BOOK REVIEW
Evolution's Many Branches
Douglas Erwin
Assembling the Tree of Life. Edited by Joel Cracraft and
Michael J. Donoghue. xvi + 576 pp. Oxford University Press, 2004. $59.95.
Big problems often require big science. For example, synchrotrons,
cyclotrons, linear accelerators and interplanetary spacecraft all
cost too much for single investigators. Thus high-energy
experimental physicists and planetary geologists have been forced to
come up with communal approaches to research and to acquire the
political skills needed to get their projects funded. The advent of
farms of DNA sequencers and of the necessary computational power to
make them useful has brought molecular biology into the era of big
science as well with a variety of efforts to sequence whole genomes,
including the Human Genome Project. Related developments include
various proteomic and structural biology projects and the sprouting
of new departments of systems biology, in which computer scientists
and underemployed physicists mix freely with biologists. But big
science is not driven exclusively by problems requiring large,
expensive instruments. Innovation in instrumentation can also be a
factor, opening up new avenues for exploration and reopening old
questions that had been abandoned because progress was so difficult.
The effort to produce a Tree of Life—a "correct and
verifiable family tree" of all life, both living and
extinct—is unquestionably big science. This formidable
undertaking will require the mobilization of vast numbers of
systematists to acquire and analyze reams of morphological,
behavioral and molecular data, and the informatics involved in the
unenviable task of classifying and storing phylogenetic information
is complex. The end result will show the evolutionary relationships
between all organisms, from those whose diversity is still poorly
explored—such as crenarchaea (sulfur-metabolizing organisms
that thrive at high temperatures) and heterokonts (which include
water molds, diatoms and brown algae)—to nematodes, probably
the most species-rich group of animals.
What might be the justifications for constructing a Tree of Life? E.
O. Wilson nominates simply having a complete accounting of life on
Earth, promoting conservation, searching for new biological products
and improving our understanding of community assembly (how species
coadapt to live together in a given spot). Not surprisingly, Wilson
makes an effective case that a Tree of Life will revolutionize
ecology by marrying NASA-like technology with old-fashioned
fieldwork to allow rapid characterization of broad swaths of the
members of a community.
Assembling the Tree of Life grew out of a 2002 symposium
(sponsored jointly by the American Museum of Natural History, Yale
University, The International Union of Biological Science and the
international biodiversity science program DIVERSITAS) that produced
a synthesis of knowledge about evolutionary relationships among the
major branches of the Tree of Life. The book is the first
comprehensive scientific attempt to consider the tree of life since
the publication in 1989 of The Hierarchy of Life (the
proceedings of a Nobel Symposium, edited by Bo Fernholm and others).
Although we are not yet within reach of Wilson's dream, we are much
closer to it than one might have expected in 1989. The
"debate" between molecules and morphology, which was a
centerpiece of The Hierarchy of Life, has vanished with the
recognition that no one source of information can provide an
infallible guide; rather, a variety of combined-evidence approaches
are required. Editors Joel Cracraft and Michael J. Donoghue have two
ambitious and at times contradictory aims: to demonstrate to readers
outside the field of systematics the broader significance of
building the Tree of Life (that is, to explain why systematists need
big science) and to provide a current assessment of phylogenetic
efforts across the tree.
The contributors, who include nearly 100 systematic biologists, are
more successful in achieving the latter goal. There is no question
that individual chapters will be useful for those seeking a meaty
overview of a particular clade. Most of the systematic chapters are
written by the premier experts in the area and include a brief
synopsis of the morphology of the group, some anatomical highlights
and comments on diversity. The better chapters include detailed and
critical commentaries on previous phylogenetic analyses and discuss
where they might have gone wrong; the chapter by Maureen A. O'Leary
and colleagues on mammalian phylogeny is particularly noteworthy in
this regard.
The early chapters on microbial phylogeny provide a wealth of
information on the problems of phylogenetic reconstruction in the
face of an unknown degree of lateral gene transfer and serve as an
excellent primer on the evolutionary history of these groups. The
differing perspectives on phylogenetic relationships offered by
Sandra L. Baldauf and colleagues, by Norman R. Pace and by W. Ford
Doolittle illustrate the magnitude of the problem. The chapter on
early algal evolution by Charles F. Delwiche and others provides an
outstanding illustration of how critical a phylogenetic perspective
is to unraveling the evolutionary history of a clade. It is
particularly unfortunate that the authors of many of the chapters on
animal groups did not make a similar effort; many of those authors
appear to view a phylogeny as an end in itself, rather than as a
tool to advance other questions.
The volume is a treasury of unexpected information. Who knew that in
the 1830s France imported 50 million leeches annually for medicinal
bloodletting (and that the government collected a tax of one franc
per thousand leeches)? And who would have guessed that the key to
unraveling lepidopteran phylogeny lies in "an almost infinite
variety of small, drab moths from multiple evolutionary lineages"?
Curiously, different authors in the volume appear to mean different
things by the Tree of Life. Most appear to be principally concerned
with the topology of the tree—with the relationships between
various subclades and species that the tree depicts. Others,
including Wilson, seem more concerned with taxonomic descriptions,
databases of images of types and the like—an effort that has
also been described as the Encyclopedia of Life. The distinction
between these disparate views is critically important for
identifying the scope and likely cost of the project as well as for
deciding how it should be carried out. If a tree alone is the goal,
the new effort at DNA bar-coding might be all that is required. One
can even imagine the whole process being automated, with organisms
fed into a hopper at one end and the critical sequences isolated,
sequenced and added to GenBank (the genetic sequence database of the
National Institutes of Health) as the biological exudates are heaped
on a growing recycling pile. Of course most of the contributors to
this volume are not really interested in topology alone,which would
provide us with none of the critical information that accompanies
proper systematic treatments—information about functional
adaptations essential for understanding evolutionary pattern and
process, for example.
What is missing from the volume? Understandably, most clades are not
treated in much detail, with the exception of one subclade of
aberrant, highly encephalized primates. Cnidarians (such as
jellyfish and corals) get short shrift, which is rather unusual
given that significant advances have been made recently in
understanding the group. But Douglas J. Eernisse and Kevin J.
Peterson, in their detailed discussion of metazoan phylogeny, do
discuss the recent evidence that the calcareous and siliceous
sponges arose independently. Happily, arthropods and their
ecdysozoan relatives have been allotted just five chapters (some
enthusiasts will doubtless be disappointed).
There are two more telling omissions. Although the editors'
introduction provides a very brief historical synopsis of
tree-building, complete with Darwin's canonical figure from The
Origin of Species and Ernst Haeckel's 1866 Tree of Life, a
contribution by a historian of science on evolving approaches to the
subject would have been most welcome. The second omission is more
procedural or methodological: Few contributions explicitly address
our current abilities to actually produce a rigorous,
well-substantiated tree of life. This is far from a simple matter,
as the most serious discussion of this shortcoming (in the chapter
by O'Leary and others) makes clear. Building supertrees is more
complicated than adding up previously published trees or building a
massive character matrix. The initial steps in resolving this
problem have been quite positive, but ultimately the viability of
the Tree of Life enterprise requires addressing these and related
methodological issues. Some might suggest that it would have been
inappropriate to include such dirty laundry in this volume, but I
would argue that to have done so might have gotten additional
computer scientists and mathematicians interested in these problems.
Reviewers are expected to offer some platitudinous comments on the
appropriate readership for a volume, although whether this is for
the edification of librarians or the gratification of the publisher
continues to elude me. The question is particularly apt in the case
at hand, for the intended readership is as poorly resolved as some
of the topologies. In their introduction, the editors note the
impact of Fernholm's The Hierarchy of Life. My observations
suggest that it is one of the more widely stolen library volumes,
which is a sort of impact metric, and perhaps Assembling the
Tree of Life will also disappear from shelves. It should
probably be required reading for first-year biology graduate
students, but otherwise the effort to demonstrate the significance
of the Tree of Life project is largely preaching to the converted.
The first three chapters (by Terry L. Yates and others, Rita R.
Colwell and Douglas J. Futuyama) provide some stimulating insights
into how the Tree of Life could be useful in human health,
conservation and agriculture. The 26 systematic chapters are most
likely to appeal to specialists in related areas and to students and
teachers seeking a solid, tree-based introduction to specific clades.
So why do we need to construct the Tree of Life? Because ultimately
it is, like a synchrotron or a spaceship, a tool that will allow
future generations of scientists to address a whole new set of
questions—about the ecological and evolutionary processes that
have produced the diversity of life on Earth.