At the root of animal diversity: Evolvability etc. 21

Сборник трудов Зоологического музея МГУ им. М.В. Ломоносова Archives of Zoological Museum of Lomonosov Moscow State University Том / Vol. 54 Cтр. / Pр. 21–41

AT THE ROOT OF ANIMAL DIVERSITY: EVOLVABILITY, , AND HOMOLOGY Alessandro Minelli Department of Biology, University of Padova; [email protected]

We cannot satisfactorily explain the origin of biological diversity wi- thout considering that can only test the that the developmental system is able to produce. In other terms, we must con- sider its evolvability, the scenario of possible, likely and (apparently at le- ast) forbidden changes that a species’ developmental system can accept. Evolvability is sometimes systemic, but more frequently modular, in terms of morphological organization but also of the temporal articulation of de- velopmental processes, as shown by . This has important con- sequences on our possible approach to homology. The traditional “all-or- nothing” notion of homology must be replaced by a combinatorial approach that abandons the Owenian requirement of conservation of sameness. Conservation of homology across evolutionary transitions between en- vironmentally controlled and genetically encoded traits encourages ap- proaching the issue of homology from the perspective of evolutionary de- velopmental biology, including the appreciation of the multiplicity of paths along which inheritance of traits is carried across generations.

У ИСТОКОВ РАЗНООБРАЗИЯ ЖИВОТНЫХ: ЭВОЛЮИРУЕМОСТЬ*, МОДУЛЬНОСТЬ И ГОМОЛОГИЯ Алессандро Минелли

Мы не можем удовлетворительно объяснить происхождение биоло- гического разнообразия без учёта того, что естественный отбор может только «испытывать» фенотипы, которые способна производить онто- генетическая система. Другими словами, мы должны рассматривать её эволюируемость, а именно совокупность возможных, вероятных

© A. Minelli, 2016. *Прим. ред. В рунете нам не удалось найти русскоязычный эквивалент термина «evolvability», означающего способность (склонность) эволюировать. Калька «эволюбельность», аналогичная при- жившимся «вариабельности» или «мутабельности», не слишком благозвучна. Термин «эволюируе- мость» здесь использован по аналогии с принятыми «мутируемостью» и «растворимостью», озна- чающими способность мутировать и растворяться, соответственно. 22 A. Minelli

и (по крайней мере явно) запрещённых изменений, которые может включать онтогенетическая система вида. Рассматриваемая с точки зрения морфологической организации, а также различных временны́х членений процессов онтогенеза, эволюируемость может быть систем- ной (затрагивает весь организм), чаще же она модульная, иллюстра- цией чего служит гетерохрония. Это имеет важное значение для раз- работки возможных подходов к трактовке гомологии. Традиционное понимание гомологии по принципу «всё или ничего» должно быть заменено комбинаторным подходом, подразумевающим отказ от оуэ- новского требования сохранения тождественности. Сохранение гомо- логии между контролируемыми средой и генетически кодируемыми признаками при их эволюционных изменениях побуждает исследо- вать гомологию с точки зрения эволюционной биологии развития, включая признание множественности путей, по которым осуществля- ется наследование признаков в цепочке поколений.

1. Introducing evo-devo pecially geographical) barriers is very often into comparative biology the cause of the origin of new species, but the Our approach to biodiversity necessar- latter is not necessarily associated with the ily includes comparisons. Comparisons of of new . For example, morphological features, of genes and perhaps many genera of terrestrial animals with poor of , of life cycle characteristics, of dispersal ability, such as fl ightless ground- behavioural schedules. Comparisons targeted living insects or land snails, include large to the identifi cation of traits shared by differ- or very large numbers of allopatric species ent species, or suggesting their phylogenetic confi ned either to individual islands or to relationships, as well as to diagnostic traits individual mountaintops or to isolated cal- differentiating them. Comparisons leading careous outcrops surrounded by substrates of to the untangling of the alpha-diversity in a different nature. In terms of , rapid biological community or an ecosystem, but and richly ramifi ed patterns of speciation are also to an appreciation, albeit necessarily a often explained (tentatively at least) in terms subjective one, of disparity, i. e. of the extent of adaptations made possible by the presence, of morphospace occupied by the representa- in a megadiverse clade, of some “key” trait tives of a taxonomic group within that com- the of which would have opened munity or ecosystem. With a further step, the possibility to undergo abundant specia- the analysis of biodiversity proceeds in the tion. Globally, this explanation of biological direction of functional ecology, or towards a diversity is looked for in the usual terms of historical and causal analysis of biodiversity, evolution as the survival of the fi ttest. perhaps in the traditional terms of Neodar- However, this is only one side of the coin: winian , that is, looking the other side is the arrival of the fi ttest. Natu- for processes of adaptation and speciation. ral selection can only test the fi tness of the Not necessarily these two aspects of evo- phenotypes that the developmental system lution — adaptation and speciation — are is able to produce (de Vries, 1905). Investi- tightly coupled together. Prolonged inter- gating the arrival of the fi ttest is one of the ruption of gene fl ow caused by external (es- qualifying targets of evolutionary develop- At the root of animal diversity: Evolvability etc. 23 mental biology, or evo-devo (e. g., Gilbert tives of which were until recently thought to et al., 1996; Alonso, 2008; Wagner, 2011). have either 21 or 23 pairs of legs, a species with 39 or 43 pairs of legs, Scolopendropsis 2. Evolutionary transitions duplicata, has evolved, probably derived in Remarkably, the landscape of the possible a single step from an ancestor very similar to morphological transitions from an existing Sc. bahiensis, a species with either 21 or 23 to a novel is neither isotropous pairs of legs (Minelli et al., 2009). Back in the nor necessarily continuous (e. g. Theißen, evolutionary history of the Chilopoda, a tran- 2006, 2009; Minelli et al., 2009). Evolution sition from 15 pairs of legs, the plesiomor- offers examples of discontinuities (Frazzetta, phic number of leg pairs within centipedes, 2012) that cannot be explained as the long- to 21 pairs was an innovation of the lineage term result of a prolonged accumulation of leading to the extant Scolopendromorpha. small adaptive changes. Within the latter, a further increase in the Morphological discontinuities that do not number of leg-bearing segments occurred appear to be bridgeable by a series of micro- repeatedly, always with discontinuous tran- are, for example, the torsion of the sitions: from 21 to 23 in the Scolopocryp- visceral sac that distinguishes the gastropods topidae (Vahtera et al., 2013), from 21 to 23 from all other molluscs, or the sudden transi- and again to 39 and 43 in Scolopendropsis tion from bilateral symmetry to directional (Minelli et al., 2009). From an ancestor be- asymmetry in the fl atfi shes. In the evo-devo longing to the stem-group Scolopendromor- literature, the most popular case is the pecto- pha (that is, likely provided with 21 or 23 ral girdle of turtles, which is encased within pairs of legs) evolved also the lineage leading the ribs, rather than external to them as in all to the extant Geophilomorpha, within which other tetrapods, a change that Gilbert et al. the leg pair number was fi rst in the order (2001) and Rieppel (2001) did not hesitate of 41, 43 or 45, but subsequently diverged to describe as saltational. between 27 and 191 leg pairs. Besides the Many phenotypes, often very similar to obvious saltations, the whole evolutionary existing and even successful ones, are not radiation of centipedes is characterized by present in nature, not because of a very low a remarkable constraint: the number of leg fi tness, but because the existing developmen- pairs is always limited to odd values (Minelli, tal systems cannot be modifi ed easily, or at Bortoletto, 1988). all, in such a way as to produce them. For These examples suggest that phenotypic example, there is no reason to expect that a evolution is shaped in part by the constraints scolopendromorph centipede with 22 pairs under which developmental systems operate of legs would be functionally impaired, com- (Arthur, 2004). We are therefore prompted to pared to those possessing either 21 or 23 pairs investigate evolvability, that is, the scenario of legs, nevertheless not a single specimen of possible, more probable, less probable (not to say, a species) of scolopendromorph and (apparently, at least) forbidden changes with 22 pairs of legs has ever been recorded, that a species’ developmental system can ac- whereas the “neighbouring” phenotypes with cept, thus imposing a bias on the production either 21 or 23 pairs of legs are characteristic of new phenotypes. Discussing evolvability of hundreds of species each (Minelli, 2009). will bring us to discuss briefl y the modularity But this is not the whole story. Within the of animal organization as well as the com- clade Scolopendromorpha, all representa- plexity of the relationships that may link an 24 A. Minelli organism’s genotype to the phenotypes it is 394) defi nition of evolvability as “the capac- eventually able to express. Modularity is not ity of a developmental system to evolve [..] limited to the morphological organization largely [as] a function of the developmental of the animal, but extends to the temporal system’s ability to generate variation”. sequence of individual developmental pro- Potentially adaptive phenotypes do not cesses or stages along an animal’s , show up because their production is very and this brings us to heterochrony and its diffi cult, think of mammals with a number consequences for the notion of homology. of cervical vertebrae other than seven: this The two concepts of homology and number is conserved between animals with evolvability are interrelated. Brigandt (2007, so differently elongated necks as hippo and p. 710) defi nes a homologue “as a unit of giraffe. Yet, as soon as variation appears in morphological evolvability, i. e., as a part of the system, alternative phenotypes evolve an organism that can exhibit heritable pheno- rapidly, often departing amazingly from the typic variation independently of the variation original phenotype. Here are two examples. that the organism’s other homologues can un- The axoneme of cilia and fl agella of eu- dergo,” or as “a unit of heritable phenotypic karyotic cells is generally composed of two variability — a structural unit being able to central microtubules surrounded by a circle phenotypically vary in response to genetic of nine doublets (the usual 9+2 arrangement). variation”. Similar concepts, but framed in This structure is remarkably stable among more comprehensive terms, that is, in respect the vast majority of eukaryotes provided both to ontogeny and evolution, have been with cilia or fl agella, but it is far from being proposed e. g. by Laubichler (2000), New- universal. Exceptions to the rule, however, man and Müller (2000), Müller (2003, 2007), are not limited to the phenetically closest and Jamniczky (2008) (cf. Pavlinov, 2012). arrangements such as 9+1, 9+0 or 9+9+2. Departures from 9+2 have occasionally 3. Evolvability opened the way towards the evolution of Some architectural aspects are very stable an extraordinary diversity of arrangements, throughout the animal kingdom, or a sizeable most conspicuous being those in the sperm part of it, well beyond any reasonable expla- fl agellum of gall midges (Diptera: Cecid- nation in terms of adaptation. In many such omyidae), with up to 2500 doublets in As- instances, it is not natural selection that must phondylia ruebsaameni (Lanzavecchia et al., be regarded as responsible for evolutionary 1991; Mencarelli et al., 2000). stasis, but the lack of selectable variation. Another example of diversity obtained Constraints must be searched for at the level by abandoning a previously very stable phe- of developmental systems. notype is provided by the antennal articles This approach can be expressed in terms of the Coleoptera. Here, the plesiomorphic of evolvability, a term that has unfortunately number is 11, widely conserved throughout been applied by different authors to a plu- the order, despite very conspicuous vari- rality of different concepts (e. g., Dawkins, ation in the relative size and shape of the 1988; Alberch, 1991; Hansen, 2003, 2006; individual articles and thus in the overall Schlichting, Murren, 2004; Klingenberg, shape of the appendage. Nevertheless, this 2005; Wagner, 2005; Colegrave, Collins, number has been repeatedly reduced, e. g. to 2008; Pigliucci, 2008; Brookfi eld, 2009). I ten articles in 46 families, to nine in 31, to will follow here Hendrikse et al.’s (2007, p. eight in 22, and even to three articles in fi ve At the root of animal diversity: Evolvability etc. 25 families and to two articles in two families 4. Genes and homology (Minelli, 2004). However, variation in the Are there genes individually “responsi- opposite direction has been rare and mostly ble for” a given feature, as we have possibly limited to one extra article and very rarely learned from schoolbooks introducing el- extended to numbers higher than 20 (only ementary Mendelian genetics? A one-to-one in some Lampyridae, Cerambycidae and correspondence between genes and pheno- Rhipiceridae). typic features would help consolidating our With the advent of evolutionary devel- appreciation of homology relationships, but opmental biology, evolvability has taken a this path of enquiry would bring us nowhere. central role in explanations of evolutionary change and its study is even regarded as the Two problems must be acknowledged. core feature of evolutionary developmental First, there is abundant evidence of ho- biology (Hendrikse et al., 2007). Study- mologous morphological features controlled, ing evolvability has caused an increasing in different species, by nonhomologous appreciation of the complex relationships genes or networks of genes, and vice versa, linking the genotype to the phenotype (the i.e. nonhomologous morphological features so-called “genotype→phenotype map”), controlled by homologous genes (discussed which are now largely acknowledged to e. g. in Wray, Abouheif, 1998). It is remark- be mostly non-linear and far from uni- able that this mismatch between genotype form (e. g. Alberch, 1991; Wagner, Alten- and phenotype has eventually caused one of berg, 1996; West-Eberhard, 2003; Draghi, the 19th century scientists who most contrib- Wagner, 2008; Pigliucci, 2010). In simple uted to the conceptual development of com- terms, rarely, if ever, does one gene cor- parative biology to eventually take distance respond to one , and vice from the very notion of homology (de Beer, versa. As a rule, the expression of one ge- 1971; Pavlinov, 2012). ne affects a diversity of phenotypic traits Evidence from developmental genet- (ple iotropy), and indistinguishable pheno- ics can assist in identifying homologous types can be under the control of different morphological structures but does not pro- genes, or genetic cascades (convergence vide necessary nor suffi cient conditions for and/or redundancy). determining structural homology (Galis, To some extent, what we register as plei- 1999; Bolker, Raff, 2003; Minelli, 2003a; otropy is nothing but a consequence of the Pavlinov, 2012). way we describe a phenotype as a sum of Acknowledging this frequent mismatch characters: this articulation into units is per- between genes and morphology, Nielsen and haps reasonable in terms of morphology, but Martinez (2003) recognized, under the new does not necessarily correspond to distinct term of homocracy, the correspondence be- developmental processes responsible for tween organs or structures organized through the individual characters, or to the expres- the expression of the same patterning genes, sion patterns and the functions of as many irrespective of whether these structures can genes. Strictly speaking, the only uniquely be regarded or not as homologous in terms of controlled phenotype corresponding to a comparative morphology. A related concept given gene is perhaps its primary mRNA stressing the conservation throughout phy- transcript, previous to any post-transcrip- logeny of genetic networks underlying the tional editing. production of eventually diverging organs 26 A. Minelli has been suggested by Shubin et al. (2009) However, the very existence of master con- under the evocative but controversial name of trol genes is questionable. Davidson (2001, deep homology (briefl y discussed in: Minelli, p. 27), for example, disposed of the idea of Fusco, 2013a). a linear hierarchical control sequence begin- Second, hundreds and hundreds of genes ning with a hypothetical master gene describ- are differentially expressed in each body ing it as just a “fantasy of earlier days”. Evi- parts. For example, in the mouse, many genes dence suggests instead that involved in limb initiation and patterning are is controlled by complex networks of signal part of regulatory networks common to both systems and transcriptional regulators (Da- forelimbs and hindlimbs: many of them are vidson, 1993). This revised interpretation of differentially expressed between the grow- the molecular circuitry controlling morpho- ing and differentiating anterior vs. posterior genesis is certainly more realistic than the appendages, and contribute to the specifi ca- older one, based on the putative existence of tion of limb-type identity (Logan et al., 1998; master control genes. However, compared to Logan, 2003). the morphological features they presumably Admittedly, the extent to which a given control, even Davidson’s gene regulative net- body feature is controlled by any one of the works are not conservative enough to repre- genes whose expression in the Anlage of that sent the genetic or mechanistic counterparts body part is somehow different from its ex- of homologues. The evolvability of regula- pression elsewhere in the organism will not tory cascades is shown by examples where be the same for all genes. One might argue the same molecule is regulated by different that the molecular (genotypic) unit corre- genes in different species or even within the sponding to a morphological homologue is same organism (Larsen, 2003). For exam- not a gene whatsoever, but a master control ple, the gene hedgehog, which is involved in gene, that is, a gene responsible for a major establishing the antero-posterior axis of the switch in the expression of a large number of embryonic segments and in patterning the downstream genes. This fashionable concept larval imaginal discs, is controlled by bicoid was fi rst introduced by Lewis (1992) for the in Drosophila and by caudal in the beetle homeotic genes of the Bithorax complex in Tribolium (Dearden, Akam, 1999), whereas Drosophila, but was mostly championed by , whose expression is critically Gehring (for an historical perspective, see: important in fi xing segmental boundaries in Gehring, 1998). Again in Drosophila, tar- Drosophila, is regulated by paired in some geted expression of the eyeless (ey) gene can cells but by fushi tarazu in others, a few cell result in the production of ectopic eyes (e. g., diameters apart (Manoukian, Krause, 1992). on a tibia) and this result has been regarded as an experimental proof that ey is the master 5. Parallelism and convergence control gene for eye morphogenesis. Homo- Lack of selectable variation is some- logues of the Drosophila ey gene (generally times responsible for discontinuities in the known as the Pax6 genes) are involved in the occupancy of the morphospace, but biased production of eyes in metazoans as different evolvability is also involved in the opposite as a squid and a vertebrate; as a consequence, phenomenon, that is, in the occurrence of Pax6 genes have been interpreted as mas- “privileged” phenotypes evolved in multiple ter control genes in the production of eyes lineages as the effect of parallel or conver- throughout the Metazoa (Halder et al., 1995). gent evolution. In those instances, selective At the root of animal diversity: Evolvability etc. 27 advantage is likely involved, but a bias in biology, whereas modularity and correlated the landscape of evolvable forms is prob- progression are briefl y discussed below. ably much more frequent than generally In many rapid radiations, the explosion acknowledged. In recent times, strict focus of phenotypes is essentially restricted to on phylogeny reconstruction has caused ho- large variation in a well circumscribed mod- moplastic features to be simply regarded as ule. According to Klingenberg (2005, p. 6), noise contrasting the phylogenetic signal “Modules are assemblages of parts that are provided by synapomorphies, but parallel- tightly integrated internally by relatively ism and convergence deserve to be studied as many and strong interactions but relatively important evolutionary phenomena. Towards independent of one another because there the turn of the century, Moore and Willmer are only relatively few or weak interactions (1997) provided a detailed overview of the between modules”. occurrence of convergent evolution in in- Examples of extensive radiations based vertebrates; soon thereafter, Conway Morris on rapid and diversifi ed change in a single (e.g. 2003a, b, 2006) went so far as to regard module are those based on the copulatory convergence as a major feature of evolution structures, especially the male ones, of many and to acknowledge that it allows some pre- insect groups, and those of the helmintho- dictions of long-term evolutionary trends. morph millipedes (Minelli, 2015a). It is by now nearly one century since The latter case is a unique example of Vavilov (1922) proposed a law of homolo- modularity in which a tiny fraction of a long, gous variation, according to which the simi- apparently homogeneous series of modular larity of developmental pathways in related units undergoes a dramatic metamorphosis, species causes the appearance of similar vari- whereas all remaining, initially identical ants. Translated into the current language of units do not undergo ontogenetic modifi ca- evo-devo, this means that the recurrent evo- tions other than growing. Adult helmintho- lution of similar phenotypes among closely morph millipedes (a clade to which the ma- related species is suggestive of positively jority of the Diplopoda belong) have between biased evolvability of some developmental 32 and 375 pairs of legs, according to species. modules (Inge-Vechtomov, 2004). In the females, and also in male juveniles, all leg pairs are morphologically identical 6. Modularity except for the smaller size of the fi rst pair, To address evolutionary change in terms or the fi rst few pairs. New pairs of legs are of evolvability, we must identify operation- added, with a number of post-embryonic ally sensible units of change. Functional in- moults, to those already present in the pre- tegration of the phenotype must be preserved vious stage. Eventually, however, the eighth for the change to have a chance of success pair of legs, and often also the ninth, regress over evolutionary time. According to Kemp totally, to be fi nally replaced in the adult male (2016), three categories of mechanisms can by gonopods, specialized and generally very account for the maintenance of phenetic inte- complex sexual appendages used as claspers gration during the course of extensive evolu- or to transfer sperm. To stress the strict lo- tionary transition: developmental homeosta- calization of the ontogenetic changes these sis, modularity and correlated progression. I appendages undergo, the term “non-systemic will not discuss here developmental homeo- metamorphosis” has been introduced (Drago stasis, a topic that pertains to developmental et al., 2008). In all but a few genera, the go- 28 A. Minelli nopods of helminthomorph millipedes are them to prevent loss of adequate integra- by far the most diversifi ed module of these tion and therefore of fi tness. But no part can arthropods’ architecture and are therefore the change unless and until appropriate com- main morphological resource for millipede pensatory change in the functionally linked taxonomy. parts have accumulated over evolutionary Much of the species-level diversity within time […] Unlike […] modularity, correlat- Onthophagus, a huge genus of dung beetles ed progression as a process sets no limit to with close to 2000 described species, is also how much evolution can occur in a lineage, concentrated in a couple of modules, the ce- that is to say how far through morphospace phalic and prothoracic horns. These horns are it can travel”. a conspicuous morphological novelty in the evolution of which is 8. Permissive and generative also involved (e. g. Wasik, Moczek, 2011), apomorphies as discussed in the section 13. In cladistic reconstructions of phyloge- netic relationships, clades are defi ned by 7. Correlated progression apomorphies shared by their members (syna- Is evolvability always dependent on mod- pomorphies), but it is not granted if and how ularity? A number of conceptual arguments those characters may have contributed to the and empirical examples suggests that this is clade’s diversity. not the case. According to Kemp (2016, p. In this respect it is sensible to distinguish 177), “Despite the popularity of modularity between permissive and generative apomor- as an explanation of evolvability, its role […] phies (Minelli, 2015b). is necessarily limited to relatively short-term Permissive apomorphies have only an evolution. In principle this is because much indirect effect on the rate of speciation, the of the phenotype is not modular but consists latter being mainly dependent on the spe- of functional processes integrated with the cifi c geographic and ecological context in rest of the organism. Empirically, modules which the clade is evolving. For example, are demonstrably transient, with their com- many birds and insects of oceanic islands ponents changing over evolutionary time, so have reduced wings, a trait positively adap- they cannot be long-term evolutionary units”. tive in that geographic context, where stormy There is also another reason not to expect winds would severely affect the chance of that a developmental module can maintain survival of winged animals (e. g. Carlquist, its autonomy for long. In principle at least, 1965, 1974). However, the effects of fl ight- natural selection operates on organisms as lessness on speciation are clearly indirect. wholes (Kemp, 2007). If so, it is reasonable Flighlessness involves reduced vagility, thus to accept that the evolution of the different reduced gene fl ow between populations and parts of the body is subjected to correlated their eventual divergence in a classic allopat- progression (Lee, 1996; Budd, 1998; Kemp, ric scenario. Only in this very indirect sense 1999, 2007). is wing reduction or loss responsible for the Kemp (2016, p. 177) defi nes correlated remarkable species diversity of many gen- progression as “the mechanism of change by era, e. g. of rails (Rallidae) among the birds which small modifi cations to single parts of and ground beetles (Carabidae) and weevils the phenotype are acceptable because there (Curculionoidea) among the beetles, all well is enough functional flexibility between represented on oceanic islands. At the root of animal diversity: Evolvability etc. 29

In contrast, a generative apomorphy the genus Megaphragma, where 95% of the provides a clade with the access to an envi- ca. 4,600 neurons forming the brain are anu- ronmental resource positively involved in cleate (Polilov, 2012). speciation, e. g. the access to a new exclusi- A different kind of systemic change is ve food source such as in many parasites or paramorphism (Minelli, 2000, 2003b), i. e. parasitoids. the evolution of new axes initiated and pat- terned by the iteration of existing develop- 9. Systemic change mental dynamics previously responsible for The evolutionary effects of change in the the production of the main body axis — pos- developmental schedule can be either modu- sibly followed by divergence and specializa- lar or systemic. tion further ahead in evolution. The existence A fi rst group of systemic changes occur of a systemic coupling between the different by reduction, often as the effect of progenesis body axes is suggested by a large number of (Westheide, 1987): reproductive maturity is examples (Minelli, 2000), e.g. the presence reached at a stage corresponding in morphol- of segmented appendages in segmented ani- ogy to an embryonic or larval stage of their mals, whereas unsegmented animals have (if relatives. These animals lack many of the any) unsegmented appendages. In this case, parts or organs usually found in the members the systemic nature of is addi- of the group to which they belong. Reduc- tionally suggested by the fact that segmenta- tion, however, is sometimes accompanied by tion has very likely evolved independently the expression of novel traits. Examples are in arthropods, annelids and vertebrates, nev- offered by Buddenbrockia and Polypodium, ertheless all three lineages have eventually two miniaturized and morphologically very evolved segmented, rather than unsegment- unusual representatives of the Hydrozoa. ed, appendages. These small cnidarians do not exhibit any Systemic phenotypic changes do not nec- trait characteristic of a polyp or a medusa. essarily depend on large genetic differences, Buddenbrockia, a parasite of freshwater as shown by the change from left-handed to bryozoans, is worm-like, without tentacles right-handed shell, or vice versa, in some or other appendages (Jiménez-Guri et al., gastropod taxa. In the only gastropod spe- 2007), whereas Polypodium, a parasite of cies in which this phenomenon (enantiomor- sturgeon’s eggs, is an irregular mass of jelly phism) has been studied, i. e. in Lymnaea with fi nger-like projections (Raikova et al., stagnalis, shell’s chirality depends on a sin- 1994). gle maternally inherited factor (reviewed in: Other systemic morphological transi- Asami et al., 2008). Yet, according to Gitten- tions are based on evolutionary changes in berger (1988), inversion of chirality has con- the structure of the cells of which the whole tributed to speciation e. g. in Partula, a genus animal is formed. This is the case of the Lo- including numerous (about 150 species) and ricifera, minuscule but anatomically complex showy snails from the islands between New metazoans formed by a high number of cells Guinea and French Polynesia. of extremely small size: these are the only an- imal cells known not to possess mitochondria 10. Modular vs. systemic evolvability (Danovaro et al., 2010). Not less dramatic, along the life cycle but modular rather than systemic, is the case Along the animal’s life cycle, some stag- of the minuscule trichogrammatid wasps of es, or some periods, are more conservative 30 A. Minelli than others. A high degree of conservation alized parts of the phenotype”. The legitima- has been often claimed to extend to the ma- cy of a homology of process is quite largely jority of the members of a given phylum, to acknowledged (e. g., Laubichler, 2000; Gil- such extent that a phylotypic stage can be bert, Bolker, 2001; Scholtz, 2005; Pavlinov, recognized, in arthropods and vertebrates at 2012) although these authors’ statements of least (Sander, 1983; Slack et al., 1993; Raff, principle are seldom accompanied by actual 1994; Galis and Metz, 2001). Richardson examples, not to say by demonstrations of (1995), however, rightly claimed that indi- the heuristic importance of process homol- vidual stages, per se, are actually less con- ogy. More or less explicitly, however, pro- servative than the phylotypic concept would cess homology is implied in the identifi cation imply, and therefore suggested to speak of of heterochronies. This can be seen even in a phylotypic period, rather than phylotypic the crudest form of heterochrony, the tradi- stage, at least in the case of vertebrates. tional growth heterochrony (e. g., de Beer, The existence of a phylotypic stage (but 1930, 1940; Gould, 1977) where the tem- also, to some extent, the existence of a phy- poral deployment of somatic development lotypic period) suggests a degree of temporal was contrasted with the temporal course of modularity of evolvability. In other instances, development towards sexual maturity. How- however, evolvability is ontogenetically sys- ever, it must be acknowledged that somatic temic, that is, it affects most of the life cycle. development and sexual maturation are both For example, in the Cycliophora, minuscule far too complex to legitimately qualify as animals that live on the appendages of the developmental modules. Norwegian lobster (Nephrops norvegicus), Very different, and actually cognate to our the whole life cycle is represented by an un- discussion of developmental modularity, is usual sequence of unusually shaped stages the more recent approach to heterochrony, for some of which no term was available in currently known as sequence heterochrony zoology, previous to the recent discovery of (e. g., Smith, 1997, 2001, 2002, 2003; Vel- this phylum; as a consequence, new terms hagen, 1997; Bininda-Emonds et al., 2007). such as the Pandora larva and the Prometheus Here, a number of developmental processes larva were introduced (Obst, Funch, 2003). are singled out and the temporal schedule according to which these processes begin 11. Heterochrony or end is compared between two or more The modularity of developmental pro- animal species. Basic condition to this kind cesses legitimates searching for homology of comparison is the identity (operational at between modules. Indeed, acknowledging least) of individual developmental processes a degree of individuality of developmental among the species compared. processes is the ontological background of Quite different from homology of devel- Wagner’s biological concept of homology. opmental processes is the homology of de- In Wagner’s (1989a, p. 62) original formu- velopmental stages that Scholtz (2005, 2008) lation, “[s]tructures from two individuals or recognizes as morphologically constrained from the same individual are homologous if and independently evolving units and thus they share a set of developmental constraints, as legitimate units of comparison, but this caused by locally acting self-regulatory perspective is questionable. Bininda-Emonds mechanisms of organ differentiation. These et al. (2003) presented quantitative evidence structures are thus developmentally individu- against the existence of a strictly defi ned At the root of animal diversity: Evolvability etc. 31 phylotypic stage in vertebrate development an “all-or-nothing” relationship, as tradi- and a comparative review on the periodiza- tionally accepted. Sometimes, however, the tion of arthropod post-embryonic develop- “sameness” of homologues is explicitly men- ment (Minelli et al., 2006a) suggested that tioned: “There are numerous examples of developmental stages are not necessarily corresponding characters between species for conserved in evolution, especially when dis- which it is hard to escape the conclusion that tantly related taxa are compared, but some- organisms from different species are clearly times even between species classifi ed in the composed of the same building blocks, such same genus. as heads, limbs and brains” (Wagner, 2014, The individuality of post-embryonic de- p. 40). These are the body parts we can trace velopmental stages is often blurred in deca- as homologous in comparing one species to pods crustaceans (Minelli and Fusco, 2013b). another: “Any character that can be homolo- For example, the very short larval stage of gized is assumed to have continuity in terms Upogebia savignyi is a kind of “advanced of its existence in a lineage of descent, as zoea” with several pairs of appendages like well as persistence of differences from other those of the adult (Gurney, 1937). The rea- parts of the body (individuality)” (Wagner, sons for classifying a developmental stage 2014, p. 42–43). There are problems also (or phase) as larval (e. g., megalopa) or post- with the appeal to a specifi c “lineage of de- larval (e. g., zoea) are sometimes completely scent” because continuity through descent arbitrary. For example, in the shrimp of the is, in Darwin’s words, “common descent genus Macrobrachium, Shokita (1977) dis- with modifi cation”. The idea that characters tinguished a “megalopal phase” from a sec- can “remain themselves” throughout an in- ond “zoeal phase” despite the fact that the defi nite number of evolutionary modifi ca- “megalopa” exhibits some zoeal characters tion suggests an idealistic interpretation of combined with many more postlarval ones. how organisms evolve (Minelli et al., 2006b; Minelli, Fusco, 2013a). 12. Character individuality The problem becomes more critical when and the emergence of novelties the so-called evolutionary novelties are in- Sixteen years before the publication of volved. Müller and Wagner (1991, p. 243) Darwin’s Origin, Owen (1843, p. 379) de- defi ned a morphological novelty as “a struc- fi ned homologue as “the same organ in dif- ture that is neither homologous to any struc- ferent animals under every variety of form ture in the ancestral species nor homonomous and function”. Nowadays, in a scientific to any other structure of the same organism”. context dominated by an evolutionary per- The same authors (Müller, Wagner 2003, p. spective on life, it seems diffi cult that we 218–219) redefi ned evolutionary innovation can still be satisfi ed with the ill-defi ned and as “a specifi c class of phenotypic change that subjective criterion of “sameness” to which is different from adaptive modifi cation [such Owen appealed. Nevertheless, the notion as] the origin of new body parts [or] major that homologous features can be recognized organizational transitions” and distinguished as “the same” survives, in different ways, in as novelties those innovations that “introduce many current approaches to the problem of new entities, units, or elements into pheno- homology. The assumption of “sameness” typic organization”. is mainly implicit, although arguably obli- Eventually, the line of arguments appar- gate, so long as homology is conceived as ently follows this path. First, there are body 32 A. Minelli parts, or features, among which there is ho- working on complex animal structures. The mology; in other terms, homology, despite mainstream attitude towards the issue of all diffi culties to recognize it in practice, is homology still rotates around the historical accepted as given. If homology can be predi- concept of homology: “homologous features cated of structures, or features, then there (or states of features) in two or more organ- must be a way in which these structures, or isms are those that can be traced back to the features, can be predicated to be “the same”. same feature (or state) in the common an- The problem then is, on which foundation cestor of those organisms” (Mayr, 1969, p. this sameness can be predicated. Eventually, 85). This was reformulated by Bock (1974, this question has been answered in three dif- p. 881) in the following terms: “Features (or ferent ways: (i) in terms of “universal laws” conditions of a feature) in two or more organ- of form, (ii) as the product of common an- isms are homologous if they stem phyloge- cestry, or (iii) in terms of proximal causes netically from the same feature (or the same responsible for the emergence of conserved condition of the feature) in the immediate developmental modules. common ancestor of these organisms)”, a A search for universal laws of form has defi nition that opens the door to a revisitation surfaced several times in the history of biol- of the historical concept of homology in con- ogy, at least since the time of Wilhelm Roux’s sequent phylogenetic terms (Hennig, 1966). developmental mechanics (Entwicklungs- Other researchers have been searching in- mechanik) (cf. Goodwin, 1977). In the last stead for a proximal-cause concept of homol- decades, it has shown up again, in two dif- ogy (this term was introduced by: Minelli, ferent forms at least. On the one side, in the Fusco, 2013a), perhaps in terms of conti- abstract, formalized terms of process struc- nuity or commonality of information (e. g., turalism (Resnik, 1994; Webster, Goodwin, Osche, 1973, 1982; van Valen, 1982; Roth, 1996); on the other, in terms of the physico- 1984, 1988; Minelli, Peruffo, 1991; Minelli, chemical properties of living matter. The 1996). Eventually, Wagner (1989a,b) intro- latter perspective has been championed by duced the so-called biological concept of Stuart Newman, who has suggested that in early stages of the evolution of metazoans the homology, with the defi nition quoted above. shapes of the emerging multicellulars were The underlying concept of developmental essentially determined by mechano-elastic individualization has been revisited in subse- properties, arguably suffi cient to produce quent papers (Wagner, Misof, 1993; Wagner, a set of generic forms, e. g. hollow spheres 1994), and eventually rephrased in terms of and segmented beads (Newman, Comper, independent units of developmental control, 1990; Forgacs, Newman, 2005; Newman et due to either morphogenetic or morphostatic al., 2006). constraints, although, in a later revisitation Unfortunately, these physicalist and stru- of the problem, Müller and Wagner (1996, cturalist approaches only apply to the simple p. 4) adopted a less deterministic approach, geometric structure of early embryos, but suggesting “some degree of independence the whole range of shapes studied by com- of structural homology from its genetic and parative morphology remain beyond reach. developmental makeup”. This may explain the limited audience these Of course, one may argue that homology approaches have found among anatomists, is not an “all-or-nothing” correspondence: embryologists and especially systematists in this case, no sameness is implied, and At the root of animal diversity: Evolvability etc. 33 an explanation of homology in terms of ei- (hth), three genes otherwise involved in the ther common ancestry or common proximal specifi cation of the proximal-distal axis of causes can be advocated. But this requires insect legs (Moczek, Nagy, 2005; Moczek, abandoning the Owenian requirement of Rose, 2009). Thus, the horns of these beetles, sameness, to follow instead the path of rea- while representing an evolutionary novelty, soning suggested in the next section. in the sense that similar structures were not present in their (even quite recent) ancestors 13. Towards a combinatorial and have no equivalent in many of their close approach to homology relatives, are not totally new. The legs and the horns of these beetles are historically non “Nirgends ist Neubildung, sondern nur homologous despite the involvement in their Umbildung“ — “nowhere is there new development of similar (“serial”) patterns of formation [= origin of fully new parts], expression of homologous genes. there is only transformation”. This iconic Evolutionary change is a continuous pro- characterization (von Baer, 1828, p. 156) cess based on the never ending remolding of ontogenetic change could be used also of pre-existing features or, to adopt Jacob’s to describe evolution, including the “ori- (1977) well-known metaphor, of the never gin” of evolutionary novelties. According ending tinkering with the genetic networks to West-Eberhard (2008, p. 198), a nov- that regulate and control their development. elty is indeed a “phenotypic trait that is This has induced several authors (e. g., Van new in composition or context of expres- Valen, 1982; Gans, 1985; Roth, 1984; Sat- sion relative to established ancestral traits” tler, 1992, 1994; Haszprunar, 1992; Shubin, (italics mine, — A.M.). Indeed, even the Wake, 1996; Meyer, 1998; Minelli, 1998, best characterized feature is a mosaic, or a 2003a; Abouheif, 1999; Wake, 1999; Pigli- mixture, of a multiplicity of traits, some of ucci, 2001; Minelli, Fusco, 2013) to aban- which can be traced to homologous traits don the traditional “all-or-nothing” notion of remote ancestors, others are more recent of homology in favour of a different one, one, while total novelty cannot be predi- described as either partial or relative. An im- cated of any feature as a whole (Minelli, portant distinction is in order. As remarked Fusco, 2005; see also Moczek, 2008; Hall, by Endress (2011, p. 122), “a sensible evo- Kerney, 2012). It is therefore difficult to lutionary question in the detailed compari- establish where homology ends and novel- son of two parts is not by what percentage ty begins, if and when establishing that they are homologous, but in which respects bo undary makes sense at all. they are homologous”. This corresponds to A fi ne dissection of the old and new as- my suggestion (Minelli, 1998, p. 344) that pects of an evolutionary novelty has been “we must proceed, in every assessment of carried on by Armin Moczek and colleagues. homology, by specifying fi rst the structural Target of their studies were the head and pro- layer, or the developmental control, or the thoracic horns of the scarab beetles of the gene (or complex of genes), which identifi es genus Onthophagus, already mentioned in the morphological or developmental unit on section 6 above. During pre-pupal and pu- which we are focusing […] that amounts to pal stages, the development of these horns is adopting a combinatorial approach to homol- regulated through the expression of Distal- ogy (Bachmann, 1989; Minelli, 1992, 1996; less (Dll), dachshund (dac) and homothorax Haszprunar, 1992)”. 34 A. Minelli 14. Homology in the context example, 23 out of the 54 species discussed of phenotypic plasticity by Susoy et al., 2015) are dimorphic for the armature of the mouth: the two phenotypes, Phenotypic plasticity (reviewed, e. g., in the stenostome and the eurystome one, are Schlichting, Pigliucci, 1998; Pigliucci, 2001; differentially successful in exploiting differ- West-Eberhard, 2003; Fusco, Minelli, 2010) ent kinds of prey. In the vast majority of the is “a property of individual genotypes to pro- dimorphic species, this trait is phenotypi- duce different phenotypes when exposed to cally (and adaptively) plastic, the alternative different environmental conditions” (Pigli- phenotypes being preferentially expressed in ucci et al., 2006, p. 2363). It is through phe- the presence (the stenostomous one) or in the notypic plasticity that the different castes are absence (the eurystomous one) of bacteria. generally produced among social insects, but Starvation and population density also affect this applies also to the environmental deter- the relative proportions in which the two phe- mination of sex in reptiles such as the alliga- notypes are produced. But in isogenic lines tor and the origination of predator-induced of the genus Pristionchus both phenotypes morphs in water fl eas and frogs. are also expressed in the absence of any en- It has been suggested (Nijhout, 2003) that vironmental stimulus, a stochastic develop- non-adaptive or just incidentally adaptive mental property that phylogeny shows to be phenotypic plasticity is likely the primitive a derived, genus-specifi c trait. character state for most if not all traits. Even- A comparative zoologist may ask if, in tual fi xation can result either by progressive what sense or to what extent the stenosto- reduction of plasticity, thus ensuring the pro- mous (or the eurystomous) phenotypes of duction of a stable phenotype irrespective of a Pristionchus species can be regarded as environmental variation, or by evolution of a homologous to the morphologically equiva- genetic polymorphism by genetic assimila- lent phenotype expressed by a nematode tion of multiple phenotypes. However, evo- where the trait’s expression is not subjet to lution likely occurs through repeated cycles stochasticity. The same comparative zoolo- along which plasticity and genetic fi xation gist may also ask if, in what sense or to what alternate. There is arguably no evidence that extent a trait expressed under strict genetic genes — to use the terminology of Schwan- control is to be considered homologous to a der and Leimar (2011) — must necessarily morphologically equivalent trait expressed be leaders or followers in respect to environ- under environmental control by a related mentally directed change. species exhibiting plasticity for the same In addition, there is also evidence of sto- trait. Once more, to answer these questions chastic production of alternative phenotypes we need to specify the perspective (e. g., (without intermediates) even in the absence strictly morphological, or developmental, or of genotypic differences and in strictly uni- genetic-mechanistic) from which we want to form, standardized environmental condi- address them. tions. Within the diplogastrid nematodes, a Evolutionary of- conspicuous example of stochastic polymor- fers some useful suggestions, encouraging phism appears to be a condition secondarily to scrutinize evolved in a line previously showing poly- — the evolvability of morphological phenism (environmentally induced polymor- traits, of developmental processes and also phism). Many members of this family (for of inheritance systems, which are not limited At the root of animal diversity: Evolvability etc. 35 to the conventional, DNA-encoded informa- Bachmann K. 1989. Homologie bei Pflanzen. tion carried by chromosomes (Helanterä, — Zoologische Beiträge, Neue Folge, 32 Uller, 2010), (3): 479–486. Bininda-Emonds, O.R.P., Jeffery, E.J., Richard- — the modularity of developmental pro- son, M.K., 2003. Inverting the hourglass: qu- cesses, which nevertheless is never total and antitative evidence against the phylotypic sta- exclusive, and itself evolves, ge in vertebrate development. — Proceedings — the mosaic nature of virtually all mor- of the Royal Society of London, Ser. B, Bio- phological traits we are able to isolate within logical Sciences, 270 (1513): 341–346. the phenotype, and of the developmental Bininda-Emonds O.R.P., Jeffery J.E., Sánchez- stages through which the ontogeny proceeds. Villagra M.R. et al. 2007. Forelimb-hindlimb developmental timing across tetrapods. — This articulated perspective requires a BMC Evolutionary Biology, 7: 182. combinatorial approach to homology and the Bock W.J. 1974. Philosophical foundations of rejection of the traditional “all-or-nothing” classical evolutionary classification. — Sys- perspective. The latter would be suitable if te matic Zoology, 22 (4): 375–392. the object we need to compare were natural Bolker J.A., Raff R.A. 2003. Developmental ge- kinds, but the contingent, historical nature of netics and traditional homology. — BioEs- says, 18 (6): 489–494. living organisms does not fi t into this philo- Brigandt I. 2007. Typology now: homology and sophical category. developmental constraints explain evolva- Acknowledgements bility. — Biology and Philosophy, 22 (5): 709–725. I am grateful to Igor Ya. Pavlinov for in- Brookfield J.F.Y. 2009. Evolution and evolva- viting me to contribute to this volume and bility: celebrating Darwin 200. — Biology for stimulating me to familiarize with the Letters, 5 (1): 44–46. great Russian tradition in comparative mor- Budd G.E. 1998. Arthropod body-plan evolu- tion in the Cambrian with an example from phology. anomalocaridid muscle. — Lethaia, 31 (3): References 197–210. Carlquist S. 1965. Island life. A natural history Abouheif E. 1999. Establishing homology criteria of the islands of the world. New York: The for regulatory gene networks: prospects and Natural History Press/Garden City. 451 p. challenges. — Bock G.R., Cardew G. (eds). Carlquist S. 1974. Island biology. New York: Co- Homology. Chichester: Wiley. P. 207–221. lumbia Univ. Press. 660 p. Alberch P. 1991. From genes to phenotype: dy- Colegrave N., Collins S. 2008. Experimental namical systems and evolvability. — Genet- evolution: and evo- ica, 84 (1): 5–11. lvability. — Heredity, 100 (5): 464–470. Alonso C.R. 2008. The molecular biology un- Conway Morris S. 2003a. The Cambrian ‘ex- derlying developmental evolution. — Mine- plosion’ of metazoans and molecular biolo- lli A., Fusco G. (eds). Evolving pathways: gy: would Darwin be satisfied. — Interna- Key themes in evolutionary developmental tional Journal of Developmental Biology, 47 biology. Cambridge: Cambridge Univ. Press. (7/8): 505–515. P. 80 –99. Conway Morris S. 2003b. Life’s solution: inevi- Arthur W. 2004. Biased embryos and evolution. table humans in a lonely universe. Cambri- Cambridge: Cambridge Univ. Press. 233 p. dge: Cambridge Univ. Press. 464 p. Asami T., Gitternberger E., Falkner G. 2008. Conway Morris S. 2006. Evolutionary conver- Whole-body enantiomorphy and maternal gence. — Current Biology, 16 (19): R826– inheritance of chiral reversal in the pond R827. snail Lymnaea stagnalis. — Journal of He- Danovaro R., Dell’Anno A., Pusceddu A. et redity, 99 (5): 552–557. al. 2010. The first metazoa living in per- 36 A. Minelli

mane ntly anoxic conditions. — BMC Bi- Bock G.R., Cardew G. (eds). Homology. Chi- ology, 8: 30. chester: Wiley. P. 80–94. Davidson E.H. 1993. Later embryogenesis: reg- Galis, F., Metz, J.A.J. 2001. Testing the vulner- ulatory circuitry in morphogenetic fields. — ability of the phylotypic stage: on modu- Development, 118 (3): 665–690. larity and evolutionary conservation. — Davidson E.H. 2001. Genomic regulatory sys- Journal of Experimental Zoology, Part B, tems: development and evolution. San Di- Molecular and Developmental Evolution, ego: Academic Press. 261 p. 291 (2): 195–204. Dawkins R. 1988 The evolution of evolvability. — Gans C. 1985, Differences and similarities: com- Langton C.G. (ed.). Artificial life. SFI Studies parative methods in mastication. — Ameri- in the sciences of complexity. Redwood City can Zoologist, 25 (2): 291–301. (CA): Addison-Wesley. P. 201–220. Gehring W.J. 1998. Master control genes in de- de Beer G.R. 1930. Embryology and evolution. velopment and evolution. New Haven: Yale Oxford: Clarendon Press. 116 p. Univ. Press. 236 p. de Beer G.R. 1940. Embryos and ancestors. Ox- Gilbert S.F., Bolker J.A. 2001. Homologies of ford: Cla rendon Press. 108 p. process and modular elements of embryon- de Beer G.R. 1971. Homology: an unsolved pro- ic construction. — Wagner G.P. (ed.). The blem. London: Oxford Univ. Press. 16 p. character concept in evolutionary biology. de Vries H. 1905. Species and varieties, their or- San Diego: Academic Press. P. 437–456. igin by . Chicago: The Open Court Gilbert S.F., Loredo G.A., Brukman A., Burke Publishing Company. 847 p. A.C. 2001. Morphogenesis of the turtle shell: Dearden P., Akam M. 1999. Developmental evo- the development of a novel structure in tet- lution: axial patterning in insects. — Cur rent rapod evolution. — American Zoologist, 31 Biology, 9 (16): R591–R594. (4): 616–627. Draghi J., Wagner G.P. 2008. Evolution of evo- Gilbert S.F., Opitz J.M., Raff R.A. 1996. Re- lvability in a developmental model. — Evo- synthesizing evolutionary and developmen- lution, 62 (2): 301–315. tal biology. — Developmental Biology, 173 Drago L., Fusco G., Minelli A. 2008. Non-sys- (2): 357–372. temic metamorphosis in male millipede ap- Gittenberger E. 1988. Sympatric speciation in pendages: long delayed, reversible effect of snails. A largely neglected model. — Evolu- an early localized positional marker? — Fro- tion, 42 (4): 826–828. n tiers in Zoology, 5: 5. Goodwin B.C. 1977. Mechanics, fields and sta- Endress P.K. 2011. Changing views of flower tistical mechanics in developmental biolo- evolution and new questions. — Wanntorp gy. — Proceedings of the Royal Society of L., Ronse De Craene L.P. (eds). Flowers on London, Ser. B, Biological Sciences, 199 the tree of life, Cambridge: Cambridge Univ. (1136): 407–414. Press. P. 120–141. Gould S.J. 1977. Ontogeny and phylogeny. Cam- Forgacs G. Newman S.A. 2005. Biological phy- bridge (MA): The Belknap Press. 501 p. sics of the developing embryo. New York: Gurney R. 1937. Notes on some decapod Crus- Cambridge Univ. Press. 337 p. tacea from the Red Sea I. The genus Pro- Frazzetta T.H. 2012. Flatfishes, turtles, and bo- cessa. II. The larvae of Upogebia savignyi ly erine snakes: evolution by small steps or Strahl. — Proceedings of the Zoological So- large, or both? — Evolutionary Biology, 39 ciety of London, 107B: 85–101. (1): 30–60. Halder G., Callaerts P., Gehring W.J. 1995. In- Fusco G., Minelli A. 2010. Phenotypic plastici- duction of ectopic eyes by targeted expres- ty in development and evolution. — Philo- sion of the eyeless gene in Drosophila. — sophical Transactions of the Royal Society, Science, 267 (2505): 1788–1792. Ser. B, Biological Sci., 365 (1540): 547–556. Hall B.K., Kerney R. 2012. Levels of biologi- Galis, F. 1999. On the homology of structures cal organization and the origin of novelty. and Hox genes: the vertebral column. — — Journal of Experimental Zoology, Part At the root of animal diversity: Evolvability etc. 37

B, Molecular and Developmental Evolution, grímsson B., Hall B.K. (eds). Variation: a 318 (6): 428–437. central concept in biology. Burlington (MA): Hansen T.F. 2003. Is modularity necessary for Elsevier. P. 219–247. evolvability? Remarks on the relationship Lanzavecchia S., Dallai R., Bellon P.L., Af- between and evolvability. — Bio- zelius B.A. 1991. The sperm tail of a gall- Systems, 69 (2/3): 83–94. midge and its microtubular arrangement Hansen T.F. 2006. The evolution of genetic ar- studied by two strategies of image analysis chitecture. — Annual Review of Ecology (Cecidomyiidae, Diptera, Insecta). — Jour- and Systematics, 37: 123–157. nal of Structural Biology, 107 (1): 65–75. Haszprunar G. 1992. The types of homology Larsen E. 2003. Genes, cell behavior, and the and their signifi cance for evolutionary biol- evolution of form. — Müller G.B., Newman ogy and phylogenetics. — Journal of Evolu- S.A. (eds). Origination of organismal form. tionary Biology, 5 (1): 13–24. Cambridge (MA): MIT Press. P. 119–131. Helanterä H., Uller T. 2010. The Price equation Laubichler M.D. 2000. Homology in develop- and extended inheritance. — Philosophy ment and the development of the homolo- and Theoretical Biology, 2, e101. gy concept. — American Zoologist, 40 (5): Hendrikse J.L., Parsons T.E., Hallgrímsson B. 777–788. 2007. Evolvability as the proper focus of evo- Lee M.S.Y. 1996. Correlated progression and lutionary developmental biology. — Evolu- the origin of turtles. — Nature, 379 (6568): tion & Development, 9 (4): 393–401. 812–815. Hennig W. 1966. Phylogenetic systematics. Ur- Lewis E.B. 1992. Clusters of master control bana (IL): Univ. of Illinois Press. 263 p. genes regulate the development of high- Inge-Vechtomov S.G. 2004. [Block principle in er organisms. — Journal of the American the theory of evolution: Prospects and par- Medical Association, 267 (11): 1524–1531. adoxes]. — Fundamental’nye zoologiches- Logan M. 2003. Finger or toe: the molecular kie issledovaniya [Fundamental Zoological basis of limb identity. — Development, 130 Studies]. Moscow: KMK Sci Press. P. 74– (26): 6401–6410. 87. (in Russian) Logan M., Simon H.G., Tabin C. 1998. Differ- Jacob F. 1977. Evolution and tinkering. — Sci- ential regulation of t-box and ence, 196 (4295): 1161–1166. transcription factors suggests roles in con- Jamniczky H.A. 2008. Phenotypic integration trolling chick limb-type identity. — Devel- patterns support an account of homology as a opment, 125 (15): 2825–2835. manifestation of evolvability. — Evolution- Manoukian A.S., Krause H.M. 1992. Concen- ary Biology, 35 (4): 312–316. tration-dependent regulatory activities of Jiménez-Guri E., Okamura B., Holland P.W.H. the even-skipped protein in Drosophila 2007. Origin and evolution of a myxozoan embryos. — Genes & Development, 6 (9): worm. — Integrative and Comparative Bi- 1740–1751. ology, 47 (5): 752–758. Mayr E. 1969. Principles of systematic zoology. Kemp T.S. 1999. Fossils and evolution. Oxford: New York: McGraw Hill. 428 p. Oxford University Press. 284 p. Mencarelli C., Lupetti P., Rosetto M., Dallai R. Kemp T.S. 2007. The concept of correlated pro- 2000. Morphogenesis of the giant sperm ax- gression as the basis of a model for the oneme in Asphondylia ruebsaameni Ker- evolutionary origin of major new taxa. — tesz (Diptera, Cecidomyiidae). — Tissue Proceedings of the Royal Society, Ser. B, Bio- and Cell 32 (2): 188–197. logical Sciences, 274 (1618): 1667–1673. Meyer A. 1998. We are devo-evo. — Trends in Kemp T.S. 2016. The origin of the higher taxa. Genetics, 14 (12): 482–483. Oxford: Oxford Univ. Press & Chicago– Minelli A. 1992. Towards a new comparative London: Univ. of Chicago Press. 201 p. morphology of myriapods. — Berichte des Klingenberg C.P. 2005. Developmental const- naturwissenschaftlichen-medizinischen raints, modules and evolvability. — Hall- Verein Innsbruck, Supplement, 10: 37–46. 38 A. Minelli

Minelli A. 1996. Segments, body regions and Minelli A., Fusco G. 2005. Conserved vs. in- the control of development through time. — novative features in animal body organiza- Memoirs of the California Academy of Sci- tion. — Journal of Experimental Zoology ences, 20: 55–61. (Molecular and Developmental Evolution), Minelli A. 1996. Some thoughts on homology 304B (6): 520–525. 150 years after Owen’s definition. — Me- Minelli A., Fusco G. 2013a. Homology. — Kam- morie della Società Italiana di Scienze Na- pourakis K. (ed.). The philosophy of biolo- turali e del Museo Civico di Storia Naturale gy: A companion for educators. Dordrecht: di Milano, 27: 71–79. Spri nger. P. 289–322. Minelli A. 1998. Molecules, developmental mo- Minelli A., Fusco G. 2013b. Arthropod post-em- dules and phenotypes: A combinatorial ap- bryonic development, in Arthropod biolo- proach to homology. — Molecular Phyloge- gy and eolution. — Minelli A., Boxshall G., netics and Evolution, 9 (3): 340–347. Fusco G. (eds). Molecules, development, mor- Minelli A. 2000. Limbs and tail as evolution- phology. Heidelberg: Springer. P. 91–122. arily diverging duplicates of the main body Minelli A., Negrisolo E., Fusco G. 2006b. Re- axis. — Evolution & Development, 2 (3): constructing animal phylogeny in the light 157–165. of evolutionary developmental biology. — Minelli A. 2003a. The development of animal Hodkinson T.R., Parnell J.A.N., Waldren S. form: ontogeny, morphology, and evolution. (eds). Reconstructing the tree of life: Tax- Cambridge: Cambridge Univ. Press. 323 p. onomy and systematics of species rich taxa. Minelli A. 2003b. The origin and evolution of Boca Raton (FL): Taylor & Francis/CRC appendages. — International Journal of De- Press. P. 177–190. velopmental Biology, 47 (7/8): 573–581. Minelli A., Peruffo B. 1991. Developmental path- Minelli A. 2004. A segmental analysis of the ways, homology and homonomy in metame- beetle antenna. — Studi Trentini di Scienze ric animals. — Journal of Evolutionary Bi- Naturali. Acta Biologica, 81: 91–101. ology, 4 (3): 429–445. Minelli A. 2009. Forms of becoming. Prince- Moczek A.P. 2008. On the origins of novelty in ton: Princeton University Press. 228 p. Minelli A. 2015a. Morphological misfits and development and evolution. — Bioessays, 30 the architecture of development. — Serre- (5): 432–447. lli E., Gontier N. (eds). Macroevolution: ex- Moczek A.P., Nagy L.M. 2005. Diverse devel- planation, interpretation, evidence. Cham: opmental mechanisms contribute to differ- Springer. P. 329–343. ent levels of diversity in horned beetles. — Minelli A. 2015b. Species diversity vs. morpho- Evolution & Development 7 (3): 175–185. logical disparity in the light of evolutionary Moczek A.P., Rose D.J. 2009. Differential recruit- developmental biology. — Annals of Bot- ment of limb patterning genes during devel- any, in press, http://dx.doi.org/10.1093/aob/ opment and diversification of beetle horns. mcv134. — Proceedings of the National Academy of Minelli A., Bortoletto S. 1988. Myriapod meta- Sciences of the United States of America, 106 merism and arthropod segmentation. — Bi- (22): 8992–8997. ological Journal of the Linnean Society, 33 Moore J., Willmer P. 1997. Convergent evo- (4): 323–343. lution in invertebrates. — Biological Re- Minelli A., Brena C., Deflorian G. et al. 2006a. views, 72 (1): 1–60. From embryo to adult. Beyond the conven- Müller G. B. 2003. Homology: The evolution tional periodization of arthropod develop- of morphological organization. — Müller ment. — Development Genes and Evolu- G.B., Newman S.A. (eds). Origination of tion, 216 (7/8): 373–383. organismal form. Cambridge (MA): MIT Minelli A., Chagas-Júnior A., Edgecombe G.D. Press. P. 51–69. 2009. Saltational evolution of trunk seg- Müller G.B. 2007. Evo-devo: extending the evo- ment number in centipedes. — Evolution & lutionary synthesis. — Nature Reviews Ge- Development, 11 (3): 318–322. netics, 8 (12): 943–949. At the root of animal diversity: Evolvability etc. 39

Müller G.B., Wagner G.P. 1991. Novelty in evo- Surgeons. London: Longman Brown Green lution: restructuring the concept. — Annu- and Longmans. 392 p. al Review of Ecology, Evolution and Sys- Pavlinov I.Y. 2012. The contemporary concepts tematics, 22: 229–256. of homology in biology: a theoretical re- Müller G.B., Wagner G.P. 1996. Homology, Hox view. — Biology Bulletin Reviews, 2 (1): genes, and developmental integration. — 36–54. American Zoologist, 36 (1): 4–13. Pigliucci M. 2001. Phenotypic plasticity: Be- Müller G.B., Wagner G.P. 2003. Innovation. yond nature and nurture. Baltimore: John Hall B.K., Olson W.M. (eds). Keywords and Hopkins University Press. 328 p. concepts in evolutionary developmental bi- Pigliucci M. 2008. Is evolvability evolvable? — ology. Cambridge (MA): Harvard Univ. Nature Reviews, Genetics, 9 (1): 75–82. Press. P. 218–227. Pigliucci M. 2010. Genotype→phenotype map- Newman S.A., Comper W.D. 1990. Generic ping and the end of the ‘genes as blueprint’ physical mechanisms of morphogenesis metaphor. — Philosophical Transactions of and . — Development, the Royal Society, Ser. B, Biological Sci- 110 (1): 1–18. ences, 365 (1540): 557–566. Newman S.A., Forgacs G., Müller G.B. 2006. Pigliucci M., Murren C.J., Schlichting C.D. Before programs: the physical origination 2006. Phenotypic plasticity and evolution of multicellular forms. — International by . — Journal of Ex- Journal of Developmental Biology, 50 (2/3): perimental Biology, 209 (12): 2362–2367. 289–299. Polilov A.A. 2012. The smallest insects evolve Newman S.A., Müller G.B. 2000. Epigenetic anucleate neurons. — Arthropod Structure mechanisms of character origination. — and Development, 41 (1): 29–34. Journal of Experimental Zoology, Part B, Raff RA. 1994. Developmental mechanisms in Molecular and Developmental Evolution, the evolution of animal form: origins and 288 (4): 304–317. evolvability of body plans. — Bengston S. Nielsen C., Martinez P. 2003. Patterns of gene (ed.). Early life on earth. New York: Colum- expression: Homology or homocracy? — bia Univ. Press. P. 489–500. Development Genes and Evolution, 213 (3): Raikova E.V. 1994: Life cycle, cytology, and mo- 149–154. rphology of Polypodium hydriforme, a coe- Nijhout H.F. 2003. Development and evolution lenterate parasite of the eggs of acipenser- of adaptive . — Evolution & iform fishes. — Journal of Parasitology, 80 Development, 5 (1): 9–18. (1): 1–22. Obst M., Funch P. 2003. Dwarf male of Sym- Resnik D. 1994, The rebirth of rational mor- bion pandora (Cycliophora). — Journal of phology: process structuralism’s philosophy Morphology, 255 (3): 261–278. of biology. — Acta Biotheoretica, 42 (1): Osche G. 1973. Das Homologisieren als eine 1–14. grundlegende Methode der Phylogenetik. Richardson M.K. 1995. Heterochrony and the — Aufsätze und Reden der Senckenbergi- phylotypic period. — Developmental Biol- schen naturforschenden Gesellschaft, 24: ogy, 172 (2): 412–421. 155–165. Rieppel O. 2001. Turtles as hopeful monsters. Osche G. 1982. Rekapitulationsentwicklung — BioEssays, 23 (11): 987–991. und ihre Bedeutung für die Phylogenetik. Roth, V.L. 1984. On homology. — Biological Wann gilt die “Biogenetischc Grundregel”? Journal of the Linnean Society, 22 (1): 13– — Verhandlungen des naturwissenschaftli- 29. chen Vereins in Hamburg, Neue Folge, 25: Roth, V.L. 1988. The biological basis of homo- 5–31. logy. — Humphries C.J. (ed.). Ontogeny Owen R. 1843. Lectures on the comparative and systematics London/New York: British anatomy and physiology of the invertebrate Museum (Natural History)/Columbia Univ. animals, delivered at the Royal College of Press. P. 1–26. 40 A. Minelli

Sander K. 1983. The evolution of patterning me- Smith K.K. 1997. Comparative patterns of cra- chanisms: gleanings from insect embryoge- niofacial development in eutherian and me- nesis and spermatogenesis. — Goodwin B.C., tatherian mammals. — Evolution, 51 (5): Holder N., Wylie C.C. (eds). Development 1663–1678. and evolution. Cambridge: Cambridge Univ. Smith K.K. 2001. Heterochrony revisited: the Press. P. 137–159. evolution of developmental sequences. — Sattler R. 1992. Process morphology: Struc- Biological Journal of the Linnean Society, tural dynamics in development and evolu- 73 (2): 169–186. tion. — Canadian Journal of Botany, 70 (4): Smith K.K. 2002. Sequence heterochrony and 708–714. the evolution of development. — Journal of Sattler R. 1994. Homology, homeosis, and pro- Morphology, 252 (1): 82–97. cess morphology in plants. — Hall B.K. Smith K.K. 2003. Time’s arrow: heterochrony (ed.). Homology: the hierarchical basis of co- and the evolution of development. — Inter- mparative biology. San Diego: Academic national Journal of Developmental Biology, Press. P. 423–475. 47 (7/8): 613–621. Schlichting C.D., Murren C.J. 2004. Evolvabil- Susoy V, Ragsdale E.J., Kanzaki N., Sommer ity and the raw materials for adaptation. — R.J. 2015. Rapid diversification associated Cronk Q.C.B. et al. (eds). Plant adaptation: with a macroevolutionary pulse of devel- Molecular genetics and ecology. Ottawa: opmental plasticity. — eLIFE, 4, e05463. NRC Research Press. P. 18–29. http://elifesciences.org/content/4/e05463v1. Schlichting C.D., Pigliucci M. 1998. Phenotyp- Theißen G. 2009. Saltational evolution: hopeful ic evolution: a perspective. Su- monsters are here to stay. — Theory in Bio- nderland (MA): Sinauer. 387 p. sciences, 128 (1): 43–51. Scholtz G. 2005. Homology and ontogeny: Pat- Vahtera V., Edgecombe G.D., Giribet G. 2013. tern and process in comparative develop- Phylogenetics of scolopendromorph centi- mental biology. — Theory in Biosciences, pedes: can denser taxon sampling improve 124 (2): 121–143. Scholtz G. 2008. On comparisons and causes an artificial classification? — Invertebrate in evolutionary developmental biology. — Systematics, 27 (6): 578–602. Minelli A., Fusco G. (eds). Evolving path- Van Valen L.M. 1982. Homology and causes. ways. Cambridge: Cambridge Univ. Press. — Journal of Morphology, 173 (3), 305–312. P. 144 –159. Vavilov N. I. 1922. The law of homologous se- Schwander T., Leimar O. 2011. Genes as lead- ries in variation. — Journal of Genetics, 12 ers and followers in evolution. — Trends in (1): 67–87. Ecology and Evolution, 26 (3): 143–151. Velhagen W.A. Jr. 1997. Analyzing developme- Shokita S. 1977. Abbreviated metamorphosis of ntal sequences using sequence units. — Syste- landlocked fresh-water prawn, Macrobrachi- matic Biology, 46 (1): 204–210. um asperulum (Von Martens) from Tai wan. von Baer K.E. 1828. Entwicklungsgeschichte — Annotationes Zoologicae Japonense, 50 der Thiere: Beobachtung und Reflexion, B. (2): 110–122. 1. Königsberg: Bornträger. 271 S. Shubin N., Tabin C., Carroll S. 2009. Deep ho- Wagner A. 2005. and evolvability mology and the origins of evolutionary nov- in living systems. Princeton (NJ): Princeton elty. — Nature, 457 (7231): 818–823. Univ. Press. 367 p. Shubin N., Wake D. 1996. Phylogeny, variation, Wagner A. 2011. The molecular origins of evo- and morphological integration. — Ameri- lutionary innovations. — Trends in Genet- can Zoologist, 36 (1): 51–60. ics, 27 (10): 397–410. Slack J.M., Holland P.W., Graham C.F. 1993. Wagner G.P. 1989a. The biological homology The zootype and the phylotypic stage. — concept. — Annual Review of Ecology and Nature, 361 (6412): 490–492. Systematics, 20: 51–69. At the root of animal diversity: Evolvability etc. 41

Wagner G.P. 1989b. The origin of morphologi- G.R., Cardew G. (eds). Homology. Chiches- cal characters and the biological basis of ho- ter: Wiley. P. 24–33. mology. — Evolution, 43 (6): 1157–1171. Wasik B.R., Moczek A.P. 2011. Decapentaple- Wagner G.P. 1994. Homology and the mecha- gic (dpp) regulates the growth of a morpho- nisms of development. — Hall B.K. (ed.). logical novelty, beetle horns. — Developm- Homology: the hierarchical basis of compa- ent Genes and Evolution, 221 (1): 17–27. rative biology. New York: Academic Press. Webster G., Goodwin B. 1996, Form and trans- P. 273 –299. formation. Generative and relational prin- Wagner G.P. 2014. Homology, genes, and evolu- ciples in biology. Cambridge: Cambridge tionary innovation. Princeton (NJ): Prince- Univ. Press. 287 p. ton Univ. Press. 478 p. West-Eberhard M.J. 2003. Developmental plas- Wagner G.P., Altenberg L. 1996. Complex ad- ticity and evolution. New York: Oxford Univ. aptations and evolution of evolvability. — Press. 794 p. Evolution, 50 (3): 967–976. West-Eberhard M.J. 2008. Toward a modern re vi- Wagner G.P., Misof B.Y. 1993. How can a char- val of Darwin’s theory of evolutionary no velty. acter be developmentally constrained despite — Philosophy of Science, 75 (5): 899–908. variation in developmental pathways? — Jour- Westheide W. 1987. Progenesis as a principle in nal of Evolutionary Biology, 4 (3): 449–455. meiofauna evolution. — Journal of Natural Wagner, G.P. 1989b. The origin of morphologi- History, 21 (4): 843–854. cal characters and the biological basis of ho- Wray G.A., Abouheif E. 1998. When is ho- mology. — Evolution, 43 (6): 1157–1171. mology not homology? — Current Opin- Wake D. 1999. Homoplasy, homology and the ion in Genetics and Development, 8 (6): problem of ‘sameness’ in biology. — Bock 675–680.