RESEARCH NEWS & VIEWS
the complication of convergent evolution.
The work is another success story in the
young discipline of phylogenomics — which
attempts to resolve evolutionary history by
genomic comparisons — and is one of the first
aimed at probing deep within a phylum. Struck
The origin of the annelids is buried in distant evolutionary time. A molecular
and co-workers have sequenced about a thouphylogeny resolves their deep family interrelationships and provides a picture of
sand expressed-sequence tags (complementheir ‘urannelid’ ancestor. See Letter p.95
tary DNA library clones) from 17 members of
annelid families and complemented these collections with existing data, yielding a total repDETLEV ARENDT
eat sediment or surface deposits, or filter the resentation of 34 annelid species. They selected
surrounding water with their tentacle crowns. 231 genes common to at least one-third of the
ome animals are so familiar that we can
In 1865, a similar grouping, but excluding total taxa, and aligned and concatenated them
scarcely believe we know little about their the earthworms and leeches, was put forward into a supermatrix of 47,953 amino acids.
origins. Take earthworms, for example, by Jean Louis Armand de Quatrefages de Bréau.
Next, they reconstructed a phylogenetic tree
common to everyone’s garden, which belong to However, it was dismissed by later authors, who using refined methods capable of handling
a large phylum of invertebrates — the annelids considered the similarities in lifestyle to be con- the diversity of amino-acid substitution pro(ringed worms). The hitherto shrouded evolu- vergent adaptations due to similar ecological cesses in such a supermatrix. Many of the
tionary history of annelids is now illuminated constraints. Ironically, Struck and colleagues’ nodes in their tree, especially that separating
by Struck et al. on page 95 of this issue1.
study1 reveals that the older classification was the Errantia from the Sedentaria, had remarkAnnelids are global players in terrestrial closer to the truth, thus ‘revising the revision’. ably high support values (contrasting with
and freshwater environments, and in marine The authors’ phylogeny demonstrates that those of previous annelid phylogenies based on
ecosystems, where they live in and on the sea broad features of lifestyle and morphology, single genes4), making it highly likely that this
floor. But the identity of their nearest relatives even if sometimes challenging to quantify, grouping is definitive.
(maybe molluscs, maybe flatworms), and can be at least as informative as ultrastrucThe case of the annelids exemplifies both the
even their affinities within the phylum, has tural or fine morphological characteristics, beauty and the pitfalls of phylogeny reconstrucremained a puzzle. This lack of understand- and are not necessarily much more prone to tion when applying the principle of parsimony,
ing has another edge, given that variwhich settles on the tree minimizing gain
ous annelid species serve as model
or loss of particular characteristics. At the
organisms for the investigation of
molecular level, this approach has proved
basic biological processes. Embryolovery powerful, and it has been further
gists and neuroscientists have studied
enhanced by the advent of phylogenomleeches for decades. And the ragworm
ics. But it is becoming increasingly
Platynereis has recently emerged as a
obvious that, on the basis of morphologivaluable model for studying developcal characteristics alone, there is a serious
ment, evolution and neurobiology 2,
problem: the apparent ease with which
along with Capitella, Hydroides and
such characteristics are lost.
other marine species3. Just imagine
This point is illustrated by a morphoworking on vertebrate models such as
logical parsimony analysis of annelid
fish, mice and frogs without knowing
phylogeny5 that established a group,
their evolutionary interrelationships.
the Palpata, whose members were
Struck et al.1 have made a signifidefined by the presence of specific head
cant advance in the reconstruction of
appendages (palpae) — the implication
annelid phylogeny, having resolved
being that other groups without palpae
their internal affinities. Using molecuhad never had them. The new annelid
lar techniques, they have studied the
phylogeny instead indicates independrelationships between various annelid
ent loss of palpae in errantian and
families and orders, and have obtained
sedentarian groups, as was previously
a surprisingly clear result. It turns out
suggested6,7 by two of the co-authors of
that annelids are deeply subdivided
the current paper1. This example corinto two main groups, the Errantia (to
roborates the general idea of frequent
which the ragworm belongs) and the
and independent loss of traits during
Sedentaria (to which the leech, earthanimal evolution.
worm, Capitella and Hydroides belong).
Finally, the new work1 nicely illustrates
As the names suggest, the sub­
how, once the (molecular) phylogeny has
division of annelids into the Errantia
been solved, matrices of morphological
and Sedentaria matches their overcharacteristics can be used to reconstruct
all lifestyles (Fig. 1). Members of the
the common ancestors of the respective
Errantia are free to move about, and Figure 1 | An illustration of the Errantia and Sedentaria by
groups. If a given characteristic is found
Ernst Häckel, dated 1904. To take just two examples, top right is
crawl, swim or burrow. Many are Eunice magnifica (Grube, 1866), Eunicidae, an errantian; top left
in both branches resulting from a node,
predators or feed on macroalgae. By is Sabellastarte spectabilis (Grube, 1878), Sabellidae, a sedentarian.
it must have been present in the common
contrast, representatives of the Sed- Errantians often have especially prominent lateral appendages
ancestor. In this way, we can infer a lot
entaria are hemi-sessile burrowers or with bristles, for undulatory crawling. Many sedentarians exhibit
about the ‘urannelid’, the last common
tube dwellers (apart from the highly beautiful tentacle crowns for filtering plankton and other food
ancestor of all annelids. Most signifispecialized, parasitic leeches). They particles from the water.
cantly, it was an animal richly equipped
E VOLUTIO NARY B IO LO GY
Annelid who’s who
4 4 | NAT U R E | VO L 4 7 1 | 3 M A RC H 2 0 1 1
© 2011 Macmillan Publishers Limited. All rights reserved
MARIXVERLAG
S
NEWS & VIEWS RESEARCH
with sensory organs. It had chemosensory
nuchal organs and palpae; a pair of two-celled
larval eyes for phototaxis2; and possibly a pair
of more elaborate, multicellular adult eyes with
an alternating arrangement of rhabdomeric
photoreceptors and shading pigment cells. The
latter combination is found in extant errantians8
and in the sipunculans9, which represent an outgroup to both errantians and sedentarians (see
Fig. 1 of the paper1).
The urannelid was segmented, a detail that
is clear from the nested position of two unsegmented taxa, the echiurans and sipunculans,
within segmented groups10. This ancestor
probably lived on the sea floor, using its relatively complex lateral appendages for undulatory crawling (as seen in today’s Errantia and
for example in the Spionidae, which lie in the
basal part of the Sedentaria branch of the tree).
Given the power of phylogenomics, we might
soon know what the urannelid mollusc- or
flatworm-like relatives looked like in the
ancient oceans. ■
Detlev Arendt is in the Developmental
Biology Unit, European Molecular Biology
Laboratory, 69012 Heidelberg, Germany.
e-mail: [email protected]
Struck, T. H. et al. Nature 471, 95–98 (2011).
Jékely, G. et al. Nature 456, 395–399 (2008).
McDougall, C. et al. Parasite 15, 321–328 (2008).
Struck, T. H. et al. BMC Evol. Biol. 7, 57 (2007).
Rouse, G. W. & Fauchald, K. Zoologica Scripta 26,
139–204 (1997).
6. Purschke, G., Hessling, R. & Westheide, W. J. Zool.
Syst. Evol. Res. 38, 165–173 (2000).
7. Bleidorn, C. J. Zool. Syst. Evol. Res. 45, 299–307 (2007).
8. Purschke, G., Arendt, D., Hausen, H. & Müller, M. C. M.
Arthropod Struct. Dev. 35, 211–230 (2006).
9. Hermans, C. O. & Eakin, R. M. Z. Zellforsch. Mikrosk.
Anat. 100, 325–339 (1969).
10.Kristof, A., Wollesen, T. & Wanninger, A. Curr. Biol.
18, 1129–1132 (2008).
1.
2.
3.
4.
5.
C LIM ATE CHANGE
Another Antarctic
rhythm
A novel explanation for the long-term temperature record in Antarctic ice cores
invokes local solar radiation as the driving agent. This proposal will prompt
palaeoclimate scientists to pause and to go back to basics. See Letter p.91
KO J I F U J I TA
A
nalysis of ice cores retrieved from the
Antarctic and Greenland ice sheets
is one of the main sources of our
understanding of past climate. A component
of that understanding is that, on timescales
of 20,000 years and more, climate change in
Antarctica is determined by the amount of
solar radiation (insolation) reaching high
northern latitudes in summer. On page 91
of this issue, Laepple et al.1 call into question
some of the evidence for that view.
Precisely dated polar ice cores have allowed
examination of the ‘bipolar see-saw’ relationship of air temperatures between the
hemispheres on millennial timescales2, as well
as of longer-term, glacial–interglacial climate
change paced by variations in Earth’s orbit —
the Milankovitch forcing of ice ages3. In these
studies, the use of isotopes that are stable in
water, in the form of the ratios of oxygen and
deuterium isotopes, is well established. These
ratios constitute the fundamental proxy measurements for estimating past temperatures
from ice cores at both poles2–4.
Because ice cores consist of ice, the stableisotope ratios in the ice stem from those contained in precipitation (snow, which becomes
compacted to ice). In other words, if there is
no precipitation, no isotopic signal remains
in the ice core. This simple principle has been
acknowledged in interpreting the Greenland
ice-core record5. Subsequent studies6,7 have
described how changes in the seasonal pattern of precipitation during glacial–inter­glacial
cycles have significantly biased the isotopic
temperature record in Greenland. But it was
thought that the effect in Antarctica was probably minor because of its comparatively stable
precipitation seasonality.
Laepple and co-authors1 apply this idea of
precipitation seasonality to the Antarctic icecore record. However, they do not deal with
changes in seasonal patterns, as the previous
studies did, but instead consider the situation
in which seasonality is itself unchanging and in
which snow accumulation over inland Antarctica is maximal in winter and minimal in summer. This seasonality in snowfall has various
causes, such as the strong radiative cooling that
induces clear-sky precipitation and increased
moisture transport in winter, and sublimation
of ice into water vapour in summer.
By assuming that this seasonal pattern of
snow accumulation has persisted throughout
glacial–interglacial cycles, and that the local air
temperature has fluctuated according to the
present-day relationship between temperature and insolation, the authors1 produce an
© 2011 Macmillan Publishers Limited. All rights reserved
accumulation-weighted insolation signal as a
record of temperatures in Antarctic ice cores.
They find that it has the opposite phase to the
orbital-precession component (determined
by long-term changes in the orientation of
Earth’s rotational axis) of the local summerinsolation signal — and so, surprisingly, that
it is in phase with summer-insolation intensity
in the Northern Hemisphere.
If the Antarctic local temperature is determined by local insolation, the precession
component in the ice-core temperature signal
should be out of phase with Northern Hemisphere insolation, because the precession
component is out of phase between the two
hemispheres. However, the precession component filtered from the isotopic temperature
record in the Antarctic ice cores is coherent
and in phase with the Northern Hemisphere
insolation intensity3 — seemingly supporting the Milankovitch theory, according to
which southern climate is driven by insolation
changes at high northern latitudes.
But does the close phasing necessarily support a causal relationship? Perhaps not. Laepple and co-authors1 have rethought how the
signals of temperature change are produced.
Their accumulation-weighted insolation
record suggests that a precession rhythm synchronized with — but not caused by — the
Northern Hemisphere could be generated if
the local temperature fluctuated in line with
local insolation conditions in the Southern
Hemisphere. The unveiling of this ‘pseudorhythm’ strikes at the foundation of temperature estimates gleaned by analysing isotope
ratios in ice cores. Does it mean, as Laepple
et al. suggest, that the evidence from Antarctic
ice cores cannot be used to support or refute
the Milankovitch theory?
This theory is supported not just by temperatures inferred from Antarctic ice cores,
but also by sea surface temperatures recorded
in sediment cores from the Southern Ocean.
In these cores, the orbital-precession rhythm is
often found to be in phase with summer insolation in the Northern Hemisphere and therefore opposing the local summer insolation8.
The seasonality of snow accumulation does
not affect sediment processes in the ocean.
Furthermore, the existence of shorter (millennial timescale) but strong bipolar see-saw connections between the two hemispheres implies
that there are indeed mechanisms for the interhemispheric propagation of climate signals
through the ocean and/or atmosphere2. There
is no reason to believe that such mechanisms
have not operated over longer timescales.
A caveat regarding the results themselves
is that Laepple and colleagues’ insolationbased air-temperature estimate shows a
rather small amplitude (around 0.7 °C peak
to peak) compared with that derived from
ice cores (3 °C peak to peak). This is probably because the authors’ use of local insolation as the temperature proxy means that they
3 M A RC H 2 0 1 1 | VO L 4 7 1 | NAT U R E | 4 5
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.