many “ specialists ", those “ disciplines ”, which

had been artificially set up in order to render the

study of extremely complex phenomena easier,

or even simply possible, hâve finally become

“ sciences ” as such. Nothing can be more dan-

gerous than this attitude for the future of

biology. Fortunately, a salutary reaction against

it is now developing, and some biologists try to

restore a comprehensive, synthetic approach to

biology that takes into account ail the spécifie

attainments contributed by each of these disci¬

plines: works such as The Growth of Biological

Thought by Mayr (1982 a) or the Traité du

Vivant by Ruffié (1982), testify to the reality of

this movement. Such attempts at synthesis, even

though they cannot but remain incomplète and

imperfect for the time being, can only be carried

out within the framework of an evolutionary

conception of biological facts, and it is only in

such a perspective that the unity of biology may

eventually be reestablished.

Although the “ synthetic theory of évolution ”

has been discussed for a long time, the science of

évolution itself has long remained a discipline

separated from the other disciplines of biology,

and the synthesis is not complété yet. There still

remains today a wide gap between the approach

of the study of évolution through population

genetics on one hand, and the study of macro-

evolutionary phenomena which refers in particu¬

lar to the recent notions concerning the genes of

régulation on the order hand: this gap clearly

shows for instance in the complété absence of

any link between the two parts of the book

Evolution published by Hermann about ten years

ago (Petit, 1976; Zuckerkandl, 1976 b; see

Dubois, 1982 b: 372-373). The synthesis in this

field has only just started, with works such

as Ontogeny and Phylogeny by Gould (1977),

Macroevolulion by Stanley (1979), or Embryos,

Genes and Evolution by Raff & Kaufman (1983).

The latter authors offer an interesting historical

account that makes it possible to understand

how the divorce between genetics and embry-

ology came about at the beginning of the

Source : MNHN, Paris

ALAIN DUBOIS

twentieth century. The two disciplines are now

becoming reconciled, in particular around the

concept of genetic and developmental régulation,

thus opening the way to a new field of research,

viz. the genetics of development, which is of

utmost interest for ail biologists and especially

for those who are trying to understand the

modalities of évolution. As for Stanley (1979),

he proposes a brief historical account of the

divorce that long separated paleontology from

the study of the mechanisms of évolution. It is

interesting to note that in both cases a large rôle

was played by the “ setting-off effect ” of Gold-

schmidt’s (1940) théories on the part acted in

évolution by “ hopeful monsters due to the

blatantly erroneous nature of the genetic model

proposed by this author, the “ modem synthe-

sis ” of the theory of évolution has rejected not

only that model, but also the indéniable evolu-

tionary reality that had inspired it, namely the

fact that évolution proceeds at least partly by

sudden phenomena, what Simpson (1944, 1953)

has named quantum évolution; in spite of some

“ prophetical ” articles (among which, that of

Mayr, 1954, in particular), it is only recently

that the importance of this type of évolution,

and above ail the fact that it is closely related to

the phenomena of spéciation, hâve really been

perceived.

It is surprising that, apart from a handful of

isolated searchers, so many biologists should

hâve been interested in morphology and its

évolution within groups of animais without

considering the processes of morphogenesis. Now,

the adult forms of living beings that systematists

compare between one another no doubt are the

products of evolutionary processes (phylogene-

sis), but they are also the results of processes of

development (ontogenesis). They can therefore

not be compared with each others as objects, or

“ completed products ”, can, without their growth

being taken into account.

It is now clear that the évolution of the

morphology of adults can be grasped only

through the évolution of morphogenetic pro¬

cesses. The recent works on the biology of

development (see e.g. Raff & Kaufman, 1983)

hâve revealed a certain number of fundamental

processes the understanding of which calls for

the notions — that are sometimes old but that

one is only beginning to perceive clearly — of

genetic régulation, canalization, induction, molec-

ular and cellular interactions, pleiotropy, epis-

tasy, etc. The ontogenesis of an individual now

appears like a chain of interdependent processes

influencing each other and following each other

in sériés, etc. Any disruption in one of those

processes (e.g. any change in a growth rate) may

hâve great conséquences as to the morphology of

the adult, provided it remains compatible with

the life of the animal. The morphological changes

will be ail the more important as the disrupting

action has taken place at an earlier stage of

development, as the whole chain of interactions,

inductions, etc., posterior to that stage will be

modified in conséquence. It is therefore easy to

understand why simple genetic alterations, bearing

upon few genes or even a single regulatory gene,

may give birth to a new adult morphology:

“ Macroevolutionary changes in development need not be extreme. We

propose that in fact the initial steps for rapid, and ultimately. large

evolutionary transitions require only that key regulatory genes be few in

number and accessible to nonlethal genetic alterations in their functions.

Initial. ' easy ’ genetic changes, which may hâve significant effects on the

organism and become established in a small population, are of necessity

viable, and présent open avenues for sélection of successive genetic changes.

Profound change may be rapid in this way without recourse to any

instantaneous hopeful monsters. ” (Raff & Kaufman. 1983: 163).

Among the evolutionary mechanisms that are

beginning now to be well-known and that enable

such spectacular alterations at little cost (in

terms of mutation), let us mention the genes

having pleiotropic effects, the mutations that

hâve conséquences as to the rate of development

or the sexual maturation (aneuchrony), and the

homoeotic mutations (see e.g.: Ouweneel, 1976;

Gould, 1977; Dubois, 1979 b, 1987 a; Raff &

Kaufman, 1983). The existence of genes having

pleiotropic effects, for instance, has been known

for a long time, but that has not prevented a

great number of theoreticians of évolution from

using the “ one gene, one character ” postulate.

Today, two types of pleiotropic effects are

recognized, viz. direct pleiotropy and relational

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

pleiotropy, the study of which is rich in informa¬

tion (see e.g. Raff & Kaufman, 1983).

Another relatively recent idea, at least for neo-

Darwinian evolutionists, which had already been

expressed in the form it could take at the time by

authors such as Goldschmidt (1940), is the very

simple one that the ontogenesis of an individual,

being a complex, integrated process with numer-

ous interactions, obeys a certain number of

constraints, and that not ail modifications are

possible, the development being for the great

part “ canalized ” (see e.g.: Alberch, 1980, 1982;

Wake, 1982 a, 1982 b; Wake, Roth & Wake,

1983). Similar ideas had, it is true, been mentioned

incidently by the most “ synthetic ” theoreticians

of évolution, yet no real discussion had been

devoted to them, as is shown for instance in the

following quote:

“ The students of development hâve various terms for these regulatory

powers, such as buffering, canalizalion, and developmenlal homeoslasis.

These terms apply to models that help us to visualize the action of genes in

the developmental process, but they should not blind us to our basic

ignorance of the exact mechanisms by which the universally observed

régulation during development is achieved. (For further details on the

physiology of différentiation of tissues and organs in relation to gene action,

refer to books on epigenetics.) ” (Mayr, 1970: 168).

Recently, Mayr (1975, 1982 b) has stressed

the importance of these notions, and that of

concepts such as the “ unity ” or “ cohésion ” of

the génotype. They shed a new light upon the

phenomena of macroevolution, which had so far

eluded scientific interprétation to a large extent

and given room to numerous spéculations.

In the light of the recent Works on the biology

of development, Raff & Kaufman (1983) hâve

shown how biological évolution could only be

possible in some directions, because of the

constraints imposed by the mechanisms of onto¬

genesis:

“ If the notion of developmental constraints limiting evolutionary direc¬

tions has any meaning, it is in the sense that modifications of already

existing developmental processes provide the most readily available route

for evolutionary change. Once a modification becomes established, it in turn

makes acceptance of changes in certain directions more feasible than others.

But if existing developmental patterns constrain, they also provide opportu-

nities for rapid evolutionary departures when sélection pressures on

morphology change because of their dissociability and apparently simple

genetic Controls. ” (Raff & Kaufman, 1983: 355).

Besides, Raff & Kaufman (1983) insist upon

the fact that the regulatory genes which play an

important rôle in the control of the morpho-

genesis are probably in small number by com-

parison to the structural genes that corne into

action during it :

“ In both the fly Drosophila and the sea urchin Strongylocentrotus, a

relatively large proportion of the genes expressed at some time during the

life cycle are expressed in a spécifie manner during ontogeny. The crucial

question of how many of these genes control morphogenesis is simply

unanswerable at présent. The overall proportion of genes concerned with

morphogenesis may be great, but paradoxically the number of genes that

actually regulate morphogenesis may not be. Many structural genes required

for morphological ontogeny provide essential products without which

particular morphological entities could not be assembled. Yet these genes

provide little in the way of regulatory information: They are instead

regulated in their action. Genes of this type should not be thought trivial,

however, because the products of some of them, as for exemple, tubulins,

actins, or cell surface proteins, provide the actual machinery for cell shape-

change and cell movements directly underlying morphogenesis. Much of the

control exerted by regulatory genes, those genetic gray eminences, must be

devoted to orchestrating the expression of ontogeny-specific structural

genes. If regulatory genes were very large in number, interactions between

Source : MNHN, Paris

ALAIN DUBOIS

them would be so complex as to render viable evolutionary changes nearly

impossible. " (Raff & Kaufman, 1983: 299).

" Ontogeny involves the activity of many genes expressed in a whole set of

very stable processes In Drosophila about one-third of the total number of

détectable genes are expressed in a developmentally spécifie manner. and are

needed for successful completion of spécifie developmental stages. Neverthe-

less, the number of switches is small, and changes in switch functions may

hâve correspondingly great effects in morphogenesis. It is important to note,

however, that évolution is not a single-step aflfair. The chief significance of

alterations in genes with regulatory functions may be to produce changes in

ontogeny that provide the raw material for further changes in a new

direction. Further change and consolidation of the novel direction occur

through mutational events in genes modifying the principal regulatory gene.

Canalization and intégration can be retained in the midst of evolutionary

transitions in morphogenesis. ” (Raff & Kaufman, 1983: 344).

Raff & Kaufman’s (1983) work is enthralling

and will no doubt prove very useful to ail the

biologists who want to know more about the

présent State of our knowledge in the genetic

determinism of morphogenesis, particularly in

order to better understand the relations between

the latter and évolution. However, although it

begins with a criticism of the partitioning of

biology which had long separated genetics from

embryology, this book is not yet the synthesis

that one may be expecting and that the title of its

last chapter seems to be heralding: “ Regulatory

hiérarchies and évolution: a synthesis The

reason for it is simple: just as Goldschmidt

( 1940), to whom they dedicate their work, Raff

& Kaufman (1983) do not understand that

biological évolution is not an évolution from

organism to organism, from individual to individ-

ual, but that it consists on the contrary in a

process that has to do with populations. In this

respect, the lack of any référencé to Mayr’s

works in their bibliography, as well as the lack of

any discussion of the fundamental phenomena of

populations genetics, of the genetic révolution, or

even of spéciation in general, testify to a serious

shortcoming. As Mayr (1942, 1963, 1970,

1982 b, 1982 c), Rensch (1959) or Stanley

( 1979), for example, hâve emphasized, no theory

of évolution can évadé the central problem of

spéciation:

“ (•••) I feel that it is the very process of creating so many species which

leads to evolutionary progress. Species, in the sense of évolution, are quite

comparable to mutations. They also are a necessity for evolutionary

progress, even though only one out of many mutations leads to a significant

improvement of the génotype. Since each coadapted gene complex has

different properties and since these properties are, so to speak. not

predictable. it requires the création of a large number of such gene

complexes before one is achieved that will lead to real evolutionary advance.

Seen in this light, it appears then that a prodigious multiplication of species

is a prerequisite for evolutionary progress. (...)

The evolutionary significance of species is now quite clear. Although the

evolutionist may speak of broad phenomena, such as trends, adpatations,

specializations. and régressions, they are really not separable from the

progression of entities that display these trends. the species. The species are

the real units of évolution, as the temporary incarnation of harmonious,

well-integrated gene complexes. And spéciation, the production of new gene

complexes capable of ecological shifts, is the method by which évolution

advances. Without spéciation there would be no diversification of the

organic world, no adaptive radiation, and very little evolutionary progress.

The species, then, is the keystone of évolution. ” (Mayr, 1963: 621).

Raff & Kaufman’s (1983) “ synthesis ” remains

therefore very incomplète. They show us con-

vincingly enough that great alterations in the

morphology may be produced by only a few

mutations affecting the regulatory genes, but

they are not concerned with the mechanisms that

may be responsible for the appearance and

fixation of such mutations in naturel populations,

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

whereas it is only when such mechanisms are

known that we shall really be able to begin to

understand the phenomena of spéciation and of

macroevolution.

I think it is only through a synthesis of the

modem data related to the genetics of develop¬

ment and to the study of spéciation (with the

help of the concepts and techniques of popula¬

tions genetics, among others), that apprehending

evolutionary phenomena will prove possible, as

Mayr (1970, 1982 b) has already stressed it:

“ Much that is now explained as ‘ epistatic interactions between different

loci ’ might well be due to the activities of regulatory genes. (...)

The day will corne when much of population genetics will hâve to be

rewritten in terms of the interaction between regulator and structural genes.

This will be one more nail in the coffin of beanbag genetics. It will lead to a

strong reinforcement of the concept that the génotype of the individual is a

whole and that the genes of a gene pool form a unit. ” (Mayr. 1970: 183).

We now know that the génotype, even though ail of it is composed of

DNA, consists of highly heterogeneous classes of DNA, each of which is

likely to hâve a somewhat or altogether different function. Those of us who

for a long time hâve been on the road toward the explanation of spéciation

and évolution and who thought that we were nearing the goal now feel

suddenly like the player in a parlor game who is told to go back to position

zéro. Indeed as far as our understanding of the genetics of spéciation is

concerned we are almost at position zéro." (Mayr, 1982b: 1124).

Another recent attempt at a synthesis is that

by Stanley (1979). Contrary to Raff & Kauf¬

man (1983), this author grants the study of

spéciation ail the importance it deserves, and he

relates it to the notion of genetic régulation, but

other evolutionary phenomena are underrated

this time, namely those of sélection and adapta¬

tion on the scale of populations. Stanley would

probably agréé with Mayr (1978: 478), when the

latter remarks with surprise

“ how little population genetics has contributed to our understanding of

spéciation ”.

Yet, whatever its importance, spéciation is not

everything in évolution, and the lack of any

concern for the results of populations genetics in

Stanley's (1979) work restricts the interest of

this book within the limits of a study of

macroevolutionary phenomena (as its title indi-

cates).

The overall synthesis of what we know about

ail evolutionary phenomena remains to be writ-

ten, and I cannot share Stebbins & Ayala’s

(1981), optimism in this respect, who consider

that such a synthesis would only call for a small

transformation of the “ modem synthesis ”, or

that of Mayr (1982 b) who does not deem a

transformation at ail necessary — see Gould &

Lewontin (1979), Gould (1980), Wake, Roth

& Wake (1983), etc.

The preceding remarks will enable us to

consider the problem of the modalities of appear-

ance of new généra in évolution: indeed it is

typically a field in which the various types of

phenomena above mentioned meet (at the level

of the génotype and of development; at the level

of populations).

Phyletic gradualism AND QUANTUM EVOLUTION:

ARE GENERA DISCONTINUOUS?

As far as the genus, first of the higher

categories, is concerned, one may ask oneself, in

a simple and almost testable way, the fundamen-

tal question of the study of macroevolution: do

evolutionary innovations, i.e. new types of mor¬

phologies, appear in a strictly progressive way.

Source : MNHN, Paris

ALAIN DUBOIS

without disruption, without interruption, as some

of the theoreticians of the synthetic theory of

évolution claim (e.g.: Thaler, 1982; Charles-

worth. Lande & Slatkin, 1982; Barton &

Charlesworth, 1984), or does their rise need

some sort of disruption, a sudden évolution, of

the quantum type, such as Simpson (1944, 1953)

or Stanley (1979) define it? According to the

second hypothesis, two possibilities still remain:

either quantum évolution requires passage through

a “ hopeful monster ” as defined by Gold-

schmidt (1940), as the authors of the theory of

punctuated equilibria seem to think; or this

évolution occurs on the occasion of a genetic

révolution, as Mayr (1954) defines it, or of other

similar mechanisms.

In a very interesting work, Lemen & Freeman

(1984) hâve recently regarded this problem in an

original manner, that is by studying, in three

families of microchiropterans, the way species

were distributed in a given hyper-space defined

by a multivariate analysis of their morphology,

in which the “ size ” and “ shape ” components

of the latter were dissociated by means of

allometrical curves of growth (the successive

morphologies along the same curve were inter-

preted as being the same “ biological shape ”,

though differing between each other only by the

“ size ” factor, whereas the changes perpendicu-

lar to the curve correspond to a change in

“ shape ”). They compared these data gathered

from three real groups of animais to the data

obtained through simulation from three evolu-

tionary models based upon different postulâtes:

(1) a “ uni-modal ” model, compatible with a

graduai évolution, in which the morphological

changes of a character through time hâve a

normal distribution; (2) a “ decoupled ” model,

in which there exist two different types of

evolutionary events causing the morphological

alteration, some being linked to size, and the

others not (“ decoupled ”); (3) lastly, a “salta-

tional ” model, in which there also exist two

types of evolutionary events, some linked to size,

and the others of the saltational type (alterations

of a great amplitude, but in which size and shape

remain correlated).

Regarding the three families of bats studied,

Lemen & Freeman (1984) hâve shown that the

généra such as systematists acknowledge them

today correspond to groups of species of similar

“ shapes ” but of variable “ sizes ”; conversely, a

significant alteration in “ shape ” may be observed

from one genus to another. Comparing those

results to those obtained with the three models

described above, Lemen & Freeman (1984) noticed

that only one of these models, namely the

“ decoupled ” one, yielded similar results, while

the other two models did not produce such

groups of species “ variable in size but homoge-

neous in shape”. Lemen & Freeman's (1984)

conclusion is that these results are consistent

with the hypothesis according to which évolution

would proceed in two successive stages; first,

diversification in “ size ” within a group of

species of similar “ shapes "; then, dissociation

of characters previously correlated, and appear-

ance of a new group of species with a different

“ shape ”. The authors deduce from this that size

and shape do not diversify in the same way, and

that the two processes should be considered as

different evolutionary events: the interaction

between these two types of évolution would

produce the groups of species one observes, that

are homogeneous as far as “ shape ” is con-

cemed, and greatly heterogeneous as far as

“ size ” is concerned; and such groups are those

that are generally considered as généra by sys¬

tematists. They can be holophyletic or para-

phyletic (that is to say that groups defined thus

can rarely be formed by convergence of several

independent lineages). The importance of the

“ distance ” that séparâtes such groups may

vary, and it dépends upon that of the “decou¬

pled jumps ” that enable the passage of one

“ shape group ” to another, or upon the nature

of the adaptive zones: hence, there does not

always exist a gap between these groups, but

there always exists a discontinuity. Finally, the

respective rates of “ correlated ” or “decou¬

pled ” events in the évolution of a given group

will détermine the number of species in each

genus, and the diversity in shapes of that group.

As a conclusion to their work, Lemen &

Freeman (1984: 1236-1237) write:

“ We can speculate that the evolutionary mechanism that makes shape-

conservative généra may work at higher taxonomie levels as well. This idea

leaves us to wonder to what extent the typological concept of discrète

hierarchical categories in systematics might hâve originally hinged on the

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

shape groups produced by the interaction of two different processes, the

évolution of size and the évolution of shape. ”

Lemen & Freeman’s (1984) paper does not (1) What définition of the genus do these

answer some of the questions that its reading authors use, or advocate? They do not make a

may raise (Dubois, 1988): clear choice between today’s various théories of

zoological classification:

“ We take no stand on how généra are actually formed, or on how généra

should be formed ” (Lemen & Freeman, 1984: 1220).

As a matter of fact, they seem to opt for an

empirist conception of classification, since they

consider

“ the actual formation of généra difficult and perhaps a matter of art in

science” (Lemen & Freeman, 1984: 1236).

(2) What do they mean by expressions such as

“ real généra ” or “ the real world ”, which

frequently appear in their text? Do they mean

généra that are “ real ” in the taxinomie practice

(by opposition to an “ idéal ” définition or

conception of the genus, or to the artificial

groups that may produce the computer simula¬

tions such as the ones they use in their work), or

généra that are “ real ” in nature, and exist

independently from the systematists’ idea of

them? A close reading of their article reveals that

the expression “ real généra ” takes either of

these two meanings alternatively in various parts

of the text. It is true that the two meanings do

not necessarily exclude one another: it is indeed

quite possible to claim, as I am precisely doing in

this paper, that there exist in nature, as a

conséquence of biological évolution, “ real enti-

ties ”, real groups of species to which the

category of genus can be applied; the system-

atists’ task would then be to recognize or identify

such entities in nature rather than to try to

construct artificial groups. It seems that such an

idea is in the back of Lemen & Freeman’s (1984)

minds, for instance when they write:

“ It is the interaction of the évolution of size and shape that produces the

shape-conservative groups that can vary greatly in size. ” (Lemen &

Freeman, 1984: 1236).

However, if such an hypothesis is made, it

should be clearly stated. Moreover it entails

other conséquences: for example, if the généra

exist, and must be recognized, in nature, it

cannot simply be a matter of “ art ”, but scien-

tific rules must be proposed in order to reach

such a goal, contrarily to what Lemen & Free¬

man (1984) write.

(3) Lemen & Freeman (1984) do not question

the nature of the genetic phenomena likely to be

responsible for the two fundamentally different

evolutionary processes that they think account

for the rise of the groups, homogeneous in

“ shape ” and variable in “ size ”, that they hâve

fourni. What can these mechanisms be? That is

what I am now going to try to deal with.

However, let me first note that the discontin-

uity between généra, clearly expounded in Lemen

& Freeman’s (1984) work, has been known by

systematists for a long time. It can be shown by

various methods of study of morphology, but

also, in quite a different way, by the study of

hybridization: as we hâve already seen above, the

study of the lists of species likely to generate

viable hybrids makes it possible to acknowledge

the existence of closed groups, of varied sizes, not

that of a continuum of species hybridizable step by

step. Moreover, these groups of potentially

hybridizable species happen to hâve long been

recognized as systematic units, although the rank

given to those taxa may vary from one group to

another (genus, family, etc.): my proposition to use

the criterion of hybridizability to define généra

boils down to choosing a level of standardization

and making the two types of discontinuities

Source : MNHN, Paris

ALAIN DUBOIS

(according to morphology and to hybridizability)

coincide. The fact that the two criteria can

coincide rather easily shows that the discontinuities

between généra are mainly due to the particular

conditions in which the généra appear, rather

than to the extinctions of so-called intermediate

species which some authors refer to, maybe for

fear they should hâve to believe in “ hopeful

monsters ” instead, in order to be able to

account for the phenomena of “ saltation ” in

évolution. It is therefore interesting now to

consider the processes involved in the birth of

généra.

TRANSILIENCE, GENETIC REVOLUTION AND GENIATION

Geniation

In 1981, I proposed the use of the new term

geniation (from the Latin genus) to describe the

“ appearance or birth of a new genus ” (Dubois,

1981 c: 508). The use of such a term implies that

one admits that there are indeed entities in

nature that one can call by the name of genus,

and that the entities in question are not only

créations of the human mind. If one admits that

these entities do exist, it is legitimate to devote

attention to the mechanisms responsible for their

birth or appearance. However, the term “genia¬

tion ” in itself does not imply any mechanism a

priori : one can envisage a graduai geniation, i.e. a

progressive one, and a quantic geniation, i.e. a

rapid and sudden one.

If we are to believe Lemen & Freeman (1984),

généra appear in nature as a resuit of “decou-

pled events ", during which the “ size ” and

“ shape ” factors of the morphology of the

organisms happen to be decoupled or separated

for some time. As we hâve seen, it is therefore a

question of discontinuous events, of the quantic

type and not of the graduai one.

Lemen & Freeman (1984: 1221) call the model

that describes such events by the name of

“ decoupled/adaptive zone model ”. They thus

refer to the concept of adaptive zone, as it was

formulated by Simpson (1944, 1953). Each genus

may be considered as a group of species occu-

pying a given adaptive zone. The basic idea upon

which this conception is grounded is the fol-

lowing one: adaptive zones are discontinuous,

and the passage from one to another requires

important genetic alterations, that are irréver¬

sible at little cost (Dubois, 1975, 1976, 1981 c,

1982 a). The passage into a new adaptive zone

requires the Crossing of a gap of adaptive disequi-

librium which séparâtes it from the previous one

(Simpson, 1944, 1953). The question is to know

how this gap can be crossed.

Concerning this, Simpson’s hypothèses remain

vague and quite debatable: he proposed the

formula of quantum évolution to describe this

type of events, but the mechanism suggested

(fragmentation of a large population into small,

isolated populations, then passage of the latter

through a “ non-adapted ” phase before “ac-

costing ” in a new adaptive zone) is not very

likely (Pasteur, 1982: 512). Moreover, Simpson

did not propose any genetic model to account for

this quantum évolution (Dubois, 1982 b: 398).

MaYR'S MODEL OF GENETIC REVOLUTION

The first cohérent model proposed in this

respect is that of the genetic révolution of Mayr

(1954, 1963, 1970, 1975). It is a particular model

of spéciation, which belongs to the more general

category of spéciation called “ peripatric spécia¬

tion ” by Mayr (1982 b, 1982 c), in which the

emphasis is laid upon: (1) the isolation, in

adverse environmental conditions, of one or

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

several founder individuals from the initial popu¬

lation (large, panmictic and with an important

genetic polymorphism); (2) the reorganization of

the génotype on new bases; (3) the passage into a

new adaptive zone. This model cornes as a

complément to Simpson’s quantum évolution

(1944, 1953), with which it is quite compatible

(Dubois, 1982 b), and Pasteur (1982: 512) has

suggested combining the two théories under the

general name of “ Simpson-Mayr model of

transspecific évolution The concept of genetic

révolution would thus make it possible to

account for the cases of “ sudden ” appearance

of completely new types of organization within

homogeneous groups which hâve drawn the

evolutionists’ attention for quite a while.

The genetic révolution would not so much

consist in the appearance of new mutations as in

a re-organization after a new mode, of the genes

already présent in the initial stock. In his work of

1954, Mayr already insisted upon the fact that

the most important aspect of this event would be

the great rise in the degree of homozygosity in

the small isolated founder population. In this

small population, the homozygosity would be

maintained, and even increased through généra¬

tions. It would affect the sélective value of many

genes, as well as the overall internai equilibrium

of the génotype. Under the effect of the heavy

natural sélection that would affect this popula¬

tion, the génotype would be profoundly allered,

before reaching a new stade of equilibrium. The

population could thus go from one “ adaptive

peak ” to another, to take up Wright’ s (1932)

image. Mayr (1954: 169-170) does not write that

ail the genes would be directly modified, but that

they would at least be “ affected ” in their

“ genetic environment ” and their sélective value:

“ We corne thus to the important conclusion that the mere change of the

genetic environment may change the sélective value of a gene very consider-

ably. Isolating a few individuals (the ‘ founders ’) from a variable population

which is situated in the midst of the stream of genes which flows ceaselessly

through every widespread species will produce a sudden change of the

genetic environment of most loci. This change, in fact, is the most drastic

genetic change (except for polyploidy and hybridization) which may occur in

a natural population, since it may affect ail loci at once. Indeed, it may hâve

the character of a véritable ' genetic révolution ’. Furthermore, this - genetic

révolution ’, released by the isolation of the founder population, may well

hâve the character of a chain reaction. Changes in any locus will in turn

affect the sélective values at many other loci, until finally the System has

reached a new State of equilibrium. ”

One understands therefore why various authors close scrutiny, and that Mayr himself (1982 b:

ascribed to Mayr the idea that most genes would 1124) daims he never held:

be altered, an idea that will not hold up under a

“ 1 did not claim in the least that every founder population expériences a

genetic révolution. Neither did I claim that ail or even most genes were

genetically affected. Ail I claimed was that by changing their genetic milieu

the phenotypic expression and hence the sélective value of many genes

would be affected. "

The process described above, which would

occur in some isolated populations but not in ail,

might lead to the appearance of morphological

innovations and enable passage into a new

adaptive zone.

Mayr (1982 a, 1982 b, 1982 c) has recently

proposed a slightly modified phrasing of his

“ the gene pool of a small either founder or relict population is rapidly.

and more or less drastically, reorganized, resulting in the quick acquisition

of isolating mechanisms and usually also in drastic morphological modifica¬

tions and ecological shifts. It involves populations that pass through a

bottleneck in population size. ” (Mayr, 1982 c: 4).

theory, which takes into account recent develop-

ments of genetics and of the study of spéciation.

The most important characteristic of what he

now calls peripatric spéciation is the reorganiza¬

tion of the génotype on new bases, without.

however, most loci being modified:

Source : MNHN, Paris

ALAIN DUBOIS

Mayr's hypothèses on genetic révolution hâve being probably the fact that this model is based

given rise to numerous criticisms, some of which upon conditions that are mutually contradictory

are important (Lewontin, 1965; Lande, 1980; in populations genetics:

Carson & Templeton, 1984), the main one

“Genetic révolution requires a significant increase in homozygosity

relative to the ancestral condition. By emphasizing that the founders corne

primarily from peripheral demes, however. Mayr makes it more difficult to

satisfy this requirement in the many species in which peripheral demes are

already characterized by inbreeding and increased homozygosity. More

damaging is the fact that a population’s ability to respond to intense

sélection is directly proportional to the amount of genetic variation it has.

Yet the genetic révolution model demands a rapid and effective response to

sélection precisely when genetic variation is at a minimum — conditions that

make a rapid and effective response impossible. Thus, Mayr’s genetic

révolution model is based upon mutually contradictory population-genetic

conditions.” (Carson & Templeton, 1984: 119).

OTHER MODELS OF GENETIC REVOLUTION

Other models of spéciation by founder-effect

were proposed after Mayr’s (1954). Thus Car¬

son (1975, 1982) proposed the “ founder-flush

spéciation theory ” (Powell, 1978), recently

rediscussed by Carson & Templeton (1984),

which, as its name indicates, calls on a founder-

effect followed by a demographical explosion.

In a fundamental paper, Carson (1975) has

suggested that every diploid species has two

distinct Systems of genetic variability. The “ open ”

System consists in ail the genes which are

frequently polymorphie and which can recom¬

bine freely without this having important consé¬

quences on the viability: he mentions as examples

of such genes those that intervene in enzymatic

polymorphism, in clinal and subspecific variabi¬

lity. These genes may be introgressed from one

species into the other in the case of species that

can occasionally hybridize in nature (Sene &

Carson, 1977). On the contrary, the “ closed ”

System consists of “ internally balanced gene

blocks ” forming coadapted complexes. Such

supergenes (Darlington & Mather, 1949: 46)

can be preserved from dissociation by recombi¬

nation for instance by the presence of inversions

(see Wasserman, 1968). Their dissociation by

crossing-over leads to an important réduction in

viability in the normal conditions of natural

sélection. These blocks are stable within a species

but different from one species to another. They

cannot be introgressed by hybridization from

one species to another (Sene & Carson, 1977).

Spéciation therefore requires that the existing

blocks be broken and new ones established.

Recent studies on the structure and the work-

ing of the genome of Eucaryotes hâve made it

possible to State more precisely the nature of the

supergenes which make up the “ closed ” genetic

System such as Carson (1975, 1982) conceives it.

Here is how Demarly (1979) defines the concept

of linkat.

“ The linkat is presented as:

1. A set of loci which aggregated in a same chromosomal sector during

species différentiation. These clusters show strong epistasy and generally

represent coadapted functions.

2. Each of the loci are constituted by a sériés (of) duplicated transcrip-

tional units. Their expression has some flexibility caused either by hierarchi-

cal repression or derepression between slightly differentiated duplicates or

by rearrangements of introns to exons after DNA transcription, which

breaks the dogma ‘ one gene one polypeptide chain '.

3. On these chromosomal segments the allelic arrangements which

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

possess the highest adaptive value hâve been stabilized by génie and epigenic

factors lowering the rate of recombination between them. Therefore they are

inherited as a semistable block.

4. In some case it could be postulated that these arrangements contain

inside them antimutator factors which give a longer perennity to the

clusters.

Therefore linkats appear to be semi-stable functional units, the expression

of which having some flexibility following environmental corrélations and

genetic background. This concept is included in a genetic System which

minimizes genetic load. " (Demarly, 1979: 258).

In the “ founder-flush ” model of spéciation, a

founder population is isolated from an ancestral

polymorphie and coadapted population. The

genetic drift that follows the foundation event

starts desorganizing the ancestral coadapted gene¬

tic complex. As the population settles into its

new environment, it goes through a phase of

démographie explosion, in which, due to the

slackening of natural sélection, the genetic vari-

ability of the ancestral population is not only

preserved, but also increased by phenomena of

recombination and of alteration of the pleiotropic

equilibria. At the end of this phase of démogra¬

phie explosion the population is therefore highly

polymorphie. The environment becoming satu-

rated the sélective forces appear again, and they

can entail a new phase of mass mortality, which

can lead to the surviving of only one or a few

individuals, in which the initial balanced and

coadapted genetic Systems may happen to hâve

been modified and reorganized in a different

way. Those atypical individuals, characterized by

a new coadapted “ closed ” genetic System, can

be at the origin of a new species.

The model of “ genetic transilience ” proposed

by Templeton (1979, 1980 a) and recently re-

discussed by Carson & Templeton (1984) is

close to the preceding one, in so far as it does not

call on a sharp increase in the degree of homozy-

gosity. However, in this model, the main factor

of sélection, instead of being external (high

sélection after the period of démographie explo¬

sion), is endogenous, viz. it is a modification at

random, in a very limited population, of the*

frequencies of some rare “ major alleles ”, i.e.

genes with important pleiotropic effects. The

alteration of the initial frequencies of these

alleles can end up in the fixation of some of them

in the homozygous State. The resulting transfor¬

mation of the genetic environment leads to a

“ there is not one founder principle

ton, 1980 a: 1030).

fundamental change in the sélective value of the

génotype, and the population enters a new phase

of sélection. If the founder population has a high

genetic variability at numerous loci (and there¬

fore a high heterozygosity), it may happen to

react to this sélection in rapidly shifting towards

a new State of genetic equilibrium (a new

coadapted génotype).

The validity of Carson’s (1975, 1982) and

Templeton’s (1980 a) models is supported, not

only by a theoretical study of them (Carson &

Templeton, 1984), but also by laboratory works

(Powell, 1978; Wallace, 1978; Templeton,

1979; Arita & Kaneshiro, 1979; Ahearn, 1980),

as well as by the study of the spéciation processes

in certain groups of animais, the most spectacu-

lar of which in this respect being that of Hawaii

drosophils (Carson & Kaneshiro, 1976). In

quite a different group, that of geckos, Pasteur

(1964, 1977, 1982) has also shown that some

phenomena of spéciation can obviously be ex-

plained by such founder-effects.

In their recent works, Templeton (1980 a,

1980 b, 1981, 1982) and Carson & Templeton

(1984) hâve not merely described the process of

genetic révolution, but they hâve also endeav-

oured to incorporate theoretical concepts of

populations genetics into the study of the process

in question. They hâve thus shown that genetic

révolution can only take place in some very

précisé conditions, not only ecological and geo-

graphical, but also genetic: in other words, only

some species are a good “ ground ” for such an

event. Therefore these authors do not claim in

the least that ail the events of spéciation occur by

genetic révolution, but only that some of those

events do so. Moreover, they insist upon the fact

that there exist several distinct types of genetic

révolution:

in spéciation, but several ” (Temple-

Source : MNHN, Paris

ALAIN DUBOIS

As for Pasteur (1982), he gave a list of the

properties (prédispositions, preadaptation, other

properties) of an ancestral species, or of some of

its populations, that will render easier or possible

the birth out of it of a daughter species by effect

of founder and genetic révolution.

In this respect, it is interesting to note that

one’s agreement with the theory of genetic

révolution (spéciation by founder-effect) does

not at ail imply that one should automatically

agréé with the recent theory of punctuated

equilibria (Eldredge & Gould, 1972; Gould &

Eldredge, 1977; Stanley, 1979; Gould, 1982).

First of ail, results compatible with this model

can be obtained by other models than that of

spéciation by founder-effect (Carson & Temple-

ton, 1984). Secondly, it is not true that, as

Stanley (1979) or Gould (1982) thought, the

référencé to a model of genetic révolution should

imply that sélection and adaptation in a given

species (“ phyletic gradualism ”) do not play any

evolutionary rôle. Indeed, the main rôle of

genetic révolution is to free the species from the

epistatic constraints of its coadapted “ closed ”

System, but, once that done, a new coadapted

“ closed ” System will still hâve to be buill again,

which cannot be an instantaneous phenomenon.

From this point of view, there is agreement

between authors who developed different models

of genetic révolution, such as Mayr, Carson

and Templeton:

“ The évolution of a new coadapted gene complex (the event actually

associated with the development of a new species) generally occurs after the

genetic révolution, and it occurs via the normal operation of sélection,

mutation, drift, and so on within a single breeding population. The inference

that microevolutionary processes are unimportant in spéciation because of

genetic révolution is totally unfounded. " (Carson & Templeton, 1984:

126).

" What is crucial is the fact that prior epistatic and regulatory Systems are

broken up during a genetic révolution in the founder population, making

room for new ones. This greatly facilitâtes and speeds up the acquisition of

new adaptations. These are, of course, not acquired by single steps, and

sélection for their improvement continues. It may even be accelerated by the

establishment of descendant founder populations. It is unknown and

presumably variable whether such an evolutionary shift requires a few,

scores, hundreds, or thousands of générations, but it is certainly by several

orders of magnitude faster than the traditional phyletic évolution described

in the paleontological literature as requiring millions of years. Even so,

évolution through changes in founder populations is not a process of

saltation but one of graduai évolution. The most important departure in the

new way of thinking is to treat it as a populational phenomenon. ” (Mayr,

1982 a: 618).

" One of the major effects of the disorganization described above is that it

often may bring the relevant population close to extinction. Numbers

become small; adaptations are impaired by stochastic effects. The mean

fitness of the population is lowered as the various balanced genetic

components of the gene pool are destabilized. If the population is to survive

the threatened extinction, then, the générations that immediately follow the

disorganization phase become crucial. Under these circumstances, a change

in ambient environment is not a necessary prerequisite for genetic change. It

is not a matter of the details of the génotype slavishly tracking the

environment. What has happened is that the former genetic organizations of

the gene pool, its old epistases and balances, are suddenly in disarray.

Accordingly, sélection begins to actively form new balances, using the

remnant genetic éléments segregating in the depauperate gene pool, which

may continue to hâve a small effective size.

The ensuing one hundred to one thousand générations are considered

crucial in the building of the organization of the new gene pool, and the

synthesis of the new adaptations. In fact. this stage in the life history of the

species. in this reductionist view, is the most important one from the point

of view of progressive, significant genetic change per unit time. It is during

this time that the adaptations characteristic of the species as a whole are

forged by mutation, sélection, and recombination along with other corre-

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

lated morphological, behavioral, and physiological novelties of the new

species. Basically, it is a graduai, anagenetic intrapopulational process; there

is nothing saltational, rectangular, punctuated, concerted, or instantaneous

about it. Macromutations and mutations profoundly affecting development

are not required. As the gene pool expands in size and gradually

équilibrâtes, the rate of genetic change is gradually reduced. In most diploid

organisms, what has been achieved is considered to be a new complex

dynamic balance, not a new fixed homozygous State. The biggest change

may well be a change in internai genetic environment and interaction

between the many component genes. ” (Carson, 1982: 423-424).

For the paleontologist, some events may seem

to hâve been “ instantaneous ”, while they took

place over a span of many générations: neon-

tologists and paleontologists work on different

“ scales ”, which accounts for the basic différence

in their appréciation of the more or less “grad¬

uai ” nature of evolutionary phenomena (see e.g.

Mayr, 1982 b).

Besides, Templeton (1980 a) has emphasized

that numerous fossil groups, the history of which

had been interpreted as supporting the theory of

punctuated equilibria, probably did not meet the

requirements enabling spéciation by founder-

effect:

“ Consequently, founder-induced spéciation models do not provide either

a general theory of macroevolution or a general interprétation framework

for the fossil data. " (Carson & Templeton, 1984: 126).

The theory of punctuated equilibria suffers

also from other difficultés or incoherencies, that

Mayr (1982 b), for instance, has analysed. There

exist several versions of this theory. The two

extreme ones are, on one hand, a moderate,

Mayrian or Simpso-Mayrian one, which acknowl-

edges that genetic révolution is a graduai and

populational, albeit very rapid, phenomenon,

and on the other hand a drastic, or Goldschmid-

tian one, which refers back to notions such as

“ systemic mutations ” or “ hopeful monsters ”.

The latter overlooks the populational aspect of

evolutionary phenomena, and is only an inac¬

ceptable simplification of the observed facts.

Genetic révolution and chromosomal rearrangements

It is tempting to try to “ visualize ” genetic

révolution, in particular at the chromosomal

level. Thus, Wilson, Sarich & Maxson (1974)

hâve suggested that a genetic révolution could be

caused by a rearrangement of the position of the

genes on the chromosomes; Wilson et al .'s

(1975) data go in the same direction. According

to these authors, such chromosomal rearrange¬

ments would be particularly frequent and rapid

in the groups in which the effective size (Wright,

1931) of the reproductive populations is low

(Wilson et al., 1975, 1977; Bush et al.. 1977),

such as the founder populations in the models

above mentioned. The chromosomal rearrange¬

ments in question would entail alterations in the

Systems of genetic régulation, without a modifi¬

cation of the structural genes, but with changes

in the rates of the different types of molécules

that regulate genetic activity, and, consequently,

in the quantitative relations between the activi¬

tés of various genes, crossings of thresholds

(Zuckerkandl, 1979, 1980). These changes in

the Systems of genetic régulation could hâve far-

reaching conséquences as regards both morpho-

logy, and postzygotic isolation from the initial

stock. If it is clear that ail the cases of spéciation,

including those by genetic révolution, do not fit

within the framework of this model, it is difficult,

for the time being, to guess the proportion of

cases of spéciation that do fit in it. According to

White (1978: 324), more than 90 % of the cases

of spéciation would be accompanied by chromo¬

somal rearrangements, but the nature of the

implications of those rearrangements, particu¬

larly what has to do with the mechanisms of

genetic révolution, is still very poorly known:

Source : MNHN, Paris

ALAIN DUBOIS

" In fact, each chromosomal rearrangement — whether fusion or

dissociation, translocation, inversion, gain or loss of heterochromatin —

must be regarded as a unique event whose conséquences will be almost

impossible to predict in the présent State of our knowledge. " (White, 1978:

336).

“ It seems unlikely that the chromosomal rearrangements that lead to

changes in chromosome number or in the number of chromosome arms

would themselves directly produce regulatory genetic changes. ” (White,

1982: 88).

“ The introduction of the concept that changes in gene régulation may

hâve greater evolutionary significance than changes in the genes themselves

has had a major impact on evolutionary studies in the last decade. (...) By

potentially altering the m-acting regulatory circuitry, a chromosome

rearrangement may affect gene régulation, and thus organismal phenotype.

(...) There is, however, little hard evidence on the types of chromosomal

rearrangements observable by standard cytogenetic techniques that supports

this view. On the contrary, systematic studies of rodents hâve discovered

numerous examples of cryptic chromosomal ‘ species ", many of which

involve substantial reorganization of the karyotype. (...) In these cases,

cytological rearrangements hâve had no discernible phenotypic effects ;

those that do produce noticeable pathologies would be rapidly eliminated

from natural populations. (...) Phenotypic changes clearly are not a general

conséquence of karyotypic change. ” (Patton & Sherwood, 1983: 149).

“ There is a reciprocal relationship between chromosome structure and

gene function. The rôle of genes in determining the behavior, function, and

even structure of chromosomes has been almost entirely neglected and is

absent from discussions of the rôle of chromosome change in population

divergence. The data available to date suggest that chromosome change may

well be of secondary importance in processes of spéciation and phyletic

divergence." (Patton & Sherwood, 1983: 152).

“ Our own view is that genomic reorganization is crucial to morphologi-

cal évolution. However, these changes are achieved by mechanisms more

subtle than gross chromosomal rearrangement, and gross changes are not a

necessary component of spéciation and morphological change. ” (Raff &

Kaufman, 1983: 82).

It is therefore probable that there does not

exist a straightforward relation between chromo¬

somal rearrangements and the évolution of the

Systems of genetic régulation which is itself

associated with spéciation and morphological

évolution. This independence is stressed by the

now well-known fact that spéciation can occur

without rearrangements, as for instance certain

species of Hawaiian drosophils show (Carson,

Clayton & Stalker, 1967; Carson & Kane-

shiro, 1976; etc.).

Genetic révolution as a mode of spéciation among others

Some general conclusions can be drawn from

what has just been said.

First, it is certain that, in small, isolated

populations, a fundamental reorganization of the

génotype may occur under certain circumstances,

which can lead to an alteration in the morpho-

logy, to the passage into a new ecological niche,

and to the rise of a new species which may be the

starting point for a new genus.

Several mechanisms hâve been proposed to try

to explain how such a reorganization of the

génotype can take place and, above ail, how it

can be fixed in the founder population. Some of

these mechanisms are not very likely, others are

more so, but the concrète data, based as much

upon experimental facts as upon study of natural

populations and species, hâve so far remained

too scanty and lacking in details to make it

possible to know which mechanisms really oper-

ate in nature, and what is the relative importance

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

of each of them at the global scale of évolution.

However, it is quite probable that there does not

exist only one mechanism likely to produce such

a resuit, but several, which are not ail known as

yet (even as “ models ”).

It would be advisable in this respect not to

repeat the mistakè that had been made in the

case of the general study of spéciation, where

controversies went on for years: these controver-

sies were partly distorted because in particular

they were based upon the erroneous hypothesis

whereby ail the cases of spéciation had to obey

similar mechanisms. Various recent works hâve

on the contrary led to the conclusion that there

certainly exist several well distinct types of

spéciation (see e.g.: Scudder, 1974; Bush, 1975;

Carson, 1975, 1982; Endler, 1977; Pasteur,

1977, 1982; White, 1978; Templeton, 1980 a,

1980 b, 1981, 1982; Barigozzi, 1982; Rose &

Doolittle, 1983; Carson & Templeton, 1984;

Barton & Charlesworth, 1984). These differ¬

ent modes of spéciation are in particular related

to the type of ecology of the ancestral species

(type of environment, size and structure of the

populations), with their genetic structure and

with the geographical conditions (see e.g. the

various chapters in Barigozzi, 1982):

“ Quite often it has been concluded that one aspect of evolutionary

change is the most important one with respect to spéciation, such as

karyotypic évolution (...), or that certain levels of genetic divergence

correspond to certain levels of taxonomie status. (...) However, the evidence

has dashed ail these hopes: Spéciation can occur in the absence of, or is

uncorrelated in some groups with, karyotypic change (...), significant DNA

sequence divergence (...), significant isozyme différentiation (...), morpholo-

gical change (...), and shifts in niche or habitat (...). These studies do not

imply that these factors are never involved in spéciation, simply that one

factor is not critical or necessary for ail modes of spéciation. Because of the

failure of individual éléments to identify a universal marker of spéciation,

some workers hâve investigated joint patterns of two or more of these

différences and their relation to spéciation. For example, in some verte-

brates, karyotypic and morphological évolution are positively correlated

with each other and with spéciation rates, whereas protein évolution is

uncorrelated with ail the others (...). However, other studies do not support

this pattern (...). Thus, there is also no universal joint pattern relative to

spéciation. However, predictable patterns and différences do emerge for

particular groups of organisms (...), and population-genetic considérations

are apparently important déterminants of these patterns (...). ” (Templeton,

1981: 24).

Despite this diversity, it seems possible to

classify the different modes of spéciation into

two main categories. The spéciations that belong

to the first category are slow phenomena in

which genetic différences are gradually accumu-

lated between separated populations; when the

latter are brought in contact again, reproductive

isolation already exists between them, or becomes

progressively established. The species originating

from such a type of spéciation can be separated

only by a few “ minor ” genetic alterations,

bearing only upon a few structural loci. The

morphology of the two species can be very

similar or even identical (notion of “ sibling

species” or dualspecies; see Bernardi, 1980),

and so can it be as regards the structure of their

chromosomes, their behavior, their ecology, etc.

(except eventualiy for some behavioral différ¬

ences working as pre-ejaculatory mechanisms of

isolation). These species often remain able to

give birth to viable hybrids, at least in experi¬

mental conditions.

In the cases of spéciation belonging to the

second category, on the contrary, the genetic

alterations are more important and sudden.

Although they do not affect ail the génotype as it

had first been thought, the modifications can be

of a different nature, since they can concern the

genetic regulatory Systems themselves, and not

only the structural genes. Spéciations of this type

probably occur mainly in small isolated founder

populations. They sometimes, but not always,

produce species that are quite different in their

morphologies, behaviors, écologies, etc.

The two categories of spéciation, the “ graduai ”

one and the “ quantic ” one, are fundamentally

distinct and occur in very different conditions.

Moreover each of them includes several distinct

modes of spéciation (Templeton, 1980 b, 1981,

1982).

Source : AANHN, Paris

ALAIN DUBOIS

Questions of terminology

How should one call the cases of sudden

spéciation that take place in small isolated

founder populations, and in which the génotype

undergoes a fundamental reorganization? Several

terms hâve been proposed, which might suit this

type of spéciation: “ transilience ” (Galton,

1894; a term used again with a modified meaning

byTEMPLETON, 1979, 1980 a, 1980 b, 1981, 1982);

“genetic révolution” (Mayr, 1954); spéciation

by " catastrophic sélection " (Lewis, 1962); “ quan¬

tum spéciation ” (Grant, 1963; phrase taken up

by Stanley, 1979); “ founder-flush spéciation ”

(Powell, 1978); “ regulatory révolution ” (Tem-

pleton, 1979); “ rapid spéciation” and “salta-

tional spéciation" (Ayala, 1982); “ founder-

induced spéciation " (Carson & Templeton,

1984); etc.

Assuredly, these various terms are not exactly

synonymous with each other. Most of them were

precisely coined by their authors because the

mechanisms (in particular genetic ones) imagined

for this type of spéciation were different from

those postulated by the previous authors. How-

ever, it is clear that the various concepts are

akin, since they ail describe cases of sudden

spéciation, in opposition to the phenomena of

graduai and slow spéciation which had long been

considered as the only ones existing.

Mayr’s (1954) phrase “ genetic révolution ”

was used for a long time to indicate the cases of

sudden spéciation by founder effect in small

isolated populations. Templeton (1979, 1980 a)

having proposed the new term “ genetic transi¬

lience ”, elicited the following comment from

Mayr (1982 a: 885-886):

“ Templeton assumed that his modified interprétation of genetic révolu¬

tions would require the introduction of a new term (' genetic transilience ').

However, this change of interprétation is far less than between the species of

Linnaeus, the gene of Johannsen, the mutation of de Vries, and the current

concepts designated by these terms. We would drown in terminology if a

new term were introduced every time a scientific concept was modified.

Furthermore, Galton coined the term ‘ transilience ’ for a major saltation

in a single individual. ”

I agréé with Bernardi (1956, 1980) and Mayr

(1982 a) that only the rule of priority should be

used to choose between various “ synonymous ”

terms. However, 1 think that the terms “transi¬

lience " and “ genetic révolution ” are not syn¬

onymous, but that the second one describes only

one particular case among ail the phenomena

concerned by the first one. Similarly, “ genetic

révolution ” is only one of the possible types of

“ peripatric spéciation ”, i.e. which occur in

small isolated populations (Mayr, 1982 b, 1982 c).

Galton (1894: 368) defined the term transi¬

lience as opposed to divergence :

" The phrase of organic stability must not as yet be taken to connote

more than it actually dénotés. Thus far it has been merely used to express

the well-substantiated fact that a race does sometimes abruptly produce

individuals who hâve a distinctly different typical centre, in the sense in

which those words were defined. The inference or connotation is that no

variation can establish itself unless it be of the character of a sport, that is,

by a leap from one position of organic stability to another, or as we may

phrase it, through ‘ transilient ' variation. If there be no such leap the

variation is, so to speak, a mere bend or divergence from the parent form,

towards which the off-spring in the next génération will tend to regress ; it

may therefore be called a ‘ divergent ' variation. Thus the unqualified word

variation comprises and confuses what I maintain to be two fundamentally

different processes, that of transilience and that of divergence, and its use

destroys the possibility of reasoning correctly in not a few important

matters. The interval leapt over in a transilience may be at least as large as it

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

has been in any hitherto observed instance, and it may be smaller in any less

degree. Still, whether it has been large or small, a leap has taken place into a

new position of stability. ”

Of course, as Mayr (1982 a, 1982 c) observes,

the “ genetic model ” evoked by Galton (1894),

which consists of a sudden évolution by macro¬

mutation in a single individual, belongs to the

same type as Goldschmidt’s (1940), “ hopeful

monster ” model, and it cannot be supported

anymore today. However such a model was not

absurd back in 1894, before Mendel’s laws were

rediscovered and populations genetics was born.

Galton’s (1894) merit was to make a distinction

between two fundamental types of évolution —

by divergence and by transilience. This distinc¬

tion is still valid today, although other terms

hâve sometimes been used to mark it: “ phyletic

gradualism ” and “ quantum évolution ” (e.g.:

Simpson, 1944, 1953; Stanley, 1979), “géographie

spéciation ” and “ quantum spéciation ” (Ayala,

1982), etc. With Templeton (1980 b, 1981, 1982),

I deem that Galton’s (1894) terms divergence

and transilience must be kept to name the two

major categories of spéciation modes.

Besides, as Templeton (1980 b, 1981, 1982),

has shown, the transilience category, just like

that of divergence, is not homogeneous. This

author distinguishes between four fundamental

modes of spéciation within the first category:

“ genetic transilience ”, “ chromosomal transi¬

lience ”, “hybrid maintenance ” and " hybrid

recombination ”, Some of these modes of spécia¬

tion do not call for a founder population of

small size in the least. For instance, spéciation by

polyploidization can occur in sympatry and in a

single génération (see Dubois, 1977 b and Bogart,

1980), and yet the resulting polyploid species

may hâve no allele different from the diploid

species (or from the two diploid species, in the

case of allopolyploidy) it dérivés from: the new

polyploid species may produce perfectly viable

hybrids with the ancestral diploid species, but

these hybrids produce aneuploid gametes and

their descent shows signs of deep chromosomal

imbalance and is not viable (see e.g. Dubois,

1977 b: 195). In such cases it is clear that there

was indeed spéciation by “ transilience ”, but no

“genetic révolution” at ail. It is only after a

long period of séparation that the polyploid

species and its ancestral species will hâve suffi-

ciently diverged to manifest différences at the

génie, and not only at the chromosomal, level.

Among the four types of transilience acknowl-

edged by Templeton (1980 b, 1981, 1982), only

the one that he calls “ genetic transilience ”, and

that Carson & Templeton (1984) later called

“ founder-induced spéciation ”, corresponds to

the phenomenon we are here concerned with.

Moreover, as we hâve seen, this category itself is

not homogeneous. Obviously, the first term

available for this category is spéciation by gene¬

tic révolution (Mayr, 1954). I therefore suggest

keeping this term to call one of the types, and

one only, of the larger category of spéciation by

transilience. I propose to use this term in a

strictly descriptive perspective, to describe sudden

spéciations in isolated populations, which does

not imply an agreement with the model Mayr

(1954) proposed in a purely spéculative fashion

to account for the mechanism at work in such

spéciations. When mechanisms are at stake, 1

deem it better to call them, as Barton &

Charlesworth (1984), for example, do, by

names such as “ Mayr’s (1954) model (or

theory) ”, “ Carson’s (1975) model ”, “Temple-

ton's (1980a) model”, etc.

Although, according to the rules proposed by

Bernardi (1956, 1980), a term should not be

rejected because it is “ improper ”, let us remark

that the term “ genetic révolution ”, which was

criticized in this respect, seems to me to hâve

been very well chosen:

“ It has been questioned, with some justification, whether the term

• révolution ' was not too strong. The student of history, however, knows

that many révolutions hardly touched any other institution of a country

except the form of its government. Furthermore, nothing ever occurs in

other kinds of populations that even approaches the drastic genetic turnover

of those founder populations that expérience a genetic révolution. ” (Mayr,

1982 b: 1124).

Source : MNHN, Paris

ALAIN DUBOIS

Moreover, Gould & Eldredge (1977) are quantum évolution, and particularly to the theory

obviously right when they account for the resis- of genetic révolution, in terms of ideological

tance of many biologists to the notion of factors, and I associate myself with their plea in

favor of a “ general philosophy of change

" We believe that a cohérent, punctuational theory, fully consistent with

Darwinism (though not with Darwin’s own unnecessary preference for

gradualism), will be forged from a study of the genetics of régulation,

supported by the résurrection of long-neglected data on the relationship

between ontogeny and phylogeny (see Gould 1977). Ager (1973, p. 100)

(...) speaks in simile of the tempo that we support as most characteristic of

the way our world works: ‘ The history of any one part of the earth, like the

life of a soldier, consists of long periods of boredom and short periods of

terror '. ” (Gould & Eldredge, 1977: 147).

Genetic révolution and geniation

As we hâve just seen, it is now clear that there

is not one, but several, modes of spéciation.

There is also good reason to think there must

exist several modalities of spéciation by genetic

révolution: Carson’s (1975, 1982) and Temple-

ton’s (1980 a) models already présent two possi-

bilities (Carson & Templeton, 1984). Without

further spéculation on these models of popula¬

tions genetics, which I think prématuré, 1 would

like, by way of conclusion, to grant some

reflection to the relations that may exist between

genetic révolution and geniation.

Let me stress first that I do not think that the

two phenomena are always linked. In other

words, 1 think: (1) that there can occur a genetic

révolution followed by a “ simple spéciation ”,

without appearance of a new genus; (2) that, in

some cases, a new genus can appear progres¬

sive^, gradually, in particular in a lineage sub-

mitted to a strong rate of anagenesis. However 1

consider that (3) in most cases , geniation occurs

on the occasion of a genetic révolution. I

therefore consider that even if the two funda-

mental types of geniation (by divergence and by

transilience) exist, the latter is much more fre¬

quent than the former.

The fundamental distinction that I think sépa¬

râtes geniation from “ simple spéciation ” is the

fact that in the latter the modifications of the

génotype bear solely, or mainly, upon structural

genes, whereas in geniation they affect mainly

regulatory genes (Carson's (1975, 1982) and

Sene & Carson’s (1977) “ closed ” genetic Sys¬

tem).

The fact that the modification of the genetic

regulatory Systems may lead to radical alter¬

ations in the génie expression (in particular during

the development), and therefore in the charac-

teristics of the morphogenesis and, lastly, in the

adults’ morphology, physiology and ecology, has

been mentioned several times here above. It has

been discussed in detail in Raff & Kaufman’s

(1983) work. They insist upon the fact that

regulatory genes, which play a great evolution-

ary rôle, are in relatively small number: there¬

fore, the fixation, on the occasion of a genetic

révolution, of one, or only some, mutations

bearing upon such genes, in an isolated founder

population of small size, may prove sufficient to

lead to a “ decoupled ” change, in Lemen &

Freeman’s (1984) sense, and to the passage into

a new adaptive zone.

In the light of what précédés, the following can

be asserted:

(1) The birth of a new genus is not a simple

and frequent event, because of both genetic and

developmental constraints, and of ecological

constraints. The constraints of the first type hâve

been known for a long time, and expressed

through concepts such as canalization, coadapta¬

tion, epistasy, etc. They hâve been evoked from

various viewpoints, for instance in Mayr’s (1975,

1982 b) discussions on the “ unity ” or “cohé¬

sion ” of the génotype, in Carson’s (1975, 1982)

discussions on the notion of “ closed ” genetic

System, in Alberch’s (1980, 1982), Wake (1982 a,

1982 b), Wake, Roth & Wake’s (1983) and many

Source : AANHN, Paris

THE GENUS IN ZOOLOGY

others’ discussions on the rôle of developmental

constraints in evolutionary processes, etc. These

constraints are examined in detail in Raff &

Kaufman’s (1983) book. As for the ecological

constraints, they are mentioned for instance in

Simpson’s (1944, 1953) works on the discontinuity

of adaptive zones, separated by nonadaptive zones.

(2) Because of these constraints, the birth of a

new genus is likely mainly in exceptional condi¬

tions. These conditions can be met with in a

founder population of small size submitted to a

new environment: there can then occur a genetic

révolution, in which the “ cohésion of the géno¬

type ” is broken, the “ closed ” genetic System

decoupled by recombination or by mutation, and

the Systems of genetic régulation profoundly

altered and reorganized. The phenomenon of

genetic révolution takes place on the scale of a

population and not on that of an isolated

individual. It consists in the fixation, by popula¬

tions genetics’ phenomena (sélection, adaptation,

etc.), i.e. graduai, even if rapid, of new regula-

tory genes or of new modalities of interaction

between those genes. The model proposed for

this process reconciles the molecular, develop¬

mental and populational approaches of evolu¬

tionary processes.

(3) Not ail genetic révolutions give birth to a

new genus, but some do.

Pasteur (1982) proposed distinguishing between

two types of events due to founder effects: one,

for which he uses the term genetic transilience,

would be sudden, almost instantaneous; the

other, which he calls genetic révolution, would

spread out more in time. He suggested that the

birth of a new genus would require a process of

genetic révolution extended over a long span of

time, making it possible to bridge the gap

between two adaptive zones. For this process, he

proposed the term “ hopeful transilience ” (Pas¬

teur, 1982). As an illustration, he gives the

convincing example of the Malagasy gekkonid

genus Millotisaurus, for which he had used, as

early as his work back in 1964, whose discussion

is worth re-reading, the Simpsonian phrase “quan¬

tum genesis of a taxon of higher category ”

(Pasteur, 1964: 105).

However we need not necessarily call for a

long phase of instability in order to explain ail

the cases of founder geniation. The particularity

of genetic révolution is precisely that is dissociâtes

the “ closed ’’ genetic System and makes the

reconstruction of a new coadapted genetic com-

plex possible, so that precisely this process

enables the rapid passage from one adaptive

zone to another, without any “ lingering ” in the

intermediate inadaptive zones. Moreover, as

Carson (1982) stresses, young species, which

hâve not had time to reconstruct a “ closed ”

genetic System, are more likely than others to re¬

enter phases of imbalance leading to new spécia¬

tions:

“ It may well be that an old mature species becomes so locked into

obligatory balances that this condition is not conducive to the formation of

new species, since the genetic System is résistant to the disorganization

phase. Such old species thus may not be competent for the budding off of

new ones; they may be looked upon as having essentially become inert from

the evolutionary point of view.

Conversely, a fairly young species that has perhaps been through only

several thousand générations of organizational balance may be capable of

early budding ofT populations capable of disorganization and reorganiza-

tion. This may account for the repeated observation, in the contemporary

fauna and flora, of clusters of very closely related species (‘ explosive

spéciation ’). I refer to species clusters found in some freshwater lakes (eg

Lake Baikal) or species in clusters such as are found in Hawaiian

drosophilids. ” (Carson, 1982: 425).

For the appearance of Millotisaurus as for that

of other similar cases, we must therefore suppose

a sériés of spéciations by successive genetic

révolutions rather than a long period of imbal¬

ance “ between ” two généra.

The genetic révolutions that occur in the

geniation process must be of a particular type, or

“ important ” enough in tenus of genetic rear¬

rangements, to hâve the three following consé¬

quences, which characterize the birth of a new

genus (according to the genus conception that

was developed here above): (a) change in mor-

phology, in which the “ shape ” factor happens

to be decoupled from the “ size ” factor for a

Source : MNHN, Paris

ALAIN DUBOIS

while (“ decoupled event ” of Lemen & Freeman,

1984); (b) postzygotic genetic isolation from the

ancestral species, with which hybridization be-

comes impossible, because of the incompatibility

of the genetic Systems of the two species during

the development of the hybrid embryo; (c) change

in other dimensions of the holomorph, in partic-

ular change of ecology and passage into a new

adaptive zone. I consider these three phenomena

not to be independent processes, but, together,

the results of one event of “genetic révolution

These three characteristics make geniation differ

from “ simple spéciation ”, i.e. phenomena of

spéciation that only lead to a multiplication of

species within the same genus, which corre¬

sponds to what Lemen & Freeman (1984: 1234)

call “ diversification in size within one shape

group “ Simple spéciation ” is not accompa-

nied by as important a change in morphology

and ecology as that which séparâtes two généra

after my définition. In some instances, the loss of

the ability to hybridize can occur during events

of “simple spéciation that is not in contradic¬

tion with what précédés, insofar as the inability

of two species to hybridize may be caused by a

few genetic factors only, sometimes by only one

gene. This inability has therefore no particular

evolutionary or systematic meaning. Conversely,

the fact that two species should remain suscep¬

tible of giving birth to viable adult hybrids

testifies to the fact that their Systems of genetic

régulation hâve remained compatible, therefore

very akin, so that the two species still belong to

the same genus.

(4) We hâve seen that, in a synthetic concep¬

tion of classification, généra can be defined by

three types of criteria, which represent the three

sides of the same reality: genetic, phylogenetic

and ecological units as they are, généra are

evolutionary units which exist as such in nature.

In the light of what précédés, the genus appears

as a basic category, which expresses the fact that

a species has left the adaptive zone of the

ancestral species and has begun to conquer a new

milieu. Thus the genus is the first of the higher

categories, and not only a “ group of related

species ”. Généra so defined are both clades and

grades. The birth of a new genus is a phenome-

non that is qualitatively different from “ simple

spéciation once a species has crossed a “ gap ”

of adaptive imbalance and “ landed ” in a new

adaptive zone, there may occur a new explosion

of spéciation. In this respect, and first under the

form of one species only, the genus is “préexis¬

tent ” to the species that will constitute it; its

appearance will enable their multiplying. After

the arrivai in the new grade, radiation may give

birth to better and better adapted species, and

the ancestral species of the genus may disappear,

though it made ulterior spéciation possible. In

this respect each genus clearly is a natural taxon,

expressing the existence of a real phenomenon in

nature (see also Schaefer, 1976).

(5) Généra so conceived can hâve an extre-

mely variable size, some being monotypic, others

being very large (several hundreds or even

thousands species). Rather than trying to artifi-

cially break up the généra that are “ too large ”

and group together the généra that are “ too

small ”, for instance by requiring that the size of

the “ gaps ” separating généra be inversely pro-

portional to the size of the latter (Mayr, Linsley

& Usinger, 1953; Mayr, 1969), I think this

disparity must be respected, for it expresses a

real phenomenon. The “ large ” généra are those

that hâve “ succeeded ”, that hâve conquered a

large adaptive zone. The small ones on the

contrary are in adaptive zones that are either

narrower or already partly occupied by species

of a different phylogenetic origin. We would

considerably deprive the notion of genus of its

meaning if we systematically broke up large

généra. It is then useful to acknowledge taxino¬

mie sub-units within généra: the next chapter of

this work gives them a brief look.

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

Conclusion

Although this paper deals mainly with the

discussion of the criteria one can use to recognize

généra in zoology, and in particular the new

criterion of hybridizability, I hâve felt it useful to

study the mechanisms likely to account for the

geniation process. The forerunning discussion is

based upon the présent State of our knowledge

about animal spéciation: as we hâve seen, it is

grossly incomplète, and the share of hypothèses

is still important. Part of these may well hâve to

be altered, or even totally abandoned in the

future. Let us stress however that these possible

modifications should not invalidate my main

propositions. In other words, if it is true that the

criterion of hybridizability to define geneça has

the advantage of being in agreement with the

model of geniation proposed above (which I

think grants it its biological and evolutionary

value), the two can however be disconnected: it is

not necessary to accept this model of geniation

to accept the criterion of hybridizability, which

entails many practical advantages, independently

from the biological and evolutionary meaning I

gave it. Such an agreement is not necessary either

to accept the term geniation: this term is purely

descriptive; it désignâtes an undeniable evolu¬

tionary phenomenon, whatever the models pro¬

posed to account for it may be. If we agréé with

Gould & Eldredge (1977: 139)

“ that the importance of a phenomenon is not recognized unless it has a

spécial name ",

we must then admit that the lack, until 1981, of a

proper term to describe the birth of a new genus

testifies to the lack of interest among evolu-

tionists for this type of events until today: I hope

that the présent work will incite new reflections

and studies in this respect.

Source : MNHN, Paris

THE TAXINOMIC CATEGORIES

BETWEEN THE GENUS AND THE SPECIES

SUPERSPECIES, ULTRASPECIES AND SUPRASPECIES

It is not useful to return here to the définition,

the history and the synonymy of the categories

superspecies and ultraspecies, which were very

clearly discussed by Bernardi (1980). This author

showed that the many categories created in the

past by systematists to take place between the

subgenus and the species could be reduced to a

few only, of which Mayr’s superspecies (a group

“ to transform taxinomie categories

instead of being an obstacle to

translation mine).

For the same reason it is necessary to rec-

ognize, between the genus and the species,

several categories, which do not necessarily fit

into each other as is the case in the rest of the

Linnaean hierarchy: it is thus perfectly possible

for a group of species to be both a superspecies

and an ultraspecies.

In the same book in which Bernardi’s (1980)

paper appears, Génermont & Lamotte (1980)

proposed a new supraspecific and infrasubgene-

ric category, that of supraspecies, which groups

in fact ail the supraspecific categories defined by

Bernardi (1980). The proposai of these authors

is therefore in opposition to that of Bernardi

(1980): they suggest grouping under a same

of Birula's prospecies) and Kiriakoff’s ultra¬

species (a group of Pryer's dualspecies) are the

two most important ones. Let us hope that this

review will be accepted as an authority and that,

starting from this work, ail systematists and

evolutionists will use the same words for the

same phenomena.

The purpose of Bernardi’s work is

into a tool for the study of évolution

this study.” (Bernardi, 1962: 333;

name, in a same category, sets of species which

represent very different evolutionary phenomena,

instead of reserving a distinct term to each of

these phenomena. Génermont & Lamotte’s

(1980) superspecies is a taxinomie category which

is devoid of précisé evolutionary meaning and

which principally has a “ practical ” interest.

While the use of the categories discussed by

Bernardi (1980) should prompt authors to

refine their analysis of evolutionary phenomena

as much as possible, the use of the supraspecies

would rather tend to discourage such an enter-

prise, and for this reason does not seem advis-

able.

Source : MNHN, Paris

102

ALAIN DUBOIS

Species complex AND SPECIES GROUP

For the evolutionist, the most interesting of

the categories situated immediately above the

species is certainly that of the superspecies. Very

schematically, one may conceive the superspecies

as follows: when a species has a relatively vast

area distribution, and when the latter is subdi-

vided into several discontinuous zones, there

may occur a divergent évolution, in allopatry, of

the various sets of populations; these may

remain members of a single species, of which

they constitute the subspecies, but they may also

reach spécifie status. These various sets of popu¬

lations may then be designated as the prospecies

of a single superspecies. When a secondary

contact zone appears between two prospecies, a

“ zone of overlap and hybridization ” may become

established in this région, and from that moment

the two species hâve a parapatric distribution

(see e.g. Dubois, 1977 b).

The parapatric distribution is maintained as

such, during a certain time, thanks to mecha-

nisms which are often poorly known: it seems

that the simple presence of each species may be

the proper barrier which precludes the other one

from spreading beyond the zone of contact. But

this is a transitory situation, which cannot persist

during long geological periods. Two fundamen-

tal sorts of évolution may occur then: either the

two species continue to exclude each other

mutually in the régions that they occupy, but the

contact zone between them progressively moves,

until one of the two species, rejected against a

naturel - barrier, is eliminated (Dubois, 1977 b:

173); or the genetic and eco-behavioral divergence

between the two species increases, progressively

allowing these species to become sympatric, at

least in certain régions.

In this latter case, it is no longer possible to

speak of superspecies. It is then possible to speak

of species complex or of species group. These two

categories are often used indiscriminately, in a

relatively informai way, by various authors. It

seems to me however that it could be useful to

apply these categories to two slightly different

evolutionary situations, and by doing so to go

further into the work of terminological clarifica¬

tion started by Bernardi (1980).

The species complex could correspond to the

first evolutionary stage which cornes after the

superspecies once a (at least partial) sympatry

has been established between two (or several)

species. At this stage hybridization may still

occur, albeit rather exceptionally, in nature. Let

us however note that the hybrids obtained are

then either nonviable, or stérile, and are there-

fore not at the origin of an effective gene flow

(with introgression) between the two species, for

otherwise we would be in the situation of having

two entities which hâve not really reached the

status of species but which correspond to what

Bernardi (1980) calls quasispecies or vicespecies.

Afterwards, naturel hybridization tends to

disappear, to be totally or almost totally absent

in the case of the species group, of which

furthermore the species may be largely sympa¬

tric, and may not retain much trace of their

previous allopatric or parapatric distribution.

The species of a species group, however, still

remain morphologically very similar to the unique

ancestral species from which they descend, which

gives them this “ family likeness ” mentioned by

Pasteur (1964: 118), who further remarks that

species groups

“ are entities having essentially a phylogenetic meaning which may not be

utilizable for détermination: they can be defined positively by the affinities

which connect certain species one with another, but not dichotomically and

negatively by diagnostic criteria; they may not necessarily be discriminated

one from another in an absolute way. " (Pasteur, 1964: 97; translation

mine).

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

103

Several degrees of complexification exist, after

additional spéciations within a group, and, in a

purely practical aim, it may be useful to rec-

ognize additional subdivisions: species subgroup,

super-group, etc. (see e.g. Dubois, 1976, 1977 c;

Dubois & Khan, 1980; Dubois & Matsui, 1983).

These various categories hâve no formai status in

systematics and are only useful conventions.

Species groups are much less diversified taxa

than généra. Just as a species having a large

distribution may be subdivided into subspecies, a

genus having a large distribution may be subdi¬

vided into species groups. Species groups are

formed more often, but not always, in allopatry:

they correspond then to a geographical différen¬

tiation within a genus. But if the adaptive zone

of the genus does not show major changes in the

whole area of the genus, there will occur little

divergence between the species groups, in partic-

ular no ecological divergence, each group playing

a similar rôle in each région. A good example in

this respect is the cosmopolitan genus of anuran

amphibians Bufo (Blair, 1972 a).

Any spéciation implies however a certain

ecological divergence, at the level of the species.

allowing several species groups to become sym-

patric. The number of species groups sympatric

in a given région remains however limited, as is

also illustrated here by the genus Bufo (Blair,

1972 a).

It may be useful to briefly discuss the mode of

notation of the different categories of evolu-

tionary taxinomy which hâve just been mentioned.

At the moment, any author who wants to

express the fact that a species belongs to a

species complex or to a species group is obliged

to hâve recourse to a périphrase, of the type:

“ Rana palustris, of the Ranci pipiens group ”. The

mode of notation exposed by Bernardi (1980:

413-414), and which was recently integrated in

the International Code of Zoological Nomencla¬

ture (Anonymous, 1985), allows one to lighten

this notation. The preceding example could thus

be written: Rana (gr. pipiens) palustris. The four

supraspecific and infrasubgeneric categories briefly

discussed above could be distinguished, in such a

notation, by the use within parenthesis of one of

the four following abbreviations: supersp. for

superspecies, cplx. for species complex, gr. for

species group, and ultrasp. for ultraspecies.

Synklepton

In the last twenty years, various works hâve

demonstrated the existence in nature of particu-

lar animal “ forms ” which cannot be considered

as “ true species ”, such as the unisexual, gynoge-

netic and hybridogenetic forms of fishes of the

généra Poeciliopsis and Poecilia (see e.g.

Schultz, 1977), and the hybridogenetic forms of

anurans of the genus Rana (see e.g. Dubois,

1977 b, 1982 c).

AU these forms, despite their diversity, hâve

the following characteristics in common:

(1) They are of hybrid origin. Some of them

dérivé from hybridizations between two “ good ”

species, others from hybridizations between a

“ good ” species and a hybrid form.

(2) These forms do not behave genetically like

“ good ” species, but hâve the genetic character¬

istics of clones. Hybridogenetic forms are hemi-

clones, and gynogenetic forms are full clones.

(3) These forms cannot survive alone in

nature. They need to “ steal ” gametes from

“ good ” species to realize their reproduction,

thus having recourse to a kind of “ sexual

parasitism ”.

Insisting upon the fact that such forms cannot

be considered as true “ biological species ” (which

reproduce and evolve independently one from

another, and are characterized by a biparental

sexual reproduction, with a génie flow which is

potentially free between ail members of the

species, recombinations between the parental

genomes during the meiosis, etc.). Dubois &

Günther (1982) proposed giving such forms the

name of kleptons, and to call synkleptons the

Source : MNHN, Paris

104

ALAIN DUBOIS

groups consisting of two “ good species ” (or

more) and one (or several) klepton(s) issued

from the hybridization of these species. They

proposed considering kleptons as taxa of the

species-group, belonging to a third taxinomie

category distinct from that of species and of

subspecies, and they suggested rules for the

nomenclature of these forms. Thus, within the

esculenta synklepton of the genus Rana, occur-

ring in ail Europe, where it consists of more than

ten distinct species and kleptons (Dubois, 1982 c),

and where furthermore several difTerent types of

populations do exist (Günther, 1983), the names

suggested for the various existing forms are of

the type Rana (synkl. esculenta) lessonae for the

species, and Rana (synkl. esculenta) kl. esculenta

for the kleptons.

Synkleptons and kleptons are undeniable evo-

lutionary units in nature: they are phylogenetic

and genetic units (within which genetic exchanges

continue to occur between separated forms) and

ecological units (see Dubois & Günther, 1982).

Kleptons may persist as such in nature for long

periods (sometimes thousands of years), but they

are not necessarily evolutionary dead-ends: they

may constitute intermediate stages leading to

other forms, such as polyploid bisexual “ good ”

species (see in particular Dubois, 1977 b, and

Bogart, 1980).

For the time being, kleptons and synkleptons

are known with certainty only in vertebrates,

but, as we hâve suggested (Dubois & Günther,

1982), it is very possible that similar situations

also exist in invertebrates, where they hâve not

yet been recognized as such. It is likely that these

situations are much more abundant in nature

than it has been believed until now, and that

various groups considered until now as “ species

groups ” will prove in the future to be synklep¬

tons, composed of “ good ” species and of

kleptons.

The subgenus

Introduction

Although it is officially recognized in the

International Code of Zoological Nomenclature

(Anonymous, 1985), the subgenus category is

used in a very unequal way in the different

branches of zoology. For many authors, it is

only a “ small genus ” or a “ large species

group The subgenus is rejected by some

authors, in particular those who consider embar-

rassing the presence of a Latin name, which is

later liable to be elevated to generic rank (Dunn,

1943; Duellman, 1977). Such a conception

seems to imply that it is classification which

reflects nomenclature, not the reverse, which is

inacceptable in theory and very disturbing if it

happens in practice.

Within the framework of the définition of the

genus here proposed, it seems that the subgenus

may be conceived as a category distinct from

both the genus and the species group, and which

would allow one to underline the existence of

evolutionary phenomena of a different type. To

illustrate these différences in concrète terms, I

will largely make use of examples from the

amphibians.

The subgenus could be used in two particular

situations:

(1) It first seems indicated to recognize sub-

genera when one can detect, within a genus, a

manifest tendency towards progressive improve-

ment or refinement of the adaptation of the

species to the adaptive zone of the genus: the first

species “ landing ’’ in the zone are still rather

poorly adapted to it, the following ones are more

finely adapted. There may then exist a tendency

to the replacement of the first ones by the

following ones (the subgenera being then succes¬

sive), just as it is possible, in certain conditions,

that two subgenera should subsist together,

possibly in different régions. A good example of

Source : MNHN, Paris

THE GENIUS IN ZOOLOGY

105

this is that of the two subgenera of the Asiatic

genus of anuran amphibians Scutiger, which

show two successive stages, which by the way are

not clearly separated by a discontinuity, in the

adaptation to life in high altitude torrents (Dubois,

1979 a, 1980 b).

(2) A second case where the notion of sub-

genus may apply is that where the adaptive zone

of a genus is large or diversified enough to allow

a subdivision into several adjacent subzones, in

which species groups specialize (which does not

exclude other species or species groups from

retaining a less specialized ecology, which may

cover the whole adaptive zone or several sub¬

zones). These subgenera are then contempora-

neous and sometimes sympatric, sharing among

them resources and niches. A good example in

this case is the genus Ram, the adaptive zone of

which is exceptionally wide and has become

divided into several subzones, which correspond

to as many subgenera (Boulenger, 1918, 1920;

Dubois, 1975, 1976, 1981 b, 1984 a, 1984 e).

This specialization within a zone, which implies

no discontinuity, is distinct from the shift to a

different zone. It is possible to postulate (Dubois,

1975, 1976, 1981 c, 1982 a) that, in this case, the

genetic changes which hâve occurred are minor

and therefore possibly réversible , that these changes

would not constitute a real genetic révolution.

THE CRITER1A OF THE SUBGENUS

Distinction between subgenus and genus

Such a conception of the subgenus entails the

possibility of using three types of criteria to

distinguish subgenus from genus.

Hybridizability

The species of various subgenera of a same

genus, although they may exhibit relatively impor¬

tant différences between them as far as morpho-

logy and ecology are concerned, may be liable to

give viable adult hybrids. The use of this crite-

rion of hybridizability, which was dicussed at

length above, would allow one to definitely

résolve many systematic problems which hâve

long divided the authors. As a matter of fact,

many cases do exist where it is clear that two

groups of species are very close, but at the same

time show significant différences. Such cases are

not rare in particular in amphibians where

furthermore, starting with Noble (1924), the

subgenus category has fallen into disgrâce. Work-

ers are then confronted with the following

alternative: either two généra, or two species

groups, should be recognized. With some authors

insisting upon the différences and others on the

resemblances, one has often observed, without

any justification due to the discovery of new

facts, vacillation between these two attitudes,

which is deleterious to the stability of nomencla¬

ture. In many cases, the intermediate attitude,

which consists in considering the two groups as

subgenera of a same genus, seems best able to

solve the problem, in asserting at the same time

both the resemblance (same genus) and the

différence (different subgenera).

A very good example of this is that of the

problem of the status of the American “ tree

frogs ” grouped under the name Pseudacris (see

Dubois, 1982 a, 1984 b). This name has long

been used in northern America and for this

reason authors conserve it as a generic name,

although the characters which separate Pseud¬

acris from Hyla are very weak. Thus Duellman

(1970: 642) writes:

“ The frogs of the genus Pseudacris differ from most North and Middle

American Hyla by having small dises and greatly reduced webbing on the

the feet. No other external features will distinguish them from Hyla. If these

frogs occurred in South America, they probably would not hâve been

recognized generically. ”

Source : MNHN, Paris

106

ALAIN DUBOIS

As a matter of fact, the species grouped under

Pseudacris are obviously phylogenetically close

to certain species groups of Hyla, as is shown by

the study of mating calls (Blair, 1958, 1959), of

osteology (Gaudin, 1974) and of albumins (Max-

son & Wilson, 1975), as well as the fact that

some species combinations may give hybrids

(Ralin, 1970). However it is true that Pseudacris

has a different overall morphology and a pecu-

liar type of ecology, these species being described

as “ terrestrial ” or “ terrestrial-fossorial ” by

Ralin (1970: 44). There exist therefore good

arguments to consider that these species are

members of the genus Hyla and that they are

“ not like other ” members of this genus. It

appears therefore very justified to treat Pseud¬

acris as a subgenus of Hyla, which until now,

despite abundant discussions on the problem of

“ the validity of Pseudacris ”, no author seems to

hâve contemplated doing, the problem being

always set in terms of the wrong alternative: “ it

is a genus or nothing”.

A second example may be borrowed from the

urodelan amphibians, in which the success of the

hybridization between the European species Pleu-

rodeles waltl and the Asiatic species Tylototriton

verrucosus (Ferrier, Beetschen & Jaylet, 1971)

is enough in itself to consider, in my opinion, the

two contemporaneous species of Pleurodeles and

the four species of Tylototriton (Thorn, 1969;

Nussbaum & Brodie, 1982) as belonging to a

single genus Pleurodeles, although to two distinct

subgenera, to which a third subgenus Echinotri-

ton should be added (see Nussbaum & Brodie,

1982; Frost, 1985; Dubois, 1987 b).

Evolutionary reversibility of characters

A second interesting criterion is that of the

evolutionary reversibility of adaptive characters.

The fact that this reversibility remains possible

indicates that these characters are determined by

a very small number of genes, possibly by a

single regulatory gene. Some examples may be

found in this respect among amphibians.

The presence or absence of digital dises has

long been considered as an important character,

allowing to define généra, if not families, of

anurans. However it is easy to notice that such

dises appeared independently and in parallel in

several lineages of anurans. Species of a same

genus, sometimes very close to each other, may

differ between themselves in this character: thus

some species are “ intermediate ” in this respect

between the subgenera Rana and Hylarana of the

genus Rana (Boulenger, 1920), or some species

which obviously belong to the subgenus Hyla¬

rana, like Rana galamensis and Rana darlingi in

Africa (Laurent, 1956) or like Rana malabarica

in Asia (Dubois, 1981 b) are devoid of dises. It

seems that a single mutation or a very low

number of mutations may be enough to déter¬

mine the presence of terminal dilatations at the

tips of digits and toes in a species which is

usually devoid of such dilatations (Smith & List,

1951). Even if, as is probably the case, such

“ dilatations ” are not identical with true dises, it

seems clear that the presence or absence of dises

is in anurans a very labile character, liable to

appear or disappear independently in different

lineages, and which cannot in itself be used to

separate généra.

Similarly, the presence of intercalary pha¬

langes, although considered by some authors as

an adéquate feature with which families of

anurans may be defined, also seems a character

of little interest, since supernumerary phalanges

may appear as anomalies in species which do not

usually hâve them (Dubois, 1974 b). The same is

true for the presence of nuptial spines on the

breast and forearms of reproductive males, these

characters having appeared independently in

various families and being liable to lack in

species very close to species which hâve them, as

is the case with the almost sibling species Rana

liebigii and Rana vicina (Dubois, 1980 a). Let

us finally cite the absence of a toe on the hind

limbs, which has sometimes been considered as

a generic character, e.g. for separating Sala-

mandrella from Hynobius, while ectrodactyly

may occur in some populations of Hynobius

(Maruyama, 1977) and that it is known to be, in

some cases, of a simple, monogenic, determin-

ism, in amphibians (Dubois, 1977 a).

The criterion of reversibility may also consid¬

ered in a négative way. It is clear that certain

morphological characters or certain ecological,

physiological, etc., adaptations, dépend on a

complex genetic determinism and do not allow a

true reversibility, i.e. a simple return to condi¬

tions strictly identical to the ancestral, plesio-

morphous, conditions, which would imply, so to

speak, a “ genetic counter-revolution ”. Thus the

ventral sucker of the rheophilous tadpoles of the

genus Amolops (Inger, 1954, 1966) does not

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

107

seem liable to be lost at little cost and it may be

supposed that the tadpole of a species issued

from the genus Amolops but which would hâve

secondarily returned to a non-torrenticolous

mode of life would keep some trace of the very

peculiar morphology of its close ancestors, and

at any rate would very probably exhibit appré¬

ciable différences with the tadpoles of the genus

Rana, from which the genus Amolops is probably

issued (Dubois, 1982 a; see also Kuramoto,

Wang & Yü, 1984).

Another interesting example in this respect is

that of the African anurans currently grouped in

the genus Nectophrynoides (Grandison, 1978).

This set of species obviously constitutes a homo-

phyletic unit, separated by a marked morpholo-

gical “ gap ” from the genus Bufo and the other

généra of bufonids (Grandison, 1978). How-

ever, despite the low number of species of this

group currently known, these show a great

diversity of types of reproduction and of devel¬

opment, which may be classed in four major

categories (Grandison, 1978; Wake, 1980). These

four groups show between them différences

which most likely translate important and irré¬

versible genetic changes, and it seems necessary

to account for this important phenomenon by

subdividing this group into four distinct généra

(Dubois, 1982 a, 1987 b): a first one would

correspond to N. osgoodi, which lays numerous

small pigmented eggs in water, which give birth

to tadpoles of the “ generalized ” type (Grandi¬

son, 1978); a second genus would contain N.

malcolmi , which lays large, not numerous, unpig-

mented eggs on the ground, and the development

of which takes place out of water (Grandison,

1978); a third genus, which would keep the name

Nectophrynoides , would group the ovoviviparous

species like N. tornieri and N. viviparus , with

large, but not numerous, eggs (Lamotte &

Xavier, 1972; Lamotte & Lescure, 1977); finally,

a fourth genus would accomodate the viviparous

species N. occidentalis and N. liberiensis, with

few small eggs (Lamotte & Lescure, 1977;

Xavier, 1979). Nothing opposes the création of

a subfamily Tornieriobatinae (Dubois, 1982 a,

1983 b, 1984 d, 1985 a, 1987 b), grouping the

four généra above and the related généra, and

emphasizing the fact that they constitute, within

the Bufonidae, a homophyletic group, but of a

higher rank than that of genus. Many other

examples of this type could be mentioned.

Absence of discontinuities between subgenera

A third and last criterion is the absence of

discontinuities between subgenera. It is not rare

to find species intermediate between two subge¬

nera, difficult to class and which must be rather

arbitrarily attached to one of them. The discov-

ery of such species may lead one to consider two

groups of species which had until then been

considered as distinct généra as subgenera of a

single genus: this was the case for example in

anurans of the subgenera Scutiger and Oreolalax

of the genus Scutiger (Dubois, 1979 a, 1980 b).

Distinction between subgenus and species group

As for this second distinction, it does not rely

at ail on a question of size of the taxon (number

of included species). A genus may be composed

either of subgenera, or of species groups. or

both, or neither (see e.g. Rosen & Bailey, 1963).

The choice between the two categories implies, in

the présent perspective, a value judgement about

the type of évolution which gave birth to the

group in question. If only phenomena of clado-

genesis (spéciation), within a given adaptive

zone, hâve occurred, one will speak of species

groups. If phenomena of anagenesis (différentia¬

tion) also hâve occurred, and in particular if that

implies an ecological specialization, it will be

more justified to recognize subgenera. The large

généra of anurans are exemplary in this respect:

while Bufo and Hyperolius only contain species

groups, ecological and morphological différentia¬

tion is on the other hand much more accentuated

within généra like Rana and Hyla, where it seems

justified to recognize subgenera, as was done by

Boulenger (1918, 1920), Dubois (1975, 1976,

1981b, 1984 a, 1984 e, 1987 b) and others for

Rana, but as has still apparently never been done

for Hyla, despite the interesting remarks of

Martin & Watson (1971), who did not clearly

consider this possibility.

While species groups, which hâve similar écol¬

ogies, are rather rarely sympatric, subgenera,

being specialized, may easily become sympatric

over vast régions: this is the case for several

subgenera of Rana in Europe and in Asia.

Because they are adaptive, the characters of a

subgenus will often be “ good ” taxinomie char-

Source : MNHN, Paris

108

ALAIN DUBOIS

acters, which may be used e.g. in dichotomie

keys and in allowing an easy identification of

specimens (Pasteur, 1964: 97). However, and in

particular because of the reversibility of cbarac-

ters, it will not always be the case: subgenera like

généra may be polythetic.

At the level of the subgenus it may be difficult

to ascertain if true homophyly is involved or if

évolutive parallelism has occurred. Thus the

“ grade ” Hylarana may hâve originated several

times, in Asia, from the “ grade ” Rana s. str.,

giving birth to the various species groups of the

subgenus Hylarana (Boulenger, 1920; Dubois,

1981 a, 1982 a). It is certainly necessary to break

up subgenera which prove to be artificial because

they are polyphyletic; however this problem is

less serious at this level than at the level of the

genus, because the subgenus, contrary to the

genus, expresses a tendency more than a break.

NOMENCLATURAL INTEREST OF THE SUBGENUS

The subgenus présents several nomenclatural

advantages which seem to hâve, at least in part,

escaped many systematists, in particular among

the specialists of amphibians, although they hâve

been stressed by a few authors (Metcalf, 1915;

Schenck, 1937; Simpson, 1943; Edwards, 1953;

Crowson, 1970; Dubois, 1981 c, 1982 a, 1984 e;

etc.).

(1) First of ail, the subgenus is conservative. It

allows one to conserve particularly well-known

old names. This would be the case e.g. if one was

led, to satisfy the hybridizability criterion, to

suppress several current généra of birds or of

bony fishes: the names of these older généra

could be kept, at least in part, for subgenera.

(2) The use of the subgeneric name, when one

désignâtes a species, is optional. This name must

be used in purely systematic or faunistic works,

and may also be used to designate, for instance,

an interesting ecological unit in a work of

ecology. This name must be totally omitted in

works which are far from these concerns: works

of embryology, physiology, biochemistry, etc.,

for which it is mostly important to know the

generic group to which the species studied

belongs (see Rosen & Bailey, 1963). The subge¬

neric name may also be deliberately omitted in

systematic works, when the allocation of a

species to a given subgenus poses some prob-

lems, e.g. nomenclatural ones (see Dubois, 1977 c).

In some révisions, an author may be led to

provisionally subdivide a subgenus into several

subgenera, without always being certain of the

validity of some of them (e.g. because of the lack

of certain types of information on certain species

at the time of the révision). If available names do

exist for these subgenera, it is possible to use

them, but otherwise one must avoid creating

names as long as the validity of the subgenera

has not been demonstrated by good arguments.

This does not raise any nomenclatural problem

since only the generic and spécifie names are

nomenclaturally indispensable.

(3) Finally, the subgeneric name is a unique,

collective , Latin name. It allows one to designate

a group by a name, without having at any time

to describe or qualify it. This may be very useful

when, e.g. in a work of systematic révision, this

group must be designated as such dozens of

times in the text (see e.g. Dubois, 1976). This

simplification of writing is, let us not forget, the

fondamental ground for a existence for a nomen¬

clature like Linnaean nomenclature.

Mayr (1969: 197) has suggested that when a

systematist hésitâtes as to the status to ascribe to

two allopatric groups of populations (species or

subspecies?), he should choose the status of

subspecies. Similarly, when one hésitâtes as to

the status of a group of species (genus or

subgenus?), it seems indicated to consider it as a

subgenus. As a matter of fact, this attitude is

conservative, allowing one to provisionally keep

the two names if they exist. Such a process

indicates at the same time both broad groups to

which are referred the species, and the existence

of a divergence; it is liable to stimulate more

thorough research on the relationships between

the two groups (Dubois, 1982 a, 1984 e).

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

109

Conclusion

The hierarchy genus/subgenus/species group is

by no means compulsory. Large généra may

allow no subdivision, while very small généra

may contain subgenera. Among the various

Latin names which may appear in the désigna¬

tion of a form (names of subspecies, species,

species group, subgenus, etc.), the generic and

spécifie names remain the two most important

ones and the only ones to be indispensable in ail

cases. The complexification of binomial nomen¬

clature expresses the increase in our knowledge,

but it should not lead us to forget that the

Linnaean binomial remains the most important

name, in particular for non-systematists, to

whom systematics must bring useful informa¬

tion. In accepting a rather “ broad ” concept of

the genus, we give pre-eminence to the major

discontinuities: for a non-systematist, the distinc¬

tion between Rana and Platymanlis (character-

ized by its “ terrestrial ” development) or between

Rana and Amolops (characterized by its very

peculiar tadpole) is more important than those

between Rana and Hylarana (Inger, 1954, 1966)

or between Scutiger and Oreolalax (Dubois,

1979 a, 1980 b), because between these latter

groups no clear discontinuity exists. This différ¬

ence must be accounted for in the classification.

Source : MNHN, Paris

Acknowledgements

For the various helps they brought me in the

préparation of this paper, I am most grateful to

John C. Avise (Athens), Roger Bour (Paris),

Jacques Daget (Paris), Pierre Dubois (Moulins),

Cari Gans (Ann Arbor), Jean-Jacques Morère

(Paris), Annemarie Ohler (Paris), Georges and

Nicole Pasteur (Montpellier), Dominique Payen

(Paris), Luc Plateaux (Paris), Denise and Donald

Russell (Paris), Louis Thaler (Montpellier),

Simon Tillier (Paris), Marie-Noëlle Uhl (Paris),

Gregory S. Whitt (Urbana) and Émile Zucker-

kandl (Menlo Park).

Paris, 7 November 1985

Source : MNHN, Paris

REFERENCES

Anonymous, 1985. — International Code of Zoological

Nomenclature. Third édition, London, International

Trust for Zoological Nomenclature : i-xx + 1-338.

Abelson, J., 1980. — A révolution in biology. Science,

209: 1319-1321.

Ager, D. B., 1973. — The nature of the stratigraphical

record. New York, Wiley: 1-114.

Ahearn, J. N., 1980. — Evolution of behavioral

reproductive isolation in a laboratory stock of

Drosophila silvestris. Experientia, 36: 63-64.

Aimar, C., 1977. — Contrôle cytoplasmique de la

segmentation chez les Amphibiens : modifications

des rythmes du clivage de l'œuf après injection de

cytoplasme hétérospécifique. C. r. Acad. Sci., (D),

284 : 215-218.

Aimar, C. & Delarue, M., 1976. — Modification, par

injection de cytoplasme hétérospécifique, de la chro¬

nologie du premier clivage de l'œuf d'Amphibien. C.

r. Acad. Sci., (D), 282 : 1737-1740.

Aimar, C., Delarue, M. & Vilain, C., 1981. —

Cytoplasmic régulation of the duration of cleavage

in Amphibian eggs. J. Embryol. exp. Morph., 64:

259-274.

Alberch. P., 1980. — Ontogenesis and morphological

diversification. Amer. Zool., 20: 653-667.

— 1982. — Developmental constraints in evolution-

ary processes. In: J. T. Bonner (ed.). Evolution and

development, Berlin, Heidelberg & New York, Sprin¬

ger-Verlag: 313-332.

Alberti, B., 1978. — Quelques considérations sur le

problème du “genre”. Linneana belgica, 7 : 175-

177.

Alphêraky, S., 1912. — Sur les notions “genre” et

“ sous-genre ” dans la classification zoologique. In:

C. Oberthür (ed.). Études de Lépidoptérologie

comparée. Rennes, 6 : 35-45.

Anderson S. & Jones, J. K., Jr., 1967. — Recent

Mammals of the world. A synopsis of families. New

York, The Ronald Press Company: i-viii + 1-453.

Ansevin, K. D. & Williams, B. J.. 1974. — Massive

cytoplasmic transfers between early blastomeres of

Rana pipiens eggs. J. exp. Zool., 190: 373-379.

Aquadro, C. F. & Avise. J. C., 1982. — Evolutionary

genetics of birds. VI. A reexamination of protein

divergence using varied electrophoretic conditions.

Evolution, 36: 1003-1019.

Arita, L. H. & Kaneshiro, K. Y., 1979. — Etholo-

gical isolation between two stocks of Drosophila

adiastola Hardy. Proc. Hawaii, entomol. Soc., 13:

31-34.

Arnold, E. N., 1973. — Relationships of the Palae-

arctic lizards assigned to the généra Lacerta. Algy-

roides and Psammodromus (Reptilia: Lacertidae).

Bull. Brit. Mus. (Nat. Hist.), Zoology, 25: 289-366.

Ashlock, P. D., 1971. — Monophyly and associated

terms. Syst. Zool., 20: 63-69.

— 1972. — Monophyly again. Syst. Zool., 21: 430-

438.

— 1980. — An evolutionary systematist’s view of

classification. Syst. Zool.. 28: 441-450.

Avise, J. C., 1983. — Commentary. In: A. H. Brush &

G. A. Clark, Jr. (eds.). Perspectives in ornithology,

Cambridge, Cambridge Univ. Press: 262-270.

Avise, J. C. & Aquadro, C. F., 1982. — A compar¬

ative summary of genetic distances in the Verte-

brates. Patterns and corrélations. Evol. Biol., 15:

151-185.

Avise, J. C., Patton, J. C. & Aquadro, C. F., 1980. —

Evolutionary genetics of birds. Comparative molecu-

lar évolution in New World warblers and rodents. J.

Hered., 71: 302-310.

Avise, J. C. & Smith, M. H., 1974. — Biochemica!

genetics of sunfish. II. Génie similarity between

hybridizing species. Amer. Nat., 108: 458-472.

Ayala, F. J., 1977. — Phylogénies and macromole-

cules. In: T. Dobzhansky, F. J. Ayala. G. L.

Stebbins & J. W. Valentine, Evolution. San Fran¬

cisco, Freeman: 262-313.

— 1982. — Gradualism versus punctualism in spé¬

ciation: reproductive isolation, morphology, gene¬

tics. In: Barigozzi (1982): 51-66.

Ball, I. R., 1983. — On groups, existence and the

ordering of nature. Syst. Zool., 32: 446-451.

Baltzer, F., 1952. — The behaviour of nuclei and

cytoplasm in Amphibian interspecific crosses. S. E.

B. Symposia, 6: 230-242.

Banks, R. C. & Johnson. N. K.. 1961. A review of

North American hybrid hummingbirds. Condor, 63:

3-28.

Barbault, R.. 1974 a. — Le régime alimentaire des

Amphibiens de la savane de Lamto (Côte d’ivoire).

Bull. I.F.A.N., 36 (A) : 952-972.

— 1974 b. — Observations écologiques dans la

Source : MNHN, Paris

112

ALAIN DUBOIS

savane de Lamto (Côte d’ivoire) : structure de

l'herpétocénose. Bull. Eco!., 5 : 7-25.

— 1980. — L’analyse de l'organisation des peuple¬

ments. étape capitale dans l’étude des écosystèmes.

Acta Oecol.. Oecol., Gener., 1 : 237-247.

— 1981. — Ecologie des populations et des peuple¬

ments. Des théories aux faits. Paris, Masson : i-vii +

1-200.

— 1984. — Stratégies de reproduction et démogra¬

phie de quelques Amphibiens Anoures tropicaux.

Oikos, 43 : 77-87.

Barigozzi, C., ed., 1982. — Mechanisms of spéciation.

New York, Liss: i-xiv + 1-546.

Bartlett, H. H., 1940. — History of the generic

concept in botany. Bull. Torrey bot. Club, 67: 349-

362.

Barton, N. H. & Charlesworth, B.. 1984. — Gene-

tic révolutions, founder effects, and spéciation. Ann.

Rev. Ecol. Syst., 15: 133-164.

Beck, W. M.. Jr. & Beck, E. C., 1968. — The concept

of genus in the family Chironomidae. Ann. zool.

fennici, 5: 14-16.

Bengtsson, B. O., 1980. — Rates of karyotype évolu¬

tion in Placental Mammals. Hereditas, 92: 37-47.

Bernardi, G., 1956. — Contribution à l’étude des

catégories taxonomiques. I. Avant-propos, nomen¬

clature et définitions. Bull. Soc. ent. Fr., 61: 194-200.

— 1962. — Discussion. In: F. A. Schilder, Das

Geographische Prinzip in der Taxonomie. In: Ver-

handl. XI. Internat. Kongr. Enlom. Wien, Band 111:

329-333.

— 1980. — Les catégories taxonomiques de la sys¬

tématique évolutive. Mèm. Soc. zool. Fr., 40 : 373-

425.

— 1983. — Le genre et le sous-genre en taxonomie

conventionnelle et évolutive. Bull. Soc. zool. Fr., 108 :

135-141.

Bigelow, R. S., 1956. — Monophyletic classification

and évolution. Syst. Zool., 5: 145-146.

Billington, W. D., James, D. A. & Kirby. D. R. S.,

1968. — Some effects of genetic dissimilarity between

mother and foetus. J. Reprod. Fert., suppl. 3: 1-7.

Blair, W. F., 1958. — Call différence as an isolation

mechanism in Florida species of hylid frogs. Quart.

J. Florida Acad. Sci., 21: 32-48.

— 1959. — Call structure and species groups in U.S.

treefrogs {Hy la). Southwest Nat., 3: 77-89.

— ed., 1972 a. — Evolution in the genus Bufo. Aus-

tin & London, Univ. Texas Press: i-viii + 1-459.

— 1972 b. — Evidence from hybridization. In: Blair

( 1972 a): 196-232 + 388-435.

Blandin. P., 1978. — Les reconstitutions phylogéné¬

tiques " sans fossiles ” : les méthodes, leurs possibi¬

lités, leurs limites. Rev. Quest. sci., 149 : 55-75.

Blondel, J. & Bourlière, F., eds., 1979. — Le

concept de niche en écologie animale. La Terre et la

Vie, 33 : 343-520.

Bock, W. J. & Farrand, J., Jr., 1980. — The number

of species and généra of recent birds: a contribution

to comparative systematics. Amer. Mus. Novit.

2703: 1-29.

Bocquet, C., Génermont, J. & Lamotte, M., eds.,

1976. — Les problèmes de l’espèce dans le règne

animal. Tome I. Mèm. Soc. zool. Fr., 38 : 1-407.

— 1977. — Les problèmes de l’espèce dans le règne

animal. Tome II. Mèm. Soc. zool. Fr., 39 : 1-381.

— 1980. — Les problèmes de l'espèce dans le règne

animal. Tome III. Mèm. Soc. zool. Fr., 40 : 1-453.

Bogart, J. P., 1972. — Karyotypes. In: Blair (1972 a):

171-195 + 361-387.

— 1980. — Evolutionary implications of polyploidy

in Amphibians and Reptiles. In: W. H. Lewis (ed.),

Polyploidy, biological relevance, New York, Plénum

Press: 341-378.

Boulenger, G. A., 1918. — Aperçu des principes qui

doivent régir la classification naturelle des espèces

du genre Rana. Bull. Soc. zool. Fr., 43 : 111-121.

— 1920. — A monograph of the South Asian,

Papuan, Melanesian. and Australian frogs of the

genus Rana. Rec. Indian Mus., 20: 1-126.

Brachet, J., 1957. — Biochemical cytology. New

York, Academie Press: i-xi + 1-516.

Briggs, R. & King, T. J., 1959. — Nucleocytoplasmic

interactions in eggs and embryos. In: J. Brachet &

A. E. Mirsky (eds.), The Cell, vol. I, New York &

London, Academie Press: 537-617.

Britten, R. J. & Davidson, E. H., 1969. — Gene

régulation for higher cells: a theory. Science, 165:

349-357.

Brun, R. & Kobel, H. R., 1977. — Observations on

the fertilization block between Xenopus borealis and

Xenopus laevis laevis. J. exp. Zool., 201: 135-137.

Bush, G. L., 1975. — Modes of animal spéciation.

Ann. Rev. Ecol. Syst., 6: 339-364.

Bush, G. L., Case, S. M., Wilson, A. C. & Patton,

J. L., 1977. — Rapid spéciation and chromosomal

évolution in mammals. Proc. nat. Acad. Sci. U.S.A.,

74: 3942-3946.

Caïn, A. J., 1954. — Subdivisions of the genus

Ptilinopus (Aves, Columbae). Bull. Brit. Mus. (Nat.

Hist.), Zoology, 2: 265-284, 2 pl.

— 1956. — The genus in evolutionary taxonomy.

Syst. Zool., 5: 97-109.

— 1958. — Logic and memory in Linnaeus’s System

of taxonomy. Proc. Linn. Soc. Lond., 169: 144-163.

Camp, W. H.. 1940. — Our changing generic concepts.

Bull. Torrey bot. Club, 67: 381-389.

Carson, H. L., 1975. — The genetics of spéciation at

the diploid level. Amer. Nat., 109: 83-92.

— 1982. — Spéciation as a major reorganization of

polygénie balances. In: Barigozzi (1982): 411-433.

Carson, H. L., Clayton, F. E. & Stalker, H. D.,

1967. — Karyotypic stability and spéciation in

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

113

Hawaiian Drosophila. Proc. nat. Acad. Sci. U.S.A.,

57: 1280-1285.

Carson, H. L. & Kaneshiro, K. Y., 1976. Drosophila

of Hawaii: systematics and ecological genetics. Ann.

Rev. Ecol. Syst., 7: 311-345.

Carson. H. L. & Templeton, A. R., 1984. — Genetic

révolutions in relation to spéciation phenomena: the

founding of new populations. Ann. Rev. Ecol. Syst.,

15: 97-131.

Champion, M. J. & Whitt, G. S., 1976. — Synchro-

nous allelic expression at the glucosephosphate

isomerase A and B loci in interspecific sunfish

hybrids. Biochem. Genel., 14: 723-737.

Charlesworth, B., Lande, R. & Slatkin, M., 1982. —

A neo-Darwinian commentary on macroevolution.

Evolution, 36: 474-498.

Chen, P. S., 1967. — Biochemistry of nucleo-cyto-

plasmic interactions in morphogenesis. In: R. Weber

(ed.), The biochemistry of animal development , vol. II,

New York & London, Academie Press: 115-191.

Cherry, L. M., Case, S. M., Kunkel, J. G. & Wil¬

son, A. C., 1979. — Comparions of frogs, humans,

and chimpanzees. Science, 204: 435.

Cherry, L. M., Case, S. M., Kunkel, J. G., Wyles,

J. S. & Wilson, A. C., 1982. — Body shape metrics

and organismal évolution. Evolution, 36: 914-933.

Cherry, L. M., Case, S. M. & Wilson, A. C., 1978. —

Frog perspective on the morphological différence

between humans and chimpanzees. Science, 200:

209-211.

Clark, A. H., 1911. — The ontogeny of a genus.

Amer. Nat.. 45: 372-374.

Clark, R. B., 1977. — Phylogenetic reconstruction.

Verh. Dtsch. Zool. Ges ., 70: 175-183.

Clarke, A. G. & Hetherington, C. M.. 1972. —

Immunogenetic aspects of maternal/fetal relations.

J. Reprod. Fert.. suppl. 15: 99-118.

Clayton, W. D., 1972. — Some aspects of the genus

concept. Kew Bull., 27: 281-287.

Commoner, B., 1969. — Quelle terre laisserons-nous à

nos enfants? Paris, Seuil : 1-206.

Cope, E. D., 1868. — On the origin of généra. Proc.

Acad. nat. Sci. Phila., 20: 242-300.

Cothran, E. G. & Smith, M. H., 1983. — Chromoso¬

mal and génie divergence in mammals. Svst. Zool,

32: 360-368.

Cousin, G., 1967. — Quelques points de vue sur

l'hybridation chez les animaux. Bull. Soc. zool. Fr.,

92 : 441-485.

Cracraft, J., 1972. — The relationships of the higher

taxa of birds: problems in phylogenetic reasoning.

Condor, 74: 379-392.

Crowson, R. A., 1970. — Classification and biology.

London, Heinemann: i-ix + 1-350.

Crump, M. L., 1974. — Reproductive strategies in a

tropical anuran community. Univ. Kansas Mus. nat.

Hist. Mise. Publ., 61: 1-68.

Cuénot, L. & Têtry, A., 1951. — L'évolution biolo¬

gique. Les faits, les incertitudes. Paris, Masson : i-ix +

1-588.

Daget, J., 1983. — Vers une conception bio-génétique

du genre chez les poissons. Bull. Mus. nain. Hist.

nat., (4), 5 (A) : 399-406.

Danielli, J. F. & Diberardino, M. A., eds., 1979.

Nuclear transplantation. Int. Rev. Cytol., suppl. 9: i-

xv + 1-307.

Darlington, C. D. & Mather, K.. 1949. — The

éléments of genetics. London, Allen & Unwin: 1-

466, 1 pl.

Davidson, E. H. & Britten, R. J.. 1973. — Organiza-

tion, transcription, and régulation in the animal

genome. Quart. Rev. Biol., 48: 565-613.

Davis, L. I. & Webster, F. S., Jr., 1970. — An

intergeneric hybrid flycatcher {Tyrannus x Musci-

vora). Condor, 72: 37-42.

De Beauchamp, P., 1965. — Classe des Rotifères. In:

P.-P. Grasse (ed.). Traité de Zoologie, Tome IV,

Fascicule 3, Paris, Masson : 1225-1379.

Delarue, M., 1977 a. — Conséquence de la greffe de

cytoplasme hétérospécifique sur la gastrulation chez

Bufo bufo L. (Amphibien Anoure). C. r. Acad. Sci.,

(D), 285: 397-400.

— 1977 b. — Influence de la greffe de cytoplasme

hétérospécifique sur la gastrulation chez Bufo bufo

(Amphibien Anoure). Mém. Soc. zool. Fr., 41 : 191-

199.

Demarly, Y., 1979. — The concept of linkat. Proc.

Conf. Broadening Genet. Base Crops, Wageningen,

1978. Wageningen, Pudoc: 257-265.

Diamond, J. M., 1983. — Taxonomy by nucléotides.

Nature, 305: 17-18.

Diberardino, M. A., 1980. — Genetic stability and

modulation of metazoan nuclei transplanted into

eggs and oocytes. Différentiation, 17: 17-30.

Dickinson, W. J.. 1980 a. — Tissue specificity of

enzyme expression regulated by diffusible factors:

evidence in Drosophila hybrids. Science. 207: 995-

997.

— 1980 b. — Complex cis-acting regulatory genes

demonstrated in Drosophila hybrids. Devel. Genet.,

1: 229-240.

Dickinson, W. J., Rowan, R. G. & Brennan, M. D..

1984. — Regulatory gene évolution: adaptive dif¬

férences in expression of alcohol dehydrogenase in

Drosophila melanogaster and Drosophila simulons.

Heredity, 52: 215-225.

Dubois, A., 1974 a. — Liste commentée d'Amphibiens

récoltés au Népal. Bull. Mus. nain. Hist. nat., (3),

213 (Zool. 143) : 341-411.

— 1974 b. — Polydactylie massive, associée à la cli-

nodactylie, dans une population de Rana graeca.

Remarques sur la polydactylie faible et la clinodac-

tylie chez Bufo bufo. (Amphibiens, Anoures). Bull.

Soc. zool. Fr., 99 : 505-521, pl. I-II.

— 1975. — Un nouveau sous-genre (Paa) et trois

Source : MNHN , Paris

114

ALAIN DUBOIS

nouvelles espèces du genre Rana. Remarques sur la

phylogénie des Ranidés (Amphibiens, Anoures).

Bull. Mus. nain. Hist. nat., (3), 324 (Zool. 231) :

1093-1115.

— 1976. — Les Grenouilles du sous-genre Paa du

Népal (famille Ranidae, genre Rana). Cahiers népa¬

lais — Documents. Paris, C.N.R.S., 6 : i-vi + 1-275.

— 1977 a. — Une mutation dominante déterminant

l’apparition de diverses anomalies digitales chez

Rana temporaria (Amphibiens, Anoures). Bull. Soc.

zool. Fr.. 102 : 197-213.

— 1977 b. — Les problèmes de l’espèce chez les

Amphibiens Anoures. Mém. Soc. zool. Fr.. 39 : 161-

284.

— 1977 c. — Morphologie et statut systématique de

Rana fansipani Bourret. 1939 (Amphibiens. Anoures).

Bull. Mus. nain. Hist. nat.. (3). 480 (Zool. 337) :

981-992.

— 1979 a. — Une espèce nouvelle de Sculiger (Amphi¬

biens, Anoures) du nord de la Birmanie. Rev. suisse

Zool.. 86 : 631-640.

— 1979 b. — Métamérie. tératologie et génétique.

Bull. Soc. zool. Fr.. 104 : 325-331.

— 1980 a. — L'influence de l'homme sur la réparti¬

tion des Amphibiens dans l’Himalaya central et

occidental. C. r. Soc. Biogéogr., 55 : 155-178.

— 1980 b. — Notes sur la systématique et la réparti¬

tion des Amphibiens Anoures de Chine et des

régions avoisinantes. IV. Classification générique et

subgénérique des Pelobatidae Megophryinae. Bull.

Soc. linn. Lyon, 49 : 469-482.

— 1981 a. — Hybridation interspécifique et notion

de genre en zoologie. C. r. Acad. Sci., (3), 292 : 201-

203.

— 1981 b. — Liste des genres et sous-genres nomi¬

naux de Ranoidea (Amphibiens Anoures) du monde,

avec identification de leurs espèces-types : consé¬

quences nomenclaturales. Monit. zool. iial., (n. s.),

15, suppl. : 225-284.

— 1981 c. — Quelques réflexions sur la notion de

genre en zoologie. Bull. Soc. zool. Fr., 106 : 503-513.

— 1982 a. — Les notions de genre, sous-genre et

groupe d’espèces en zoologie à la lumière de la

systématique évolutive. Monit. zool. ital., (n. s.), 16 :

9-65.

— 1982 b. — Interventions à la suite des conférences

de C. Petit, L. Thaler et A. Jacquard. In: R. A.

Fisher et l'histoire de la génétique des populations.

Revue de Synthèse. (3), 103-104 : 372-373, 397-399,

430-431.

— 1982 c. — Notes sur les Grenouilles vertes (groupe

de Rana kl. esculenta Linné, 1758). I. Introduction.

Alytes, 1 : 42-49.

— 1983 a. — Hybridation interspécifique, similarité

génétique, parenté phylogénétique et classification

supraspécifique en zoologie. Année biol., (4), 22 : 37-

68.

— 1983 b. — Classification et nomenclature supra-

générique des Amphibiens Anoures. Bull. Soc. linn.

Lyon, 52 : 270-276.

— 1983 c. — Note préliminaire sur le genre Leptola-

lax Dubois, 1980 (Amphibiens, Anoures), avec

diagnose d’une espèce nouvelle du Vietnam. Alytes,

2 : 163-170.

— 1984 a. — Miscellanea nomenclatorica batracho-

logica (I). Alytes, 3 : 39-43.

— 1984 b. — Miscellanea nomenclatorica batracho-

logica (II). Alytes, 3 : 83-84.

— 1984 c. — Miscellanea nomenclatorica batracho-

logica (IV). Alytes, 3 : 103-110.

— 1984 d. — La nomenclature supragénérique des

Amphibiens Anoures. Mém. Mus. nain. Hist. nat.,

(A), 131 : 1-64.

— 1984 e. — Note préliminaire sur le groupe de

Rana limnocharis Gravenhorst, 1829 (Amphibiens,

Anoures). Alytes, 3 : 143-159.

— 1985. — Miscellanea nomenclatorica batracholo-

gica (VII). Alytes, 4 : 61-78.

— 1986. — À propos de l’emploi controversé du

terme “ monophyletique ” : nouvelles propositions.

Bull. Soc. linn. Lyon, 55 : 248-254.

— 1987 a. — Neoteny and associated terms. Alytes. 4:

122-130.

— 1987 b. — Miscellanea taxinomica batrachologica.

Alytes. 5: 7-95.

— 1988. — Some comments on the genus concept in

zoology. Monit. zool. ital., (n.s.), 22: 27-44.

Dubois, A. & Günther, R., 1982. — Klepton and

synklepton: two new evolutionary systematics cate¬

gories in zoology. Zool. Jb. Syst., 109: 290-305.

Dubois, A. & Khan, M. S., 1980. — A new species of

frog (genus Rana, subgenus Paa) from northern

Pakistan (Amphibia, Anura). J. Herpet., 13: 403-

410.

Dubois, A. & Matsui, M., 1983. — A new species of

frog (genus Rana, subgenus Paa) from Western

Népal (Amphibia: Anura). Copeia, 1983: 895-901.

Duellman, W. E., 1970. — The hylid frogs of Middle

America. Monogr. Mus. nat. Hist. Univ. Kansas, 1,

vol. 2: 429-753, pl. 1-72.

— 1977. — Nepalese frogs: the subgenus in herpeto-

logical classification. Herpet. Rev., 8: 11-12.

— 1978. — The biology of an équatorial herpeto-

fauna in Amazonian Ecuador. Univ. Kansas Mus.

nat. Hist. mise. Publ., 65: 1-352, pl. 1-4.

— 1979. — The numbers of Amphibians and Rep¬

tiles. Herpet. Rev., 10: 83-84.

Duellman, W. E. & Pyles, R. A., 1983. — Acoustic

resource partitioning in Anuran communities. Copeia,

1983 : 639-649.

Dunn, E. R., 1943. — Lower categories in herpeto-

logy. Annals New York Acad. Sci., 44: 123-131.

Dupuis, C., 1979. — Permanence et actualité de la

systématique : la “ systématique phylogénétique ”

de W. Hennig (historique, discussion, choix de

références). Cah. Nat.. Bull. Nat. paris., (n. s.), 34 :

1-69.

Source : MNHN, Paris

THE GENUS IN ZOOLOGY

115

— 1984. — Willi Hennig's impact on taxonomie

thought. Ann. Rev. Ecol. Syst., 15: 1-24, 1 pl.

Edwards, F. W., 1953. — Généra and subgenera.

Syst. Zoo/., 2: 135.

Eldredge, N. & Gould, S., 1972. — Punctuated

equilibria: an alternative to phyletic gradualism. In:

T. J. M. Schopf (ed.), Models in paleobiology , San

Francisco, Freeman, Cooper & Co.: 82-115.

Elinson, R. P., 1974. — A block to cross-fertilization

located in the egg jelly of the frog Rana clamitans. J.

Embr. exp. Morph., 32: 325-335.

— 1975 a. — Viable Amphibian hybrids produced

by circumventing a block to cross-fertilization (Rana

clamitans $ x Rana catesbeiana S). J. exp. Zool.,

192: 323-329.

— 1975 b. — Fertilization of green frog {Rana clami¬

tans) eggs in their native jelly by bullfrog (Rana

catesbeiana) sperm. J. exp. Zool.. 193: 419-423.

— 1981. — Genetic analysis of developmental arrest

in an Amphibian hybrid (Rana catesbeiana. Rana

clamitans). Devel. Biol., 81: 167-176.

Endler, J. A., 1977. — Géographie variation, spécia¬

tion. and clines. Princeton, Princeton Univ. Press: i-

ix + 1-246.

Fankhauser, G., 1955. — The rôle of nucléus and

cytoplasm. In: B. H. Willier, P. A. Weiss &

V. Hamburger (eds.). Analysis of development,

Philadelphia & London, Saunders: 126-150.

Farris, J. S., 1974. — Formai définitions of paraphyly

and polyphyly. Syst. Zool., 23: 548-554.

ed., 1979. — The Willi Hennig memorial sympo¬

sium. Syst. Zool., 28: 415-519.

Ferrier, V., Beetschen, J.-C. & Jaylet, A., 1971. —

Réalisation d'un hybride intergénérique viable entre

deux Amphibiens Urodèles européen et asiatique

(Pleurodeles waltlii $ x Tylototriton verrucosus $,

Salamandridae). C. r. Acad. Sci., (D), 272: 3079-

3082, pl. I.

Findley, J. S., 1979. — Comparisons of frogs, humans,

and chimpanzees. Science, 204: 434-435.

Fischer, J.-L., 1981. — L'hybridologie et la zootaxie

du siècle des lumières à " L'origine des espèces

Revue de Synthèse, 101-102 : 47-72.

Fischer, J.-L. & Rey, R., 1983. — De l’origine et de

l'usage des termes taxinomie - taxonomie. In: Docu¬

ments pour l'histoire du vocabulaire scientifique,

Paris, CNRS, Institut national de la langue fran¬

çaise, vol. 5 : 97-113.

Flourens, P., 1845 a. — Cuvier. Histoire de ses

travaux. Paris, Paulin, seconde édition : i-viii + 1-

324.

— 1845 b. — De l'instinct et de l’intelligence des

animaux. Paris, Paulin, seconde édition : i-viii + 1-

207.

— 1856. — Cours de physiologie comparée. De l'onto¬

logie ou étude des êtres. Paris, Baillière : i-viii + 1-

184.

Fouquette, M. J., Jr. & Delahoussaye, A. J., 1977. —

Sperm morphology in the H y la rubra group (Amphi-

bia, Anura, Hylidae), and its bearing on generic

status. J. Herpet., 11: 387-396.

Frost. D. R. (ed.), 1985. — Amphibian species of the

world. Lawrence, Allen Press & Assoc. Syst. Coll.:

[i-iv] + i-v + 1-732.

Gallien, C. L„ 1970. — Recherches sur la greffe

nucléaire interspécifique dans le genre Pleurodeles

(Amphibiens Urodèles). Ann. Embr. Morph., 3 :

145-192.

Gallien, L., 1972. — Aspects des relations nucléo¬

cytoplasmiques dans l'ontogenèse des Amphibiens.

Bull. Soc. zool. Fr., 97 : 233-263, 1 pl.

Galton, F., 1894. — Discontinuity in évolution.

Mind, (n. s.), 3 : 362-372.

Gaudin, A. J.. 1974. — An osteological analysis of

holarctic tree frogs, family Hylidae. J. Herpet., 8:

141-152.

Génermont, J. & Lamotte, M., 1980. — Le concept

biologique de l’espèce dans la zoologie contempo¬

raine. Mém. Soc. zool. Fr., 40 : 427-452.

Ghigi, A., 1936. — Affinité gametica ed affinité

sistematica alla luce dell’ esperienza. Riv. Biol., 20:

1 - 10 .

Gisin, H., 1964. — Synthetische Théorie der Systema-

tik. Z. zool. Syst. Evol.-Forsch., 2: 1-17.

— 1966. — Signification des modalités de l’évolution

pour la théorie de la systématique. Z. zool. Syst.

Evol.-Forsch., 4: 1-12.

Goldschmidt, R., 1940. — The material basis of

évolution. New Haven, Yale Univ. Press: i-xi + 1-

436.

Gorham, S. W„ 1977. — A note regarding the

number of vertebrate species and the scheme of the

présent classification. J. New Brunswick Mus., 1977:

53-56.

Gould, S. J., 1966. — Allometry and size in ontogeny

and phylogeny. Biol. Rev., 41: 587-640.

— 1977. — Ontogeny and phylogeny. Cambridge,

Mass., Belknap Press: i-xvi + 1-501.

— 1980. — 1s a new and general theory of évolution

emerging? Paleobiology, 6: 119-130.

— 1982. — Darwinism and the expansion of evolu-

tionary theory. Science, 216: 380-387.

Gould, S. J. & Eldredge, N., 1977. — Punctuated

equilibria: the tempo and mode of évolution recon-

sidered. Paleobiology, 3: 115-151.

Gould, S. J. & Lewontin, R. C., 1979. — The

spandrels of San Marco and the Panglossian para-

digm: a critique of the adaptationist programme.

Proc. roy. Soc. Lond., (B), 205: 581-598.

Grandison, A. G. C., 1978. — The occurrence of

Nectophrynoides (Anura Bufonidae) in Ethiopia. A

new concept of the genus with a description of a

new species. Monit. zool. ital., (n. s.), 11, suppl.:

119-172.

Source : MNHN, Paris

116

ALAIN DUBOIS

Grant, V., 1963. — The origin of adaptations. New

York, Columbia Univ. Press: 1-606.

Gray, A. P.. 1958. — Bird hybrids. A check-list with

bibliography. Bucks, Commonwealth Agricultural

Bureaux: i-x + 1-390.

— 1972. — Mammalian hvbrids. A check-list with

bibliography. Bucks, Commonwealth Agricultural

Bureaux: i-x + 1-262, 1 pl.

Greenman, J. M„ 1940. — Généra from the stand-

point of morphology. Bull. Torrey bot. Club , 67:

371-374.

Günther, R., 1983. — Zur Populationsgenetik der

mitteleuropâischen Wasserfrôsche des Rana escu-

/en/a-Synkleptons (Anura, Ranidae). Zool. An:.,

211 : 43-54.

Guttman, S. I., 1985. — Biochemical studies of

anuran évolution. Copeia, 1985: 292-309.

Haeckel, E„ 1868. — Natürliche Schôpfungsges-

chichte. Berlin, Reimer: i-xvi + 1-568, front. +

pl. I-V11I + pp. 204 b-c.

Hafner, M. S., Remsen, J. V., Jr. & Lanyon, S. M.,

1984. — Bird versus mammal morphological diver-

sity. Evolution, 38: 1154-1156.

Haldane, J. B. S., 1922. — Sex-ratio and unisexual

sterility in hybrid animais. J. Genet., 12 : 101-109.

Hennig, W.. 1936. — Beziehungen zwischen geogra-

phischer Verbreitung und systematischer Gliede-

rung bei einigen Dipterenfamilien: ein Beitrag zum

Problem der Gliederung systematischer Kategorien

hôherer Ordnung. Zool. An:., 116 : 161-175.

— 1950. — Grund:üge einer Théorie der phylogenetis-

chen Systematik. Berlin, Deutscher Zentralverlag: 1-

370.

— 1966. — Phylogenetic systematics. Urbana, Chi¬

cago & London, Univ. Illinois Press: i-vii + 1-263.

Heyer, W. R., 1973. — Ecological interactions of frog

larvae at a seasonal tropical location in Thailand. J.

Herpet., 7: 337-361.

— 1974. — Niche measurements of frog larvae from

a seasonal tropical location in Thailand. Ecology,

55: 651-656.

— 1976. — Studies in larval amphibian habitat par-

titioning. Smithsonian Contrib. Zool., 242: i-iii + 1-

27.

— 1979. — Annual variation in larval amphibian

populations within a temperate pond. J. Wash.

Acad. Sci., 69: 65-74.

Hoyer, B. H.. McCarthy, B. J. & Bolton, E. T..

1964. — A molecular approach in the systematics of

higher organisms. Science, 144 : 959-967.

Hubbs, C. L„ 1943. — Criteria for subspecies, species

and généra, as determined by researches on fishes.

Annals New York Acad. Sci., 44 : 109-121.

— 1955. — Hybridization between fish species in

nature. Syst. Zool., 4 : 1-20.

Hubbs, C., 1970. — Teleost hybridization studies.

Proc. Calif. Acad. Sci., (4), 38: 289-297.

Hubbs, C. & Drewry, G. E., 1960. — Survival of Fl

hybrids between cyprinodont fishes, with a discus¬

sion of the corrélation between hybridization and

phylogenetic relationship. Publ. Inst. mar. Sci., 6'

81-91.

Huxley, J. S., 1958. — Evolutionary processes and

taxonomy with spécial référencé to grades. Uppsala

Univ. Arsskr., 6: 21-39.

Illies, J., 1970. — Die Gattung als ôkologische

Grundeinheit. Faun.-ôkol. Mitt. Kiel., 3: 369-372.

Inger, R. F., 1954. — Systematics and zoogeography

of Philippine Amphibia. Fieldiana: Zool., 33: 183-

531.

— 1958. — Comments on the définition of généra.

Evolution, 12: 370-384.

— 1966. — The systematics and zoogeography of

the Amphibia of Bornéo. Fieldiana: Zool., 52: 1-

402.

Inger, R. F. & Colwell, R. K., 1977. — Organiza-

tion of contiguous communities of Amphibians and

Reptiles in Thailand. Ecol. Monogr., 47: 229-253.

Jacob, F., 1970. — La logique du vivant. Une histoire

de l’hérédité. Paris, Gallimard : 1-354.

Jacob, F. & Monod, J., 1961. — Genetic regulatory

mechanisms in the synthesis of proteins. J. mol.

Biol., 3: 318-356.

James, M. T.. 1953. — An objective aid in determining

generic limits. Syst. Zool., 3: 136-137.

Johnsgard, P. A., 1960. — Hybridization in the

Anatidae and its taxonomie implications. Condor,

62 : 25-33.

— 1970. — A summary of intergeneric New World

quail hybrids, and a new intergeneric hybrid combi¬

nation. Condor, 72 : 85-88.

Jolicoeur, P. & Mosimann, J. E., 1960. — Size and

shape variation in the painted turtle, a principal

component analysis. Growth, 24: 339-354.

Kerr, M. G., 1968. — Immunological rejection as a

cause of abortion. J. Reprod. Fert., suppl. 3: 49-55.

Kessler, L. G. & Avise, J. C., 1985. — A comparative

description of mitochondrial DNA différentiation in

selected avian and other vertebrate généra. Mol.

Biol. Evol., 2 : 109-125.

King, M.-C. & Wilson, A.C., 1975. — Evolution at

two levels in humans and chimpanzees. Science,

188: 107-116.

Kinsey, A. C., 1936. — The origin of higher categories

in Cynips. Indiana Univ. Publ.. Bloomington, Sci.

Ser., 4: 1-334.

Kiriakoff, S. G.. 1954. — Chorologie et systématique

phylogénétique. Bull. & Ann. Soc. roy. Ent. Belg., 90 :

185-198.

— 1965. — La vicariance géographique et la taxo¬

nomie. C. r. Soc. Biogéogr., 41 : 103-115.

Kluge, A. G., 1966. — A new pelobatine frog from

the Lower Miocene of South Dakota with a discus¬

sion of the évolution of the Scaphiopus-Spea com-

Source : MNHN , Paris

THE GENUS IN ZOOLOGY

117

plex. Los Angeles County Mus. Conlrib. Sci., 113: 1-

26.

Kuramoto, M., Wang, C.-S. & Yü, H.-T., 1984. —

Breeding, larval morphology and experimental hybri¬

dization of Taiwanese brown frogs. Ranci longicrus

and R. sauteri. J. Herpet., 18: 387-395.

Lamotte, M. & Lescure, J., 1977. — Tendances

adaptatives à l’affranchissement du milieu aqua¬

tique chez les Amphibiens Anoures. La Terre ei la

Vie, 31 : 225-311.

Lamotte, M. & Xavier. F., 1972. — Les Amphibiens

Anoures à développement direct d'Afrique. Obser¬

vations sur la biologie de Nectophrynoides tornieri

(Roux). Bull. Soc. zool. Fr., 97 : 413-428.

Lande, R., 1979. — Quantitative genetic analysis of

multivariate évolution, applied to brain: body size

allometry. Evolution , 33: 402-416.

— 1980. — Genetic variation and phenotypic évolu¬

tion during allopatric spéciation. Amer. Nat., 116:

463-479.

Landres, P. B. & Macmahon, J. A., 1980. — Guilds

and community organization: analysis of an oak

woodland avifauna in Sonora, Mexico. The Auk,

97: 351-365.

Lanza, B., Cei, J. M. & Crespo. E. G., 1976. —

Further immunological evidence for the validity of

the family Bombinidae (Amphibia Salientia). Monit.

zool. ital., (n. s.), 10: 311-314.

Larson, A., Prager, E. M. & Wilson, A. C., 1984. —

Chromosomal évolution, spéciation and morpholo-

gical change in vertebrates: the rôle of social

behaviour. Chromosomes today, 8: 215-228.

Laurent, R. F., 1953. — Une méthode pour la

recherche des meilleurs caractères taxonomiques

fournis par les proportions. Ann. Soc. roy. zool.

Belg., 84 : 271-282.

— 1956. — Notes herpétologiques africaines. I. Revue

Zool. Bot. afr., 53 : 229-256.

— 1964. — Reptiles et Amphibiens de l'Angola.

(Troisième contribution). Publ. Cuit. Comp. Diam.

Angola, 67 : 11-165.

— 1967. — Un método para la determinaciôn de los

mejores caractères de proporciôn. Acta zool. lil-

loana, 23 : 97-109, pl. 1-IV.

— 1972. — Exploration du Parc National des Virunga

(2 e série). Fascicule 22, Amphibiens : 1-125, pl. I-XI.

— 1973. — The natural classification of the Arthro-

leptinae (Amphibia, Hyperoliidae). Revue Zool. Bot.

afr., 87: 666-678.

— 1981. — Phénogrammes d'Anoures basés sur la

morphométrie. Monit. zool. ital., (n. s.), 15, suppl.:

1 - 22 .

Lemen, C. A. & Freeman, P. W„ 1984. — The genus:

a macroevolutionary problem. Evolution, 38: 1219-

1237.

Lewis, H., 1962. — Catastrophic sélection as a factor

in spéciation. Evolution, 16: 257-271.

Lewontin, R. C., 1965. — Comment. In: H. G. Baker

& G. L. Stebbins (eds.), The genetics of colonizing

species, New York, Academie Press: 481-484.

— 1974. — The genetic basis of evolutionary change.

New York, Columbia Univ. Press: i-xv + 1-346.

Linné, C. von, 1751. — Pliilosophica Botanica in qua

explicantur Fundamenta Botanica cum Definitionibus

partium, Exemplis terminorum, Observationibus rario-

rum, adjectis Figuris aeneis. Stockholmiae & Amste-

lodami, G. Kiesewetter & Z. Châtelain: i-ii + 1-362,

pl. I-IX.

Littlejohn, M. J., 1969. — The systematic signifi-

cance of isolating mechanisms. In: Systematic Bio-

logy. Proceedings of an international Conférence,

Washington, National Academy of Sciences: 459-

482.

Lotsy, J. P., 1918. — La quintessence de la théorie du

croisement. Arch. nèerl. Sci. exact, nat., (3b), 3 :

351-353.

Lucotte, G. & Dubouch, P., 1980. — Étude électro¬

phorétique de l'hybride expérimental entre Papio

anubis et P. cynocephalus. Biochem. Syst. Ecol., 8 :

323-327.

Lynch, J. D., 1970. — Systematic status of the

American leptodactylid frog généra Engystomops,

Eupemphix, and Phvsalaemus. Copeia, 1970: 488-

496.

— 1971. — Evolutionary relationships, osteology, and

zoogeography of leptodactyloid frogs. Univ. Kansas

Mus. nat. Hist. mise. Publ., 53: 1-238.

Lynch, J. F. & Ames, P. L., 1970. — A new hybrid

hummingbird, Archilochus alexandri x Selasphorus

sasin. Condor, 72: 209-212.

Macintyre, R. J., 1982. — Regulatory genes and

adaptation. Past. présent, and future. Evol. Biol.,

15: 247-285.

Mandelbrot, B., 1956. — La distribution de Willis-

Yule, relative aux nombres d'espèces dans les genres

biologiques. C. r. Acad. Sci., 242 : 2223-2226.

Martin, A. A. & Watson, G. F., 1971. — Life history

as an aid to generic délimitation in the family

Hylidae. Copeia, 1971: 78-89.

Martin, P. G. & Hayman. D. L., 1965. — A quanti¬

tative method for comparing the karyotypes of

related species. Evolution, 19: 157-161.

Martin, R. F. & Selander, R. K., 1975. — Morpho-

logical and biochemical evidence of hybridization

between cave and barn swallows. Condor, 77: 362-

364.

Maruyama. T., 1977. — Digital variation in the hind

legs of the salamander Hynobius lichenatus. Japon J.

Herpet., 7: 10-14.

Maxson, L. R., 1976. — The phylogenetic status of

phyllomedusine frogs (Hylidae) as evidenced from

immunological studies of their sérum albumins.

Experientia, 32: 1149-1150.

— 1977. — Immunological détection of convergent

évolution in the frog Anotheca spinosa (Hylidae).

Syst. Zool., 26: 72-76.

Source : MNHN , Paris

118

ALAIN DUBOIS

Maxson, L. R., Sarich, V. M. & Wilson. A. G, 1973. —

Genetic distance and hybridization potential. Gene-

lics, 74. suppl.: S 176.

Maxson, L. R. & Wilson, A. G, 1974. — Convergent

morphological évolution detected by studying pro¬

teins of tree frogs in the H y la eximia species group.

Science, 185: 66-68.

— 1975. — Albumin évolution and organismal évo¬

lution in tree frogs (Hylidae). Syst. Zoo/., 24: 1-15.

Mayr, E., 1942. — Systematics and the origin of

species front the view point of a zoologist. New York,

Columbia Univ. Press: i-xiv + 1-334.

— 1943. — Criteria of subspecies, species and généra

in ornithology. Annals New York Acad. Sci., 44:

133-139.

— 1954. — Change of genetic environment and évo¬

lution. In: J. Huxley, A. C. Hardy & E. B. Ford

(eds.). Evolution as a process, London, Allen &

Unwin: 157-180.

— 1963. — Anima! species and évolution. Cambridge,

Mass., Belknap Press of Harvard Univ. Press: i-xiv

+ 1-797.

— 1965. — Classification, identification and sequence

of généra and species. L'Oiseau et la Revue fr.

Ornith., 35, n° spécial: 90-95.

— 1969. — Principles of systematic zoology. New

York, McGraw-Hill: i-xiii + 1-428.

— 1970. — Populations, species, and évolution. Cam¬

bridge, Mass., Belknap Press of Harvard Univ.

Press: i-xvii + 1-453.

— 1974. — Cladistic analysis or cladistic classifica¬

tion? Z. zool. Syst. Evol.-Forsch., 12: 94-128.

— 1975. — The unity of the génotype. Biol. Zbl., 94:

377-388.

— 1978. — (Analysis of White, 1978). Syst. Zool.,

27 : 478-482.

— 1981. — Biological classification: toward a syn-

thesis of opposing méthodologies. Science, 214: 510-

516.

— 1982 a. — The growth of biological thought. Diver-

sity, évolution, and inheritance. Cambridge, Mass. &

London, Belknap Press of Harvard Univ. Press: i-

xiii + 1-974.

— 1982 b. — Spéciation and macroevolution. Evolu¬

tion , 36: 1119-1132.

— 1982 c. — Processes of spéciation in animais. In:

Barigozzi (1982): 1-19.

Mayr, E., Linsley, E. G. & Usinger, R. L., 1953. —

Methods and principles of systematic zoology. New

York, McGraw-Hill: i-ix + 1-336.

Mayr, E. & Short, L. L., 1970. — Species taxa of

North American birds. A contribution to compara¬

tive systematics. Publ. Nuttall Ornith. Club, 9: 1-127.

McAllister, D. E. & Coad, B. W„ 1978. — A test

between relationships based on phenetic and cladis¬

tic taxonomie methods. Canad. J. Zool., 56: 2198-

2210.

Mecham, J. S., 1961. — Isolating mechanisms in

Anuran Amphibians. In: W. F. Blair (ed.). Verté¬

brale spéciation, Austin, University of Texas Press -

24-61.

Medawar, P. B.. 1953. Some immunological and

endocrinological problems raised by the évolution

of viviparity in Vertebrates. In: Evolution. Symposia

of the Society for Experimental Biology, Cambridge

Univ. Press: 7: 320-338.

Mertens, R., 1950. — Über Reptilienbastarde. Senc-

kenbergiana, 31: 127-144, pl. 1-3.

— 1956. — Über Reptilienbastarde, IL Senckenber-

giana biol., 37: 383-394.

— 1964. — Über Reptilienbastarde, III. Senckenber-

giana biol., 45: 33-49.

— 1968. — Über Reptilienbastarde, IV. Senckenber-

giana biol., 49: 1-12.

— 1972. — Über Reptilienbastarde, V. Senckenber-

giana biol., 53: 1-19.

Metcalf, M. M., 1915. — Genus and subgenus.

Science, 42: 796-797.

Michener, C. D„ 1957. — Some bases for higher

categories in classification. Syst. Zool., 6: 160-173.

Milstein, P. le S., 1979. — The evolutionary signifi-

cance of wild hybridization in South African high-

veld ducks. The Ostrich, suppl. 13: 1-48.

Moenkhaus, W. J., 1910. — Cross fertilization among

fishes. Proc. Indiana Acad. Sci., 1910: 353-393.

Montalenti, G., 1938. — L'ibridazione interspecifica

degli Anfibî Anuri. Archivio zoologico, suppl. au

vol. 25, vol. 4: 157-213.

Moore. L, 1976. — Splitting well defined généra and

the use of cryptic characters for delimiting généra.

Syst. Zool., 25: 81.

Moore, J. A., 1955. — Abnormal combinations of

nuclear and cytoplasmic Systems in frogs and toads.

Adv. Genet., 7: 139-182.

— 1967. — Diploid and haploid hybridization of

different populations of the Rana pipiens complex.

IL Experiments with females from Oklahoma. J.

exp. Zool., 165: 461-474.

Nei, M., 1972. — Genetic distance between popula¬

tions. Amer. Nat., 106: 283-292.

Nelson, J. S., 1984. — Fishes of the world. Second

édition. New York, Wiley: i-xvii + 1-523.

Noble, G. K., 1924. — Contributions to the herpeto-

logy of the Belgian Congo based on the collection of

the American Muséum Congo Expédition, 1909-

1915. Part III. Amphibia. Bull. Amer. Mus. nat.

Hist., 49: 147-347. pl. XXIII-XLII.

Nussbaum, R. A. & Brodie, E. D., Jr„ 1982. —

Partitioning of the salamandrid genus Tvlototriton

Anderson (Amphibia: Caudata) with a description

of a new genus. Herpetologica, 38: 320-332.

Oliver. C. G., 1979. — Genetic différentiation and

hybrid viability within and between some Lepido-

ptera species. Amer. Nat., 114: 681-694.

Ortega, A., Maury, M. E. & Barbault, R., 1982.

Source : MNHN , Paris

THE GENUS IN ZOOLOGY

119

Spatial organization and habitat partitioning in a

mountain lizard community of Mexico. Acta oecol.,

Oecol. gener., 3: 323-330.

Ouweneel, W. J., 1976. — Developmental genetics of

homoeosis. Adv. Genet., 18: 179-248.

Paclt, J., 1957. — A protest against the atomizing of

généra, with spécial reference to the Hepialidae

(Lepidoptera). Syst. Zool., 6: 51-52.

Parker, H. R., Philipp, D. P. & Whitt, G. S.,

1985 a. — Gene regulatory divergence among spe-

cies estimated by altered developmental patterns in

interspecific hybrids. Mol. Biol. Evol., 2: 217-250.

— 1985 b. — Relative developmental success of inter¬

specific Lepomis hybrids as an estimate of gene

regulatory divergence between species. J. exp. Zool..

233: 451-466.

Parkes, K. C., 1961. — Intergeneric hybrids in the

family Pipridae. Condor. 63: 345-350.

Pasdar, M., Philipp, D. P., Mohammad, W. A. &

Whitt, G. S., 1984. — Différences in tissue expres¬

sions of enzyme activities in interspecific sunfish

(Centrarchidae) hybrids and their backcross prog-

eny. Biochem. Genet., 22: 931-956.

Pasdar, M„ Philipp, D. P. & Whitt, G. S.. 1984. —

Enzyme activities and growth rates in two sunfish

species and their hybrids. J. Hered.. 75: 453-456.

Pasteur, G., 1964. — Recherches sur l'évolution des

Lygodactyles, lézards afro-malgaches actuels. Trav.

Inst. sci. chérif. (Zool.). 29 : 1-132, pl. I-X1I, I carte

h.-t.

— 1976. — The proper spelling of taxonomy. Syst.

Zool.. 25: 192-193.

— 1977. — Quelques commentaires à propos de l’es¬

pèce chez les Lygodactyles. Mém. Soc. zool. Fr.. 39 :

335-358.

— 1982. — Spéciation and transspecific évolution.

In. Barigozzi (1982): 511-538.

— 1985. — Les paramètres statistiques communé¬

ment utilisés dans l’exploitation des résultats de

l’électrophorèse des protéines et leur avenir en

systématique. Mém. Soc. zool. Fr.: 141-180.

Pasteur, G. & Pasteur, N., 1980. — Les critères

biochimiques et l'espèce animale. Mém. Soc. zool

Fr., 40 : 99-150.

Patton, J. L. & Sherwood, S. W.. 1983. — Chromo¬

some évolution and spéciation in rodents. Annual

Rev. Ecol. Syst., 14: 139-158.

Pépin, H.. Moreau, G., Marazzato, S. & Géry, J.,

1970. — Biométrie d'un hybride naturel de poissons

Cyprinidae, la brème de Buggenhagen. Ann. Hydro-

biol., 1 : 43-54.

Petit. C., 1976. — Génétique des populations et

évolution. In: Petit & Zuckerkandl (1976) : 1-195.

Petit, C. & Zuckerkandl, E„ 1976. — Evolution.

Paris, Hermann : i-xviii + 1-278.

Philipp, D. P., Childers, W. F. & Whitt, G. S.,

1979. — Evolution of patterns of differential gene

expression: a comparison of the temporal and

spatial patterns of isozyme locus expression in two

closely related fish species (northern largemouth

bass, Micropterus salmoides salmoides, and small-

mouth bass, Micropterus dolomieui). J. exp. Zool.,

210: 473-487.

Philipp, D. P.. Parker, H. R. & Whitt. G. S„ 1983. —

Evolution of gene régulation: isozymic analysis of

patterns of gene expression during hybrid fish

development. In: M. C. Rattazzi, J. G. Scanda-

lios & G. S. Whitt (eds.), Isozymes: current topics

in biological and medical research, New York, Liss,

10: 193-237.

Pi a, J., 1920. — Zur Kritik des Gattungsbegriffes.

Ver h. zool.-bot. Ges. Wien, 70: 145-152.

Plateaux, L„ 1981. — Critère mixiologique et notion

de genre. Bull. Soc. zool. Fr., 106 : 513-520.

Platnick, N. I.. 1977. — Paraphyletic and polyphy-

letic groups. Syst. Zool., 26: 195-200.

Powell, J. R., 1978. — The founder-flush spéciation

theory: an experimental approach. Evolution, 32:

465-474.

Poynton, J. C., 1964. — The Amphibia of Southern

Africa: a faunal study. Ann. Natal Mus., 17: 1-334.

— 1976. — Classification and the Arthroleptinae

(Amphibia). Rev. Zool. afr., 90: 215-220.

Prager, E. M., Brush, A. H., Nolan, R. A., Naka-

nishi, M. & Wilson, A. C., 1974. — Slow évolution

of transferrin and albumin in birds according to

micro-complement fixation analysis. J. mol. Evol., 3:

243-262.

Prager, E. M. & Wilson, A. C., 1975. — Slow

evolutionary loss of the potential for interspecific

hybridization in birds: a manifestation of slow

regulatory évolution. Proc. nat. Acad. Sci. U.S.A.,

72: 200-204.

Prevosti, A., Ocana, J. & Alonso, G., 1975. —

Distances between populations of Drosophila subobs-

cura, based on chromosome arrangement frequen-

cies. Theor. appt. Genet., 45: 231-241.

Rafe, R. A. & Kaufman, T. C., 1983. — Embryos.

genes, and évolution. The developmental-genetic basis

of evolutionary change. New York & London,

Macmillan: i-xiii + 1-395.

Ralin. D. B., 1970. — Genetic compatibility and a

phylogeny of the temperate North American hylid

fauna. Thesis, Austin: i-v + 1-117.

Rensch. B., 1959. — Evolution above the species level.

London, Methuen: i-xvii + 1-419.

Ripley, S. D., 1945. — Suggested terms for the

interprétation of spéciation phenomena. J. Wasli.

Acad. Sci., 35: 337-341.

Rogers, J. S., 1972. — Measures of genetic similarity

and genetic distance. Stud. Genet., 7: 145-153.

Rose, M. R. & Doolittle, W. F., 1983. — Molecular

biological mechanisms of spéciation. Science, 220:

157-162.

Rosen, D. E. & Bailey. R. M., 1963. — The poeciliid

fishes (Cyprinodontiformes), their structure, zoo-

Source : MNHN , Paris

120

ALAIN DUBOIS

geography and systematics. Bull. Amer. Mus. nal.

Hist., 126: 1-176.

Ross, H. H., 1975. — Biological systematics. Reading.

Addison-Wesley: 1-345.

Rostand, J., 1933. — Sur l'hybridation du crapaud et

de la grenouille rousse. C. r. Soc. Biol.. 112 : 1042-

1043.

Rowell, A. J., 1970. — The contribution of numer-

ical taxonomy to the genus concept. In: E. L.

Yochelson (ed.), Proc. N. Amer, paleontol. Conven¬

tion, Chicago 1969 , Lawrence, I, pt. C: 264-293.

Ruffié, J., 1982. — Traité du vivant. Paris, Fayard : 1-

795.

Savage, J. M.. 1958. — The iguanid lizard généra

Urosaurus and Uta. with remarks on related groups.

Zoologie a , 43: 41-54.

Schaefer, C. W., 1976. — The reality of the higher

taxonomie categories. Z. zool. Syst. Evol.-Forsch..

14: 1-10.

Schenck, H. G., 1937. — The subgenus as a taxono¬

mie category. Proc. geol. Soc. Amer.. 1936 (1937):

398.

Schultz, R. J., 1977. — Evolution and ecology of

unisexual fishes. Evol. Biol.. 10: 277-331.

Scott, N. J., Jr.. ed., 1982. — Herpetological commu-

nities. Wildlife Research Report. Washington, Uni¬

ted States Department of the Interior, Fish and

Wildlife Service. 13: i-v + 1-239.

Scudder, G. G. E., 1974. — Species concepts and

spéciation. Canad. J. Zool.. 52: 1121-1134.

Sene, F. M. & Carson, H. L„ 1977. — Genetic

variation in Hawaiian Drosophila. IV. Close allozy-

mic similarity between D. silvestris and D. hetero-

neura from the Island of Hawaii. Genetics. 86: 187-

198.

Serov, O. L., Zakijan, S. M. & Kulichkov, V. A.,

1978 a. — Allelic expression in intergeneric fox

hybrids ( Alopex lagopus x Vulpes vulpes). III.

Régulation of the expression of the parental alleles

at the Gpd locus linked to the X chromosome.

Biochem. Genet ., 16: 145-157.

— 1978 b. — Analysis of mechanisms regulating the

expression of parental alleles at the Gpd locus in

mule érythrocytes. Biochem. Genet.. 16: 379-386.

Sherff, E. E., 1940. — The délimitations of généra

from the conservative point of view. Bull. Torrey

bot. Club, 67: 375-380.

Short, L. L., 1969. — Taxonomie aspects of avian

hybridization. The Auk, 86: 84-105.

Sibley, C. G.. 1957. — The evolutionary and taxono¬

mie significance of sexual dimorphism and hybridi¬

zation in birds. Condor. 59: 166-191.

— 1962. — The comparative morphology of protein

molécules as data for classification. Syst. Zool.. 11:

108-118.

— 1965. — Molecular systematics: new techniques

applied to old problems. L 'Oiseau et la Revue fr.

Ornith., 35, n° spécial: 112-124.

Sibley, C. G. & Ahlquist, J. E., 1982. — The rela-

tionships of the yellow-breasted chat (Icteria virens)

and the alleged slowdown in the rate of macromole-

cular évolution in birds. PostiUa, 187: 1-19.

Simpson, G. G., 1937. — Supra-specific variation in

nature and in classification. Amer. Nal., 71: 236-

267.

— 1943. — Criteria for généra, species, and subspe-

cies in zoology and paleozoology. Annals New York

Acad. Sci.. 44: 145-178.

- 1944. — Tempo and mode in évolution. New York,

Columbia Univ. Press: i-xviii + 1-237.

— 1945. — The principles of classification and a

classification of mammals. Bull. Amer. Mus. nat.

Hist.. 85: i-xvi + 1-350.

— 1951. — The species concept. Evolution. 5: 285-

298.

— 1953. — The major features of évolution. New

York, Columbia Univ. Press: i-xx + 1-434.

— 1961. — Principles of animal taxonomy. New

York, Columbia Univ. Press: i-xii + 1-247.

— 1962. — Primate taxonomy and recent studies of

nonhuman primates. Annals New York Acad. Sci.,

102: 497-514.

Smith, H. M. & List, J. C., 1951. — The occurrence

of digital pads as an anomaly in the bullfrog.

Turtox News, 29: 118-119.

Sneath, P. H. A., 1962. — The construction of taxo¬

nomie groups. In: Microbial classification. Symposia

Soc. Gen. Microbiol., Cambridge Univ. Press, 12:

289-332.

Sneath, P. H. A. & Sokal, R. R., 1973. — Numerical

taxonomy. San Francisco, Freeman: i-xv + 1-573.

Stanley, S. M., 1979. — Macroevolution. Pattern and

process. San Francisco, Freeman: i-xi + 1-332.

Stebbins, G. L.. 1958. — The inviability, weakness,

and sterility of interspecific hybrids. Adv. Genet.. 9:

147-215.

Stebbins, G. L. & Ayala, F. J„ 1981. Is a new

evolutionary synthesis necessary? Science, 213: 967-

971.

Stoyan, D„ Stoyan, H. & Fiksel, T., 1983.

Modelling the évolution of the number of généra in

animal groups (Yule’s problem revisited). Biometr.

J., 25: 443-451.

Sturtevant, A. H., 1979. — A new species closely

resembling Drosophila melanogaster. Psyché, 26:

153-155.

— 1920. — Genetic studies on Drosophila simulons.

I. Introduction. Hybrids with Drosophila melanogas¬

ter. Genetics, 5: 488-500.

Subtelny, S., 1974. — Nucleocytoplasmic interaction

in development of Amphibian hybrids. Int. Rev.

Cytol., 39: 35-88.

Suchetet, A., 1897. — Des hybrides à l’état sauvage.

Règne animal. Tome premier. Classe des Oiseaux.

Paris, Baillière : i-clii + 1-1002.

Takamura, T. & Watanabe, T. K., 1980. — Further

Source : MNHN , Paris

THE GENUS IN ZOOLOGY

121

studies on the Léthal hybrid rescue (Lhr) gene of

Drosophila simulons. Japon J. Genet., 55: 405-408.

Templeton, A. R., 1979. — The unit of sélection in

Drosophila mercatorum. II. Genetic révolution and

the origin of coadapted genomes in parthenogenetic

strains. Genetics, 92: 1265-1282.

— 1980 a. — The theory of spéciation via the founder

principle. Genetics, 94: 1011-1038.

— 1980 b. — Modes of spéciation and inferences

based on genetic distances. Evolution, 34: 719-729.

— 1981. — Mechanisms of spéciation — a popula¬

tion genetic approach. Annual Rev. Eco!. Syst., 12:

23-48.

— 1982. — Genetic architectures of spéciation. In:

Barigozzi (1982): 105-121.

Thaler, L., 1982. — La génétique des populations

dans la théorie moderne de l’évolution. Revue de

Synthèse, 103-104 : 381-403.

Thorn, R., 1969. — Les salamandres d'Europe. d’Asie

et d'Afrique-du-Nord. Paris, Lechevalier : i-iv + 1-

376, 11 cartes h.-t., pl. I-XVI.

Tortonese, E., 1962. — Généra and subgenera in

classification and nomenclature. In: Verli. XI. Inter¬

nat. Kongr. Entom., Wien, 17-25 August I960, Wien,

Band 3: 339-343.

Turesson, G., 1922. — The genotypical response of

the plant species to the habitat. Hereditas, 3: 211-

350.

Twitty, V. C., 1961. — Second-generation hybrids of

the species of Taricha. Proc. natn. Acad. Sci. U.S.A.,

47: 1461-1486.

— 1964. — Fertility of Taricha species-hybrids and

viability of their offspring. Proc. natn. Acad. Sci.

U.S.A., 51: 156-161.

— 1966. — Of scientists and salamanders. San Fran¬

cisco, Freeman: 1-178.

Van Gelder, R. G., 1977. — Mammalian hybrids and

generic limits. Amer. Mus. Novit., 2635: 1-25.

— 1978. — A review of canid classification. Amer.

Mus. Novit., 2646: 1-10.

Van Valen, L., 1973. — Are categories in different

phyla comparable? Taxon, 22: 333-373.

Viot, C.-R., 1985. — Étude de génétique d'Oiseaux

subantarctiques et antarctiques utilisant l’électropho¬

rèse des protéines. Thesis, Montpellier : i-viii + 1-

209.

Voous, K. H., 1964. — Het genusbegrip in de zoologie

in théorie en praktijk. Vakblad voor Biologen, 44:

139-149.

Wake, D. B., 1982 a. — Functional and develop-

mental constraints and opportunities in the évolu¬

tion of feeding Systems in Urodeles. In: D. Mossa-

kowski & G. Roth (eds.), Environmental adaptation

and évolution. Stuttgart & New York, G. Fischer:

51-66.

— 1982 b. — Functional and evolutionary morpho-

logy. Persp. Biol. Med., 25: 603-620.

Wake, D. B., Roth, G. & Wake, M. H., 1983. — On

the problem of stasis in organismal évolution. J.

theor. Biol., 101: 211-224.

Wake, M. H., 1980. — The reproductive biology of

Nectophrynoides malcolmi (Amphibia: Bufonidae),

with comments on the évolution of reproductive

modes in the genus Nectophrynoides. Copeia, 1980:

193-209.

Walker, P. M. B., 1979. — Genes and non-coding

DNA sequences. Ciba Found. Symp., 66: 25-45.

Wallace, B., 1978. — The adaptation of Drosophila

virilis to life on an artificial crab. Amer. Nat., 112:

971-973.

Wallace, D. G., King, M.-C. & Wilson. A. C.,

1973. — Albumin différences among ranid frogs:

taxonomie and phylogenetic implications. Syst. Zool.,

22: 1-13.

Wallace, D. G., Maxson, L. R. & Wilson, A. C.,’

1971. — Albumin évolution in frogs: a test of the

evolutionary clock hypothesis. Proc. natn. Acad.

Sci. U.S.A., 68: 3127-3129.

Wasserman, M., 1968. — Recombination induced

chromosomal heterosis. Genetics, 58: 125-139.

Watanabe, T. K., 1979. — A gene that rescues the

léthal hybrids between Drosophila melanogaster and

D. simulons. Japan J. Genet., 54: 325-331.

Watson, G. F. & Martin, A. A., 1968. — Postmating

isolation in the Hyla ewingi complex (Anura:

Hylidae). Evolution, 22: 664-666.

White, M. J. D., 1978. — Modes of spéciation. San

Francisco, Freeman: i-viii + 1-455.

— 1982. — Rectangularity, spéciation, and chromo¬

some architecture. In: Barigozzi (1982): 75-103.

Whitt, G. S., 1985. — Personal communication (letter

of 17 September 1985).

Whitt, G. S., Childers, W. F. & Cho, P. L., 1973. —

Allelic expression at enzyme loci in an intertribal

hybrid sunfish. J. Hered.. 64: 54-61.

Whitt, G. S., Philipp, D. P. & Childers, W. F..

1977. — Aberrant gene expression during the

development of hybrid sunfishes (Perciformes, Tele-

ostei). Différentiation, 9: 97-109.

Wiley, E. O., 1979. — An annotated Linnean hier-

archy, with comments on natural taxa and compe-

ting Systems. Syst. Zool., 28: 308-337.

— 1981. — Phylogenetics. The theory and practice of

phylogenetic systematics. New York, Wiley: i-xv +

1-439.

Williams, C. B.. 1951. — A note on the relative sizes

of généra in the classification of animais and plants.

Proc. Linn. Soc. Lond., 162: 171-178.

Williamson, F. S. L., 1957. — Hybrids of the Anna

and Allen hummingbirds. Condor, 59: 118-123.

Wilson, A. C., 1975. — Evolutionary importance of

gene régulation. Stadler Symp., 7: 117-134.

Wilson, A. C., Bush, G. L., Case, S. M. & King, M.-C.,

1975. — Social structuring of mammalian popula-

Source : MNHN , Paris

122

ALAIN DUBOIS

tions and rate of chromosomal évolution. Proc,

nain. Acad. Sci. U.S.A.. 72: 5061-5065.

Wilson, A. G, Carlson, S. S. & White, T. J.. 1977. —

Biochemical évolution. Annual Rev. Biochem.. 46:

573-639.

Wilson, A. C., Kunkel, J. G. & Wyles, J. S., 1984. —

Morphological distance: an encounter between two

perspectives in evolutionary biology. Evolution , 38:

1156-1159.

Wilson, A. C., Maxson, L. R. & Sarich. V. M..

1974. — Two types of molecular évolution. Evi¬

dence from studies of interspecific hybridization.

Proc. nain. Acad. Sci. U.S.A.. 71: 2843-2847.

Wilson, A. C., Sarich, V. M. & Maxson, L.R.,

1974. — The importance of gene rearrangement in

évolution: evidence from studies on rates of chro¬

mosomal, protein, and anatomical évolution. Proc,

nain. Acad. Sci. U.S.A., 71: 3028-3030.

Wilson, A. C., White, T. J.. Carlson, S. S. & Cherry,

L. M.. 1977. — Molecular évolution and cytogene-

tic évolution. In: R. S. Sparkes, D. E. Comings &

C. F. Fox (eds.), Human cyiogenelics. ICN-UCLA

Symposia on Molecular and Cellular Biology. New

York, San Francisco & London, Academie Press,

vol. 7: 375-393.

Wright, S., 1931. — Evolution in Mendelian popula¬

tions. Gene tics, 16: 97-159.

— 1932. — The rôle of mutation, inbreeding, cross-

breeding and sélection in évolution. Proc. Sixth

internat. Congr. Genetics: 356-366.

Wyles, J. S.. Kunkel, J. G. & Wilson, A. C., 1983. —

Birds, behavior, and anatomical évolution. Proc,

natn. Acad. Sci. U.S.A., 80: 4394-4397.

Xavier, F., 1979. — Une espèce nouvelle de Necto-

phrynoides (Anoure. Bufonidé) des Monts Nimba,

N. liberiensis n. sp. I. Description de l'espèce. Bull.

Soc. zool. Fr., 103 : 431-441.

Zepp, H. D„ Conover, J. H., Hirschhorn, K. &

Hodes, H. L., 1971. — Human-mosquito somatic

cell hybrids induced by ultraviolet-inactivated Sen-

dai virus. Nature New Biology, 229: 119-121.

Zink, R. M., 1982. — Patterns of génie and morpho¬

logie variation among sparrows in the généra Zono-

trichia, Melospiza, Junco, and Passerella. The Auk

99: 632-649.

Zuckerkandl, E., 1964. — Controller gene diseases:

the operon model as applied to B-thalassemia,

familial fêtai hemoglobinemia and the normal switch

from the production of fêtai hemoglobin to that of

adult hemoglobin. J. moi Biol., 8: 128-147.

— 1976 a. — Programs of gene action and progres¬

sive évolution. In: M. Goodman & R. E. Tashian

(eds.), Molecular anthropology, Genes and proteins in

the evolutionary ascent of the primates, New York,

Plénum Press: 387-447.

— 1976 b. — L’évolution moléculaire. In: Petit &

Zuckerkandl (1976) : 197-271.

— 1979. — Personal communication ( verbo, Novem-

ber 1979).

— 1980. — Personal communication (letter of 17

February 1980).

Zuckerkandl, E. & Pauling, L., 1962. — Molecular

disease, évolution, and génie heterogeneity. In: M.

Kasha & B. Pullman (eds.). Horizons in biochem-

istry, New York & London, Academie Press: 189-

225.

Source : MNHN , Paris

Date de distribution : 18 novembre 1988.

Dépôt légal : novembre 1988.

IMPRIMER IE NATIONALE

7 564 010 T 57

Source : MNHN , Paris

Source : MNHN, Paris

DERNIERS TITRES PARUS

RECENTLY PUBLISHED MEMOIRS

Série A (Zoologie) :

Tome 139 : Alain Dubois, 1988 — Le genre en zoologie : essai de systématique théorique. 132 pp. (ISBN

2-85653-152-0) 150 FF.

Tome 138 : Christian Erard, 1987 — Écologie et comportement des gobe-mouches (Aves : Muscicapinae,

Platysteirinae, Monarchinae) du Nord-Est du Gabon. 256 pp. (ISBN 2-85653-142-3) 250 FF.

Tome 137 : Jacques Forest, 1987 — Résultats des campagnes Musorstom, volume 3. Les Polychelidae ou

« Pagures symétriques » (Crustacea Coenobitoidea). 274 pp. (ISBN 2-85653-141-5) 300 FF.

Tome 136 : Claude Monniot et Françoise Monniot, 1987 — Les Ascidies de Polynésie française. 156 pp.

(ISBN 2-85653-139-3) 200 FF.

Tome 135 : Michel Lemire, 1986 — Contribution à l’étude des fosses nasales des Sauriens. Anatomie

fonctionnelle de la glande « à sels » des Lézards déserticoles. 119 pp. (ISBN 2-85653-137-7) 200 FF.

Série B (Botanique) :

Tome 31 : Odile Poncy, 1985 — Le genre Inga (Légumineuses, Mimosoideae) en Guyane française.

Systématique, Morphologie des formes juvéniles. Ecologie. 124 pp. (ISBN 2-85653-135-0) 210 FF.

Tome 30 : Lucile Allorge, 1985 — Monographie des Apocynacées — Tabernaemontanoïdées américaines.

216 pp. (ISBN 2-85653-132-6) 280 FF.

Tome 29 : Monique Keddam-Malplanche, 1985 — Le Pollen et les stomates des Gardéniées (Rubiacées) du

Gabon. Morphologie et tendances évolutives. 109 pp. (ISBN 2-85653-13l-C-8) 220 FF.

Tome 28 : Marie-France Roquebert, 1981 — Analyse des phénomènes pariétaux au cours de la conidiogenèse

chez quelques Champignons microscopiques. 79 pp. (ISBN 2-85653-116-4) 130 FF.

Tome 27 : Auteurs multiples, 1981 — Entretiens du Muséum : Palynologie et Climats, 16-18 octobre 1979.

302 pp. (ISBN 2-85653-110-5) 300 FF.

Série C (Sciences de la Terre) :

Tome 53 : Donald E. Russell, Jean-Pierre Santoro and Denise Sigogneau-Russell, 1988 — Teeth

Revisited : Proceedings of the Vllth International Symposium on Dental Morphology. 462 pp. (ISBN

2-85653-148-2) 625 FF.

Tome 10 : Jacques Roger, 1962 (Réimpression/Reprint 1988) — Les Époques de la nature. Éditions critique.

344 pp. (ISBN 2-85653-160-1) 100 FF.

Tome 52 : Donald E. Russell and Zhai Ren-Jie, 1987 — The Paleogene of Asia: mammals and stratigraphy.

490 pp. (ISBN 2-85653-140-7) 300 FF.

Tome 51 : Mireille Gayet, 1986 — Ramallichtys Gayet du Cénomanien inférieur marin de Ramallah (Judée),

une introduction aux relations phylogénétiques des Ostariophysi. 81 pp. (ISBN 2-85653-138-5) 130 FF.

Tome 50 : Jean-Marie Rouchy, 1983 — La genèse des évaporites messiniennes de Méditerranée. 267 pp.

(ISBN 2-85653-122-9) 290 FF.

Prix hors taxe, valides jusqu’à juin 1989. Frais de port en sus. Vente en France : TVA 4%.

Prices in French Francs are valid until June 1989. Postage not included.

Source : MNHN, Paris