Evodevos identity from model organisms to developmental types

ronald a. jenner evo-devo's identity

Evo-devo studies the evolution of development, and how changes in development influence phenotypic evolutionary change. The evolution of novelties and body plans are considered as the most distinctive research areas of evo-devo (Wagner 2000, 2001, Wagner et al. 2000, Müller and Newman 2005). Nevertheless, there seems to be little consensus about evo-devo's disciplinary identity. It has been regarded as a branch of developmental biology, part of evolutionary biology, a revision of evolutionary theory or an independent new synthetic discipline (Gilbert et al. 1996, Arthur 2000, 2002, 2004a,b, Hall 2000, Raff 2000, Wagner 2000, Robert et al. 2001, Gould 2002, Wilkins 2002, Baguna and Garcia-Fernandez 2003, Gilbert 2003, Kutschera and Niklas 2004, Amundson 2005, Müller and Newman 2005). Similarly, there has been skepticism about evo-devo's promise in both the literature (Wagner 2000, 2001, Richardson 2003, Wagner and Larsson 2003, Coyne 2005) and at meetings such as the one in 2006 in venice, at which the present book was conceived.

Although various factors are at play, I think that current skepticism partly results from a failure to articulate evo-devo's conceptual foundation properly. This issue comes into focus when it is observed that the papers outlining evo-devo's research agenda almost exclusively link the promise of evo-devo to discovering general concepts and rules.

Evolving Pathways: Key Themes in Evolutionary Developmental Biology, ed. Alessandro Minelli and Giuseppe Fusco. Published by Cambridge University Press. # Cambridge University Press 2008.

Arthur (2002: 757), for example, expresses concern when he writes that we are currently in a situation 'where it almost seems that anything goes, that is, any developmental gene, its expression pattern and the resultant ontogenetic trajectory can evolve in any way. If this were true, no generalisations would be possible, let alone universally applicable laws'. A senior developmental biologist in my institution expresses it thus: 'I am left thinking that there are no rules, hence nothing for evo-devo to discover'.

Here I discuss the identity of evo-devo from the perspective of an important, but neglected, epistemological dualism: idiographics vs. nomothetics. By grounding evo-devo's identity in this framework I show that the above pessimism is misguided, and bring into focus a bias in what evo-devo is generally expected to contribute to biology. This perspective is also vital for understanding the role of model organisms in evo-devo ( Jenner 2006a), and for salvaging the status of developmental types, which have increasingly fallen into disrepute in the recent literature.

generality and uniqueness in science: nomothetics and idiographics

An epistemological distinction can be made between idiographic and nomothetic aspects of science. The description of unique and historically contingent particulars is the domain of idiographics, while the discovery of law-like regularities or generalities falls under the rubric of nomo-thetics. This epistemological dualism is predicated on the individuality thesis, which distinguishes between individuals and classes (Ghiselin 1997). Idiographics is strictly concerned with the description of unique, concrete individuals, while nomothetics is concerned with formulating generalisations for abstract classes of which individuals may be members. Such generalisations can be formulated with respect to traits shared between the members of a class. Evo-devo embodies both principles, and like other historical sciences such as anthropology, paleo-biology and evolutionary biology, evo-devo strives to relate the detailed description of particulars to law-like regularities (Gould 1980, Ghiselin 1997, 2005, Lyman and O'Brien 2004). Specifically, evo-devo aims to understand how the unique evolutionary histories of particular body plans, or origins of novelties, relate to the involvement of different classes of evolutionary developmental mechanisms that are embodied in the nomothetic conceptual categories at the core of evo-devo's research agenda (Table 6.1).

Table 6.1 A sample of evolutionary developmental mechanisms and the evo-devo concepts based on them, as well as developmental types with examples of model organisms.

Although listed separately in this table, developmental types can be formulated for any class of organisms sharing particular developmental traits, including specific phenotypes, or exemplifying particular evolutionary developmental mechanisms.

Evo-devo concept

Example

Model system

References

Gene level evolutionary developmental mechanisms

Epigenetic evolutionary developmental mechanisms Cell level evolutionary developmental mechanisms

Tissue/organ level evolutionary developmental mechanisms

Organism level evolutionary developmental mechanisms General evolutionary developmental mechanisms

Gene regulatory networks (GRN)

Gene/genome duplication and divergence Gene regulation by transposable elements Competence

Cell condensations Induction

Segmentation

Developmental bias

Redundancy

Sea urchin endomesoderm development

Drosophila engrailed family genes

Mouse coat colour, teleost Fundulus heteroclitus differential gene expression Nematode vulva development

Drosophila (imaginal discs) Cavefish Astyanax mexicanus eye loss

Arthropods, onychophorans, annelids, chordates

Caenorhabditis elegans body size variation

Drosophila antero-posterior axis development

Davidson and Erwin (2006)

Richards (2006), Biemont and Vieira (2006) Rudel and Sommer (2003), Hong and Sommer (2006)

Hall (2003b) Jeffery et al. (2003)

Minelli and Fusco (2004)

Arthur (2004b)

Rudel and Sommer (2003)

Modularity Co-option

Evolvability

Developmental constraint Evolutionary novelties

Relationship between micro and macroevolution

Genetic assimilation

Canalisation and cryptic genetic variation o w

General themes addressed by study of evolutionary developmental mechanisms

Cavefish Astyanax mexicanus sense organs Franz-Odendaal and Hall (2006)

Head and thoracic horns in Onthophagus Moczek (2006) beetles

In silico cell lineage evolution Azevedo et al. (2005)

Gastropod Cerion Sea anemone NematosteUa vectensis (bilateral symmetry, triploblasty) Three-spined stickleback skeletal evolution, Heliconius butterfly wing patterns

Bicyclus butterfly seasonal wing pattern differences Drosophila phenotypic variation increase during Hsp90 impairment

Gould (2002) Darling et al. (2005)

Rudel and Sommer (2003), Joron et al. (2006)

Pigliucci (2005)

Flatt (2005)

Continued

Evo-devo concept

Example

Model system

References

Developmental types

Developmental/phenotypic plasticity, polyphenism

Organisms with reduced characters

Animals with set-aside cells

Animals with type I development Animals with split or dispersed Hox clusters, and exhibiting Hox gene loss

Ant caste polyphenism, caste determination by primordial germ cells in parasitic wasp Copidosoma floridanum Wing loss in sterile ant castes, pelvic reduction in fish, eye loss in cavefish

Polychaetes, sea urchins

Nematodes, tunicates

Drosophila, the ascidian Ciona, the larvacean Oikopleura, C. elegans

Abouheif and Wray (2002), Extavour (2004)

Abouheif and Wray (2002), Jeffery et al. (2003), Tanaka et al.

(2005), Shapiro et al. (2006) Peterson et al. (1997), Ransick et al. (1996), Blackstone and Ellison (2000) Davidson (1991)

idiographics and nomothetics in evo-devo

Evo-devo's most outspoken practitioners have presented evo-devo as unabashedly nomothetic in its promise (Gilbert et al. 1996, Arthur 2000, 2002, 2004a,b, Hall 2000, Raff 2000, Baguna and Garcia-Fernandez 2003, Gilbert 2003), a view explicitly accepted by those who see evo-devo as an important contribution to, or corrective of, evolutionary theory (Gould 2002, Kutschera and Niklas 2004, Amundson 2005, Stoltzfus 2006). At the core of evolutionary biology, neo-Darwinian evolutionary theory supplies a set of nomothetic principles with respect to which evo-devo has staked out its conceptual territory. However, I think that a misleadingly biased perspective has been established in the literature by downplaying the importance of idiographics. For example, Arthur (2002: 759) labels evolutionary biology 'a conceptually driven discipline'. It is unlikely that Arthur simply means that evolutionary biology is based on hypothetico-deductive methodology characterised by the interplay of concepts and empirical evidence. Instead, it is clear that he refers specifically to nomothetics, to which evo-devo can make 'a conceptual contribution'. Yet, nomothetic insights are epistemologically accessible only through the study of idiographic details.

evo-devo ' s nomothetic aspects

There are two issues that require detailed examination. Firstly, what is evo-devo's potential contribution to neo-Darwinian evolutionary theory? Secondly, what is the explanatory range of evo-devo's central nomothetic concepts?

Evo-devo's relation to neo-Darwinian evolutionary theory

Developmental biology is reclaiming its appropriate place in evolutionary theory. Gilbert et al. 1996: 368

The clamour to revise neo-Darwinism is becoming so loud that hopefully most practising evolutionary biologists will begin to pay attention.

Pigliucci 2005: 566

We suspect that most evo-devoists are not concerned with enhancing, completing, modifying or overturning the modern synthesis of evolution.

Robert et al. 2001: 958

Although a 'new synthesis' has been repeatedly announced in recent years, those announcements are premature. Wilkins 2002: 34

As the above quotes show, evo-devo's status and ambitions with respect to neo-Darwinian evolutionary theory are controversial. Which quote is most accurate? Consideration of the hierarchical organisation of biology provides an important insight.

Evo-devo's main challenge is to codify the relationship of its organismlevel focus with the population-level focus of neo-Darwinian evolutionary theory (Wagner 2000, 2001, Gilbert 2003, Wagner and Larsson 2003, Amundson 2005). Evo-devo does not provide a new component to evolutionary theory, but instead draws attention to a previously neglected level. Evo-devo focuses on the origin and nature of the material substrate of evolution. Within the neo-Darwinian framework this rich topic was blackboxed under the rubric 'variation,' which was considered a mere boundary condition for the operation of the population-level processes deemed most important in determining the direction of evolutionary change, such as drift and selection. Specifically, any potential for shaping the direction of evolutionary change inherent in the nature of variation itself has been codified as 'constraint' in evolutionary theory (Maynard Smithet al. 1985, Arthur 2000,2004b, Reifet al. 2000, Gould2002, Stoltzfus 2006). This is the proper locale of evo-devo's empirical and theoretical contribution to evolutionary theory. As Gould (2002: 82) summarised it: 'the revolutionary empirics and conceptualisations of evo-devo [are] united by a common goal: to rebalance constraint and adaptation as causes and forces of evolution.' As far as evo-devo contributes general conceptual sub-themes that can be categorised under the rubric of variation, it should be considered a genuine contribution to a previously neglected part of evolutionary theory (Arthur 2000, 2004a,b, Stoltzfus 2006).

Critically, this means that the term 'evolution' itself is understood differently by evo-devoists and neo-Darwinians. The standard neo-Dar-winian understanding of evolution is a population-level process of the sorting of variation that is brought about by genetic recombination and mutation (Reif et al. 2000, Kutschera and Niklas 2004). Strikingly, as Stoltzfus (2006) points out, the neo-Darwinian perspective does not consider the processes of the origin of variation to be a part of evolution! As the Encyclopedia of Evolution states, 'Darwin's theory is peculiar in that evolution is not an extension of the mutational process' (Ridley 2002: 800). Under the evo-devo perspective of evolution, the processes generating variation are very much part of what evolution is.

The explanatory range of evo-devo's nomothetic components

The best answer to any question about evolution is the lawyer's answer to any general question about the law: 'It depends on the jurisdiction'.

Lewontin 2002: 17

The explanatory range of a concept can be defined as the class size over which it rules, i.e. the range of taxa, or events, or facts, over which generalisations can be made. For evo-devo, the relevant classes are defined with respect to the taxonomic range of organisms with particular developmental properties. Evo-devo's main criterion of theoretical importance is the extent to which evolutionary developmental mechanisms can influence the direction of phenotypic evolution. Evo-devo mechanisms are defined (Hall 2003a) as mechanisms operating during development that can be modified during evolution, thereby affecting phenotypic evolution (Table 6.1). Again, consideration of biology's hierarchical organisation is helpful.

Populations and organisms exemplify two distinct focal levels in life's constitutive hierarchy (Vrba and Gould 1986, Valentine and May 1996, Gould and Lloyd 1999, Gould 2002). By focusing on individual organisms evo-devo complements the neo-Darwinian focus on populations. A formal property of biology's hierarchy is a marked asymmetry of the interactions between levels that can be summarised as follows: the lower level proposes and the higher level disposes.

This sheds light on the relative importance of phenomena that occur on different hierarchical levels. Any phenomenon on the level of individual organisms or their parts has to be filtered through the population level if it is to have an effect on the direction of phenotypic evolutionary change. Any change on the higher level automatically has an impact on the lower levels, but the reverse is not true. Since individual organisms define the focal level of evolutionary developmental mechanisms, phenomena on this level will always be subject to the population level processes of natural and sexual selection, and drift. Natural and sexual selection are considered the most important forces governing phenotypic evolutionary change because they define the competitive economic context in which all evolution takes place (Ghiselin 1995, 1999, Vermeij 2004). Consequently, the maximum explanatory range of evo-devo mechanisms can logically never exceed the explanatory range of population-level processes. At best they can be equal partners. For this reason I disagree with authors who seem to try to overextend evo-devo's explanatory umbrella. For example, Gilbert (2003: 350): 'It may even be the case that the population genetics model turns out to be placed within a developmental framework', and Hall (2003a: 494): 'Evolutionary developmental mechanisms also include interactions between individuals of the same species, individuals of different species, and species and their biotic and/or abiotic environment'. These statements seem to imply that evo-devo's scope can encompass the traditional neo-Darwinian arena. I think this may obscure the legitimate complementary roles of organism- and population-level perspectives.

What then is the precise explanatory range of particular evo-devo concepts? According to Arthur (2004a,b) developmental bias represents evo-devo's most important challenge to a strict neo-Darwinian view of life. Developmental bias describes how the direction of evolutionary change is influenced by the non-random structure of variation. The potential explanatory range of developmental bias is enormous, because logic alone dictates the 'null model of zero bias' as 'inherently improbable anyhow' (Arthur 2004b: 284). However, the true extent to which developmental bias will direct evolutionary change is a matter of historical contingency, to be determined independently and idiogra-phically for the evolution of each character in each population. So it is too with the explanatory range of other evo-devo mechanisms on the hierarchical level of the individual and below (Table 6.1).

Therefore the study of the theoretical importance of evo-devo concepts falls into the same category as the older study of general evolutionary trends, rules or laws of evolution, such as Cope's rule (phyletic size increase), Bergmann's rule (temperature dependence of body size), Williston's rule (reduction in number and specialisation of repeated body parts) and ecological rules regulating the evolution of r-versus K-strategies (based among others on rapid development and high fecundity versus long development and low fecundity, respectively). These rules are not universal as their realisation depends on the relative strength of different selection pressures and taxon-specific characteristics that may aid or constrain a certain outcome; and Williston's rule depends also on developmental mechanisms for the multiplication and/or specialisation of parts. For example, Kingsolver and Pfennig (2004) showed that in many populations with variable body sizes there is positive individual-level selection (both natural and sexual) for increased body size, providing a potential explanation for instances of Cope's rule. In many situations larger body size is selectively advantageous, which may lead to broad, but not universal, predictions of when Cope's rule will obtain. Similarly, with respect to evo-devo mechanisms or mechanisms of genomic change in general (Ryan 2006), generalisations may be formulated that may ascribe different probabilities to particular kinds of events, or even allow (probably much more rarely and difficult to study) predictions of what will happen if certain circumstances pertain. For example, on the principle that gene duplication can have important consequences for evolvability (Carroll 2005), assessing the relative frequencies of different fates of duplicated genes (neofunctionalisation, subfunctionalisation or loss of function by becoming pseudogenes) for different taxa may lead to general insights or broad predictions about the evolvability of different taxa.

the roles of evo-devo's idiographics

Am I overly critical by claiming that the identity and promise of evo-devo has been presented in a biased way by overemphasising its nomothetic aspects? Surely, it is at least implicitly realised that any generalisations can only be built upon a rich idiographic foundation. However, as mentioned at the beginning of the chapter, some workers think that if anything goes, and if there are no general rules, then evo-devo has nothing to discover.

Extensive documentation of the unique contributions of evo-devo mechanisms (Table 6.1) to the origin of novelties and body plan evolution is a central idiographic goal of evo-devo. Yet, this goal seems almost pejoratively dismissed as 'merely filling in some missing details' (Arthur 2002: 757). Perhaps this is merely an unsurprising remnant of the pervasive tradition for the status ranking of scientific disciplines in which the arrow of arrogance unfailingly soars from the nomothetic domain to impale innocent idiographers (Jenner 2006b). Nevertheless, the documentation of evo-devo's unique phenomenology is integral to both evo-devo's idiographic and nomothetic goals. The central question then is how best to mine evo-devo's idiographics by the judicious choice of model organisms.

model bias equals model strength

. . . reasoning via model organisms, in a sense, has become the lingua franca of biologists. Ankeny 2001: S259

Choice of evo-devo model organisms has been discussed in detail elsewhere ( Jenner and Wills 2007). Here I restrict discussion to trait bias in model organisms. There are two extreme strategies for choosing new model organisms: (1) minimising character overlap between model organisms by maximising phylogenetic diversity; (2) maximising character overlap between model organisms by explicitly choosing them on the basis of shared developmental traits.

The first strategy maximises the amount of unique idiographic detail captured by models. It is generally recommended that a wider phylogenetic range of taxa, including satellite species that allow the easy transfer of experimental techniques, should be a prime guideline for choosing new models (Bolker 1995, Bolker and Raff 1997, Hughes and Kaufman 2000, Raff 2000, Simpson 2002, Wilkins 2002, Minelli 2003, Rudel and Sommer 2003, Sommer 2005). This perspective considers the trait bias (such as short generation time, rapid and stereotypical development) of established model organisms, especially those inherited from molecular developmental biology, as an important drawback because any general conclusions one might draw on the basis of these species 'are not universally true beyond our models' (Bolker and Raff 1997: 36).

Although broad phylogenetic sampling is important for assessing the extent of developmental variation, it is largely a distant idiographic goal. Given limited time and resources it is not the most efficient route to general nomothetic insights. A more pressing immediate goal is to establish the value of important evo-devo themes, such as developmental and phenotypic plasticity, canalisation, genetic assimilation and evol-vability. These topics as still labelled 'controversial', or 'too esoteric for mainstream consideration' (Gibson and Dworkin 2004, Sniegowski and Murphy 2006) and allegedly supported only by 'anecdotal evidence' (Leroi 1998, Sniegowski and Murphy 2006).

For fulfilling evo-devo's ultimate idiographic goal of documenting the diversity of evo-devo mechanisms, and their unique roles in the evolution of novelties and body plans, each idiographic particular - each species - is equally valuable. In contrast, the value of organisms for empirically grounding evo-devo's nomothetic themes is for each model based on possessing particular developmental characteristics that provide independent support for a particular concept. Importantly, there is a trade-off between explanatory range and explanatory force at a given sample size. Maximising the amount of unique idiographic detail captured by new models minimises the ability to draw general conclusions from them. In contrast, by sampling taxa that share particular characters, one can maximise explanatory force (a measure of explanatory or predictive reliability), which is the basis of general insights. Thus nomothetic profits for each idiographic investment are maximised by the coordinated choice of models with the potential to shed independent light on each theme. An efficient search for general nomothetic insights on the way to fulfilling evo-devo's ultimate goals depends on a biased search for models. The general value of bias in model organisms becomes clear when they exemplify developmental types.

developmental types: the basis of evo-devo' s nomothetics

Developmental types can be considered a special kind of body plan or Bauplan, differing only in the number or nature of body parts, or developmental aspects of form they refer to. Consequently, in the following discussion the terms 'body plan' and 'developmental type' can be interchanged without disturbing the logic of argument. Apart from this basic statement about the nature of developmental types, the literature is rife with confusion, and I know of no proper treatment of this important issue. Fitch and Sudhaus (2002: 243) pinpoint the problem when they note that the Bauplan suffers from 'uncertain ontology'. On the one hand, body plans are interpreted as concrete entities (Hall 1999, Amund-son 2005, Rieppel 2006), while on the other hand they may be conceptualised as abstractions (Scholtz 2004, Rieppel 2006). However conceived, body plans and developmental types are widely considered detrimental to evolutionary research because of supposed typological connotations (Arthur 1997: 30, Richardson et al. 1999, Fitch and Sudhaus 2002: 243, Baguna and Garcia-Fernandez 2003: 708, Scholtz 2004: 5, Amundson 2005: 256, Hübner 2006: 379, Rieppel 2006: 531). This creates a paradoxical situation as Amundson (2005: 235) writes that 'Bauplans are taken very seriously within evo-devo'. The evolution of body plans is the overarching theme of evo-devo. It is therefore crucial to understand the nature of body plans and developmental types.

Body plans and developmental types aren't what they seem

Baupläne and developmental types both refer to phenotypic traits shared among taxa. For the sake of this discussion, they may refer to any geno-typic, epigenetic or phenotypic traits, as well as the functional organisation that results from the interaction among organismal parts, on any hierarchical level, from the parts of individual organisms to mono-phyletic high-level taxa.

The nature of body plans can be clarified by being very clear about the fundamental ontological distinction between classes and individuals, which is the very foundation of natural science (Ghiselin 1997). Everything is either an individual or a class, and this distinction is very useful in addressing central evo-devo issues (Jenner 2006a). For the present discussion the following distinctions are important. Individuals are concrete and spatio-temporally restricted, while classes are abstract, spatio-temporally unrestricted concepts. The ontology of individuals is the part/whole relation, in which lower-level parts (cells) form a higher-level individual (multicellular animal). In contrast, the ontology of classes is the membership relation. Class membership is defined on the basis of possessing certain traits stipulated by the class's definition. In contrast, individuals do not have defining properties.

Baupläne can be conceived as both individuals and classes. The former seems widely favoured as body plans are said to evolve, and not to represent mere abstractions (Hall 1999, Amundson 2005, Rieppel 2006). For a body plan to be concrete it needs to refer to all parts that make up a higher-level whole. As soon as a body plan refers to only some abstracted parts of a whole, it is a class. This is intuitively obvious when considering the part/whole relationship of a multicellular individual. I am the sum of all my lower-level parts, so in total they make up a single concrete individual. However, if I refer to only some of my lower-level parts, for example my epithelial cells, this collection of parts together no longer constitutes a single concrete individual. Instead it specifies a class of traits of which parts of a higher-level individual are members. It represents an abstraction based on some specified characteristics.

In analogy, the Bauplan of a taxon is a concrete entity (individual) only when it refers to all parts making up the whole. The Bauplan consisting of all parts is then synonymous with the high-level whole. In contrast, when a Bauplan refers to only a selection of traits, such as major organ systems or developmental genes present in all of its organisms, or all traits that are shared by some but not all organisms of the clade, the Bauplan is an abstract class with an intensional definition stipulating the possession of certain traits. It seems that Baupläne are usually defined as classes, but construed as if they were individuals. This reifica-tion of a class as an individual is perniciously typological. The common view of body plans and developmental types strictly in terms of homologies and monophyletic taxa (Hall 1996, 1999, Arthur 1997: 29, Valentine 2004, Amundson 2005: 232, Hübner 2006: 370) seems to lend concreteness to an abstract concept, but this is unnecessarily restrictive. By properly defining a body plan as a class of characters, taxa can share a body plan even if it has independently evolved. Such a body plan can form the basis of generalisations that can go beyond particular monophyletic taxa. The recognition of Baupläne as classes is perfectly legitimate. Indeed, the recognition of classes is 'not really based on a historical analysis' (Scholtz 2004: 4); it is completely ahisto-rical. Importantly, exactly this ahistorical formulation of Baupläne and developmental types as classes provides the necessary basis for any nomothetic insights into evolution.

The heuristic value of developmental types

Developmental types attain importance in evolutionary research by specifying a class of organisms with shared properties of development. They may exemplify one or a combination of evo-devo mechanisms, or they may refer to the possession of a particular developmental genetic or morphological phenotype, such as animals with gene regulatory networks or larvae with set-aside cells (Table 6.1). This allows developmental types to function in generating nomothetic insights into the evolution of development. For example, animals with set-aside cells have functioned in several general evolutionary hypotheses. Set-aside cells are cells from which the adult body develops in animals with distinct larval stages (Peterson et al. 1997). They have been implicated in the evolution ofgerm-line sequestration as a mechanism to mediate conflict between cell lineages that may result in lowered organismal fitness (Blackstone and Ellison 2000, Michod and Roze 2001). Possession of set-aside cells also has general implications for the timing of germline formation (Ransick et al. 1996). Recently Peterson et al. (2005) also implicated set-aside cells as a general solution to the problem of alleviating the danger of predation in a sensitive phase of the life cycle. For these hypotheses it does not matter whether or not set-aside cells are homologous across groups.

Another developmental type comprises organisms with reduced morphologies, either uniformly across individuals of a species, or only characterising certain morphs, such as castes in social hymenopterans. It may be expected that a single evolutionary origin of organ loss may be reflected in an identical change at the level of developmental regulation, while convergent loss may be reflected in obvious differences in developmental regulation. However, the unique evolutionary loss of wings in non-reproductive castes of different ant species is reflected in a diversity of genetic regulatory changes in different species (Abouheif and Wray 2002), while in several populations of Mexican cavefish with independently degenerated eyes, the developmental mechanisms of eye reduction are surprisingly similar (Jeffery et al. 2003). This shows the value of studying independent instances exemplifying a common evo-devo theme.

The point here is not that new models should solely be chosen on the basis of a bias in developmental traits, because general patterns also need testing by potentially falsifying evidence. However, in cases where there is insufficient supporting evidence for the value of general concepts, it makes sense to focus first on documenting confirming cases, which is most efficiently achieved by the coordinated choice of model organisms to illuminate evo-devo's nomothetic themes.

conclusion: from model organisms to developmental types

Evo-devo is an ambitious young discipline with both idiographic and nomothetic goals. Idiographically, evo-devo aims to document the unique effects of changes in evolutionary developmental mechanisms on the origin of novelties and the evolution of body plans. Nomothetically, evo-devo attempts to establish the general effects of evolutionary developmental mechanisms on determining the overall direction of phenotypic evolution. Ultimately these aspects can be combined into an evolutionary narrative that relates the description of unique particulars to broad generalisations. On the long road towards fulfilment of evo-devo's ultimate aims, the coordinated choice of models to illuminate nomothetic evo-devo themes is a more efficient route to general insights than choosing new model organisms based solely on the criterion of maximising phylogenetic spread, which tends to maximise the amount of unique idiographic detail. Evo-devo's idiographics are most efficiently translated into nomothetic insights when model organisms are judiciously chosen to aid the discovery of developmental types, based on the models sharing certain developmental traits. The diversity of our models should therefore reflect the diversity of the general questions that interest us.

Love (2006: 95) observed that there is currently a mismatch between evo-devo's broad research agenda and 'the predominance of particular experimental tools', which are biased towards 'current consensus methodologies derived from genetic regulatory mechanisms'. This narrowing of evo-devo's research agenda is unsurprising insofar as our most important model systems were chosen to function within the genetic paradigm of developmental biology (Gilbert 2001). Consequently, it should be one ofevo-devo's central goals to promote hitherto neglected research topics into fully fledged research programs by judicious choice of new model organisms. Important topics, such as the evolutionary ecology of plasticity, may be 'logistically cumbersome and tedious' (Pigliucci 2005: 485), and are therefore in urgent need of more empirical work. This necessitates the selection of appropriate model organisms, even if they are not selected to function in evo-devo's current genetic paradigm. This requires evo-devoists to communicate their needs clearly to granting agents to prevent the exclusive funneling of funds into narrow research areas. For example, the British

Biotechnology and Biological Sciences Research Council (BBSRC) currently requests evo-devo proposals thus: 'Applications are encouraged to make comparisons between the genetic basis of development in different organisms'. For the moment, to conclude that evo-devo 'hasn't quite lived up to expectations' (Richardson 2003: 351) is simply to expect too much too soon.

0 0

Post a comment