Biol Theory (2011) 6:4–15 DOI 10.1007/s13752-011-0002-6

ORIGINAL PAPER

Animal Development, an Open-Ended Segment of Life Alessandro Minelli

Received: 21 May 2011 / Accepted: 7 July 2011 / Published online: 9 November 2011 Ó Konrad Lorenz Institute 2011

Abstract No comprehensive theory of development is available yet. Traditionally, we regard the development of animals as a sequence of changes through which an adult multicellular animal is produced, starting from a single cell which is usually a fertilized egg, through increasingly complex stages. However, many phenomena that would not qualify as developmental according to these criteria would nevertheless qualify as developmental in that they imply nontrivial (e.g., non degenerative) changes of form, and/or substantial changes in gene expression. A broad, comparative approach is badly needed. In the Cnidaria, for example, even the boundary between generations is problematic. Describing their life cycle in terms of metagenesis (alternation between polyp generation and medusa generation) or in terms of metamorphosis (polyp as larva or juvenile) are matters of semantics more than biology. The life cycle of other metazoans, described in textbooks in terms of larva-to-adult metamorphosis, is hardly different from a typical metagenetic life cycle of cnidarians. This applies to holometabolous insects and to marine invertebrates like sea urchins, where most of the larval cells are discarded at metamorphosis. The uncertain temporal and spatial boundaries of individual development are also shown by the widespread lack of a strict correspondence between adult and mature. A comprehensive theory of development should start with a zero principle of ‘‘developmental inertia,’’ corresponding to an indeterminate local self-perpetuation of cell-level dynamics. Indeterminate growth, scale-invariance, segmentation, and regeneration provide examples of developmental dynamics close to that.

A. Minelli (&) Department of Biology, University of Padua, Padua, Italy e-mail: [email protected]

123

Keywords Adult
Adultocentrism
Larva
Metagenesis
Metamorphosis
Theory of development
Zero model of development

‘‘Development’’ is a seemingly uncontroversial term, one of those words whose erudite origin has been largely forgotten, together with its original connection with a specific view of nature, in this case, preformism. Development has long been a word used in everyday language, and its technical use in science has never fueled debates comparable to those surrounding, for example, the term ‘‘species,’’ another word of erudite origin nowadays used both in popular and in scientific language. ‘‘Development,’’ however, is likely to become a controversial term whenever we are called to determine precisely the spatial and especially the temporal boundaries of the phenomena to which it applies. For example, is it legitimate to speak of development of a unicellular organism? Or, is isometric growth (i.e., growth without change in form) an aspect of development? If not, does any slight deviation from isometry turn growth into a developmental process? All these questions, and the likely uncertainty we may experience in trying to answer them, betray a fundamental, but probably unexpected cause. Our troubles with them derive from our standing lack of an explicit theory of development. Arguably, we won’t dream of finding a set of absolutely valid criteria for distinguishing developmental phenomena from those, e.g., of metabolism, or those of reproduction. All of these are the products of evolution and their very mechanisms, and their mutual relationships, are obviously subject to change. We can expect that any theory of development will eventually incorporate mechanisms responsible for nontrivial (e.g., non degenerative) changes of form, mainly

Animal Development

based on changes in gene expression substantially more important than those accompanying metabolism. As a preliminary step toward a future comprehensive theoretical account of development, I present in this article a sample of comparative evidence from a diversity of animal taxa that may help in conceiving of development as other than simply the way an egg is turned into an adult. Eventually, we may hope to bring together a set of defining criteria for developmental dynamics that will minimize controversy in categorizing and explaining the biology of organisms other than humans, or vertebrates generally.

5

between the two categories is not necessarily so neat as we generally assume. In addition, conditions bridging the unicellular with the multicellular organization are found in small eukaryotes such as the choanoflagellates, the closest relatives of the true animals, or metazoans (e.g., Fairclough et al. 2010). In this article, I will restrict examples and discussion to metazoans, but any future comprehensive theory of development will necessarily extend to the developmental processes occurring in the unicellulars, and also in plants and fungi. Development Does Not Necessarily Start from an Egg

What Development is Not It is usually taken for granted that development is a sequence of changes through which a multicellular adult is produced, through an increase in complexity more or less strictly programmed in its genes, starting from a single cell which in most instances is a fertilized egg. However, broad-scope comparison of a diversity of living beings will easily show that such a narrow view would leave out of consideration many phenomena that, for reasons mentioned below, deserve to be treated under the common term of development. At the same time, such a taxonomically unrestricted overview is likely to provide useful suggestions toward a comprehensive theory of development. In this article, I will not discuss the inadequacy of a strict genetic deterministic view of development or the manifold aspects of the genotype-to-phenotype map (on these topics, see for example Nijhout 1990; Alberch 1991; Keller 2000; Pigliucci 2001; West-Eberhard 2003; Kupiec 2009; Fusco and Minelli 2010; Pigliucci and Mu¨ller 2010). Instead, I will argue that development (1) is not restricted to the multicellular organisms, (2) does not necessarily start from an egg, (3) does not necessarily start from a single cell, (4) does not necessarily imply an increase in structural complexity, and (5) does not necessarily end with the achievement of sexual maturity. Development is Not Restricted to Multicellular Organisms Many eukaryotic unicellular organisms have complex life cycles. For example, many parasitic protozoans alternating between two hosts (e.g., human and mosquito, in the case of the malaria parasite Plasmodium) exhibit different morphologies and reproduce through different means in either host. These morphological and functional changes are no less predictable and conspicuous than those displayed by metazoans along their life cycle. In either case, these changes deserve to be classified as developmental rather than purely metabolic, although the boundary

I will argue below that the unicellular condition of which the egg is an example is a suitable, but not undisputable candidate for the role of starting point of an animal’s development. But we must preliminarily address the point, that in a sizeable minority of cases the individual development does not pass through (much less, begin with) an egg stage. This happens indeed in the case of vegetative reproduction, i.e., of reproduction by means of somatic cells rather than gametes; as, for example, the freshwater polyp Hydra does by producing buds. This may represent an occasional, or even a common occurrence among species that nevertheless reproduce also by sexual means, thus, with the intervention of gametes, but in a few cases the sexual reproduction is totally unknown and has been arguably lost. Vegetative reproduction has been recorded in most animal phyla, including arthropods and vertebrates (in the form of polyembryony, i.e., the production of two or more twins out of one fertilized egg), but is more common among sponges, cnidarians, flatworms, and annelids (reviewed in Minelli 2009a). The fact that the cellular progeny of the egg goes on diverging, morphologically and functionally, to eventually include the whole range of cell types we recognize in an adult animal, is likely to suggest that the egg represents an undifferentiated and perhaps primitive cell. Such a conclusion, however, would be ill-advised. With its enormous size, its yolk content, its specialized envelopes, the egg is one of the most specialized cell types evolved in the animal lineage (Boyden and Shelswell 1959). This is also suggested by the patterns of gene expression reported, for example, for the oocytes of the sea urchin Strongylocentrotus purpuratus, which are not suggestive of a ‘‘default’’ condition of generic or primitive metazoan cell, but specifically include genes involved in egg growth, meiotic recombination and division, storage of yolk, and fertilization (Song et al. 2006). If we are in search of a living proxy of a primitive metazoan cell, it is better to consider stem cells (see Laplane 2011), rather than eggs, and to look into the phenomena of wound repair, regeneration, and

123

6

vegetative budding rather than in the conventional beginnings of the life cycle in a sexually reproducing species. A further argument against the presumed unique role of the egg as the starting point of development is provided by animals that can reproduce both asexually and by sexual means. One of these is Hydra, whose polyps occasionally produced from fertilized eggs do not differ from those produced by budding. Moreover, this equivalence between the eventual products of embryogenesis (the developmental sequence starting from a fertilized egg) and blastogenesis (the developmental sequence starting from a bud of vegetative cells) is not limited to animals of simple anatomical structure, like the freshwater polyp, but extends to quite more complex organisms. A well-studied example is the colonial sea-squirt Botryllus (Manni and Burighel 2006; Brown et al. 2009). Sponges and other ‘‘simple’’ metazoans lack a dedicated gametogenic cell population (Extavour and Akam 2003; Extavour 2008) but even in those animals most cells are not able to produce gametes. The singularity of the egg (of the gametes, more generally) is better understood in negative terms: as rightly remarked by Extavour (2008), the evolutionary beginnings of the germ line are basically a consequence of the loss of gametogenic potential from the majority of the organism’s cells. Development Does Not Necessarily Start from a Single Cell If we want to circumscribe in an adequately inclusive way the class of phenomena about which a theory of development should be articulated, we cannot restrict our attention to the developmental sequences that include an egg stage. Instead, we must additionally take into account the developmental sequences associated with vegetative reproduction. This is not simply because in the latter case no ‘‘unique’’ cell such as the egg is involved, but—perhaps more interestingly—because in the phenomena of vegetative reproduction it is usually impossible to restrict the umbilical cord between parent and offspring to one cell only. For example, the buds by which sponges often reproduce are multicellular (Brien 1973). In most instances, it is probably impossible to say whether a bud has a unicellular or a multicellular origin, the latter anyway being likely the condition in the case of the sponges (Brien 1973). To decide between the two alternatives, we should be able to fix exactly when a cell, or a group of cells, is irreversibly committed to become a bud. It is true, indeed, that within the parent organism all the cells eventually forming the bud are traceable, by moving back along the cell lineage, to a single ancestral cell. I mean, here, the most recent common ancestor of all the cells in the bud. However, many if not most of the cells

123

A. Minelli

derived from that ancestor are likely to be eventually included in the parent’s body rather than in the bud. Interestingly, the latter circumstance has a precise equivalent in both the embryonic and the post-embryonic segments of the development of animals issued from an egg. In fact, of the cells deriving from the latter, a variable number is often discarded by apoptosis (for example, 131 out of 1,090 cells in the standard development of Caenorhabditis elegans). More conspicuously, an adult fly and an adult sea urchin derive from a small subset of the cellular offspring of the egg, because in either case a majority of the larval cells get lost at metamorphosis. These facts cast a shadow over the widespread opinion that the lack of a unicellular bottleneck between subsequent generations, usually represented by the egg, is important as a means to avoid the risk of building a new individual out of a cluster of cells including one or more mutant cells, and their potentially disruptive offspring (Grosberg and Strathmann 1998; Wolpert et al. 2007), or to avoid starting with a cell that does not faithfully mirror the parent’s genome, whose fitness has been already positively tested by natural selection (Gerhart and Kirschner 1997).

Development Does Not Necessarily Imply an Increase in Structural Complexity A decrease in morphological complexity is an expected evolutionary trend in lineages evolving from a free-living state to a parasitic life style. To a lesser extent, morphological reductions are also visible in animals that have adopted a subterranean life style, which is usually accompanied by a reduction of eyes (both in terrestrial and aquatic animals) and wings (in insects). If we consider instead development rather than evolution, a regressive trend is probably less expected. However, examples of dramatic decrease in morphological complexity are well documented along the life cycle of several animals. Again, this is often associated with the adoption of a parasitic life style, as in the case of many copepod and rhizocephalan crustaceans, where the reproductive phase shows an overall morphological complexity quite lower than the preceding free-living larvae. An extreme case is the rhizocephalan Loxothylacus panopaei. When its cyprid larva, provided with articulated appendages as is typical for crustaceans and arthropods generally, finds a suitable host (a crab), it changes into a worm-like mass that penetrates into the host and eventually splits into two dozen cells that move about independently in the host’s hemolymph (Glenner and Høeg 1995). These cells are not the parasite’s gametes, they are a postembryonic unicellular stage eventually growing into a poorly structured multicellular adult.

Animal Development

Adulthood Does Not Necessarily Coincide with Sexual Maturity A major impediment to the production of a sensible theory of development is arguably the adultocentric perspective from which this biological phenomenon is usually considered (Minelli 2003). Adultocentrism means to regard all embryonic, larval, and juvenile stages, and the developmental processes in which these are involved, simply as steps, or means, required to eventually obtain an adult. This perspective, most explicitly linked to the 17thcentury preformist views of development, survives in the modern concept of development as the deployment of a genetic program whose final target is the production of the adult. Good reasons for taking distance from adultocentrism are its unconfessed but undeniable finalistic flavor and its reliance on a strict genetic determinism whose inadequacy has been amply demonstrated. But even at a less philosophical and more factual level, the legitimacy of adultocentrism is negated by the contribution to reproduction provided, in many species including our own, by developmental stages other than the adult, that is, by embryos, or larvae, or juveniles. These phenomena show that we cannot uniformly equate ‘‘adult’’ with ‘‘sexually mature.’’ A look into arthropod development will provide the evidence we need for a critical discussion of this point. The post-embryonic development of arthropods is punctuated by molts. In the vast majority of insects, and also in many representatives of the other major arthropod lineages, the adult cannot molt any more. As a consequence, its external shape is essentially fixed, the only nontrivial change still possible being an enlargement of the female abdomen by a stretching of the flexible intersegmental membranes, thus accommodating a sometimes huge mass of eggs. In the holometabolous insects, the adult is preceded by a largely or completely immobile stage, the pupa, during which the larval structures are demolished and replaced by the adult ones, but even in those insects that do not undergo such a dramatic metamorphosis, the final instar is generally quite different from the previous ones, especially when the adult is provided with wings. But the final instar is not simply a structurally singular developmental stage; in the vast majority of cases, it is also the only reproductive stage in the insect’s life cycle. Thus, we may be hastily tempted to equate ‘‘adult’’ with ‘‘mature,’’ but this is generally imprecise and often wrong. In insects, and in arthropods generally, there is no universal and precise correspondence between the molt to adult and the first availability of mature gametes. Mature gametes, especially male ones, are often available days or weeks before the molt to adult, as in the case of the males of the cellar spider Pholcus phalangioides (Michalik and Uhl 2005) as well as in the males and females of mayflies, stoneflies, and many

7

lepidopterans. Sometimes, the heterochronic dissociation between the production of mature gametes and the development of somatic structures is pushed to such a degree that the reproductive instar is still a larva, morphologically and developmentally. This phenomenon of paedogenesis is known in the beetle Micromalthus debilis (Pollock and Normark 2002) and in several gall midges such as Heteropeza and Miastor (Cecidomyiidae) (Surhone et al. 2010). In some species of cecidomyiids, paedogenesis occurs in the second larval instar, in others in the third, and in still other species in the pupa (Nikolei 1961; Wyatt 1961, 1964). On the other hand, there are arthropods where the beginning of a reproductive phase is not accompanied by the end of molting. In these animals, there are thus multiple adult instars. This is the case, for example, of the horseshoe crabs, of many spiders (but not those belonging to some ‘‘derived’’ families), and also of many bristletails and silverfish among the insects and even of one group of winged insects. These are the mayflies, most of which first take wings as subimagoes, which eventually molt to equally winged, and generally very similar, imagoes. This terminology, contrasting subimago to imago, is obviously adultocentric, as it suggests that only one of the two winged instars is the real, legitimate adult, whereas the other is just a preparatory one. In fact, several mayfly species reproduce during the first rather than the second winged instar, thus demonstrating the arbitrary nature of the current terminology. Our categories are still further challenged by those male millipedes which, following an adult/mature instar, molt into a nonreproductive instar with non-functional genital appendages, which will eventually molt in turn into a second, functional reproductive instar. This is called a development by periodomorphosis (Verhoeff 1923; Sahli 1990). A similar phenomenon occurs in the crayfish Orconectes immunis, the crab Macropodia rostrata, and a few other crustaceans (Hobbs 1981; Pandian 1994). The question is, are the intercalary males, i.e., those in the nonreproductive instar between two reproductive ones, worthy of the name of adults? The answer to this question is arguably a matter of definitions. Overtly adultocentric is the term ‘‘set-aside cells’’ by which those cells are known, in the larvae of marine invertebrates such as sea urchins, sea stars, and nemerteans, out of whose progeny the whole adult body will be formed, following metamorphosis, while the remainder of the larval body will be ‘‘discarded.’’ I have suggested (Minelli 2009a) that these cells would rather deserve to be called ‘‘temporarily marginalized cells,’’ that is, cells that in the internal competition for a share of the common metabolic resources, within the developing larva, are temporally, but reversibly, outcompeted by other cell lineages. Another

123

8

reason to reject the interpretation of the set-aside cells in adultocentric terms is the fact that in some sea urchins and other echinoderms the adult sometimes derives from cells other than those that only by definition have been ‘‘setaside’’ as the material required to make the future adult (Eaves and Palmer 2003). The general features of development are thus arguably elsewhere: in the iteration of cellular dynamics increasingly tempered by constraints due to either generalized or local changes in gene expression, and mostly accompanied by morphogenetic effects.

Boundaries Between Generations In most metazoans, the beginning of a new generation can be easily, although not precisely, fixed around the time the zygote is formed and starts dividing. Easily, because the events characterizing fertilization, egg cleavage, and the subsequent steps of embryonic development will not be duplicated anymore, before the multicellular animal eventually developing out of the egg will produce in turn its own gametes and some of these will be eventually involved in fertilization, that is, in the production of the zygotes with which, or next to which, another generation will start. Not precisely, however, for all those reasons the modern debates in bioethics have discovered to exist, which put a heavy burden onto the shoulders of whoever wants to determine when the life of a new individual begins. Will this point be fertilization, i.e., the time a previously not existing genetic identity is formed by the coming together of two hitherto separated genomes? Or will it be the subsequent time at which the zygotic genome will actually start being expressed, the earliest embryonic phases proceeding only thanks to the contribution of maternally encoded proteins? Answering these questions is likely vital to informing important choices in our lives, but I think that this cannot be based on anything other than stipulation, or common agreement. This is indeed one of those difficult areas where increased knowledge of facts and additional ability to manipulate them contribute to blurring traditional, apparently clear-cut divides and add controversy to the theory and uncertainty to the practice. Others are the problems with the asexually reproducing organisms, where the boundary at which to fix the definitive separation between the budding parent and a finally independent offspring is often a matter of arbitrary decision. Similar, indeed, to another problem, where our customary concept of the biological individual equally demonstrates its inadequacy: I mean, in deciding whether a red coral is better treated as a colony of individual polyps, or as an individual within which the polyps are only tenuously connected and tenuously integrated parts.

123

A. Minelli

Recent analyses (e.g., Folse and Roughgarden 2010; Pradeu 2010) have shown that a sensible theory of individuality cannot be limited to the traditional criteria of physiological unity, genetic homogeneity, and genetic uniqueness, even if we acknowledge, with Santelices (1999), that these criteria do not necessarily bring us to recognizing identical lines of demarcation between biological individuals. Folse and Roughgarden (2010) thus suggest that individuality should be rather recognized in terms of fitness and adaptation—the usual currency in a neo-Darwinian view of life. Pradeu (2010) suggests instead that immunology can offer a theory which makes it possible to base individuality on physiological grounds. I prefer to support Dupre´’s (2010) view, that, ‘‘We should be pluralistic about how we divide the biological world into individuals’’ (p. 21). Within this view, the notion of biological individual turns out to be inadequate to serve as a foundation for a theory of development, either as the locus of developmental dynamics or, worse (see my previous comments on adultocentrism), as the target of these. Instead, from an opposite perspective, a theory of development will probably provide a ‘‘generative’’ foundation for a largely applicable concept of individuality. An effort in that direction is anyway beyond the scope of this article. Eventually, deciding about individuality and thus about the boundaries between generations is further complicated by the regular occurrence of both sexual and asexual reproduction within one and the same species. This may take different forms, the most popular being the life cycle of many cnidarians, which includes both a polyp stage and a medusa stage. The conventional description of the life cycle of these animals is as follows. Medusas behave as conventional sexually reproducing animals, i.e., they produce sperm cells and eggs. The fertilized egg develops into a multicellular ‘‘larva’’ called a planula, which eventually turns into a sedentary polyp. The latter often produces other polyps that usually do not get independent, but remain physically connected to form a branched colony. Eventually, however, a special kind of buds differentiates on the polyp, or the colony. Detaching from it, these buds develop into free-living, sexual medusas. Alternatively, the original polyp can turn into a pile of disks, each of which becomes autonomous and develops into a medusa. In either case, the whole cycle is usually described as an instance of metagenesis (a concept first introduced by Steenstrup 1845), that is, as a regular alternation between a generation of asexual polyps and a generation of sexual medusas. In other terms, between a fertilization event and the next one there would be two generations rather than one, as happens instead in the majority of metazoans. This life cycle is thus interpreted very differently from the life cycles of frogs or butterflies, where only one generation is recognized between subsequent sexual events, despite the most

Animal Development

obvious contrast, morphological and ecological alike, between the aquatic, fish-like tadpole and the air-breathing, four-legged frog, or the not less obvious contrast between the worm-like caterpillar and the winged, flying butterfly. We describe the transition from tadpole to frog as the metamorphosis of one and the same individual, and the same holds for the caterpillar turning into a butterfly. But the distinction between metagenesis and metamorphosis is far from being a neat one. To begin with, it does not apply at all to a whole group of cnidarians, the cubozoans. In these animals, the whole polyp simply turns into a medusa rather than budding off one or more medusas while retaining a residual individuality. One might argue that, at variance with the other cnidarians, the life cycle of the cubozoans is not metagenetic. In other words, that their polyp is a larva, the medusa is the (only) adult, and their mutual relationship is a metamorphosis. However, is the cubozoan life cycle really closer to that of a frog than to the life cycle of other cnidarians? Is their polyp really ‘‘other’’ than the polyp of hydrozoans or scyphozoans only because the latter produce medusas by budding, the cubozoan polyp by metamorphosis? One may still try to defend this distinction between metamorphosis and metagenesis by framing the question in terms of numbers. In cubozoans, when a polyp turns into a medusa, the number of individuals in the population remains the same, thus, there is no reproduction, whereas the number of individuals increases when a hydropolyp or a scyphopolyp gives off a multiplicity of little medusas. But this further distinction is less neat than we may wish. And a revision of the purported distinction between metagenesis and metamorphosis is likely to have a domino effect on our categorization of developmental processes. I will focus here on one aspect only, that is, the degree of material continuity we may regard as adequate to suggest that a change is, or is not, an example of metamorphosis and thus involves only one individual, or more than one, and one generation, or more than one. How much of a hydropolyp or a scyphopolyp goes into a medusa? Only a small amount, generally, and this may be used as a further argument in favor of interpreting the polyp and the medusa as different individuals, more specifically, as individuals belonging to two distinct generations. As convincing as this argument may appear on a first consideration, it would easily get us in trouble, when considering how little of what is usually called a larva is passed to what is usually called the corresponding adult. This happens all too frequently to be considered as a marginal exception we can ignore when formulating general principles about development and reproduction, individuality and generations. In many insects, for example in Drosophila and the other dipterans, the metamorphosis undergone by the larva is truly dramatic. Most of the larval

9

body is literally destroyed and all the typically adult organs including the wings, the compound eyes, and the genitalia are rapidly built from scratch, starting from the imaginal disks, small, dedicated groups of previously quiescent cells. Basically, the larval nervous system is largely saved for the adult, but most of the other larval structures fail to provide the adult with anything more than recycled molecules eventually used to set up the new organs. Not less dramatic is the metamorphosis of many marine invertebrates, where the adult, as mentioned before, is similarly formed from small groups of set-aside cells, whereas all the remaining larval structures are literally discarded (Peterson et al. 1997). This happens for example in the sea urchins and in a still more extreme form in a group of marine ‘‘worms’’ called the nemerteans. Does their life cycle still deserve to be described in terms of metamorphosis of one individual? Why not use here too the language we have long been using for the cnidarians, thus describing a nemertean ‘‘larva’’ and the corresponding ‘‘adult’’ as two distinct generations, the first of which gives rise to the second by means of asexual reproduction? A similar description would apply very fittingly to the starfish Luidia sarsi, where the larva may remain living and active for 3 months after it has dropped off the juvenile fated to grow into the conventional adult (Tattersall and Sheppard 1934; Williamson 2006). Actually, this is not a semantic shift I would generally recommend. Rather, I am presenting these examples to show how careful we must be in our search for a comprehensive model of development, as most if not all of our traditional categorizations turn out to fit only with a selected range of species (see also Laplane 2011). I close this section mentioning the unlikely case of a bivalve mollusk with something like an alternation of generations. This is Mutela bourguignati, whose larva lives temporarily on a freshwater fish as a parasite but eventually produces a bud that will give rise to the conventional adult (Fryer 1961).

What Development Possibly Is A Zero Condition of Developing Systems In mechanics, the principle of inertia states that in the absence of forces applied to it, a material point will maintain forever its condition of rest or of uniform rectilinear motion. But the phenomena in which we are actually interested are not those of inertial behavior, but those deriving from the most different deviations from inertia. In biology, an inertial system is a population in Hardy– Weinberg equilibrium (see Gayon 1992, 1998), a condition which is not so interesting per se, as is instead any

123

10

deviation from it. In the same vein, I have recently advocated a corresponding principle in developmental biology, i.e., a principle of ‘‘developmental inertia’’ (Minelli 2009a, 2011) according to which the ‘‘zero condition’’ of development is an indeterminate local self-perpetuation of celllevel dynamics (Soto and Sonnenschein 2004). The most general and possibly fundamental of these dynamics is arguably the sequence of mitotic and interphase events comprising the cell cycle. Recently, a zero model of purported universality has been suggested by McShea and Brandon (2010) as ‘‘biology’s first law.’’ According to this principle, the default behavior of any evolving system, irrespective of selection or other cause acting on it, is a steady trend toward increasing complexity and diversity. It may be useful to briefly compare this model with the concept I am suggesting here as a zero model for developmental dynamics. There are important differences, indeed, between the two principles. As mentioned, McShea and Brandon’s principle has been proposed as a foundation for a theory of evolution, rather than development, although its extension to this aspect of biology would probably offer little difficulty, and may deserve a chance. More important, however, is the fact that the principle I am advocating here has the nature of an empirical generalization, while McShea and Brandon (2010) present their ‘‘biology’s first law’’ as a universal law of nature. Both principles, however, aim to define a default behavior with which to compare real systems, which are actually subjected to a whole variety of constraining and modulatory actions which form the bulk of the subject matter of evolutionary biology and developmental biology. Another important point is that, similar to what McShea and Brandon predicate of evolution, I contend that without moving from a zero model we are inevitably prone to see in development the action of forces where none is acting, and vice versa. Animal examples of systems better approaching default developmental conditions are the sponges and the hydra, where the archaeocytes and the interstitial cells, respectively, are responsible for such an indeterminate ability to proliferate. Another example is provided by the planarians, where the neoblasts can comprise up to 30% of the total number of cells in the adult (Ellis and Fausto-Sterling 1997). Indeterminate Growth and Scale-Invariance If the zero condition for a developing system is an indeterminate replication of local cellular dynamics, the most obvious of those dynamics, i.e. mitosis, would produce, as a default condition, multicellular organisms with indeterminate growth.

123

A. Minelli

There are, indeed, examples of animals with indeterminate growth, such as the small freshwater annelid Pristina, the giant clam Tridacna and other long-lived bivalve mollusks, the sea urchin Strongylocentrotus, some decapod crustaceans, many bony fishes, several reptiles, and some big mammals like bison, giraffe, and elephant (Karkach 2006). Other animals, e.g., sponges, cnidarians, echinoderms, even appear to lack true senescence, although this ‘‘privilege’’ should not be regarded as a requisite for indeterminate growth (Sebens 1987). In the absence of constraints to its default behavior, a developmental system is also expected to show structural scale-invariance. Scale-invariance is actually widespread in biological phenomena and analyses in terms of fractals have been successfully applied to the suture lines of the ammonites (e.g., Garcı´a-Ruiz et al. 1990; Long 2005), or to the branched respiratory and vascular systems of vertebrates (e.g., Glenny and Robertson 1990; Zamir 2001). However, in mechanistic terms the question remains, whether these patterns are actually produced by iteration of one and the same morphogenetic process. This cannot be taken for granted and must be ascertained case by case. In Drosophila, for example, each trachea is a branched tube, with three subsequent levels of branching. The resulting pattern would suggest the iterated operation of one and the same machinery but, in fact, each of the three levels of branching is actually under a distinct genetic control (Samakovlis et al. 1996). Branched, grossly fractal is also the structure of the lungs of terrestrial vertebrates. In the case of the mouse, the genetic control of lung development suggests a condition intermediate between a simple iteration of branching and specific control for each branching step (Metzger et al. 2008). In the face of this evidence, caution is clearly required before accepting any specific instance of scale-invariance in a biological structure as a pure effect of simple, iterative developmental dynamics. Of course, the theoretical default behavior of developmental systems is no more than a bottom line eventually modulated by a multiplicity of constraints and controls. Already at the level of the most universal of the developmental dynamics, i.e., mitotic cell division, during an animal’s life there are times of uniform and intense activity, and times at which this comes to a temporary or definitive stop, either locally or generally. This is often associated, although not necessarily, with terminal cell differentiation. In several small-size animals, such as many nematodes and mites, mitotic activity is limited to the embryonic life. In others, the stop is limited to some tissues, e.g., the epidermis in most flatworms (Littlewood et al. 1999). Sometimes, cell proliferation is not simply stopped, but actively counteracted by programmed cell death, or apoptosis, as mentioned before.

Animal Development

Within a developing embryo, the multiplication of developmental foci and the iteration of developmental sequences is sometimes accompanied by an actual splitting of the embryo into a number of fragments, each of which continues developing as an independent individual. This way, more than one embryo is eventually obtained from a single egg (zygote). This process (polyembryony) is not very common in our species, but it is the rule in others, such as the armadillos of the genus Dasypus. For example, the well-known nine-banded armadillo (the one occurring also in the southern USA) gives birth to sets of four identical twin (Loughry et al. 1998). This number pales in comparison to the thousand and more embryos deriving from one egg in some tiny parasitic wasps, also polyembryonic (Zhurov et al. 2007). Developmental Modules The default, indeterminately iterative behavior of developmental systems is obviously only a starting point to which we must realistically add a lot of specifications, in order to approach a minimally sensible description of the real course of development in organisms of some complexity. Within a developing organism, we can usually recognize a number of more or less independent ‘‘modules’’ (Schlosser 2002; Schlosser and Wagner 2004; Callebaut and Rasskin-Gutman 2005) within which development goes on differently from what happens nearby. Some developmental modules are temporary features of a developmental system, as are the mitotic domains described by Foe (1989) in the early embryos of Drosophila. All the cells belonging to the same mitotic domain initiate mitosis synchronously after a phase of generalized mitotic arrest corresponding to the transition from a syncytial to a cellular condition of the blastoderm. At a later stage, the same embryo will be subdivided into polyclonal compartments each of which is exclusively formed by the progeny of a distinct set of founder cells (Garcia-Bellido et al. 1973). Similar clonal restrictions separate the cell populations of the different ‘‘segments’’ (rhombomeres) of the vertebrate hindbrain (Guthrie et al. 1993). However, the most popular units of restriction of developmental unitary behavior are arguably the germ layers, including the neural crest of vertebrates, that Hall (1998, 1999) has convincingly added to the list of these sheets of cells that selectively transcribe a few diagnostic genes by whose expression their identity is specified. For example, the mesoderm is the germ layer expressing snail and twist. Similarly, units of restriction of developmental unitary behavior are the imaginal disks of insects and the set-aside cells in the larvae of many marine invertebrates (Minelli 2003, 2009a).

11

Many hydrozoans provide excellent examples of diffuse multiplication of growth and differentiation centers. In these cnidarians, new buds can appear in the most diverse body parts and along all stages of the conventional life cycle (Bouillon et al. 2006). Production of buds is not restricted to polyps. Many medusas are also capable of vegetative reproduction and the buds borne on them (virtually at any possible location) can eventually turn into daughter medusas or even into polypoid structures. It is only in the textbooks that the polyps of Scyphozoa are obliged to give off medusas. In the case of Aurelia, the polyp can alternatively bud off new polyps, or produce stolons out of which new typical polyps (scyphistomes) will be generated, but it can also produce larvae that will later turn into scyphistomes (Franc 1993). Another animal group that provides conspicuous examples of multiple focal points of development is the Digenea, or flukes. For example, the mother sporocyst of Lechriorchis primus has a tree-like structure within which each branch has its own center of proliferation of germinal cells (Galaktionov and Dobrovolskij 2003). Focusing on the multiplicity of centers of developmental dynamics concurrently active within a developing animal should not obscure the obviously integrated nature of animal development. Nevertheless, there are examples of such a degree of functional and even physical separation between developmental modules that a continuity of overall developmental control over the whole system can be reasonably doubted. For example, at the 8-cell stage the embryo of the freshwater planarians of the genus Dendrocoelum resolves into separate cells that migrate individually within the mass of yolk cells and proliferate. Only at a later stage will their progeny gather together again, eventually continuing development into a single embryo. This phenomenon (Hallez 1887) has been long known under the sensible term of blastomere anarchy, or chaotic segmentation (Dawydoff 1928). Still more impressive is the split-embryo development of small freshwater fish (Cynolebias) where the zygote often gives rise to two separate ‘‘twin embryos’’ that unite again at gastrulation (Carter and Wourms 1993). Fusion of two separately developing organisms into one may even involve postembryonic stages, as in the freshwater sponges Ephydatia and Spongilla, where a single individual sometimes results from the fusion of two or more larvae (Brien 1973). Segments Segmentation is a widespread feature of body organization, sometimes very conspicuous on the animal’s external aspect, as in the case of earthworms and millipedes, in other instances less obvious but nevertheless distinct in anatomical features such as the vertebral column of

123

12

vertebrates. Many other serial features of body organization, including the eight plates of the dorsal shell of chitons, or the proglottids of tapeworms, are usually denied the name of segments, but from our perspective, mainly centered on developmental dynamics and independent of any claim of homology, this makes little difference. Segments, and serial features generally, are the site, or the product, of parallel, or equivalent, worksites (Minelli 2009b). Within each of these growth and differentiation foci, identical or at least similar developmental sequences go on, either simultaneously or in regular temporal progression. Different developmental dynamics may operate more or less simultaneously over broadly overlapping spatial domains, thus giving rise to a series of integrated morphological and eventually functional units, such as the serial body units of many conspicuously segmented animals. In other terms, the production of ganglia, body septa, excretory organs, muscle masses, etc., is likely to derive from the activity of different kinds, and series, of local developmental centers, but all their products are spatially aligned and eventually will result into ‘‘true segments.’’ Often, however, these different dynamics are not in pace, and the outcome of their operation is a body architecture with segmental mismatches (Minelli 2003; Minelli and Fusco 2004; Fusco 2005). A popular example is the trunk of millipedes, most of which is composed of body rings, each of which is associated with two pairs of legs. How to count segments, in this case? Do segments correspond to the body rings, or to the individual leg pairs? Recent evidence from developmental genetics shows that the question can only be answered by stipulation, because nothing in these animals corresponds to ‘‘the segment’’ as an integrated developmental unit. In the pill millipede Glomeris marginata, Janssen et al. (2004) have shown that the serial arrangement of sclerites on the dorsal and the ventral side of the animal are under the control of mainly different genes, and even those genes that are involved both dorsally and ventrally do not necessarily have an identical role in either body side. From the point of view of the standing or iterative activation of developmental centers, it is worth remarking that a developmental increase in segment number is not necessarily a diachronic event due to the continuing activity of a (generally posterior, subterminal; Jacobs et al. 2005) proliferation zone. In several groups, there is evidence of a hierarchical production of segments, with an early (eventually diachronic) establishment of a series of primary segments, each of which will later split, mostly in a taxon-specific stereotyped way, into two or more serial units of secondary level (Minelli 2000, 2001). Such a mechanism of double segmentation is suggested by the subdivision into 3, or 5, or more annuli of each body segment of leeches, but in an earlier developmental phase

123

A. Minelli

and with eventually less obvious results this mechanism also operates in the Arthropoda. This is suggested, in particular, by the idiosyncratic distribution of the number of leg-bearing segments in the Chilopoda, or centipedes, where only odd numbers (2N - 1) are found (Minelli and Bortoletto 1988). Most impressive is the recent discovery (Chagas Jr. et al. 2008) of a species of scolopendromorph centipedes whose specimens have either 39 or 43 pairs of legs, whereas in its closest relative only 21 or 23 pairs of legs are found, these latter being the only numbers previously recorded in the whole of the Scolopendromorpha (about 700 species). Segment number in Scolopendropsis duplicata has been interpreted (Minelli et al. 2009) as an example of ‘‘saltational evolution’’ caused by an extended operation of the segmental duplication dynamics otherwise responsible for the idiosyncratic distribution of segment numbers in these arthropods. Regeneration The last class of phenomena to be mentioned here is regeneration, a peculiar set of developmental processes that—by providing ‘‘local replacement copies’’ of lost products of the ordinary developmental processes— uniquely approximates a default behavior of developmental systems. Regeneration is known in many animals groups, including sponges, cnidarians, flatworms, mollusks, annelids, arthropods, echinoderms, and vertebrates (Minelli 2009a; see also Vervoort 2011). However, the extent to which the animal can replace its lost parts varies extensively, sometimes even within one major lineage. For example, within the Annelida, many earthworms and polychaetes have extensive power of regeneration, but leeches have none. Among vertebrates, salamanders regenerate legs and tail, lizards the tail only, while other vertebrates, humans included, do hardly more than seal wounds. In arthropods, regeneration is limited to the appendages. Regeneration is one of those biological phenomena in front of which the category of the individual may require quite a lot of stretching to fit with the facts we are describing. To what extent can we safely regard the regenerated organism as the same individual as the original one? A difficult case is for example the colonial sea squirt Botryllus, where regeneration causes a complete turnover of the original components (Lauzon et al. 2002), thus adding to the problems with circumscribing individuals already caused by colonial organization. Moving from a zero model of developmental dynamics, it is arguably easier (Minelli 2011) to understand the effectiveness of regeneration in front of the unpredictability of site, nature, and amount of physical damage an

Animal Development

animal may suffer (Birnbaum and Sa´nchez Alvarado 2008). The evolutionary trends toward increasingly strict constraints limiting the simply iterative behavior of developmental systems are accompanied by a parallel trend toward increasingly reduced regeneration capacities, partly because of selection against the likely maladaptive results of occasionally releasing developmental sequences ‘‘out of time.’’ Reasons why regeneration capacities have been lost in so many animal lineages are anyway a still largely open question (Bely 2010). To some extent, injury reactivates the most elementary dynamics of development, and can even allow the regenerating animal to live longer than it would have lived in the absence of damage and regeneration, as recently demonstrated in the little flatworm Macrostomum lignano (Egger et al. 2006). This suggests that we should turn upside down the popular but unnecessarily finalistic view, that many animals regenerate in order to live longer, to say instead that those animals live longer because they are able to regenerate (Minelli 2009a). Often but not always (e.g., Zattara and Bely 2011), regeneration in a given animal species has much in common with the ‘‘ordinary’’ developmental processes occurring in the same species. This should not be construed as meaning that regeneration is nothing but the reactivation of processes otherwise occurring during embryonic development, or an opportunistic deployment of developmental mechanics otherwise involved in vegetative reproduction, if any. However, there is abundant evidence of developmental mechanisms shared, even if at different degrees in the different animal groups, between embryogenesis, asexual reproduction, and regeneration. To argue that only the first of these processes can be categorized as development would imply a notion of development based on the nature of the products of the process rather than on its mechanisms. Exemplary is Cardona et al.’s (2005) comparison between embryogenesis and regeneration in planarians. Along the embryonic development of these animals it is not possible to recognize germ layers or organ primordia like in most other animals. Rather, a planarian embryo has quite a lot in common with a regenerating blastema. Nevertheless, the origin of the cells involved and the cellular environment within which the two processes occur are different between embryogenesis and regeneration. The context within which the embryo develops is only patterned by the spatial distribution of maternal determinants, whereas a regenerating blastema develops within the frame of already differentiated body parts. In the case of planarians, regeneration and embryogenesis similarly start with the production of a provisional epidermis, followed by the activation of neoblasts that proliferate and eventually produce the definitive epidermal cells. Also similar in

13

embryogenesis and regeneration are the next developmental phases including the differentiation of muscle tissue and the production of the ventral nerve cords.

Conclusions There are many reasons, empirical and theoretical alike, to regard development as a segment of life whose two ends can be fixed by stipulation only. Arguably, we may also have good practical reasons to conventionally fix the origin of a new individual at a specified biological transition, be it fertilization, or the beginning of translation of the zygotic genome, or some other. However, from a theoretical point of view, we must acknowledge that our choice is not irrevocably dictated by the internal logics of the phenomena to which we need to apply a periodization. Still more, opening our views to the whole range of developmental sequences of the most diverse animal species (not to speak of the other living beings) invites such a relaxation of our customary anthropocentric (or at least vertebratocentric) views, to suggest the need for a currently wanting comprehensive theory of development. As with other scientific disciplines, it is likely that developmental biology would benefit from anchoring its theory to a small set of ‘‘zero principles’’ corresponding to an ideal, not periodizable iteration of local cellular dynamics. To be sure, in the actual developmental schedule of all living organisms there is more, often much more than such an elementary behavior. But this does not mean that actual developmental schedules must necessarily start with the egg, or with fertilization, and necessarily end when the last fertile gamete has been eventually spawned by the adult. Acknowledgments My sincere thanks to Thomas Pradeu for inviting me to contribute to this Thematic Section. To Thomas, as well as to Lucie Laplane and Antonine Nicoglou, I am much indebted for insightful comments on an earlier version of this article. JeanJacques Kupiec has kindly shared with me his views on inertial systems.

References Alberch P (1991) From genes to phenotype: dynamical systems and evolvability. Genetica 84:5–11 Bely AE (2010) Evolutionary loss of animal regeneration: pattern and process. Integr Comp Biol 50:515–527 Birnbaum KD, Sa´nchez Alvarado A (2008) Slicing across kingdoms: regeneration in plants and animals. Cell 132:697–710 Bouillon J, Gravili C, Page`s F, Gili JM, Boero F (2006) An introduction to Hydrozoa. Me´moires du Muse´um national d’Histoire naturelle, vol 194. Muse´um National d’Histoire Naturelle, Paris Boyden A, Shelswell EM (1959) Prophylogeny: some considerations regarding primitive evolution in lower Metazoa. Acta Biotheor 13:115–130

123

14 Brien P (1973) Les de´mosponges. Morphologie et reproduction. In: Grasse´ PP (ed) Traite´ de Zoologie, vol 3(1). Masson, Paris, pp 133–461 Brown FD, Tiozzo S, Roux MM, Ishizuka K, Swalla BJ, De Tomaso AW (2009) Early lineage specification of long-lived germline precursors in the colonial ascidian Botryllus schlosseri. Development 136:3485–3494 Callebaut W, Rasskin-Gutman D (2005) Modularity: understanding the development and evolution of natural complex systems. MIT Press, Cambridge, MA Cardona A, Hartenstein V, Romero R (2005) The embryonic development of the triclad Schmidtea polychroa. Dev Genes Evol 215:109–131 Carter CA, Wourms JP (1993) Naturally occurring diblastodermic eggs in the annual fish Cynolebias: implications for developmental regulation and determination. J Morphol 215:301–312 Chagas A Jr, Edgecombe GD, Minelli A (2008) Variability in trunk segmentation in the centipede order Scolopendromorpha: a remarkable new species of Scolopendropsis Brandt (Chilopoda: Scolopendridae) from Brazil. Zootaxa 1888:36–46 Dawydoff C (1928) Traite´ d’embryologie compare´e des inverte´bre´s. Masson, Paris Dupre´ J (2010) The polygenomic organism. Sociol Rev 58(s1):19–31 Eaves AA, Palmer AR (2003) Widespread cloning in echinoderm larvae. Nature 425:146 Egger B, Ladurner P, Nimeth K, Gschwentner R, Rieger R (2006) The regeneration capacity of the flatworm Macrostomum lignano— on repeated regeneration, rejuvenation, and the minimal size needed for regeneration. Dev Genes Evol 216:565–577 Ellis CH, Fausto-Sterling A (1997) Platyhelminths, the flatworms. In: Gilbert SF, Raunio AM (eds) Embryology: constructing the organism. Sinauer Associates, Sunderland, MA, pp 115–130 Extavour CGM (2008) Urbisexuality: The evolution of bilaterian germ cell specification and reproductive systems. In: Minelli A, Fusco G (eds) Evolving pathways: key themes in evolutionary developmental biology. Cambridge University Press, Cambridge, pp 321–342 Extavour CG, Akam M (2003) Mechanisms of germ cell specification across the metazoans: epigenesis and preformation. Development 130:5869–5884 Fairclough SR, Dayel MJ, King N (2010) Multicellular development in a choanoflagellate. Curr Biol 20:R875–R876 Foe VE (1989) Mitotic domains reveal early commitment of cells in Drosophila embryos. Development 107:1–22 Folse HJ, Roughgarden J (2010) What is an individual organism? A multilevel selection perspective. Q Rev Biol 85:447–472 Franc A (1993) Classe des Scyphozoaires. In: Grasse´ PP (ed) Traite´ de zoologie, vol 3(2). Masson, Paris, pp 597–884 Fryer G (1961) The developmental history of Mutela bourguignati (Ancey) Bourguignat (Mollusca: Bivalvia). Philos Trans R Soc Lond B 244:259–298 Fusco G (2005) Trunk segment numbers and sequential segmentation in myriapods. Evol Dev 7:608–617 Fusco G, Minelli A (2010) From polyphenism to complex metazoan life cycles. Philos Trans R Soc B 365:545–690 Galaktionov KV, Dobrovolskij AA (2003) The biology and evolution of trematodes. An essay on the biology, morphology, life cycles, and evolution of digenetic trematodes. Kluwer Academic, Dordrecht Garcia-Bellido A, Ripoll P, Morata G (1973) Developmental compartmentalisation of the wing disc of Drosophila. Nat New Biol 245:251–253 Garcı´a-Ruiz JM, Checa A, Rivas A (1990) On the origin of ammonite sutures. Paleobiology 16:349–354 Gayon J (1992) Darwin et l’apre`s-Darwin: une histoire de l’hypothe`se de se´lection naturelle. Kime´, Paris

123

A. Minelli Gayon J (1998) Darwinism’s struggle for survival. Cambridge University Press, Cambridge Gerhart JC, Kirschner MW (1997) Cells, embryos and evolution. Blackwell Science, Boston Glenner H, Høeg JT (1995) A new motile, multicellulars stage involved in host invasion by parasitic barnacles (Rhizocephala). Nature 377:147–150 Glenny RW, Robertson HT (1990) Fractal properties of pulmonary blood flow: characterization of spatial heterogeneity. J Appl Physiol 69:532–545 Grosberg RK, Strathmann RR (1998) One cell, two cell, red cell, blue cell: the persistence of a unicellular stage in multicellular life histories. Trends Ecol Evol 13:112–116 Guthrie S, Prince V, Lumsden A (1993) Selective dispersal of avian rhombomere cells in orthotopic and heterotopic grafts. Development 118:527–538 Hall BK (1998) Germ layers and the germ-layer theory revisited: primary and secondary germ layers, neural crest as a fourth germ layer, homology, demise of the germ-layer theory. Evol Biol 30:121–186 Hall BK (1999) The neural crest in development and evolution. Springer, New York Hallez P (1887) Embryoge´nie des dendrocoeles d’eau douce. Baillie`re, Paris Hobbs HH Jr (1981) The crayfishes of Georgia. Smithson Contrib Zool 318:1–549 Jacobs DK, Hughes NC, Fitz-Gibbon ST, Winchell CJ (2005) Terminal addition, the Cambrian radiation and the Phanerozoic evolution of bilaterian form. Evol Dev 7:498–514 Janssen R, Prpic NM, Damen WGM (2004) Gene expression suggests decoupled dorsal and ventral segmentation in the millipede Glomeris marginata (Myriapoda: Diplopoda). Dev Biol 268:89–104 Karkach AS (2006) Trajectories and models of individual growth. Demogr Res 15:347–400 Keller EF (2000) The century of the gene. Harvard University Press, Cambridge, MA Kupiec J-J (2009) The origins of individuals. World Scientific, Singapore Laplane L (2011) Stem cells and the temporal boundaries of development: toward a species-dependent view. Biol Theory 6. doi:10.1007/s13752-011-0009-z Lauzon RJ, Ishizuka KJ, Weissman IL (2002) Cyclical generation and degeneration of organs in a colonial urochordate involves crosstalk between old and new: a model for development and regeneration. Dev Biol 249:333–348 Littlewood DTJ, Rohde K, Clough KA (1999) The interrelationships of all major groups of Platyhelminthes: phylogenetic evidence from morphology and molecules. Biol J Linn Soc 66:75–114 Long CA (2005) Intricate sutures as fractal curves. J Morphol 185:285–295 Loughry WJ, Prodohl PA, McDonough CM, Avise JC (1998) Polyembryony in armadillos. Am Sci 86:274–279 Manni L, Burighel P (2006) Common and divergent pathways in alternative developmental processes of ascidians. BioEssays 28:902–912 McShea DW, Brandon RN (2010) Biology’s first law: the tendency for diversity and complexity to increase in evolutionary systems. The University of Chicago Press, Chicago Metzger RJ, Klein OD, Martin GR, Krasnow MA (2008) The branching programme of mouse lung development. Nature 453:745–750 Michalik P, Uhl G (2005) The male genital system of the cellar spider Pholcus phalangioides (Fuesslin, 1775) (Pholcidae, Araneae): development of spermatozoa and seminal secretion. Front Zool 2:12

Animal Development Minelli A (2000) Holomeric vs. meromeric segmentation: a tale of centipedes, leeches, and rhombomeres. Evol Dev 2:35–48 Minelli A (2001) A three-phase model of arthropod segmentation. Dev Genes Evol 211:509–521 Minelli A (2003) The development of animal form. Cambridge University Press, Cambridge Minelli A (2009a) Perspectives in animal phylogeny and evolution. Oxford University Press, Oxford Minelli A (2009b) Forms of becoming. Princeton University Press, Princeton Minelli A (2011) A principle of developmental inertia. In: Hallgrı´msson B, Hall BK (eds) Epigenetics: linking genotype and phenotype in development and evolution. University of California Press, San Francisco, pp 116–133 Minelli A, Bortoletto S (1988) Myriapod metamerism and arthropod segmentation. Biol J Linn Soc 33:323–343 Minelli A, Fusco G (2004) Evo–devo perspectives on segmentation: model organisms, and beyond. Trends Ecol Evol 19:423–429 Minelli A, Chagas A Jr, Edgecombe GD (2009) Saltational evolution of trunk segment number in centipedes. Evol Dev 11:318–322 Nijhout HF (1990) Metaphors and the role of genes in development. BioEssays 12:441–446 Nikolei E (1961) Vergleichende Untersuchungen zur Fortpflanzung der heterogenen Gallmu¨cken unter experimentellen Bedingungen. Z Morphol Okol Tiere 50:281–329 Pandian TJ (1994) Arthropoda-Crustacea. In: Adiyodi KG, Adiyodi RG (eds) Reproductive biology of invertebrates, VI(B) Asexual propagation and reproductive strategies. Wiley, Chichester, pp 39–166 Peterson KJ, Cameron RA, Davidson EH (1997) Set-aside cells in maximal indirect development: evolutionary and developmental significance. BioEssays 19:623–631 Pigliucci M (2001) Phenotypic plasticity: beyond nature and nurture. The Johns Hopkins University Press, Baltimore Pigliucci M, Mu¨ller GB (eds) (2010) Evolution: the extended synthesis. MIT Press, Cambridge Pollock DA, Normark BB (2002) The life cycle of Micromalthus debilis LeConte (1878) (Coleoptera: Archostemata: Micromalthidae): historical review and evolutionary perspective. J Zool Syst Evol Res 40:105–112 Pradeu T (2010) What is an organism? An immunological answer. Hist Philos Life Sci 32:247–268 Sahli F (1990) On post-adult moults in Julida (Myriapoda, Diplopoda). Why do periodomorphosis and intercalaries occur in males? In: Minelli A (ed) Proceedings of the 7th international congress of myriapodology. Brill, Leiden, pp 135–156 Samakovlis C, Hacohen N, Manning G, Sutherland DC, Guillemin K, Krasnow MA (1996) Development of the Drosophila tracheal system occurs by a series of morphologically distinct but

15 genetically coupled branching events. Development 122: 1395–1407 Santelices B (1999) How many kinds of individual are there? Trends Ecol Evol 14:152–155 Schlosser G (2002) Modularity and the units of evolution. Theory Biosci 121:1–80 Schlosser G, Wagner GP (eds) (2004) Modularity in development and evolution. University of Chicago Press, Chicago Sebens KP (1987) The ecology of indeterminate growth in animals. Ann Rev Ecol Syst 18:371–407 Song JL, Wong JL, Wessel GM (2006) Oogenesis: single cell development and differentiation. Dev Biol 300:385–405 Soto AM, Sonnenschein C (2004) The somatic mutation theory of cancer: growing problems with the paradigm? BioEssays 26:1097–1107 Steenstrup JJS (1845) On the alternation of generation or the propagation and development of animals through alternate generations. Ray Society, London Surhone LM, Timpledon MT, Marseken SF (2010) Paedogenesis. VDM, Saarbru¨cken Tattersall WM, Sheppard EM (1934) Observations on the bipinnaria of the asteroid genus Luidia. In: Daniel RJ (ed) James Johnstone memorial volume. University Press of Liverpool, Liverpool, pp 35–61 Verhoeff KW (1923) Periodomorphose. Zool Anz 56(233–238): 241–254 Vervoort M (2011) Regeneration and development in animals. Biol Theory 6. doi:10.1007/s13752-011-0005-3 West-Eberhard MJ (2003) Developmental plasticity and evolution. Oxford University Press, New York Williamson DJ (2006) Hybridization in the evolution of animal form and life-cycle. Biol J Linn Soc 148:585–602 Wolpert L, Jessell T, Lawrence P, Meyerowitz E, Robertson E, Smith J (2007) Principles of development. Oxford University Press, Oxford Wyatt IJ (1961) Pupal paedogenesis in the Cecidomyidae (Diptera) I. Proc R Entomol Soc Lond A 36:133–143 Wyatt IJ (1964) Immature stages of Lestremiinae (Diptera: Cecidomyidae) infesting cultivated mushrooms. Trans R Entomol Soc Lond 116:15–27 Zamir M (2001) Fractal dimensions and multifractility in vascular branching. J Theor Biol 212:183–190 Zattara EE, Bely AE (2011) Evolution of a novel developmental trajectory: fission is distinct from regeneration in the annelid Pristina leidyi. Evol Dev 13:80–95 Zhurov V, Terzin T, Grbic´ M (2007) (In)discrete charm of the polyembryony: evolution of embryo cloning. Cell Mol Life Sci 64:2790–2798

123