The Lancelet

M Dale Stokes & Nicholas D Holland. American Scientist. Volume 86, Issue 6. Nov/Dec 1998.

Lancelets have long occupied a key place in a famous paternity suit: the question of how the vertebrates evolved from the invertebrates. Resembling a small, colorless anchovy fillet without obvious eyes or lateral fins, the lancelet is classified as a cephalochordate, a member of the phylum Chordata. Vertebrates are also considered to be chordates because, along with the lancelet, they have a stiff cartilaginous rod, called a notochord, which supports the body at some stage during the animal’s development. Although the living lancelets are obviously not ancestors to modern vertebrates, some scientists have long believed that the immediate ancestor of the vertebrates resembled a lancelet-like creature that lived some 500 million years ago.

Because of its putative evolutionary importance the lancelet has played a key role in the history of biology, falling in and out of favor as a subject of study for well over a hundred years. The last decades of the l9th century saw the publication of hundreds of papers on lancelets, mostly on their embryonic and adult microscopic anatomy. This activity was driven partly by the passion of the German biologist Ernst Haeckel, who relentlessly promoted the phylogenetic importance of lancelets in his books, and partly by rapid improvements in histological techniques. Then, around the beginning of the 20th century, the rate of publication on lancelet biology declined sharply and did not pick up again until about 1960, when advances in electron microscopy stimulated a new surge of interest. A second surge of interest, which is still in full swing, was triggered about 10 years ago when lancelets began to be studied by the methods of molecular biology. The techniques of comparative molecular genetics, which we shall consider later, are now being used to address classical questions about the evolutionary origin of the vertebrates.

In spite of their importance, lancelets are inconspicuous creatures that usually live out of sight in shallow burrows beneath coastal marine sediments. Most biologists have never seen a living lancelet, and for this reason we also provide a brief review of their curious little lives. In particular, we have been studying the ecology of the Florida lancelet (Branchiostoma floridae), which lives in hip-deep water in Tampa Bay. This subtropical population is probably typical of the 29 known species of lancelets, most of which are distributed along the warm coastal waters of the world.

It would be an oversight to neglect the issue of the lancelet’s other name, amphioxus. Unfortunately the word amphioxus has no convenient plural form (amphioxi is pedantic, and amphioxuses is unpronounceable) and tends to be used as a valid generic name (which it isn’t). Here we avoid the term amphioxus, but with some regrets because it is a name that has been celebrated in song-It’s a Long Way from Amphioxus by Philip H. Pope-and story, The Angel at the Grave by Edith Wharton. At the climax of Wharton’s short story, an old manuscript on amphioxus saves a long-dead professor’s career from a fate worse than death: anonymity. Such drama is typical of the lancelet’s history

Raising Haeckel’s Hackles

The first scientists to study the lancelet weren’t always sure what to make of the creature. The 18th-century German biologist, Peter Pallas, described a European species (Branchiostoma lanceolatum), believing it to be a molluscan slug. In the 1830s, the Italian naturalist Gabriele Costa anticipated modern views of the lancelet when he classified it as a lower vertebrate akin to the jawless fishes. At about the same time the English biologist William Yarrell, working independently, also concluded that the lancelet was a fish.

Although the lancelet is indeed fishlike-having a pharynx perforated with gill slits, segmented axial muscles, a notochord, a dorsal hollow nerve cord and a post-anal tail-it was eventually transferred from the vertebrates to the cephalochordates in the early part of this century. The chief reason for the lancelet’s “demotion” was the lack of anything resembling a vertebral column. During the years between Charles Darwin’s publication of The Origin of Species in 1859 and the appearance of Haeckel’s General Morphology in 1866, lancelets were increasingly studied, but had not yet taken center stage. In General Morphology, Haeckel enunciated his law of recapitulation, predicting that the tree of life could be revealed by embryologic descriptions of the phyla. In reviewing the available information on animal phylogeny, however, Haeckel had to admit that “the origin of the vertebrates is still shrouded in deep darkness.” The very next year the shroud was lifted for Haeckel when he read two papers by Aleksandr Kovalevskii, a young Russian who had done pioneer work on the embryology of lancelets and tunicates (both are invertebrate chordates). Kovalevsk showed that the early development of these animals has invertebrate characteristics (a hollow blastula invaginates, forming a gastrula of the sea-urchin sort) but that later development results in the production of vertebrate-like features. Haeckel interpreted Kovalevskii’s discoveries as revealing an evolutionary link between the lower invertebrates and the vertebrates (where he had placed the lancelet).

Haeckel believed that the lancelet was the “last of the Mohicans,” representing a lower class of vertebrate that had largely vanished from the fossil record owing to the absence of a solid skeleton. His high esteem for the lancelet was evident in the book The Evolution of Man (1874), wherein he admonished his readers to regard lancets “with special veneration as that animal which alone of all extant animals can enable us to form an approximate conception of our earliest Silurian vertebrate ancestors.” The lancelet was now central to Haeckel’s theory that “ontogeny recapitulates phylogeny.”

In a reaction to Haeckel’s enthusiasm, a church newspaper complained that reverence for the lancelet caused the “dignity of humanity to be trodden underfoot, and the divine rational conscience of man to be grievously hurt.” Haeckel responded that the lancelet “has a better right to be an object of profoundest admiration and of devoutest reverence than any one in that worthless rabble of so-called ‘saints’ in whose honour our `civilized and enlightened’ cultural nations erect temples and decree processions.”

Haeckel’s devotion to the lancelet brought him into conflict not only with the church but also with two German zoologists, Karl Semper and Anton Dohrn. Both men proposed a rival theory of origins in which vertebrates were derived from a segmented wormlike creature. Semper believed that the lancelet was some sort of protomollusc, and Dohrn believed it was derived from the early vertebrates by degeneration (having lost its vertebral column in evolution). Feelings ran so high that Haeckel published an attack on Semper accusing him of a “defective education and insufficient acquaintance with the whole subject of zoology.” For his part, Dohrn got back at Haeckel by referring to the lancelet as “degraded and condemned to a miserable existence in the sand, cast there from his usurped ancestral throne.” The three men never resolved their differences.

A Modern Phylogenetic View

Perhaps it’s not surprising that the scientists of the 19th century could not agree on the lancelet’s evolutionary relationships. Limited to light microscopes and simple tools for phyletic analysis, these men were simply not properly equipped for the task. Even so, they came close to anticipating some of the modern views on the lancelet and the origin of the vertebrates. Molecular biological studies have now shown that the lancelet is the closest living invertebrate relative of the vertebrates. A phylogenetic tree based on the DNA sequence of the large ribosomal subunits in various animal subphyla supports the view that the cephalochordates are the sister group of the vertebrates (Wada and Satoh 1994). This is also consistent with views based on classical morphological studies of the lancelet indicating that these animals share a number of features common to vertebrates (a dorsal hollow nerve cord, segmented muscles, a digestive system that ends at the anus and a circulatory system that passes through the gills). Recent work also suggests that the lancelet nephridia are simple versions of the vertebrate kidney

Other clues to early vertebrate evolution have recently come from studies of developmental genes, which are remarkably conserved throughout the animal kingdom. This conservation was originally discovered for the homeotic genes, which define the identity of the regions along the anteroposterior axis of animals. In lancelets and vertebrates the homeotic genes are called Hox genes. In the lancelet the dozen or so Hox genes are all clustered on a single chromosome (Garcia-Fernandez and Holland 1994). In contrast, the Hox genes of vertebrates are arranged as four separate clusters, each located on a different chromosome. In evolutionary terms, it is easy to explain how a single cluster of Hox genes might become four clusters as a result of chromosome duplication events, but it is difficult to imagine a simple mechanism for reducing four Hox clusters to a single cluster. In this light, it seems extremely unlikely that the ancestral vertebrates degenerated to give rise to the cephalochordates. On the contrary, modem lancelets and vertebrates apparently had a cephalochordate-like ancestor in common.

Such an ancestor may have resembled some recently discovered Cambrian fossils. The Burgess Shale fossil Pikaia, originally described as a worm, has since been reinterpreted as a cephalochordate. Its exquisitely preserved soft parts are lancelet-like, although, disconcertingly, no obvious gill slits appear to be present. Pikaia has been widely accepted as a cephalochordate, although a few experts still have their doubts. The Chinese fossils Cathaymyrus and Yunnanozoon in the Chengjiang Cambrian deposits have also been described as cephalochordates, although some paleontologists still favor alternative interpretations.

A further understanding of the cephalochordate-like ancestor may be gleaned from developmental genes by unraveling possible homologies between the body parts of lancelets and vertebrates. (According to one current conception of homology, body parts in different animals are homologous if they are derived from an equivalent part in the animals’ common ancestor.) Two factors come into play here. First, developmental genes typically have conserved base sequences that can be used to identify closely related genes in different species. Second, homologous developmental genes are often transcribed in similar places and at similar embryonic stages in different kinds of animals. This information is especially useful when the overall body plans of the animals under comparison are relatively similar (as they are in lancelets and vertebrates) and when used in conjunction with detailed knowledge of embryology and morphology.

Consider the question of whether the lancelet has a brain. The central nervous system of the lancelet is a dorsal, hollow nerve cord with a small anterior region (characterized by a somewhat dilated neural canal) called the cerebral vesicle. The lancelet neural tube is not divided into obvious regions by constrictions such as those delimiting the forebrain, midbrain and hindbrain of vertebrates. As a result there have been numerous ideas about what, if anything, constitutes the lancelet brain. There have been “no-brain theories,” “little-brain theories” and “big-brain theories.”

Three-dimensional reconstructions of the cerebral vesicle in a larval lancelet suggest that this part of the nerve cord corresponds to the diencephalic part of the vertebrate forebrain (which includes the thalamus and the hypothalamus), but lacks a homologue of the vertebrate telencephalic forebrain (Lacalli, Holland and West 1994). There is also a relatively extensive hindbrain, as well as less compelling evidence that a posterior part of the cerebral vesicle corresponds to the vertebrate midbrain.

In vertebrates the neural expression of the genes dlx and otx is limited to the developing forebrain and midbrain, whereas the homologous genes of lancelets are expressed in the developing cerebral vesicle of the dorsal nerve cord. The gene-expression evidence supports the homologies derived from morphological studies and suggests a diencephalic forebrain is present in lancelets. Similar comparisons of Hox gene expression in the developing central nervous system of lancelets and vertebrates indicate that lancelets also have a relatively large hindbrain (Holland, Holland, Williams and Holland 1994). All together, these homologies support the idea that the cephalochordate-like ancestor of the vertebrates had a relatively extensive brain, which at least included a diencephalon and a hindbrain.

Developmental genes can also be used to compare body parts between animals with markedly divergent body plans. Although it may be difficult to establish convincing homologies, owing to the increased chance of mistaking convergently evolved features for homologies, such observations can suggest new hypotheses that can then be tested by further study. A good example of a comparison spanning a wide stretch of the animal kingdom is the discovery that both the developing lancelet and the developing fruit fly express the gene called engrailed in stripes along the anteroposterior axis (Holland, Kene, Williams and Holland 1997). Engrailed is involved in establishing and maintaining the segments along the length of an animal’s body. Along with recent genetic information suggesting a reversal of the dorsoventral axis between fruit flies and vertebrates (De Robertis and Sasai 1996), a common role of engrailed would be consistent with the idea that vertebrates are derived from an annelid-like or arthropod-like body plan.

This notion revives some 19th-century views, including Semper and Dohrn’s theory that annelid-like creatures gave rise to the vertebrates. Although Semper and Dohrn did not think cephalochordates played any part in the origin of the vertebrates, the molecular genetic data indicate that ancestral cephalochordates could very well have been evolutionary intermediates. Interestingly, the German biologist Theodor Boveri reached the latter conclusion in 1890, when he discovered that lancelet kidney tubules were structurally intermediate between annelid nephridia and the vertebrate pronephros. In the same paper, Boveri added that “The lancelet kidney should be welcome to defenders of the annelid theory” In effect, Boveri was suggesting to his good friend Dohrn that vertebrates evolved from lancelet-like creatures that in turn had evolved from annelid-like creatures. Regrettably, Dohrn, the great defender of the annelid theory, never did modify his views. To do so would have meant conceding a point to his archenemy Haeckel by admitting that lancelets were not degenerate vertebrates.

A Lancelet’s Life-style

In some respects the modern lancelet may serve as a useful stand-in for the immediate invertebrate ancestor of the vertebrates. Of course, some odd anatomical features of contemporary lancelets (such as the muscle cells of the notochord) probably represent independent developments within the cephalochordate lineage. Yet the basic apparatus of the vertebrate ancestor may have been quite similar, so it stands to reason that some aspects of the lancelet’s life-style may shed light on the habits of the creature that gave rise to the first vertebrates about 500 million years ago. The modern lancelet is a creature of the seas. An adult lancelet may be found anywhere between the shore and the edge of the continental shelf, and this distribution seems to vary among species. The adults spend most of their time in shallow burrows filter feeding on small particles, especially phytoplankton. Although most textbooks claim that lancelets live only in relatively coarse sand, the Florida lancelet can flourish in sand as fine as powdered sugar. Undisturbed lancelets tend to stay in the same burrow for weeks at a time.

Despite their relatively sessile lifestyle, adult lancelets are capable of swimming through the water by rapid, eel-like undulations powered by their extensive trunk musculature. They can attain speeds up to 20 body lengths per second. There is some possibility that they use their swimming speed to escape predators. In Tampa Bay, the most conspicuous predators on the lancelets are stingrays that ingest both sand and lancelets. The sand is blown out through the gill slits while the larger lancelets are retained in the pharynx by the gill rakers.

Lancelets also swim while spawning. Each year, the breeding season of the Florida lancelet lasts from late May through early September, a period when water temperatures are relatively warm. Laboratory observations of the Chinese lancelet show that the males and then the females leave their burrows and swim through the water, shedding gametes. After a few minutes the animals swim down back to the bottom and burrow into the sand again. After spawning the Florida lancelet’s gonads can refill with gametes in a little over a week, allowing the animals to spawn several times during a season.

After fertilization, development proceeds rapidly. The morning after the adults spawn, embryos covered with cilia hatch and swim through the water like large ciliate protozoans. A few hours later, as the ciliated embryos are elongating, they also commence brief, intermittent episodes of swimming by muscular undulation. After about a day and a half of development, the mouth opens, an event that is arbitrarily taken to divide the embryonic and larval stages. The larvae are conspicuously asymmetric in the left-right axis. For example, the mouth opens on the left side of the head, and what later prove to be the right gill slits develop from tissues derived from the left side of the body. It is generally believed that this asymmetry is an adaptation ensuring a large enough mouth for effective suspension feeding.

The larvae spend most of their time using a wavelike beating of their epidermal cilia to hover in the water with the head oriented approximately upward. Pre-metamorphic larval life lasts three to four weeks and is followed by a metamorphic period, which lasts for just under a week. Three important morphological changes occur during metamorphosis. First, the highly asymmetric features of the larva become almost bilaterally symmetrical. Some asymmetries remain, but these are subtle-for example, the anus is displaced slightly to the left of the midline. Second, flanges of body wall grow over the gill slits on either side and fuse ventrally to form an atrial chamber. Third, the cilia become so reduced that the animals are no longer capable of ciliary hovering and settle to the bottom as juveniles. Juveniles and adults must rely on their muscular undulations to swim, which is about 10 times more expensive metabolically than ciliary hovering.

At least some lancelet species can evidently produce two alternative kinds of larvae. One kind settles at a respectably small size, but the other seems to remain swimming in the plankton for extended periods of time and may grow to a centimeter in length without entering metamorphosis. Such oversized larvae were originally described as distinct species of cephalochordates belonging to a separate genus, Amphioxides. Occasionally, they were even found with a row of immature gonads along the right side of the body.

One extreme view held that Amphioxides represented a permanently planktonic, neotenic (juvenile-like adult) cephalochordate that was supposedly much like the ancestor of the vertebrates. The more likely explanation is that Amphioxides represents an alternative larval type produced by some benthic lancelets. If this is so, then current methods of molecular taxonomy should permit the assignment of each Amphioxides “species” to a known species of lancelet.


Molecular biology has now shown that Haeckel was correct in placing a lancelet-like creature at the base of the vertebrates and in stressing that, of all living animals, the lancelets were the most likely to provide insights into the origin and earliest evolution of the vertebrates. In the end it may turn out that Haeckel and his enemies (Dohrn and Semper) each supplied part of the right answer about the origin of the vertebrates-if additional molecular data support the notion that a lancelet-like creature was an evolutionary intermediate between an annelidlike precursor and the first vertebrate. Thus lancelets seem to have come full circle in the history of biology. After years of neglect, they have now regained the prominence that they enjoyed in the last decades of the l9th century when they were the centerpiece of Haeckel’s recapitulation theory.