Thalassas, 27 (2): 193-224 An International Journal of Marine Sciences

FROM “TREE-THINKING” TO “CYCLE-THINKING”: ONTOGENETIC SYSTEMATICS OF MOLLUSCS

ALEXANDER V. MARTYNOV(1)

Key words: ontogeny, evolution, phylogenetics, ontogenetic systematics, synthesis, ontogenetic cycles, heterochrony, paedomorphosis, evolutionary models, molluscs, , nudibranchia,

ABSTRACT degree independently developing major biological fields, i.e. ontogeny, evolution and , the new During the last decade traditional morphological conception of ontogenetic systematics is therefore paradigm of evolutionary biology has been suggested; the practical usefulness of the new concept challenged. Molecular systematics and morphology- is illustrated by some examples from nudibranch based phylogenetics were considered as “advanced” molluscs. Such re-formulation of apparently well- fields compared to the “old-fashioned” traditional known and obvious biological knowledge implies systematics. At the same time, an enormous body of also a great challenge for current phylogenetics and the practical and theoretical methods of “traditional” systematics: an understanding of the necessity to biology was considered usually in a minimal degree. consider not only evolutionary “lines” and “branches” It is here demonstrated that the current evolutionary of the “Tree of Life”, but also its cycle nature, since paradigm in the “phylogenetic era” lacks a theory of ontogenetic cycles are indispensable and active parts how organisms change their shape. The links between of the process of evolution. evolution, ontogeny, systematics and phylogenetics are prima facie obvious, but similarly greatly FROM SYSTEMATICS TO EVOLUTION AND underestimated currently, though the field of “evo- VICE VERSA: THE CURRENT PARADIGM devo” is continuously growing. As a synthesis (or more exactly, re-synthesis) of the still in considerable After collapse of the classical biological science following by two world wars, traditional systematics underwent rapid evolution leading to dramatic changes in this field. Major agenda of the unprecedented challenge for the systematics (1) Zoological Museum, Moscow State University, Bolshaya become the “search for objectivity”. Following this Nikitskaya str. 6, Moscow 125009 Russia. way several schools pretending on an objective E-mail: [email protected] systematics emerged consequently. Phenetics refused

193 ALEXANDER V. MARTYNOV

classical systematic hierarchy in favour of “direct” phylogenetic systematics – reconstructing/showing the measurement of a dimension between characters historical sequence of the morphological apomorphies (Sokal & Sneath, 1963). Initial enthusiasms of was soon almost completely disregarded by some phenetists soon was declined and replaced by the authors (e.g. Scotland et al., 2003). Though few authors next paradigm – phylogenetic systematics (Henning, attempted to advocate importance of morphology 1966). Phylogenetic systematics became the most itself as dataset for reconstructing phylogeny prolific and successful branch of the “objectivity- (e.g. Wiens, 2004), and though a so-called “bio- driven” taxonomy, first of all because it was actually ontological” approach has been developed (see e.g. a stricter reformulation of the classical heritage Ramírez et al., 2007; Vogt, 2009), modern systematics of Darwin’s (1859) and Haeckel’s (1866) texts. soon was transformed into “phylogenies without I.e. phylogenetic systematics, exactly in Darwinian synapomorphies” (see Mooi & Gill, 2010). sense, converted traditional taxonomic hierarchy into a consequence of the evolutionary events. Heated Even as promisingly claimed often recently, that debates around theoretical foundations of taxonomy a reasonable study should include both morphology featured 1970S and 1980S and caused formation of and molecules (see e.g. major topics for the the modern paradigm (see e.g. Wiley, 1981; Ridley, planning in October 2011 a large-scale international 1986; Pesenko, 1989; Wägele, 2005; and many conference Deep Metazoan Phylogeny 2011 at others). Very shortly outlined, this paradigm implies Ludwig-Maximilians-Universität München), this tree-thinking instead of taxon-thinking. does not prevent for transforming the until quite recently extremely fashionable morphology-based In other words, there are no more “statical” taxa phylogenetic systematics into an auxiliary discipline, of classical systematic hierarchy, but instead any in favour of the molecule-based “fourth great school (monophyletic) taxon should be regarded as a part of of systematics” – the “It-Doesn’t-Matter-Very-Much an endless “Tree of Life”. To produce this, traditional school” (see Felsenstein, 2004, p. 145). Finally, most hierarchy should be “ripped” and separate characters of the recent researchers will eagerly answer “yes” afterward need to be “extracted” and reanalyzed on a question whether or not “A new and general under various methods (e.g. Wiley et al., 1991). Then, theory of molecular systematics emerged” (Edwards, a new, apparently objective hierarchy of characters 2009) (see e.g. any content of any journal where the (and taxa) is expected to appear. The breakthrough word “evolution” is mentioned; few examples could in computer technologies almost not leaved a space be cited – e.g. Dunn et al., 2008; Lartillot, Philippe, for critics of such an approach and “new, easy, 2008). Substantial revision of traditional systems (e.g. and objective” phylogenetics celebrated their victory Halanych, 2004), apparently caused by the new field over all 1990S and earlier 2000S. Both modern and emergence, forces even traditional taxonomists and classical zoological periodicals then rapidly showed evolutionary morphologists, especially (still) well- thousands of trees of strict dichotomic forms. established in Russia, for uncritical acceptance of the results of molecular phylogenetics and claims for a The major target of phylogenetic systematics – “revolution in systematics” (see e.g. Shatalkin, 2005). reconstruction of a definite “Tree of Life” apparently The general current enthusiasm about molecular was so close, but real results soon demonstrated a phylogenetics is so overwhelming (even despite on completely different picture. A principal theme of most numerous problematic issues – see e.g. Philippe et of the phylogenetic studies instead became “absence of al., 2011a), that any opposite or just more balanced resolution of phylogenetic trees” (see almost any paper views could be easily interpreted as an “old-fashion” that contains morphological phylogenetic analyses, adherence of an author with old ideas. Nevertheless, e.g. Dayrat, Gosliner, 2005). Major advantages of the one of such attempts will be performed here.

194 FROM “TREE-THINKING” TO “CYCLE-THINKING”: ONTOGENETIC SYSTEMATICS OF NUDIBRANCH MOLLUSCS

Considering the above outlined few key-points of the famous Ecdysozoa vs. Coelomata controversy: the “modern” systematics development it is possible currently phylogenetists claim that the new data clearly to conclude that several generations of “objective” in favour of the former (e.g. Holton, Pisani, 2010), but approaches have substituted each other during numerous previous contradicting studies most likely the last 50 years. The current one is molecular suggested that is not a final conclusion. In turn, newest phylogenetics, and it actually includes elements of nudibranch phylogenies based, for the first time, on a all previously “eaten by each other” approaches: substantial taxon sampling, resulted, however, mostly analysis of numerous separate characters apart from in a conclusion, that “our results do not resolve all any hierarchy concept of phenetics and the cladistic the relationships within the ” (see Pola approach of Hennigian phylogenetic systematics & Gosliner, 2010, p. 931). In the meanwhile, on two that, in contrast, is based on the at least idealistic main problems of the molecular phylogenetics — assumption of homology of characters. incompleteness and absence of a definite resolution have been pointed soon after the new field became Many shortcomings of the traditional systematics a mainstream (see, e.g. Jenner, 2000). It is quite leave of course a possibility for some kind of clear that obtaining of the large molecular data sets “misuse” of its methods, leading e.g. to so called and developing of the sophisticated molecular tools splitting and lumping, arbitrary- and authority-driven is of much benefit for the independent checking of classificatory schemes, a potential possibility for the evolutionary-morphological hypotheses. However, almost “endless” split a higher taxon into subtaxa, there are still so scarce data in favour of the true etc., and partially provoked the XX-century race objectivity of such approach. Some remarkable for “objective systematics”. But is the molecular examples of controversies, e.g. on the phylogenetic phylogenetics exactly the long-expected substitution position of with extremely simple morphology, for the “fabulously arbitrary” classical systematics? like Acoela, where molecular phylogenetic methods As it was admitted by molecular phylogenetists prima facie unquestionably superior over traditional themselves, “Phylogenetics is still a difficult and approaches, revealed however, that the range of controversial field, because no foolproof method is discussions even in the newest publications (i.e. basal yet available to avoid systematic errors” (Lartillot vs. derived simplified position of acoelomorph – & Philippe, 2008, p. 1469). Special bioinformatics Mwinyi et al., 2010 vs. Philippe et al., 2011b) is literature is full of debates of the usefulness of comparable with the past debates in the evolutionary the different mathematical models of the nucleotide morphology field (e.g. Ivanov, Mamkaev, 1973). And substitutions and alignment algorithms. All after 15 years of active developing of sophisticated the enormously branched field of the molecular programs for molecular phylogeny reconstructions, systematics is constructed on the probabilistic professional taxonomists recommend “use biology, not principles that are not “hided”, but instead, widely algorithms to make homology decisions” (see Mooi & advertised (see Nei, Kumar, 2000; Felsenstein, 2004; Gill, 2010, p. 26). Lukashov, 2009; and many others). Thus, as a result of more than 40 years of However, even the phylogenetic theorists theoretical discussions and about 20 years of practical and bioinformatics specialists are in doubt while applications, the field of phylogenetic systematics, attempting to distinguish “truth” and “statistics” in the while performing their great quest for objectivity, molecular phylogenetics (see e.g. Wägele et al., 2009; have provided us with: 1. New arbitrariness, 2. Novel Kumar, 2010; Philippe et al., 2011a). The exponential preconceptions, 3. Absence of the strict resolution, 4. growth of the new publications in this field already Doubts in monophyly of every taxa. That was therefore has led to the hardly avoidable contradictions. E.g., almost entirely missed in the apparently exhaustive

195 ALEXANDER V. MARTYNOV

Figure 1: New model of the dorid evolution. Ancestral notaspid and dorid ontogenetic cycles (external features). Elaboration (prolongation) of the notaspid ontogeny led to appearing of the complicated cryptobranch ontogenetic cycle, whereas it further heterochronic modifications caused appearing of the variously juvenilized phanerobranch dorid families up to strongly paedomorphic group. The prognostic features of this model imply presence of the cryptobranch postlarval stages in the newly discovered Onchimira cavifera, despite that real ontogeny of this taxon is completely unknown yet. Drawing by T.A. Korshunova based on A.V. Martynov originals. theoretical discussions, which are included perhaps Despite of the methods of traditional systematics thousands of papers and hundreds of monographs, have slightly changed since the Linnean time, they still and that have led to the current situation? The answer successfully allow to describe these independently- is very simple, quite disappointing, and somewhat existed phenomena – taxa of the systematic hierarchy. contradicting even with the first lines of the present Very important for the aims of the present paper paper. This almost entirely forgotten field is the is the fact that traditional systematics still acted traditional systematics. This, quite humble and clearly independently from any phylogenetic study itself endangered (even the recent great rising of “Zootaxa” (does not matter, morphological or molecular), despite could not completely prevent traditional systematics on repeated claims of the tree-priority over the taxa from the label of “old-fashioned”) in the modern time, themselves. In a most paradoxical way, the traditional scientific tribe of professional taxonomists almost taxonomy in the second part of XX century became 300 years already has performed their really great a real “keeper” of true objectivity in the study of task – finding a way how to describe independently organism diversity. No traditional systematist will from our consciousness existing patterns of the ever place a dandelion into a phylum Arthropoda and organisms in a most consistent manner. to the family of the harvest mites (Trombiculidae),

196 FROM “TREE-THINKING” TO “CYCLE-THINKING”: ONTOGENETIC SYSTEMATICS OF NUDIBRANCH MOLLUSCS

but instead into a division of the flowering plants the traditional systematics. Nevertheless, a system (Angiospermae or Magnoliophyta) and to the family generated of the knowledge of traditional systematists, Asteracea. In a similar way, it is hardly expected, that even methodologically almost lacking any notions on a bivalve molluscs will be considered by traditional the evolutionary process, comes much more close systematist as a polychaet, or that a nudibranch to a yet not existing “ideal” of the biologically- mollusk will be named as a cephalaspid. based theory of the historical transformations of the organisms, than any modern taxa and ranking-free Thus, prima facie absolutely combinative and statistics-aided study of nucleotide substitutions in the “typological” method of the taxa formation actually molecular phylogenetic field. This is because they are have led to discovery of the independent from a able to distinguish and define (name) any organisms particular observer natural phenomena — really structurally and use the hierarchical principle. existing groups of organisms, united by a set of unique characters. In any country, a student, professional There are no doubts though, that structural expert or amateur unequivocally and independently approach itself has many restrictions – that were a will confirm existence of such groups, e.g. bivalve driving force for the past and present challenges for molluscs, polychaet worms or mammals. The the theory of systematics. Traditional systematics independent multiple confirmations is a base for any offers mighty methods for exactly recognizing scientific knowledge and in this respect, paradoxically, organismal patterns, but almost do not consider traditional systematics comparable with exact science. causes of emerging of such patterns, i.e. evolution. In other words, traditional systematists in course of Therefore, finding, for instance, some aberrant taxa almost 300-years of the taxonomy development have within a well-established taxonomical group (e.g. learned perfectly how to structurally attributed any, opisthobranch order Doridacea), which are, similar previously unknown specimen to a known taxon of the to the apparently ancestral (for dorids) systematic hierarchy. by gill patterns (i.e. dorid family Corambidae), traditional systematics may relatively easily be And if a taxonomist will find a mollusk misled in assessing their relationship. However it with crawling sole, radula and also, with a spiral will be too simple and premature labeled the classical protoconch, but with a two-valved shell, he will systematics as a non-evolutionary and “typological” definitely consider this unusual taxon as an aberrant (e.g. Mayr, 1963; Hennig, 1966). Because, as it is member of the class , not as an aberrant sometimes eluding from the current thinking, one member of the class Bivalvia. Moreover, several of the major evidence for the evolution itself became fine, complex “enough” morphological features exactly existence of the systematical hierarchy of the two-valved snails, including radula and (Darwin, 1859). In another word, the properties of reproductive system unambiguously point to the the traditional systematics allow for relatively easy subclass Opisthobranchia, and even more exactly, to converting it into a sequence of the evolutionary the order Sacoglossa. Understanding that Julia and event (Haeckel, 1866). And though it sounded so Berthellinia are the members of the class Gastropoda sensational in the 1970-80S, Hennigian phylogenetic and subclass Opisthobranchia at the level of soft body systematics was actually a quite obvious and much morphology (Kawaguti & Baba, 1959) then helps postponed almost direct implication of the Darwin’s to paleontologists recognize their initial mistake and Haeckel’s key texts. in assessing of the fossil shells of the two-valved sacoglossans to the class Bivalvia. Is it an ideal of a Thus, we have an obvious, but still not solved scientific knowledge? Professional taxonomists are in a most consistent way, contradiction: traditional of course well aware about numerous “dangers” of systematics very well works with the recognition

197 ALEXANDER V. MARTYNOV

of unique taxa, but almost does not include an already quite evident for the Ancient Egyptians evolutionary approach. Whereas phylogenetic (see Hennig, 1966), and became one of the most systematics (including molecular), instead, built basic elements of the continuously growing modern on strict evolutionary principles, but considerably field of the evolutionary developmental biology disregard the fact that any living thing exists not only (commonly known as “evo-devo”). Few essentially as a “ line” or “branch”, but forms also particular, similar quotations may be included that “Morphology separate organisms, quite “closed” systems called considers organism not only in the adult condition, , which are able to exist millions of years just but also in all preceding stages of their development” slightly changing. Moreover, all organization of the by classic of the comparative and evolutionary living organisms demonstrates a hardly disputable morphology, Carl Gegenbaur (1859, S. 1) and “Any conservatism: most of the metazoan phyla have been organism does not possess an ontogenetic cycle, traced already in Cambrian, that means therefore, that but instead part of it” by modern biologist John not only the morphological, but also physiological, Bonner (1965). In this respect, special emphasis on molecular and ontogenetic basis of any modern the relation between ontogeny and evolution was put animals have not changed significantly in course of in Russian science by comparative and evolutionary the last half of the billion years! morphologist Vladimir Beklemishev (1925, 1969), who featured his understanding of any organism Is it possible to suggest some set of theoretical not as a static structure like morphologists and and practical instruments that may challenge current systematists, but instead as the morphoprocess, e.g. paradigm that can be shortly named as “tree- continuous flow of various processes, and first of all, thinking”? Because the rapidly growing field of the ontogenetic cycle itself (see e.g. Korotkova, 1979; “phylogeny reconstruction” is almost not interested Ivanova-Kazas, 1995 and many others). Beklemishev in the fundamentals, on which any evolutionary study specially highlighted that exactly the life cycle of any is based, and the molecular phylogenetics is not an species is the unit of the comparative morphology exception. Is it therefore possible, that suggestions (see Beklemishev, 1969, introduction). Benedictus how to link the “immobile” traditional systematics Danser (1950, p. 142), Dutch plant taxonomist, quite and idea of the historical transformation of the taxa unequivocally formulated this principle for the field (i.e. evolution) in most consistent and natural way, of systematics: “The life-cycle with its multiformity from “point of view” of the real biological processes, is the smallest unit of classification”. and not the statistical probability calculation, already have emerged in the history of biology, but a working In this respect, major deficiency of the current method based on such ideas was not formed? And phylogenetics (in any from) in “low resolution” of taxa some modern reformulation, of apparently already relationship appears not due of the not enough degree quite well established ideas is therefore required? of the advancements of algorithms and software, but in fundamentally incorrect basic concept regarding ONTOGENETIC CYCLE — THE UNIT OF the most important biologically-based properties of SYSTEMATICS AND EVOLUTION the organisms – the ontogenetic cycle. As it was already mentioned above, modern phylogenetics it is Perhaps among one of such apparently obvious a somewhat hybrid between evolution-free phenetics and self-evident ideas, that is simultaneously and, strictly evolutionary phylogenetic systematics. underestimated by both traditional systematics Both books of Sokal & Sneath (1963) and Hennig and modern phylogenetics, is the principle that (1966) thus became an indispensable basis of the any organism can be understood only as a part of modern phylogenetics. And in both monographs a particular ontogenetic (or life) cycle. This was properties of the ontogenetic cycles were either

198 FROM “TREE-THINKING” TO “CYCLE-THINKING”: ONTOGENETIC SYSTEMATICS OF NUDIBRANCH MOLLUSCS

not considered at all (Sokal & Sneath, 1963) or 1866). Nevertheless, we still should thank Haeckel not specifically interpreted (Hennig, 1966). Already on only for one of the most consequential defending of the first pages of Hennig’s fundamental publication, the evolutionary theory, but also for the formulation the entire ontogenetic cycle was actually “ripped” into of the terms ontogeny and phylogeny, facts which independent stages of the “semaphoronts” (Hennig, are most likely well remembered only by a handful 1966), “character bearers”, and the major problems to of modern molecular phylogenetists. It is quite hard assign different developmental stages to one species, to understand now, but in the second part of the XIX without direct observations, was highlighted. Such and first half of the XX centuries ontogeny was an methodological approach was important for Hennig important and actually integral part of many of the in order to find stricter bases of his theory. However, systematic and evolutionary studies. The famous widely cited and too superficially and incorrectly Haeckel’s agenda “ontogeny recapitulates phylogeny” interpreted, most likely, they opened a way for further has received controversial acceptance, and there were consideration not only taxonomy, but also ontogenetic many repeated attempts to challenge the biogenetic cycle itself as a predominantly combinative fields, law (e.g. Sedgwick, 1909), but a comprehensive theory lacking their own strict patterns. The Sokal & Sneath of the interaction of the ontogeny and evolution for the “Numerical Taxonomy” (1963), apparently built first time was suggested by the Russian morphologist their theoretical apparatus on a completely different and embryologist Alexei Severtsov (also spelled ground than Hennig, and just not considered almost as Sewertzoff) (1912, 1931). Ten years prior to the any biological properties (including ontogenetic well-known paper of Garstang (1922, p. 81, claimed cycle) at all, but instead formulated an “operational that “no one has presented [until now] a complete taxonomical unit” (OTU), as super-formal ground for theoretical scheme capable of replacing Haeckel’s as the taxa construction. However, these both so different an explanation of the relations between ontogeny and books have produced in total very similar effect, phylogeny”), Severtsov exactly and unequivocally which can be most well described as “founder effect”: suggested such a theory that finally allowed to the key importance of the ontogenetic cycle was reformulate the biogenetic law: not phylogeny is the almost vanished from entire field of the phylogenetics source for ontogeny, but instead ontogeny creates and systematics. And a modern important review on evolution (Severtsov, 1912; Garstang, 1922; Levit et the methods of the phylogeny reconstruction easily al., 2004, p. 349–353). avoided even one mention (!) of the fundamental biological term – ontogeny (see Felsenstein, 2004), What is the important difference in such thus perfectly recapitulates one of their own most re-formulation? The key-importance for the entire important ancestor – the Sokal & Sneath’s book evolutionary field is due to the very clear understanding “Numerical Taxonomy”. that not some obscure historical processes along the endless branches of the Tree of Life are responsible The importance of the ontogeny for the for the evolution, but instead, the routine “miracle” understanding of evolution has not disappeared from of each new ontogeny formation and their slight or the scientific publications completely, but rather pronounced modifications solely feature the evolution transformed into a special field. The modern founder itself. This new understanding of Haeckel’s law of the field is of course famous German biologist opened new horizons in evolutionary studies and have Ernst Haeckel and his immediate predecessor, Fritz lead to formation of a particular field of the studies Müller. The latter suggested basic principles of the of the interaction between ontogeny in evolution. ontogenetic and evolutionary interactions (Müller, For many decades, this field was developed quite 1865) prior to the Haeckel’s first monumental independently in Russia (USSR) and in USA. In both monograph “Generelle Morphologie” (Haeckel, countries independent sets of theoretical instruments

199 ALEXANDER V. MARTYNOV

Figure 2: Comparision of the morphological patterns of opisthobranchs molluscs of the group Notaspidea, ancestral to all dorids (gill, lag and anal opening, an are ventro-lateral, rhinophores, rh — without pockets, their bases united) with adult (gills, dg and anal opening, an are dorso-terminal, rhinophores, rh have pockets and diverged) and postlarval (gills absent, anal opening ventro-terminal, rhinophores without pockets and united) morphology of the cryptobranch dorids. A, B, C — Berthellina citrina (Ruppell et Leuckart, 1828) (Notaspidea), adult specimen 23 mm length; D — Сadlina laevis (L., 1767) (Doridacea), adult specimen 25 mm length; E — C. laevis, postlarval stage ca. 500 μm length; F — C. laevis, preceding early postlarval stage ca. 400 μm length. Scale bars: B, C — 1 mm; E, F — 100 μm (on lower inset, E — 30 μm). Photos: O.V. Savinkin (A), T.A. Korshunova (D). SEM micrographs (B–F): A.V. Martynov. Further abbreviations: fn — frontal part of notum; lag — lateral gill; no — notum; ppl — primary posterior notal lobes.

200 FROM “TREE-THINKING” TO “CYCLE-THINKING”: ONTOGENETIC SYSTEMATICS OF NUDIBRANCH MOLLUSCS

for the description of the phylogenetic effects of the biogenetic law and Severtsov’s phylembriogenesis ontogenetic shifts have been established. In Russia it theory with genetic addition. The rise of “evo-devo” became the phylembriogenesis concept of Severtsov was mostly coincided with the declining of the previous (1912; 1934) and his successors (e.g., Schmalhausen, ontogenetic-evolutionary concepts in both, Severtsov’s 1938, 1969 and many others). In USA, instead and Gould’s version. However, seeking for a new wide- dominated attempts of re-formulating concepts of scope biological theory and pretending to describe Haeckel’s initially auxiliary term heterochrony ontogenetic mechanisms (i.e. organisms’ shape (De Beer, 1930; Gould, 1977; Alberch et al., 1979; transformation), evolutionary developmental biology McNamara, 1997 and many others). Both schools have almost “forgot” about true indispensable “data base”, used basically the same principle, i.e. the “ontogeny almost three hundred years have accumulating the defines evolution”, but make different accents and information about the non-random organism diversity use somewhat different methodological tools. For patterns – traditional systematics, that deals not with instance, in the Russian theory of phylembryogenesis, some “theoretical” but instead with most really practical researchers persisted in using a quite “rigid” scheme, organism shape patterns. But the most responsible field describing the alteration of the ontogeny leads to for such “ontogenetic oblivion” is of course modern the evolutionary effects only at three levels: earlier phylogenetics, as a discpline that claimed repeatedly to ontogenetic changes (archallaxis), middle ontogenetic be the strict substitution of the traditional systematics. stages (deviation) and, finally, most recent additions In the 1970-80S phylogenetically-orientated theorists to a given ontogeny (anaboly). The main idea referred virtually re-discovered the biogenetic law theme (see to the wholeness of organisms and their ontogenetic Fink, 1982; Kluge, 1985; De Queiroz, 1985; Mishler, cycles (e.g. Schmalhausen, 1938). 1988; Weston, 1988; Bryant, 1991; and others), but finally mostly concluded ontogenetic studies are On the contrary, the approach of dissociability just of auxiliary importance for the phylogenetic (Needham, 1933) featured the Gould (1977) version of inference, except for few examples, which not became the theory of interaction of the ontogeny and evolution a mainstream (e.g. Nelson, 1978). Even few relatively and allowed to consider any organism as rather recent attempts to refresh interactions between mosaic where different parts can develop relatively phylogenetics and ontogeny have met unexpected separate from each other. The dissociability concept difficulties and such studies are still far from being also allowed to construct the much more flexible widely accepted (e.g. Jaecks & Carlson, 2001; Wiens theory of heterochronic shifts of ontogeny, either et al., 2005; etc.), though importance of the ontogenetic regressive (e.g. pedomorphosis) or “progressive” approach is not completely vanished from the most (=peramorphosis) (see e.g. McNamara, 1986, 1997). recent phylogenetic publications (e.g. Struck, 2007; Box et al., 2008; Smirthwaite et al., 2009; Ji et al., An almost one century gap between genetic and 2009). phylogenetic approaches in biology (when Wilgelm Roux has proclaimed his “developmental mechanics” ONTOGENETIC SYSTEMATICS instead of the Haeckel’s biogenetic law) was filled only in 1980S, when the apparently completely new Thus, until recently both branches of the biogenetic field currently known under name of “evo-devo” has law developing have not led to the appearing of a emerged (see e.g. Raff & Kaufmann, 1983; Hall, 1992, general theory of the organism shape changing. The 1999, 2003; Gilbert et al., 1996; Minelli, 2003; Carroll, apparently promising synthesis, “evo-devo” might 2008; and many others) as a re-union of the ontogeny have become such a result of the post-Haeckelian and evolution. In reality, “evo-devo” should be rather re-thinking of the ontogeny and evolution interactions considered as a modern re-formulation of the Haeckel that could have integrated major branches of biology.

201 ALEXANDER V. MARTYNOV

However, “evo-devo” remains largely a discipline ontogenetic systematics (Martynov, 2010). Despite applying modern technology on studying the ontogeny of the apparent easiness of creation of this term and of organisms. In favour of such view quite clearly deep sense, uniting both ontogeny and evolution, as points the most recently suggested term “phylo-evo- far as I know, there was only one single previous devo” (see Minelli, 2009), which was suggested as an attempt to use it (see Albert et. al., 1998). However, important novelty in one of the leading evolutionary regardless of using identical words, the latter journal. However, since ontogeny already implies publication understood under the term ontogenetic evolution and, moreover, “evo-devo” exactly means systematics not systematics itself, but instead, “evolutionary development biology”, this additional somewhat paradoxically, a molecular pathways that emphasizing on “phylo-” also makes no sense because can lead to formation of a structure during the phylogeny itself means modifications of the ontogeny ontogeny. The term ontogenetic systematics was in the historical scale. In other words, this fact of suggested independently in the mentioned author’s interacting ontogeny and evolution that is known publications. However, these two both apparently since Severtsov in 1912 is now eluding from the very different meanings of the term ontogenetic enormous recent “evo-devo” field, and the principle systematics point to this highly desirable, but yet that “ontogeny builds evolution” continues to be very far from completeness general theory of the re-discovered in modern times. organisms’ shape transformation, of course implies understand of the way how the DNA information Thus, despite on almost 150 years of the evident became the macromorphological characters. And it scientific history of the interactions of ontogenetic is a widespread modern illusion that such apparently and evolutionary studies, thousands of publications grandiose task the “evo-devo” field can performed and hundreds of various terms (most of them are alone. Instead, an approximation to such task has perfectly dead now), an important synthesis and a already been made, during almost 300 years, while term connected with is still missing in this field. the traditional systematics was in searching for a I.e., despite of the very complicated history and system how in a most unequivocal way to describe promising modern researches outlined above, organisms’ diversity. By-products of such searching current evolutionary developmental biology, classical became discovery of a real and not imaginable systematics (the science, which maximally possesses and methodologically very useful taxa hierarchy, the information on the diversity and hierarchical and then, as a direct consequence of the latter, the patterns of the morphological characters of the discovery of the evolutionary process itself. organisms) and phylogenetics have all developed mostly separate from each other. Their importance Most generally, ontogenetic systematics may be for each other is self-evident but still greatly regarded as a starting point for constructing a general underestimated in modern biology (e.g. there are no theory of the evolution of ontogenies, which largely any mentioning of the systematics on the schemes proceeded as the evolution of a limited set of basic explaining “evo-devo” synthesis – see e.g. Love, ontogenetic cycles of metazoans that originated as Raff, 2003; Olsson et al., 2010). Therefore, their new early as in the Precambrian or Early Cambrian synthesis (“re-synthesis”) might become an important and correspond to different phyla in the classic source for further understanding of the historical systematics. Such re-formulation of apparently well- succession of the organisms’ shape changing, i.e. known and obvious biological knowledge implies evolution in its original sense. also a great challenge for the current phylogenetics and systematics: an understanding of necessity of A new term that features and highlights the consideration not only evolutionary “lines” and such re-synthesis has been suggested recently as “branches” of the “Tree of Life”, but also a cycle,

202 FROM “TREE-THINKING” TO “CYCLE-THINKING”: ONTOGENETIC SYSTEMATICS OF NUDIBRANCH MOLLUSCS

since the ontogenetic cycle is an indispensable and become a true and not only a theoretical unit of active part of the process of evolution. I.e., it implies a systematics and phylogenetics. gradual shifting from current merely “tree-thinking” to what could be named as the “cycle-thinking”. Such problematic issues are perfectly highlighted by the very practical taxonomy and phylogenetics At the same time, such shift does not implies in one of the major nudibranch groups — dorids consideration of evolution as a strictly cyclic process, (Doridacea or Anthobranchia), when some aberrant the view that we also should thank to the Jena scientist, features of the dorid nudibranch family Corambidae, Ernst Haeckel. He suggested the phases of “epacme” such ventral anus and gills have been considered as (primary diversification), “acme” (“flourishing”) and archaic (basal), and therefore the entire group have “paracme” (declining) (see Haeckel, 1866, Bd. 2, S. been placed into the beginning of the nudibranch 320–322), that apparently every taxon “has passed” in classification (e.g. Odhner in Franc, 1968), and, what course of the historical development. The Haeckel’s is more notable, persisted even in recent reviews “thee-phases” evolution was then recapitulated in (e.g. Rudman, 1998). However, as it was already many of the second part of XIX and earlier XX evidently shown, also by careful cladistic analysis of centuries considerations; most notable became the morphological characters (Martynov & Schrödl, Otto Schindewolf’s (1936, 1950) “typogenesis”, in press), corambids are “just” secondary regressive “typostasis” and “typolysis”. However, it is quite descendants of the common dorid ontogenetic cycle, obvious, that despite of many taxa have been already originated due to the progenesis (see Martynov, 1994b; completely extinct, most of the basic, for instance Martynov et al., 2011; Martynov, Schrödl, in press). Metazoa structural patterns, i.e. phyla, originated The consideration of the entire ontogenetic cycle thus yet in the earlier Cambrian, and still exist in the helped to correctly recognize the homologies of gills modern biosphere. It is prevented from such simplistic and gill cavities and thus contributed to a character set representation (though, of course apparently “natural” with better signal for cladistic study. Few non-trivial in analogy with the ontogeny itself) of the taxa phylogenetic studies have already come to essentially history in the three-phase model, but do not refused similar conclusions, most importantly, including the the possibility to consider the evolution as result molecular data (see Wiens et al., 2005). It can be of the various ontogenetic shifts (e.g., by the well only guessed how many thousands or hundred of established process of heterochrony). What is very thousands particular taxa incorrect evaluation and important in such consideration is the possibility phylogeny reconstructions will be revealed exactly to predict similar, either regressive (paedomorphic) due to forgetting of the ontogenetic cycle properties. or progressive (additive) ontogenetic shifts within various descendant taxa of the same ancestral taxon, Some important (but of course not absolutely thus quite easily explaining the phenomenon of exhaustive), both theoretical and practical principles parallel evolution, especially in relatively closely of the ontogenetic systematics are listed below: related taxa. All modern phylogenetics is full of claims for the paraphyly cases, but few modern researches 1. The ontogeny defines evolution, and not vice understand that mostly responsible for such pattern versa. Therefore, — ontogeny is basis for the are exactly the indisputable cyclic properties of the phylogeny. ontogeny itself. Returning to the entire evolutionary 2. Therefore, the unit of the evolution, systematics field, considering ontogenetic cycles as a major and and phylogenetics — is the ontogenetic cycle and not just auxiliary principle thus much more increases not just a “clade”. the reliability of both morphological and molecular 3. The primary classificatory procedure is not phylogenetics. The ontogenetic cycle thus should connected directly with the historical or individual

203 ALEXANDER V. MARTYNOV

Figure 3: Heterogeneity in construction of the rhinophoral and oral apparatus in nudibranch molluscs (Nudibranchia) and in their ancestral group — opisthobranch molluscs Notaspidea (=Pleurobranchoidea). A — Berthellina citrina (Notaspidea: ), adult specimen 23 mm length; B — Cadlina laevis (Doridacea), postlarval specimen ca. 400 μm length; C — Cadlina laevis, adult specimen 15 mm length; D — postlarval specimen Cadlina laevis ca 500 μm length; E — brocki Bergh, 1897 (Notaspidea: ), adult specimen; F — Tritonia antarctica Martens et Pfeffer, 1886 (Nudibranchia), adult specimen. Scale bars: A, C, E — 1 mm; B, D — 30 μm; F — 300 μm. SEM micrographs: A.V. Martynov. Abbreviations: fn — frontal part of notum; ov — oral veil; rc — rninophoral cavity; rh — rhinophores; rp — rhinophoral pocket; rs — larval suture of the rinophoral pockets closing.

succession of the origin of the classification’s always incorporated (attributed) into already objects. A systematist attributed any just known higher taxa (species to , genus to discovered new specimen on the basis of sets of family, etc. etc.) on the basis of the sets of unique unique characters, evidently homologous within characters. Such principle could be also named as any given stable (well-established) taxonomic a main principle of the systematics. group. With only help of the classificatory scheme, 4. The evolutionary succession of the organisms’ any yet unknown specimen could unambiguously shape transformation — it is the succession of the

204 FROM “TREE-THINKING” TO “CYCLE-THINKING”: ONTOGENETIC SYSTEMATICS OF NUDIBRANCH MOLLUSCS

formation and “disruption” of various ontogenetic memory of the heredity system), in even greatly cycles. From the spawn mass of a Littorina we modified secondarily ontogenetic cycles (e.g. a expected to see a Littorina juvenile or larva modified nauplius larva in the rhizocephalous emerging, not a nudibranch veliger or a crustacean crustacea allows to infer that they still belong to nauplius. However, the exact shape of the newly the in other features completely different group formed molluscs or crustacean (e.g. fine details Cirripedia). I.e. there is a possibility to infer of shell or legs proportions) is hardly predictable. the phylogenetic events by the real, biologically- Thus ontogeny combines great conservatism with based processes. The succession of the characters possibility to some shifts, which thus leave a place also implies their “trivial”, “systematic” homology for evolution. Considerable, noticeable shifts of in any reasonable narrow-defined group (e.g. any ontogenetic cycles are e.g. paedomorphic rhinophores and gills within dorid group are clearly (then juvenile features persist on the adult homologous, but their exact degree of homology stages; same for progressive (“peramorphic”) to other opisthobranch group is less defined), shifts; then new stages are added into a given and also a “deep” homology between prima facie ontogeny). When the initial ancestral shape very different structure between far related taxa. is greatly distorted, then this could be named The unequivocal homological correspondences as “disruption” of an ontogenetic cycle. This manifested at the morphological level may have succession of the ontogeneses in the historical very complicated both molecular and ontogenetic perspective (evolution) thus, forms own theory, “recording methods” in the heredity system (see e.g. significantly independent from the procedure of Shubin et al., 1997, 2009). This not yet discovered the attributing, describing above. “DNA to morphology” transition mechanisms do 5. The biologically-based unifying theory of the not prevent us from assessing these unique sets as historical transformations of the organisms real characteristics of taxa. should in a less contradicting way integrate 8. The ontogenetic memory is maximally identical information obtained from these both main between closely related species and maximally systematic procedures: a) attributing of the yet imprecise between the very distantly related unclassified specimens according to the degree ones. These facts allow considering traditional of the correspondence to the already known evolution-free systematics within real, instead of unique sets of the characters (taxa) and b) theory only “hypothetic” historical dimension. This also of the historical transformations of these sets of implies that is senseless to search for a “succession characters (phylogeny in a strict sense). Such of all ancestors” in a given ontogeny, that is often general theory which is included both listed used as an evidence for a complete “failure” of components is therefore suggested to term as the biogenetic law. Such succession can be really ontogenetic systematics. found, but on a restricted phylogenetic distance, 6. Ontogeny — it is the only existing real keeper of the between particular taxa. Thus, only converse of the memory of the preceding historical transformations. systematic hierarchical principle into sequence of I.e. ontogeny — it is the only single available the ontogenetic transformations, from particular process allowing to link these two key-components taxa to particular taxa, it is possible to construct a of ontogenetic systematics (history–independent general theory of systematics, i.e. general theory taxa attribution and the theory of phylogeny itself) of the organism shape transformations. These in a maximally objective, biologically-, and not statements also imply several very important statistically-based way. further issues: a) Recapitulations are really 7. The succession of the characters allows to discover existing and are mighty instruments of the various traces of the ancestral ontogeny (a limited phylogeny inference, but should be used only on

205 ALEXANDER V. MARTYNOV

a restricted phylogenetic distance; b) Any well- these processes are widespread among living established taxon of traditional systematics is organisms. Any species thus is a combination of an approximation to the model of the ancestral regressive and progressive features. Many groups ontogenetic cycle — i.e. such if its ancestral also clearly demonstrate domination of the larval species had been really existed some time ago, features, thus show the paedomorphosis in narrow and had particular, functional, and not some sense. formal and abstract, properties; c) Therefore, 12. A resulting model of the phylogenetic the in the modern evolutionary and phylogenetic transformations of the ontogenetic cycles may theory almost completely abandoned term include the hierarchical order of the sequence “transitional taxon” or “missed link” is of great of the taxa appearing, with indications of the importance and should be restored and given full major succession of the morphological structures consideration as a valid and not auxiliary, “bad” (key phylogenetic characters; it is a method to term, i.e. a heritage of old, “imprecise” and distinguish a cycle within a line), graphic model “arbitrary” systematics. of the morphological transformations, and also a 9. There are firm evidences that some sets of tree-like scheme. characters (that can be called as morphogenetic networks), for instance, some “dependent” In a short conclusion, the “cycle-thinking” characters of one ontogenetic cycle (e.g. juvenile vs. “tree-thinking” lead to another important features) may became “dominating” characters implication: we should not concentrate so much on (e.g., adult, definitive), and vice versa. This the merging of any descendant taxon with ancestral indicates that evolution does not equal freely ones, as often done by current phylogenetics, but varying characters along the endless “branches” instead on separations of “new cycles” from “old of the “phylogenetic tree”, but instead, rather lines”. It is possible by detecting of the sets of the within quite limited frames of particular entire newly emerged morphogenetic networks as key- ontogenetic cycles, most basic of the have features of every new taxa, never existed in the originated already in the earlier Cambrian and previous one. Restricted recapitulations on limited then became named by traditional systematics as phylogenetic distances became instead some the metazoan phyla. obscure theoretical considerations, but a practical 10. The consideration of the evolution restricted by the system of “reference points”, unambiguously particular “technological” properties of particular indicate real, and not hypothetically constructed ontogenetic cycles lead to the possibility of direction of the evolution. Thus, the ranking system considering the diagnosis of any well-established of the traditional systematics does not lose their traditional systematic group (taxa) higher than importance: it has a crucial role indicating the species level as a minimally contradicting model succession of the modifications of the ontogenetic of the ancestral ontogenetic cycle, that imply cycles with help of the recapitulatory reference that all subtaxa included in this group should be points, but not as a continual line, but as succession therefore products of the modifications of this of sets of particular functional organizations, ancestral ontogenetic cycles. including particular properties of a given 11. Historical transformations of the ontogenetic ontogenetic cycle. For instance, there are not any cycles (evolution) might perform in a regressive crustaceans with veliger larvae and there are not way (for instance, paedomorphosis), a reduction any molluscs with the nauplius larvae. development of a given character (set) compared to the ancestral condition, and in a progressive The usefulness of the consideration of any way compared to the ancestral ontogeny. Both well-established, “narrow-enough” in definition

206 FROM “TREE-THINKING” TO “CYCLE-THINKING”: ONTOGENETIC SYSTEMATICS OF NUDIBRANCH MOLLUSCS

by the morphological characters and apparently in most species and triaulic in two species of the monophyletic in phylogenetic terms, taxon as model genus Bathyberthella. Vas deferens is fully closed. of a particular ancestral ontogenetic cycle (imagined, Copulative apparatus near female genital opening on thus as a really existed ancestral species, and not just the left side in front of the gill. a formal node of a tree) will be demonstrated on the Including families (alphabetic order): example of the three taxa that are closely related Pleurobranchidae and Pleurobranchaeidae. from the point of view of traditional systematics Key newly emerged morphogenetic networks (e.g. Thiele, 1931), morphological cladistics (e.g. (key phylogenetic characters): Notum; Rhinophoral- Wägele & Willan, 2001) or molecular phylogenetic oral veil common apparatus; Gill apparatus and anus study (e.g. Jörger et al., 2010), i.e. groups of free of mantle cavity on the body wall; “higher” opisthobranchs, notaspids (s.str.) and Model of key-transformations: Mantle nudibranchs that were considered earlier within transforms into the notum (progressive); Gills the “supertaxa” Acoela or . Based on attached directly to the body wall (regressive); sets of unique structural characters, we recognize Posterior corners of the cephalic shield are bending three potential ancestral ontogenetic cycles within forward and form thus enrolled rhinophores, enlarged this group, and thus further define them as order anterior part of the cephalic shield became oral veil Notaspidea s.str (=Pleurobranchoidea), order (progressive); Doridacea (=Anthobranchia) and Nudibranchia s.str. Phylogenetically important succession of the (=Cladobranchia) and then will shortly outline some key-characters: Rhinophores remain fused with basic properties and evolutionary potential of these the oral veil, the feature strongly points to the defined ontogenetic cycles in order to conclude cephalic shield of Cephalaspidea s.l. Lateral gill and whether they rather represent a natural, monophyletic lateral anus position strongly point to Cephalaspidea unit or not. This is an initial example and an s.l. Non-cephalic copulative organ weakly points to experiment in an almost completely new field of the Acteonidae. biologically-based ontogenetic systematics and can Recapitulations: Spiral external shell of the not be completely free from flaws. Further analysis juvenile stages points to the basal opisthobranchs like in greater details and elaboration are necessary: Cephalaspidea s.lato (including Acteonidae). Ancestral group (strong or weak inference): Model of the ancestral ontogenetic cycle of the order Weak inference for Acteonidae-like group ancestry Notaspidea s.str. (=Pleurobranchoidea) (Fig. 1) (non-head copulative apparatus). Monophyletic status: Yes. Systematic diagnosis: Notum present; Thin General model of the ancestral ontogenetic internal shell under the notum or absent; Rhinophores cycle: Pelagic larva with spiral shell > benthic juvenile are well defined, enrolled and united together with with notum enclosed the shell, formation of free oral veil into a common structure; Mantle cavity enrolled rhinophores connected with each other and absent in all species; Single true gill is always on the with oral veil > benthic adult with internal plate-like right lateral side under the notum; Anal opening on shell, entire notum both anteriorly and posteriorly, the right side behind the gill and under the notum; enrolled rhinophores connected with oral veil, latero- Genital opening in front of the gill; Jaws comprising ventral gill and anus partially covered by notum. from numerous separate elements, always present; Model of further ingroup evolution (major Radula possess numerous uniform hook-shaped features): Shortening of the life of pelagic larva radular rows; Central tooth absent; Digestive gland towards to the direct development; Shell reducing is entire, not branched; CNS with fused cerebral towards complete absence; Notum reducing (not and pleural ganglia; Reproductive system is diaulic complete); Gill reducing (not complete);

207 ALEXANDER V. MARTYNOV

The Tree and the Cycle: Interaction Adult apomorphic character

Juvenile plesiomorphic character

Descendant taxon

Adult apomorphy Deletion Juvenile plesiomorphic character

Juvenile plesiomorphy Adult plesiomorphic Ancestral taxon character

Figure 4: General scheme illustrated interactions between evolutionary trees and ontogenetic cycles. Phylogenetic consequences of the characters transformations in ontogenetic context: juvenile plesiomorphic character (red) capable to transform into adult apomorphic character (orange) many times independently in different lines underlined by the same ancestral ontogenetic cycle.

Model of the ancestral ontogenetic cycle of the differentiated and reduced in different degree; Central order Doridacea (=Anthobranchia) (Figs 1, 5) tooth present or absent; Digestive gland is entire, not branched; CNS with fused or separated cerebral and Systematic diagnosis: Notum present or reduced; pleural ganglia; Reproductive system is triaulic in Internal shell always absent; Rhinophores are well most species, except for some (but not all) species of defined, solid, completely separate from oral veil and the genus Bathydoris. Vas deferens is fully closed. enclosed by anterior notal margins; Oral veil small, Copulative apparatus near female genital opening on placed under anterior notum (if it not reduced); Mantle the left side of the body. cavity absent in all species; Single true gill is circularly Including families (alphabetic order): bent in a secondary cavity medially on the dorsal Actinocyclidae, Aegiridae, Akiodorididae, Aldisidae, side of notum or gill cavity absent; Few species have Anculidae, Bathydorididae, Chromodorididae, ventral gills or gills are completely reduced; Genital Dendrodorididae, Discodorididae, Dorididae, openings lateral; Jaws comprising from the numerous Goniodorididae, Gymnodorididae, Hexabranchidae, separate elements, or monolith, often absent; Radula Mandeliidae, Onchidorididae, , possess numerous basically hook-shaped radular rows Polyceridae, Vayssiereidae.

208 FROM “TREE-THINKING” TO “CYCLE-THINKING”: ONTOGENETIC SYSTEMATICS OF NUDIBRANCH MOLLUSCS

Key newly emerged morphogenetic networks of the dorsal cavity where then three first gills (key phylogenetic characters): Formation of developed > benthic adult without shell, entire notum separate rhinophores in closed rhinophoral pockets; both anteriorly and posteriorly, solid rhinophores not Solid rhinophores; Formation of a highly specialized connected with oral veil, dorso-terminal circumanal and complicated dorsal gill apparatus and gill cavity. gill within a special cavity. Model of key-transformations: Internal shell Model of further ingroup evolution (major completely disappeared (regressive); Separation of features): Shortening of the life of pelagic larva the fused rhinophores into the separate pockets by towards to the direct development; Notum reducing; the lateral shifting and anterior notal lobes including Posterior notal lobes appearing; Gill cavity reducing; formation of the rhinophoral pockets (progressive); Gills reducing; Rhinophoral pockets reducing. Formation of the solid rhinophores and their complete Explanatory remarks: Presence of the well- separation from oral veil (progressive); Single gill established recapitulations of the notal and and anus transit from latero-ventral to dorso-terminal rhinophoral patterns, and ventral position together position with simultaneous formation of the closed with the significant structural similarity of the adult gill cavity (progressive). stages (presence of the notum; digestive, nervous, Phylogenetically important succession reproductive systems details etc.) strongly point for of the key-characters: Circumanal gills remain the notaspidean s.str. ancestry, closely similar to associated with the anus and nephroproct strongly the recent members of the family Pleurobranchidae. point to Notaspidea s.str. and Cephalaspidea s.l. Presence of the recapitulation of the juvenile gill Well defined (though solid) rhinophores point to cavity in the juveniles of completely phanerobranch Notaspidea s.str. Triaulic reproductive system points genus , presence of the fully-functional to the pleurobranchid notaspidean of the genus to partially reduced gill cavity in the adult stages Bathyberthella. of some genera of so called phanerobranch group Recapitulations: Spiral shell of the juvenile stages (Onchimira, Calycidoris, Diaphorodoris), presence points to basal opisthobranchs like Cephalaspidea of the juvenile gill cavity in the adult meyeni, s.lato (including Acteonidae) and Notaspidea; Fused presence of the partially modified semi-open gill rhinophores in earlier postlarval stages and not cavity in adult of some species strongly enclosed by the anterior notal lobes strongly point to point to ancestral condition of the gill cavity presence Notaspidea s.str. (family Pleurobranchidae); Ventral and their further reduction as major feature of the anus in earlier postlarval stages strongly points to ingroup evolution (see below for details). Notaspidea s.str.; Posterior entire notum of middle postlarvae without gills and gill cavity strongly points Model of the ancestral ontogenetic cycles of the to Notaspidea s.str. order Nudibranchia s.str. (=Cladobranchia) Ancestral group (strong or weak inference): Strong inference for the Notaspidea s.str. Systematic diagnosis: Notum present or reduced; Monophyletic status: Yes. Internal shell always absent; Rhinophores are well General model of the ancestral ontogenetic defined, solid, connected with oral veil and partially cycle: Pelagic larva with spiral shell > benthic surrounded by lateral notal margins; Oral veil is large, juvenile completely loosing the shell, formation of the completely substitutes anterior notum. Mantle cavity solid rhinophores shifted laterally from each other by absent in all species; True gill is absent in all species; the enclosing of the lobes of anterior notum, complete Various secondary gills or papillae present in many separation of the rhinophores and oral veil, posterior species; Genital openings on the lateral side; Jaws notal lobes starts to develop in earlier postlarvae and monolith, well defined; Radula possess numerous led to dorsal anus shift and formation in juveniles basically hook-shaped radular rows differentiated and

209 ALEXANDER V. MARTYNOV

reduced in different degree (up to one teeth per row); s.lato (including Acteonidae) and Notaspidea; Central teeth present in most species; Digestive gland Separated rhinophores in earlier postlarval stages is entire in few taxa, in most is branched in various of Tritonia species and not enclosed by the anterior degree, in many taxa branches are penetrated to the notal lobes point to family Pleurobranchaeidae of dorsal papillae; CNS with fused or separated cerebral Notaspidea s.str. and pleural ganglia; Reproductive system is diaulic in Ancestral group (strong or weak inference): most species; Vas deferens is fully closed. Copulative Strong inference for the Notaspidea s.str. Weak apparatus near female genital opening on the left side inference for the particular notaspidean family of the body. Pleurobranchaeidae. Including families (alphabetic order): Monophyletic status: Yes. Aeolidiidae, , Bornellidae, Calmidae, General model of the ancestral ontogenetic Cumanotidae, Dendronotidae, Dironidae, cycle: Pelagic larva with spiral shell > benthic juvenile Doridomorphidae, Doridoxidae, Dotoidae, completely loosing the shell, formation of the solid Embletoniidae, Eubranchidae, Facelinidae, rhinophores shifted laterally from each other without Fionidae, Flabellinidae, Glaucidae, Goniaeolidiidae, enclosing of the lobes of anterior notum, integration Hancockiidae, Heterodorididae, Janolidae, of the rhinophores and oral veil > benthic adult Lomanotidae, Notaeolidiidae, Phylliroidae, without shell, with reduced notum both anteriorly Pseudovermidae, Scyllaeidae, Tergipedidae, and posteriorly, solid rhinophores integrated with Tethydidae, . the widened oral veil which is substitute the anterior Key newly emerged morphogenetic networks part of notum, true gill completely absent, weakly (key phylogenetic characters): Strong oral papillated lateral and dorsal notum. veil substitute anterior part of the notum; Solid Model of further ingroup evolution (major rhinophores; Solid jaws. features): Shortening of the life of pelagic larva Model of key-transformations: Internal shell towards to the direct development; Notum reducing; completely disappeared (regressive); Rhinophoral and Rhinophoral pockets reducing; Rhinophores oral apparatus accepted as direct modification of the secondary united together. Pleurobranchaeidae type with addition of formation of Explanatory remarks. The special similarity the solid rhinophores still integrated with the strongly between rhinophoral and oral apparatus of some basal widened oral veil substitutes the anterior part of the Nudibranchia (Doridoxa, Heterodoris, Tochuina, notum (progressive or neutral); Complete reducing Tritonia) and the particular notaspidean s.str. of the gill (regressive); Formation of the solid jaws family Pleurobranchaeidae and ontogenetic patterns based on integration of the separated elements of the of the rhinophores development point to possible notaspideans (progressive). nudibranchs s.str. ancestry within the latter group (see Phylogenetically important succession of below for details). the key-characters: Anus and nephroproct still placed laterally strongly point to Notaspidea s.str. ONTOGENETIC SYSTEMATICS OF THE Well defined (though solid) rhinophores point to DORID NUDIBRANCHS (ORDER DORIDACEA) Notaspidea s.str. Laterally shift rhinophores and widened oral veil substitutes anterior part of the As was defined above, the dorid group (here notum point to particular notaspidean s.str. family considered as order Doridacea) possesses a set of Pleurobranchaeidae. Diaulic reproductive system unique characters, including a dorsal circumanal points to the Notaspidea. gill corolla integrated within posterior part of the Recapitulations: Spiral shell of the juvenile stages notum and elaborated rhinophores, integrated within points to basal opisthobranchs like Cephalaspidea anterior part of the notum, typically within special

210 FROM “TREE-THINKING” TO “CYCLE-THINKING”: ONTOGENETIC SYSTEMATICS OF NUDIBRANCH MOLLUSCS

pockets. The oral veil of dorids is small and reduced, (including super- and suborders) and completely anterior end of the body is the notal edge, not the disregard any classical concepts of “Opisthobranchia” edge of the oral veil (Figs 2 D; 3 C). Dorids include and “Nudibranchia”, thus prima facie perfectly two major groups — Cryptobranchia (cryptobranch anticipated all further both morphological cladistics dorids) and Phanerobranchia (phanerobranch dorids). (e.g. Haszprunar, 1988) and molecular phylogenetics Cryptobranchia possess special cavity around (Jörger et al., 2010) central idea of paraphyly of any the gills, where the gills are able to completely traditional taxa. However, despite on this challenge, retracted (Fig. 2 D) and usually well defined notum the Minichev and Starobogatov systems remained a without posterior lobes and numerous processes. “blind street” compare to the modern evolutionary Phanerobarnchia, by definition is completely devoid studies first of all was extremely lapidary and of any gill cavity, instead gills are directly attached to actually “closed” for further discussion in persisting the dorsal notum and able to contract, notum usually of “once and for all” approach. For instance, the reduced and often posses posterior lobes or processes. systems were constructed only on the reproductive Gills morphology and digestive (except for radula) apparatus features, and many important external and reproductive organs are essentially similar in and digestive system features have been completely cryptobranchs and phanerobranchs. omitted. At the boundary of 1980S-1990S only few publications dealing with nudibranch macrosystems Dorid systematics is deeply rooted in the classical and evolution have been further appeared (e.g. period of taxonomy (e.g. Linnaeus, 1758; Alder & Schmekel, 1985; Wägele, 1989a; Evolutionary Hancock, 1864; Abraham, 1877; Bergh, 1892; and Biology of Opisthobranchs, 1991) closely approach many others), but serious evolutionary studies were soon happen “phylogenetic explosion”. started only recently. Various classificatory schemes and implied their evolutionary models, already have Thus, accumulated to the end of XX century been suggested in old works (e.g., Bergh, 1892; knowledge on morphology of the nudibranch molluscs Odhner, 1934, 1939), but usually without much was so immense, whereas number of published discussion and unequivocal set of strict evidences. evolutionary works was so scarce, that “phylogenetic The same author can use different schemes of explosion” appeared as absolutely unavoidable. It was classification in different works, that is especially happen mostly at the millennium boundary and was characteristic for the nudibranch classics Rudolph much inspirited also by statistical approach, cladistics, Bergh and Nils Odhner. However, Odhner’s (1939) and somewhat later, by molecular phylogenetics (e.g., system of the order Nudibranchia including four Wägele & Willan, 2000; Valdés, 2002a,b; Fahey suborders (one of which is Doridacea — dorids), & Gosliner, 2004; Fahey & Valdés, 2005, Pola et became an important base actually for all modern al., 2007 and many others), and not only consider classification and phylogenetic reconstructions. the dorids but also any other living organisms. By current widespread opinion such combination finally The only pre-1990S challenge for the classical will answer on the “main evolutionary question” Nudibranchia concept was presented by Yu.S. and presented the exhaustive “Tree of Life” — Minichev and Ya.I. Starobogatov (1979). They, from Archaea to Chordata (see e.g. Maddison et however, choose similar to the classical authors al., 2007; Philippe et al., 2011a). The challenge thus the “not much discussion” strategy, and published is only increasing the number of genes used for already “completed” systems in the abstracts of reconstruction, and improvement of the statistical Soviet malacological conferences with minimum algorithms, etc. (e.g., Ciccarelli et al., 2006; Dunn commentaries and explanatory remarks. They et al., 2008; Goloboff et al., 2009; and many others). suggested numerous new taxa of the order ranks This problematic issues especially became clear in

211 ALEXANDER V. MARTYNOV

Adult apppomorphic character Echinocorambe Corambe Doridunculus Lophodoris Juvenile plesiomorphic character

Juvenile plesiomorphic Descendant taxon characters Deletion

Adult apomorphy

Juvenile plesiomorphy

Ancestral taxon Adult plesiomorphic character

Figure 5: Example of the interactions between tree and cycle: model of the independent regressive transformation (juvenilization) of the external shape of dorid nudibranch in several families: Akiodorididae (genera Echinocorambe and Doridunculus); Goniodorididae (genus Lophodoris); Onchidorididae (genus Corambe). I.e. in three different families of phanerobranch dorids essentially similar properties of the ancestral dorid ontogenetic cycle led to independent appearing similarly constructed postlarvae-like adults. the nowadays, when importance of the traditional about importance of the biology itself in the field of systematics have been challenged in favour of the molecular biology (see Woese & Goldenfeld, 2009). phylogenetics. Numerous researchers, working in the field of phylogenetics usually have superficial All this can be applied to the modern dorid knowledge about tasks and methods of the traditional systematics and phylogenetics. A good deal of modern systematics (i.e. systematics itself), and not rarely, papers have suggested a particular consideration underlined in their work by the past evolutionary- on a priori decisions of character polarity (Valdés, morphological concepts, without understanding scale 2002a,b; Fahey & Gosliner, 2004; Fahey & Valdés , of the past discussions. At the same time, earlier 2005 and others), but scarcely offer a discussion for molecular phylogenetists, were much more careful such decisions. That in this case difference to the past about phylogeny reconstructions, leaving a “last plea” “extremely arbitrary” implications of the traditional for the morphology and ontogeny (e.g., Raff et al., taxonomy and evolutionary morphology? One of such 1989). In this respect, quite remarkable, that famous most controversy evaluation has received the key- RNA (!) researcher Carl Woese in 2009 warned feature of the entire dorid group — gill apparatus.

212 FROM “TREE-THINKING” TO “CYCLE-THINKING”: ONTOGENETIC SYSTEMATICS OF NUDIBRANCH MOLLUSCS

Besides the traditional dorid group Phanerobranchia, reconstructions. Another important implication of prima facie lacking any gill cavity, such feature also the discovery of these highly aberrant corambids characteristics for another, minor Doridacea group was unambiguous consideration about ancestral –– Bathydoridoidea. That is most important, that the pattern of the cryptobranch mode (i.e. presence of latter group traditionally is always considered as an the gill cavity). Thus, step by step, was developed an “archaic” (Minichev, 1969). This view without critical improved model of the dorid evolution (see Fig. 1), discussion have been accepted by the morphological which much better explains existence real taxa being cladistics, as a plesiomorphic character, and simultaneously “typical” phanerobranch by presence bathydoridids, therefore were “translated” as a of the radula and buccal pump, but can be also no less “most basal group” (Valdés, 2002a,b). This concept better described as cryptobranch due to presence of of morphological cladistics was then accepted by the gill cavity, even reduced. molecular one (Valdés, 2004). However, by definition “cavityless” phanerobranchs have demonstrated the Thus, the new model of the dorid evolution, which presence of quite well defined gill cavities (e.g. in considers properties of the entire ontogenetic cycle the genera Calycidoris and Diaphorodoris). This (Fig. 1) implies principally different direction of the obviously important fact has received a very scarce evolution of the order Doridacea (Martynov, 1994b, discussion in past (Abraham, 1876; Roginskaya, 1995; Martynov et al., 2011; Martynov & Schrödl, in 1972; Millen, 1985) but was not considered at all by press) than that was “reconstructed” by the modern the modern phylogenetic studies (Valdés, 2002 a,b; phylogenetic analysis with an invisible help of very Fahey & Valdés, 2005; see Martynov et al., 2009 for old and purely evolutionary morphological concept of further details). the “archaic” phanerobranchs (Minichev, 1969).

Already in 1994 were obtained further evidences According to the new model, in the ancestral for much more complicated picture of the dorid ontogenetic cycle of the dorid groups (order evolution: was described an unusual corambid species Doridacea) has emerged for the first time a key- Loy meyeni, possess dorsal gills within small cavity novelty — mechanism of the transferring of the (Martynov, 1994a) –– the feature not only unknown anus area (i.e. area, there then gills will be formed) before in any corambids, but also strongly contradicts from ventral to the dorsal position (see Figs 1, 2). with the diagnosis of Phanerobranchia. Furthermore, This mechanism implies asymmetrical growth of pattern of the posterior notal lobes of Loy meyeini and the right notal posterior lobes and further formation another species L. millenae was essentially similar to on this base the gill cavity (Martynov et al., 2011). the pattern known in the dorid postlarval, specimens Thus, the dorid as group is clear delineated by described earlier (Thompson, 1958). Thus, as it was otherwise similar to it Notaspidea s.str., by appearing became obvious, that some important issues have in their ontogenetic cycle the gill cavity and separated been completely omitted in the dorid systematics rhinophores as key-novelties. This also implies that and evolutionary studies because all efforts were all modern cryptobranch dorids should have in their put on understanding of the adult stages, whereas ontogenetic cycle the peculiar stage with the posterior postlarval and juvenile stages of the ontogenetic notal lobes lead then to the gill cavity formation. cycle were not considered at all. However, quite clearly expressed implications on the paedomorphic Such prognostic ability is another very important origin of the corambids and significant corrections differences of the ontogenetic systematics from of the major trends of the dorid evolution (Martynov, the current phylogenetic thinking. Taking into 1994b, 1995) were positive noticed (Wägele & Willan, consideration, that about 1.500 species of cryptobranch 2000) but not considered at all for the phylogeny dorid are currently known, but only 2-3 species have

213 ALEXANDER V. MARTYNOV

been studied ontogenetically (e.g. Thompson, 1958; Another very important implication of the Usuki, 1967; Martynov et al., 2011; this study), with new ontogenetic model of the dorid evolution is its help of such prognostic model we can then infer, possibility to find traces (i.e. recapitulations) of the that earlier postlarval stages of yet unknown other initial cryptobranch mode of the gill formation in cryptobranch also should possess such stage with the the earlier postlarval stages of phanerobranch dorids, posterior lobes. A possibility to infer other ontogenetic which are completely lacking any gill cavity at the stages used the known one (first of all, adults) has adult stages (Martynov et al., 2011). already been suggested in frames of comparative and evolutionary morphology (see e.g., Remane, 1955), but Finally, the new model greatly contradicts with was not then used as a routine, practical methods. Such widely accepted considerations about primary possibility to infer for any yet unknown ontogenetic condition of the gill cavity absence in the dorid stage by already known is very important feature of group Bathydoridoidea (Minichev, 1969; Wägele, the ontogenetic systematics, making the systematics 1989a; Valdés, 2002 b). There is no any ontogenetic itself not only purely descriptive science, as usually information on the bathydoridid groups currently. considered, but instead, a science of predictions, that then could be tested in different ways (e.g. by However, apart form the absence of the gill study of the ontogeny of further dorid species). Most cavity, bathydoridids demonstrated number of other importantly also, the such possibility to infer unknown reduced features: the notum is completely reduced stages is greatly contradict to the Hennig fundamental in the Bathydoridoidea, but leave a clear border principle considering any organism as a just separate between lateral body wall and dorsal side (see, e.g. semaphoront. Wägele, 1989b). This fact quite clearly points toward considering this feature as a secondary reduced instead The new model of the dorid evolution has also of a primary one. Bathydoridids also are completely several different important implications. First of all, devoid of the rhinophoral pockets — a feature that the essential similarity of the ontogenetic pathways in usually points to reduction in other phanerobranchs. the different cryptobranchs (e.g. so different groups as Chromodorididae and Discodorididae) implies Until recently bathydoridids were considered as potential possibility for further independent, parallel having diaulic reproductive system (Odhner, 1934; appearing in non directly related dorid subgroups, of Minichev, 1969; Wägele, 1989a). It was used for heterochronic regressive variants, from “moderate”, further evidence of clear separation of Bathydorididae when only gill cavity not develops in adult organisms from other dorids, and most likely also as an archaic, (i.e. most species of the family Onchidorididae), to basal feature (e.g. Wägele, 1989a; Wägele & “radical”, then clear earlier postlarval features (notal Willan, 2000). However, 9 years ago was described lobes) appeared in the adult stages. Reality of the Bathydoris spiralis Valdés, 2002 (Valdés, 2002b), independent heterochronic shifts underlying by the the first bathydoridid with a triaulic reproductive common ontogenetic cycle in dorids have been already system, i.e. as in other dorids. The gill apparatus of B. demonstrated both structurally and phylogenetically spiralis is very similar to the compact gill corolla of (Martynov, 2000; Millen & Martynov, 2005; Martynov other dorids, and quite different from the “disperse” et al., 2011): corambid group (family Onchidorididae) gill placement of most other species of the genus and the genus Echinocorambe (Akiodorididae) Bathydoris. Two groups have been already recognized have reached externally very similar juvenile-like within bathydoridids (Valdés, 2002b). One group of adult organization, whereas internal features still species of the genus Bathydoris (including diaulic preserve unique pattern of radula and buccal pumps Bathydoris clavigera and triaulic Bathydoris spiralis) of onchidoridid and akiodoridid respectively. possesses eyes and a relatively low body, whereas

214 FROM “TREE-THINKING” TO “CYCLE-THINKING”: ONTOGENETIC SYSTEMATICS OF NUDIBRANCH MOLLUSCS

the majority of other Bathydoris species, like B. At the same time, if interpreting historical abyssorum Bergh, 1884 and, B. ingolfiana Bergh, development of the group Bathydoridoidea 1899, do not have eyes and their body became very according to the new model of the dorid evolution, high, almost sphere-shaped, what is very unusual for these apparent paradoxes mostly disappeared. dorids. Intriguingly, those species having eyes and Ancestor of the bathydoridid, should most likely low body inhabit relatively shallow depth of 40–500 have gill cavity and free notal edge. Increasing m, whereas eyeless species with spherical body are degree of adaptations to the more and more deep known from much deeper environments, up to 4500 water conditions led to consequential reductions m (Bathydoris abyssorum). of the gill cavity, notum, eyes and appearing in the most deep-water species this specific sphere- In absence of an elaborated model of the dorid shaped body (possible a response for the pressure). evolution and underlying by formal approach According to this model, omniphagy of the of phylogenetic systematics, the findings of the genus Bathydoris, considered as a plesiomorphic unique triaulic Bathydoris spiralis with typical character for all dorids (Valdés, 2002b, 2004), is for the many dorids circumanal gill corolla, the only a very special adaptation of the very restricted characters which are clearly connected small group in species number Bathydoris for the poverty of the Bathydoridoidea with 2000 other dorid species, food resources in the abyssal environment, when did not lead, nevertheless for the consideration of used any available benthic material. Bathydoridids, which species as a transitional taxon or “missing thus, judged from their many unique characters is a link”. Instead main conclusion of the single available monophyletic group, but incorrect previous model phylogenetic analysis of bathydoridid including B. of their evolution led to incorrect conclusion of spiralis (Valdés, 2002b) became their paraphyly. their paraphyletic status. Shallow-water species turned to be a sister-group for the usual phanerobranch dorids, however these Thus, one of the most important implications groups share only absence of the gill cavity. I.e. the of both ontogenetic systematics and new model of resulting tree actually has showed the similarity of the the dorid evolution, it their prognostic ability, i.e., bathydoridids and all other dorids, however, implicit in the case of dorids, the possibility to predict real model of the ancestral mode of the phanerobranch existence of several transitional taxa (“missing dorids led to their incorrect interpretation. The results links”), combining of the presence of the gill were truly paradoxical and required creation within cavity (diagnostic feature of Cryptobranchia) with single genus Bathydoris, not only new genera and various specific characters of the phanerobranch families, but potentially, even suborders (!), to show dorids (lacking any gill cavity by definition). the tree pattern in the system of categories (Valdés, Despite that it will be may be considered as a 2002b). Nevertheless, some bit of absurdity of such purely hypothetic field, many facts, instead speak potential decision was clear, and in the cited paper in favour of such approach. One of such most there were not suggested any new even generic remarkable case, when already constructed model names, but the case was discussed and interpreted as of the dorid evolution has found strong support in failure of the traditional systematics, hardly allowed new independent finding of Onchimira cavifera, true, phylogenetic interpretation at the present stage having both well-defined full-functional gill of knowledge (Valdés, 2002b). The final conclusion cavity and simultaneously very special radular and to consider the genus Bathydoris as a paraphyletic buccal pump features, allowing unambiguously complex s.l. not resolved any of the problematic to place this truly cryptobranch taxon into the issues and still there no any study attempting to make phanerobranch family Onchidorididae (see further solution for this remarkable case. Martynov et al., 2009).

215 ALEXANDER V. MARTYNOV

Table 1: Prognostic table for the group Doridacea

Gill cavity Gill cavity present absent Gills present, anus dorsal Genus yet unknown (Bathydorididae), Bathydoris (Bathydorididae)

All Cryptobranchia,

Genus yet unknown (Hexabranchidae), Hexabranchus (Hexabranchidae),

Onchimira, Calycidoris, Loy , Onchidoris, Adalaria (Onchidorididae) (Onchidorididae),

Genus yet unknown (Akiodorididae) Akiodoris, Doridunculus (Akiodorididae),

All Polyceridae, Genus yet unknown (Gymnodorididae) Gymnodoris, Analogium (GymnodoriGLdae) Gills present, anus ventral Corambe (Onchidorididae) Corambe s.str. (Onchidorididae), Echinocorambe (Akiodorididae) Gills absent, anus dorsal , Vayssierea (Vayssiereidae) Phylidiella, Phyllidiopsis etc. (Phyllidiidae) Gills absent, anus ventral Impossible combination? Fryeria

All these facts are of course intriguing but there ARE THE TRADITIONAL NUDIBRANCHIA is no any (at least explicit) prognostic in the modern MONOPHYLETIC? systematics, both “traditional” or “phylogenetic”, though few exotic attempts to produce a “periodical As defined above, the traditional order systems” for taxonomy have been performed Nudibranchia is here considered as two independent in past (e.g. Schimkewitsch, 1906, 1909). Here, groups (orders): Doridacea and Nudibranchia s.str. therefore, will attempted to challenge this, and gave (equal to the Cladobranchia). What was the reason for further evidence for the validity of the ontogenetic that, except for briefly outlined above in the diagnosis systematics by presenting a special application, and the model transformations? The traditional that is almost completely absent in the modern group Nudibranchia despite of numerous attempts systematics and phylogenetics: the prognostic table to challenge it (for instance, Marcus & Marcus (see Table 1). (1967) used a system of four independent orders

216 FROM “TREE-THINKING” TO “CYCLE-THINKING”: ONTOGENETIC SYSTEMATICS OF NUDIBRANCH MOLLUSCS

corresponding to the four suborders of Nudibranchia; Interesting material for such analysis available Minichev & Starobogatov (1979) not used the name from another external key nudibranchs feature — Nudibranchia at all), nevertheless recently received rhinophores. The small relic group Notaspidea s.str. status of monophyletic according to a morphological is well established as ancestral or sister taxon for the cladistic study (Wägele & Willan, 2000), and still no Nudibranicha (Thiele, 1931; Tardy, 1970; Martynov, unambiguous molecular evidences for their paraphyly 1999; Wägele & Willan, 2000; Martynov & Schrödl, (see Jörger et al., 2010 vs. Grande et al., 2004). 2008; present work). Both adult notaspids of the family Pleurobranchidae and earlier postlarval specimens of In the morphological phylogenetic analysis the order Doridacea (Figs 2; 3 A, B) possess similarly (Wägele & Willan, 2000) indicated only 4 constructed, united together rhinophores, without autapomorphies for Nudibranchia: 1. Solid special pockets, because anterior parts of the notum yet rhinophores; 2. Absence (by loss) of the shell; 3. not enclosed it, as in the adult dorids. Thus the anterior Longitudinally situated pericardium; 4. Presence of edge of the body of adult dorids is always the anterior special vacuolated epithelium. However, uniqueness edge of the notum, whereas the oral veil is reduced and of most of these characters, their importance as remained under the anterior part of the notum. the indicators of the monophyletic group may be questioned. First of all the loss of shell is really not In this respect quite remarkable and until possible to discuss as any key-characters since it recently never considered is the fact of the significant happened many times independently within various similarity in the general rhinophoral patterns between Opisthobranchia. Formation of the solid rhinophores one of the notaspid family — Pleurobranchaeidae on the base of the enrolled chemosensory organs of the and most of the subgroups of the traditional notaspideans is also not unique for the Nudibranchia, Nudibranchia, excluding the dorids. If in Doridacea and clearly independently has taken place within anterior notal edge enclosed the rhinophores and the group Saccoglossa (e.g., Jensen, 1996) and then separate initially common structures of the acochlidians (Schrödl, Neusser, 2010). In a similar rhinophores itself and oral veil (Figs 2D; 3C), in one way, another listed nudibranch autapomorphy — the of the notaspid family, Pleurobranchaeidae, oral veil, longitudinally situated pericardium also independently instead, considerably widened, whereas anterior part appeared within Sacoglossa and acochlidians. of the notum is reduced. As result, rhinophores are Finally, the vacuolated epithelium occurrence and considerably shifted laterally, and anterior edge of the functions remains very scarcely studied regarding its oral veil (!) became anterior edge of the body instead phylogenetic importance. Even, based on exclusively of notum as in dorids (Compare Figs 3 A–D and 3 E). structural, systematical approach, traditional Nudibranchia are very heterogeneous. The group Most remarkably, that several obviously basal of dorids, from one side, and three other traditional in cladistics terms (see e.g. Schrödl et al., 2001), suborders of the Nudibranchia (Dendronotacea, but not related directly nudibranch genera, i.e. Arminacea and Aeolidacea, all three suborders also Tochuina and Tritonia (both belong to the traditional have been united under the name of Cladobranchia) suborder Dendronotacea), and Heterodoris and have so many important differences regarding both Doridoxa (traditional suborder Arminacea) have external and internal (digestive and reproductive amazingly similar to the family Pleurobranchaeidae systems) characters, that make unification these rhinophoral apparatus, including wide oral veil groups under one name problematic. Is it however and reduced anterior notal edge (Compare Figs some promising set of structures that may be used 3 E and 3 F). I.e. in the basal genera of at least for at least potential assessment of the para- or two major traditional groups of Nudibranchia — monophyletic status of Nudibranchia? Dendronotacea and Arminacea anterior edge of the

217 ALEXANDER V. MARTYNOV

oral veil is anterior edge of the body, exactly as in have separated rhinophores at adult stage but yet the family Pleurobranchaeidae (see also Martynov united in the postlarval ontogeny. Rhinophoral & Schrödl, 2008). The difference between adult development of Tritonia appeared thus as secondary morphology of the rhinophoral apparatuses of modification of the ontogeny of Pleurobranchaeidae, the notaspidean families Pleurobranchidae and but in which the stage with united rhinophores was Pleurobranchaeidae (Compare Figs. 3 A and 3 E) further deleted. If Tritonia is a descendant of the has also remarkable correspondence in the ontogeny. dorid-based ontogenetic cycle, it is more likely that In the postlarval specimens of Berthella californica stage with the united rhinophores should be persists (family Pleurobranchidae) anterior edge of the notum in it, which is not the case. further grows and partially covers rhinophores and oral veil (LaForge & Page, 2007), whereas in Thus a scenario implies that traditional postlarval specimens of Pleurobranchaea japonica Nudibranchia is heterogenous group in the terms of anterior part of notum reduced, and remained traditional systematics and paraphyletic in cladistics’ notum fused with the posterior part of the oral view (Martynov, 1999) can not be completely excluded. veil (Tsubokawa & Okutani, 1991; Gibson, 2003). For further proof of above described ontogenetic As already well established (Thompson, 1958; patterns it is highly necessary modern studies of the Usuki, 1967; Martynov et al., 2011; present study), postlarval ontogeny of some basal nudibranchs, and earlier postlarval stages of both, cryptobranch and also search for additional characters that potentially phanerobranch dorids have principally similar to can share Pleurobranchaeidae and basal Nudibranchia. the adult and postlarval notaspid of the family At least, in non-dorid nudibranchs there is the clear Pleurobranchidae pattern of the united rhinohores pleurobranchid character — lateral position of not enclosed by the anterior notum (Compare Figs anus. Basal nudibranch genera, such Heterodoris 3 A and 3 B). and Doridoxa, are without secondary respiratory structures, thus is possible to consider these taxa as On the contrary, postlarval ontogeny of modifications of the ancestral ontogenetic cycle of the Tritonia hombergi (Nudibranchia: Dendronotacea), Pleurobranchaeidae, in which regressive heterochronic rhinophores, compare to the both dorids and event of the true gill reduction took place. notaspids do not demonstrate a stage with united rhinophores at all, but instead, they appeared CONCLUDING REMARKS initially already very separate from each other, at the lateral edges of future oral veil and lateral Under approach of the ontogenetic systematics parts of the notum (Thompson, 1962). Thus, the quite senseless to speak about just “plesiomorphic” only principal difference otherwise very similar or “apomorphic” states (two basic terms of the rhinophoral apparatuses of Pleurobranchaeidae morphological phylogenetics). Instead much more and non-dorid nudibranchs — presence in the productively to specify that was common ancestral latter group of solid rhinophores. However, it is juvenile condition (i.e. juvenile plesiomorphic not prevent for further originating of the solid state) and that was ancestral adult condition (i.e. rhinophores on the base of the enrolled as it was adult plesiomorphic state) (see Figs 4, 5). Further definitely took place independently in Sacoglossa. heterochronic shifts towards juvenilization of the In this respect, the ontogenetic mechanism of the adult morphology might, for instance, produce similar rhinophores development in Tritonia, without stage juvenile adult morphology many times independently in with the unite rhinophores, may be an important different families, strongly underlined by the common evidence that such pattern is further modification of ancestral ontogenetic cycle and it is well supported, e.g., exactly Pleurobranchaeidae-like notaspids, which in dorid nudibrancs, by unique morphological markers.

218 FROM “TREE-THINKING” TO “CYCLE-THINKING”: ONTOGENETIC SYSTEMATICS OF NUDIBRANCH MOLLUSCS

Thus, ontogenetic systematics does not disregarded REFERENCES trees as useful representation of the evolutionary process, but instead makes phylogenetic terminology Abraham PS (1877). Revision of the anthobranchiate and conclusions much more precise, thus highlights nudibranchiate , with descriptions or notices the integrative interactions between the “cycle” and the of forty-one hitherto undescribed species. Proceedings “tree” (Figs 4, 5). of the Zoological Society of London, pp. 196–269. Alberch PS, Gould J, Oster GF, Wake DB (1979). Size and The “cycle-thinking” implies thus thinking in shape in ontogeny and phylogeny. Paleobiology 5: the term of newly emerged characters and particular 296–317. functional organizations in any taxa, even currently Albert VA, Gustafsson MHG., Di Laurenzio L (1998). completely extinct, with recapitulation on limited Ontogenetic systematics, molecular developmental distance as “reference points” in the evolutionary genetics, and the angiosperm petal. Molecular succession of each preceding and next ontogenetic systematics of plants II: Eds. Soltis P. S., Soltis D. E., cycle-taxon. In this respect, the above demonstrated Doyle J.J. Boston: Kluwer Academic Publishers. P. key-novelties have well characterized Notaspidea 349–374. s.str., Doridacea and Nudibranchia s.str., but failed to Alder J, Hancock A (1864). Notice of a collection of found a reasonable ground in the purely phylogenetic nudibranchiate Mollusca made in India by Walter conception “Nudipleura”, overwhelmingly merged all Elliot Esq., with descriptions of several new genera these quite well defined separate taxa and obscure and species. Transactions of the Zoological Society of potential paraphyletic status of traditional (and still London 5: 113-147, pls. 28-33. accepted as monophyletic) Nudibranchia s.l. Beklemishev VN (1925). Morphological problem of the structures. Proceedings of the Biological ACKNOWLEDGEMENTS Research University of Perm University. 3. 74 p. Beklemishev VN (1969). Principles of comparative anatomy I am sincerely grateful to T.A. Korshunova of invertebrates. Vol. 1-2. Chicago: University of (Institute of Higher Nervous Activity and Chicago Press. 490 p., 529 p. Neurophysiology, Russian Academy of Sciences, Bergh R (1892). System der Nudibranchiaten Gasteropoden. Moscow) for her generous help in research. My Semper C.G. (ed.). Reisen im Archipel der Philippinen. particular gratitude for long term collaboration is due Zweiter Theil. Wissenschaftliche Resultate. Band 2, 3. to Michael Schrödl (Zoologische Staatssammlung, Wiesbaden. S. 995–1165. München). Special thanks to organizers of the 3th Bonner JT (1965). Size and сycle. An essay on the structure I WO, ve r y suc ce ssf u l b ot h scie nt i f ical ly a nd cu lt u r al ly of biology. Princeton: Princeton Univ. Press. 219 pp. meeting, and particularly, to Jesús S. Troncoso Box MS, Bateman RM, Glover BJ, Rudall PJ (2008). (University of Vigo). Hans Bertsch (Universidad Floral ontogenetic evidence of repeated speciation via Autónoma de Baja California) is thanked for kind paedomorphosis in subtribe Orchidinae (Orchidaceae). discussions. Specialists of the Interdepartmental Botanical Journal of the Linnean Society. 157: 429–454. Laboratory of Electron Microscopy (Biological Bryant HD (1991). The polarization of character Faculty, Moscow State University), including head transformations in phylogenetic systematics: role of laboratory G.N. Davidovich, chief engineer of axiomatic and auxiliary assumptions. Systematic A.G. Bogdanov, and researchers Yu.V. Golubtsova Zoology 40: 433-445. and A.M. Kusnetsova, provided all the necessary Carroll SB (2008). Evo-Devo and an expanding evolutionary conditions for the study of morphological structures. synthesis: a genetic theory of morphological evolution. This study was supported by DFG grants no. SCHR Cell 134: 25–36. 667/6–1 and SCHR667/10–1. Ciccarelli FD, Doerks T, von Mering C, Creevey CJ, Snel

219 ALEXANDER V. MARTYNOV

B, Bork P (2006). Toward automatic reconstruction of Gibson GD (2003). Larval development and metamorphosis a highly resolved tree of life. Science 311: 1283–1287. in , with a review of Danser BH (1950). A theory of systematics. Bibliotheca development in the Notaspidea (Opisthobranchia). Biotheoretica. 4: 115–180. Biological Bulletin 205: 121–132. Darwin C (1859). On the origin of species by means of Gilbert SF, Opitz JM, Raff RA (1996). Resynthesizing natural selection, or the preservation of favoured races evolutionary and developmental biology. Developmental in the struggle for life. 1st ed. L.: John Murray. 502 pp. Biology 173: 357–372. Dayrat B, Gosliner TM (2005). Species names and Goloboff PA, Catalano SA, Mirande JM, Szumik CA, Arias metaphyly: a case study in Discodorididae (Mollusca, JS, Kallersjo M, Farris JS (2009). Phylogenetic analysis Gastropoda, , Nudibranchia, Doridina). of 73 060 taxa corroborates major eukaryotic groups. Zoologica Scripta 34: 199–224. Cladistics 25: 211–230. De Beer GR (1930). Embryology and evolution. Oxford: Gould SJ (1977). Ontogeny and phylogeny. Cambridge: Oxford Univ. Press. 108 p. Harvard Univ. Press. 501 pp. De Queiroz K (1985). The ontogenetic method for Grande CJ, Templado J, Cervera L., Zardoya R. (2004). determining character polarity and its relevance to Phylogenetic relationships among Opisthobranchia phylogenetic systematics. Systematic Zoology 34: 280– (Mollusca: Gastropoda) based on mitochondrial cox 1, 299. trnV, and rrnL genes. Molecular Biology & Evolution Dunn CW et al. (2008). Broad phylogenomic sampling 33: 378-388. improves resolution of the animal tree of life. Nature Haeckel E (1866). Generelle Morphologie der Organismen. 452: 745–750. Allgemeine Grundzüge der organischen Formen- Edwards SV (2009). Is a new and general theory of Wissenschaft, mechanisch begründet durch die von molecular systematics emerging? Evolution 63: 1-19. Charles Darwin reformirte Descendenztheorie. Bd. Evolutionary Biology of Opisthobranchs (1991). Edmunds 1–2. Berlin: G. Reimer. 574 S. 462 S. M (ed.). Malacologia 32: 203–327. Halanych KM (2004). The new view of animal phylogeny. Fahey SJ, Gosliner TM (2004). A phylogenetic analysis of Annual Review Ecology, Evolution & Systematics 35: the Aegiridae Fischer, 1883 (Mollusca, Nudibranchia, 229–256. Phanerobranchia) with descriptions of eight new species Hall BK (1992). Evolutionary Developmental Biology. and a reassessment of phanerobranch relationships. London: Chapman and Hall. 275 p. Proceedings of the Californian Academy of Science Hall BK (1999). Evolutionary developmental biology 55: 613–689. (Second ed.). Dordrecht: Kluwer Academic Publishers. Fahey SJ, Valdés Á (2005). Review of Acanthodoris Gray, 491 pp. 1850 with a phylogenetic analysis of Onchidorididae Hall BK (2003). Evo-devo: evolutionary developmental Alder and Hancock, 1845 (Mollusca, Nudibranchia). mechanisms. International Journal Developmental Proceedings of the Californian Academy of Science Biology 47: 491–495. 56: 213–273. Haszprunar G (1988). On the origin and evolution of Felsenstein J (2004). Inferring phylogenies. Sunderland: major gastropod groups, with special reference to the Sinauer Associates. 664 pp. Streptoneura. Journal of Molluscan Studies 54: 367- Fink WL (1982). The conceptual relationship between 441. ontogeny and phylogeny. Paleobiology 8: 254–264. Hennig W (1966). Phylogenetic systematics. Urbana: Garstang W (1922). The theory of recapitulation: a critical University of Illinois Press. 263 p restatement of the biogenetic law. Proceedings Linnean Ivanov AV, Mamkaev YuV (1973). Turbellaria, their origin Society, Zoology 35: 81–101. and evolution. Leningrad: Nauka Publishing House. Gegenbaur C (1859). Grundzüge der vergleichenden 221 p. Anatomie. Leipzig Wilhelm Engelmann. Verlag: 606 S. Ivanova–Kazas OM (1995). Evolutionary embryology of

220 FROM “TREE-THINKING” TO “CYCLE-THINKING”: ONTOGENETIC SYSTEMATICS OF NUDIBRANCH MOLLUSCS

animals. St. Petersburg. 565 pp. decima, reformata. Holmiæ: Salvius. 824 pp. Jaecks GS, Carlson SJ (2001). How phylogenetic inference Love AC, Raff RA (2003). Knowing your ancestors: themes can shape our view of heterochrony: examples from in the history of evo-devo. Evolution and Development thecideide brachiopods. Paleobiology 27: 205-225. 5: 327–330. Jensen KR (1996). Phylogenetic systematics and Lukashov VV (2009). Molecular evolution and phylogenetic classification of the Sacoglossa (Mollusca, Gastropoda, analysis. Moscow: Binom. 256 p. Opisthobranchia). Philosophical Transaction Royal Maddison DR, Schulz K-S, Maddison WP (2007). The Tree Society London. Ser. B 351: 91–122. of Life Web project. Zootaxa. 1668: 19–40. Ji Q, Luo ZX., Zhang X, Yuan CX, Xu L (2009). Evolutionary Marcus Ev, Marcus Er (1967). American opisthobranch development of the middle ear in Mesozoic therian mollusks Part I, Tropical American opisthobranchs, mammals. Science. 326: 278-281. Part II, Opisthobranchs from the Gulf of California. Jörger KM, Stöger I, Kano Y, Fukuda H, Knebelsberger Studies in Tropical Oceanography, Miami 6: 1-256. T, Schrödl M (2010). On the origin of Acochlidia Martynov AV (1994a). Materials for the revision of the and other enigmatic euthyneuran gastropods, with nudibranch molluscs of the family Corambidae implications for the systematics of . (Gastropoda, Opisthobranchia) Part 1. Taxonomy. BMC Evolutionary Biology 10: 323. Zoologichesky Zhurnal 73: 1–15. Kawaguti S, Baba K (1959). A preliminary note on a two- Martynov AV (1994b). Materials for the revision of the valved sacoglossan gastropod, Tamanovalva limax, n. nudibranch molluscs of the family Corambidae gen., n. sp., from Tamano, Japan. Biological Journal (Gastropoda, Opisthobranchia). Part II. The Origin. Okayama University 5: 177–184. Zoologicheskiy Zhurnal 73: 36–43. Kluge AG (1985). Ontogeny and phylogenetic systematics. Martynov AV (1995). Materials for the revision of the Cladistics 1: 13–27. nudibranch family Corambidae (Gastropoda, Korotkova GP (1979). Origin and evolution of the ontogeny. Opisthobranchia). Communication 2. Origin of the Leningrad: Leningrad State University Press. 296 pp. Corambidae. Hydrobiological Journal 31: 59–66. Kumar S (2010). Statistics and truth in phylogenomics. Martynov AV (1999). Nudibranch molluscs of North-West Molecular Phylogenetics (MolPhy-2) International part of the Sea of Japan with discussion on taxonomy Conference, Moscow State University, 18-21 May 2010. and phylogeny of the order Nudibranchia. Ph. D. thesis. LaForge N, Page LR (2007). Development in Berthella Zoological Institute RAS, St. Petersburg. 424 p. californica (Gastropoda: Opisthobranchia) with Martynov AV (2000). On the taxonomic placement of comparative observations on phylogenetically relevant the genus Echinocorambe Valdés et Bouchet, 1998 larval characters among nudipleuran opisthobranchs. (Gastropoda: Nudibranchia). XIV (5) All-Russian Invertebrate Biology 126: 318–334. Malacological Conference, St. Petersburg, pp. 50–52. Lartillot N, Philippe H (2008). Improvement of molecular Martynov AV (2010). Toward the synthesis of taxonomy, phylogenetic inference and the phylogeny of Bilateria. ontogeny, and phylogenetics: a new concept of Philosophical Transaction of the Royal Society London ontogenetic systematics. Moscow University Biological 363: 1463–1472. Sciences Bulletin, Allerton Press. 65: 145–147. Levit GS, Hoßfeld U, Olsson L (2004). The integration Martynov AV, Schrödl M (2008). The new Arctic side- of Darwinism and Evolutionary Morphology: gilled sea slug genus Boreoberthella (Gastropoda, Alexej Nikolajevich Sewertzoff (1866–1936) and the Opisthobranchia): Pleurobranchoid systematics and developmental basis of evolutionary change. Journal of evolution revisited. Polar Biology 32: 53–70. Experimental Zoology (Mol Dev Evol). 302B: 343–354. Martynov AV, Korshunova TA, Sanamyan NP, Sanamyan Linnaeus C (1758). Systema naturæ per regna tria naturæ, KE (2009). Description of the first cryptobranch secundum classes, ordines, genera, species, cum onchidoridid Onchimira cavifera gen. et sp. nov. and characteribus, differentiis, synonymis, locis. Editio of three new species of the genera Adalaria and

221 ALEXANDER V. MARTYNOV

Onchidoris (Nudibranchia: Onchidorididae) from Columbia University Press, pp. 117-136. Kamchatka waters. Zootaxa. 2159: 1–43. Mooi RD, Gill AC (2010). Phylogenies without Martynov AV, Schrödl M. In press (accepted). Phylogeny synapomorphies— a crisis in fish systematics: time to and evolution of corambid nudibranchs (Mollusca: show some character. Zootaxa 2450: 26–40. Gastropoda). Zoological Journal of the Linnean Society. Müller F (1864). Für Darwin. Leipzig: Verlag von Wilhelm Martynov AV, Brenzinger B, Hooker MYu, Schrödl Engelmann. 91 S. M (2011). 3D Anatomy of a new tropical Peruvian Mwinyi A, Bailly X, Bourlat SJ, Jondelius U, Littlewood nudibranch gastropod species, Corambe mancorensis, DTJ, Podsiadlowski L (2010). The phylogenetic position and novel hypothesis on dorid gill ontogeny and of Acoela as revealed by the complete mitochondrial evolution. Journal of Molluscan Studies 77: 129–141. genome of Symsagittifera roscoffensis. BMC Mayr E (1963). Animal species and evolution. Cambridge: Evolutionary Biology 10:309. Harvard University Press. 797 p. Needham J (1933). On the dissociability of the fundamental McNamara KJ (1986). A guide to the nomenclature of process in ontogenesis. Biological Review 8: 180–223. heterochrony. Journal of Paleontology 60: 4–13. Nei M, Kumar S (2000). Molecular evolution and McNamara KJ (1997). Shapes of time: the evolution of phylogenetics. Oxford University Press. 333 p. growth and development. Baltimore: Johns Hopkins Nelson GJ (1978). Ontogeny, phylogeny, paleontology and Univ. Press. 322 pp. the biogenetic law. Systematic Zoology 27: 324–345. Millen SV (1985). The nudibranch genera Onchidoris Odhner NH (1934). The Nudibranchiata. British Antarctic and Diaphorodoris (Mollusca, Opisthobranchia) in the (“Terra Nova”) Expedition, 1910. British Museum northeastern Pacific. The Veliger 28: 80–93. (Natural History) Natural History Report. Zoology. 7: Millen SV, Martynov AV (2005). Redescriptions of 229–310. the nudibranch genera Akiodoris Bergh, 1879 and Odhner NH (1939). Opisthobranchiate Mollusca from the Armodoris Minichev, 1972 with description of a western and northern coasts of Norway. Det Kongelige new species of Akiodoris and description of the new Norske Videnskabernes Selskabs Skrifter, pp. 1–93. family Akiodorididae. Proceedings of the Californian Odhner NH in Franc A (1968). Sous-classe des Academy of Science: 1–22. opisthobranches, pp. 608-893. In: E. Fischer, Andre Minelli A (2003). The development of animal form. Franc, Micheline Martoja, G. Termier, & H. Termier. Ontogeny, morphology and evolution. Cambridge: Traite de zoologie. Anatomie, systematique, biologie. Cambridge University Press. 323 pp. Tome V, 1083 pp. Minelli A (2009). Phylo-evo-devo: combining phylogenetics Olsson L, Levit GS, Hoßfeld U (2010). Evolutionary with evolutionary developmental biology. BMC developmental biology: its concepts and history Biology. 7: 36. with a focus on Russian and German contributions. Minichev YuS (1969). The organization of Bathydoris vitjazi Naturwissenschaften. 97: 951–969. sp. n. and the problem about the origin of nudibranch Pesenko YuA (1989). Methodological analysis of the molluscs (Opisthobranchia Nudibranchia). Vestnik systematics. I. Problem formulation, main taxonomic Leningradskogo Gosudarstvennogo Universiteta 21: schools. Proceedings of Zoological Institute, St. 51-58. Petersburg 206: 8–119. Minichev YuS, Starobogatov YaI (1979). Patterns Philippe H, Brinkmann H, Lavrov DV, Littlewood DTJ, of the reproductive systems and systematics of Manuel M et al. (2011a). Resolving difficult phylogenetic Opisthobranchia. Molluscs, main results and questions: Why more sequences are not enough. PLoS perspectives of research. Leningrdad: Nauka, pp.16–21. Biol 9(3): e1000602. Mishler BD (1988). Relationships between ontogeny and Philippe H, Brinkmann H, Copley RR, Moroz LL, Nakano phylogeny, with reference to bryophytes. In C.J. H, Poustka AJ, Wallberg A, Peterson KJ, Telford MJ Humphries (ed.), Ontogeny and systematics. N.Y.: (2011b). Acoelomorph flatworms are deuterostomes

222 FROM “TREE-THINKING” TO “CYCLE-THINKING”: ONTOGENETIC SYSTEMATICS OF NUDIBRANCH MOLLUSCS

related to Xenoturbella. Nature. 470: 255–258. Academy of Science of USSR Publishing house. 144 pp. Pola M, Cervera L, Gosliner TM (2007). Phylogenetic Schmalhausen II (1969). Problems of the Darwinism. relationships of Nembrothinae (Mollusca: Doridacea: Second edition. Leningrad: Nauka. 493 p. Polyceridae) inferred from morphology and Schmekel L (1985). Aspects of evolution within the mitochondrial DNA. Molecular Phylogenetics & opisthobranchs. In: Trueman E.R., Clark M.R. (eds.), Evolution 43: 726–742. Mollusca. Vol. 10. Evolution. L.: Academic Press, pp Pola M, Gosliner TM (2010). The first molecular phylogeny of 221–267. cladobranchian opisthobranchs (Mollusca, Gastropoda, Schrödl M, Neusser T (2010). Towards a phylogeny and Nudibranchia). Mol Phylogenet Evol. 56: 931-941. evolution of Acochlidia (Mollusca: Gastropoda: Raff RA, Kaufman TC (1983). Embryos, genes, and evolu- Opisthobranchia). Zoological Journal of the Linnean tion: the developmental-genetic basis of evolutionary Society 158: 124 –154. change. Macmillan Publishing. 395 pp. Schrödl M, Wägele H, Willan RC (2001). Taxonomic Raff RA, Field GJ, Olsen SJ, Giovannoni DJ, Lane MT, redescription of the Doridoxidae (Gastropoda: Ghiselin MT, Pace NR, Raff EC (1989). Metazoan Opisthobranchia), an enigmatic family of deep water phylogeny based on analysis of 18S ribosomal RNA. nudibranchs, with discussion of basal nudibranch In: Fernholm B., Bremer K., Jörnvall H. (eds.). The phylogeny. Zoologischer Anzeiger 240: 83–97. Hierarchy of life. Molecules and morphology in Scotland RW, Olmstead RG, Bennett JR (2003). Phylogeny phylogenetic analysis. Amsterdam: Excerpta Medica/ reconstruction: The role of morphology. Systematic Elsevier. 247–260 p.p. Ramirez MJ, Coddington JA, Biololgy 52: 539–548. Maddison WP et al. (2007). Linking of digital images Sedgwick A (1909). The influence of Darwin on the study to phylogenetic data matrices using a morphological of animal embryology. Seward A.C. (ed.). Darwin and ontology. Systematic Biology 56: 283–294. Modern Science. Cambridge: Cambridge University Remane A (1955). Morphologie als Homologienfosrchung. Press, pp. 171–184. Verhandl. Deutsch. Zool. Gesel. 18: 159–183. Severtsov AN (1912). Essays on the theory of evolution: Ridley M (1986). Evolution and classification. The ontogeny and evolution. Kiev: Tipogr. Imp. Univ., Sv. reformation of cladism. L., N.Y. 201 pp. Vladimira. 300 pp. Roginskaya IS (1972). Calycidoris guentheri (Gastropoda, Severtsov AN (1931). Morphologische Gesetzmassigkeiten Nudibranchia). Taxonomy and distribution. der Evolution. Jena: G. Fischer Verlag. 371 S. Zoologicheskii Zhurnal 51: 913–918. Shatalkin AI (2005). Molecular phylogenetics – a Rudman WB (1998). Suborder Doridina. In: Beesley, P.L., revolutionary breakthrough in the systematics. In Ross, G.J.B. & Wells, A (eds), Mollusca - The Southern Vorobieva E.I., Striganova B.R. (eds.), Evolutionary Synthesis. Fauna of Australia, Vol 5. CSIRO Publishing: factors of the formation of diversity of animals. Melbourne. 990-1001 pp. Moscow: KMK Scientific Press, pp. 30-42. Schimkewitsch W (1906). Über die Periodizität in den Shubin N, Tabin C, Carroll S (1997). Fossils, genes and the System der Pantopoda. Zool. Anz. 30: 1–22. evolution of animal limbs. Nature. 388: 639–648. Schimkewitsch W (1909). Nochmals über die Periodizität Shubin N, Tabin C, Carroll S (2009). Deep homology and the in dem System der Pantopoden. Zool. Anz. 34: 1–13. origins of evolutionary novelty. Nature. 457: 818–823. Schindewolf OH (1936). Paläontologie, Entwicklungslehre Smirthwaite JJ, Rundle SD, Spicer JI (2009). The use of und Genetik. Kritik und Synthese. Berlin. 108 S. developmental sequences for assessing evolutionary Schindewolf OH (1950). Grundfragen der Paläontologie. change in gastropods. American Malacological Bulletin Geologische Zeitmessung – Organische 27: 105–111. Stammesentwicklung – Biologische Systematik. 506 S. Sokal RR, Sneath PHA (1963). Principles of numerical Schmalhausen II (1938). Organism as whole in the individual taxonomy. San-Francisco and London: W.H. Freeman and historical development. Moscow–Leningrad: and Co. 359 pp.

223 ALEXANDER V. MARTYNOV

Struck TH (2007). Data congruence, paedomorphosis and Journal of Molluscan Studies 55: 343– 364. salamanders. Frontiers in Zoology. 4: 22. Wägele JW (2005). Foundations of phylogenetic systematics. Tardy JP (1970). Contribution a l’étude des métamorphoses München: Verlag Dr. Friedrich Pfeil. 365 p. chez les . Annales des Sciences Naturelles Wägele JW, Letsch H, Klussmann-Kolb A, Mayer C, Misof (Zoologie), Paris 12: 299–370. B, Wägele H (2009). Phylogenetic support values are Thiele J (1931). Handbuch der Systematischen not necessarily informative: the case of the Serialia Weichtierkunde. Band 1. 778 pp. hypothesis (a mollusk phylogeny). Frontiers of Zoology Thompson TE (1958). The natural history, embryology, 6: 12. larval biology and post-larval development of Wägele H, Willan R (2000). Phylogeny of the Nudibranchia. Adalaria proxima (Alder and Hancock) (Gastropoda Zoological Journal of the Linnean Society 130: 83–181. Opisthobranchia). Philosophical Transactions of the Weston PH (1988). Indirect and direct methods in systematics. Royal Society of London (B) 242: 1-58. In Ontogeny and systematics, C. J. Humphries, (ed.). Thompson TE (1962). Studies on the ontogeny of Tritonia N.Y.: Columbia Univ. Press, 27–56. hombergi Cuvier (Gastropoda Opisthobranchia). Phil. Wiens JJ (2004). The role of morphological data in phylogeny Trans. R. Soc. Lond. Ser. B. 245: 171–218. reconstruction. Systematic Biology 53: 653–661. Tsubokawa R, Okutani T (1991). Early life history Wiens JJ, Bonnet RM, Chippindale PT (2005). Ontogeny of Pleurobranchaea japonica Thiele, 1925 discombobulates phylogeny: paedomorphosis and (Opisthobranchia: Notaspidea). Veliger. 34: 1–13. higher-level salamander relationships. Systematic Usuki I (1967). The direct development and the single Biology 54: 91–110. cupshaped larval shell of a nudibranch, Glossodoris Wiley EO (1981). Phylogenetics: the theory and practice of sibogae (Bergh). Sci. Rep. Niigata Univ. Ser. D (Biol.). phylogenetic systematics. N.Y. 439 pp. 4: 75–85. Wiley EO, Siegel-Causey D, Brooks DR, Funk VA (1991). Valdés Á (2002a). A phylogenetic analysis and systematic The Compleat Cladist: a primer of phylogenetic revision of the cryptobranch dorids (Mollusca, procedures. Lawrence, Kansas: Museum of Natural Nudibranchia, Anthobranchia). Zoological Journal History, Dyche Hall, The University of Kansas. 158 pp. Linnean Society 136: 535–636. Woese CR, Goldenfeld N (2009). How the microbial world Valdés Á (2002b). Phylogenetic systematics of ‘Bathydoris’ saved evolution from the Scylla of molecular biology and s.l. Bergh, 1884 (Mollusca, Nudibranchia), with the the charybdis of the modern synthesis. Microbiology description of a new species from New Caledonian deep and Molecular Biology Reviews 73: 14–21. waters. Canadian Journal of Zoology 80: 1084–1099. Valdés Á. (2004). Phylogeography and phyloecology of dorid nudibranchs (Mollusca, Gastropoda). Biological Journal of the Linnean Society 83: 551-559. Vogt L (2009). The future role of bio-ontologies for developing a general data standard in biology: chance and challenge for zoo-morphology. Zoomorphology 128: 201–217. Wägele H (1989a). Die Gattung Bathydoris Bergh, 1884 (Gnathodoridacea) im phylogenetischen System der Nudibranchia (Opisthobranchia, Gastropoda). Zeitschr. Zool. Syst. Evol. 27: 273–281. Wägele H (1989b). A revision of the Antarctic species of Bathydoris Bergh, 1884 and comparison with other known bathydorids (Opisthobranchia, Nudibranchia).

224