Content uploaded by Gerd B. Müller
Author content
All content in this area was uploaded by Gerd B. Müller on Feb 14, 2018
Content may be subject to copyright.
The Innovation Triad: An EvoDevo Agenda
GERD B. MU
¨LLER
1,2
AND STUART A. NEWMAN
3
1
Department of Theoretical Biology, University of Vienna, A-1090 Vienna
2
Konrad Lorenz Institute for Evolution and Cognition Research,
A-3422 Altenberg, Austria
3
Department of Cell Biology and Anatomy, New York Medical College,
Valhalla, New York 10595
ABSTRACT This article introduces a special issue on evolutionary innovation and morpholo-
gical novelty, two interrelated themes that have received a remarkable increase of attention over the
past few years. We begin with a discussion of the question of whether innovation and novelty
represent distinct evolutionary problems that require a distinct conceptualization. We argue that the
mechanisms of innovation and their phenotypic results—novelty—can only be properly addressed if
they are distinguished from the standard evolutionary themes of variation and adaptation, and we
present arguments for making such a distinction. We propose that origination, the first formation of
biological structures, is another distinct problem of morphological evolution, and that together with
innovation and novelty it constitutes a conceptual complex we call the innovation triad. We define a
problem agenda of the triad, which separates the analysis of the initiating conditions from the
mechanistic realization of innovation, and we discuss the theoretical problems that arise from
treating innovation as distinct from variation. Further, we categorize the empirical approaches
that address themes of the innovation triad in recognizing four major strands of research: the
morphology and systematics program, the gene regulation program, the epigenetic program, and the
theoretical biology program. We provide examples of each program, giving priority to contributions
in the present issue. In conclusion, we observe that the innovation triad is one of the defining topics
of EvoDevo research and may represent its most pertinent contribution to evolutionary theory. We
point out that an inclusion of developmental systems properties into evolutionary theory represents
a shift of explanatory emphasis from the external factors of natural selection to the internal
dynamics of developmental systems, complementing adaptation with emergence, and contingency
with inherency. J. Exp. Zool. (Mol. Dev. Evol.) 304B:487– 503, 2005.r2005 Wiley-Liss, Inc.
Evolutionary theory treats morphological evolu-
tion primarily from the point of view of variation
and adaptation of characters. This approach
provides satisfactory explanations of the targeted
phenomena, but the adaptationist program has
also been criticized for its limited scope (e.g.,
Williams, ’66; Gould and Lewontin, ’79; Emlen
et al., ’98). Such criticisms concern the exclusion
of characters that are non-adaptive in their origin,
as well as the exclusive selectionism and the gene
centrism of the received theory, but rarely has an
expanded or alternative research program that
would overcome these shortcomings been sug-
gested. Over the past several years, much as a
consequence of a rising interest in understanding
the role of embryonic development in evolution
(EvoDevo), new programmatic goals have been
addressed, such as constraints, modularity, or
epigenetic factors in the origin of organismal body
plans. This expanded approach to evolution has
also brought to the fore a new set of related themes
termed ‘‘novelty’’, ‘‘innovation’’, and ‘‘origination’’
(e.g., Carroll et al., 2001; Wagner, 2001; Gottlieb,
2002; Hall and Olson, 2003; Mu¨ller and Newman,
2003a; West-Eberhard, 2003) that together repre-
sent a conceptual complex that we term the
‘‘innovation triad’’. This special issue explores
the plurality of views and positions in innova-
tion research characteristic of a nascent area of
scientific attention. The articles represent a selection
of up-to-date accounts of empirical and theoretical
approaches to innovation. In this introductory
Published online in Wiley InterScience (www.interscience.wiley.
com). DOI: 10.1002/jez.b.21081.
Received 12 August 2005; Accepted 19 September 2005
Correspondence to: Gerd B. Mu¨ller, Department Fu¨r Theoretische
Biologyie, Biozentrum, Althanstrasse 14, A-1090 Wien, Austria.
E-mail: gerhard.mueller@univie.ac.at
r2005 WILEY-LISS, INC.
JOURNAL OF EXPERIMENTAL ZOOLOGY (MOL DEV EVOL) 304B:487–503 (2005)
chapter, we will survey the conceptual foundations
of the innovation triad, the problem agenda of
a research program on innovation, the practical
research initiatives taken so far, and the theoretical
consequences of this approach for evolutionary
theory.
The rapid spread of the terminology of innova-
tion revived a long-standing discussion about
whether or not the referenced phenomena and
their underlying causal mechanisms are accom-
modated by the neo-Darwinian framework. Some
authors argue that the appearance of new traits or
characters that had no antecedent in the history of
a taxon, and hence exhibit neither a phenotypic
nor a genetic variation that is specific for the new
trait, is not—or not fully—addressed by the
received theory (Wagner et al., 2000; Love, 2003).
Others see no necessity to invoke anything but the
standard mechanisms of variation and selection to
account for novelties (e.g., Hall, 2005, this issue).
But there is a widespread consensus that the
generative mechanisms of morphological evolution
are underrepresented in the standard theory,
which is one reason EvoDevo has expanded so
explosively in the past decade. The EvoDevo
program may in fact be able to answer perennial
calls for an expansion of neo-Darwinian
theory if it can demonstrate that it is able to
solve biological problems that are not solved
by the traditional discipline or by one of its
subdomains, such as population genetics or devel-
opmental genetics (Mu¨ller, 2006). The innovation
triad could become one of the key themes in
this quest.
In exploring the potential of the innovation triad
to represent a distinct domain of EvoDevo re-
search, a number of questions arise. Among these
the most important are: What is actually meant by
‘‘origination’’, ‘‘innovation’’, or ‘‘novelty’’, i.e., are
there unambiguous definitions and examples of
these phenomena? At what levels can these
problems be addressed, and can this lead to
distinct empirical research projects? Do such
projects generate new insights, and are there good
examples of experimental results? And what are
the theoretical consequences: is anything ex-
plained that cannot be explained by the conven-
tional theory, that is, is there at all a need for a
distinct conceptualization of morphological no-
velty and the processes of origination and innova-
tion? There is no agreement as yet on most of
these questions, but clearly it only makes sense to
concentrate on these issues if it can be shown that
we deal with a sufficiently distinct problem, i.e.,
that the innovation triad is not part of the
adaptation complex.
ADAPTATION VS. NOVELTY—
VARIATION VS. INNOVATION
To make a distinction between adaptation and
novelty implies that novelties represent a class of
phenotypic change that differs from adaptive
variation and hence constitutes a separate evolu-
tionary problem in the same way that speciation is
an evolutionary problem that is distinct from
adaptation (see also Wagner and Lynch, this
issue). Two possibilities can justify a distinction
between adaptation and novelty: (1) the mechan-
isms involved in the generation of novelty may be
different from those underlying variation and
adaptation, and (2) the consequences of novelties
could influence the dynamics of phenotypic evolu-
tion in a way that differs from variational change.
In order to appreciate these distinctions, it is
necessary to recapitulate briefly the essence of
evolution through variation and adaptation.
Adaptations are generally understood as the
evolutionary improvement (with regard to a
particular performance) of an existing unit of
phenotypic organization, based on the processes
of heritable variation and natural selection
(Futuyma, ’86). It is widely agreed that adapta-
tions augment the reproductive success of their
bearers. Thus, favorable modifications will spread
in a population and contribute to its overall
ecological suitedness. The study and explanation
of this aspect of phenotypic evolution constitutes
what has been called the ‘‘adaptationist program’’.
It treats phenotypic characters from the point of
view of their adaptive significance. The key
elements of this approach are: (1) the choice of a
given character or trait for which an evolutionary
adaptation is to be considered; (2) the demonstra-
tion of heritable variation for that trait; and
(3) the demonstration that natural selection was
able to act on this variation. The latter is generally
regarded as the sufficient causal factor in the
generation of adaptations. Hence natural selection
represents the key element of all theoretical
explanations of adaptations. Although heritable
variation is sometimes difficult to demonstrate,
and the determination of the precise role of
natural selection can be problematic in a specific
case, there is ample field and experimental
evidence that these processes are active in
adaptive evolution (Futuyma, ’86; Rose and
Lauder, ’96). Further support comes from the
MU
¨LLER AND NEWMAN488
predictability of the expected changes of traits
under selection.
To distinguish novelty from adaptation would
mean that there is a difference in one or several of
the key elements noted above. One of these
differences already appears in step one: the choice
of the character or character complex that is to be
studied. In adaptation studies, the character is
usually chosen for its capacity to demonstrate
quantitative variation and adaptive modification.
This requires that the character is already in
existence in the chosen species; otherwise, it
cannot exhibit any degree of variation. A novelty,
by contrast would be a character that does not
belong to the constitutive range of variation of a
phenotypic precursor. Hence it is not usually
chosen in such kinds of studies. Further differ-
ences lie in steps two and three, namely that
heritable variation cannot be demonstrated at the
incipient stage, and therefore selection could not
act directly on the character since it is not yet in
existence. Note that this does not automatically
mean that selection has no role in the processes of
innovation, since indirect effects are not excluded.
In the case of novelty, a number of definitions
have been proposed. Early definitions have linked
novelty to function. Mayr (’60) proposed that
‘‘tentatively, one might restrict the designation
‘evolutionary novelty’ to any newly acquired
structure or property which permits the assump-
tion of a new function’’. Such a broad definition,
however useful at a taxonomic level, makes it
difficult to distinguish between quantitative var-
iational change (which may permit a new func-
tion), and qualitative structural novelty (which
may also permit a new function) although Mayr
favored the latter. Subsequent definitions have
focused on the fact that novelties usually repre-
sent deviations from quantitative variation and
constitute qualitative differences, as stated by
Mu¨ller (’90) ‘‘a qualitatively new structure with
a discontinuous (developmental) origin’’ or by
West-Eberhard (2003) ‘‘a novel trait (based on)
a qualitatively distinct developmental variant’’.
Both definitions address the mechanistic mode of
novelty generation by linking its origin to devel-
opment, and they emphasize the qualitative
change, although the former excludes a variational
origin whereas the latter does not. Yet other
definitions are associated with the term ‘‘key
innovation’’. Here the definition of novelty is
linked to its role in macroevolutionary processes
(see below), when a new trait permits the
exploitation of a new adaptive zone or facilitates
species diversification (Liem, ’74, ’90; Galis and
Drucker, ’96).
The definitions of novelty provided above,
although useful in their respective contexts, fail
to provide an operational criterion that can
distinguish novelty from variation at the pheno-
typic level. A suitable criterion can be based on a
comparative character definition of morphological
traits, such as provided by the homology concept.
The homology criterion is part of a definition
proposed by Mu¨ller and Wagner (’91), which says
‘‘a novelty is a new constructional element in a
bodyplan that neither has a homologous counter-
part in the ancestral species nor in the same
organism (serial homologue)’’. This definition
excludes characters that lie within the normal
range of variation, or deviate only quantitatively
from the ancestral morphological condition. But it
includes those cases in which a new homologue
has arisen through individualization of a preexist-
ing serial element. The definition leaves the
mechanistic mode of novelty origination open,
i.e., it does not restrict it to developmental
mechanisms alone. Although it can be argued that
this definition is narrow, it at least permits one to
identify unambiguous cases of novelty and is easy
to apply (but see West-Eberhard, 2005, this issue).
The use of homology criteria in the novelty
definition has sometimes led to the assumption
that novelty is synonymous with various forms of
apomorphy, as used in systematics, e.g., with
synapomorphy (Stone and Hall, 2004) or auta-
pomorphy (Hall, 2005, this issue). However, such
a conflation is difficult to justify. Attributions of
autapomorphy and synapomorphy are statements
about the distribution of derived characters in
taxonomical groups. But novelties represent the
source or initial character from which these (syn-
or aut-)apomorphies are derived. Feathers were a
novelty when they first arose, but are not a novelty
in extant birds, for which feathers are autapo-
morphic. The relevant EvoDevo problem of no-
velty is not the nature of the derived state (syn- or
autapomorphic distribution among taxa) but it is
one of mechanistic (developmental and evolution-
ary) origin. Although we may use the tables of
apomorphies in a search for interesting novelties,
the two categories are not equivalent, as empha-
sized earlier (Mu¨ller and Wagner, ’91; see also
Minelli and Fusco, 2005, this issue). Some apo-
morphies are rooted in a novelty, others are not.
This is because the absence of a character,
quantitative variations of size and shape, or a
specific combination of traits can also qualify as an
INNOVATION TRIAD: AN EVODEVO AGENDA 489
apomorphy in the taxonomical sense. A second
reason why apomorphies should not be regarded
as equivalent with novelty is precisely because in
doing so the differences in their causal origin are
blurred. Apomorphies are easily explained by
variation, novelties are not. Hence it does not
help clarify the issue to equate novelty with
apomorphy or one of its subcategories.
Just as a distinction can be made between
adaptation and novelty as two classes of pheno-
typic outcome, it might prove necessary to make a
distinction at the mechanistic level as well,
regarding the processes that bring about such
changes. As suggested above, the mechanisms
underlying standard variation cannot be consid-
ered the source of phenotypic novelty if we accept
that novelties represent characters for which no
ancestral homologue has existed. The innovation
mechanisms that introduce a new homologue into
an existing, lineage-specific character assembly
(bauplan) could be multifold. The possibilities
include mutational and recombinatorial events,
but other explanations favor a role for specific
responses of developmental systems either to
natural selection or to direct environmental
induction, as discussed below.
Although the terms innovation and novelty are
often used synonymously (e.g., Hall, 2005, this
issue, but cf. Love, 2003; Mu¨ ller and Wagner,
2003), the distinctions emphasized above suggest
that innovation should be used preferentially to
refer to the evolutionary modes and mechanisms
underlying novelty generation, whereas novelty
should designate a phenotypic outcome, such as
‘‘morphological novelty’’, ‘‘physiological novelty’’,
or ‘‘behavioral novelty’’. But since innovation is
frequently used in a non-specific sense so as to
include the origins of major taxonomical groups as
well, the use of qualifiers such as ‘‘morphological
innovation’’ ( 5novelty) or ‘‘functional innova-
tion’’ would also be sufficient to distinguish the
products from the processes of innovation. In the
usage adopted here, innovation pairs with varia-
tion at the level of evolutionary mechanism and
novelty pairs with adaptation at the level of
phenotypic result.
We have suggested elsewhere that ‘‘origination’’
is yet another distinction that should be made in
dealing with the generation of novelty (Mu¨ller and
Newman, 2003b). Origination refers to the specific
causality of the generative conditions that under-
lie both the first origins and the later innovations
of phenotypes. But whereas innovation and novelty
designate the processes and results of introducing
new characters into already existing pheno-
typic themes of a certain architecture (bodyplans),
origination emphasizes the very first beginnings of
phenotypes, e.g., the origin of mutlicellular assem-
blies, of complex tissues, and of the generic forms
that result from the self-organizational and physi-
cal principles of cell interaction (Newman, ’92,
’94). Because of presumably less integrated genetic
underpinnings, greater phenotypic flexibility is
suggested to have characterized the organismal
forms of early stages of metazon evolution
(Newman, ’94; Newman and Mu¨ller, 2000). In
this ‘‘pre-Mendelian’’ world, genetic inheritance
and phenotypic realization would have been less
strictly linked than in extant organisms (Newman
and Mu¨ller, 2000; Newman, 2005). Hence it was
argued that the predominant causalities for the
origin of primary morphological forms would have
depended less on programs of gene expression and
more directly on the inherent physico-chemical
properties of the involved cell and tissue assemb-
lies. Our use of the term origination therefore
refers to this principle (Mu¨ller and Newman,
2003b; Newman and Mu¨ller, 2005a, this issue)
from which will have resulted an early repertoire
of generic organismal structures, such as various
kinds of tubes, rods, hollow, multilayered and
segmented tissue masses that could later be
elaborated upon by natural selection.
So far we have discussed the mechanistic
arguments that support the adaptation vs. novelty
and the variation vs. innovation distinction. As
mentioned at the beginning of this section, a
second justification for this distinction comes from
the special role that novelties may have in the
dynamics of phenotypic evolution. This argument
is much more straightforward. Several well-stu-
died examples indicate that morphological innova-
tions can lead to the exploitation of new ecological
niches and to rapid and multiple speciation events.
This distinct effect of certain novelties is embodied
in the concept of key innovation (Liem, ’74, ’90).
The innovation of pharyngeal jaws in African
cichlid fish provides a good example of such effects
on the diversification and speciation of a taxon
(Galis and Drucker, ’96). The effect of key
innovations is probably what Mayr meant by
‘‘evolutionary avalanche’’ in his 1960 paper. But
although key innovations provide important ma-
terial for the study of evolutionary dynamics (and
for functional morphology), the particular macro-
evolutionary consequences of innovations do
not inform us about the causation of novelty.
The understanding, however, of the ecological
MU
¨LLER AND NEWMAN490
conditions that are associated with key innovations
and major macroevolutionary events do assist in the
explanation of novelty (Jablonski, 2005, this issue).
THE PROBLEM AGENDA OF THE
INNOVATION TRIAD
If we assume that origination, innovation, and
novelty represent a distinct complex of interre-
lated problems of phenotypic evolution, a number
of theoretical issues arise in attempting to account
for them. In particular, as with all biological
phenomena, it will be necessary to distinguish
what earlier had been called ultimate and prox-
imate causation (Mayr, ’61), and what may also be
referred to as general initiating conditions vs.
specific mechanistic realization. West-Eberhard
(2003) addresses the same point by distinguishing
‘‘initiation’’ vs. ‘‘sources’’ of novel traits. Con-
straint, integration, and fixation represent further
important issues that arise from regarding the
innovation triad as distinct, and, finally, there
is the problem of the transitional functionality
of novelties.
Initiating conditions (ultimate causes)
A number of initiating causes for the origination
of novelties have been proposed in the past, such
as accumulative genetic variation, environmental
influences, behavioral change, or change of func-
tion, depending on the point of view or preferred
explanatory paradigm. Under strictly neo-Darwi-
nian assumptions, for instance, accumulative gene
mutation had been considered as the primary
factor, as mentioned by Mayr (’60): ‘‘The problem
of the emergence of evolutionary novelties then
consists in having to explain how a sufficient
number of small gene mutations can be accumu-
lated until the new structure has become suffi-
ciently large to have selective value’’. Although
Mayr (’60) reaches the conclusion that ‘‘mutation
pressure plays a negligible role in the emergence
of evolutionary novelties’’, this had little impact
on the general opinion that regarded molecular
change as primary and selection as secondary. But
since a phenotypic novelty often requires develop-
mental modifications that are not within the
mutational reach of the ancestral character state,
alternative scenarios had to be considered. In
more recent treatments, the notion of indepen-
dent, incremental single locus mutations has been
replaced by gene duplications, evolution of gene
regulatory circuits by, e.g., promoter mutations,
and gene recruitment, but the primacy of genetic
change is maintained. The proposal that novelties
could arise from pleiotropic by-products of genetic
change, offered by Mayr (’60) is rarely voiced
today.
Natural selection as the primary initiating cause
has been invoked as frequently as genetic variation,
despite Darwin’s (1859) early warning that ‘‘char-
acters may have originated from quite secondary
causes, independently from natural selection’’.
The ‘‘selection paradox’’, the fact that selection
cannot act on characters that are not yet in
existence, and hence cannot directly cause novelty,
had been an early point of critique of Darwinism
(Mivart, 1871), but any precaution resulting from
it was largely abandoned in the neo-Darwinian
era. A particular version of the selection problem
arises with the concept of innovation through
symbiosis (Margulis and Fester, ’91). The central
tenet of this view is the integration of smaller
replicative units into higher level replicators, such
as in the origin of the eukaryote cell from
symbiosis with bacteria, or the association of cells
that produce a multicellular organism, or the
origin of colonies and eu-social communities in
which only a few individuals reproduce (Buss, ’87;
Maynard-Smith and Szathmary, ’95; Michod, ’99).
Such major transitions require cooperation among
replicative units (cells or individuals) to form the
higher level units by sacrificing the reproductive
success of the lower level units. In these kinds
of innovation, the problem is how the lower and
the higher levels of selection are reconciled in the
evolutionary process.
Concepts that assign the key locus of novelty
initiation to the epigenetic properties of develop-
mental systems involve selection in a more
indirect way. Such proposals argue that incipient
novelties arise as developmental side effects of
selection that acts on other organismal parameters
such as shape, size, or proportions, through the
alteration of developmental processes, e.g., the
modification of cell behaviors or of developmental
timing (Hanken, ’85; Mu¨ller, ’90; Mu¨ller and
Wagner, ’91; Newman and Mu¨ller, 2000; see also
Wagner and Lynch, 2005, this issue). In these and
earlier views (Schmalhausen, ’49), selection is
regarded as a facilitating factor or as a general
boundary condition rather than a direct cause of
novelty. In these scenarios selection is non-specific
with regard to the arising novelty; the specificity
of the phenotypic outcome is provided by
the developmental system under modification.
INNOVATION TRIAD: AN EVODEVO AGENDA 491
Threshold effects that occur in developmental
systems that undergo continuous variation are
seen as one of the key factors in such epigenetic
modes of novelty generation and could also
account for seemingly ‘‘saltatory’’ appearances of
novelty in phenotypic evolution although the
conditions for its generation may have preexisted
embryologically for a long time.
Another mode of novelty initiation based on
embryonic development has received renewed
attention, namely direct environmental influences
on developmental processes (Gilbert, 2001; Hall
et al., 2003; Mu¨ller, 2003a; West-Eberhard, 2003).
Here neither selection nor novel genetic variation
represent the initiating agent but rather the
immediate influence of external physical and
chemical factors. This view does not imply a
Lamarckian mechanism but a response of the
developmental system to an external (environ-
mental) perturbation that goes beyond the species-
specific range of variation (reaction norm), a
mechanism termed phenotypic accommodation
(West-Eberhard, 2005, this issue). The heritability
of the environmentally induced change may then
be consolidated by subsequent natural selection.
The conditions that permit the system to over-
come the limits of the reaction norm need further
clarification, but based on an extensive survey of
examples it was recently concluded that ‘‘the most
important initiator of evolutionary novelties is
environmental induction’’ (West-Eberhard, 2003).
West-Eberhard points out that a major advantage
of environmental induction over mutational
change is that it can affect many (or all) members
of a population at once.
An earlier statement of this evolutionary me-
chanism (Baldwin, ’96) was termed the ‘‘Baldwin
effect’’ by Simpson (’53) in a paper that argued for
its marginality to the neo-Darwinian framework.
Simpson’s critique, however, assumed a straight-
forward genotype–phenotype relationship and did
not consider either the extensive phenotypic plas-
ticity of modern forms (West-Eberhard, 2003) or
the possibility that ancient forms were even more
plastic (Newman and Comper, ’90; Newman and
Mu¨ller, 2000). The fact, moreover, that organisms
and their tissues at all stages of their evolution, in
common with any material system, have inherent
morphogenetic and other phenotypic properties
that govern their response to external effects
(Newman and Comper, ’90; Newman and Mu¨ller,
2005b), indicates that accounts based on environ-
mental considerations and those based on epige-
netic considerations are inextricably connected.
The correspondences between macroevolution-
ary innovation rates and large-scale ecological
patterns that seem to promote or sustain novelty
origination (Jablonski, 2005, this issue) support
the environmental connection, although this does
not mean that the environmental effect is direct in
all cases (see below). A special case of environ-
ment-induced novelty generation was described
for the vertebrate skeletal system, where new
skeletal elements can arise from altered embryo-
nic motility that is easily influenced by external
conditions (Mu¨ller, 2003a).
Behavioral change as an initiator of morpholo-
gical novelty had been considered a primary factor
in early discussions (Mayr, ’58, ’60). Behavioral
change was proposed to act via the intensification
or the change of a particular function, such as gait
or flight. Sewertzoff (’31) had given special
attention to the principle of intensification of
function, although it appears that this would
mostly result in the modification of existing
structures, such as shifts of proportions, fusion,
or loss of elements, but would rarely lead to the
emergence of a new structural character. Mayr
(’60) regarded not intensification but the change
of function ‘‘by far the most important principle’’
in the origin of novelty and provides a thorough
analysis of the conditions that are required for
shifts of function to take place. However, he also
reached the conclusion that the ‘‘new’’ structure
resulting from shifts of function ‘‘is merely a
modification of a preceding structure’’. This again
circumvents the novelty problem. However, note
the possibilities of a direct influence of the
environment on embryonic behavior mentioned
above (Mu¨ller, 2003a). Functional shift and func-
tional decoupling have recently been evaluated
(Galis, ’96; Ganfornina and Sanchez, ’99). The
‘‘behavioral change comes first’’ position has also
been reemphasized and elaborated on in psychol-
ogy. Behavioral flexibility based on developmental
plasticity is thought to result in behavioral
neophenotypes that in turn cause morphological
innovation followed by genetic integration (John-
ston and Gottlieb, ’90; Gottlieb, ’92).
Mechanistic realization
(proximate causes)
If the initiating causes for innovation are
unspecific and general, acting at the population
level, the conditions for the physical realization of
a specific novelty must be sought in development.
This is the domain of EvoDevo and the critical
MU
¨LLER AND NEWMAN492
question is: What are the specific developmental
mechanisms that underlie particular instances of
novelty generation? Currently, the bulk of atten-
tion is focused on the evolution of new gene
regulatory interactions and the recruitment of
genes and gene circuits into new developmental
functions, as seen in a growing number of
examples (Shapiro et al., 2004; Colosimo et al.,
2005). Understanding the kinetics (Bolouri and
Davidson, 2003), dynamics (Cinquin and Demon-
geot, 2005), and topological aspects (von Dassow
et al., 2000; Salazar-Ciudad et al., 2001a; 2003) of
developmental gene regulation and their correla-
tion with morphogenetic events will be central in
this endeavor.
Applying such dynamical concepts to gene
regulatory networks, including the possibilities of
non-linear effects, raises the specter of gene
changes acting as ‘‘macromutations,’’ i.e., the
‘‘hopeful monster’’ scenario (Goldschmidt, ’40),
long considered to have been retired from active
consideration in evolutionary biology. Contrary to
neo-Darwinian expectations, however, genes of
large effect are, in fact, present in Drosophila
populations (Tautz, ’96) as well as in plants, where
they are apparently involved in speciation (Got-
tlieb, ’84; Bradshaw and Schemske, 2003). While
such phenomena are rare (for the classical neo-
Darwinian reasons) in populations of modern-day
organisms, in the less developmentally canalized
forms of earlier evolutionary periods they are
likely to have been more prevalent (Newman and
Mu¨ller, 2000; Newman, 2005).
The preceeding consideration is not intended to
resurrect a macromutational scenario for the
origins of innovation and novelty, but rather
to highlight that any gene or gene regulatory
changes involved in innovation only become
morphogenetically relevant in the context of
molecular, cellular, and tissue-level interactions
that physically generate the new character. Since
each level of biological organization includes its
own emergent properties, it is a specific task of the
EvoDevo research program to identify the relative
importance of these mechanisms for the explana-
tion of particular instances of innovation (Wagner
et al., 2000). Candidate processes discussed in this
volume include the establishment of new inductive
interactions and cell specification mechanisms
(Cebra-Thomas et al., 2005, this issue; Fe
´lix and
Barrie
`re, 2005, this issue; Hall, 2005, this issue;
Minelli and Fusco, 2005, this issue), the redeploy-
ment of plesiomorphic molecular signaling path-
ways in hierarchically organized morphogenetic
modules (Prum, 2005, this issue), the establish-
ment of genetic individualization in existing organ
primordia (Kramer and Jaramillo, 2005, this issue;
Wagner and Lynch, 2005, this issue), and the
genetic modulation of self-organizing cell and
tissue properties followed by developmental
autonomization (Newman and Mu¨ller, 2005a, this
issue). These approaches will be discussed further
in the section on research programs below.
Constraints
Phenotypic evolution is limited and biased by
developmental constraints that have accumulated
in the history of all taxa (Maynard Smith et al.,
’85; Schwenk and Wagner, 2003). Such constraints
are another central EvoDevo principle. Innova-
tions seem to arise in spite of such constraints and
may sometimes require a relaxation or breaking
up of constraints that prevailed in the ancestral
state. Whether and how this occurs is an un-
resolved empirical question that needs to be
addressed in future studies of innovation. Shifts
of constraints have been described (Rienesl and
Wagner, ’92); other possibilities for constraint
relaxation include mutational events or neutral
phenotypic drift (Mayr, ’60). Even in closely
studied cases, the interpretation of whether or
not developmental constraints had to be overcome
for the origination of an innovation can differ
substantially (Eberhard, 2001; Wagner and
Mu¨ller, 2002). But constraints should not be
understood only in a negative way, as mere
limitations to phenotypic variation, since they
can also provide taxon-specific opportunities for
novelties to arise. A novelty can result from a
constrained developmental system, not because
genetic or developmental variation is relaxed, but
precisely because the system is unable to respond
by variation and is forced to transgress a develop-
mental threshold. This can provide heightened
potentialities for innovation in particular areas of
phenotypic character space (Roth and Wake, ’89;
Arthur, 2001; Rasskin-Gutman, 2003).
Integration and fixation
If innovation is considered as an event that is
not based on the continuous variation of a
preexisting, integrated character but arises
through any of the mechanisms discussed above,
then the problem arises as to how the new
character can be accommodated into the preexist-
ing, constructional, developmental, and genetic
systems of a taxon, in order to ensure functionality
INNOVATION TRIAD: AN EVODEVO AGENDA 493
and inheritance. The problem is not an immediate
one, because if a novelty arises from a develop-
mental response in one form or another, the same
response will be elicited in every new generation
and the novelty can remain epigenetically inte-
grated for an extended period of time before
genetic integration takes place. It is possibly a
standard rule that epigenetic integration precedes
the genetic integration of novelties (Newman
and Mu¨ller, 2000). Similar concepts that empha-
size epigenetic integration are generative
entrenchment (Wimsatt, ’86) and epigenetic traps
(Wagner, ’89).
As anticipated by Waddington (’56, ’62), genetic
integration follows from selection acting on the
genetic variation that will arise with the spreading
of a novel character. This will include orthologous
and paralogous regulatory circuits that acquire
new developmental roles over the course of
evolution (Wray, ’99; Wray and Lowe, 2000;
Carroll et al., 2001). Evolving structure–function
interrelationships (Galis, ’96) integrate novel
characters at the phenotypic level but will also
contribute to genetic integration, with selection
favoring the genetic linkage of functionally
coupled characters (Wagner, ’84; Bu¨rger, ’86).
The evolving genome can thus gain control over
the epigenetic conditions that prevail during the
origination of novelties. Since epigenetic integra-
tion will usually come first, its patterns can
provide the templates for both phenotypic and
genetic integration (Newman and Mu¨ller, 2000;
Mu¨ller, 2003b). Genetic integration will increas-
ingly stabilize and overdetermine the generative
processes, resulting in an ever-closer mapping
between genotype and phenotype. Such transi-
tions can be interpreted as a change from
emergent to hierarchical gene networks (Salazar-
Ciudad et al., 2001a,b). The sum of these processes
locks in the novel characters that arose as a
consequence of the mechanisms discussed above
and thus will generate the stable, heritable
building units of the phenotype that compose
organismal body plans.
Transitional functionality
As mentioned above, morphological novelties are
often associated with radically new functions that
were not present in the primitive condition. It has
therefore been assumed that a new function can
be responsible for the origin of a novelty. The
problem with this argument is that the derived
function usually requires that the new organ is
already fully developed, because it could not have
executed the new function in its incipient state.
The origin of insect wings is a good example. Small
or rudimentary wings do not allow any form of
flight and thus cannot be optimized for flight by
natural selection. The solution to this problem was
proposed to lie in a transfer of function, i.e., the
new structure would initially have evolved serving
a different function and would have become co-
opted into performing the derived function only
subsequently (Mayr, ’60). In the case of insect
wings, for instance, it is assumed that they
originally evolved as external gills of aquatic
arthropods (Averof and Cohen, ’97), which may
then have served as ‘‘sails’’ for surface skimming
(Marden and Kramer, ’94), and became used
and selected for flight only in a third transition
of function.
The innovation triad, in its emphasis on epige-
netic determinants, environmental influences and
plasticity, provides new ways of conceptualizing
abrupt morphological transitions. In particular,
thresholds and other nonlinear effects in develop-
mental systems, coupled with the potential of
externally induced changes to affect populations
and not only individuals, can render moot ques-
tions of phenotypic gradualism and thus transi-
tional functionality (see below).
RESEARCH PROGRAMS ADDRESSING
THE INNOVATION TRIAD
The innovation topic has been implicitly ad-
dressed in a number of research initiatives, but
often without directly invoking themes of the
innovation triad. Recently, more empirical studies
explicitly address innovation issues, and a number
of different strands of work can be discerned that we
will here call ‘‘programs’’ for convenient distinction,
although not many of the involved research groups
would see their work as belonging to one program or
another, since they all intersect. In accordance with
an earlier categorization of EvoDevo projects (Mu¨l-
ler, 2005), we distinguish four principal programs of
innovation research (Table 1).
The morphology and systematics program
The factual basis of the occurrence of novelty in
phenotypic evolution is provided by morphological
and systematic investigations, many of which
are paleontological. These data demonstrate
the kinds, the frequency, and the pervasiveness
of novelty generation in different phylogenetic
lineages as well as the relation of these occur-
MU
¨LLER AND NEWMAN494
rences with anatomical, developmental, and
environmental conditions. Another important
function of the comparative program is the
identification and characterization of novelties.
Because of the intimate interlace of conserved and
novel features in complex organismal bodyplans, it
is necessary to define operational criteria by which
complex characters can be disentangled and
novelties can be recognized based on character
theory (Mu¨ller and Wagner, ’91; Wagner, 2001;
Minelli and Fusco, 2005, this issue).
Both temporal and ecological patterns of in-
novation emerge from paleontological surveys
(Jablonski and Bottjer, ’90; Nitecki, ’90; Erwin,
’93; Eble, ’99; Jablonski, 2005, this issue). In these
contexts innovation is predominantly defined in a
macroevolutionary sense, emphasizing speciation
and the origin of higher taxa, but in many cases
these events are concurrent with the appearance
of morphological novelty. Such studies indicate
that novelties do not originate at random but are
often related to instances of environmental change
and to heterogeneous ecological domains. Patterns
of origination of major marine invertebrate taxa,
for instance, are clearly different in the onshore
vs. offshore venues (Jablonski, 2005, this issue).
This points towards different large-scale driving
mechanisms, such as sea-level changes, salinity or
oxygen changes, and other oceanographic influ-
ences. The onshore bias might be attributed to a
higher frequency of such habitat disturbances that
can in turn affect nutrient supply, population
genetic dynamics, phenotypic plasticity, and other
life history parameters. Other paleontological
surveys show differences in the ratio of major to
minor morphological innovations (Eble, ’99).
Jablonski (2005, this issue) suggests that an
observed origination bias allows two interpreta-
tions: either direct environmental influence on the
genetic–developmental systems of the affected
taxa, or a preferential preservation of innovations
in the onshore environment.
Functional morphology has played an important
role in the first conceptualizations of EvoDevo, in
particular with regard to the origin of innovations.
The vertebrate feeding apparatus is one of the best
analyzed functional systems including studies in
fish (Liem, ’74, ’90; Galis and Drucker, ’96),
amphibians (Roth and Wake, ’85, ’89), reptiles
(Frazzetta, ’75; Jayne et al., 2002), and many
other taxa. Recent non-vertebrate examples in-
clude the pulsatile organs in insects (Pass, 2000)
and the moveable abdominal lobes in male sepsid
flies (Eberhard, 2001). Although the functional
morphology approach to innovation is less promi-
nent today, many of the functional explanations of
novelty include the mobilization of developmental
processes (Roth and Wake, ’85; Galis, ’96).
The gene regulation program
The bulk of present innovation studies concen-
trates on the evolution of developmental control
genes and gene regulation networks (Carroll et al.,
2001; Davidson, 2001; Wilkins, 2002). This aims at
demonstrating the specific genetic changes that
are associated with novelty generation, with a
major focus on the Hox genes. Well-documented
cases include the innovation of eye spot patterns
in butterfly wings (Keys et al., ’99; Nijhout, 2001).
Hox gene-associated changes in axial differentiation
TABLE 1. Principal programs and issues
of innovation research
Research programs Issues addressed
Morphology and Analysis of Morphological Novelty:
Systematics Program Identification and characterization
Phylogenetic patterns of
occurrence
Frequencies of occurrence
Ecological distribution
Functional analysis
Gene Regulation Analysis of Genetic Innovation:
Program Comparative gene expression
Gene regulatory evolution
Co-option and recruitment
Regulatory modules
Genetic individualization
Epigenetic Program Analysis of Developmental
Causation:
Physical and other generic
properties of cells and tissues
Tissue self-organization
Dynamics of tissue interactions
Tissue geometry and architecture
Influences of external factors
Phenotypic plasticity
Experimental testing
Theoretical Biology
Program
Conceptualization and
Formalization:
Genotype–phenotype mapping
Measurement and quantification
Computational modeling
Simulation and prediction
Theory development
INNOVATION TRIAD: AN EVODEVO AGENDA 495
(Burke et al., ’95; Gaunt, 2000; Nowicki and
Burke, 2000), the vertebrate limb (Sordino et al.,
’95; Shubin et al., ’97; Crawford, 2003; Wagner and
Chiu, 2003), and innovation in the cephalopod
neural system and brachial crown (Lee et al.,
2003). In plants, good evidence for the genetic basis
of innovation is obtained from the evolution floral
organs (Kramer and Jaramillo, 2005, this issue).
Besides the ubiquitous documentation of chan-
ging gene expression patterns associated with
novelty and of evolving gene regulation, a number
of themes emerge as key topics in the field. One is
the cooption or recruitment of regulatory circuits
and signaling pathways that were already estab-
lished in the primitive condition before a mor-
phological novelty arose. Cooption can include
transcription factors, paracrine signaling proteins,
cell adhesion molecules, and other proteins in-
volved in the regulation of cell behavior. Cooption
has been documented in many instances of
innovation, such as the Shh–Bmp2 interaction in
the development of avian feather germs, which is
necessary for a suite of morphological innovations
in feather structure (Harris et al., 2002; Prum,
2005, this issue). In cephalopod evolution, it has
been shown that Hox orthologues have been
recruited multiple times and in divergence from
the ‘‘colinearity rule,’’ which describes the usual
correspondence between chromosomal order and
spatiotemporal expression of these genes, during
the formation of novel structures (Lee et al., 2003).
Modularity is another frequent theme in the
novelty-related evolution of gene regulation, em-
phasizing the fact that gene recruitment often
concerns not individual genes but entire sets of
regulatory pathways, such as the Notch–Delta
signaling pathway. But modularity also has wider
implications since it encompasses not only genetic
but also epigenetic components and higher levels
of organismal organization (Schlosser and
Wagner, 2004; Callebaut et al., 2005). Modularity
is a way to increase evolvability, i.e., a clade’s
capacity to generate variation (and innovation),
and regulatory modules seem to be able also to
evolve in a (nearly) neutral way, without a direct
effect on the related morphology (Fe
´lix and
Barrie
`re, 2005, this issue). The general modular
nature of genetic and epigenetic developmental
regulation therefore seems not only a means to
generate morphological novelty but also contri-
butes to phenotypic stability in the face of
changing developmental regulation (von Dassow
and Munro, ’99). Hence one of the central issues in
EvoDevo, the relationship between gene regula-
tion and the evolving phenotype, requires an
analysis of modularity in the genotype–phenotype
map (Altenberg, 2005).
Genetic individualization is an emerging topic in
the gene regulation approach to novelty. The
recruitment of molecular signaling modules and
their interaction with organ identity gene net-
works seem to be characteristic modes for estab-
lishing the individuality of morphological
novelties. The establishment of organ identity
networks that are dedicated to the individuation of
structures is required in the evolutionary con-
solidation of novelties, because it confers the
quasi-independence (Lewontin, ’78; Wagner, this
volume) necessary for selection to act on them.
Two examples discussed in the present issue
(Kramer and Jaramillo, 2005; Wagner and Lynch,
2005) suggest that complex patterns of conserva-
tion and divergence characterize the evolution of
identity programs. An unavoidable general short-
coming of the evolutionary genetics program is
that (with few exceptions) we can only study the
genetics of extant organisms. There is no guaran-
tee that the genes associated with present devel-
opmental processes are also the ones that were
causally responsible for the innovations which
they regulate today.
The epigenetic program
The epigenetic approach investigates the generic
properties of developmental systems in the origi-
nation of novelty, explicitly addressing the non-
programmed aspects of development (Newman
and Mu¨ller, 2000). This includes the physical
properties of biological materials, the self-organi-
zational capacities of cell and tissue assemblies,
the dynamics of developmental interactions, the
role of geometry and tissue architecture, the
influence of external and environmental para-
meters, and all other factors that affect the
development of organismal form—regardless of
whether their role in generating novelty is
associated with concurrent changes in the genetic
hardwiring or not. (Epigenetic is here used in its
original meaning, as derived from ‘‘epigenesis’’
and not from ‘‘epigenetics’’ in the sense of
non-DNA-based modulation of gene activity; see
Mu¨ller and Newman, 2003b.)
Epigenetic novelty origination can be tested by
experimental procedures in vivo or in vitro
through the alteration of the epigenetic context
in which a developmental process takes place.
Local and global perturbations of development
MU
¨LLER AND NEWMAN496
yield information about the generative capacity of
developmental systems. Cell dissociation and
aggregation or tissue recombination fall into this
category, as do many classical excision and
transplantation experiments, or alterations of
environmental conditions, which all produce alter-
native morphologies while keeping the genetic
background constant. In vitro recombinations of
different tissue types, for instance, demonstrate
the role of cell surface tension in tissue arrange-
ments (Steinberg, ’63, 2003), and the influence of
extracellular and structural cues (Bissell et al.,
2003). Under natural conditions, recombination of
tissues seems to be a frequent mode of innovation,
as in the origin of external cheek pouches of
geomyoid rodents, where a minor shift of epithe-
lial invagination results in a new tissue interaction
that produces a fur-lined external pouch from an
internal one (Brylski and Hall, ’88a,b).
An exemplary case of novelty generation
through heterotopic tissue recombination is the
turtle carapace (Burke, ’89; Gilbert et al., 2001;
Loredo et al., 2001; Cebra-Thomas et al., 2005, this
issue). Whereas in the primitive condition the
reptilian ribs do not interact with the dermis, in
carapace-forming turtles the ribs and dermal
tissues interact. The initiating condition could
have been selection favoring proportional body
changes that brought the embryonic ribs closer to
the dermis. Once in contact with the dermis the
ribs act as initiating centers for dermal ossifica-
tion. Rib and carapace growth may have subse-
quently become coordinated and stabilized
through an epidermal–mesenchymal signaling
center, the carapacial ridge, recruited from a
preexisting signaling complex. This model pro-
posed by Cebra-Thomas and coworkers contains
all the previously postulated elements for stepwise
novelty generation: epigenetic generation, devel-
opmental integration, and genetic fixation (Mu¨ller
and Newman, ’99). Whether this suite of events
actually took place phylogenetically requires
further corroboration.
Origination and innovation of the vertebrate
limb skeleton is another case in support of
epigenetic initiation discussed in this issue. Most
current models of limb development based on the
genetic regulation of extant limbs sidestep the
mechanistic question of how the skeletal patterns
arose initially and how new elements were added.
Newman and Mu¨ller, 2005a (this issue) describe a
model in which the self-organizational properties
of precartilage mesenchymal tissue provide a basic
template for the limb skeletal pattern that could
later be elaborated upon by evolution. A network
of cell and molecular interactions in limb bud
precartilage cells has been shown both in vitro and
in vivo to provide a core mechanism for the
generation of patterns of discrete aggregates,
leading to nodules and bars of cartilage. In silico
embodiments of this mechanism, employing ‘‘re-
action’’ (i.e., gene regulation and biosynthesis of
key products) and ‘‘diffusion’’ (i.e., local spread of
secreted morphogens), demonstrate that this net-
work is capable of producing ‘‘bare-bones’’ skele-
tal patterns in a spatiotemporally regulated
fashion without the need for programmed control
over the emerging pattern (Hentschel et al., 2004).
By the recruitment of additional cellular and
molecular processes for developmental regulation,
the basic patterns could have become refined,
stabilized, and integrated at both epigenetic and
genetic levels. The subsequent innovation and
addition (or loss) of skeletal elements is proposed
to have occurred also as a consequence of the
biosynthetically reactive chondrogenic mesen-
chyme, in particular when thresholds defined by
the dynamics of the limb skeletogenic system were
exceeded. Other strands of the epigenetic program
that focus on environmental induction, phenotypic
plasticity, phenotypic accommodation, hormonal
influences, and other effects of external and
ecological factors have been, in part, discussed
above. Excellent overviews and detailed examples
can be found in several books and reviews
(Schlichting and Pigliucci, ’98; Gilbert, 2001;
Pigliucci, 2001; Hall et al., 2003; West-Eberhard,
2003).
The theoretical biology program
Theoretical biology serves two functions: The
conceptual characterization of biological problems
(including the history of its ideas, theory relations,
definitions, etc.) and their formalization (using
mathematics, modeling, and simulation). In both
areas the treatment of innovation has made
significant progress over the past decades. Since
Mayr’s (’60) paper, in which innovation was
perceived from an exclusively gradualistic and
variational perspective, the conceptualization of
the innovation problem has greatly diversified.
Individual treatments sharpened the appreciation
of its theoretical importance to a number of
subfields (Frazetta, ’70; Futuyma, ’86; John and
Miklos, ’88; McKinney, ’88). The 1988 symposium
and ensuing volume (Nitecki, ’90) were a first
attempt to integrate the different approaches,
INNOVATION TRIAD: AN EVODEVO AGENDA 497
dealing with innovation and novelty in a variety
of different disciplines including development
(Mu¨ller, ’90; Raff et al., ’90). As a consequence,
the novelty problem was reconceptualized and
placed in an EvoDevo framework (Mu¨ller and
Wagner, ’91). With the spread of EvoDevo,
innovation and novelty became increasingly reg-
ular parts of the theoretical characterization of
several different areas (Gottlieb, ’92; Raff, ’96;
Gerhart and Kirschner, ’97; Carroll et al., 2001;
Wagner, 2001; Hall and Olson, 2003a; Minelli,
2003; Mu¨ller and Newman, 2003a; West-Eberhard,
2003; Callebaut et al., 2005). A growing number of
metatheoretical treatments reflect this intensified
interest (Love, 2003; Robert, 2004; Laubichler and
Maienschein, 2006).
Formalization has also progressed. In particular,
the innovation problem has been addressed in the
fields of biometrics, multivariate statistics, com-
putational modeling, and other areas of bioinfor-
matics. One approach aims at quantifying the
novelty-associated dynamics of gene, cell, and
tissue interactions by developing computational
tools for the three-dimensional representation of
gene expression and other markers of cellular
activity (Streicher et al., 2000; Sharpe et al., 2002;
Weninger and Mohun, 2002), and new algorithms
are being designed for the analysis of such data
(Fontoura Costa et al., 2004).
‘‘Systems biology’’ strategies are also leading to
insights into necessary and sufficient conditions
for phenotypic innovations. For example, dynami-
cal systems-based modeling of cell clusters, such as
would have existed at the beginnings of multi-
cellularity, has disclosed a new physical principle,
‘‘isologous diversification,’’ that may bear on the
origination of cell differentiation (Furusawa and
Kaneko, 2000). In addition, experimentally moti-
vated modeling of specific developmental systems,
such as tooth development (Jernvall, 2000; Jernvall
et al., 2000; Salazar-Ciudad and Jernvall, 2002)
and limb development (Hentschel et al., 2004), has
permitted the devising of simulations that can
illustrate and predict how the differential acti-
vation of genes and gene products influences
morphogenesis and innovation. The question of
innovation has also motivated a recategorization
of mechanisms of developmental pattern forma-
tion in terms of their likehood, under mutational
change, to produce morphological novelties (Salazar-
Ciudad et al., 2001a,b).
Consideration of the dynamical and topological
properties of the gene networks associated with
developmental mechanisms provides additional
insight into sources of innovation and its comple-
ment, developmental, and evolutionary robust-
ness. In silico (Salazar-Ciudad et al., 2001a) or
artificial biochemical–genetic (Isalan et al., 2005)
models for such networks have disclosed generic
properties that promote the formation of novel
patterns. Computational models for developmen-
tal genetic networks can be caused to undergo
simulated evolution and can help identify net-
work constructional principles that render the
system susceptible to significant deviations from
an initial pattern (i.e., novelty) or resistant to
such deviations (i.e., robustness) (von Dassow
et al., 2000; Salazar-Ciudad et al., 2001a; Ingolia,
2004).
Modeling analyses have also led to questioning
the widely assumed gradualism of phenotypic
innovation, the neo-Darwinian default which,
until recently, had little challenge from experi-
mentally based theoretical models. Salazar-Ciudad
and Jernvall, 2005 (this issue), considering tooth
development as an evolving ‘‘morphodynamic’’
process, show that gradual variation is only
expected to occur in relatively simple phenotypes
whereas complex phenotypes exhibit less gradual
variation and have a tendency towards early
accelerations vs. late decelerations of innovation
rates during phylogenetic diversification. These
predictions seem to correlate with the evidence
from the fossil record (Erwin, ’93).
The morphospace concept provides another
fertile theoretical approach to innovation. ‘‘Gene-
rative morphospaces’’ in particular provide a tool
by which the range of possible patterns that are
produced from a set of developmental rules can be
compared with forms that did or did not appear in
natural systems (Thomas and Reif, ’93; Eble, 2001;
Rasskin-Gutman, 2003). These models can be used
to detect general rules that underlie the patterns of
phenotypic variation and innovation and to derive
predictions about the generative capacities of a
given developmental system. Morphospace model-
ing indicates that only a limited number of
phenotypic solutions can be obtained from a given
developmental system, even in the presence of
ample genetic variation. But these effects are not
only limitational. Certain morphological solutions
are more likely to arise than others, independent of
the molecular and genetic circuitry associated with
their generation, pointing to inherent properties
of the developmental system involved (McGhee,
’99; Stadler et al., 2001). Similar results arise from
the modeling of RNA genotype–phenotype maps
(Fontana, 2001, 2002).
MU
¨LLER AND NEWMAN498
IMPLICATIONS OF A RESEARCH FOCUS
ON INNOVATION
The multitude of concepts and programs re-
viewed above indicates a growing awareness in
evolutionary biology of the problems of the
innovation triad. This change reflects a shift of
attention from the phenomenon of variation and
the modes of its preservation toward the phenom-
enon of innovation and the modes of its genera-
tion. At the same time there is a general hesitation
to accept the consequences of such a change of
conceptual focus. There is concern that treating
innovation as a distinct problem might undermine
the basic neo-Darwinian framework (e.g., Hall,
2005, this issue). This need not be the case,
though. Selection has a causal role both in
adaptation and in innovation. But since morpho-
logical novelty denotes a qualitative deviation
from the purely quantitative phenomena of varia-
tion (the focus of the neo-Darwinian framework),
a shift of explanatory weight is required. In
adaptation, the motive force resides in natural
selection acting on an underlying substrate—
heritable variation—the necessary boundary con-
dition. In innovation, natural selection represents
the boundary condition, whereas the properties of
developmental systems provide the motive force
for the ensuing change.
As we become more familiar with the themes of
the innovation triad, and more information is
generated on its various subtopics, the advantages
of making the variation–innovation distinction
will become clearer. On the one hand, it esta-
blishes a new epistemological position in the study
of the phenotypic level of organismal evolution, a
position that is not limited by the necessity to
conform to the variation–adaptation paradigm.
The limitations implicit in the requirement to
perceive all morphological change as a gradual
transition series, e.g., feathers from scales, or
limbs from lateral folds, has impeded the under-
standing of phenotypic evolution. On the other
hand, it assigns a central role to internal causa-
tion, as advocated earlier (Bateson, 1894; Baldwin,
1896; Roth and Wake, ’85), and thus liberates
evolutionary theorists from the requirement to
search for purely external causes of phenotypic
evolution, as implied by the adaptationist para-
digm. As a consequence, a suite of new empirical
research programs concentrating on the mechan-
istic causes of innovation become possible and
complement other guiding themes of EvoDevo
such as constraint, modularity, and homology.
Therefore, the variation–innovation distinction
does not oppose neo-Darwinism but, on the
contrary, has a distinctly synthesizing effect on
evolutionary theory, because innovation requires
the inclusion of development (Love, 2003).
The new general tenets that will be added to
evolutionary theory by a focus on innovation may
be called ‘‘emergence’’ and ‘‘inherency’’. Emer-
gence has for a long time been seen as an essential
property of evolving complex systems, such as
development, without formal representation in
the theory of evolution. Emergence denotes the
non-foreseeable element in the processes of origi-
nation and innovation, for example, when the
recombination of two (or more) preexisting com-
ponents leads to unpredicted results. In his
discussion of innovation, and criticism of the
received terminology, Lorenz (’73) had suggested
calling the evolutionary recombination phenom-
enon ‘‘fulguration’’, emphasizing the suddenness
and historical singularity of such events. While his
suggestion is unlikely to be followed, it emphasizes
the long-felt need to account for emergence in
evolutionary theory. A mechanistic concept of
innovation could fill this void by moving beyond
the neo-Darwinian focus on variation–selection
dynamics which implies a pervasively gradualistic
model of evolution.
Inherency is a second general property of evolving
systems. It complements the contingency empha-
sized by the neo-Darwinian side of evolutionary
theory. Whereas historical contingency denotes the
lawful dependency of evolutionary change on earlier
conditions that involved a large component of
chance, inherency represents the tendency to
organize and change along preferred routes, leading,
unless inhibited, to predictable outcomes (Newman
and Mu¨ller, 2005b). Emergence and inherency
represent those generative principles that are miss-
ing from the standard evolutionary framework and
which are now in the process of being incorporated
into a more complete theory by EvoDevo.
ACKNOWLEDGMENTS
G.B.M. acknowledges support from the Konrad
Lorenz institute for Evolution and Cognition
Research, and S.A.N. acknowledges support from
the National Science Foundation (IBN-0083653
and IBN-0344647).
REFERENCES
Altenberg L. 2005. Modularity in evolution: some low-level
questions. In: Callebaut W, Rasskin-Gutman D, editors.
INNOVATION TRIAD: AN EVODEVO AGENDA 499
Modularity understanding the development and evolution of
natural complex systems. Cambridge, MA: MIT Press.
p 99–128.
Arthur W. 2001. Developmental drive: an important determinant
of the direction of phenotypic evolution. Evol Dev 3:271–278.
Averof M, Cohen SM. 1997. Evolutionary origin of insect
wings from ancestral gills. Nature 385:627–630.
Baldwin JM. 1896. A new factor in evolution. Am Nat
30:441–451, 536–553.
Bateson W. 1894. Materials for the study of variation treated
with especial regard to discontinuity in the origin of species.
London: Macmillan.
Bissell MJ, Mian S, Radisky D, Turley E. 2003. Tissue
specificity: structural cues allow diverse phenotypes from a
constant genotype. In: Mu¨ller GB, Newman SA, editors.
Origination of organismal form. Boston, MA: MIT Press.
p 103–117.
Bolouri H, Davidson EH. 2003. Transcriptional regulatory
cascades in development: initial rates, not steady state,
determine network kinetics. Proc Natl Acad Sci USA
100:9371–9376.
Bradshaw HD, Schemske DW. 2003. Allele substitution at a
flower colour locus produces a pollinator shift in monkey-
flowers. Nature 426:176–178.
Brylski P, Hall BK. 1988a. Epithelial behaviors and threshold
effects in the development and evolution of internal and
external cheek pouches in rodents. Z Zool Syst Evol 26:144–154.
Brylski P, Hall BK. 1988b. Ontogeny of a macroevolutionary
phenotype: the external cheek pouches of geomyoid rodents.
Evolution 42:391–395.
Bu¨ rger R. 1986. Constraints for the evolution of functionally
coupled characters: a nonlinear analysis of a phenotypic
model. Evolution 40:182–193.
Burke AC. 1989. Development of the turtle carapace: implica-
tions for the evolution of a novel Bauplan. J Morphol
199:363–378.
Burke AC, Nelson CE, Morgan BA, Tabin C. 1995. Hox genes
and the evolution of vertebrate axial morphology. Develop-
ment 121:333–346.
Buss LW. 1987. The evolution of individuality. New York:
Columbia University Press. 201p.
Callebaut W, Rasskin-Gutman D. 2005. Modularity: under-
standing the development and evolution of complex natural
systems. Cambridge, MA: MIT Press.
Carroll SB, Grenier JK, Weatherbee SD. 2001. From DNA to
diversity. Malden: Blackwell Science.
Cebra-Thomas J, Tan F, Sistla S, Estes E, Bender G, Kim C,
Riccio P, Gilbert SF. 2005. How the turtle forms its shell:
a paracrine hypothesis of carapace formation. J Exp Zool B
(Mol Dev Evol) 304B:558–569.
Cinquin O, Demongeot J. 2005. High-dimensional switches
and the modelling of cellular differentiation. J Theor Biol
233:391–411.
Colosimo PF, Hosemann KE, Balabhadra S, Villarreal G Jr,
Dickson M, Grimwood J, Schmutz J, Myers RM, Schluter D,
Kingsley DM. 2005. Widespread parallel evolution in
sticklebacks by repeated fixation of Ectodysplasin alleles.
Science 307:1928–1933.
Crawford M. 2003. Hox genes as synchronized temporal
regulators: implications for morphological innovation.
J Exp Zool B Mol Dev Evol 295:1–11.
Darwin C. 1859. On the origin of species. London: John Murray.
Davidson EH. 2001. Genomic regulatory systems: develop-
ment and evolution. San Diego: Academic Press.
Eberhard WG. 2001. Multiple origins of a major novelty:
moveable abdominal lobes in male sepsid flies (Diptera:
Sepsidae) and the question of developmental constraints.
Evol Dev 3:206–222.
Eble GJ. 1999. Originations: land and sea compared. Geobios
32:223–234.
Eble GJ. 2001. Multivariate approaches to development and
evolution. In: Minugh-Purvis N, McNamara KJ, editors.
Human evolution through developmental change. Balti-
more, MD: Johns Hopkins University Press.
Emlen JM, Freeman DC, Mills A, Graham JH. 1998. How
organisms do the right thing: the attractor hypothesis.
Chaos 8:717–726.
Erwin DH. 1993. Early introduction of major morphological
innovations. Acta Palaeontol Polon 38:281–294.
Fe
´lix M-A, Barrie
`re A. 2005. Evolvability of cell specification
mechanisms. J Exp Zool B (Mol Dev Evol) 304B:536–547.
Fontana W. 2001. Novelty in evolution. Bioevolutionary
Concepts for NASA, BEACON.
Fontana W. 2002. Modelling ‘evo-devo’ with RNA. Bioessays
24:1164–1177.
Fontoura Costa L, Barbosa MS, Manoel ET, Streicher J, Mu¨ller
GB. 2004. Mathematical characterization of three-dimensional
gene expression patterns. Bioinformatics: 20:1653–1662.
Frazetta TH. 1970. From hopeful monsters to bolyerine
snakes? Am Nat 104:55–72.
Frazzetta TH. 1975. Pattern and instability in the evolving
premaxilla of bolyerine snakes. Am Zool 15:469–481.
Furusawa C, Kaneko K. 2000. Complex organization in multi-
cellularity as a necessity in evolution. Artif Life 6:265–281.
Futuyma DJ. 1986. Evolutionary biology. Sunderland: Sinauer.
Galis F. 1996. The application of functional morphology to
evolutionary studies. Tree 11:124–129.
Galis F, Drucker EG. 1996. Pharyngeal biting mechanics in
centrarchid and cichlid fishes: insights into a key evolu-
tionary innovation. J Evol Biol 9:641–670.
Ganfornina MD, Sanchez D. 1999. Generation of evolutionary
novelty by functional shift. Bioessays 21:432–439.
Gaunt SJ. 2000. Evolutionary shifts of vertebrate structures and
Hox expression up and down the axial series of segments: a con-
sideration of possible mechanisms. Int J Dev Biol 44:109–117.
Gerhart J, Kirschner. 1997. Cells, Embryos, and Evolution.
Oxford: Blackwell Science.
Gilbert SF. 2001. Ecological developmental biology: develop-
mental biology meets the real world. Dev Biol 233:1–32.
Gilbert SF, Loredo GA, Brukman A, Burke AC. 2001.
Morphogenesis of the turtle shell: the development of a
novel structure in tetrapod evolution. Evol Dev 3:47–58.
Goldschmidt RB. 1940. The material basis of evolution. New
Haven: Yale University Press.
Gottlieb G. 1992. Individual development and evolution:
the genesis of novel behavior. Oxford: Oxford University
Press.
Gottlieb G. 2002. Individual development and evolution: the
genesis of novel behavior. Mahwah, NJ: Erlbaum.
Gottlieb LD. 1984. Genetics and morphological evolution in
plants. Am Nat 123:681–709.
Gould SJ, Lewontin RC. 1979. The spandrels of San Marco
and the Panglossian paradigm: a critique of the adap-
tationist programme. Proc R Soc Lond B Biol Sci 205:
581–598.
Hall BK. 2005. Consideration of the neural crest and its
skeletal derivatives in the context of novelty/innovation.
J Exp Zool B (Mol Dev Evol) 304B:548–557.
MU
¨LLER AND NEWMAN
500
Hall BK, Olson WM, editors. 2003. Keywords and concepts
in evolutionary developmental biology. Cambridge, MA:
Harvard University Press.
Hall BK, Pearson BJ, Mu¨ller GB, editors. 2003. Environment,
development, and evolution. Cambridge, MA: MIT Press.
Hanken J. 1985. Morphological novelty in the limb skeleton
accompanies miniaturization in Salamanders. Science 229:
871–874.
Harris MP, Fallon JF, Prum RO. 2002. Shh-Bmp2 signaling
module and the evolutionary origin and diversification of
feathers. J Exp Zool 294:160–176.
Hentschel HG, Glimm T, Glazier JA, Newman SA. 2004.
Dynamical mechanisms for skeletal pattern formation in the
vertebrate limb. Proc R Soc Lond B Biol Sci 271:1713–1722.
Ingolia NT. 2004. Topology and robustness in the Drosophila
segment polarity network. PLoS Biol 2:805–815.
Isalan M, Lemerle C, Serrano L. 2005. Engineering gene
networks to emulate Drosophila embryonic pattern forma-
tion. PLoS Biol 3:488–496.
Jablonski D. 2005. Evolutionary innovations in the fossil
record: the intersection of ecology, development, and
macroevolution. J Exp Zool B (Mol Dev Evol) 304B:504–519.
Jablonski D, Bottjer DJ. 1990. The ecology of evolutionary
innovation. In: Nitecki MH, editor. Evolutionary innova-
tions. Chicago: The University of Chicago Press. p 253–288.
Jayne BC, Voris HK, Ng PK. 2002. Snake circumvents
constraints on prey size. Nature 418:143.
Jernvall J. 2000. Linking development with generation of
novelty in mammalian teeth. Proc Natl Acad Sci USA 97:
2641–2645.
Jernvall J, Keranen SV, Thesleff I. 2000. Evolutionary
modification of development in mammalian teeth: quantify-
ing gene expression patterns and topography. Proc Natl
Acad Sci USA 97:14444–14448.
John B, Miklos LG. 1988. The eukaryote genome in develop-
ment and evolution. London: Allen & Unwin.
Johnston TD, Gottlieb G. 1990. Neophenogenesis: a develop-
mental theory of phenotypic evolution. J Theor Biol 147:
471–495.
Keys DN, Lewis DL, Selegue JE, Pearson BJ, Goodrich LV,
Johnson RL, Gates J, Scott MP, Carroll SB. 1999. Recruit-
ment of a hedgehog regulatory circuit in butterfly eyespot
evolution. Science 283:532–534.
Kramer EM, Jaramillo MA. 2005. Genetic basis for innova-
tions in floral organ identity. J Exp Zool B (Mol Dev Evol)
304B:526–535.
Laubichler M, Maienschein J, editors. 2006. From embryology
to evo-devo: a history of embryology in the 20th century.
Cambridge, MA: MIT Press, in press.
Lee PN, Callaerts P, De Couet HG, Martindale MQ. 2003.
Cephalopod Hox genes and the origin of morphological
novelties. Nature 424:1061–1065.
Lewontin RC. 1978. Adaptation. Am Sci 239:156–169.
Liem KF. 1974. Evolutionary strategies and morphological
innovations: cichlid pharyngeal jaws. Syst Zool 22:
425–441.
Liem KF. 1990. Key evolutionary innovations, differential
diversity, and symecomorphosis. In: Nitecki MH, editor.
Evolutionary innovations. Chicago: University of Chicago
Press. p 147–170.
Loredo GA, Brukman A, Harris MP, Kagle D, LeClair EE,
Gutman R, Denney E, Henkelman E, Murray BP, Fallon JF,
Tuan RS, Gilbert SF. 2001. Development of an evolutiona-
rily novel structure: fibroblast growth factor expression in
the carapacial ridge of turtle embryos. 1–24.
Lorenz K. 1973. Die Ru¨ckseite des Spiegels. Mu¨nchen: Piper
Verlag.
Love AC. 2003. Evolutionary morphology, innovation, and the
synthesis of evolutionary and developmental biology. Biol
Philos 18:309–345.
Marden JH, Kramer MG. 1994. Sufrace-skimming stone flies:
a possible intermediate in insect flight evolution. Science
266:427–430.
Margulis L, Fester R, editors. 1991. Symbiosis as a source of
evolutionary innovation. Cambridge, MA: MIT Press.
Maynard-Smith J, Szathmary E. 1995. The major transitions
in evolution. Oxford, New York; Heidelberg: W.H. Freeman.
Maynard Smith J, Burian R, Kauffman S, Alberch P, Camp-
bell J, Goodwin B, Lande R, Raup D, Wolpert L. 1985.
Developmental constraints and evolution. Quart Rev Biol
60:265–287.
Mayr E. 1958. Behavior and systematics. In: Roe A, Simpson
GG, editors. Behavior and evolution. New Haven: Yale
University Press. p 341–362.
Mayr E. 1960. The emergence of evolutionary novelties. In:
Tax S, editor. Evolution after Darwin. Cambridge, MA:
Harvard University Press. p 349–380.
Mayr E. 1961. Cause and effect in biology: kinds of causes,
predictability, and teleology are viewed by a practicing
biologist. Science 134:1501–1506.
McGhee GR. 1999. Theoretical morphology. New York:
Columbia University Press.
McKinney FK. 1988. Multidisciplinary perspectives on evolu-
tionary innovations. Tree 3:220–222.
Michod RE. 1999. Darwinian dynamics: evolutionary transi-
tions in fitness and individuality. Princeton, NJ: Princeton
University Press.
Minelli A. 2003. The development of animal form: ontogeny,
morphology, and evolution. Cambridge: Cambridge Univer-
sity Press.
Mivart SG. 1871. On the genesis of the species. London:
Mcmillan.
Minelli A, Fusco G. 2005. Conserved versus innovative
features in animal body organization. J Exp Zool B (Mol
Dev Evol) 304B:520–525.
Mu¨ ller GB. 1990. Developmental mechanisms at the origin of
morphological novelty: A side-effect hypothesis. In: Nitecki
MH, editor. Evolutionary innovations. Chicago: The Uni-
versity of Chicago Press. p 99–130.
Mu¨ ller GB. 2003a. Embryonic motility: environmental influ-
ences and evolutionary innovation. Evol Dev 5:56–60.
Mu¨ ller GB. 2003b. Homology: the evolution of morphological
organization. In: Mu¨ ller GB, Newman SA, editors. Origina-
tion of organismal form. Boston: MIT Press. p 51–69.
Mu¨ ller GB. 2005. Evolutionary developmental biology. In:
Wuketits FM, Ayala FJ, editors. Handbook of evolution.
Weinheim: Wiley-VCH. p 87–115.
Mu¨ ller GB. 2006. Six memos for EvoDevo. In: Laubichler MD,
Maienschein J, editors. From embryology to evo-devo: a
history of embryology in the 20th century. Cambridge: MIT
Press, Dibner Series, in press.
Mu¨ ller GB, Newman SA. 1999. Generation, integration,
autonomy: three steps in the evolution of homology.
Novartis Found Symp 222:65–73, discussion 73-69.
Mu¨ ller GB, Newman SA, editors. 2003a. Origination of
organismal form. Cambridge, MA: MIT Press.
INNOVATION TRIAD: AN EVODEVO AGENDA 501
Mu¨ ller GB, Newman SA. 2003b. Origination of organismal
form: The forgotten cause in evolutionary theory. In: Mu¨ller
GB, Newman SA, editors. Origination of organismal form.
Boston: MIT Press. p 3–10.
Mu¨ ller GB, Wagner GP. 1991. Novelty in evolution: restruc-
turing the concept. Annu Rev Ecol Syst 22:229–256.
Mu¨ ller GB, Wagner GP. 2003. Innovation. In: Hall BK, Olson
W, editors. Keywords and concepts in evolutionary develop-
mental biology. Cambridge, MA: Cambridge University
Press. p 218–227.
Newman SA. 1992. Generic physical mechanisms of morpho-
genesis and pattern formation as determinants in the
evolution of multicellular organization. In: Mittenthal JB,
Baskin AB, editors. Principles of organization in organisms,
SFI studies in the sciences of complexity. Reading, MA:
Addison-Wesley. p 241–267.
Newman SA. 1994. Generic physical mechanisms of tissue
morphogenesis: a common basis for development and
evolution. J Evol Biol 7:467–488.
Newman SA. 2005. The pre-Mendelian, pre-Darwinian world:
shifting relations between genetic and epigenetic mechan-
isms in early multicellular evolution. J Biosci 30:75–85.
Newman SA, Comper WD. 1990. ‘Generic’ physical mecha-
nisms of morphogenesis and pattern formation. Develop-
ment 110:1–18.
Newman SA, Mu¨ ller GB. 2005b. Origination and innovation
in the vertebrate limb skeleton: an epigenetic perspective.
J Exp Zool B (Mol Dev Evol) 304B:593–609.
Newman SA, Mu¨ller GB. 2000. Epigenetic mechanisms of
character origination. J Exp Zool (Mol Dev Evol) 288:
304–317.
Newman SA, Mu¨ller GB. 2005b. Inherency, interaction, and
integration in the evolution of developmental mechansims.
In: Rehmann-Sutter C, Neumann-Held E, editors. Genes in
development rereading the molecular paradigm, Duke
University Press, in press.
Nijhout HF. 2001. Origin of butterfly wing patterns. In:
Wagner GP, editor. The character concept in evolutionary
biology. San Diego: Academic Press. p 511–529.
Nitecki MH, editor. 1990. Evolutionary innovations. Chicago:
The University of Chicago Press. 304p.
Nowicki JL, Burke AC. 2000. Hox genes and morphological
identity: axial versus lateral patterning in the vertebrate
mesoderm. Development 127:4265–4275.
Pass G. 2000. Accessory pulsatile organs: evolutionary
innovations in insects. Rev Entomol 45:495–518.
Pigliucci M. 2001. Phenotypic plasticity: beyond nature and
nurture. Baltimore: Johns Hopkins University Press.
Prum RO. 2005. Evolution of the morphological innovations of
feathers. J Exp Zool B (Mol Dev Evol) 304B:570–579.
Raff R. 1996. The shape of life. Chicago: Chicago University
Press.
Raff RA, Parr BA, Parks AL, Wray GA. 1990. Heterochrony
and other mechanisms of radical evolutionary change in
early development. In: Nitecki MH, editor. Evolutionary
innovations. Chicago: The University of Chicago Press.
p 71–98.
Rasskin-Gutman D. 2003. Boundary constraints for the emer-
gence of form. In: Mu¨ller GB, Newman SA, editors. Origina-
tion of organismal form. Boston: MIT Press. p 305–322.
Rienesl J, Wagner GP. 1992. Constancy and change of
basipodial variation patterns: a comparative study of crested
and marbled newts—Triturus cristatus,Triturus marmor-
atus—and their natural hybrids. J Evol Biol 5:307–324.
Robert JS. 2004. Embryology, epigenesis, and evolution:
taking development seriously. Cambridge: Cambridge Uni-
versity Press.
Rose MR, Lauder GV, editors. 1996. Adaptation. San Diego:
Academic Press.
Roth G, Wake DB. 1985. Trends in the functional morphology
and sensorimotor control of feeding behavior in salaman-
ders: an example of the role of internal dynamics in
evolution. Acta Biotheor 34:175–191.
Roth G, Wake DB. 1989. Conservatism and innovation in the
evolution of feeding in vertebrates. In: Wake DB, Roth G,
editors. Complex organismal functions: integration and
evolution in vertebrates. Chichester: John Wiley & Sons.
p 7–21.
Salazar-Ciudad I, Jernvall J. 2002. A gene network model
accounting for development and evolution of mammalian
teeth. Proc Natl Acad Sci USA 99:8116–8120.
Salazar-Ciudad I, Jernvall J. 2005. Graduality and innovation
in the evolution of complex phenotypes: insights from
development. J Exp Zool B (Mol Dev Evol) 304B:619–631.
Salazar-Ciudad I, Newman SA, Sole RV. 2001a. Phenotypic
and dynamical transitions in model genetic networks. I.
Emergence of patterns and genotype–phenotype relation-
ships. Evol Dev 3:84–94.
Salazar-Ciudad I, Sole RV, Newman SA. 2001b. Phenotypic
and dynamical transitions in model genetic networks. II.
Application to the evolution of segmentation mechanisms.
Evol Dev 3:95–103.
Salazar-Ciudad I, Jernvall J, Newman SA. 2003. Mechanisms
of pattern formation in development and evolution. Deve-
lopment 130:2027–2037.
Schlichting C, Pigliucci P. 1998. Phenotypic evolution: a reaction
norm perspective. Sunderland MA: Sinauer Associates.
Schlosser G, Wagner GP, editors. 2004. Modularity in
development and evolution. Chicago: University of Chicago
Press.
Schmalhausen II. 1949. Factors of evolution. Chicago: The
University of Chicago Press.
Schwenk K, Wagner GP. 2003. Constraint. In: Hall BK, Olson
W, editors. Keywords and concepts in evolutionary develop-
mental biology. Cambridge: Cambridge University Press.
p 52–61.
Sewertzoff AN. 1931. Morphologische Gesetzma
¨Xigkeiten der
Evolution. Jena: Fischer.
Shapiro MD, Marks ME, Peichel CL, Blackman BK, Nereng
KS, Jonsson B, Schluter D, Kingsley DM. 2004. Genetic and
developmental basis of evolutionary pelvic reduction in
threespine sticklebacks. Nature 428:717–723.
Sharpe J, Ahlgren U, Perry P, Hill B, Ross A, Hecksher-
Sorensen J, Baldock R, Davidson D. 2002. Optical projection
tomography as a tool for 3D microscopy and gene expression
studies. Science 296:541–545.
Shubin N, Tabin C, Carroll S. 1997. Fossils, genes and the
evolution of animal limbs. Nature 388:639–648.
Simpson GG. 1953. The Baldwin effect. Evolution 7:110–117.
Sordino P, van der Hoeven F, Duboule D. 1995. Hox gene
expression in teleost fins and the origin of vertebrate digits.
Nature 375:678–681.
Stadler BM, Stadler PF, Wagner GP, Fontana W. 2001. The
topology of the possible: formal spaces underlying patterns
of evolutionary change. J Theor Biol 213:241–274.
Steinberg MS. 1963. Reconstruction of tissues by dissociated
cells. Science 141:401–408.
MU
¨LLER AND NEWMAN
502
Steinberg MS. 2003. Cell adhesive interaction and tissue
self-organization. In: Mu¨ller GB, Newman SA, editors.
Origination of organismal form. Boston: MIT Press. p
137–163.
Stone JR, Hall BK. 2004. Latent homologues for the neural
crest as an evolutionary novelty. Evol Dev 6:123–129.
Streicher J, Donat MA, Strauss B, Spo
¨rle R, Schughart K,
Mu¨ller GB. 2000. Computer based three-dimensional visuali-
zation of developmental gene expression. Nat Genet 25:
147–152.
Tautz D. 1996. Selector genes, polymorphisms, and evolution.
Science 271:160–161.
Thomas RDK, Reif W-E. 1993. The skeleton space: a finite set
of organic designs. Evolution 47:341–360.
von Dassow G, Munro E. 1999. Modularity in animal
development and evolution: elements of a conceptual
framework for EvoDevo. J Exp Zool 285:307–325.
von Dassow G, Meir E, Munro EM, Odell GM. 2000. The
segment polarity network is a robust developmental module.
Nature 406:188–192.
Waddington CH. 1956. Genetic assimilation. Adv Genet 10:
257–290.
Waddington CH (editor). 1962. New patterns in genetics and
development. New York: Columbia University Press.
Wagner GP. 1984. Coevolution of functionally constrained
characters: prerequisites for adaptive versatility. BioSys-
tems 17:51–55.
Wagner GP. 1989. The biological homology concept. Annu Rev
Ecol Syst 20:51–69.
Wagner GP, editor. 2001. The character concept in evolu-
tionary biology. San Diego: Academic Press.
Wagner GP, Chiu C. 2003. Genetic and epigenetic factors in
the origin of the tetrapod limb. In: Mu¨ller GB, Newman SA,
editors. Origination of organismal form. Cambridge, MA:
MIT Press. p 265–285.
Wagner GP, Mu¨ ller GB. 2002. Evolutionary innovations overcome
ancestral constraints: A re-examination of character evolution
in male sepsid flies (Diptera: Sepsidae). Evol Dev 4:1–6.
Wagner GP, Lynch VJ. 2005. Molecular evolution of evolu-
tionary novelties: the vagina and uterus of placental
mammals. J Exp Zool B (Mol Dev Evol) 304B:580–592.
Wagner GP, Chiu C, Laubichler M. 2000. Developmental
evolution as a mechanistic science: The inference from
developmental mechanisms to evolutionary processes. Am
Zool 40:819–831.
Weninger WJ, Mohun T. 2002. Phenotyping transgenic
embryos: A rapid 3D screening method based on episcopic
fluorescence image capturing. Nat Genet 30:59–65.
West-Eberhard MJ. 2003. Developmental plasticity and
evolution. Oxford: Oxford University Press.
West-Eberhard MJ. 2005. Phenotypic accommodation: adap-
tive innovation due to developmental plasticity. J Exp Zool B
(Mol Dev Evol) 304B:610–618.
Wilkins A. 2002. The evolution of developmental pathways.
Sunderland, MA: Sinauer Associates.
Williams G. 1966. Adaptation and natural selection. Prince-
ton, NJ: Princeton University Press.
Wimsatt WC. 1986. Developmental constraints, generative
entrenchment, and the innate-acquired distinction. In:
Bechtel W, editor. Integrating scientific disciplines. Dor-
drecht: Martinus Nijhoff Publishers. p 185–208.
Wray. 1999. Evolutionary dissociations between homologous
genes and homologous structures. In: Bock GR, Cardew G,
editors. Homology. Chichester: Wiley. p 189–203.
Wray GA, Lowe CJ. 2000. Developmental regulatory genes
and echinoderm evolution. Syst Biol 49:151–174.
INNOVATION TRIAD: AN EVODEVO AGENDA 503
