Download as a PDF

yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Life wikipedia , lookup

Cell culture wikipedia , lookup

Embryonic stem cell wikipedia , lookup

Somatic evolution in cancer wikipedia , lookup

Evolutionary history of life wikipedia , lookup

Koinophilia wikipedia , lookup

Human embryogenesis wikipedia , lookup

Cellular differentiation wikipedia , lookup

Dictyostelium discoideum wikipedia , lookup

Cell (biology) wikipedia , lookup

Somatic cell nuclear transfer wikipedia , lookup

Adoptive cell transfer wikipedia , lookup

Saltation (biology) wikipedia , lookup

Organ-on-a-chip wikipedia , lookup

Cochliomyia wikipedia , lookup

Chimera (genetics) wikipedia , lookup

Cell theory wikipedia , lookup

Biology wikipedia , lookup

Marine larval ecology wikipedia , lookup

Amitosis wikipedia , lookup

Microbial cooperation wikipedia , lookup

State switching wikipedia , lookup

Developmental biology wikipedia , lookup

seminars in CELL & DEVELOPMENTAL BIOLOGY, Vol. 11, 2000: pp. 395–402
doi: 10.1006/scdb.2000.0192, available online at on
Functional design in the evolution of embryos and larvae
Richard R. Strathmann
contain most or all of the cell types of an adult.
Animals in at least eight phyla divide longitudinally
or transversely or separate from marginal fragments,
with the pieces regenerating missing parts.2–4 Colonial animals, such as sponges, cnidarians, bryozoans,
pterobranchs, and ascidians, produce multicellular
buds with initially few cell types but many cells, and
many also divide into large fragments. Starting life
with more cells can result in more rapid development. Starting with numerous differentiated cells
confers capacity for diverse functions. Why should
multicellular organisms, with all the advantages
of numerous and differentiated cells, so regularly
reduce themselves to a single cell?
Suggested hypotheses fall into two overlapping
categories. (1) A unicellular stage exposes offspring
that are each uniformly of one genotype to selection,
thereby purging deleterious mutations (and more
rarely fixing beneficial ones). (2) A unicellular
stage reduces conflicts of interest among genetically
different replicators within an organism.5–10
For the first hypothesis, Kondrashov10 modelled
mutation load with asexual propagation at the
equilibrium between mutation rate and selection.
If all cells are recent descendants of one cell, the
mutation load does not increase with the number of
cells in the propagule, but if distantly related initiating cells form the propagule, then the mutation
load approaches one as the number of cells in the
propagule increases. In multicellular animals, cells
from lineages separated by a large number of mitotic
divisions can be included in one propagule. Thus the
equilibrium mutation load is expected to be larger
for animals with multicellular propagules than for
plants, whose cells grow and change shape but do not
move to distant parts of the body.
Conflicts among genetically distinct replicators
are of varied sorts. Some conflicts arise from mutation, as in cancer. By inclusion in multicellular
propagules, cancers could persist through several
generations. Pathogens are also genetically distinct
replicators whose interests conflict with their host’s.
A function of development is to put the right kind of cells in
the right place at the right time. Other functional analyses
help define what is right. As examples, functional analyses
offer explanations for the unicellular bottleneck in life
histories that necessitates embryos, evolutionary divergences
in embryonic cell cycles, conditions permissive of loss of
larval structures and consequent change in embryonic
development, and the decoupled development of larval bodies
and juvenile rudiments. Functional analyses also reveal
the specifications required of morphogenesis, hence defining
developmental phenomena to be explained.
Key words: ascidian / cell cycle / echinoid / plasticity /
c 2000 Academic Press
Most studies of the evolution of development describe change or stasis in developmental processes.
The studies seldom explain why a change occurred
or why stasis has been so prolonged. Analysis of the
functional requirements for embryos and larvae is a
necessary part of such an explanation. Such analyses
are rare in developmental biology.1 The following
examples illustrate several ways that functional requirements shape development. In these examples,
it should become apparent that developmental
processes have also shaped functional requirements.
Why are there embryos?
Developmental biologists ask how embryos develop.
Life history theorists ask why embryos exist. Many
organisms produce multicellular propagules that
From the Friday Harbor Laboratories and the Department of Zoology,
University of Washington, 620 University Road, Friday Harbor, WA
98250, USA. E-mail: [email protected]
Academic Press
1084–9521 / 00 / 060395+ 08 / $35.00/0
R. R. Strathmann
in the right types of cells in the right numbers in the
right places in the organism. Another functional requirement is that multicellular function be restored
rapidly. The urgency depends on the vulnerability of
the embryo. Planktonic embryos receive less protection than brooded or encapsulated embryos. Comparisons of mortality rates of planktonic larvae and protected aggregations of embryos indicate higher mortality rates for planktonic larvae.16
For embryos at risk from predators or other external causes of mortality, one would expect selection
for rapid development. Rapid development reduces
the period of vulnerability. Selection for rapid development includes selection for rapid cell cycles in embryos. Selection should decrease cell cycle duration
until the costs of an even shorter cell cycle balance
the advantages of a further reduction in the period
of vulnerability. Differences in rates of early development could arise from trade-offs between risks from
external sources of mortality and the costs associated
with fast cell cycles.
Jennifer Staver and I tested the hypothesis that
early embryonic cell cycles are more rapid for
planktonic embryos than for embryos in protected
aggregations (unpublished observations). In six out
of eight independent evolutionary divergences in
protection, the time from first to second cleavage was
shorter for the planktonic embryos. The evolutionary
divergences in protection included examples from
five phyla (molluscs, phoronids, brachiopods, echinoderms, ascidians). In most of the paired comparisons,
the protected embryos had longer cell cycles. In no
paired comparison did the protected embryos have
a shorter cell cycle. Other comparisons from other
phyla confirm this pattern. In the familiar mouse
and frog, the interval from first to second cleavage is
longer in the more protected embryos of the mammal than in those of the amphibian. The differences
in cell cycle duration in these comparisons is not the
result of temperature, nor are the more slowly developing protected embryos consistently from larger
eggs. In some cases, as in mouse versus frog, the
protected embryos are smaller, warmer, and slower.
To my knowledge, the trade-offs governing the
evolution of cell cycle durations are unknown. Hypotheses include (but are not limited to) costs with
short cell cycles in less protection against molecular
damage or less effective repair, greater quantities of
rate-limiting materials that are maternally supplied
to eggs, or less early transcription.
Differences in risk to embryos and the differing
environments for protected and planktonic embryos
Unicellular propagules restrict opportunities for
vertical transmission, as compared to multicellular
propagules, especially when embryos are not closely
associated with a parent. Fusion chimaeras occur
in several phyla of animals, including the sponges,
cnidarians, bryozoans, and even chordates.11 Fusion
chimaeras open the possibility of competition among
cells for inclusion in the germ line, with cells of some
genotypes shirking their somatic responsibilities
but contributing disproportionately to the next
generation.12 Such differential reproductive success
of cell lineages in chimaeras has been demonstrated
for ascidians.13 A unicellular stage in the life history
is one of the ways of reducing parasitism of germ line
cells of one genotype on the soma of another.7
The disadvantages of multicellular propagules apply to asexual multicellular propagules, but are especially evident for sexual reproduction. Processes of recombination produce genetically distinct cells. Combining numbers of these cells into a multicellular organism would result in a highly chimaeric offspring.
Syngamy that involved multiple pairs of cells and that
followed meiosis or involved multiple partners in a
mating would also produce genetic diversity within
offspring. Although the disadvantages of multicellular propagules are acute and more obvious for sexual reproduction, a unicellular stage nevertheless persists with amictic parthenogenesis. Asexual production of embryos has even persisted in animals that can
also produce larger multicellular propagules by fission or fragmentation,14, 15 although the number and
lineages of parental cells in these embryos remains to
be demonstrated.
The relative importance of pathogens or deleterious mutations as agents of selection for persistence of
unicellular stages has not been demonstrated. Fusion
chimaeras should be important where they occur,
but they appear to be rare in non-colonial animals.
Analyses of the costs and benefits of multicellular and
unicellular propagules depend on combined studies
of functional morphology, population genetics, and
developmental processes. Key phenomena that affect
the costs and benefits of unicellular and multicellular
propagules include cell–cell recognition, defenses
against pathogens, effectiveness of apoptosis in
eliminating mutant cells, and cell motility.
Evolution of embryonic cell cycles
A function of an embryo is to restore a multicellular
organism that gains the advantages of differentiated
cells, specialized for different functions. Studies of development have emphasized the processes that result
Function of development
Evolutionary loss of tails of ascidian tadpoles
contrasts with the conservatism of the vertebrate
‘pharyngula’ stage. Similarities among vertebrate
pharyngulas was an inspiration for von Baer’s and
Haeckel’s laws on evolutionary conservatism of
early stages in development. The conservatism of the
pharyngula has been questioned20 and HeLa cells (of
human origin) have been described as new species
(without vertebrate anatomy),21 but it is nevertheless
true that all animals classified as vertebrates retain a
notochord, dorsal nerve cord, and axial body muscles
at some stage in their development. How is it that
ascidian species have ‘unchordated’ in their evolution while vertebrates have not, despite the immense
diversification of vertebrate species? Differences in
function of these structures are an essential part of
the explanation, although developmental differences
enter as well.
It is instructive to compare the tailless ascidians
to the floating head and no tail mutants in zebra
fish. These unfortunate zebra fish demonstate that
mutations interfering with differentiation of the
notochord are possible, with consequent drastic
reduction of dorsal nerve cord and axial muscles.22
Mutations that go part way in removing vertebrate
traits occur, but they are lethal for functional reasons.
These mutant fish demonstrate that the conservatism
in the vertebrate body plan is not purely developmental. It is the result of functional constraints associated
with vertebrate ways of life.
This example illustrates the importance of functional burden in evolutionary stasis and change in
development. The functional burden on notochorddependent structures in the ascidian is low because
these structures appear only in the tadpole stage.
These structures are lost at metamorphosis, and
their absence in the tadpole makes little difference
to the postmetamorphic juvenile. It makes little
difference because the tadpole does not feed or grow
(a functional reason) and because development of
the pharynx does not depend on the notochord
(a developmental reason). The functional burden
on notochord-dependent structures is high in vertebrates because these structures are necessary for
feeding and escape at nearly all postembryonic
Burden can increase or decrease during evolution.23 If the ancestral chordate was like an
ascidian,24 then the functional burden on structures dependent on the notochord has increased
in the vertebrates. If the ancestral chordate was
like a larvacean or cephalochordate (requiring the
may combine to affect synchrony or asynchrony
in cleavages. Van den Biggelaar and Haszprunar17
compared the number of cells in gastropod embryos
at the time when the mesentoblast is formed. Snails
with planktonic embryos form the mesentoblast when
there are 63 cells. Although snails with encapsulated
embryos form the mesentoblast when there are fewer
cells, their formation of the mesentoblast is not
accelerated. The authors do not give developmental
schedules, but the snails with protected embryos
have longer embryonic cell cycles than the snails
with planktonic embryos. In protected embryos, the
mesentoblast is formed when there are fewer cells
because the lineages of animal pole micromeres
(fated to become ectoderm) are retarded even
more than the vegetal pole macromeres (one of
which eventually produces the mesentoblast). The
overall retardation has occurred in association with
protection. A functional hypothesis for the greater
retardation of the presumptive ectoderm is that the
encapsulated embryos have no immediate need for
ciliated bands for swimming or apical sense organs
(which are ectodermal), whereas the planktonic
embryos begin swimming much earlier in their
development, using locomotory cilia at an early stage.
Burden and loss of function: transitions in
modes of larval development
Evolutionary loss of function is often accompanied by
reduction or loss of structures. The first structures
formed in larvae are those used in swimming or
feeding. It is therefore no surprise that the loss or
reduction of these structures is often accompanied by
changes in cell fates and morphogenetic processes.
Loss of the larval tail in ascidians is a spectacular
example that involves a profound change in the
chordate body.18 Tailless ascidians have lost the
notochord, dorsal nerve cord, and axial body muscles
characteristic of chordates. Tails have been lost multiple times and such losses have been concentrated
within the family Molgulidae.19 Thus molgulids have
‘unchordated’ several times in their evolution. Retention of pharyngeal openings hardly qualifies them
as chordates. The hemichordates have pharyngeal
openings but lack of other chordate features put
them in another phylum. Tailless ascidians are still
classified as chordates. They have not even changed
genus. The taxonomic conservatism is due to respect
for ancestors and perhaps due to adults being more
conspicuous than larvae.
R. R. Strathmann
between two different cell lineages.28 The ciliary
band forms a continuous loop that encloses the oral
field and mouth. The form of this ciliary band is a
compromise between requirements of feeding and
swimming.29, 30 Feeding requires an effective stroke
of cilia perpendicular to the band and away from
the oral field. This brings food particles toward the
ciliary band, and the particles induce local reversals of
ciliary beat that retain the food on the oral field and
aid transport toward the mouth. Swimming requires
that the direction of the effective stroke of cilia have
a component toward the posterior so that the larva
moves forward.
In sea urchin larvae, this requirement is met
by loops of the band on arms that are directed
anteriorly with a slight outward tilt and by transverse
sections of the band between the arms. The larval
arms are supported by skeletal rods. Interactions of
mesenchyme and ectoderm are responsible for the
form and position of these skeletal rods. The form
of the pluteus arms and ciliary bands depends in part
on the skeleton.
In contrast, larval seastars and sea cucumbers (some
other echinoderms) have no skeletal rods or anteriorly directed arms. Instead the loops of the band
are everywhere slightly tilted so that there is a posterior component to the ciliary currents. These tilts of
the band are easily overlooked but are necessary for
its function and require a more precise positioning
of cells than one would first suppose.1 The morphogenetic processes that produce these subtle tilts of the
bands of seastar and sea cucumber larvae are as yet
Evolution of larger eggs is associated with loss
of the necessity for feeding,31 and eventually the
larval ciliation evolves into transverse bands or ciliary
fields as arrangements more suited for the sole
requirement of swimming.32 Extreme reduction
of the ciliary band has occurred in the sea urchin
Heliocidaris erythrogramma (with a non-feeding larva)
which has diverged from its sister species Heliocidaris
tuberculata (which has retained a feeding larva).
H. erythrogramma provides an extensively studied
example of changes in cell fates and morphogenetic
processes associated with loss of function.33, 34 Maternal H. e. × paternal H. t. hybridization is possible and
offers an opportunity to explore the genetic basis of
profound evolutionary changes in morphogenesis.33
A remarkable result of the maternal H. e. × H. t.
cross is restoration of a continuously looping ciliary
band, but the skeletal rods do not develop normally
in the hybrid.33 The hybrid’s superficial resemblance
notochord and notochord-dependent structures for
feeding and growth), then the functional burden on
structures dependent on the notochord decreased in
the ascidians when they evolved a sessile habit during
the feeding stages, with consequent loss of the tail at
metamorphosis. If a population of fishes had evolved
a sessile habit, so that at an early stage they became
attached to the bottom and fed with a pharyngeal
filter, then the functional burden on notochorddependent structures would have decreased for these
vertebrates, and we might see viable fish in which
a mutation for taillessness had spread to become
characteristic of the species.
Loss of tails also illustrates a deficiency in many
claims of developmental constraint. In cases where
mutations producing the phenotype occur but the
phenotype is disadvantageous, there is a functional
constraint. Developmental processes might or might
not be combining with such a functional constraint
to restrict evolutionary change, but the environment
affects what enhances survival and reproduction and
what does not. With appropriate culture conditions,
vertebrates evolved into protists, with HeLa cells an
especially successful example.25
Before leaving tailless ascidians, it is worth noting
the multiple evolutionary loss of tails in molgulids
and very rare loss in other families of ascidians.
Berrill suggested a functional hypothesis: taillessness
evolved frequently in molgulids because they live
in sandy habitats where larval habitat selection has
little advantage. This is inconsistent, however, with
a tailless molgulid on a rocky shore26 and with
selective settlement by swimming larvae of other
animals in sandy habitats. Sandy habitats are by
no means uniform for settling larvae.27 The traits
that have facilitated tail loss in molgulids are as yet
Setting specifications for a developmental
Functional analyses show the degree of developmental specification that is required of a structure, thus
revealing aspects of morphogenesis that are otherwise
overlooked. An example is the orientation of ciliary
bands of echinoderm larvae. Overlooking the necessary specification for larval forms has misled interpretation of an otherwise extraordinarily informative hybridization experiment.
The ciliary bands of echinodem larvae consist of
columnar ciliated cells that differentiate at the border
Function of development
proceed with little effect on the juvenile sea urchin.
Modularity in development confers flexibility in evolution and accounts for numerous violations of von
Baer’s law. Evolutionary divergence among animals is
not consistently greater at later developmental stages.
What accounts for this modularity?
One suggested origin of modularity is sequential
evolution of the modules. The separate and different
development of larval body and echinus rudiment in
sea urchins suggested the hypothesis that ancestral
bilaterian animals resembled the present day larvae,
and the juvenile and adult stages evolved via the
origin of cells set aside from development of the larval
body.36, 37
This hypothesis identified important differences
among animals’ embryos and larvae, but as originally
stated, it did not account for patterns of evolutionary
divergence in what juvenile structures develop from
separate rudiments and what larval structures are
retained through metamorphosis. For example, in
the sea cucumbers (relatives of the sea urchins),
the ‘set-aside’ cells with no apparent function in
the larval stage are simply parts of the coelom and
mesenchyme, and the body axes and site of the
mouth are retained through metamorphosis. In the
cidaroid sea urchins, which are the sister group of
euechinoids (other living sea urchins), the juvenile
mouth forms at the larva’s lower left side, but the
juvenile rudiment is not internalized.38 Thus the
extensive setting aside of cells for postlarval structures
in euechinoids appears to be a derived condition,
and much less setting aside of cells the ancestral
condition. Moreover, numbers of cell divisions in
the larval tissues are not constrained as originally
proposed.36 Echinoderm larval bodies vary greatly in
size and thus cell number, both among species and
as a developmental response to food supply. Explicit
analyses of the functional consequences of juvenile
rudiments can improve hypotheses on modularity
and set-aside cells.
Functional advantages of modularity appear to be
diverse. For bottom-living marine animals with planktonic larvae, structures suited to life at the sea bed
can be unsuited to life in the plankton and vice versa.
Hence metamorphosis occurs at about the time the
animals change habitat. One advantage of internalized or localized rudiments of juvenile structures is
to keep them out of the way while they are functionless.39, 40 Formation of the juvenile mouth at a new
site on the lower left side of the sea urchin larva permits continued larval feeding during development of
the juvenile, right up to the time of settlement and
to a larval seastar or sea cucumber suggested the
hypothesis that the early hybrid larva resembled
‘the dipleurula-type larva typical of other classes
of echinoderms and is considered to represent the
ancestral echinoderm larval form’. In the photos,
however, there is no indication of the tilts of the
ciliary band required by larval seastars or sea cucumbers. The H. e. × H. t. hybrid is more simply
interpreted as a larva that has formed the ciliary band
but without development of the skeletal rods that
would produce the form of a sea urchin pluteus.35
The resemblance to larval seastars or sea cucumbers
is explicable by subtraction of the skeletal rods rather
than resurrection of the ontogenetic pathway that
produces the larval ciliary bands of sea cucumbers or
seastars. The feeding larvae of echinoderms (pluteus,
bipinnaria, and auricularia) are, at early stages,
sufficiently similar, that a diagram of their common
features (drawn 30 years ago) resembles the H. e. ×
H. t. hybrid, but with a stronger ventral flexure.30
The form of the hybrid’s ciliary band can be
interpreted in relation to divergence in its immediate
parents’ development (as other features of the hybrid
were interpreted by the authors). From the evidence
presented thus far, there is no apparent need to
suppose a ‘surprisingly long term retention of an
ancient pattern of larval development’. There was
a long term retention of an ancient pattern in H.
tuberculata because it retained a feeding larva, but
the retention is not surprising because the ciliated
band developed each generation and mutations were
eliminated by selection during its entire evolutionary
This story should in no way diminish interest in
what is a remarkable evolutionary divergence in embryos and larvae, and the potential of the hybrids for
probing the developmental processes involved.33 The
moral is that attention to functional morphology can
help developmental biologists interpret evolutionary
divergence of developmental processes.
Evolution of modularity, along with some
set-aside cells
Development of body parts in separate, largely
independent modules permits evolutionary divergence in one part with little divergence in another.
Evolutionary loss of the ascidian tail can proceed
with little effect on the pharynx. Evolutionary loss
of larval ciliary bands, skeletal rods, and mouth can
R. R. Strathmann
completion of metamorphosis.30 Secondly, if developing juvenile structures are tucked away, they can be
kept longer in a less differentiated state, with advantages for growth41 as they are nourished by the larva.
Thirdly, development of juvenile structures to an advanced state before metamorphosis permits a rapid
transition to a new form and habitat.
Setting some cells aside as they develop into postlarval structures is a means of preparing the sea urchin
for a rapid change in form and habitat. Functional
requirements of larvae, juveniles, and transitions in
form and habitat help explain which cells are ‘set
aside’. Endoderm forming midgut becomes functional in feeding larvae and continues to be midgut
through and past metamorphosis. These cells also
store nutritional reserves that are carried through
metamorphosis. Continuity of midgut is common to
most marine invertebrate larvae, whereas the extent
to which mesodermal and ectodermal cells are ‘set
aside’ varies greatly. In echinoderms and hemichordates, some mesodermal cells (some mesenchyme
and parts of the coelom) lack apparent function in
the larva and contribute to the postlarval juvenile,
but the hydrocoel (a premier example of presumed
set-aside cells) functions as a larval excretory organ42
and mesenchyme contributes to larval muscles. A
plausible hypothesis is that metamorphoses evolved
many times in the marine bilateria and that the
setting aside of cells that form juvenile structures
accompanied divergence of larval and juvenile forms
and the elaboration of metamorphic transitions.
An additional advantage of modularity can be
developmental plasticity, in which developmental
timing, size of structures, or form of structures are
altered in response to environmental cues. Developmental plasticity in sea urchin larvae is an example
that involves the cells set aside in the echinus rudiment. The echinus rudiment develops within the
pluteus larva and becomes most of the sea urchin’s
body wall at metamorphosis. Timing of development
of the rudiment relative to the larval body depends
on the supply of food for the larva.43, 44 At low concentrations of food (algal cells), the formation of the
echinus rudiment is delayed. Even the invagination of
ectoderm towards the hydrocoel is delayed. Instead,
the larva grows longer arms and a longer ciliated
band. A hungry pluteus develops eight long arms but
almost no rudiment. A satiated larva develops stubby
arms, but accelerated development of the rudiment
is already apparent by the six-armed stage.
These changes in both timing and size appear to be
adaptive. The longer arms mean a longer ciliary band
and thus faster processing of water for food when
food is limiting.45 The larva first increases its ability
to catch food and then allocates growth to the rudiment. When food is abundant, however, higher rates
of capture do not increase growth rates. The larva
begins forming the juvenile structures immediately,
reducing the time for its development. The plasticity
appears suited to minimizing the development times
over a range of food concentrations. Rapid development is presumably advantageous because mortality
rates of these tiny larvae are high.46 A shorter larval
period should yield higher survival. Decoupling of
the development of the rudiment from structures
for larval feeding permits this developmental plasticity. Conversely, selection for this developmental
plasticity may have contributed to this decoupling in
A cautionary tale: stasis and change in a
seemingly simple and much studied trait
Tests of functional hypotheses can be based on the
consequences of experimental change in a trait, comparisons of performance of different organisms, or
both.47 Unsurprisingly, tests often reveal unsuspected
complications and raise new questions.
Comparison of 3 million years of evolutionary
divergence of eggs of echinoderms on the two sides
of the Isthmus of Panama revealed the predicted
trend of larger eggs in the ocean with a lower
concentration of food.48 Eggs were predicted to be
larger where there was slower larval growth or greater
larval death rates. Production of smaller eggs permits
production of more eggs, but the smaller eggs are at
greater risk because more growth and hence a longer
development is required in early life. The predicted
optimum size is a compromise between fecundity of
the mother and mortality of the offspring.
Confirmation of a prediction by replicated evolutionary events is gratifying, but the comparison also
revealed that the genus was a better predictor of egg
size than the ocean. The importance of ancestry in
the evolution of egg size suggests that selection on
egg size depends on other traits of these animals,
not just their environment, and that the traits differ
among genera. At present, these generic differences
are anyone’s guess. Experimentally decreasing sea
urchin egg volume by half is not directly lethal.47 The
generic differences could involve the habitats occupied by the adults (despite interocean differences),
Function of development
differences in larval forms, or other consequences
of coadapted gene complexes or pleiotropic effects
that involve egg size. The interocean comparison
suggests that in each genus there is a well integrated
phenotype that results in stabilizing selection on egg
size in a wide range of habitats.
Evolutionary divergence and stasis in development
is not simple. Analyses of the functional demands
met by embryos and larvae are an essential part of
explanations but do not tell the whole story.
A function of developmental processes is putting
the right kinds of cells in the right places at the
right times. The criterion for ‘right’ is survival and
For this essay, I gave some examples of functional
hypotheses that predict what is right. These were
qualitative statements, due to the limits of a short
essay and my abilities. Quantitative predictions are
more vulnerable to tests than qualitative predictions.
Mathematical skills are useful and, for many of us, a
reason for collaboration.
Converting ‘evodevo’ from descriptions of evolutionary change to explanations of evolutionary
change will require a connection to survival and
reproduction. Functional analyses are necessary for
making this connection.
Larry McEdward encouraged pompous pontification
and, with Peter Marko, Amy Moran, Jennifer Staver
and Greg Wray, provided useful suggestions for this
paper. Support was from NSF grant OCE9633193 and
the Friday Harbor Laboratories of the University of
1. Strathmann RR (1988) Functional requirements and the
evolution of developmental patterns, in Echinoderm Biology
(Burke RD, Mladenov PV, Lambert P, Parsley RL, eds) pp. 55–
61. A. A. Balkema, Rotterdam
2. Adiyodi KG, Adiyodi RG (1993) Reproductive Biology of Invertebrates. Asexual Propagation and Reproductive Strategies, vol 6A, p. 410, John Wiley & Sons, Chichester
3. Adiyodi KG, Adiyodi RG (1994) Reproductive Biology of In-
vertebrates. Asexual Propagation and Reproductive Strategies, vol 6B, p. 432, John Wiley & Sons, Chichester
Gilbert SF, Raunio AM (1997) Embryology. Constructing the
Organism, Sinauer, Sunderland
Dawkins R (1982) The Extended Phenotype, W. H. Freeman,
Crow JF (1988) The importance of recombination, in The
Evolution of Sex: An Examination of Current Ideas (Michod
RE, Levin BR, eds) pp. 56–73. Sinauer, Sunderland
Grosberg RK, Strathmann RR (1997) One cell, two cell,
red cell, blue cell: the persistence of a unicellular stage in
multicellular life histories. Trends Ecol Evol 13:112–116
Maynard SJ (1988) Evolutionary progress and levels of selection, in Evolutionary Progress (Nitecki MH, ed.) pp. 219–230.
University of Chicago Press, Chicago
Bell G, Koufopanou V (1991) The architecture of the life cycle
in small organisms. Phil Trans R Soc Lond B 332:81–89
Kondrashov AS (1994) Mutation load under vegetative reproduction and cytoplasmic inheritance. Genetics 137:311–318
Grosberg RK (1988) The evolution of allorecognition specificity, in Invertebrate Historecognition (Grosberg RK, Hedgecock D, Nelson K, eds) pp. 157–167. Plenum, New York
Buss LW (1987) The Evolution of Individuality, Princeton
University Press, Princeton, NJ
Stoner DS, Weissman IL (1996) Somatic and germ cell parasitism in a colonial ascidian: possible role for a polymorphic
allorecognition system. Proc Natl Acad Sci USA 93:15254–
Ayre DJ, Resing JM (1986) Sexual and asexual production of
planulae in reef corals. Mar Biol 90:187–190
Jackson JBC, Coates AG (1986) Life cycles and evolution of
clonal (modular) animals. Phil Trans R Soc Lond B 313:7–22
Strathmann RR (1985) Feeding and non-feeding larval development and life history evolution in marine invertebrates.
Annu Rev Ecol System 16:339–361
van den Biggelaar JAM, Haszprunar G (1996) Cleavage
patterns and mesentoblast formation in the gastropoda: an
evolutionary perspective. Evolution 50:1520–1540
Swalla BJ, Jeffery WR (1996) Requirement of the Manx gene
for expression of chordate features in a tailless ascidian larva.
Science 274:1205–1208
Hadfield KA, Swalla BJ, Jeffery WR (1995) Multiple origins
of anural development in ascidians inferred from rDNA
sequences. J Mol Evol 40:413–427
Richardson MK (1995) Heterochrony and the phylotypic
period. Dev Biol 172:412–421
Van Valen LM, Maiorana VC (1991) HeLa, a new microbial
species. Evol Theory 10:71–74
Halpern ME (1997) Axial mesoderm and patterning of the
zebrafish embryo. Am Zool 37:311–322
Riedl R (1978) Order in Living Organisms, John Wiley &
Sons, Chichester
Whittaker JR (1997) Chordate evolution and autonomous
specification of cell fate: the ascidian embryo model. Am Zool
Strathmann RR (1991) From metazoan to protist via competition among cell lineages. Evol Theory 10:67–70
Young CM, Gowan RF, Dalby J Jr, Pennachetti CA, Gagliardi D
(1988) Distributional consequences of adhesive eggs and anural development in the ascidian Molgula pacifica (Huntsman,
1912). Biol Bull 174:39–46
Woodin SA, Lindsay SM, Wethey DS (1995) Process-specific
recruitment cues in marine sedimentary systems. Biol Bull
R. R. Strathmann
28. Cameron RA, Britten RJ, Davidson EH (1993) The embryonic
ciliated band of the sea urchin, Strongylocentrotus purpuratus
derives from both oral and aboral ectoderm. Dev Biol
29. Strathmann RR, Jahn TL, Fonseca JRC (1972) Suspension
feeding by marine invertebrate larvae: clearance of particles
by ciliary bands of a rotifer, pluteus, and trochophore. Biol
Bull 142:505–519
30. Strathmann RR (1971) The behavior of planktotrophic echinoderm larvae: mechanisms, regulation, and rates of suspension feeding. J Exp Mar Biol Ecol 6:109–160
31. Hart MW (1996) Evolutionary loss of larval feeding: development, form and function in a facultatively feeding
larva, Brisaster latifrons. Evolution 50:174–187
32. Emlet RB (1994) Body form and patterns of ciliation in
nonfeeding larvae of echinoderms: functional solutions to
swimming in the plankton? Am Zool 34:570–585
33. Raff EC, Popodi EM, Sly BJ, Turner FR, Villinski JT, Raff RA
(1999) A novel ontogenetic pathway in hybrid embryos between species with different modes of development. Development 126:1937–1945
34. Emlet RB (1995) Larval spicules, cilia, and symmetry as
remnants of indirect development in the direct developing
sea urchin Heliocidaris erythrogramma. Dev Biol 167:405–415
35. Pennington JT, Strathmann RR (1990) Consequences of
the calcite skeletons of planktonic echinoderm larvae for
orientation, swimming, and shape. Biol Bull 179:121–133
36. Davidson EH, Peterson KJ, Cameron RA (1995) Origin
of bilaterian body plans: evolution of developmental and
regulatory mechanisms. Science 270:1319–1325
37. Peterson KJ, Cameron RA, Davidson EH (1997) Set-aside
cells in maximal indirect development: evolutionary and
developmental significance. Bioessays 19:623–631
38. Emlet RB (1988) Larval form and metamorphosis of a “primi-
tive” sea urchin, Eucidaris thouarsi (Echinodermata: Echinoida: Cidaroida), with implications for developmental
and phylogenetic studies. Biol Bull 174:4–19
Wilson DP (1932) On the mitraria larva of Owenia
fusiformis Delle Chiaje. Phil Trans R Soc B 221:231–334
Emlet RB, Strathmann RR (1994) Functional consequences of
simple cilia in the mitraria of oweniids (an anomalous larva of
an anomalous polychaete) and comparisons with other larvae,
in Reproduction and Development of Marine Invertebrates
(Wilson WH Jr, Stricker SA, Shinn GL, eds) pp. 235–257.
Johns Hopkins University Press, Baltimore
Ricklefs RE, Shea RE, Choi IH (1994) Inverse relationship
between functional maturity and exponential growth rate of
avian skeletal muscle: a constraint on evolutionary response.
Evolution 48:1080–1088
Ruppert EE, Balser EJ (1986) Nephridia in the larvae of
hemichordates and echinoderms. Biol Bull 171:188–196
Bertram DF, Strathmann RR (1998) Effects of maternal and
larval nutrition on growth and form of planktotrophic larvae.
Ecology 79:315–327
Strathmann RR, Fenaux L, Strathmann MF (1992) Heterochronic developmental plasticity in larval sea urchins and
its implications for evolution of nonfeeding larvae. Evolution
Hart MW, Strathmann RR (1994) Functional consequences
of phenotypic plasticity in echinoid larvae. Biol Bull 186:291–
Rumrill SS (1990) Natural mortality of marine invertebrate
larvae. Ophelia 32:163–198
Sinervo B, McEdward L (1988) Developmental consequences
of an evolutionary change in egg size: an experimental test.
Evolution 42:885–899
Lessios HA (1987) Temporal and spatial variation in egg size
of 13 Panamanian echinoids. J Exp Mar Biol Ecol 114:217–239