A Blog by Carl Zimmer

In Search of the First Animals

All animals–from corgis to Greenland sharks, from dog ticks to toucans to you–descend from a common ancestor. The fossil record of animals, which runs back over 600 million years, can help us travel back some of the way through animal evolution towards the origin of the kingdom. But those early rocks contain precious few remains of animals, and so fossils alone can’t tell us what our common animal ancestor looked like.

Scientists can add to their supply of clues by studying living animals. And it now looks as if some of the most important clues to how animals got their start come from a beautiful creature called the comb jelly. This video from the Monterey Bay Aquarium is a good introduction to their luminescent loveliness.

In the past, a lot of scientists would not have put so much importance on comb jellies. If you asked them (as I did) how animals evolved, they’ve sketch out a version of events that runs like this:


1. Before multicellular animals evolved, their ancestors were single-celled protozoans that may have formed colonies. Our DNA shows that our closest non-animal relatives are critters called choanoflagellates. I wrote about our single-celled cousins in the New York Times.

2. Our ancestors then crossed the line from colonial life to life as multicellular creatures. They became the first animals.

3. The animal lineage then started to split into new branches. Many branches are now extinct. Among living animals, the first split divided the ancestors of today’s sponges from all other species.

4. That could tell you something important about what the earliest animals were like. Sponges have no nervous system and no muscles, for example. All other animals, from comb jellies to starfish to clams to us, do. So the early animals didn’t have muscles and neurons yet, and didn’t evolve them until after sponges split off on their own. You might even go so far as to say that our own direct ancestors were sponges–animals anchored to the sea floor, filtering food through small pores.

5. Later, the common ancestor of the non-sponge animals evolved muscles and neurons. These animals began to move around–as pulsating jellyfish, crawling worms, and, eventually, swimming fish.

6. Later still, the muscle-and-neuron-carrying animals split apart into two main lineages. Jellyfish belong to a lineage called cnidarians. The other branch is known as bilaterians. It includes all the animals with heads, brains, and tails, from insects to mammals.

Comb jellies seemed to many researchers to simply be a cousin lineage to cnidarians. And that meant they weren’t important to understanding how animals first evolved.

But a few years ago something weird started to happen. When scientists compared the DNA from more and more species, some of them would end up with animal trees in which the comb jellies split off first, not the sponges. (Here’s a piece I wrote about that work for the Boston Globe in 2008.)

A lot of debate ensued. Drawing evolutionary trees is no simple task, especially when you’re looking at branches that split off from each other hundreds of millions of years ago. Teasing apart the order of the branches can be as tricky as teasing apart the stars in a distant corner of the galaxy. So evolutionary biologists built themselves a better telescope.

A new paper today in Science is that telescope. In the past, scientists have compared limited segments of DNA from different species to work out the animal tree. A team of researchers at the National Institutes of Health and elsewhere decided that a more powerful view of the tree might come from looking at the entire genomes of many animals. And it just so happened that of all the major groups of animal species (known as phyla), only the comb jellies were without a sequenced genome.

The scientists rectified the matter, sequencing the first comb jelly genome, belonging to a species called Mnemiopsis leidyi. They then compared the genomes of 13 species, and then did a second comparison of smaller pieces of DNA from 58 species. The studies all pointed in the same direction: comb jellies, not sponges, split first. And the statistical support for that split was very strong.

This diagram, published in Science, sums up the findings nicely:


Having sequenced a comb jelly genome, the scientists could also march through its catalog of genes to see how many genes for different functions it shares in common with us. Despite the fact that comb jellies have muscles, they lack many of the genes essential for muscles in other animals. This would suggest that muscles evolved twice in the animal kingdom.

On the other hand, comb jellies have a lot of nervous system-related genes in common with bilaterians and cnidarians. It’s therefore possible that the common ancestor of all living animals had a simple nervous system. Sponges lost their nervous system and muscles entirely as they adapted to a quiet existence as filter feeders. (The placozoans shown on this tree are an obscure group of weird animals that are just tiny sheets of cells that creep across the sea floor. If the new study is right, they lost their nervous system and muscles too.)

I asked Antonin Rokas, an expert on animal relationships at Vanderbilt University, what he thought of the research. ” I think it’s a step in the right direction,” he told me, “but I doubt that it will silence those who have championed sponges as the earliest branching animal phylum.” The idea of comb jellies as belong to such an ancient lineage runs counter to a lot of thinking in zoology for over a century. It may take studies of more genomes to convince the tougher skeptics.

Aside from basic curiosity, there are other good reasons to settle the debate–and they help explain why the National Institutes of Health spearheaded this study. The NIH sponsors research on animals as models for human diseases. That’s because we share a lot of genes in common with them. Of particular interest to health researchers are so-called “disease genes”–human genes that are associated with diseases if they acquire mutations.

It turns out that over half of known disease genes are present in comb jellies–including many that are missing from species such as fruit flies that scientists use a lot to study human diseases. A lot of our diseases may arise from damage to the fundamental system for building an animal that evolved some 700 million years ago.

(For more on comb jellies, read this feature from earlier this year by Amy Maxmen in Science News.)

Stefan Siebert, Brown University
Stefan Siebert, Brown University

14 thoughts on “In Search of the First Animals

  1. The new tree does not look very parsimonious to me, since they show loss of nervous system and gain of mesoderm twice. I suppose they have identified other characters that override that. One odd thing: They use ‘gain of mesoderm’ in the chordate lineage but not in the cnidarians, so I suppose this annotation is meant to indicate the origin of triploblasts. But they also use the same annotation for mesoderm in the ctenophores. Like cnidarians, ctenophores are traditionally considered diploblasts not triploblasts. Is that wrong now?

  2. Great article. One minor quip: “The other branch is known as bilaterians. It includes all the animals with heads, brains, and tails, from insects to mammals.” Although it technically doesn’t, it does seem to imply that all bilaterians have heads. Depending of course on how you define head, this isn’t really true.

  3. A few naive questions:
    1. what’s the boundary between a colony and an organism?
    2. where do plants come in?
    3. What about external boundary? ectoderm? (which ought to give rise to CNS, at least in vertebrates).

  4. I really enjoyed reading your beautifully written article about the Lobed Comb Jelly! Thanks for including the lovely video from the Monterey Bay Aquarium- it was delightful!

  5. Fascinating! Is it possible to add the approximate dates of the various splits in the branching diagram shown in the article?

  6. The idea that ctenophores are the outgroup to the rest of the animals is not new. Dunn et al. introduced this tree in 2008 (doi:10.1038/nature06614), which was then found to be flawed by Pick et al in 2010 (doi: 10.1093/molbev/msq089). Many of the same authors have now returned in 2013 to defend their 2008 paper with more evidence. But, I’m sure in a couple of years someone will come out with another paper that changes the tree back again. The problem is, yes there is some genetic support (depending on how you run the analysis), but having ctenophores as an outgroup does not create the most parsimonious tree when you consider all factors. Loss of nervous system twice? Gain of mesoderm twice? And you mention that muscles would have had to evolve twice, but cnidarians have muscles that also evolved independently of other phyla (doi:10.1038/nature11180), so moving the cnidarians away from the ctenophores means muscles would have had to evolve three times! Not to mention the obvious morphological similarities between poriferans and choanoflagellates. And that the placozoans are poorly understood on all levels, so their placement anywhere on the tree is guesswork. Also, keep in mind that ctenophores are actually a rather diverse group morphologically (and probably genetically as well) and all of the trees built use genes from a single species out of probably 200 ctenophore species. I’d love to see an analysis including a platyctenid genome as well since the phylogeny of the ctenophora is currently based only on rRNA.

  7. New imaginary family trees come and go.
    Dr. Tim White is an anthropologist at the University of California in Berkeley:
    ”Perhaps no area of science is more contentious than the search for human origins. Elite paleontologists disagree over even the most basic outlines of the human family tree. [So-called] New branches grow amid great fanfare, only to wither and die in the face of new fossil finds.” (Robert Locke, “Family Fights” Discovering Archaeology, July/August 1999, p. 36-39)

  8. The “tree produced by maximum-likelihood analysis of gene content” in the original article is quite crazy when it comes to relationships within Bilateria. Suffice it to say that it makes Chordata paraphyletic with respect to a (snail + polychaete) clade and to echinoderms (all nodes with 100% bootstrap support). Don’t such results disqualify the method itself?

  9. Perhaps the rate of molecular evolution is faster in the ctenophores and their branch has simply slid down the reconstructed tree as a result.

  10. Hopefully in the year 2014, humanity and Science will finally evolve away from the trajectory towards continued use of meaningless, “art” terms like “kingdom” and “phylum.” Maybe — just “maybe,” I don’t have my hopes up — we’ll get away from the use of ranks as they do nothing but occlude our understanding of evolution.

  11. Although parsimony is a useful tool for scientists it is not required in nature. If a feature is know to arise once, it is not a big stretch to think it could happen twice.

Leave a Reply

Your email address will not be published. Required fields are marked *