683
The origin of the pelagobenthic metazoan life cycle:
what’s sex got to do with it?
Sandie M. Degnan1 and Bernard M. Degnan
School of Integrative Biology, The University of Queensland, Brisbane, Queensland 4072, Australia
Synopsis The biphasic (pelagobenthic) life cycle is found throughout the animal kingdom, and includes gametogenesis,
embryogenesis, and metamorphosis. From a tangled web of hypotheses on the origin and evolution of the metazoan
pelagobenthic life cycle, current opinion appears to favor a simple, larval-like holopelagic ancestor that independently settled
multiple times to incorporate a benthic phase into the life cycle. This hypothesis derives originally from Haeckel’s (1874)
Gastraea theory of ontogeny recapitulating phylogeny, in which the gastrula is viewed as the recapitulation of a gastraean
ancestor that evolved via selection on a simple, planktonic hollow ball of cells to develop the capacity to feed. Here, we propose
an equally plausible hypothesis that the origin of the metazoan pelagobenthic life cycle was a direct consequence of sexual
reproduction in a likely holobenthic ancestor. In doing so, we take into account new insights from poriferan development and
from molecular phylogenies. In this scenario, the gastrula does not represent a recapitulation, but simply an embryological
stage that is an outcome of sexual reproduction. The embryo can itself be considered as the precursor to a biphasic lifestyle,
with the embryo representing one phase and the adult another phase. This hypothesis is more parsimonious because it
precludes the need for multiple, independent origins of the benthic form. It is then reasonable to consider that multilayered,
ciliated embryos ultimately released into the water column are subject to natural selection for dispersal/longevity/feeding that
sets them on the evolutionary trajectory towards the crown metazoan planktonic larvae. These new insights from poriferan
development thus clearly support the intercalation hypothesis of bilaterian larval evolution, which we now believe should be
extended to discussions of the origin of biphasy in the metazoan last common ancestor.
Introduction
Extant metazoans exhibit a wide range of life cycles, but
the biphasic life cycle, which includes distinct larval and
adult phases, is by far the most common and widespread. Indeed, nearly all metazoan phyla—including
basal metazoans, such as poriferans and cnidarians, and
a wide range of bilaterian phyla—have marine representatives characterized by a specific kind of biphasy,
namely the pelagobenthic life cycle (Pechenik 2004).
This is defined by a benthic, sexually reproductive
adult phase that generates gametes which are either
broadcast and fertilized in the water column, or fertilized and brooded internally. The resultant embryos
and pelagic larvae, which are typically microscopic,
live for a variable time in the plankton before settling
down onto the benthic substrate. At this time, they
undergo metamorphosis—which varies in speed and
magnitude—back into a benthic adult form, which is
usually sessile or sedentary.
Given the extremely widespread nature of the pelagobenthic biphasic life cycle throughout the Metazoa,
its origins have received remarkably little attention in
the literature. This is in contrast to the extensive and
excellent body of literature that goes a long way toward
explaining diversification of extant larval forms, and
which is not addressed here. There exists a relatively
small, and somewhat tangled, web of hypotheses on the
origins of biphasy that continues to be discussed and
debated (for example, see reviews in Bishop and
Brandhorst 2003; Sly and others 2003). From these
hypotheses, 2 general schools of thought are
emerging—the “terminal addition” school (for example, Nielsen 1979, 1985, 2000, 2001, 2003; Nielsen and
Norrevang 1985; Davidson and others 1995; Peterson
and others 1997, 2000) and the “intercalation” school
(for example, Wolpert 1999; Valentine and Collins
2000; Sly and others 2003). A major limitation of
these ideas is that they have been formed almost
exclusively by perspectives and data from bilaterian
metazoans, with only a cursory nod toward the
basal, non-bilaterian phyla. It is now well accepted
that Metazoa is a monophyletic group (Fig. 1;
Collins 1998; Borchiellini and others 2001; Medina
and others 2001; Glenner and others 2004) and that
From the symposium “Metamorphosis: A Multikingdom Approach” presented at the annual meeting of the Society for Integrative and
Comparative Biology, January 4–8, 2006, at Orlando, Florida.
1 E-mail: [email protected]
Integrative and Comparative Biology, volume 46, number 6, pp. 683–690
doi:10.1093/icb/icl028
Advance Access publication August 15, 2006
Ó The Author 2006. Published by Oxford University Press on behalf of the Society for Integrative and Comparative Biology. All rights reserved.
For permissions please email: [email protected].
684
Fig. 1 Alternative hypotheses of inferred relationships
of the 3 sponge classes with each other and with other
metazoans based on recent molecular studies. (A)
Monophyletic Porifera. (B) Paraphyletic Porifera with
the Calcarea more closely related to eumetazoans, in
particular cnidarians (dashed line). (C) Ancient
divergence of sponge classes such that phylogenetic
relationships between these classes, cnidarians, and
bilaterians are unresolved (for example, Rokas and
others 2005).
the basic concept of biphasy predates metazoan cladogenesis and already existed in the last common ancestor
(LCA) of all extant animals. As such, we argue that
there is little relevance to a debate about the ancestry
of biphasy in the Bilateria. Instead, the focus should be
on the origin of biphasy relative to the LCA of the
Metazoa.
To this end, current weight of opinion (for example,
Martindale 2005) on the origin of metazoan (as opposed
to bilaterian) biphasy lies with a simple, larval-like holopelagic ancestor that independently settled multiple
times to incorporate by “terminal addition” a benthic
phase into the life cycle, as illustrated predominantly
S. M. Degnan and B. M. Degnan
by the Trochaea hypothesis (Nielsen 1979, 2001; Nielsen
and Norrevang 1985). This hypothesis derives in the
general sense from Haeckel’s (1874) Gastraea theory
of ontogeny recapitulating phylogeny, an idea which
has had a profound influence on considerations of animal evolution. Haeckel viewed the gastrula of all extant
animals as the recapitulation of a gastraean ancestor that
evolved invagination via selection on a simple, planktonic hollow ball of cells (Blastaea) LCA to develop the
capacity to feed. The Gastraea theory fell from favor
for almost a century, until it was resuscitated and elaborated by Jägersten (1955, 1972) and especially by
Claus Nielsen’s Trochaea theory in a series of thoughtful, and thought-provoking, papers beginning with
Nielsen (1979) and Nielsen and Norrevang (1985)
and continuing to the present day (for example,
Nielsen 2001, 2003, 2005). These hypotheses require
that several conditions are met. First, they require multiple, independent incorporations of a “terminally
added” benthic phase into an ancestral holopelagic
life cycle. Moreover, the reproductive capacity of the
holopelagic ancestor must have been transferred in
each case into the new benthic phase, because this is
the sexually reproductive stage in all extant metazoans
that exhibit a pelagobenthic life cycle. Second, they
require that the LCA to extant metazoans took the
form of a small and simple, blastula-like, hollow ball
of cells, and that the ontogeny of this ancestral Blastaea
recapitulated metazoan evolution and phylogeny.
Third, they require that the embryonic process of gastrulation, which is observed in all extant metazoans
(Leys 2004; Martindale 2005), represents a recapitulation of the origin of the gastraea as a response to selection
on a hollow balls of cells to develop the capacity to feed.
As exciting new data are now accumulating, hypotheses on the origin and evolution of biphasy can finally
begin to take greater account of basal metazoans. Here,
we present data from early branching (basal) metazoans, in particular sponges, that we believe shed significant light on this central issue of animal evolution.
Although the phylogenetic relationships of the basal
metazoans are still poorly resolved (see Fig. 1 for current
alternative hypotheses), there is now general consensus
that the Porifera diverged first from the rest of the
Metazoa and is the most ancient of the extant animal
phyla (Fig. 1A). Some recent analyses of 18S rRNA and
protein coding genes suggest (albeit with rather weak
support) that sponges may be paraphyletic, with the
calcareous sponges (Calcarea) more closely related to
the Eumetazoa (Cnidaria, Ctenophora, and Bilateria)
than to the other sponge classes (Collins 1998; Zrzavy
and others 1998; Borchiellini and others 2001; Medina
and others 2001). In this scenario, siliceous sponges
(demosponges and hexactinellids) comprise the most
Origin of pelagobenthic life cycle
ancient lineage (Fig. 1B). In terms of considerations of
the origin of metazoan biphasy, the basal position of
the Porifera places the interest squarely on sponges.
Indeed, the possible paraphyly of Porifera relative
to the Eumetazoa raises the provocative idea that the
ancestor of all metazoans may have been a sponge-like
animal.
The demosponge Amphimedon
and reconstruction of the LCA to
extant Metazoa
Amphimedon queenslandica (Hooper and Van Soest
2006; Porifera, Demospongiae, Haplosclerida,
Haplosclerina, Niphatidae; formally referred to as
Reniera sp.) is a demosponge and thus a representative
of the most ancient animal lineage. Its biphasic pelagobenthic life cycle (Fig. 2) includes a ciliated parenchymella larva that is planula-like in form. Like most
demosponges, Amphimedon has internal fertilization
and broods its embryos in discrete brood chambers
(Leys and Degnan 2001, 2002). The swimming
Amphimedon larva (Fig. 3) has a photosensory system
at its posterior end (relative to swimming direction)
that consists of a set of ciliated pigment cells organized
into a ring (Fig. 3B and C). These cells respond to light
685
in a stereotypic manner, such that the resultant effect is
a rudder-like structure that directs movement of the
larvae away from areas of higher light (Leys and
Degnan 2001; Leys and others 2002). The surface of
Amphimedon larvae also appears to have both chemosensory and mechanosensory cells that contribute to
the detection of an appropriate location for settlement
(Fig. 3D–F; Jackson and others 2002). Together, these
sensory systems indicate that the Amphimedon larva is
more complex than would be predicted by current
theories of the origin of metazoan biphasy.
The Amphimedon larva consists of a large number (at
least 11) of different cell types that are organized into
3 discrete layers—an outer epithelial layer, a middle cell
layer, and an inner cell mass (Fig. 3A; Leys and Degnan
2001, 2002). Some of these cell types, of which the
pigment ring cells are the most conspicuous, are patterned along an anterior–posterior axis relative to the
swimming direction. Importantly, detailed analyses of
Amphimedon embryology and larval formation are
demonstrating beyond doubt that sponge development
includes many of the hallmarks of eumetazoan embryogenesis (Leys and Degnan 2002; Degnan and others
2005; Larroux and others 2006). Of particular significance is the fact that fertilization is followed by a period
of cell division that yields distinct cell populations that
Fig. 2 The life cycle of the demosponge Amphimedon queenslandica (A) Underwater photograph, (B and F)
photomicrographs, (C–E) scanning electron micrographs. (A) Adult Amphimedon inhabit pieces of coral rubble on the
reef flat of Indo-Pacific coral reefs. (B) This sponge broods embryos and larvae at all times. The brood chambers are
localized to the basal region of the adult and contain 50–150 embryos that develop asynchronously. Embryos possessing
a pigment ring are older. (C) The parenchymella larva with anterior to the right. The dark pigment ring is located at the
posterior end and consists of cells possessing a long cilium. (D) Settlement with larval anterior end attached to the
substratum. The long posterior cilia (LPC) seen facing upwards. (E) The first phase of metamorphosis is characterized by
the spreading of the body over the substratum, ingression of the pigment ring cells and transdifferentiation of other
epithelial cells. (F) Three days after induction, the first osculum is formed.
686
S. M. Degnan and B. M. Degnan
Fig. 3 Sensory cells and putative sensory cells of Amphimedon. (A) Transmission electron micrograph of the Amphimedon
larva showing 3 cell layers: the inner cell mass (ICM), middle layer consisting of subepithelial cells (SEC), and the outer
layer consisting predominantly of columnar epithelial cells (CEC). At the posterior pole (PP) is a pigment ring (PRg),
from which long posterior cilia (LPC) emanate, and spicules (sp) in the inner cell mass. The outer layer is entirely
ciliated except at posterior and anterior (AP) poles. (B) Scanning electron micrograph of the PP showing the LPC
associated with the posterior pigment cells. Arrow points to a mucous cell; these cells populate the surface inside the
ring. (C) The LPC on the posterior portion of a bisected larva respond to abrupt increases and decreases in light
intensity, being bent over the pigment ring with decreased light intensity and straightened (arrowhead) with increased
light. (D) Flask-shaped epithelial cells that populate the surface of the Amphimedon larva are differentially localized to the
anterior end. These cells possess a large centrally located nucleus (n), a cilium that arises from a deep invagination in
the cell (arrowheads) and a concentration of basal vesicles, hallmarks of a sensory cell. Bar ¼ 2 mm. (E) A scanning
electron micrograph of the bare, anterior pole of an Amphimedon larva. At settlement, this region contacts and attaches
to the substratum. Short lateral cilia (LC) are seen on the surface. (F) A transmission electron micrograph of a region
near the edge of the anterior pole. The short, LC mark the end of the columnar epithelial cells. The anterior-most cells
are cuboid and generally not ciliated. Mucous cells (mc) and mitochondria (m) are evident (see Leys and Degnan 2001
for further details).
are allocated through a gastrulation process (Leys
2004) into different cell layers, and even patterned
within these layers along an anterior–posterior axis.
Further, gastrulation can lead to the formation of simple tissue-like structures such as the pigment ring (Leys
and Degnan 2002). Remarkable and unexpected similarities between the development of this demosponge
and that of eumetazoans leads us to infer that the
metazoan LCA was likely to have required asymmetric
cell division, organizers, morphogen gradients, and
populations of cells with differing competencies to
construct its body plan (Degnan and others 2005).
In addition to this suite of conserved cellular behaviors, a range of genes previously shown to be involved
in eumetazoan development are expressed during
Amphimedon embryogenesis and larval development
(Larroux and others 2006). The Amphimedon genome
in fact contains the same basic repertoire of developmental gene families as found in higher metazoans.
These are expressed during embryogenesis and metamorphosis, suggesting that the regulatory architecture
underlying eumetazoan development already existed
in the metazoan LCA (Larroux and others 2006).
Basal metazoan data are at odds
with the Trochaea theory
The ability of sponge cells to differentiate, migrate, and
pattern during gastrulation in a manner akin to that in
all other metazoans has important implications for the
Trochaea theory (Nielsen 1985; Nielsen and Norrevang
1985). Central to this hypothesis is the phylogenetic
placement of the first gastrula-like form at the LCA
of cnidarians and bilaterians, after sponges had split
from the Eumetazoa (Nielsen 1985, 2000, 2001, 2005).
Contrary to this, our analysis of sponge development
leads us to infer that the LCA to all extant metazoans already had the capacity to form a multilayered
embryo through the process of gastrulation (Leys and
Degnan 2002; Leys 2004; Degnan and others 2005).
Furthermore, sponge larvae that develop through
the “gastrea stage” do not have any capacity to feed;
there is no development of a gut in either sponge larvae
or indeed in sponge adults. There simply is no evidence
for an association between gastrulation and development of a gut, and no indication of phylogeny recapitulating ontogeny in this sense.
Origin of pelagobenthic life cycle
In general, sponge embryogenesis appears to be
directed by the same developmental gene families as
those of eumetazoans. Indeed, the LCA of the Metazoa
appears to have had all of the features of body plan
currently shared by extant eumetazoans and sponges,
namely tissue formation via gastrulation, sensory
systems, and a metazoan developmental regulatory network (Fig. 4). Together, these features clearly demonstrate that the LCA was of a vastly more complex
phenotype than widely appreciated, and that there is
a shared developmental ancestry between the Porifera
and the Eumetazoa (Fig. 4). The presence of this complexity and shared developmental ancestry preclude the
notion of a simple, hollow ball of cells as a larval-like
ancestor in which ontogeny recapitulates phylogeny. As
such, there is no support for this aspect of the Trochaea
theory.
The intercalation theory meets the
basal metazoans
If the Trochaea theory is currently the most generally
adopted explanation for the origin of metazoan
biphasy, and if our new insights from sponges are at
odds with this theory, where should we turn to seek
alternative hypotheses? We need look no further than
another feature clearly present in the LCA to all
Metazoa—the existence of sexual reproduction via
687
meiotically derived eggs and sperm. Although there
is no dispute that the origin of sexual reproduction
predates metazoan cladogenesis, and that meiosis
may be as ancient as the eukaryote, most current discussions on the origin and evolution of biphasy surprisingly do not explicitly incorporate these facts. The
fertilized egg is the first step in embryogenesis which, in
all multicellular animals, includes a process of gastrulation that generates a multilayered embryo (Leys 2004;
Martindale 2005). At some point in its development,
the embryo must separate from the adult, and this
separation can itself be considered as the precursor
to a biphasic lifestyle, wherein the embryo represents
one phase and the adult another phase. Under this
scenario, it is most likely that the first metazoan life
cycles were direct—that is, they had no larval stage that
required a metamorphosis back to the adult state and
thus no larval stage that required a capacity to feed
(Fig. 5A). It is also likely, although not essential to
this hypothesis, that the ancestral adult form was
benthic. This precludes the need to hypothesize multiple, independent origins for the “terminally added”
sexual benthic form, because the LCA already exhibited
a precursory pelagobenthic life cycle.
As soon as embryos became independent from the
adult phase, selection would have acted differentially
on size and habitat differences that exist between these
2 phases. The multilayered nature of the first metazoan
Fig. 4 Inferred characteristics of the last common ancestor to living metazoans based on characters that are shared
between sponges and eumetazoans (sponges are monophyletic in the depicted phylogeny). These metazoan-specific
characteristics are likely to have evolved after the metazoan and choanoflagellate lineages split, although some
components of the metazoan regulatory network may predate this split. The ancestor of all extant animals possessed
most, if not nearly all, gene families used during the course of metazoan development. Most of these developmental gene
families (encoding transcription factors and components of signaling pathways) do not have clear orthologs in fungal,
plant, or protist genomes. The origin of metazoan sexual reproduction, which includes the production of haploid sex
cells, specifically large eggs that could be fertilized by motile sperm, predates metazoan cladogenesis. The developmental
program of the new individual runs in a stereotypic manner, displaying a suite of predictable cell behaviors during the
course of development. Underlying these behaviors is the localized expression of developmental genes whose
expression is contingent upon inputs from the regulatory network. The sorting of cells into 2 or more layers allows for a
further division of function, with each layer acting in a semiautonomous manner. This is the one of the first steps in the
evolution of modularity in the metazoan body plan, and enables the evolution of the biphasic life cycle.
688
S. M. Degnan and B. M. Degnan
Fig. 5 Evolution of the biphasic life cycle through the intercalation of novelties into a life cycle without a larval stage.
(A) Early metazoans (prior to the sponge-eumetazoan split) evolved haploid sex-specific cells that could come together
at fertilization to create a new individual. Early development was characterized by the cleaving of the zygotic cytoplasm
into smaller cellular units. Through asymmetric cell division arose populations of cells with differing competencies and
proclivities. These were sorted into multiple layers via the process of gastrulation and could be localized into simple
tissue-like structures. Included amongst these is a cell territory that gives rise to gametes. At some point in the
evolution of this direct-developing life cycle, the new individual was separated from the parent. In the scenario depicted,
the adult is benthic and the embryo pelagic/planktonic. (B) The multilayered individual is inherently modular in design,
enabling evolvability. If embryos and adults inhabit different environments, such as the water column and benthos as
depicted here, they will be subjected to different selective forces. As development proceeds modularity can increase,
enabling the further intercalation of novelty into the life cycle. Through these developmental mechanisms evolved the
biphasic life cycle with an early life cycle stage adapted to life in the water column and to dispersal, and later stages
adapted for growth and reproduction.
embryos would have enabled elaboration of different
layers of cells, which in turn would have facilitated
intercalation of novel morphologies and functions
into a direct life cycle. This is because the multiple
cell layers enable greater modularity in body plan,
which then would facilitate evolvability (Gerhart
and Kirshner 1997) and thus would permit a
greater range of morphological diversity and further
evolution of dispersal and of phase/habitat-specific
inventions (Fig. 5B). It is therefore reasonable to
consider that embryos ultimately released into the
water column from their putatively benthic parent
would have been subject to natural selection for
dispersal/longevity/feeding (which reduces susceptibility to local extinction) that would have set them on the
evolutionary trajectory towards the crown metazoan
planktonic larvae, as already discussed in an extensive
and excellent body of literature on the evolution of
larval forms (Wolpert 1999; McEdward 2000; Rouse
2000; Strathmann 2000; Collins and Valentine 2001;
Bishop and Brandhorst 2003; Sly and others 2003).
Each elaboration, in turn, would allow new modules
to evolve, which could be incorporated into either the
later larval or adult phase and further increase bauplan
complexity.
These later novelties would have included the localized formation of larval sensory systems that allow for
selection of a suitable habitat for settlement. As seen in
extant sponges, such a sensory system does not require
integration via neurons (Leys and Degnan 2001;
Jackson and others 2002). Natural selection on these
systems would be particularly strong, as any individual
who settles in unsuitable habitat and/or away from congeners is not likely to contribute to future generations.
689
Origin of pelagobenthic life cycle
The existence of sophisticated sensory systems in all
extant metazoan larvae is compatible with the metazoan
LCA having a pelagobenthic life cycle with a sensitive
larva. The wide range of extant adult phenotypes
reflects the increase in evolvability that is inherent in
this second phase of the life cycle. Of course, the modular
design of the metazoan body plan enables the heterochronic shifting of embryonic, larval, juvenile, and
adult developmental programs relative to each other,
and to metamorphosis, in an enormous variety of
ways, as displayed by numerous extant taxa (for
example, Wray and Lowe 2000; Sly and others 2003;
Byrne 2006).
Conclusion
In summary, our sponge data permit discussions on the
origin of biphasy of the most basal metazoans, beyond
that previously possible. These data clearly dismantle
the proposal that the LCA was a simple, hollow ball of
cells. In contrast, the LCA appears to have been morphogenetically more complex than widely appreciated,
employing a developmental regulatory repertoire similar to that used by all extant animals (Fig. 4). The LCA
had a complex, multilayered embryo, which formed
through an array of entrained intercellular interactions
that were encoded in the genome and activated during
embryogenesis. Furthermore, there is no evidence that
the formation of these cell layers during gastrulation
bears any relationship to the origin of feeding. As
such, this new view precludes the need for ontogeny
to recapitulate phylogeny.
We conclude then that gastrulation is more profitably considered as an ancestral embryonic stage that is a
direct consequence of sexual reproduction. The modular design inherent in the multilayered gastrula
enables evolvability from that stage onwards. We suggest that natural selection on the embryo would have
acted in the first instance to select for a capacity for
dispersal, rather than for feeding. Our data strongly
support the intercalation hypotheses discussed previously in the context of bilaterian larval origins and
evolution (for example, Wolpert 1999; Valentine and
Collins 2000; Sly and others 2003) but not the gastraea/
trochaea hypotheses (nor then the set-aside cell
hypothesis of Davidson and others 1995; Peterson
and others 1997, 2000) of the terminal addition school.
Further, we suggest that the ancestral metazoan adult
was more likely to have been benthic rather than pelagic, with direct rather than indirect development. We
hope that these postulations provide a new starting
point for consideration of the evolution of metazoan
larval forms and embryonic development.
Acknowledgments
We thank Sally Leys and Claire Larroux for providing
images shown in this article. We are grateful to the
Society for Integrative and Comparative Biology
(SICB) for promoting and partially funding the
symposium in which these ideas were first presented.
We are very grateful to Andreas Heyland, Jason
Hodin, Corey Bishop, and Leonid Moroz for organizing
this symposium and for inviting us to contribute, and
to all audience-members from the platform and
associated-sessions for constructive discussions.
Furthermore, we would like to thank the following
organizations for their generous financial support:
the University of Florida, The Whitney Laboratory
for Marine Biosciences, the American Microscopical
Society (AMS), and the SICB Division of
Evolutionary Developmental Biology (DEDB).
Research into the biology of Amphimedon sp. 2456 (formerly known as Reniera sp) has been supported by
grants from the Australian Research Council and the
United States Department of Energy Joint Genome
Institute to B.M.D.
Conflict of interest: None declared.
References
Bishop CD, Brandhorst BP. 2003. On nitric oxide signaling,
metamorphosis, and the evolution of biphasic life cycles.
Evol Dev 5:542–50.
Borchiellini C, Manuel M, Alivon E, Boury-Esnault N, Vacelet J,
LeParco Y. 2001. Sponge paraphyly and the origin of Metazoa.
Evol Biol 14:171–9.
Byrne M. 2006. Life history diversity and evolution in the
Asterinidae. Integr Comp Biol 46:243–54.
Collins AG. 1998. Evaluating multiple alternative hypotheses for
the origin of the Bilateria: an analysis of 18S rRNA molecular
evidence. Proc Natl Acad Sci USA 95:15458–63.
Collins AG, Valentine JW. 2001. Defining phyla: evolutionary
pathways to metazoan body plans. Evol Dev 3:432–42.
Davidson EH, Peterson KJ, Cameron RA. 1995. Origin of bilaterian body plans: evolution of developmental regulatory
mechanisms. Science 270:1319–25.
Degnan BM, Leys SP, Larroux C. 2005. Sponge development and
antiquity of animal pattern formation. Integr Comp Biol
45:335–41.
Gerhart J, Kirshner M. 1997. Cells, embryos, and evolution.
Malden, MA: Blackwell Science.
Glenner H, Hansen AJ, Sorensen MV, Ronquist F,
Huelsenbeck JP, Willerslev E. 2004. Bayesian inference of
the metazoan phylogeny: a combined molecular and morphological approach. Curr Biol 14:1644–9.
Haeckel E. 1874. Die Gastrea-Theorie, die Phylogenetische
Clasification des Thierreichs und die Homologie der
Keimblatter. Jena Z Naturw 8:1–55.
690
Jackson D, Leys SP, Hinman VF, Woods R, Lavin MF,
Degnan BM. 2002. Ecological regulation of development:
induction of marine invertebrate metamorphosis. Int J Dev
Biol 46:679–86.
Jägersten G. 1955. On the early phylogeny of the Metazoa: the
bilaterogastraea theory. Zool Bidr Uppsala 30:321–54.
Jägersten G. 1972. Evolution of the metazoan life cycle: a comprehensive theory. New York: Academic Press. p 1–282.
Larroux C, Fahey B, Liubicich D, Hinman VF, Gauthier M,
Gongora M, Wörheide G, Leys SP, Degnan BM. 2006.
Developmental expression of transcription factor genes in a
demosponge: insights into the origin of metazoan multicellularity. Evol Dev 8:150–73.
Leys SP. 2004. Gastrulation in sponges. In: Stern C, editor.
Gastrulation: from cells to embryo. Cold Spring Harbor:
Cold Spring Harbor Press. p 23–31.
Leys SP, Cronin TW, Degnan BM, Marshall JN. 2002. Spectral
sensitivity in a sponge larva. J Comp Physiol A 188:199–202.
Leys SP, Degnan BM. 2001. Cytological basis of photoresponsive
behaviour in a sponge larva. Biol Bull 201:323–38.
Leys SP, Degnan BM. 2002. Embryogenesis and metamorphosis
in a haplosclerid demosponge: gastrulation and transdifferentiation of larval ciliated cells to choanocytes. Invert Biol
121:171–89.
Martindale MQ. 2005. The evolution of metazoan axial properties. Nat Rev Genet 6:917–27.
McEdward LR. 2000. Adaptive evolution of larvae and life cycles.
Semin Cell Dev Biol 11:403–9.
Medina M, Collins AG, Silberman J, Sogin ML. 2001. Evaluating
hypotheses of basal animal phylogeny using complete
sequences of large and small subunit rRNA. Proc Natl Acad
Sci USA 98:9707–12.
Nielsen C. 1979. Larval ciliary bands and metazoan phylogeny.
Fortschr Zool Syst Evolutionsforsch 1:178–84.
Nielsen C. 1985. Animal phylogeny in the light of the trochea
theory. Biol J Linn Soc 25:243–99.
S. M. Degnan and B. M. Degnan
Nielsen C. 2000. The origin of metamorphosis. Evol Dev
2:127–9.
Nielsen C. 2001. Animal evolution: interrelationships of the
living phyla. 2nd edition. Oxford: Oxford University Press.
Nielsen C. 2003. Defining phyla: morphological and molecular
clues to metazoan evolution. Evol Dev 5:386–93.
Nielsen C. 2005. Larval and adult brains. Evol Dev 7:483–9.
Nielsen C, Norrevang A. 1985. The trocheae theory: an example
of life cycle phylogeny. In: The origins and relationships of
lower invertebrates. Oxford: Oxford University Press. p 28–41.
Pechenik J. 2004. Biology of the invertebrates. 5th edition.
New York: McGraw-Hill.
Peterson KJ, Cameron RA, Davidson EH. 1997. Set-aside cells in
maximal indirect development: evolutionary and developmental significance. Bioessays 19:623–31.
Peterson KJ, Cameron RA, Davidson EH. 2000. Bilaterian origins: significance of new experimental observations. Dev Biol
219:1–17.
Rokas A, Krüger D, Carroll S. 2005. Animal evolution and the
molecular signature of radiations compressed in time. Science
310:1933–8.
Rouse GW. 2000. The epitome of hand waving? Larval feeding
and hypotheses of metazoan phylogeny. Evol Dev 2:222–33.
Sly BJ, Snoke SS, Raff RA. 2003. Who came first—larvae or
adults? Origins of bilaterian metazoan larvae. Int J Dev
Biol 47:623–32.
Strathmann RR. 2000. Functional design in the evolution of
embryos and larvae. Semin Cell Dev Biol 11:395–402.
Valentine JW, Collins AG. 2000. The significance of moulting in
Ecdysozoan evolution. Evol Dev 2:152–6.
Wolpert L. 1999. From egg to adult to larva. Evol Dev 1:3–4.
Wray GA, Lowe CJ. 2000. Developmental regulatory genes and
echinoderm evolution. Syst Biol 49:28–51.
Zrzavy J, Mihulka S, Kepka P, Bezdek A. 1998. Phylogeny of the
Metazoa based on morphological and 18S ribosomal DNA
evidence. Cladistics 14:249–85.