<<

Contributions to , 66 (1) 3-41 (1996)

SPB Academic Publishing bv, Amsterdam

Conceptual issues in phylogeny, , and nomenclature

Alexandr P. Rasnitsyn

Paleontological Institute, Russian Academy ofSciences, Profsoyuznaya Street 123, J17647 Moscow, Russia

Keywords: Phylogeny, taxonomy, phenetics, , phylistics, principles of nomenclature, type

concept, paleoentomology, Xyelidae (Vespida)

Abstract On compare les trois approches taxonomiques principales

développées jusqu’à présent, à savoir la phénétique, la cladis-

tique et la phylistique (= systématique évolutionnaire). Ce Phylogenetic hypotheses are designed and tested (usually in dernier terme s’applique à une approche qui essaie de manière implicit form) on the basis ofa set ofpresumptions, that is, of à les traits fondamentaux de la taxonomic statements explicite représenter describing a certain order of things in nature. These traditionnelle en de leur et particulier son usage preuves ayant statements are to be accepted as such, no matter whatever source en même temps dans la similitude et dans les relations de evidence for them exists, but only in the absence ofreasonably parenté des taxons en question. L’approche phylistique pré- sound evidence pleading against them. A set ofthe most current sente certains avantages dans la recherche de réponses aux phylogenetic presumptions is discussed, and a factual example problèmes fondamentaux de la taxonomie. ofa practical realization of the approach is presented. L’auteur considère la nomenclature A is made of the three main taxonomic que taxonomique comparison ap- s’appuie, au fond, sur plusieurs principes de base. Neuf tels proaches hitherto developed, viz., phenetics, cladistics, and principes sont explicitement formulés, dont six sont taxono- phylistics (= evolutionary ). The latter term denotes les trois autres étant taxono- that tries the features miquement indépendants, an approach explicitly to represent basic miquement dépendants (donc compatibles seulement avec of traditional taxonomy and particularly its use of evidence certains concepts taxonomiques). La pratique taxonomique derived from both a similarity and the relatedness ofthe taxa courante montre taxon qu’un n’est ni une classe, ni un individu, involved. The phylistic approach has certain advantages in the

mais bien un continuum (notion combinant des traits distinctifs answering ofthe basic aims oftaxonomy. de la classe et de l’individu) de subtaxons, continuum séparé Taxonomic nomenclature is found to rely ultimately on a d’autres continua lacune. On considère le par une que concept few basic principles. Nine of these principles are formulated du type est le meilleur instrument à notre disposition s’il s’agit explicitly: six ofthem are taxonomically independent, and three de travailler avec le concept taxon-continuum. are taxonomically dependent, that is, they are only compatible with taxonomic particular concepts. Judging from current taxonomic practice, a taxon is neither a class nor an individual, but continuum notion a (a combining some features ofboth the Introduction class and the of subtaxa that is delimited individual) by a gap

it from other such continua. separating The type concept is scientists do not take much notice of found be the best Frequently, to available tool to operate within the concept the of their How- of the taxon-continuum. methodological grounds work.

ever, when the standard means and methods of the

work time after time fail to yield satisfactory re-

Résumé sults, one is tempted to raise a kind of scientific

revolution (Kuhn, 1970), i.e., to abandon those tra-

and methods with Les hypothèses phylogénétiques sont élaborées et vérifiées ditional means along the under-

(normalement de manière implicite) sur la base d’une série de lying methodology and to replace them with some- d’autres d’assertions décrivant présomptions- en mots, un different. scientific thing Revolutions, even ones, certain ordre de choses dans la nature. Ces assertions doivent in rarely pass bloodless, that they usually result in être des à acceptées comme telles, indépendamment preuves rejecting certain approaches that are in fact work- leur mais seulement en l’absence de soli- appui, preuves assez

able, albeit in a somewhat different while des contre leur validité. On discute une série de présomptions domain, phylogénétiques des plus usuelles, et on présente un exemple the data obtained by means of these approaches concret de réalisation en pratique de cette manière d’aborder le still be and may meaningful worthy of being reas- problème. - 4 A.P. Rasnitsyn Phylogeny, taxonomy, and nomenclature

sessed rather than rejected. That is why it seems Phylogeny

preferable to examine the methodological back-

of Each in ground a scientific approach prior to abandoning step cognition, while knowledge is ac-

in the and its meth- it, hope to recognize improve quired, in my opinion proceeds through four suc-

and thus the cessive known odological deficiency to safeguard ap- stages. First, we start seeking for

and the results of This is the elements in proach its application. an essentially unknown (i.e., not yet of objective the present publication. understood) pattern. On the basis of previous expe-

and/or the rience results of a preliminary confron-

issues forced Two separate me to approach the tation with the pattern, we then sort these known

elements methodological problems of descriptive biology. as either important (i.e., relevant to the

One the crisis of that In was long-lasting taxonomy goal set), or as unimportant. this context, the

resulted in the continual between "tradi- known conflicts elements of an essentially unknown pattern

and the - the tional" rival are considered when that taxonomy approaches important we suppose our

the Because of it phenetic and, later, cladistic ones. of knowledge permits us to discern (i.e., is corre-

of of an intuitive a traditional lated or otherwise reflects in feeling superiority with, one way or an-

I in taxonomy, was involved this controversy, al- other) important features of the pattern. In turn, the without much though success (Ponomarenko & important features are those that have a more pro-

found influence Rasnitsyn, 1971; Rasnitsyn, 1972, 1983a, 1987a, on, or are correlated with, or

& 1988b, 1992a; Rasnitsyn Dlussky, 1988). Re- otherwise permit us to appreciate and foresee, the sults of my attempts to develop my arguments are structure of the pattern and its behaviour under

below in the presented chapters devoted to various circumstances. For example, when an en-

and phylogeny taxonomy. tomologist wishes to instate an in his opinion en- As the is different regards nomenclature, cause tirely new group as a separate taxon, he must also

in that it is rooted in deal with my paleontological experi- the classificatory aspects of his proposal

ence. Despite the long history of taxonomy and no- (I mean only the internal classification, not the

menclature, I believe the basic principles of no- general taxonomie position of that taxon). In this

menclature are only partly understood and formu- treatment he will disregard features, how signifi-

lated as such. Some of these are sim- that discriminate principles so cant per se, males and females, or

and self-evident that their existence and adults and larvae it ple appli- (except when comes to subdi-

cation is the in perceptible only in peripheral areas of tax- vide taxon question and to classify each sub-

When one has to difficult and onomy. classify objects group separately eventually to compare the re- deficient in taxonomically important characters (or sults with other data). In addition he will tend to

one is often forced break the even lacking them), to give preference to morphological similarities over rules of in nomenclature. This is common practice such features as colour patterns. in the of immature paleontology, study parasitic The second stage of cognition begins here. One

of where worms, or asexual stages of fungi, a stu- creates a hypothesis concerning the particular dent cannot normal escape violating taxonomie similarity between the patterns under examination practice and breaking the rules of nomenclature. and a class of patterns that previously had been An of this little experience sort gives satisfaction studied already but seemed to be dissimilar to the until the student to discern begins more deep- patterns in question. The supposition ought to be seated reasons why he has not followed the it should gener- predictive, i.e., propose the existence of a

rules. the ally accepted The reason why is that un- particular mechanism that is responsible for the derlying principles of nomenclature are incompat- similarity, or it must clarify why there is also a ible with the deficient material that taxonomically similarity in many other characters. This subse- nevertheless must be classified. This controversy quently allows us to hypothesize that the similarity renders these basic thus principles discernible, per- can predict the behaviour of the pattern in question their identification and mitting study (Rasnitsyn, in various environments and conditions. To con- 1986; 1992b). tinue with the above example, the taxonomist will Contributions to Zoology, 66 (1) - 1996 5

that the in possibly pose similarity morphological tion are impossible, we can never be certain of any

characters could indicate a higher (say, generic) of our observations and considerations concerning

level of the overall while the similar col- similarity, the nature of things. If so, how can we live and

our pattern concerns a lower one (say, the species work in a world which offers nothing definite to us

In other the taxonomist would be- level). words, hy- and thus seems so unreliable? This is possible

that the inferred tentative classification I skilful pothesize cause suppose we are enough at making

is more natural, e.g., less burdened with homo- correct decisions with insufficient information

the that plasies, than alternative one gives more (when information is sufficient and everything

the colour be almost weight to pattern. seems to plain sailing, we act automati-

The two above-mentioned stages of cognition cally and do not admit that we are making a deci-

depend much on the personal experience and intui- sion). We do make mistakes, but they are (or,

tion of the investigator, whose formal cognitive rather, were until now) not bad nor numerous

nature is still obscure, and I shall not discuss these enough to lead mankindto extinction.

in here. stages detail The explanation of how we make such decisions

The is third stage to test the proposed assump- has existed for a long time (Rasnitsyn, 1988b;

tions. To reach that end we should propose the Rasnitsyn & Dlussky, 1988). It has been found in

longest and most diverse chain of implications, in courts of justice where information is often insuffi-

order to have a better opportunity to compare them cient but where the "correct" decision is vital,

with the results of observations and for our experi- given that the time needed seeking a necessary

details ments (for see below). "generalization" has been sufficient. This "gener-

The fourth and last of is make is step cognition to a alization" termed a presumption. This means that

decision the of each concerning availability pro- for a particular class of problems there exists a de-

whether be cision class of that is posed hypothesis: any supposition can (or a decisions) considered to

retained for further testing, or must be rejected. be generally the most likely or otherwise the safest.

The is be Examples and a discussion follow. decision consequently to accepted no

The third step of cognition has been well studied matter what tangible evidence for it exists, but only

by Karl Popper (1959), who clarified the fact that in the absence of reasonably sound tangible evi-

there is no dence it. Otherwise way to verify a hypothesis definitively. against the alternative decision

it is Moreover, not possible to falsify a hypothesis is to be accepted. The well-known presumption of

completely, since this would mean the verification innocence is based on the observation that for the

of alternative viz. that an hypothesis, the falsifying health ofour society it is more dangerous to punish

result arises neither by chance nor is due to ne- the innocent than to leave a criminal unpunished.

external glected influences (circumstances) (cf. That is why a suspect is not generally consideredto

if that be unless the accusation is Farris, 1983). Indeed, you claim all swans a criminal proved, no

white and encounter black to matter are yet a swan, prior how strong the suspicion may be.

claim should the abandoning your you reject possi- We can use the above conclusions to understand

bilities that black is (1) your swan not a swan (e.g., how we study phylogeny. Phylogeny is commonly

it modified skilful is a or a as the of a of convergently goose, interpreted history particular group

mystification, or something like that), and (2) your living organisms, their history being described in

black black swan is not naturally (e.g., artificially terms of a diminishing genealogical relatedness

stained black). 1 believe that these possibilities are between its subgroups. In addition, it is commonly

always a matterof setting a likelihood and not just agreed that evolution proceeds too slow for imme- of a It is for this diate that giving yes-or-no reply. reason that perception, so phylogeny is perceptible the fourth of from its stage cognition appears necessary only ultimateresults, not from the process when we consider competing hypotheses and as- itself. These results consist of traits of form and

relevant sess falsifying evidence in order to reach a function of organisms, including morphological, decision. physiological, and behavioural characters, those

Since both an ultimate verification and falsifica- concerning attitude of organism to biotic and 6 A.P. Rasnitsyn - Phylogeny, taxonomy, and nomenclature

in and close abiotic factors, their distribution space time, changes. The problem certainly deserves ex- and the like. amination, which, however, lies beyond the scope

in of article. The first and central presumption phyloge- the present netic analysis is the postulation of "knowability" of When a group is meticulously studied, both with

claims: shouldbe to well as phylogeny. It any similarity con- regard paleontological as contemporary

evolu- sidered as inherited unless and until the reverse is material, it is evident that extensive parallel reasonably proved. Hennig ( 1966) calls this state- tion resulting in an abundance of homoplasies, is ment "the auxiliary principle", but it is far from often characteristic of both higher taxonomie lev- being auxiliary. The central dictum of phylo- els and species. Vertebratesmay serve as an exam-

is the claim is for show of tran- genetic analysis that phylogeny gen- ple, they numerous cases parallel erally knowable. Indeed, "if homoplasy [indepen- sition from the level of organization (grade) of

is the charac- that of dently gained similarity] universal, to amphibians (i.e., independent ac- ters imply nothing about the genealogy" (Farris, quiring by groups of the characters generally char-

1983: 14). acteristic of amphibians) and further to the level of

This is not evident at once. A man does engender organization of reptiles and (or ). In

and not an under but the last decades this old theme has been under a man, ape, any conditions, two water can be equally obtained by either burning close examination by Russian students of evolution hydrogen or by oxidizing lipids inside organisms. (Tatarinov, 1976, 1987; Shishkin, 1987; Vorobye-

The difference between the two cases evidently va, 1992). Under the old term aristogenesis (Os- lies in the different level of complexity of chemi- born, 1934), this pattern is discussed by Rautian cals vs. living beings. That is why living beings (1988).

in close their while Parallel evolution in related is change dependence on history, closely groups

in of changes chemicals are essentially governed by commonly described as a manifestation particu- circumstances and rules their his- lar tendencies intrinsic to simple (not by evolutionary a group. tory). The tendencies can be understood as an epiphe-

It of the Living beings are of different complexity. has nomenon underlying synapomorphy been suggested that in procaryotes the amount of (Saether, 1979, 1986; Sluys, 1989; a term of simi-

lar albeit identical is independently acquired similarity is so vast that not meaning key apomorphy, traditional methods are often of little help in dis- proposed by Iordansky, 1977, 1979). The notion of cerning procaryote phylogeny (Zavarzin, 1987). underlying synapomorphy implies an apomorphy

Observations ofthis sort quite evidently impose re- (any change in structure or function) that has been

the domain of the and strictions on phylogenetic analysis. gained by the common ancestor of group

There is another restriction. that is in members of the Generally speaking, present some or many the of of lower order takes because the divergence taxa a place group, probably genetic capacity to in is a less remote past than that of higher taxa. Hence develop the feature not expressed in every mem- it leave clearer of how such di- ber of the For the should evidence a group (cf. Sluys, 1989). instance, vergence has proceeded. Nevertheless, we often acquisition of the resilient longitudinal "string" feel that the genealogical relationships of species, (the chord and its successor, the backbone) that

of especially those in larger genera, are much less permits to accumulate the mechanic energy the manifest and more difficult to sort out than rela- body bending, has been the common base for vari- tionships of, say, families or subfamilies (cf. ous chordate to obtain independently the

I be- Saether, 1986). am aware of the explanation of -like general appearance whenever they this phenomenon, viz., the hypothesis that homo- come a strong swimmer (whales, , ich-

is much Another is the plasy a more common event than students thyosaurs, tadpoles, etc.). case cos- often think. This of particularly concerns groups of tal breathing, being a remote cause the inde- closely related and thus highly similar forms whose pendent development of homoiothermy, as dis- almost identical organization enables them to react cussed by Iordansky (1977, 1979). Unlike these in a similar fashion to similar environmental examples, however, we are rarely aware of the - Contributions to Zoology, 66 (1) 1996 7

concrete causes that are responsible for the re- feet cladograms by filling them with unresolved

of feature peated acquisition a particular by a par- (polytomous as opposed to dichotomous) ,

ticular of believe but otherwise considerations group organisms. Nevertheless, I the above change but

that a real common cause, albeit yet unknown (but little in the actual practice of phylogenetic study.

for underlies see Sluys, 1989, suggestions) the ma- Further phylogenetic presumptions can be ar-

of of jority cases massive parallel evolution. This ranged in two gross classes according to whether

results in the methods cause a tendency to develop apo- the respective phylogenetic are directed

character - hence morphic the term "underlying toward the analysis of history of either groups or

2 characters synapomorphy". .

Wide distribution of parallel evolution does not

permit us to think of taxa as simple products of di-

of vergence their common ancestor. Parallel adap- Group analysis

tation of closely related forms to similar niches

could contribute as if it is com- well, especially The main method of group analysis is the paleon-

bined with reduction of the "archaic which has been redefined a diversity" tological one, recently as

(term proposed by Mamkaev, 1968). The process stratophenetics (Gingerich, 1979, 1990). It is based

of evolution whole be described parallel as a can as on the presumption that among two apparently

taxon the accumulation of similar related the earlier maturation, i.e., closely groups one appearing in

changes combined with the elimination of unbal- the fossil record should be considered ancestral

in the of anced (inadaptive sense Rasnitsyn, 1987b) unless and until sound contradictory evidence is

intermediate This leads back to groups. approach presented. We shall call it the paleontological pre- the old and well-known of hypothesis Darwin sumption for groups. The paleontological method

(1859) that gaps between extant taxa are in part is not universal in application because the fossil

due extinction of in- (probably to a large extent) to record is known to be incomplete. However, this

termediate due forms. This maturation often takes quite incompleteness is neither vast nor to chance,

time. That a long is why higher taxa at later stages but rather regular. There is a special paleonto-

of their evolution are often more clear-cut and logical discipline called taphonomy that deals with

seem to have a more discernible easily phylogeny the patterns of burial and fossilization of organ-

as lower themselves and it has compared to taxa, or even to at isms, already gathered a considerable

an earlier in their amount of stage history. data relevant to those patterns (see e.g.,

The involved in explaining the & In it is al- hypotheses pat- Briggs Crowther, 1990). many cases,

tern of extinction of known what of intermediates are proposed ready kinds organisms have a good

elsewhere (Rasnitsyn, 1987b). Here it is important chance of becoming buried and fossilized and

note that to the pattern described above seems to stand a good chance to be excavated later and

less difficultiesfor the which do cause phylogenetic study of not, and what kinds of fossil environ-

than the eucaryotes in case of procaryotes. Any- ments stand a better chance to contain fossils than how, we are still able to construe a other in phylogenetic ones. Additionally, deposits vary greatly tree, although an assessment of its general validity their degree of information. Using this informa- is restricted. We can never be quite certain about tion, it is often possible to infer with reasonable

and about its it the tree particularly details, so that certainty whether a particular case of absence of a becomes of pointless to speak an ancestral species group from a particular section of the fossil record of for 1 In is due say, example a family . quite a few cases, to a true non-existence or not. a thorough and sophisticated research has failed to For instance, fossil are found mostly in identify even the ordinal position of the common lacustrine and, especially in the Paleozoic, in shal-

of And this ancestor a class. may be true even if the low marine deposits (other sources of paleoento- real ancestor has already been collected and de- mological information became important only in

albeit scribed, not yet recognized among other fos- relatively late stages of the group's history). That is sils In (Vorobyeva, 1992). short, parallelisms af- why most fossil insects seem to be either flying or 8 A.P. - and nomenclature Rasnitsyn Phytogeny, taxonomy,

aquatic/semiaquatic, shore-dwelling forms. The phylogenetic and geographic, of barriers imposed

various bristle-tailed insects (Thysanura s.l.) are wingless, by physical agents (primarily connected

terrestrial, and not especially common along sea with continental drift, but also of a climatic nature,

and lake shores. The Thysanura entered the fossil etc.) on a continuousancestral distribution. The ap-

record later than the Pterygota [that is, at West- proach ignores unique events and relies solely on

vs. Namurian time; older de- the that in various that phalian findings pattern repeats groups

scribed Shear al. and al. the by et (1984) Labandeira et frequented areas under investigation. The con-

based that (1988) are on fossils are too fragmentary gruent area cladograms (the cladograms with the

be with Absence to identified sufficient certainty]. / names replaced by the names of their

of bristletails from the Lower populated areas) are interpreted there as the evi-

(Mississipian) fossil record cannot be seriously dence of the respective area-dependent genealogy

taken evidence that could as they not be the of all groups concerned.

On the other A Pterygote ancestors. hand, hymenop- cladistic biogeography seems to fit ideally the

terous insects are and their good fliers, absence objectives of the phylogenetic analysis of groups.

from the record in In pre- should figure any seeking a particular repeating pattern of area

the that until analysis using above presumption; is, cladograms, we then should simply hypothesize

we have evidence, we should avoid for each involved. strong hypo- isomorphic cladograms group

theses that for The is however. of require searching a hymenopteran matter not so easy, Dispersion

ancestor that before the is known de- among groups died out organisms to be a common event and,

3 Triassic . pending on an existing system of barriers and on

There exist other methods that the of does phylogenetic rely dispersal potential a group (which vary

first the on group analysis, of all, biogeographical but not in a disorderly way), is not necessarily so

method. the aim of the method is chaotic be sorted Indeed, to restore as to easily out as producing

the in of its history of a group terms distribution. unique area cladograms. Regular dispersal events

their distributional of Groups are evolving on routes. may produce repeating patterns area cladograms

Thus the distributional history of a group could that could be easily but deceivingly interpreted in correlate in with its a way genealogy, so that phylo- terms of cladistic biogeography. Unless research

benefit from There geny may biogeography. are could demonstrate a subordinate role for disper-

however. in the of problems, sion origin the area cladogram patterns, I The classic, dispersalist biogeography considers would refrain from using the biogeographical distribution of a group as a result of its individual method in restoring genealogy.

dispersal through various barriers. Major factors of There is another serious reason to be cautious as

distribution thus are supposed to be (1) the geo- regards the prospects of the biogeographical

of the the graphic structure space, (2) dispersal po- method as a phylogenetic tool. Both classic (dis-

tentialof the and chance. Similar distri- and cladistic little group, (3) persalist) biogeography pay at-

could bution patterns of two groups imply their tention to extinction events. And yet they are not similar history, including genealogy, depending on only very common but often consistent enough to

it similarity of the three above factors. However, concern the entire biota. The area cladograms af-

seems that all fected not particularly likely three factors by extinction can produce a pattern easily

will be should essentially similar, so we suppose but wrongly interprétable by the cladistic method.

that the often chance. For there pattern similarity is due to example, are numerous cladistic recon-

Therefore, under the classic concept of biogeogra- structions of austral ("Gondwanian") distribution

the method to be of phy, biogeographical appears patterns (for a review, see Humphries & Parenti,

doubtful value for phylogenetic analyses. 1986). I suspect that most of them are of no use,

In contrast, the cladistic concept of biogeogra- because the fossil record, if it exists, quite often phy here (the treatment based mostly on the sum- shows that at some time in the past presumed aus-

& considers the tral occurred in northern and be- mary by Humphries Parenti, 1986) groups continents of animals distribution and as a result, both came extinct there (for details, see Eskov, 1987, Contributions to Zoology, 66 (1) - 1996 9

the 1992). Austral paleontology is often less complete tropical rain forest as we know it. Unlike tropi-

boreal and there finds of aus- cal from the later Mexican than the one, yet are assemblages Paleogene

Dominican well those from the tral fossils known for groups that now seem to have and ambers, as as

boreal distribution and a strictly (same references). Miocene of Sicily northern Apennines (Zhe-

Similarly, extinction could be indeed responsible rikhin, 1993), the Burmese assemblage is com- for the amphitropical (bipolar) distribution of pletely devoid of the isopteran family Termitidae,

in- and of social and social . As many taxa. These amphitropical distributions, bees, higher was

these stead of being a heritage of a hypothetical Pacifica demonstratedby Zherikhin (1978, 1993), are

the three that in the continent (Humphries & Parenti, 1986), are more animal groups play key roles

rain and likely to be remnants of past world-wide distri- functioning of tropical forests, possible butions that have become separated by the tropical functional analogues are not known nowadays, nor biota that originated in the equatorial zone did they occur in the past. probably during the Eocene (Razumovsky, 1971; It follows that the repeated pattern of area

Zherikhin, 1978, 1993). Unfortunately, tropical cladograms is still meaningless unless it is substan-

is than the tiated fossil record. That paleontology more poorly developed by a reasonably complete austral one. Nevertheless, it is of importance that is why I believe that we still have much to do be-

workable there are no sound indications of equatorial Late- fore it will be possible to formulate a set

Cretaceous non-marine biota comparable with the of biogeographical presumptions for the analysis tropical ones of the present (Meyen, 1987). The of taxonomie groups.

rel- only known relevant and comparatively rich fossil The same holds true for other apparently assemblage (ca. 200 specimens buried in evant methods, e.g., the parasitological approach, the Late earlier amber which the of the of the host or Paleogene suggests use cladogram

under has model for that of the from Burma), now my investigation, re- as a specialized parasite, or

which looks vealed no characteristically tropical forms. In- vice versa, depending on cladogram

the fossils there more Brooks & stead, among already identified, promising (see McLennan, 1991, are several groups very common in assemblages for an overview).

from the Late-Cretaceous fossil resins ofNorth Si- beria and Canada, e.g., the hymenopteran genera

Serphites Brues, 1937 and Palaeomymar Meunier, Character analysis

latter is extant 1901 (Rasnitsyn, 1980; the

is used either the and at present at least predominantly extratropical Character analysis to polarize and series bipolar). Additionally, there are fossil groups transformation (morphocline), or to falsify

of of that have survived with either a northern (snake- suppositions a unique origin similarity , order Raphidiida), or a bipolar (Phrys-

- There sonotus Scudder, 1885, a millipede) extratropical Polarizing transformation series. are sev-

transformation series: distribution (Condé, 1954). Phryssonotus is now eral methods for polarizing known from the Mediterranean, South Africa (1) The paleontological method is similar to that

method for (Cape Province), South Australia (New South described above as a group analysis

Wales), Uruguay, Chili, and Cuba (the Cuban (stratophenetics). These two applications are inde-

however. record is equivocal zonally, because the lowlands pendent of one another, Homoplasies are of the island are tropical and the more elevated paleontologically shown to be not only fairly com- territories in the but also often to have similar direction in subtropical). In contrast, Late mon, a

The of Eocene, as the Baltic amber specimens witness, various contemporary groups. cases mam-

Phryssonotus extended far into the northern extra- malization of the theromorph reptiles (Tatarinov,

and of the tropical zone. 1976, 1985) angiospermization gymno-

There is evidence that the some even stronger spermous plants (Krassilov, 1991) are good exam-

That is in cases we can the past Burmese forest that has yielded the fossil- ples. why many use pre- iferous amber differed from the that series fundamentally sumption a transformation should be - 10 A. P. Rasnitsyn Phytogeny, taxonomy, and nomenclature

rial for and the results polarized according to the succession of the re- phylogenetic inferences,

spective character states in the fossil record, un- should be evaluated according to their own merits,

until sound data after Bonde's less and contrary are presented. not an a priori principle. (1984) ex-

This presumption will be particularly useful in ample ofhuman paedomorphosis is instructive: we

that are well in the infer that the human skull is groups represented paleonto- paedomorphic solely

with because the of the logical record an inconclusive morphology we know topology anthropoid

(either by being simple and thus poor in characters, cladogram from other sources (primarily from

with taxonomie of various character or teeming homoplasies). patterns states, par-

(2) Another and very similar method is the ticularly from the structure of DNA, see below).

It is based on the ourselves ontogenetical one. biogenetic pre- Otherwise, basing on ontogeny per se,

sumption: Transformation series shouldbe polar- we could infer that the human skull features are

ized in agreement with the ontogenetic succession symplesiomorphic with respect to the gibbons and

character unless and until while those of could of the respective states, monkeys, the great apes be

there decide otherwise. In are serious reasons to synapomorphic.

fact, this presumption just rewords the famous Von (3) The next important method and presumption

Baer's (1828) "law" of the differentiationthrough are the method and presumption of analogy: If a

that dem the ontogeny, claiming "Aus Allgemeinsten transformation series is polarized in a group,

der results be considered valid Formverhältnisse bildet sich das weniger should as for another

allgemeine und so fort bis endlich das Speciellste unless and until sound data are , group, contrary

auftritt" (p. 224; Von Baer's italics). Or, in transla- adduced. Indeed, if we are reasonably sure that the

tion: "From the most general relationships [that is, wing is a modified leg, we can legitimately be-

beginning from synapomorphies of the most inclu- lieve it to be an example of what prevailed during

- less the of and well. sive taxa AR], develop the special ones evolution wings as

the and similar [synapomorphies of less inclusive taxa], so However, to pose a hypothesis concerning

until insect is less evident because of differ- on, the most special relationships [synapo- wings great

morphies of the terminal taxa] appear". ences between the groups and the important dis-

of their it is There is a widespread beliefthat the "biogenetic similarity wings. Nevertheless, not

law" has extraordinary importance (as a "direct ar- completely absurd. Kukalova-Peck (1978) argues

without cf. the gument") in (e.g., Nelson, 1978; (not reason: Rasnitsyn, 1981) that

Bonde, 1984). I do not believe this. Ontogenesis is insect leg and the insect wing are both modified

of biramous flexible, any of its stages can be possibly modified parts a ancestral leg, but Bitsch (1994)

secondarily or it may get lost due to paedomor- disagrees. Likewise, we know that most mamma-

lian evolved in the direction of phosis, pre-adult adaptation (caenogenesis), or in lineages increasing

the process of rationalization (streamlining) of the body size ("law of phylogenetic growth", Cope,

Therefore should infer this direction of developmental path. As a result, the biogenetic 1904). we

method, while highly effective in deciphering the evolution everywhere in mammals where no sound

evidence is found. holds for of some groups, can often be mis- contrary The same true

leading. It has no a priori advantage in comparison many other so-called empirical laws of evolution.

The with other methods, and does not work as a hard (4) well-known polarizing method using

rule but only as a presumption. Indeed, the onto- functional efficiency can be formalized as based

genetic succession of developmental stages is a re- on the presumption of functional efficiency: Of two

sult of direct observation. holds for character the The same true states one corresponding to a more

the succession should be considered stratigraphie of fossils and their efficient adaptation apomor- traits, as well as for the pattern of character states phic, until and unless there is reliable contradict-

of over the system organisms. Each of these three ing evidence. For example, viviparity is generally

sets of data leaves some traces and thus reflects, in considered as apomorphic in comparison to ovipar-

the one way or another, process of descent of living ity because it secures the safety of progeny better beings. Each of them deserves to be used as mate- than the latter (it permits viviparous animalsto sur- - Contributions to Zoology, 66 (I) 1996 11

number of Keratin known in all vive despite a lower mean offspring). ous; see below). scales are

(5) Subordinate to the above presumption is that amniotes, hair only in mammals (and in some re-

of device is lated the of complexity: a more complex more reptiles: Tatarinov, 1976; homology

of and thus ifit wasteful resources originates only the pterosaurian hairs described by Sharov, 1971,

is more efficient. That is why complex structure is still to be proved). Similarly, feathers are known

in birds in should be considered as apomorphic in relation to only (and again some related reptiles, cf.

unless and have a more simple one, until we strong Kurzanov, 1987). These cases are typically sub-

of reason to decide otherwise. Indeed, complex struc- jected to application the presumption,

tures often tend to become simpler and more effi- which claims: A character statefound only within a

cient in the course of evolution the should be considered (compare e.g., group apomorphic in respect

outside primitive therian paw and the 's ). of that distributed both within and the

Such structures com- unless and until evidence (6) secondarily simplified group, strong contrary

rudiments other of their lost The of this the monly display or signs appears. application presumption to

complexity, indicating one more polarizing pre- above transformation series should result in the hy-

the of The that both hairs and feathers sumption, viz., presumption vestiges: potheses are apomor-

character state showing any signs of being modi- phic and scales plesiomorphic.

another the lat- fied from state is apomorphic over

ter, unless and until there is strong contrary evi- Similarity analysis. - The similarity analysis was

dence (e.g., until it is found that what we took for characterized above as pursuing discrimination be-

the is in fact inherited and If vestige an Anlage). tween similarity homoplasy. we

(7) The last and most popular polarizing method consider our first presumption ("similarity is inher-

is It is ited unless and the the outgroup comparison. a probabilistic test until...") as valid, similarity

for of conformity a cladogram with our basic analysis is not necessary until the "unless and un-

til" when phylogenetic model (evolution is generally diver- applies. It is only we encounter similari-

and abundant gent, homoplasies are not so as to ties that are distributed in a contradictory way that

make similarity say nothing about genealogy: cf. we ought to analyse them. The distribution is con-

the first of presumption knowability of phylogeny, tradictory when the similarities cannot be all con-

p. 6). In this model, a character state acquired by a sidered as inherited: if we accept the similarity of

more remote ancestor has a chance to persist in a whales to other mammals in their homoiothermy as

wider range of descendants than that which ap- inherited, we must treat their swimming adapta-

later in the When tions of and vice peared same group. we can see a as acquired independently fishes,

character state both within represented a group and versa.

beyond its limits, and another one which is found There are two main methods to detect homo-

within that should infer only group, we (unless plasy: (a) The presumption of parsimony, which is there that the first are strong contrary arguments) in turn one more implication of our first presump-

character state is plesiomorphic, and the second tion. Indeed, if similarity is usually inherited and

This is when apomorphic. because, doing so, we only less commonly appears as a homoplasy, then

could infer only one evolutionary change that took "homoplasy ought not to be postulated beyond

at the of subclade of the in and the place origin a group necessity" (Farris, 1983: 8), most likely

question. Otherwise, hypothesizing the first char- cladogram is that implicating the least number of

unless and until there acter state apomorphic and the second plesiomor- homoplasies, are sound rea-

should least identical evolu- another phic, we imply at two sons indicating cladogram as the most

- 4 tionary changes one within the group, and an- likely , But why again "unless and until?" Why is

other its limits. If that the inher- and beyond we agree parsimony only a presumption not a hard rule,

ited is than It similarity more likely (more common) or even a law as many students believe? is so the all other independently acquired one, then, because characters are not equal. If we consider things being equal, we should conclude that the similarity between mimicking butterflies and count

second is less less similar in their colour hypothesis likely (or parsimoni- each element pattern sepa- 12 A.P. - and nomenclature Rasnitsyn Phytogeny, taxonomy,

rately, and then we count similarities of each mim- lasting disagreement. Unless stated otherwise, the

icking species to its relative, I am not sure that the discussion below is based on my previous relevant

will be in favour of 1988a& result always pleading true papers (Rasnitsyn, 1969, 1980, 1983b, c,

relatedness. Similarities do vary in how reliably 1990; and references therein).

kind The they indicate relatedness, depending on what hymenopterous insects (order Vespida; the

of character state is involved. We generally assess reasons for using this name instead of Hymenop-

similarity in the type of symmetry or in the gross tera are given in Rasnitsyn, 1982, 1988a, 1989)

of basic first structure organs as more important phylo- appearin the fossil record in the second halfof

genetically than, say, a colour pattern or details of the Triassic period. These Triassic species have

the surface structure. However, exceptions exist been found in both Australia (one species) and

even here, for all mammals are similar in absence Central Asia (two dozen species). These all belong

from their of blue pigments hairs, not to mention to the family Xyelidae, as indicated by their char-

the importance of differences in pigment sets for acteristic wing venation and antennal structure

and procaryote plant taxonomy. Nevertheless, a (Fig. 1). The family survives in the northern Hemi-

colour is considered relict 5 and pattern generally as only of sphere as a group comprising genera

subsidiary phylogenetic importance. about three dozen species, but not until the mid-

The above observations indicate that the phylo- Cretaceous did it become one of the leading hy-

of one and the same of - in genetic importance sort menopterous groups at least more temperate

similarity varies greatly depending on what taxa climates.

are and what characters involved. Other later than compared are hymenopterous groups appeared

Therefore, I believe, we should operate by use of the Xyelidae and are not recorded before the Early

weighted characters when discriminating homo- . This makes it possible to apply the

Hard rules be describe plasy. seem to impossible to paleontological presumption for groups and to pro-

the at least for the That is situation, present. why pose a hypothesis that Xyelidae could be ancestral

we have to rely on one more presumption: (b) the to all other hymenopterous insects. This hypothesis

of In fact there in with ofthe other observations presumption weighted similarity. is agreement most

are quite a number of such presumptions, and some that show this family to be the most plesiomorphic

of been within the order all its with them have already discussed elsewhere as regards characters,

(Rasnitsyn, 1988b; Rasnitsyn & Dlussky, 1988). the only exception discussed below. Indeed, the

However, many more efforts are needed to clarify xyelid wing venation is the richest, having the

the problem in detail, so I can only safely propose highest number (the most complete set) of veins

In the here a generalized definition: case of con- and cells, and resembles that ofother insects most:

similarities flicting (when not all similarities can be the fore wing with its RS diverging into two

as those which should be accepted inherited) con- branches in a way similar to that of various primi-

sidered inherited be as are known to more reliable tive insects and with supernumerary SC veinlets in

in other cases (especially in closely related some fossil species, and the hind wing with free Cu groups■), unless and until strong contra-arguments base. The same holds true as regards the body mor-

appear. phology and supposed groundplan bionomic fea-

tures. Particularly relevant is the suggestion that

their inside development gymnosperm staminate

Examples cones was possibly retained from the earliest

holometabolic insects (Rasnitsyn, 1980). Plesio-

How we can use the above presumptions is better morphies are cladistically unimportant, however,

with illustrated examples. I think that the early so that we have to scrutinize the nature of the

evolution of the hymenopterous insects, and par- xyelid antenna, the only character not easily to be

the of the the 5 ticularly problem phylogenetic position disregarded as of the family.

of the family Xyelidae, seems appropriate, The xyelid antenna (Figs. 2, 3a) is evidently

since is the of in it a difficult case and subject long- apomorphic respect to a normal insect antenna, Contributions to Zoology, 66 (1) - 1996 13

Fig. 1. Triassoxyela foveolata Rasnitsyn, 1964 (Vespida, Fig. 2. Anthoxyela turgensis Rasnitsyn, 1990 (Xyelidae); Lower

Xyelidae); Middle or Late Triassic of Kirghizia (after Ras- Cretaceous ofTransbaikalia (after Rasnitsyn, 1990).

nitsyn, 1969).

with several fused basiflagellar segments into a

long and thick compound "third" antennal seg-

The latter is mounted with thread ment. apically a

of normal (thin and short) flagellar segments. The

apomorphic nature of such an antenna is supported

the of number of the above by application a pre-

sumptions. The paleontological presumption for

is relevant because the closest and oldest groups

non-hymenopteran group (the extinct order Palaeo-

manteida = Miomoptera, Fig. 4) had normal anten-

nae. The respective presumption for characters is

also applicable, because all known pre-Triassic in-

sect antennae do not display xyelid apomorphy.

The same holds true for the biogenetic presump-

because the earlier tion, at developmental stage

(Fig. 3a) the antennal segmentation is more homo-

nomous than in the adult insect. Equally important

is the of because the simi- presumption analogy, Fig. 3. Antennae ofvarious , horntails, and flies (unless inflated 3rd larly (though not evidently compound) stated otherwise, of imago): a, Xyela julii Brébisson, 1818

antennal pupa); b, Blasticotoma Malaise, segment of flies (Fig. 3f) is certainly (Xyelidae, filiceti pacifica

1931 (Blasticotomidae); c, Arge ustulata (Linnaeus, 1758) (Ar- apomorphic. The outgroup presumption suggests gidae); d, Caenolyda reticulata (Linnaeus, 1767) (Pamphi- the for same conclusion, the xyelid antenna type is liidae); e, Xiphydria camelus (Linnaeus, 1767) (Xiphydriidae); found only within the Vespida while the normal f, primitive Glutops semiformis Nagatomi & Saigusa, 1970 is both within and outside this order. one common (Rhagionidae) (a—e, after Rasnitsyn, 1968; f, after Nagatomi,

= Some of the I—XII of = above presumptions seem to yield 1982). segments antenna, “III” compound “third” antennal the segment. same result when also applied intraordinally. It

be in seems to especially true outgroup compari-

because the son, normal antenna type is wide- fied xyeloid antennae. The least modified of them spread within the order, and this is not the case are the antennae of the tenthredinoid families

the This with xyeloid one. inference is probably Blasticotomidae and Argidae. Both of them have

however. the the "third" erroneous, Indeed, although among segment even hypertrophied as com-

insects the antenna in its full living xyeloid appear- pared to that of the Xyelidae themselves, but the

in that also the reduced ance occurs only family, antennae following segments are to a single one (in of insects some lower hymenopterous (sawflies some Blasticotomidae), or became completely lost

described and horntails) can be as variously modi- (Figs. 3b-c). P. - 14 A. Rasnitsyn Phylogeny, taxonomy, and nomenclature

long to sawflies and horntails (lower hymenopter-

insects Of ous comprising the suborder Siricina).

these, 120 display their antennal structure clearly

enough to render the identification of the antennal

be- type possible. In turn, 65 of these specimens

long to the Xyelidae and have the corresponding

antennal type. Among the remaining 55 fossils, 23 Fig. 4. Palaeomantina pentamera Rasnitsyn, 1977 (Palaeoman- also have xyeloid antennae, including those with tiscidae, order Palaeomanteida); Lower of Ural (after the "third" segment not so as in the Rasnitsyn, 1977). large Xyelidae but still clearly thicker and much longer than the

distal segments. The remaining 32 fossils have an-

Another source of evidence is provided by the tennae that are of the normal type or nearly so. living pamphiliid sawflies and siricoid horntails. Among the non-xyelid fossils displaying the

Pamphiliidae (except some species of Pamphilius) xyeloid antenna type there are, firstly, four genera

it have the "third" segment subdivided but is com- that share several important putative synapomor-

paratively small, several times longer than the phies with living Tenthredinoidea (Fig. 5). These

fourth and of subequal width (Fig. 3d). Less genera have the fore wing with 2r-rs meeting RS

known, though no less important, is the antenna of distally of2r-m, and 1 mcu cell enlarged so that the the horntail first abscissa of RS is short lost. primitive genusBrachyxiphus Philippi, or got Most genera

1871 (Siricoidea, Xiphydriidae) that has a rather also have the SC stock fused, albeit only distally, thick and "third" in with that its fore when forms long segment (subequal length R, so branch, present, to three more distal segments combined), and addi- an intercostal cross-vein (Pseudoxyela Rasnitsyn, tionally subdivided into 4 primary segments in 1968 is exceptional in retaining the free SC and

malesof B. grandis Philippi, 1871 (specimens were thus probably forms the of the rest of

studied in the collectionofthe Smithsonian Institu- the tenthredinoid ). The four genera also tion, Washington, D.C.). Most if not all other share synapomorphies with the Tenthredinoidea as

also have this somewhat their antennal which is thick Xiphydriidae segment en- regards flagellum,

larged (Fig. 3e), so this character state is possibly a with a reduced number of segments (supposing part of the family ground plan. that the antenna is secondarily multisegmented in

The above distribution of the compound 3rd Diprionidae and some Pterygophoridae). I con- antennal segment can be explained by its inde- sider Xyelotoma Rasnitsyn, 1968, Xyelocerus

in five different pendent development groups. That Rasnitsyn, 1968, Dahurotoma Rasnitsyn, 1990 and number doubles, however, when we consider data Pseudoxyela as forming an extinct family reported by Van Achterberg & Van Aartsen (1986: Xyelotomidae. No are recorded tables 1, and that the for the which its ancestral 2, fig. 366) suppose compos- family, implies position ite of structure the segment was never lost (cf. Fig. in respect of the other Tenthredinoidea.

extinct 11). The next five genera with the xyeloid

The of cases an independent acquisition of the type of antenna (Fig. 6) show the important syn- compound segment become even more numerous apomorphy of the superfamily Pamphilioidea, viz., if we take fossils into consideration. The fossil col- the reduced claw-like ovipositor of a characteristic lections of the Paleontological Institute, Russian general form and position in the body. The general of Academy Sciences, Moscow, are particularly appearance and wing venation are also similar to demonstrative in this that of the fos- respect. These collections living pamphilioids. Among these contain an overwhelmingly large number of Ju- sils, Xyelyda Rasnitsyn, 1968, Prolyda Rasnitsyn, rassic and Lower in- and Cretaceous hymenopterous 1968, Mesolyda Rasnitsyn, 1963 are assigned

collected all sects over the world, though mostly to the extinct family Xyelydidae which lacks auta- hailing from Kazakhstan, Siberia, and Mongolia. pomorphies and hence is supposed to be ancestral

Among them, 515 specimens of the above age be- to the other Pamphilioidea. Praesirex Rasnitsyn, Contributions to Zoology, 66 (I) - 1996 15

Fig. 5. Representatives of the sawfly family Xyelotomidae: a,

Xyelocerus admirandus Rasnitsyn, 1968; b, Xyelotoma ni-

heteroclita gricornis Rasnitsyn, 1968; c, Pseudoxyela

Rasnitsyn, 1968 (all from the Upper Jurassic of South

Kazakhstan); d, Dahurotoma robusta Rasnitsyn, 1990, Lower

Cretaceous of after Transbaikalia (a—b Rasnitsyn, 1969; c, Fig. 6. Representatives ofthe sawfly families Xyelydidae (a—c) based onRasnitsyn, 1969, fig. 82, and specimen PIN 2997/650;

and Praesiricidae (d—e): a, Xyelyda excellens Rasnitsyn, 1968; d, after Rasnitsyn, 1990).

b, Prolyda karatavica Rasnitsyn, 1968; c, Mesolyda depressa

Rasnitsyn, 1968 (all from the Upper Jurassic ofSouth Kazakh-

Praesirex hirtus stan); d, Rasnitsyn, 1968; e, Turgidontes mag- 1968 and 1990 are Turgidontes Rasnitsyn, nus Rasnitsyn, 1990 (both from the Lower Cretaceous ofTrans- with the synapomorphic living family Megalo- baikalia) (a, b, d, after Rasnitsyn, 1983b; c, original, based on

after 1990). dontidae in having wings without SC. They are as- holotype; e, Rasnitsyn, signed to the extinct family Praesiricidae which was probably ancestral to the Megalodontidae.

One more group marked by xyeloid antennae 1988c) has been found to be erroneous.

several which The last with is comprises genera (Fig. 7) are now group a xyeloid antenna rather assigned to the extinct family Sepulcidae (Rasnit- loosely constructed. I include here Protosirex Ras-

The in 1969 which syn, 1993). Sepulcidae are placed the super- nitsyn, (Fig. 8) seems to belong to the

because of family their synapomorphy horntails (Siricoidea) because of its general ap- with the living Cephidae (Rasnitsyn, 1988c). Both pearance and some siricoid plesiomorphies (e.g., families have their fore with the wings costal space the presence of a free SC). Alternatively, it may moderately narrow and band-like (secondarily represent a group ancestral to both siricoids and

with the SC fused with with few other very narrow or basally lost), cephoids. Along a Jurassic fossils,

R (except sometimes only basally), and lacking the Protosirex represents the extinct family Gigasiri-

SC branch the costal The cidae. Other siricoids with apical crossing space. a xyeloid antenna are

lack the characteristic venational Sepulcidae syna- Syntexyela media Rasnitsyn, 1968, and S. inversa

and 1968 pomorphies of the living Cephidae were prob- Rasnitsyn, (Fig. 9a-b). They surely belong to

the their ably cephid ancestors. As to the Cephidae the Siricoidea, as witnessed by characteristic themselves, the observation of their most ancient synapomorphy (a strong, needle-like ovipositor).

the (Early Cretaceous) genus Mesocephus Rasnitsyn, In particular they represent predominantly

1968, as having xyeloid antennae (Rasnitsyn, family Anaxyelidae (with a single living 16 A.P. Rasnitsyn - Phylogeny, taxonomy, and nomenclature

Fig. 8. The primitive horntail Protosirex xyelopterus Rasnitsyn,

1969 (Gigasiricidae) from the Upper Jurassic of South

Kazakhastan (after Rasnitsyn, 1969).

Fig. 7. Representatives ofthe sawfly family Sepulcidae: a, Xye- lula hybrida Rasnitsyn, 1969; b, Onokhoius aculeatus Rasnit- syn, 1993; c, Parapamphilius confusus Rasnitsyn, 1968; d,

Pamparaphilius mongolensis Rasnitsyn, 1993; e, Micram- philius minutus Rasnitsyn, 1993; from the Upper Jurassic of

South Kazakhstan (a, c) and Lower Cretaceous of Mongolia (b) and Transbaikalia (d) (all after Rasnitsyn, 1993).

Fig. 9. Representatives of the horntail family Anaxyelidae: a,

Syntexyela media (Rasnitsyn, 1963); b, S. inversa Rasnitsyn,

1968; c, Anaxyela gracilis Martynov, 1925; d, Brachysyntexis as confirmed the re- species), by synapomorphic micrura Rasnitsyn, 1969; all from the Upper Jurassic of South duction the SC and the 2r-m cross-vein Kazakhstan after based on of stock of (a, c, d, Rasnitsyn, 1969; b, original, in the fore wing and of lr-m in the hind wing. It is holotype).

that these have the 3rd important two species seg- ment slightly modified in comparison to the

and modified in different it insects. In here the Xyelidae, two ways: menopterous fact, we encounter seems simply diminished but seemingly entire in S. conflicting putative synapomorphies. The above-

while it but subdivided mentioned inversa,, is large clearly into genera are supposedly synapomorphic primary segments in S. media. Closely related with the Xyelidae in their antennal structure, while

that anaxyelid genera display antennae are either their various other characters seem to be apomor- normal or with the 3rd segment only slightly en- phies shared with some other sawflies and horn-

for larged (Fig. 9c-d; details see Rasnitsyn, 1969). tails. These hypotheses cannot all be correct, i.e.,

the above similarities cannot all be synapomor-

take all data Now we can the above into consid- phies. Thereforeour next task is to decide which of eration for assessing the meaning of the xyeloid the conflicting putative synapomorphies can be

with ho- antenna regard to the early evolution of hy- falsified (reassessed as symplesiomorphies or - Contributions to Zoology, 66 (1) 1996 17

arrive the and consis- the moplasies) to at most likely able as ancestral to more complete xyelid and different tent cladogram. Four hypotheses can be pamphiliid venation, because again many pre-

here: in the confirm that in the considered (1) similarity xyeloid an- sumptions a more

tenna is a synapomorphy while the alternative sets complete wing venation is plesiomorphic. similarities of of the listed genera are sym- Unlike the next three hypotheses, I cannot illus-

but the alternative the with plesiomorphies, (2) as above, trate one just discussed a relevant clado-

similarities being homoplasies, (3) the alternative gram. This is because this hypothesis infers incom-

sets of similarities are synapomorphies (the above patible polarization presumptions, as can be seen

of groups genera are monophyletic with the re- from the case of the ovipositor (see above).

spective non-xyelid superfamilies and families) The second hypothesis (2), suggesting the homo-

while the been xyeloid antenna has acquired inde- plastic nature of the similarities between the genera

and but discussed pendently (as a homoplasy), (4) as above, and the respective non-xyelid groups,

the similarity lies in having the xyeloid antenna as presents fewer difficulties. However, it does imply

a hymenopterous symplesiomorphy. a vast amount of homoplasy, because the above-

The first that all the two hypotheses (1,2) mean listed putative synapomorphies with other mem-

genera in question are monophyletic with the bers of the respective superfamilies and families

would have Xyelidae (i.e., cladistically belong to the family), to be considered as having originated

while their other 6 This similarity to groups is either (a) independently (Fig. 10, italics ). seems un-

inherited from remote ancestor of inde- since these traits have been some or (b) likely accepted by gen-

when pendent origin. Hypothesis 1, applied to all erations of taxonomists as sound diagnostic char-

those similarities, seems the least likely because it acters for families and superfamilies. These char-

that the direction of evolution of the observed implies re- acters were as not especially prone to a

spective characters is to that which is in related and opposite gen- homoplastic development groups

as normal most common and therefore with the erally accepted (i.e., agree presumption of weighted

thus satisfying the relevant presumptions). This similarity.

seems possible in some cases but hardly so in other One more reason to consider either of the above

For it is difficult ones. instance, to believe that the hypotheses as unacceptable is the incidence of in-

specifically reduced and modified pamphilioid termediate antennal types described above for two

needle-like ovipositor or the strong siricoid one are species of Syntexyela. Indeed, the antennal struc-

in of the of S. plesiomorphic respect saw-like ovipositor ture media is easily interprétable as an inter-

of the mediate of Xyelidae and Tenthredinoidea. Indeed, sev- stage transformation of a normal an- eral presumptions suggest the reverse direction of tenna into a xyeloid one. In contrast, in S. inversa

evolution. In the majority of the non-hymenopter- (and in several other congeners) the basiflagellar

ous and is rather but winged insects, particularly so among the segment, which large seemingly en-

Paleozoic is either ones (including Palaeomanteida, an or- tire, a reduced xyeloid segment (which is der ancestral the in- under the supposedly to hymenopterous impossible above hypothesis), or other-

the unless in with that of sects; Rasnitsyn, 1980), ovipositor, re- wise has nothing common S. media.

duced, is a flat leaf- or saw-like structure In both the (Fig. 4). cases, two species cannot be close

Other (non-hymenopterous) examples of the nee- relatives, and all their similarity should be homo- dle-like ovipositor, e.g., in some members of the plastic. oldest hemipteran family Archescytinidae, or in So we now move on to the last two alternative

of crickets, are clearly an apomorphic nature. This hypotheses (3,4). Bothofthese claim that the simi- involves both the paleontological presumptions, larity of the above-listed genera (those possessing the and the outgroup one, presumption of analogy, the xyeloid antennae) to the respective non-xyelid which all the nature of the is support plesiomorphic groups synapomorphic. These hypotheses differ serrate ovipositor of the sawflies. The same judge- from each other in proposing that the similarity of ment of apomorphy holds true for the reduced possessing the xyeloid antennal structure is either

which is anaxyelid wing venation, hardly accept- homoplastic or symplesiomorphic in nature. The 18 A.P. Rasnitsyn - Phylogeny, taxonomy, and nomenclature

Fig. 10. Cladogram ofthe lower hymenopterous insects (suborder Siricina) constructed consistent with the hypothesis [(2) in the text]

that similarity in the xyeloid antenna is synapomorphic, and the alternative similarities are homoplastic. Figures under the lines of

descent refer to the sets ofsynapomorphies of respective clades (or terminal taxa), as follows (based on Rasnitsyn, 1988a, and, for

Pamphiliidae, on Van Achterberg & Van Aartsen, 1986; homoplasies are marked by italics):

1 — Synapomorphies of the order Vespida (wing venation or almost all length of mesothoracic venter;

specialized, flight dipterous, haplo-diploidy, etc.: see Ras- 11 — fore wing with SC short, appressed to R, lacking fore

nitsyn, 1988a, for details); branch;

antennal 12 with costal needle-like 2 — 3rd (basiflagellar) segment compound; — fore wing space narrow; ovipositor

3 — basiflagellar segment much enlarged (antenna xyeloid); (not known with full certainty);

short with 4 — antennal flagellum (up to 5 segments beyond basi- 13 — propleurae contiguous ventromesally; fore wing RS2 flagellar segment), rather thick; fore wing with 2r-rs con- lost; male gonostylus with gonomacula subapically; etc. as

and RS with 1st described in item necting pterostigmal apex far beyond 2r-m, Rasnitsyn (1988a: 121, 13);

abscissa of RS short, and with 1mcu cell large; 14 — lower tentorial bridge intervening between hypostomae;

5 — single flagellomere beyond composite one; pronotum short mandible of cuttingtype (with cuttingedge twisted into moving

with SC lost crossvein-like with M+Cu claw- medially; fore wing except fore plane); fore wing angulate; ovipositorsmall,

branch, and with RS2 lost (for further possible synapomorphies, like; larva exophytic, silk-protected, with appendages setiform,

1988a: 120, items 23, distant see Rasnitsyn, 3); antenna from eye;

with RS lost between larva with soft inner tooth of tarsal claws 6 — fore wing submarginal cells; 15 — tibial spurs tips;

boring stem, with caudal structures modified; reduced; hypopygial depression large;

7 — flagellum lost beyond composite segment;fore wing lacking 16 — membrane offore wing more or less folded apically; larva

2r-rs; male rotated 180°; larva with and genitalia eye antenna on Angiospermae;

distant, with antenna 1-segmented, with mandible lacking 17 — fore wing with SC lacking fore branch; femora widened;

incisive molar flange, and with suprapedal and subspiracular hypopygial depression medium-sized;

lobes for further with and merged; numerous synapomorphies, see 18 — vertex grooves deep subparallel anteriorly;

Rasnitsyn, 1988a: 120, items 5, 6, 8, 9); tarsal claws with two long bristles medially;

tentorium with 1st abscissa of RS with 1r-rs 8 — propleurae contiguous ventromesally; issuing 19 — fore wing very short,

anterior arms below tentorial with RS2 short and and with dark from bridge;fore wing very thin, conspicuous patches;

lost; with tarsal 20 — malar space differentiated setose depression;

9 — ovipositor small, claw-like modified; claws with acute lobe;

10 — both wings lacking SC, pseudosternum extending over all 21 — ovipositor sheath with large stylus [possibly secondarily - Contributions to Zoology, 66 (J) 1996 19

first that the in lost in with normal (3, Fig. 11) means genera question independently groups antennae.

have acquired the compound "third" segment inde- Both hypotheses seem more likely in comparison

while the second with the first of because pendently, hypothesis (4, Fig. 12) pair hypotheses, they pro-

this supposes that character has been uniquely ac- pose little more than numerous homoplasies in a

and later has character in in loss quired by a common ancestor, been single (either acquisition or of

lost in betulae (Linnaeus, and P. with distant Pamphilius 1758) festivus protected, appendagessetiform, antenna from eye;

Pesarini & Pesarini, 1984]; 33 — both wings lacking SC, and pseudosternum extendingover

22 — female fore wing with distal third of costal cell glabrous; all length of mesothoracic venter; antenna pectinate; fore wing

23 — female with inner tooth of hind claw shorter than apical with basal sections ofRS and M forming entire straight line, and

with behind larva tooth; A2+3 straight, running area aspera; feeding

with costal needle- herbaceous 24 — fore wing space narrow; ovipositor on angiosperm plants;

like; with and lower 34 — vertex grooves deep subparallelanteriorly;

25 — fore wing and hind wing with single cross-vein r-m each; tentorial bridge intervening between hypostomae; mandible of

with 26 — head capsule hypostomae contiguous and with cutting type (with cutting edge twisted into moving plane);

postgenae subcontiguous between oral cavity and occipital tarsal claws with two long bristles medially; female with inner

foramen; pronotum short medially, dorsally; propleurae neck- tooth of hind claw shorter than apical tooth; fore wing with

like claw-like; elongate; mesoscutum bearing transscutal suture and M+Cu angulate; ovipositorsmall, larva exophytic on

incipient adlateral lines; prepectus concealed under postero- Angiospermae, silk-protected, with appendages setiform,

lateral tibia with hind distant edge ofpronotum; fore (outer) spur antenna from eye; membrane offore wing more or less

with rudimentary (or lost); mesofurca fore arms long, fusedfor folded apically;

some distance;fore wing with SC lost except crossvein-like fore 35 — ovipositor sheath with large stylus [possibly secondarily

branch; male gonostylus with gonomacula subapical; larva lost in P. nemorum (Gmelin, 1790)];

dead feeding on angiosperm wood conditioned by symbiotic 36 — frons with pair ofswellings; ovipositor sheath with stylus

fungi, with abdominal sterna lacking prolegs and longitudinal glabrous; larva makes specialized leafroll;

and oblique sulci; larval salivary gland covered by common 37 — ovipositor sheath with stylus large, glabrous; — with ductus in section de- 38 tibia with hind envelope, quadrangular (for further fore (outer) spur rudimentary (or lost);

1988a; items 19 & other with tails, see Rasnitsyn, 122, 22, possible mesofurca fore arms long, fused for some distance; fore

listed there item with costal larva synapomorphies at 21—22); wing space narrow; xylophagous on living

27 with RS2 lost; — fore wing plants, modified as described by Rasnitsyn (1988a: 122, item

28 — pronotum short medially; fore wing with 2r-rs connecting 19);

and pterostigmalapex R-RSfar beyond 2r-m, with 1st abscissa 39 — fore wing with SC lost and with Cu straight within 1mcu

of RS short, and with 1mcu cell large); male genitalia rotated cell;

larva with and with mandible membranous fore with 180°; eye antenna distant, lacking 40 — mesonotum except laterally; wing

incisive molar also numerous de- costal in basal third flange; synapomorphies as space very narrow or half; ovipositor long

scribed in Rasnitsyn (1988a: 120, items 2—3, 3, 5); extending;

29 — fore wing lacking 2r-rs, hind wing lacking m-cu cross- 41 — metanotum with cenchri lost; fore wingwith 1st abscissa of

larva with antenna and with and RS short and with cell 1mcu with cross-vein 1r-rs vein; 1-segmented, suprapedal very large,

lobes with A2 and with subspiracular merged; long, straight, area aspera lost; abdomen

sulci tibial male larva in 30 — pseudosternal lost; preapical spurs lost; compressed; boring angiosperm plants;

with sternum 8 visible, excised larva needle-like; hardly strongly apically; 42 — ovipositor larva feeding on dead wood con-

with prolegs 2-segmented, with subanal appendages lost, and ditioned by symbiotic fungi, modified as explained in Rasnitsyn

with ductus salivary gland margined with 2 rows of glandular (1988a; 122, item 22); — cells; 43 fore wing and hind wing with single cross-vein r-m each;

31 — head capsule with hypostomae contiguous between oral 44 — head capsule with postgenae contiguous; mesoscutum

and cavity occipital foramen; tentorium issuing anterior arms hearing transscutal suture, prepectus concealed under postero-

from below tentorial ventro- lateral bridge; propleurae contiguous edge ofpronotum;

mesally; male gonostylus with gonomacula subapical; larva 45 — mesoscutum with transscutal suture bent cephalad

with abdominal sterna lacking prolegs and longitudinal and laterally; fore wing with basal abscissa ofRS subvertical;

oblique sulci, with subanal appendages segmented, shifted 46 — tentorial bridge narrow stripe-like, n-like bent; mesonotum

basally toward base of anal slit; larval salivary glandcovered with adlateral lines; fore wing with SC lost; 1st abdominal ter-

with ductus modified into mesosomal larva by common envelope, quadrangularin section; gum propodeum; parasitic; etc.,

claw-like 32 — ovipositor small, modified; larva exophytic, silk- as listed in Rasnitsyn (1988a: 123, item 28). 20 A. P. Rasnitsyn - Phylogeny, taxonomy, and nomenclature

Fig. 11. Ciadogram ofthe lower hymenopterous insects (suborder Siricina) constructed consistent with the hypothesis [(3) in the text] that similarity in the xyeloid antenna is homoplastic, and the alternative similarities are synapomorphic. Otherwise as in Fig. 10, with the following synapomorphies inferred:

with with subanal lost, and 1 — Same as in Fig. 10; prolegs 2-segmented, appendages

and with ductus with 2 of 2 — basiflagellar (3rd antennal) segment compound very salivary gland margined rows glandular large; cells;

with 2r-rs and RS with between oral 3 — fore wing connecting pterostigmal apex 12 — head capsule hypostomae contiguous far beyond 2r-m, with 1st abscissa of RS short, and with 1mcu cavity and occipital foramen; tentorium issuing anterior arms

ventro- cell large (for further possible synapomorphies see Rasnitsyn, from below tentorial bridge; propleurae contiguous

1988a: 120, item 23); mesally; male gonostylus with gonomacula subapical; larva

and with abdominal sterna and and 4 — basiflagellar (3rd antennal) segment compound very lacking prolegs longitudinal large; flagellum short (up to 5 segments), rather thick; oblique sulci, with subanal appendages segmented, shifted

short toward base of anal larval covered 5 — pronotum medially; fore wing with SC lost except basally slit; salivary gland by crossvein-like fore branch; common envelope, with ductus quadrangular in section;

6 — single flagellomere beyond composite one; fore wing with 13 — ovipositor small, claw-like modified; larva exophytic, silk-

RS between with distant from lost submarginal cells; larva boring fern stem, with protected, appendages setiform, antenna eye; caudal and structures modified; 14 — basiflagellar (3rd antennal) segment compound very

7 — male genitaliarotated 180°; large;

larva with and all 8 — fore wing lacking 2r-rs; eye antenna distant, 15 — both wings lacking SC; pseudosternum extending over with antenna 1-segmented, with mandible lacking incisive or almost all length of mesothoracic venter;

and molar flange, and with suprapedal and subspiracular lobes 16 — basiflagellar (3rd antennal) segment compound very merged; for numerous further synapomorphies see Rasnitsyn, large;

1988a: 120, items 5, 6, 8); 17 — antenna pectinate; fore wing with basal sections ofRS and

and M entire and with A2+3 9 — basiflagellar (3rd antennal) segment compound very forming straight line, straight, running

lost larval behind larva herbaceous large; flagellum beyond composite segment; leg area aspera; feeding on angiosperm secondarily 2-clawed with 2nd claw modified; plants;

10 — hind wing lacking m-cu cross-vein; 18 — lower tentorial bridge intervening between hypostomae;

sulci tibial male mandible of twisted into 11 — pseudosternal lost; preapical spurs lost; cutting type (with cutting edge moving with sternum 8 hardly visible, strongly excised apically; larva plane); fore wing with M+Cu angulate; Contributions to Zoology, 66 (1) - 1996 21

such the hypertrophied "third" segment) and do not im- originate with great difficulty; thus structures

ply any falsification of the existing and apparently are less likely to have multiple origins as compared

it well-based ideas regarding the evolutionary path- to their multiple loss. Equally seems logical to

ways of other characters. assume that for a structure ofenigmatic function, a

in The two last hypotheses are not equivalent unique origin and subsequent multiple reduction is

than mul- terms of being the best founded. Hypothesis 3, of a generally more likely to have occurred a

homoplastic origin of the xyeloid antenna from a tiple origin. The compound antennal segment is an

with rather and of ob- homonomously segmented one, results in more organ a complex structure

abundant of with the its and cases homoplasy as compared scure function. This makes multiple widely

alternative hypothesis (4) of a homoplastic loss of asynchronous homoplastic origin difficult to ex-

the character state (cf. italics in the captions to plain. The hypothesis of a repeated loss of the com-

antennal needs Figs. 11 and 12). This makes it possible to apply pound segment only a single and

the presumption of parsimony, which will confirm quite remote event of acquisition of the enigmatic

the multiple loss hypothesis. However, applying structure, and thus it seems superior compared to

the obscure the presumption of weighted similarity seems to the hypothesis of a repeated gain of

if takes into give still more clear results. adaptation. This is especially so one

Complex structures are generally considered to consideration the above-mentioned case of Syn-

tibial and 19 — basiflagellar segment compound, enlarged; spurs 34 — basiflagellar (3rd antennal) segment compound very

with soft tips; inner tooth of tarsal claws reduced; hypopygial large;

with cenchri with costal depression large; 35 — metanotum lost; fore wing space

membrane less folded larva with SC with 1st abscissa RS short and cell 20 — fore wing more or apically; on narrow, lost, of very

Angiospermae; 1mcu large, with cross-vein 1r-rs long, with Cu straight before

with with A2 and with abdomen 21 — basiflagellar segment compound, enlarged;fore wing 1m-cu, straight, area aspera lost;

SC lacking fore branch; femora widened; hypopygial de- compressed; larva boring in angiosperm plants;

with costal needle- pression medium-sized; 36 — fore wing space narrow; ovipositor

with and tarsal larva dead wood conditioned 22 — vertex grooves deep subparallel anteriorly; like; feeding on by symbiotic

claws with two long bristles medially; fungi, modified as explained in Rasnitsyn (1988a: 122, item

23 — basiflagellar segment compound, enlarged; fore wing with 22);

1st abscissa ofRS with 1r-rs short and and and very short, very thin, 37 — basiflagellar (3rd antennal) segment compound very

with conspicuous dark patches; large;

malar with dif- and hind with cross-vein 24 — basiflagellar segment enlarged; space 38 — fore wing wing single r-m each;

ferentiated setose depression; tarsal claws with acute basal 39 — scutellum rounded basally;

lobe; 40 — basiflagellar segment enlarged though entire externally;

25 — frons with pair ofswellings; female with inner tooth of hind 41 — basiflagellar segment compound, enlarged;

claw shorter than apical tooth; ovipositor sheath with stylus 42 — several basal flagellar segments enlarged about to form

larva makes glabrous; specialized leafroll; compound, enlarged basiflagellar segment; ovipositor scarcely

26 — basiflagellar segment enlarged; fore wing of female with extending;

distal third of costal cell glabrous; 43 — head capsule with postgenae subcontiguous; mesoscutum

27 — female with inner tooth of hind claw shorter than apical bearing transscutal suture; prepectus concealed under postero-

of tooth; lateral edge pronotum;

28 — ovipositor sheath with large stylus (possibly secondarily 44 — head capsule with postgenae contiguous; mesoscutum

lost in P. nemorum (Gmelin); larva makes specialized leafroll; with transscutal suture bent cephalad laterally; fore wing with

abscissa of 29 — basiflagellar segmentenlarged; basal RS subvertical;

with adlateral with 30 — basiflagellar segment enlarged; 45 — mesoscutum incipient lines; fore wing

tibia with SC lost crossvein-like 31 — fore hind (outer) spur rudimentary (or lost); except fore branch;

mesofurca with fore fused for larva 46 arms long, some distance; basiflagellar segment compound,slightly enlarged; prono- xylophagous on living plants, modified as described by tum short medially, dorsally; propleurae neck-like elongate;

Rasnitsyn (1988a: 122, item 19); larva feeding on angiosperm wood;

32 — fore wing with SC short, appressed to R, lacking fore 47 — head capsule with postgenae contiguous; tentorial bridge branch; narrow stripe-like, n-like bent; mesonotum with adlateral lines

membranous fore with well fore with SC 1st 33 — mesonotum except laterally; wing developed; wing lost; abdominal tergum costal space very narrow in basal third or half; ovipositor long modified into mesosomal propodeum; larva parasitic; etc., as extending; listed in Rasnitsyn (1988a: 123, item 28). 22 A.P. Rasnitsyn - Phytogeny, taxonomy, and nomenclature

Fig. 12. Cladogram ofthe lower hymenopterous insects (suborder Siricina) constructed consistent with the hypothesis [(4) in the text]

that in the is similarity xyeloid antenna symplesiomorphic, and the alternative similarities are synapomorphic. Otherwise as in Figs. 10—

11, with the following synapomorphies inferred:

1 — Same as in Figs. 10—11, and additionally, basiflagellar (3rd segmented, with subanal appendages lost, and with salivary antennal) segment compound and very large; gland ductus margined with 2 rows of glandular cells;

2 — antennal flagellum short (up to 6 segments beyond basi- 10 — head capsule with hypostomae contiguous between oral

fore flagellar segment); wing with 2r-rs connecting pterostig- cavity and occipital foramen; tentorium issuing anterior arms mal and RS far with 1st abscissa RS from below tentorial apex beyond 2r-m, of short, bridge; propleurae contiguous ventro- and with 1mcu cell large (for further possible synapomorphies, mesally; male gonostylus with gonomacula subapical; larva see Rasnitsyn, 1988a: 120, item 2—3); with abdominal sterna lacking prolegs and longitudinal and

short 3 — pronotum medially; fore wing with SC lost except oblique sulci, and with subanal appendagessegmented, shifted cross-vein-like fore branch; basally toward base of anal slit; larval salivary gland covered by

4 — single flagellomere beyond composite one; fore wing with common envelope, with ductus quadrangular in section;

RS lost between submarginal cells; larva boring fern stem, with 11 — ovipositor small, claw-like modified; larva exophytic, silk- caudal structures with distant from modified; protected, appendages setiform, antenna eye;

5 — male genitalia rotated 180°; 12 — both wings lacking SC; pseudosternum extending over all

2r-rs; larva with and almost all of mesothoracic 6 — fore wing lacking eye antenna distant, or length venter; with with mandible incisive antenna 1-segmented, lacking 13 — antenna pectinate, with basiflagellar segment seemingly molar flange, and with suprapedal and subspiracular lobes ordinary; fore wing with basal sections of RS and M forming merged; for numerous further synapomorphies see Rasnitsyn, entire straight line, and with A2+3 straight, running behind area

1988a: items 5, 6, 8); larva herbaceous 120, aspera; feeding on angiosperm plants;

7 — lost larval subdivided flagellum beyond composite segment; leg 14 — basiflagellar segment into primary segments

2-clawed with 2nd secondarily claw modified; and diminished in size (subequal in width to, though much

8 — basiflagellar segment seemingly ordinary; hind wing longer than, following segments);

cross-vein; with soft inner tooth of tarsal claws lacking m-cu 15 — tibial spurs tips;

9 — basiflagellar segment seemingly ordinary; pseudosternal reduced; hypopygial depression large; sulci lost; tibial lost; preapical spurs male with sternum 8 hardly 16 — membrane offore wing more or less folded apically; larva visible, strongly excised apically; larva with prolegs 2- on Angiospermae; - Contributions to Zoology, 66 (1) 1996 23

the loss ofthe texyela. It forces us to make a choice between hypothesis proposing a multiple xyeloid

different hypotheses that claim that the enigmatic antenna within the order Vespida is mine (Ras-

either reduction multi- structure evolved by or by a nitsyn, 1968, 1969, 1980, 1988a).

ple origin. The option again seems to be clearly in As portrayed here, the last hypothesis evidently

favour of the repeated loss hypothesis. looks preferable, but the opinions of Ross and

all weak. The of The comparison of hypotheses is still not com- Königsmann are not at shortage

plete, for we have omitted one relevant presump- knowledge about (Ross) or of respect for (Königs-

above the fossil record in- tion. Indeed, the considerations are mostly mann) provides only a partial

rather obvious, especially those concerning the hy- sight. The point of view of Ross and Königsmann

the of all is based the of the ir- pothesis on groups possessing additionally on presumption

xyeloid antennae. As far as I know, they have reversibility of evolution. We often treat irrevers-

been and rule never proposed before, an attempt to ana- ibility as a or even as a law, but exceptions are

known that lyse them looks a bit like creating a straw man. I to be so numerous it is hardly more than

need them, however, to complete the picture. The just another presumption. In the framework of our

other have been de- it is two hypotheses actually hotly approach just one more particular case of

bated. The one considering the xyeloid antenna as the presumption of analogy. Indeed, it is general-

homoplastic can be found in Ross (1937) and, in a ly agreed that within a particular transformation

more explicit form, in Königsmann (1976). The series, directions of evolutionary changes are

17 — fore wing with SC lacking fore branch; femora widened; Rasnitsyn (1988a: 122, item 22);

hypopygial depression medium-sized; 30 — basiflagellar segment somewhat diminished in size; fore

with and tarsal and hind with cross-vein 18 — vertex grooves deep subparallel anteriorly; wing wing single r-m each;

claws with two long bristles medially; 31 — basiflagellar segment dissociated into primary flagellar

with 1st abscissa of RS with 1r-rs 19 — fore wing very short, segments (still widened); ovipositor scarcely extending;

short and and with dark 32 be very thin, conspicuous patches; basiflagellar segment seemingly ordinary (unknown to

20 — basiflagellar segment seemingly entire, thin (though still either diminished in size as a whole, or to dissociate into

disproportionally long); primarily segments of ordinary size);

with differentiated tarsal base 21 — malar space setose depression; 33 — scutellar rounded;

claws with acute basal lobe; 34 — basiflagellar segment subdivided into primary flagellar

22 — basiflagellar segment seemingly ordinary; segments;

23 — ovipositor sheath with large stylus; 35 — basiflagellar segment reduced in size as a whole (still

24 — basiflagellar segment seemingly ordinary; disproportionally large);

tibia with hind diminished 25 — fore (outer) spur rudimentary (or lost); 36 — basiflagellar segment in size (subequal in

mesofurca with fore fused for larva width much head arms long, some distance; to, though longer than, following segments); xylophagous on living plants, modified as described by capsule with postgenae subcontiguous; mesoscutum bearing

Rasnitsyn (1988a: 122, item 19); transscutal suture; prepectus concealed under posterolateral

26 — fore wing with SC short, appressed to R, lacking fore edge of pronotum;

branch; 37 — basiflagellar segment seemingly ordinary; head capsule

27 — basiflagellar segment seemingly ordinary; mesonotum with postgenae contiguous; mesoscutum with transscutal suture

membranous fore with costal bent fore with basal abscissa of RS except laterally; wing space very cephalad laterally; wing

narrow in basal third or half; ovipositor long extending; subvertical;

28 — basiflagellar segment seemingly ordinary; metanotum 38 — mesoscutum with incipient adlateral lines;fore wing with

with cenchri with costal with 1st SC lost crossvein-like lost; fore wing space narrow, except fore branch;

abscissa of with and short RS very short, cross-vein 1r-rs long cell 39 — pronotum medially, dorsally; propleurae neck-like

1mcu large, with Cu straight before 1m-cu, with A2 straight, elongate; larva feedingon angiosperm wood;

and with abdomen larva in head area aspera lost; compressed; boring 40 — basiflagellar segment seemingly ordinary; capsule

angiosperm plants; with postgenae contiguous; tentorial bridge narrow stripe-like,

with costal n-like with adlateral lines well 29 — fore wing space narrow; probably (unless bent; mesonotum developed;

appearing later as a synapomorphy of all remaining Siricoidea fore wing with SC lost; 1st abdominal tergum modified into

and Vespina) ovipositor needle-like; larva feeding on dead mesosomal propodeum; larva parasitic; etc., as listed in

wood conditioned by symbiotic fungi, modified as explained in Rasnitsyn (1988a: 123, item 28). - 24 A. P. Rasnitsyn Phylogeny, taxonomy, and nomenclature

with concise and available form. Each of knows commonly asymmetrical, one direction ap- us pearing to be realized more often than alternative much about, say, a sparrow (of course including

Another func- ones. Therefore, when one can note unidirectional knowledge of higher taxa of birds).

is of changes as definitely realized, one ought to prefer tion to predict properties taxa not yet studied

in full detail. few insect the cladogram supposing no changes in the oppo- Indeed, a very species (in site direction, or the least of their number, unless relation to their full number) have ever been stud-

there evidence. ied and and until is strong contrary bionomically, cytologically, genetically.

insect consists of Indeed, contrary evidence does exist, as I have And yet we can be sure that any

that the final decision de- cells normal and tried to show above, so of a eucaryote structure function, pends on what we consider to be the stronger and with chromosomes, mitosis and meiosis, mitochon- sounder arguments. I argue in favour of the higher dria, DNA, particular kinds of RNA, etc. We can

where probability of the multiple loss vs. a gain of the predict with equally reasonable confidence complex structure of enigmatic function when a particular fossil insect or a contemporary one more than one mode of change occurred, and known from a single museum specimen lived and against the irreversibility of evolution. I hope that what it feeded on. The above considerations are

rather need mention them the example selected, per se convincing or not, commonplace. I to only does permit us to assess the proposed approach to to make the background clearer for further consid- cladogram reconstructions. erations.

The system ofphylogenetic presumptions is out- There are three main trends in the taxonomy of

lined in Table I. the last decades - pheneticism, cladism, and the

third, once called with a misleading name "evolu-

tionary systematics" to distinguish it from "phylo-

Taxonomy genetic systematics", an original and equally mis-

leading name for what most people call cladistics

We need a taxonomie classification of living be- (indeed, what is the difference between the terms

in and the ings to arrange the biological diversity such a "evolutionary" "phylogenetic" relevant to way that it facilitates our activity concerning that present context?) There are more names proposed

7 There are kinds of classifications for the third alternative: diversity. many "eclectic", "synthetic", created to fulfil particular functions and to meet "syncretistic" systematics, etc. (Farris, 1979: 497,

of kinds of These not terms but requirements particular users (e.g., a 518). are, however, just euphe- system of pests according to the type of harm they misms (replacing "bad taxonomy") to qualify the

little tries reveal cause). They present only a problem to bio- approach. This version of taxonomy to logical taxonomy, so that I shall concentrate on the and make explicit the principles of traditional tax-

whose of existence is be system, very purpose to onomy. It is not a full equivalent of the latter, how- universally accepted as a common (interdiscipli- ever, and thus it deserves a distinct name. Once I

all called it is evi- nary) reference system for users. tradistics (Rasnitsyn, 1992b), which

To meet the above conditions, the system must dently an awkward name. Phylistics seems a more

taxa that are for provide meaningful as many users adequate qualification. as possible. For example: "bird" is a definite notion for a taxonomist, and also for a hunter, a cook, and an artist. And to be so, the system must present its Phenetics taxa as within-group as homogeneous and as be- tween-group as heterogeneous as possible. In other Of the three competing approaches, phenetics

it have words, must the majority of all possible seems to be the most straightforward in its goals characters similar within and that the be taxa different be- and means. Its claim is system must ex- tween them. This property allows the system to ful- plicitly constructed in terms of similarity, and this fil two more functions of lesser to avoid in the simi- no importance. way permits one subjectivity

One ofthem is to store the information in the most larity calculation. The methods of calculation are Contributions to Zoology, 66 (1) - 1996 25

Table I. Phylogenetic presumptions.

Scope Term Statement

General "Rnowability" of phylogeny Any similarity should be considered as inherited, unless and

until the reverse is reasonably proven

related the the Group analysis Paleontological Among two apparently closely groups, oneentering

fossil record earlier should be considered ancestral, unless and until

sound contrary evidence is presented

Character analysis

Dissimilarity analysis (polarizing transformation series)

Paleontological Transformation series should be polarized according to the

succession of the respective character states in the fossil record,

unless and until sound contrary data are presented

Biogenetic Transformation series should be polarized in agreement with the

ontogenetic succession ofthe respective character states, unless and

until there are serious reasons to decide otherwise

Of If transformation series is in the results should analogy a polarized a group,

considered valid for another unless and until sound be as group,

contrary data are presented

Of irreversibility When we can see one direction ofchanges as definitely realized, we

to the should prefer the cladogram supposing no changes opposite

direction, or the least their number, unless and until we have strong

contrary evidence

Of functional efficiency Oftwo character states the one corresponding to a more efficient

adaptation should be considered apomorphic, unless and until

reliable contradicting evidence occurs

Ofcomplexity A complex structure should be considered as apomorphic in relation

to a more simple one, unless and until we have strong reason to

decide otherwise

of modified from Of vestiges The character state showing any signs being

another state is apomorphic over the latter, unless and until strong

evidence exists contrary

A character found within should be considered Outgroup state only a group

apomorphic in respect to that distributed both within and outside this

unless and until evidence group, strong contrary appears

Similarity analysis (identifying homoplasies in case ofconflicting similarities)

Of parsimony The most likely cladogram is that implicating the least number of

homoplasies, unless and until there are sound reasons indicating

another cladogram as the most likely

Of weighted similarity Of conflicting similarities, that which should be considered as

inherited is that known to be more reliable in other cases (especially

in closely related groups), unless and until strong contra-arguments

appear

called collectively "taxometry" or, more com- and a determination of the most adequate strategy monly though less adequate, "numerical tax- is not easy and cannot be made in a sufficiently

is onomy". Unfortunately, the methods have never objective manner yet. However, this a problem been developed to the proclaimed level. In my adherent to any approach (including cladistics: see

this it needs hard opinion, was not because the level has been e.g., Mickevich, 1978), and evidently proved to be unattainable but because phenetics labour to evaluate what is the natural domain of a

been forced before it had method and which has back by cladistics particular are its strong and proceeded far enough in its useful work. Indeed, weak sides. Phenetics proceeded far in this way

of it the methods taxometry are numerous and diverse, before was "consumed" by more vigorous - 26 A.P. Rasnitsyn Phytogeny, taxonomy, and nomenclature

of far the be discipline cladistics, but not enough to render Second, pattern can estimated by sampling

a conscious rather than a chance determination of characters and finding replicated, internested sets

of an adequate clustering method possible. of synapomorphies. Third, our knowledge evo-

The above failure does not seem to be the most lutionary history, like our classifications, is derived

important weak point of phenetics, however. Prob- from the hierarchic pattern thus hypothesized

is that the studied ably more important similarity (Platnick, 1979: 538). The pattern is hence claimed is by phenetics not precisely the same as what the to be primary and phylogeny secondary. This ap-

needs. with simi- raises system Pheneticists were dealing proach a question: what sets the pattern into of the characters which larity studied, must be rea- nature? or, in other words, what could be the gen-

numerous but not more than that. On the of the sonably erating process pattern if not the phylo-

for contrary, the quest similarity which qualifies genetic process? Being unanswered, the question the is of another sort. It is a kind of balance leaves to system no way analyse the proposal indepen- between characters and that all properties are dently of classic cladistics, so that I shall be con- shared and not shared by the taxa compared, in- cerned only with the latter. cluding the characters not yet studied (and maybe The first question to be raised is what the central never to be investigated). This kind of similarity is cladistic claim (taxon legitimized solely by a rather abstract thing, not directly available but synapomorphy) means. Synapomorphy is com-

important enough to justify attempts to evaluate it monly defined as a character state acquired by the in indirect This is what any possible way. precisely stem species of a group and thus characteristic of both cladistics and to do. all and members of that This definition phylistics try only group.

is not correct, however, because any acquired

be while character state can lost, an apomorphy is

Cladistics universally treated as unloosable. (Snakes have

lost but their their legs not "legness" as a

The of cladistics differs the that is meaning depending on synapomorphy.) This means an apomorphy of different appreciation students (cf. e.g., Hennig, not a character state but thefact that the particular

that 1966; Platnick, 1979) so significantly, De character state ever has been acquired (a fact of

Queiroz & Donoghue (1990a) apparently proposed history of a group). to re-instate the term original "phylogenetic sys- The above considerations are still inadequate to tematics" to denote Hennig's version, and to apply explain the cladistic approach, because there exists the term "cladistics" to its transformed only ver- a problem with groups lacking apomorphies. Tech- sion (the has not such proposal gained general accept- nically groups are termed paraphyletic or

so that I shall continue the latter whether ance yet, using polyphyletic depending on they are de- term for the entire this fined field). Despite splitting, by symplesiomorphy or homoplasy, respec- there exists a stable and, by inference, central tively. The general cladistic policy in that respect is cladistic "Phenetic classifications of both into marked statement: are to split groups sorts subgroups constructed similar- by grouping according to raw by apomorphies (that is monophyletic or, rather, ity, while phylogenetic ones are recognized by holophyletic, see note 9 below). This policy meets

to grouping according putative synapomorphy" a problem in metataxa, paraphyletic taxa resisting

(Farris, 1979: 487). It is the original and most the above procedure, i.e., those which cannot be widely accepted explanation by Hennig (1966) that split so as to give subgroups that are all demonstra- the taxa marked with make the ble apomorphies sys- as holophyletic (De Queiroz & Donoghue, tem to phylogeny and thus the ef- fossil isomorphic most 1990b). Many hymenopterous groups at- fective. That is the why system was termed tested to above (pp. 14-16) as ancestral in respect phylogenetic. There is an alternative of view, of other point Hymenoptera may serve as examples. however. First, nature is ordered in of a single Most cases this kind could be settled by suppos- specifiable which can be of pattern represented by a ing the difficulties as only temporary as a result hierarchical branching diagram or classification. our incomplete knowledge (this is especially - Contributions to Zoology, 66 (1) 1996 27

in the of The of for differ- tempting case fossils). refuge igno- different species (but see Sluys, 1984, a rance is not secure, however, because parallel evo- ent view ofwhat Hennig himselfactually meant by lution, as we made sure above, is prohibitive to a his deviationrule). This proposal seems to me a lit-

in tle but it is at least consistent. the cladogram being deciphered all necessary de- strange, It gives it tails. Besides, obligatory splitting of non-holo- system a double advantage. Firstly, becomes

leads the it is phyletic taxa inescapably us from stem clearly monistic, as usually opposed to the

groupto the stem species which cannot be split any claimed syncretistic nature of phylistics (but see further in this & the way (De Queiroz Donoghue, below). Secondly, system appears as objective

1990a). and as well grounded as the respective cladogram,

Some students again take refuge in ignorance though not more than that.

be If here by claiming that the stem species can never The last observation is important. some (possi- verified as such (Nelson, 1973a). Thanks to Popper bly many) species do not differ from their ances- it became clear that no scientific can be do have theoretical hypothesis tors, we not even any possibil- finally verified at all. Moreover, we are aware of ity to identify cladogram topology correctly. In-

factors from arrive limited no biological preventing a species sur- stead, we can expect only to at some

birth viving in some of its populations after giving degree of approximation. This makes advantages to another species, and then to do the same again of the cladistic approach less evident in compari- and Such identified in number in the again. species are son with alternatives. Indeed, contrary to claim

8 some more detailed paleontological researches. above, it seems possible to invert the proposition

This is not at all a surprise, because the total and to consider the cladistic system as ultimately number of ancestral species that ever existed is definable through sequenced apomorphies, with precisely equal to the number of terminal species the divergence events used only as a tool of the

because clade has been (minus one), every holophyletic sequencing. The approach already used must have its own ancestral species. by Brothers (1975) in his classical study of the

ancestral than and of aculeate A species to more a single taxon phylogeny taxonomy wasps. A seems to impose insuperable theoretical problems similar position apparently allowed Nelson (1989: for cladistics unless it makes the central cladistic that "descent without modifica- 280-281) to say statement (taxa definable only by apomorphy) less tion, however, is insufficient explanation of taxa".

has been made is rigid. This necessary step already Since a divergence event not necessarily cou-

himself. by Hennig He considered apomorphy as pled with a gain of apomorphy, that gain can easily legitimizing a taxon not by definition but only by proceed unaccompanied by a divergence. This re-

solid in of implication of another, more (basic, central) sults a complete knowledge a fully sequenced statement that is an isomorphy between the system array of apomorphies being unattainable even in and "The duration of That is the two at least phylogeny. temporal a spe- theory. why approaches are cies is determined of their value for by two processes speciation: comparable in phylogenetic study. the one to which it owes its origin as an indepen- This is not to say that they are precisely of equal

but and of dent reproductive community, and the one that di- value, the advantages deficiencies either vides it into two or more reproductive communi- seem to deserve special study. For example, be- ties" 1966: This of sides the of in (Hennig, 66). way reasoning impossibility being fully placed a makes divergence itself the only process that gen- sequence, apomorphies are hardly available just erates while becomes for in the framework of Brothers' taxa, apomorphy just a counting ap- marker of of divergence and, by inference, a taxon. proach. They cannot be usually hypothesized as

Anotherimplication of this point of view is that the appearing in one step (during a single speciation ancestral be considered dis- in species is to as having event) their full form, for most apomorphies are appeared after each divergence event. Even if one phylogenetically complex, not unitary. This causes of the of does ofthe products a divergence event not problems for the identification system of in- show the slightest difference from its ancestor (no ternested taxa created by successive steps in acqui-

it be considered sition As been apomorphy acquired), must as a of a complex apomorphy. has dis- A.P. - 28 Rasnitsyn Phylogeny, taxonomy, and nomenclature

cussed above as regards phenetics, we are as yet ally do when encountering conflicting data. Some

do this for in of not able to cladistics a fully objec- cases will show similarity by failing

tive manner, so that the approach adds more un- to resist more sophisticated testing. That was the

On the other with certainty to classification. hand, case the insect order Neuroptera, which has

Brothers' version seems richer in information con- long been used to comprise groups which later

thatof for it takes into have been tent as compared to Hennig, found to be only remotely related. For

consideration both relatedness and the amount of instance, Sharp (1895-1899) still comprised under

see this from the change acquired (but below). name, apart Neuropteroidea in the

with its undoubtable cla- Together advantages, current sense, also the Embioptera, Isoptera, Pso-

distics (in both its versions) also has shortcomings. coptera, Mallophaga, Odonata, Plecoptera, Ephe-

be considered later in They will on comparison meroptera, Panorpata (= Mecoptera), and Trichop-

with that of phylistics. tera. Similarly, the order Pachydermata

was long in use during the 18th and 19th centuries

to cover the , , rhinos, hippos, and

Phylistics , until eventually it has been discarded in fa-

vour of more natural (monophyletic) orders. In

tries to combine of both other the of Phylistics advantages cases, hypothesis polyphyly will be

phenetic and cladistic approaches in its attempts to falsified, as in the case of the hares and their rela-

considerationboth and take directly into similarity tives. They were once established as the separate relatedness. this To reach result, the phylistic taxon order Lagomorpha, only convergently similar to

is to be phenetically within-group homogeneous the true rodents (Gidley, 1912), but nevertheless all

and between-group heterogeneous and at the same of them are now considered to form the mono-

time it must be monophyletic. This kind of taxon phyletic taxon Glires (Wilson, 1989). In theory, it

has been termed a monophyletic continuum (Pono- is quite possible also to find cases where both hy-

marenko & is used Rasnitsyn, 1971). Monophyly potheses will appear equally likely. In practice,

there as in 9. Continuum real for the explained note means: a there is no danger there, possibility to

continuous subordinate chain (branching or not) of deepen our knowledge in both the similarity level

with each of them taxa being phenetically more and phylogenetic relations of the taxa studied are

similar of its direct than exhaustible. to any neighbours to mem- not We can never be fully certain of bers of other continua. any The definition seems either aspect of the system, and thus we rely more

clear and close traditional of to the understanding on probabilistic and intuitive criteria. That is why

the term "taxon". It looks in in eclectic, however, case of contradiction we can easily reassess our

criteria of and results and the conflict. combining independent similarity escape

relatedness. There are examples, however, when traditional

An is it eclectic approach a bad one because uses taxonomyretains taxa in spite of their undisputable

define independent criteria to one and the same no- polyphyly, namely when they cannot be replaced

the criteria tion. Eventually we shall find as con- by monophyletic ones without a significant loss in

flicting and thus impossible to use. So we should usefulness. I refer to various kinds of parataxa in

make clear whetherthe fields inquire into such cases to it such as paleontology, parasitology, mycol-

eclecticism is real or only apparent. We have seen ogy, as well as many procaryote taxa (see the next

and for above that cladistics is also seemingly eclectic, section details). I consider this as evidence that that able this we are to escape eclecticism in two the traditional approach relies ultimately on simi-

versions of ways that result in two alternative larity, while relatedness (monophyly) is used as an cladistics. indirect clue to allow us to assess similarity in the

The eclecticism of is also phylistics escapable. total array of characters including those yet un-

the Indeed, phylistic criteria are in conflict when a known. The relatedness is used here similarly to continuum is polyphyletic. When we encounter show how apomorphy is used in Hennigian cladis-

such a we should to and in the Brothersian result, try falsify it, as we usu- tics divergence events one. Contributions to Zoology, 66 (1) - 1996 29

all three consistent be be claimed As a result, are equally (monis- less important or less evident. It can

tic, not eclectic): Hennigian cladistics relies ul- that phylistics records both similarity and related-

succession of the Farris timately on divergence events, ness and, therefore, stores more information.

and uses apomorphies to mark and identify them. (1979) objects to this in that the cladistic system in

Brothersian cladistics relies ultimately on the suc- its fullest form specifies origin of all characters

cession of is informative of events acquiring apomorphy, and available and thus more than

uses the divergence events to sequence the apo- phenetics (and, by inference, phylistics) is. This

relies claim is the fullest morphies. Phylistics ultimately on similarity disputable, however, because

(in the most complete array of characters) and uses form is not attainable. The more information we

relatedness (as indicated by apomorphies) to im- acquire about polarization of the transformation

prove results of study of incomplete sets of charac- series available, the more new characters we find

with the series ters. To make the list complete, remember phe- not yet polarized. As a result, we are

netics: it is monistic and straightforward in relying never able to polarize the whole transformation se-

both proximately and ultimately on similarity ries that is available, so that cladistics has to oper-

alone. ate with an incomplete set of characters. Phylistics,

Now we can return to the question of compara- however, is able to incorporate into its system the

tive advantages and disadvantages of cladistics and total amount of information available for cladis-

phylistics. In my opinion, the most important tics, and to add the information about similarity of

is that it relies the which is connected with the phylistic advantage ultimately on taxa involved, non-

that characters such. is similarity, is, on as It may be polarized characters. That why the phylistic sys-

satisfied with raw similarity alone when no tem is potentially richer in its information content.

data of reasonable There is another side the of informa- phylogenetic any reliability ex- to problem

ist. In this case, it will be as empiric as phenetics is tion content of a hierarchic system. The above dis-

fact it will be like in cussion in fact the informationthat is (in just phenetics this respect). concerns not

will It use any reliable phylogenetic data when stored in the system per se but in its description or,

and because it the evi- in its the they exist, uses only more rather, background. As to system as such, dent it is the ones, possible that resulting system is it contains information solely on the topology of not exceedingly hypothetical. Opposite to this, the the respective dendrogram and in the length ofthe

and internodes. total infor- cladistic system relies both primarily ulti- dendrogram As a result, the mately on the relatedness that is knowable solely mation content of the system per se is that con- from character It tained in the of the dendro- interpretation. is based on infer- geometry respective ences concerning the evolutionary succession of gram, and as such it is limited. For instance, a sym- character states. As a result, the cladistic system metrical dichotomous dendrogram with 4 terminal

inference contains of a lower level of hypothesiz- taxa contains 3 bits of information and not a bit ing in comparison to phylistics. more, unless rank information is added. Unless

Another side of the above result is that phylistics ranking is involved, and beyond that the amount of is "omnivorous" while cladistics is nearly highly correlated similarity and relatedness, we can intro-

that duce into the "stenophagous", is, cladistics has require- system an additional amount of

for characters that much restric- cladistic ments are more information only at the expense of any tive. That is why it has to consider as parataxa phenetic information, and vice versa.

whose characteristics Additionalinformation ("plesions") many groups can be added only by us-

than are less complete those available for their ing a ranking system, especially a highly divided

One is forced the fact That is the informative relatives. to do so, despite one. why most arrangement that the characterized respective groups are well is the original Hennigian system with its incredible enough to be considered as orthotaxa (normal taxa) number of ranks that is supposed to reflect the suc- in the phylistic system. This makes the domain of cession in time of divergence events. However, cladistics such found and it narrow. a practice was impractical, now

Other of is abandoned in favour of putative advantages phylistics seem to virtually traditional Lin- 30 A.P. Rasnitsyn - Phytogeny, taxonomy, and nomenclature

The and nean ranking (Wiley, 1979). This form of ranking symbiotic hybrid taxa are polyphyletic,

little cladistic because is contains information, being used as their lower boundary crossed by more

and thus it than of it is in a completely arbitrary way, a single line ancestry. At the same time,

transforms post-Hennigian cladistics into an eclec- they can satisfy the definition of a holophyletic

tic concept. Artificial conventions like Nelson's taxon: a nearest common ancestor of all terminal

and all descendants of that phyletic sequencing (Nelson, 1973b; Cracraft, subtaxa, ancestor, may

1974) are of little help. We can accept a convention easily be included in that taxon, which thus be-

that a comb-like cladogram should be transformed comes defined by the synapomorphy(ies) gained as in of of and the rank in direct result of the a group taxa one same listed a hybridization or symbiosis.

with the a sequence taxon diverged most early Hence, polyphyly and holophyly are not mutually

and listed the first, the pair diverged most lately exclusive. This is no surprise, for they refer to dif- listed the last (in an arbitrary order). However, we ferent aspects of the history oftaxa (either to their cannot distinguish this list from the one, which is past, or to their future, respectively), which are not

derived sequenced arbitrarily by being from a mutually exclusive. As a result, we have to choose cladogram either more complex in form than a sim- between these aspects to rely on them ultimately as

criteria of I ple comb, or representing an unresolved . availability of a group as a taxon. vote

Another and deviation from in of if still more striking favour the future: we accept a taxon which the basic claim is the cladistic popular concept of is either strictly holophyletic (as a cladistic taxon),

(equivalent of the *taxon by Hennig) or holo- or paraphyletic (as a phylistic taxon), and and stem minus *taxon Hen- if exclude those which neither of the group (taxon sensu we are two,

the is nig). Indeed, stem group (see e.g., Ax, 1985) we shall avoid polyphyletic taxa in the traditional definition by a paraphyletic taxon which is explic- sense, and not so in the sense of hybrid and symbi-

itly prohibited in cladistics. otic taxa.

has been claimed that It phenetics and phylistics, Nevertheless, hybrid and symbiotic taxa do pose in contrast to cladistics, use taxa characterized only a problem for taxonomy. This problem does not af- by an absence of characters (Platnick, 1979), that fect these taxa themselves but their ancestral taxa is, paraphyletic taxa characterized by the absence (the complete chain between both parent species

it of apomorphy. However, is shown above (p. 26) and their common ancestor!) that are paraphyletic that apomorphy is neither a character nor a charac- by definition. However, this problem concerns

but rather the his- of ter state a hypothesis concerning only cladistics, not phylistics, nor, course, phe-

of tory the taxon. Absence of an apomorphy does netics. not mean absence of a character. Lizards lack the Other considerations relevant to the comparison

of have apomorphy snakes: they legs instead. of the three leading taxonomie concepts certainly exist. Nevertheless, the result already obtained

The problem of symbiotic and hybrid taxa must seems sufficient to conclude that each system has also discussed here. The is and The balance be problem important, its own advantages disadvantages.

of because both kinds taxa are common enough not between them does not seem equal, however. In to be disregarded as something exotic. The first my opinion, phenetics and cladistics demonstrate category is represented by no less than the entire deficiencies important enough to consider the two

cell de- taxon Eucaryota (their organelles being to be inferior to phylistics. These weak points are: rived from originally free-living organisms), while the shortsightedness of phenetics, which fails to the latter is for of the of and the very characteristic, example, use prognostic power phylogeny, higher plants. As reviewed by Tzvelev (1993), in fastidiousness of cladistics, which rejects too large the the grass family (Poaceae), Tribe Triticeae a fraction ofavailable characters and also produces consists of 500 300 which species, of are caryo- taxonomie constructions with a high hypothetical typically certain to be intergeneric hybrids, and the content. I believe this makes the whole construc-

in the tribe consist of largest genera hybrid species. tion somewhat more shaky than the phylistic sys- A similar claimed pattern is to hold true for many tem. Therefore, I expect that phylistics will appear other higher plant taxa. the winner of the contest. Contributions to Zoology, 66 (1) - 1996 31

Nomenclature Rasnitsyn, 1982, 1986, 1989, 1991).

3. Identical names of different taxa must not be

Nomenclature is collection conventions The a of on used as valid names (ICZN Article 52). scope

the of is how to create and use names taxa. It com- of the principle of homonymy is arbitrarily re-

monly felt to be a field of sophisticated details and stricted so as not to concern (i) the species group

basic This is the different ho- simple principles. not exactly names belonging to genera, (ii)

case, for the principles are not all appreciated and monymy with (and between) higher taxa and (iii)

between explicitly formulated. Some of them are mentioned homonymy taxa belonging to animals and

in the Codes (e.g., ICBN, 1994; ICZN, 1985) but to plants.

not basic In of these 4. Selection of the name as principles. addition, some proper among compet-

basic principles are not true conventions by being ing ones, either synonyms or homonyms, must be

taxonomie earlier Ar- only a reflection of a particular concept, made in favour of that proposed (ICZN

inference of evolution- of or obtained by a particular ticle 23). This is the principle priority.

ary theory. Indeed, the Linnean principle of bino- 5. Principle of a standardized and rank-indicat-

minal nomenclature is conventional and base fully eas- ing form of the taxon name. It presents a for

ily coupled with any kind of taxonomy accepting a the long array of rules on the language, grammar

hierarchy formed by genus and species. In con- and syntax of the taxon name, including the princi-

trast, the type principle, as is discussed below, is ple of binominal nomenclature (ICZN Articles 4-

essentially different. The distinction between the 6, 11,25-34).

kinds of that The last the two principles seems useful, so they 6. among taxonomy-independent

deserve their own names and will be further re- principles is the principle of superior priority ofthe

ferred and InternationalCommission Nomenclature which to as taxonomy-independent taxonomy- on

dependent ones, respectively. can rule on any case contrary to the Code (ICZN:

of the follow- Because my personal experience, xiv).

ing considerations are based on the rules of zoo-

nomenclature.This make little dif- logical seems to

ference, for various Codes differ largely in details, Taxonomy-dependent principles

not in underlying principles.

There three are taxonomy-dependent principles.

1. The type concept. The principle of name--

is the and demon- Taxonomy-independent principles ing types most important most

strative with respect to the restrictions imposed by

These are often plain and need but little comment, taxonomy on nomenclature.According to the Code

first. the if at all, and they are discussed here "The name-bearing type provides objective

which 1. "The Code refrains from infringing upon taxo- standard of reference by the application of nomie judgment, which must not be made subject the name it is determined, no matter how the

boundaries of Arti- to regulation or restraint" (ICZN: xiii). This state- the taxon may change" (ICZN

ment is not fully correct because of the existence of cle 61a). In other words, a taxon can be introduced

that validate be taxonomy-dependent principles only into the system ultimately only referring to its particular cases in taxonomy. The statement should type. To assess this decision we should consider probably be completed with the words "beyond re- alternative possibilities.

strictions explicitly imposed by the present Code". The first thing to consider is the way to introduce

2. In zoology, the area of principles of nomen- a taxon by referring to its characters. This iden-

that clature is restricted to suprapopulational level tifies a taxon as a class provided the charac- groups from subspecies up to superfamily (for de- ters are defining ones and not merely diagnostic tails see ICZN Article 1). Exclusion ofhigher taxa (Ghiselin, 1974, 1987). There are several possibili- from the scope of the Code seems to me a mistake, ties of doing this; the simplest is a combinatorial

is for - with each though it a subject another discussion (see system a multidimensional matrix - 32 A.P. Rasnitsyn Phylogeny, taxonomy, and nomenclature

combi- their relatedness. becomes the compartment corresponding to a particular Relatedness thus definition. The nation of characters and thus harbouring a separate Lubischew parameter by meaning

table and of the been taxon. A version of this is an identification potentialities cladistic system have

- draw sometimes used by taxonomists a rectangular discussed above. Here I only want to atten-

with of characters and left col- tion to the fact that the does not need its matrix an upper row system

the inter- umn of taxa, and with character states at taxa to be "typified" because it is enough to referto

sections. This sort of system is really simple and the parameter to introduce a taxon into the system.

with the if refer sometimes quite useful. It does not agree Indeed, "names are synonymous they to

aims of the general system of organisms, however. clades stemming from the same ancestor" (De

inflexible: combina- 1990: It is too we cannot improve a Queiroz & Gauthier, 307). However, De

torial system locally to fit a particular taxon, for Queiroz and Gauthier failed to draw the self-evi-

deletion alternation of character dent inference from their that cladistic any addition, or a proposal

other will affect many taxa. That is why such a sys- taxonomy should abandon the type concept. This

identification next has been taken tem is not very common, even as an step by Sundberg & Pleijel

tool. (1994). I am not aware of any practical application

A flexible is with the iden- of their and I this will be more system one taxa effort, am afraid that a

tified by ranked characters, as in the standard key difficult task. Their proposal is that cladistic no-

with attach for identification, characters being more menclature should a taxonomie name as a

highly ranked the earlier they occur in the key. It is tag to an ancestor of a clade. However, as we have

it is not as as the the cladistic does not to quite handy, although good gen- seen, approach permit us

eral system because of the evident arbitrariness of think about an ancestor as a real group (taxon) that

the character in somewhat be and studied. the rankings. Nevertheless, could identified Instead, ances-

modifiedform few of rank (that a the highest char- tor is thought about only as a collection of syna-

acters are to the this of the clade in of enough shape system) ap- pomorphies question plus a sum

has remained taxonomists of proach popular among synapomorphies all more inclusive clades. All

since Linnaeus (1751), who based his system of these synapomorphies do not result from direct ob-

plants primarily on characters of fructification. servation, as the type specimen does, but exist only

This has proposal been formulated most explicitly as hypotheses concerning which character states

by Lubischew (1923, 1966), who called for a have been acquired at what divergence event. So

search for characters a few highest rank (para- we are effectively being told to attach a taxonomie

meters) that could determine the distribution of all name to a collection of hypotheses, which I do not

other like the nuclear deter- believe will be to taxonomists characters, charge an easy job persuade

the features of and the mines atoms position of re- to do.

elements in the Mendeleev This spective system. The class, as a group defined by the characters of

would result in discovering a parametric system, its members, has the individual as its logical alter-

is able characters native. The consider the which to predict all important of suggestion to taxon as an

the taxa involved from these key characters (the individual (Ghiselin, 1974, 1987, and bibliography

The task Lubischew parameters). proposed by therein) is thus quite natural. As an individual, a

seemed hopeless as applied to taxonomy, and taxon must then be able to be introduced by refer-

failed Lubischew himself to resolve it. However, ring just to its name. For this, it must have its integ-

meanwhile has a solution been found, albeit in a rity and spatiotemporal wholeness developed

place other than where he sought for it. Ironically, enough to permit the taxon to be born and to die. It

it was discovered in a field which Lubischew, as a cannot be broken down into parts equal enough so

convinced considered antiselectionist, as false. that more than one of them might pretend to inherit

field is the The cladistic system with its central the taxon name. This makes a difference between

claim of in- in- that the characters organisms and, by the ostensive definitionof an individual (direct

the location of their in the of and the the ference, taxa system, can dication its member/part) use of type be best determined one be by only characteristic, viz., concept. Any part of an individual may equally - Contributions to Zoology, 66 (1) 1996 33

intended and hard rules of the features of both the class used for the definition, deed, taxon possesses how to select and to use the taxonomie types are and the individual: as a class it possesses charac- not necessary to introduce individuals into a sys- ters, as attested to by its diagnosis, and it has mem- tem. bers (contrary to Ghiselin, Alex Rasnitsyn is not

"Taxon" therefore is a very special kind of indi- only a part but also a specimen [example] ofHomo

if it is in the to what is vidual, anything (see e.g., comments sapiens)., At same time, similarly so

and the issue deserves further for the has in Ghiselin, 1981), con- individual, a taxon parts (populations sideration. The paradigmatic case of an individual, relation to species), and, what is more important, it the has its and its individu- organism, integrity spatiotemporal can develop (evolve) and still not lose wholeness based continuous interaction of its While retains its wholeness on ality. evolving, a taxon parts. This also has been proposed to be true for in time, in the multidimensional space of all its

due to fact that their is to in species, the divergence sup- characters, and an extent geographical space pressed by gene flow (Mayr's biological species (see e.g., Schram, 1980). Because of this whole-

is limited in the is able be after its concept). The model its scope, for ness, taxon to appreciated

"The has biological species concept validity only name is given, as is an individual. In possessing in what I have called its 'nondimensional situa- features of both class and individual, the taxon is a

that where in notion intermediatebetween these rather it tion', is, populations are actually con- two, or tact with each other" 1988: fills the between because the class and (Mayr, 301-302). gap them,

individual Ghiselin (1981) hypothesized as well that the the appearto be two extremes of a single phylogenetic has integrity sufficient to spectrum. consider the lineage and the respective taxon as an Consequently, the taxon in the above sense is individual. There is a problem here too, for all of neither a true class nor a true individual. It is most the ancestor-descendent transitions in- whose are equally similar to a cloud spatiotemporal wholeness tegral. I am equally related to my father and to my is real, though far from being complete. A taxon

and is both its and de- be best defined in son, so a species to ancestor seems to as a continuum, agree- scendent The difference results from ment with its The species. only phylistic usage. phylistic ap- the asymmetrical distributionof common history: I proach is well suited to the task of introducing and share common with father but it is the continuum that a history my not handling taxa using types: with my son [thanks to one of my anonymous re- can be introducedinto a system ultimately by using viewers for showing me this dichotomy]. This is a a type. Indeed, unlike an individual, the possibility

of the of the cladistic continuum in- precise description essence to identify a depends on similarity system. It displays very clearly what a special kind stead of integrity, and unlike a class, a continuum of individual be identified in relation overall a cladistic taxon is: it can be born, can (in similarity) exist, and die, but it can never give birth to another to other taxa and not by means of particular charac- taxon. The only thing available to it is to segregate ters. If an individual is an integral body and a class

is in the character its internal contents into subclades (subtaxa). I can a compartment matrix, a con- see here an exaggerated, all-depressing role for tinuum is neither of the two. It is rather a cloud for

it history: nothing in the world, not a single evolu- which is possible to change its shape and com-

how "brilliant and in- tionary invention, no matter posing elements, but nevertheless it persists as an fluential" in determining the history of a clade it appreciable thing until it is dismembered or dis- could be, is comparable in its taxonomie meaning solved (becomes extinct). It is possible to charac- as is common history itself. The past history is eve- terize the continuum both by its integrity, manifest

the future is - this is in the existence of it rything, next to nothing appar- gaps delimiting from other ently a fairly correct characteristic of the cladistic continua, and by the characters of its diagnosis. view of the system. Both criteria are not fixed, however, through being

the does in the of However, class-individual dichotomy easily changeable course both evolution not exhaust the In- and that at necessarily existing options. knowledge acquisition, so any particu- 34 A.P. - and nomenclature Rasnitsyn Phytogeny, taxonomy,

lar moment they can be found insufficient to de- Reasonsfor refusing to follow the principles of no-

limit the taxon. That is why the taxon is ultimately menclature

identifiable through its name attached as a tag to its

In the of nomenclatural type. case taxonomy-independent principles,

the The same result can be achieved by means of a reasons for a deviation are necessarily subjec-

differentand shorter path ofreasoning. Indeed, the tive in the sense that there is no taxonomically

aim of any classification is to accommodate all of compelling reason to do so. That is why these cases

discussed here. The with the the diversity to be classified by a system of taxa are not problems tax-

that should neither overlap each other, nor leave onomy-dependent principles seem more important

uncovered and deserve consideration. As have been scru- any space (unclassified residue). This they

aim of classification tinized elsewhere will be implies that the main a is to (Rasnitzyn, 1986), they

treated in a fashion here. trace natural gaps ("to cut nature at her joints", to only cursory

Reasons for the of use Plato's metaphor as cited by Hull, 1983: 186) not following principles no-

in the field under classification, and to fix and to menclature arise when the shortage of information

taxonomie into rank them as boundaries (in case of an necessary to classify a particular group comes

conflict with the artificial classification, we just impose such boun- necessity nevertheless to classify it. The arises from scien- daries instead of tracing them). Class does not fit necessity the applied or

this it is defined tific of the As the in- procedure, for by a character, for importance group. regards the which it is quite natural to overlap with other char- complete information, reasons are usually con-

nected with some of the ma- acters. For individuals, we are aware of no natural specific imperfection

could them leave terial involved. This is well known in forces that pack so tightly as to no paleontology

in and in the of with a space between, unless individuals are defined in taxonomy groups complicated

as parasitic worms and a sophisticated manner, as is done by cladists. In ontogeny (such fungi).

The least is taxon incertae contrast, the taxon-continuum meets perfectly the disturbing case a

sedis. It differs from orthotaxon demands of the classification procedure, because an (normal taxon) in that its characteristics make it impos- the only legal way to define a continuum is to trace incomplete

sible to its at of and to rank border lines between it and other taxa. specify position a particular level

the hierarchy. For instance, a genus incertae sedis

can be assigned to an order, but not, at least for the Besides the type concept, there are two more moment, to any particular family. This practice can taxonomy-dependent of nomenclature, principles be interpreted as a local rejection of the principles ICZN. both not appreciated as such by They are of and subor- hierarchy synonymy. Indeed, a genus

also an of the continuum nature of a implication dinated directly to an order can be understood as

taxon. attributed to a new unnamed family, which can 2. The of continua, principle hierarchy. Being easily be a synonym of an existing family.

taxa be in other in the cannot grouped any way than Other and more serious cases concern only the

next more inclusive continua The (higher taxa). principle of synonymy and can all be attributed to a

latter be that in- must fully inclusive, is, they must special field called parataxonomy (the definition

clude the subordinate continua in as a whole, not published by Melville, 1979, is rather narrow in

the parts. Equally, taxon-continuum, even as a comparison with practical usage). There exist sev-

whole, cannot be a member of two or more higher eral distinct kinds of parataxa. The term formal

taxa, for otherwise the latter would and broader overlap taxon was proposed (Rasnitsyn, 1986) as a thus would the not obey definition of a continuum. interpretation of the largely botanical term form-

As a result the takes hierarchic Article formal be system a fully form. genus (ICBN 3). A taxon can

3. The of If principle synonymy. a taxon can be a treated as if it is an orthotaxon, but only in the

member of only a single higher taxon, it can only framework of a special system which is parallel to bear for a single name, extra names being destined the general one and completely independent of it,

invalidation because of synonymy. except for the principle of homonymy. These are Contributions to Zoology, 66 (1) - 1996 35

in created fos- Table II. ofnomenclature. for example taxa systems to classify Principles sil detached leaves, or seeds, or elytra, fos- Taxonomically independent silized animal traces, and other works (e.g., caddis 1 unrestricted taxonomie freedom beyond the restrictions larval of cases), or stages living parasitic worms, or imposed by taxonomically dependentprinciples even unassociated males in highly dimorphic in- 2 restricted area of application 3 with female-based clas- homonymy sect groups a traditionally 4 priority sification. Parataxa differ from orthotaxa only in 5 standardized and rank-indicating names that of them any are possibly (sometimes even cer- International Commission 6 superior authority of the on tainly) a synonym of some orthotaxon, and yet they Nomenclature should not be synonymized with the latter. Impor- tant considerations on the nature and functions of Taxonomically dependent

not 1 (with no good reasons for following it) be found in type parataxa can Meyen (1990). 2 hierarchy (not to be followed for taxa incertae sedis) Some cases of this sort are claimed to involve the 3 be followed for all synonymy (not to parataxa except type concept as well. Some of them are difficult to those belonging to one and the same particular system) the ICZN Article explain (e.g., case of ichnotaxa,

66). Others are known to arise as a means of de-

fending parataxa from the persistent drive of some reet. According to the above definition, the collec-

in the taxonomists to synonymize parataxa, however ar- tive taxon is equivalent its scope to respec-

members bitrary such synonymy might be. As an example I tive higher taxon and thus its are ulti- refer to the of the Laberius identifiable the same A case genus Kieffer, mately by referring to type.

1914 (see Rasnitsyn, 1986, for details), which was species of Cercaria can be identified as such sim- created to house male from the tribes because it is similar to larva of the wasps ply enough a

Dryinini and Gonatopodini (Dryinidae), whose ge- trematode type species. It does not matter that this neric and tribal is unknown because of the I have little position type species is not as yet appreciated.

female-oriented of these taxonomy groups. Never- doubt that higher taxa will be eventually typified in theless, the genus Laberius has been synonymized zoology, as has been done in . Another ex- with of the orthotaxa within the one Gonatopodini. ample is the collective genus Carabilarva Pono-

As A.G. Ponomarenko explained to me, it was this marenko, 1985, proposed (in Rasnitsyn, 1985) to

which had stimulated him house larvae of Mesozoic of the case to create detypified super- parataxa for fossil beetles in Rasnitsyn (1985: 47- family Caraboidea that are impossible to attribute

does Because of the of 81). Detypification prevent ungrounded syn- to a family. this, type species

is onymy. Yet it prevents any synonymy, thus being Carabilarva the type of the superfamily Cara- perhaps a case in which the medicine is more dan- boidea, that is Carabus granulatus Linnaeus, 1758.

than the illness itself. Thus Carabilarva is gerous I hope that a better formally a junior objective understanding of the aims and functions of para- synonym of Carabus and yet the two must not be

will be defence their taxa a against arbitrary syno- synonymized as long as they belong to separate nymization. systems.

The last and kind of col- considerations the poorest parataxon is a The preceding on principles lective taxon. It is a parataxon, usually of generic of nomenclature are listed in Table II.

that be but rank, can assigned to a higher taxon

be cannot organized there in a special system ofthe above sort. For example, “Cercaria O.F. Müller, Summary

1773, established for a genus of worms [...], is used in the scientific present state of knowledge as a collective- Phylogenetic inference, like any work, group name for Trematode larvae that cannot be relies on (1) observations (including experiments), placed with certainty in known genera..." (ICZN (2) search for analogies, (3) creating of hypotheses

Article 67m). The collective taxon is also stated about the underlying patterns and mechanisms, (4) there to be detypified, but this viewpoint is incor- attempts to falsify these hypotheses (mostly A.P. - and nomenclature 36 Rasnitsyn Phylogeny, taxonomy,

through their implications), and (5) assessing re- tinuum, where discontinuities (gaps) are used to

set of delimit and is considered sults of the above attempts using a presump- taxa, monophyly as a

tions. means to assess whether the resulting taxon is cor-

The set of phylogenetic presumptions covers the rectly delimited, or the system deserves re-investi-

presumption of "knowability" of phylogeny, and gation.

of more Taxonomic nomenclature is claimed to be built an array particular presumptions possible to be segregated into two groups, involved eitherin on the basis of 9 principles, 6 of which are purely

in character conventional in the that group analysis or analysis (Table I). sense they are independent

The ancestor-descendant of taxonomic and 3 group analysis applies to a concept employed, are relationships and relies primarily on the paleonto- taxonomically dependent principles, as is shown in

for The logical presumption groups. presumptions Table II. Among the taxonomically dependent used in the character analysis are further separable principles, the most important one is the type con-

those relevant of and the that less stabi- into to the study differences, cept, which is only one more or those used in the analysis of similarity. The former lizes the name of the taxon, that is designed as a

the taxonomi- subgroup covers presumptions that help us to monophyletic continuum. Two other polarize transformation series, i.e., to discriminate cally dependent principles, those of hierarchy and the and of be not followed in of plesiomorphic (ancestral) apomorphic (de- synonymy, may cases rived) character states. This group comprises a taxonomically deficient material, which then wide of the most of should be incertae array presumptions, important organized either in a taxon which are the paleontological presumption for sedis, or in a special (independent) system of a par-

the the ticular kind of characters, biogenetic presumption, pre- objects (e.g., of particular detached of sumption analogy, those of the irreversibility of fossil organs, or of a particular developmental

of functional of of of There evolution, efficiency, complexity, stage a living group organisms). are of vestiges, and the outgroup presumption. The neither valid reasons found to violate the type con-

in the violate of the inde- presumptions engaged similarity analysis cept, nor to any taxonomically concern the problem of the inherited vs. indepen- pendent principles, which are conventionally de-

These first of in order make the taxonomic dently acquired similarity. are, all, signed solely to prac- the presumption of parsimony, and the presump- tice more stable and uniform. tion of weighted similarity.

The objective of taxonomy is to create a system

for whose taxa are meaningful the maximum diver- Acknowledgements

of sity possible users, i.e., taxa should reveal maxi-

The version of mum within-group homogeneity and between- present paper represents an expanded a text

written in which circulated 1989-90, was among colleagues in group heterogeneity. Of three main rival taxo- Russia and abroad. I have benefitted from the input of nomic relies many approaches, pheneticism solely upon colleagues, but I am particularly in debt to Prof. James M. available characters as such and thus appears to be Carpenter (Museum of Comparative Zoology, Harvard Uni- too empirical). Another extreme myopic (purely versity, Cambridge, MA), Dr. Michael T. Ghiselin (California represents the cladistic approach, which employs Academy of Sciences, San Francisco, CA), and Prof. Frederick

Schram for and only apomorphies (inferences about the evolution- R. (Institute Systematics Population Biology, University of Amsterdam) for their extremely helpful discus- ary history of character states). As a result sions, for advice and help concerning the pertinent literature, is cladistics fastidious, and the cladistic system is and for assistance with linguistic and other stylistic problems. overburdened with content. The Criticism and of reviewers of hypothetical suggestions two anonymous were

is intermediate in Its ul- in visible mistakes and phylistic approach a sense. most importance making many weakly

based I also thankful for the discussions and timate goal is the system thatreflects a comprehen- statements. am advice of Prof. Charles D. Michener of Kansas, sive balance of similarities and dissimilarities, (University Lawrence, KA), Dr. K.Yu. Eskov, Dr. A.G. Ponomarenko, Dr. while the relatedness is considered to be an impor- S.A. Rautian, Dr. M.A. Shishkin, Dr. V.V. Zherikhin (Paleon- tant heuristic method to approach this goal. tological Institute, Russian Academy of Sciences, Moscow),

Phylistics defines a taxon as a monophyletic con- and Dr. I.Ya. Pavlinov (Moscow State University, Moscow). - Contributions to Zoology, 66 (I) 1996 37

Very important was the great effort by Prof. Schram, Drs. eludes weighting of the phylogenetic evidence (Farris,

Florence F.J.M. Pieters, and Dr. R. Sluys (Amsterdam) to edit 1983). This approach makes the notion of the most parsi-

fit it the rules of and of the Contribu- monious ofthe best my text to English style hypothesis just a synonym hypothesis

ofMr. Andrew J. Curator of and thus Farris himself has said tions to Zoology. By courtesy Ross, unnecessary (as on an-

fossil insects, I had the opportunity to study the collection of other occasion, "If everything is 'phenetics', there is no

insect inclusions in Burmese amber, kept in the Department of need for the term 'phenetics'; 'everything' would do just

Invertebrate Paleontology, Natural History Museum, London. as well"; 1979: 488).

[An abridged version of the present account is published in 5. An apomorphic similarity of the living xyelids in the sub-

Russian (Rasnitsyn, 1992a, b).] divided galea (a lobe of the mouthparts) is possibly of

post-Triassic origin. It is proved only for the subclade

comprising the subfamilies Xyelinae and Macroxyelinae

and is not known for the Triassic Archexyelinae. Notes 6. To be more persuasive, the cladograms presented here

possibly should be prepared using one of the popular

This is cladists who consider 1. problem ignored by an an- computerized cladistic methods. I have some experience

cestor only as a collection ofplesiomorphies and refuse to in this matter (Rasnitsyn, 1995) and could do it. However,

and real ancestor known try identify a among organisms, it should be a special, and rather extensive work, for it

either extinct or extant. I do not believe this approach is would be senseless without taking into consideration all

fruitful the data and ideas available and without the (see below). new respec-

2. The fact that their characters is ir- tive réévaluation of all and character involved. we trace groups using taxa states

relevant to the to discriminate I believe that the data here if possibility group analysis presented as they are, even

and character The is discussed at outdated, are to how the analysis. problem length slightly appropriate explain pro-

below (pp. 7-9). In short, I follow Linnaeus who said posed approach works, and how it has permitted me to

"Scias Characterem non constituere Genus, sed Genus reach the conclusions presented here.

Characterem" that is the character 7. Griffiths restrict the classifica- [1751: 169], "Know, (1974) proposes to term

does not constitute the but the the character" not genus, genus tion as applicable only to ordering into classes, into

(translation mine, corrected by M.T. Ghiselin). systems. I do not follow this restriction, because the

3. It can be seen from these that the classes and in in easily examples using systems (individuals, current usage taxo-

for like nomie such clear-cut paleontological presumption groups, any pre- theory) are not things as common

is sumption, not an easy and straightforward thing. It belief suggests (see main text, below).

needs thorough knowledge of, and considerable experi- 8. E.g. researches on bivalve molluscs inhabiting the Para-

ence in, the and related and eventhen mis- subject topics, tethys, a system of epicontinental seas of southwestern

takes are unavoidable. Of course, there is no faultless Eurasia during the later Tertiary (Neogene). A compli-

method of at all. there cated environmental cognition However, are two differ- history of the seas, with rapid, fre-

ent strategies possible here: we could prefer either to quent, and often reversible changes in hydrology and pat-

avoid to them. first mistakes, or correct In the case, we tern of straits inside and outside the system, had a strong

would consider only the most safe methods (those least effect on the evolution ofthe biota. The bivalve molluscs

of dangerous giving incorrect results). This strategy ofthe Paratethys have been a subject ofintensive research

would reduce dramatically the diversity of our scientific for more than a century because of the economic impor-

and thus lowers the approaches demands for the abilities, tance of oil-bearing strata, and currently they are rather

knowledge, and skillfulness of scientists. More people well known. Besides, bivalves have a special meaning in

would be able to participate in scientific work, albeit us- this context. At lower taxonomie levels their classifica-

ing a reduced diversity of methods, while high abilities of tion relies on their shell characters, both in fossil and in

few would This the results ofthese particularly gifted persons rest unapplied. living populations. Unfortunately, ex-

ideal of a mass, standardized science is not mine. I prefer tensive studies have mostly been published in Russian.

the science which is free to use a broad variety of meth- They have been recently summarized, together with rel-

ods, dares to make and correct mistakes (and to turn the evant literature, by Nevesskaya et al. (1986).

mistakes to their of and 9. I follow Ashlock term very good, as a source experience (1971) using monophyly as a cov-

discoveries), and permits everybody to work at his maxi- ering both and holophyly, the latter being

That is I and in cladistic mum efficiency. why am trying to develop equivalent to monophyly usage. I refuse to

the which both follow the cladistic for the Taxa support approaches promise interesting usage following reasons.

and mistakes. be classified new findings pitiful can phylogenetically depending on how

4. I use the traditional of than discernable of interpretation parsimony following many (one vs. more one) lines ances-

Ockham's and related the notion of of their razor to economy, try (successions of ancestors) cross lower border

the least of reach the aim. line. classifiable necessary amount means to Equally they are according to presence

There exists wide absence of of the another, more understanding which in- vs. descendants group, that is whether A.P. - and nomenclature 38 Rasnitsyn Phytogeny, taxonomy,

its border line is crossed of de- 1974. models and classification. upper (side) by line(s) Cracraft, J., Phylogenetic

scendants or not. A taxon with more than a single ances- Syst. Zool., 23: 71-90.

tral line is universally called polyphyletic irrespective of Darwin, C., 1859. On the origin of species by means of natu-

possessing descendant line(s). A taxon with a single an- ral selection, or the preservation of favoured races in the

cestral line is called traditionally monophyletic also irre- struggle for life: i-x, 1-502 (John Murray, London).

spective of its further fate (whether it has descendants or Eskov, K.Y., 1987. A new archaeid spider (:

only members). Cladists use the latter term exclusively Araneae) from the Jurassic of Kazakhstan, with notes on

for taxa with ancestral line and without other the so-called "Gondwanian" of recent taxa. Neues a single any ranges

descendant ones, while taxa with a single ancestral line Jb. Geol. Paläont. Abh., 175: 81-106.

and existing descendant line(s) are termed paraphyletic. Eskov, K.Y., 1992. Archaeid spiders from Eocene Baltic am-

leaves for the notions ber Araneida: with remarks This usage no term two combined, (Chelicerata: Archaeidae) on

for the traditional The cladistic the so-called "Gondwanian" of Neues i.e., monophyletic taxon. ranges recent taxa.

approach needs no term of so broad a meaning, for Jb. Geol. Paläont. Abh., 185: 311-328.

the is cladistically paraphyletic taxon a nontaxon and a Farris, J.S., 1979. The information content of the

combination of and is This is 28: 483-519. nontaxon taxon nonsense. phylogenetic system. Syst. Zool.,

not the case for phylistics (or phenetics) which considers Farris, J.S., 1983. The logical basis of phylogenetic analysis.

paraphyletic taxa as valid and thus badly needs a term In: N.I. Platnick & V.A. Funk (eds.), Advances in

covering all taxa with a single ancestral line. The broad cladistics, 2: 7-36 (Columbia Univ. Press, New York).

of has the and 1974. radical solution the usage monophyly priority over narrow one, Ghiselin, M.T., A to species prob-

there exists a replacement term, holophyly, for the latter. lem. Syst. Zool., 23: 536-544.

That is I follow Ashlock in the ofthese terms. 1981. life and why usage Ghiselin, M.T., Categories, thinking.

Behavioral and Brain Sciences, 4: 269-313.

Ghiselin, M.T., 1987. Species concepts, individuality and ob-

jectivity. Biology and Philosophy, 2: 127-143. References 1912. The order. Gidley, J.W., Lagomorpha, an independent

Science, 36: 285-286.

Achterberg, C., van & B. van Aartsen, 1986. The European Gingerich, P.D., 1979. The stratigraphie approach to

Pamphiliidae (Hymenoptera: Symphyta), with special ref- phylogeny reconstruction in paleontology. In:

erence to The Netherlands. Zool. Verh. Leiden, 234: 1-98. J. Cracraft & N. Eldredge (eds.), Phylogenetic analysis

Ashlock, P.D., 1971. Monophyly and associated terms. Syst. and paleontology: 41-77 (Columbia Univ. Press, New

Zool., 20: 63-69. York).

Ax, P., 1985. Stem species and the stem lineage concept. Gingerich, P.D., 1990. Stratophenetics. In: D.E.G. Briggs &

Cladistics, 1: 279-287. P.R. Crowther (eds.), 1990. Palaeobiology. A synthesis: Über Baer, K.E. von, 1828. Entwickelungsgeschichte der 437-442 (Blackwell Scientific Publications, Oxford, etc.).

Thiere. und 1: the foundations of Beobachtung Reflexion, i-xxii, [xxiii- Griffiths, G.C.D., 1974. On biological sys-

xxiv], 1-271, 2 tables, pis. I— III (Gebr. Bornträger, tematics. Acta Biotheor., 23: 85-131.

Königsberg). Hennig, W., 1966. Phylogenese systematics: i-vi, 1-263

Bitsch, J., 1994. The morphological groundplan of (University ofIllinois Press, Urbana).

Hexapoda: a critical review of recent concepts. Annls. Hull, D.L., 1983. Karl Popper and Plato's metaphor. In: N.I.

Soc. ent. Fr., (N.S.) 30: 103-129. Platnick & V.A. Funk (eds.), Advances in cladistics, 2:

Bonde, N., 1984. Primitive features and ontogeny in 177-189 (Columbia Univ. Press, New York).

phylogenetic reconstructions. Vidensk. Meddr. dansk Humphries, C.J. & L.R. Parenti, 1986. Cladistic biogeogra-

naturh. Foren., 145: 219-236. phy. Oxford Monographs on Biogeography, 2: i-xii, 1-98

Briggs, D.E.G. & P.R. Crowther (eds.), 1990. Paleobiology. (Clarendon Press, Oxford).

A synthesis: i-xiii, 1-583 (Blackwell Scientific Publica- ICBN, 1994. International code of botanical nomenclature

tions, Oxford, etc.). (Tokyo Code) adopted by the fifteenth International Bo-

Brooks, D.R. & D.A. McLennan, 1991. Phylogeny, ecology, tanical Congress, Yokohama, August-September 1993.

and behavior. A research in and edited Greuter program comparative biology: Prepared by W. et al.; Regnum

i-xii, 1-434 (Univ. Chicago Press, Chicago & London). vegetabile, 131: i-xviii, 1-389 (Koeltz Scientific Books,

Brothers, D.J., 1975. Phylogeny and classification of the Koenigstein).

aculeate Hymenoptera, with special reference to ICZN, 1985. International code of zoological nomenclature

Mutillidae. Univ. Kansas Sei. Bull., 50: 483-648. (3rd ed.): i-xx, 1-338 (International Trust for Zoological

Condé, B., 1954. Les Diplopodes Pénicillates de l'ambre et Nomenclature in association with British Museum (Natu-

de la faune actuelle. Bull. Soc. zool. France, 79: 74-78. ral History), London & Univ. California Press, Berkeley/

Cope, E.D., 1904. The primary factors of organic evolution: Los Angeles).

i-xvi, 1-547 (Open Court Publ. Co., Chicago). Iordansky, N.N., 1977. Unevenness of the rates of - Contributions to Zoology, 66 (1) 1996 39

microevolution and the key apomorphoses. Priroda [Na- 27: 143-158.

6: 36-46 1982. The of J. ture], Moscow, [in Russian], Nagatomi, A., genera Rhagionidae (Diptera).

Iordansky, N.N., 1979. Fundamentals of the evolutionary nat. Hist., 16: 31-70.

1-190 Press, theory: (Prosveshcheniye Moscow) [in Rus- Nelson, G.J., 1973a. "Monophyly again?" a reply to P.D.

sian], Ashlock. Syst. Zool., 22: 310-312.

Königsmann, E., 1976. Das phylogenetische System der Hy- Nelson, G.J., 1973b. Classification as an expression of

menoptera, 1: Einführung, Grundplanmerkmale, Schwester- phylogenetic relationship. Syst. Zool., 22: 344-359.

und Fossilfunde. Dt. 23: 253-279. 1978. and gruppe ent. Z., (N.F.) Nelson, G.J., Ontogeny, phylogeny, paleontology,

of Krassilov, V.A., 1991. The origin angiosperms; new and the biogenetic law. Syst. Zool., 27: 324-345.

old problems. Trends in Ecology and Evolution, 6: 215— Nelson, G., 1989. Cladistics and evolutionary models

220. Cladistics, 5(3): 275-289.

Kuhn, Th.S., 1970. The structure of scientific revolutions Nevesskaya, L.A., I.A. Gontsharova, L.B. Iljina, N.P.

(2nd ed.): i-xii, 1-210 (Univ. Chicago Press, Chicago/ Paramonova, S.V. Popov, E.V. Babak, K.G. Bagdasarian

London). & A.A. Voronina, 1986. History of the Neogene molluscs

Kukalova-Peck, J., 1978. Origin and evolution of insect in Paratethys. Trudy paleont. Inst. Acad. Sei. USSR, 220:

and their relation to wings metamorphosis as documented 1-208 [in Russian].

by the fossil record. J. Morphol., 156: 53-125. Osborn, H.F., 1934. Aristogenesis: the creative principle in

Kurzanov, S.M., 1987. Avimimidae and the problem of the the origin of species. Amer. Natur., 68: 193-703.

origin of birds. Trans, joint Sov.-Mongol. Paleontol. Platnick, N.I., 1979. Philosophy and the transformation of

Exped., 31: 1-92 [in Russian]. cladistics. Syst. Zool., 28: 537-546.

Labandeira, C.C., B.S. Beall & F.M. Hueber, 1988. Early in- Ponomarenko, A.G. & A.P. Rasnitsyn, 1971. On phenetic and

sect diversification: evidence from a Lower phylogenetic systems. Zool. Zh., 50: 5-14 [in Russian].

bristletail from Québec. Science, 242: 913-916. Popper, K.R., 1959. The logic of scientific discovery: 1-480

Linnaeus, 1751. botanica in C., Philosophia qua explicantur (Hutchinson, London).

botanica definitionibus de fundamenta cum partium, Queiroz, K. & M.J. Donoghue, 1990a. Phylogenetic sys-

exemplis terminorum, observationibus rariorum, adjectis tematics or Nelson's version of cladistics? Cladistics, 6:

figuris aeneis: [i-viii], 1-362, pis. I-XI (G. Kiesewetter, 61-75.

K. de & M.J. 1990b. Stockholm). Queiroz, Donoghue, Phylogenetic sys-

Lubischew, A.A., 1923. On the form of natural system of or- tematics and species revisited. Cladistics, 6: 83-90.

Proc. biol. Res. Inst. Univ., 2(3): 99- ganisms. Permsky Queiroz, K. de & J. Gauthier, 1990. Phytogeny as a central

110 [in Russian; reprinted, also in Russian, in: Lubischew, principle in taxonomy: phylogenetic definition of taxon

Problems of form, A.A., 1982. taxonomy and evolution of names. Syst. Zool., 39: 307-322.

organisms: 24-36 (Nauka Press, Moscow)]. Rasnitsyn, A.P., 1968. New Mesozoic Symphyta (Hymenop-

Lubischew, A.A., 1966. Taxonomy and evolution. In: tera). In: B.B. Rohdendorf (ed.), Jurassic insects of

Intraspecific variability of terrestrial vertebrate animals Karatau: 190-236 (Nauka Press, Moscow) [in Russian],

and microevolution. Trans. All-Union Conf. Sverdlovsk: Rasnitsyn, A.P., 1969. Origin and evolution ofthe Lower Hy-

45-57 [in Russian; reprinted, also in Russian, in: menoptera. Trudy paleont. Inst. Acad. Sei. USSR, 123: 1-

Lubischew, A.A., 1982. Problems of form, taxonomy and 196 [in Russian; published in English in 1979 by Amerind

evolution of organisms: 67-82 (Nauka Press, Moscow)]. Publ. Co., New Delhi],

Mamkaev, Ju.V., 1968. Comparative study of morphological Rasnitsyn, A.P., 1972, On the taxonomie analysis and some

differences in lower and of uniform other taxometric methods. Zh. higher groups an obshch. biol., 33: 60-76 [in

phylogenetic stock. Zh. obshch. biol., 29: 48-55 [in Rus- Russian],

sian]. Rasnitsyn, A.P., 1977. New Paleozoic and Mesozoic insects.

Mayr, E., 1988. Toward a new philosophy of biology: obser- Paleontol. Zh., 1: 64-77 [in Russian, translated into Eng-

vations of an evolutionist: i-xi, 1-564 (Belknap Press, lish in paleontol. J., (1977) 1978, 11: 60-72].

Cambridge/London). Rasnitsyn, A.P., 1980. Origin and evolution of Hymenoptera.

Melville, R.W., 1979. Further proposed amendments to the Trudy paleont. Inst. Acad. Sei. USSR, 174: 1-192 [in Rus-

International code of zoological nomenclature. Bull. zool. sian].

Nom., 36: 11-15. Rasnitsyn, A.P., 1981. A modified paranotal theory of insect

Meyen, S.V., 1987. Fundamentals of paleobotany: 1-432 wing origin. J. Morphol., 168: 331-338.

Hall, (Chapman & London & New York). Rasnitsyn, A.P., 1982. Proposal to regulate the names of taxa

1990. Paleobotanical and nomencla- above the Meyen, S.V., taxonomy family group. Bull. zool. Nom., 39: 200-207.

is In: S.V. Theo- ture: new approach necessary. Meyen, Rasnitsyn, A.P., 1983a. Phylogeny and taxonomy. In: L.Yu.

retical problems of paleobotany: 49-76 (Nauka Press, Zykova & E.N. Panov (eds.), Theoretical issues of the

Moscow) [in Russian], modern biology: 41-49 (A.N. Severtzov Institute of ani-

Mickevich, 1978. Taxonomie mal and M.F., congruence. Syst. zool., evolutionary morphology ecology, Acad. Sei. - 40 A.P. Rasnitsyn Phytogeny, taxonomy, and nomenclature

USSR & Research Computing Centre, Acad. Sei. USSR, cal and laurophyllous floras. Bull. Head botan. Garden,

Pushchino) [in Russian], Acad. Sei. USSR, 82: 43-51 [in Russian],

1983b. Fossil of the 1937. A classification of the nearctic Rasnitsyn, A.P., Hymenoptera super- Ross, H.H., generic

family Pamphilioidea. Paleont. Zh., 2: 54-68 [in Russian; sawflies (Hymenoptera, Symphyta). Illinois biol. Mo-

in 1-174. translated into English Paleont. J., 1983, 17(2): 56-70]. nogr., 15(2):

and Rasnitsyn, A.P. (ed.), 1985. Jurassic insects of Siberia and Saether, O.A., 1979. Underlying synapomorphies anage-

Mongolia. Trudy paleont. Inst. Acad. Sei. USSR, 211: 1— netic analysis. Zoologica Scr., 8: 305-312.

192 (Nauka Press, Moscow) [in Russian]. Saether, O.A., 1986. The myth of objectivity- post-

deviations. 1-13. Rasnitsyn, A.P., 1986. Parataxon and paranomenclature. Hennigian Cladistics, 2:

Paleont. Zh., 3: 11-21 [in Russian; translated into English Schram, F.R., 1980. On a relativistic and quantum mechani-

in Paleont. J., 1986, 20(3): 25-33]. cal approach to evolution. Zh. obshch. Biol., 41: 557-573

Rasnitsyn, A.P., 1987a. Importance of [not] being a cladist. [in Russian].

Sphecos, 14: 23—25. Sharov, A.G., 1971. New flying reptiles from the Mesozoic

Rasnitsyn, A.P., 1987b. Tempo of evolution and the theory of of Kazakhastan and Kirghizia. In: D.V. Obruchev & V.N.

evolution (hypothesis of the adaptive compromise). In: Shimansky (eds.), Current issues of paleontology. Trudy

L.P. Tatarinov & A.P. Rasnitsyn (eds.), Evolution and paleont. Inst. Acad. Sei. USSR, 130: 104-113 [in Rus-

biocenotic crises: 46-64 (Nauka Press, Moscow) [in Rus- sian],

sian], Sharp, D., 1895/1899. Insects, Pt. I, II. The Cambridge Natu-

Rasnitsyn, A.P., 1988a. An outline of evolution of the hy- ral History, 5, 6: i-ix, 83-584; i-xii, 1-626 (Macmillan

London/New menopterous insects. Orient. Insects, 22: 115-145. and Co., York).

Rasnitsyn, A.P., 1988b. Phylogenetics. In: V.V. Menner & Shear, W.A., P.M. Bonamo, J.D. Grierson, W.D.I. Rolfe, E.L.

V.P. Makridin (eds.), Modern paleontology, 1: 480-497 Smith & R.A. Norton, 1984. Early land animals in North

America: evidence from Devonian from (Nedra Press, Moscow) [in Russian]. age

Rasnitsyn, A.P., 1988c. Sepulcidae (Hymenoptera, Cephoi- Gilboa, New York. Science, 224: 492 494.

dea) and origin of Cephidae. Trans. All-Union Entomol. Shishkin, M.A., 1987. Evolution of the ancient Amphibia.

Soc., 70: 68-73 [in Russian]. Trudy paleont. Inst. Acad. Sei. USSR, 147: 1-144 (Nauka

Rasnitsyn, A.P., 1989. Vespida vs. Hymenoptera. Sphecos, Press, Moscow) [in Russian],

18: 8. Sluys, R., 1984. The meaning and implications of genealogi-

Rasnitsyn, A.P., 1990. Order Vespida. In: A.P. Rasnitsyn cal tree diagrams. Z. zool. Syst. Evolut.-forsch., 22: 1-8.

(ed.), Late Mesozoic insects of eastern Transbaicalia. Sluys, R., 1989. Rampant parallelism: an appraisal of the use

Trudy paleont. Inst. Acad. Sei. USSR, 239: 177-205 [in of nonuniversal derived character states in phylogenetic

Russian]. reconstruction. Syst. Zool., 38: 350-370.

Rasnitsyn, A.P., 1991. Typified names of higher taxa once Sundberg, P. & F. Pleijel, 1994. Phylogenetic classification

again. Sphecos, 22: 5-7. and the definition of taxon names. Zoologica Scr., 23: 19-

Rasnitsyn, A.P., 1992a. Principles of phylogenetics and tax- 25.

Zh. evolution of therio- onomy. obshch. biol., 53: 176-185 [in Russian], Tatarinov, L.P., 1976. Morphological

Rasnitsyn, A.P., 1992b. Principles of nomenclature and na- donts and general issues of the phylogenesis: 1-258

ture of the taxon. Zh. obshch. biol., 53: 307-313 [in Rus- (Nauka Press, Moscow) [in Russian].

ofthe ofthe sian], Tatarinov, L.P., 1985. The present state problem

Rasnitsyn, A.P., 1993. New taxa of Sepulcidae. In: A.G. ancestry of mammals. Acta zool. Fennica, 170: 145-147.

Ponomarenko (ed.), Mesozoic insects and ostracods from Tatarinov, L.P., 1987. Essays on the theory of evolution: 1—

Asia. Trudy paleont. Inst., Russian Acad. Sei., 252: 80-99 250 (Nauka Press, Moscow) [in Russian],

(Nauka Press, factor of the increase Moscow) [in Russian]. Tzvelev, N.N., 1993. Hybridization as a

1995. sawflies of the tribe of and the criterion of in Rasnitsyn, A.P., Tertiary Xyelini biodiversity genome genera

(Insecta: Vespida = Hymenoptera: Xyelidae) and their re- higher plants. In: B.A. Yurtsev (ed.), Biological diversity:

lationship to the Mesozoic and modern faunas. Contr. Sei. the approaches to the study and conservation: 193-201

Los Angeles, 450: 1-14. (Russian Academy of Sciences: Zoological Institute, Bo-

Rasnitsyn, A P. & G.M. Dlussky, 1988. Methods and princi- tanical Institute & Russian Botanical Society, Sanct-

ples of reconstruction of phylogeneses. In: A.G. Pono- Peterburg).

marenko (ed.), Cretaceous biocenotic crisis and evolution Wiley, E.O., 1979. An annotated Linnean hierarchy, with

of the insects: 5-15 (Nauka Press, Moscow) [in Russian], comments on natural taxa and competing systems. Syst.

Rautian, A.S., 1988. Paleontology as a source of knowledge Zool., 28: 308-337.

on the patterns and factors of evolution. In: V.V. Menner Wilson, R.W., 1989. Rodent origin. In: C.C. Black & M.R.

& V.P. Makridin (eds.), Modern paleontology, 2: 76-118 Dawson (eds.), Papers on fossil rodents. Sei. Ser. Los An-

(Nedra Press, Moscow) [in Russian]. geles Mus., 33: 3-6.

1971. On the and of the 1992. The of the terrestrial verte- Razumovsky, S.M., origin age tropi- Vorobyeva, E.I., problem Contributions to Zoology, 66 (1) - 1996 41

brate origin: 1-344 (Nauka Press, Moscow) [in Russian]. and Chelicerata). Trudy paleont. Inst. Acad. Sei. USSR,

Zavarzin, G.A., 1987. Pattern of the procaryote evolution. In: 165: 1-200 (Nauka Press, Moscow) [in Russian].

L.P. Tatarinov & A.P. Rasnitsyn (eds.), Evolution and the Zherikhin, V.V., 1993. History of tropical rain forest biome.

biocenotic crises: 144-158 (Nauka Press, Moscow) [in Zh. obshch. biol., 54: 659-666 [in Russian].

Russian],

Zherikhin, V.V., 1978. Development and changes of the Cre-

taceous and Cenozoic faunistic assemblages (Tracheata Final draft received 10 July 1995