Cnidarians are the phylum of marine predators that includes anemones, corals, jellyfish, hydra and siphonophores such as the Portuguese man-o’-war. You will come across them – or they will come across you – on almost any seaside holiday, and once encountered they are not easily forgotten. Corals are known for their extravagant colours and extravagant shapes. Coral reefs constitute the largest structures made by any group of animals and harbour the most diverse concentrations of marine life on the planet. Jellyfish too have their superlatives. Lion’s mane jellyfish reach widths of over 2 metres, with tentacles extending downwards for over 30 metres, while some siphonophores exceed 40 metres in length. The stings of a box jellyfish can kill a human being in four minutes.

A phylum is the highest level of classification after kingdom and usually – though devotees of the theory of evolution will not agree – encompasses a variety of unrelated groups. Be that as it may, Cnidaria (we don’t pronounce the ‘c’) is a phylum whose members do seem to be genealogically related. There are around 10,000 species today, representing diversification from a common ancestor on a very large scale. A diagram such as the one below can give but the barest summary. As so often, the highest-level group divided early on in history into two radically different subgroups, the Anthozoa (literally ‘flower animals’) and the Medusozoa, the first comprising corals and sea anemones, the second, jellyfish and hydras. Possibly the extinct, difficult-to-classify conulariids also belong to this phylum.

As observed today, the cnidarian body plan comes in two forms, polyp and medusa. Anthozoa take the stationary polyp form, Medusozoa the planktonic bell-like form, with the life cycle of most Medusozoa alternating between the two. First the adult jellyfish produces eggs and sperm, then the fertilised egg develops into a larva, the larva attaches itself to a hard surface and grows into a polyp, and the polyp buds off larvae which grow into jellyfish. As if to obscure relationships further, coral polyps organise themselves into colonies, for it is in concert that they marvellously assume the shape of fingers, mushrooms, cups, stag’s horns, brains, and so on, collectively looking nothing like a polyp. They secrete calcium carbonate to maintain such shapes, staying retracted within the skeleton during the day and feeding more openly at night. A few Medusozoa, such as Millepora, or fire coral, also develop calcareous skeletons at the polyp stage. Others, such as Portuguese men-o’-war, are free-floating colonies of polyps and medusae that look like a single medusa.

Polyps and medusae have the same basic body plan, where two tissue layers sandwich a jelly-like layer called a mesoglea and enclose a central digestive cavity. An opening at the end of the cavity ingests food and expels digestive waste, served by a fringing ring of tentacles. Most distinctive are the stinging cells on each tentacle, called cnidocytes (‘nettle cells’, also called nematocytes). The cnidocyte contains a coiled tubular filament, which shoots out when triggered by anything brushing against it and injects the intruder with a potent toxin. In addition to the great variety of toxins, the cnidocytes themselves are very diverse, numbering over thirty different types. Some are barbed harpoon-like structures, others stick to the victim, some wrap around it like a lasso. Described by Richard Dawkins as ‘probably the most complicated piece of apparatus inside any cell anywhere in the animal or plant kingdoms’, they are a classic example of design. Dawkins is also right in saying that cnidocytes are a rare example of an utterly unambiguous, diagnostic characteristic of a major animal group: all organisms with the cnidarian body plan have cnidocytes, and only cnidarians have them. Thus there is evidence of both creation and evolution.

Stinging cells aside, cnidarians tend to be regarded as comparatively simple, brainless animals and most biologists place them near the root of the animal tree of life, above sponges but below vertebrates and all other invertebrates. In absolute terms, however, they are far from simple, and part of their complexity can consist in the way polyps and medusae of various specialised types co-operate as superorganisms, as in the case of the siphonophores, or in the way cnidarians reproduce. For instance, the sea anemone Nematostella can come into being by four different routes:

• by sexual reproduction, where an egg develops into a larva which develops into a polyp
• by regeneration after a body is bisected
• by fission following pinching of the physa (a bulb-like organ at the base, enabling the anemone to burrow into sediment), and
• by fission from the end of the physa, where a new polyp produces mouth and tentacles first.

Any one of these reproductive processes would be enough to inspire awe, but what about the availability of all four in the same organism? The knowledge that the genetic program controls such phenomena in no way detracts from the sense that we are witnessing something amazing and miraculous. The distinction between natural and supernatural blurs.

These days complexity tends to be assessed by studying the genes rather than the morphology. When cnidarian genomes were analysed the results surprised researchers, for they were not what naturalistic assumptions led them to expect. The same signalling pathways and transcription factors involved in the development of higher animals turned out to be present and active in these lower ones. Writing in 2005, Ulrich Technau and colleagues concluded:

The resulting data set is much more complex than might be assumed, based on morphology, and implies that much of the genetic complexity commonly assumed to have arisen much later in animal evolution is actually ancestral. .. In many respects, the complexity of the anthozoan gene set does not differ substantially from that of vertebrates and frequently exceeds that of the model invertebrates Drosophila [fruitflies] and Caenorhabditis [nematode worms].

Indeed, animals higher up the tree appeared to have lost many of the genes present in the cnidarians. ‘It stands to reason,’ Elizabeth Pennisi wrote in the journal Science, ‘that the more genetically complex an organism is, the loftier its place on the evolutionary tree.’ Anthozoans having branched off from the most primitive cnidarian before medusozoans, they were expected to be amongst the simplest of animals, yet they proved not to be. “The genomic complexity of … cnidarians is much greater than expected,” commented John Finnerty, an evolutionary biologist at Boston University. “There is no simple relationship between the numbers of genes an animal possesses and its complexity at the morphological level.”

More surprises were to come. In 2007 a team led by Nicholas Putnam and Daniel Rokhsar reported the results of sequencing Nematostella’s entire genome. One surprise was its size: an estimated 18,000 protein-coding genes, comparable to the figures for other animals (human beings, at the latest count, have 20,488). Another was finding 1,292 families of genes in vertebrates such as pufferfish, frog and human as well as Nematostella that were not present in either the nematode worm or the fruitfly – despite worms and flies having supposedly evolved from animals that included these gene families and having evolved along separate paths, so that one would not have expected them both to have lost the same genes. Whole blocks of genes, in the very same order, were found in the anemone and human genomes. In short, the picture was one of ‘extensive conservation in gene content, structure, and organization between Nematostella and vertebrates’.

The sea anemone genome is complex, with a gene repertoire, exon-intron structure, and large-scale gene linkage more similar to vertebrates than to flies or nematodes, implying that the genome of the eumetazoan [all-animal] ancestor was similarly complex.

A more resounding refutation of the doctrine that vertebrates and invertebrates have a common ancestor could hardly be imagined. The implication was never entertained, however. The only interpretation on offer was that genes were ‘lost’ by some of the higher animals. It was merely ‘surprising’ that the genome of the common ancestor of cnidarians and all animals after them was so complex.

Relationship has to be established first – otherwise one is simply begging the question. Within the cnidarians it may be appropriate to talk about gene loss, because here the inference of a genealogically united group is strong. And gene loss is what one finds. At least two sequenced species, Acropora and Hydra, appear to have undergone substantial gene loss in the course of their evolution (Kortschak et al 2003, Miller et al 2007). At the morphological level the loss of several anatomical features, especially amongst the hydrozoans, is equally striking (see figure). Hydrozoans are characterised, for example, by the absence of gastric filaments and certain intra-mesogeal muscles. There were also gains in complex features, of which the most remarkable example is the multiple eyes of the box jellyfish. Also remarkable is the re-emergence of statocysts (a balancing organ) in the Leptothecata after their ancestors had lost them. Overall, it seems probable that the first cnidarian was neither more nor less complex than its modern-day descendants.

Much the same observations can be made of the other animal phyla in the fossil record, many of which appear at the same time as the cnidarians, others later. When they first appear, they are already complex, as was known even before Darwin. The hope has been that modern genetic analyses would do something to counterbalance the negative evidence and show that, deep down, the oldest animal phyla are much simpler than they look. Researchers have been disappointed.

It will be interesting to see what the project to sequence the sponge genome reveals. In the atheistic tree of life sponges are the phylum immediately below cnidarians, and the prediction – the expectation at least, though even that must be tempered by experience – is that sponges will be genetically simple. The signs are not so far promising.