New phylogenomic tools have made it possible to construct a robust phylogenetic tree of mollusks. This tree can be used to teach important evolutionary principles.

When I was studying biology in high school, my wonderful teacher, Mary Whelan, very carefully explained to us that you could not say that mollusks were characterized by the presence of shells, because there were mollusks such as the squid and octopus that did not have shells. Instead, she said, we should consider soft bodies the defining characteristic of mollusks. I have taught this to my own students ever since, but now I’ve had to reconsider this framework.

First, “having soft bodies” is not really a good way to describe mollusks. Organisms in many other phyla, like the cnidarians, also have soft bodies. And mollusks have such a tremendous diversity of traits that to say that they have soft bodies is not terribly informative.

Recent advances in genome sequencing and phylogenetic analysis, discussed below, have made a robust phylogenetic tree of mollusks possible. This enables us to use mollusks to illustrate important evolutionary principles involving the history of life, how phylogenetic trees are constructed, and the kinds of information that can be obtained from them.

A clade is a group that consists of a common ancestor and all its descendants. A clade is a natural group, which means that all organisms in a clade are more closely related to each other than to any other organism. A conserved core property is something present in the common ancestor of a clade and widely distributed among members of the clade. It does not have to be present in every single member of the clade, because some may have lost the property over evolutionary time. In almost any major group of organisms, there are a few members that do not possess a defining property of the group because, although their ancestors had it, they have lost it. For example, there are a few plants, such as the Indian pipe (Monotropa uniflora), that do not photosynthesize. Does this mean that photosynthesis is not a conserved core property of plants? Of course not. The common ancestor of all plants did photosynthesize, and just because a few groups subsequently lost this ability does not mean that it is not a conserved core property of plants. Mollusks are a particularly good example of this important general point.

The beginning of the Cambrian period was a critical time in evolution. During a period of ~25 million years, between 541 and 515 mya, at least half of the modern animal phyla suddenly appear in the fossil record (Kouchinsky et al., 2012). Perhaps the other modern animal phyla also evolved at that time, but we do not have their fossil record until later.

The first mollusks appear in the fossil record 535 mya (Kouchinsky et al., 2012). Mollusks diversified during the Cambrian, and by the end of the Cambrian, 490 mya, it is likely that all classes of mollusks had been established (Gonzalo Giribet, personal communication). Therefore, each class of mollusk, from gastropods to bivalves to chitons, has had hundreds of millions of years of independent evolution. As a result, they are quite different from each other. And yet mollusks are a clade – all mollusks are more closely related to each other than to any other animal, a fact that ultimately defines the phylum Mollusca. As such, they should share some conserved core properties that were present in their common ancestor.

Conserved Core Properties of Mollusks

Presence of a Calcium Carbonate Skeleton

All mollusks, except for a group of cephalopods such as squid and octopuses, can make some kind of calcium carbonate (CaCO3) skeleton (Figure 1 and Table 1). Most modern mollusks have calcium carbonate in some form, either as a shell or as spicules. Thus, it is very likely that the ability to make a calcium carbonate skeleton was present in the common ancestor of all mollusks. So the ability to make some kind of calcium carbonate skeleton should be considered a conserved core property of mollusks, something that was present in the common ancestor of all mollusks and is widely distributed among mollusks today.

Figure 1.

A phylogenetic tree of mollusks (modified from HMNH Exhibit; see text, under  “Sources”).

Figure 1.

A phylogenetic tree of mollusks (modified from HMNH Exhibit; see text, under  “Sources”).

Table 1.

Description of the eight classes of mollusks.

ClassDescription
Scaphopoda “Tusk shells.” These animals have curved conical shells that are narrow, pointed, and open at both ends. There are 817 known species. 
Gastropoda Snails, slugs, nudibranchs, and limpets. There are 67,690 living species listed in the WoRMS database, representing ~70% of known living mollusk species. 
Bivalvia Clams, mussels, oysters, and scallops. There are 23,889 known living species, making this the second-most-numerous class of mollusks. 
Cephalopoda Squids, octopuses, cuttlefish, and chambered nautilus. There are 1939 known living species. Only 4 of these species, all in the genus Nautilus, retain an external shell. 
Monoplacophora There are 47 living species, found mainly in the deep ocean. This class was known from the fossil record and was believed to be extinct until some living organisms were found in 1957. Smith (2011) placed the Monoplacophora as sister to Cephalopoda. This is not too surprising, since the oldest cephalopods and fossil monoplacophorans both had chambered shells. 
Aplacophora These have no shells, but their bodies are covered in tiny spike-like spicules made of calcium carbonate. There are ~540 living species. All of these are marine, and most of them live in deep water. There are two classes of Aplacophora: the Caudofoveata and the Solenogastres. 
Polyplacophora Chitons. There are 1998 living species of chitons known. They do not have shells and are “covered by overlapping plates of calcium carbonate ringed by a muscle, called a girdle” (HMNH Exhibit). 
ClassDescription
Scaphopoda “Tusk shells.” These animals have curved conical shells that are narrow, pointed, and open at both ends. There are 817 known species. 
Gastropoda Snails, slugs, nudibranchs, and limpets. There are 67,690 living species listed in the WoRMS database, representing ~70% of known living mollusk species. 
Bivalvia Clams, mussels, oysters, and scallops. There are 23,889 known living species, making this the second-most-numerous class of mollusks. 
Cephalopoda Squids, octopuses, cuttlefish, and chambered nautilus. There are 1939 known living species. Only 4 of these species, all in the genus Nautilus, retain an external shell. 
Monoplacophora There are 47 living species, found mainly in the deep ocean. This class was known from the fossil record and was believed to be extinct until some living organisms were found in 1957. Smith (2011) placed the Monoplacophora as sister to Cephalopoda. This is not too surprising, since the oldest cephalopods and fossil monoplacophorans both had chambered shells. 
Aplacophora These have no shells, but their bodies are covered in tiny spike-like spicules made of calcium carbonate. There are ~540 living species. All of these are marine, and most of them live in deep water. There are two classes of Aplacophora: the Caudofoveata and the Solenogastres. 
Polyplacophora Chitons. There are 1998 living species of chitons known. They do not have shells and are “covered by overlapping plates of calcium carbonate ringed by a muscle, called a girdle” (HMNH Exhibit). 

Notes: All descriptions are from the HMNH Exhibit, and the numbers of species are from WoRMS Editorial Board (see text, under “Sources”). The species numbers were accurate as of February 2014. Clearly, these numbers are continually changing as new species are discovered, and others have been reclassified.

Presence of a Calcium Carbonate Shell Is a Conserved Core Property of One Large Clade of Mollusks

Another major innovation occurred in the common ancestor of the clade of shelled mollusks. These are the most familiar mollusks, including Gastropoda, Bivalvia, Cephalopoda, and the less familiar Scaphopoda and Monoplacophora. The calcium carbonate skeleton is made by specialized cells that secrete calcium carbonate spicules. The common ancestor of all mollusks probably had specialized cells that produced either the spicules found in today’s Aplacophora or a covering similar to that found in today’s Polyplacophora (chitons). However, long ago in the Cambrian, when the different classes of mollusks were diverging from each other, in one group of ancestral mollusks there was a condensation of cells that secreted the spicules into a single shell gland (HMNH Exhibit; Figure 1). The descendants of this common ancestor include all the mollusks with hard shells, including the two most numerous groups, the gastropods and bivalves. This clade also includes the cephalopods.

Most Cephalopods Have Lost the Ability to Make External Shells

Among the 1939 species of living cephalopods, all but 4 species of chambered nautilus have subsequently lost the ability to make external shells, as shown in Figure 1. This is probably an adaptation that enabled them to evolve another useful trait, that of swimming quickly, though at the cost of the protection once provided by the external shell. Such is often the case with evolution – it’s a trade-off. Note, however, that some groups of cephalopods, such as cuttlefish and squids, do make internal shells.

Soft Bodies

Another characteristic of mollusks is the soft bodies many of us have taught for all these years. But we need to modify our “soft body” narrative in order to be accurate. In particular, mollusks have “soft bodies broadly divided into three parts: a muscular foot, a visceral mass containing internal organs and a mantle that surrounds the visceral mass and contains glands that secrete the animal’s shell or spicules” (HMNH Exhibit).

Mollusks are a major animal phylum. As of February 2014, 96,920 living species are listed in the WoRMS database (WoRMS Editorial Board, 2014), but zoologists estimate that there actually may be >200,000 living. Of the 96,920 known species, 67,690 are gastropods (snails, slugs, nudibranchs, and limpets), making the latter by far the most numerous mollusks. Another 23,889 are bivalves (clams, mussels, oysters, scallops, etc.), while the remaining species are found in the other classes, especially 1939 species of cephalopods and 1998 species of Polyplacophora (chitons). Only the phylum Arthropoda has more species, a fact that is likely to surprise many.

Phylogenomic Methods Made the Construction of the Mollusk Tree Possible

Phylogenetic trees are constructed by sophisticated computer programs that look for what are known as “shared derived characters.” These are traits that evolved in the common ancestor of the clade, after the clade split off from its closest relatives and before the clade diversified. For example, fur and XY sex determination are shared derived characters of mammals. Now, imagine a group of organisms, like the mollusks, that originated hundreds of millions of years ago and whose major clades diverged from each other during a brief period shortly after the group originated. Since that time, many mutations will have occurred in all the clades that evolved from that common ancestor. Situations like this are common and make it extremely difficult to tease out the shared derived characters of each clade against the tremendous background of phylogenetic “noise” due to the subsequent hundreds of millions of years of mutation.

Until recently, phylogentic trees were based on anatomy or small numbers of genes, but now they are generated with phylogenomic methods, which involve using hundreds of genes from each species. The data used in the present mollusk study were the “transcriptome,” or mRNA from each species. This means that the DNA used was predominantly protein-coding DNA. These enormous data sets make it possible to separate signal from noise and obtain robust trees. Because of recent advances in DNA sequencing, it was possible for invertebrate zoologists to obtain many mollusk sequences, and sophisticated computers and programs have been developed to handle these large amounts of data. Forty-six species of mollusks and a total of 1185 genes were used in a study conducted by Smith et al. (2011). Fruit flies (D. melanogaster) were used as the outgroup. An outgroup is a group that is closely related to the clade you are studying but is not part of the clade. Using an outgroup enables you to find the root, or base, of the tree. This extensive data set was then analyzed using three different programs for phylogenetic inference. Because all three programs, which are based on different methods, gave very similar conclusions, the tree is considered robust (i.e., likely to be close to the true tree). Of course, in science, we rarely say that something is “true,” but in phylogenetics, we make a subtle distinction. There is a true tree. The true tree is what actually happened in the evolutionary history of the organisms being studied. However, we can never know the true tree with 100% certainty. What we can hope to do is construct a tree that we are confident is very close to the true tree. The resulting mollusk tree is shown in Figure 1.

For hundreds of years, ever since Linnaeus, taxonomists have been classifying the millions of species that share the planet with us. Until the 1990s, they used traditional methods such as morphology and embryology. In the 1990s, DNA sequencing became available. As a result, it was possible to both check the work done by the earlier taxonomists and explore evolutionary relationships between different groups of organisms. In terms of classifying species into different groups, the taxonomists got it right nearly all the time. It is a tribute to the intelligence, energy, and care that they brought to their work that so much of it has been confirmed by subsequent DNA sequencing. However, morphology, embryology, and even small DNA data sets alone were not enough to discern distant evolutionary relationships. Until around 2008, most phylogenetic trees were based on a fairly small number of genes, because only limited amounts of DNA sequence were available. Since then, DNA sequencing has become much faster and cheaper. As a result, we can construct trees, like this tree of mollusks, that show evolutionary relationships that were not discernible with older methods. This is an exciting advance. DNA sequencing is providing us with deep and irrefutable evidence for evolution. We should take advantage of this in our classrooms.

Acknowledgments

Thanks to Harvard University’s Life Sciences–HHMI Outreach Program. Their lectures, and the accompanying tour of the HMNH exhibit on mollusks, provided the basis for this paper. Special thanks to Gonzalo Giribet for leading the tour of the exhibit and for many helpful comments on the manuscript. Thanks to Nadav Kupiec for expert preparation of the artwork.

Sources

  • HMNH Exhibit: All information in this article attributed to “HMNH Exhibit” was taken from an exhibit on Mollusks at the Harvard Museum of Natural History in Cambridge, Massachusetts. The exhibit’s chief scientific advisor was Gonzalo Giribet, Professor of Organismic and Evolutionary Biology at Harvard University and the Alexander Agassiz Professor of Zoology and Curator of Invertebrates at Harvard’s Museum of Comparative Zoology.

  • WoRMS Editorial Board (2014): World Register of Marine Species (http://www.marinespecies.org). This is a very useful database of all known marine species, and others as well. It can be used in the classroom to show students how many of these categories (class, order, etc.) go back hundreds of years to discoverers such as Lamarck or even Linnaeus himself.

References

References
Kouchinsky, A., Bengston, S., Runnegar, B., Skovsted, C., Steiner, M. & Vendrasco, M. (2012). Chronology of early Cambrian biomineralization. Geological Magazine, 149, 221–251.
Smith, S.A., Wilson, N.G., Goetz, F.E., Feehery, C., Andrade, S.C., Rouse, G.W. & others (2011). Resolving the evolutionary relationships of molluscs with phylogenomic tools. Nature, 480, 364–367.