The Portuguese man-o’war, Physalia physalis, is one of around 175 species that make up the order siphonophores, in the class Hydrozoa. It is not, however, a typical member of the group. While Physalia drifts on the surface, blown by winds and carried by currents, most siphonophores are active swimmers and live beneath the surface. A few – the rhodaliid family – spend their lives tethered by their tentacles to the sea bottom. Because of their habitat and their fragility, siphonophores are extremely difficult to collect, and relatively little is known about them.

Siphonophores are colonies of polyps and medusae that function together as a single organism. Arguably they are the most complex of all colonial animals. In the case of most cnidarians, an individual consists of either a polyp or a medusa; here the individual consists of many individuals that have merged their individuality into something greater. At some point in the history of the Hydrozoa, and in what can only be conceived as a quantum jump, the genome ancestral to the siphonophores reconfigured itself to code for a structure that obliged individuals budding off from each other to transcend their individuality, and become so specialised (losing many functions in the process) that they could not survive except in cooperation. Each zooid – each polyp or medusa – had to depend on all the others for what it could not do itself. Some specialised in secreting gas for buoyancy, others specialised in propulsion, predation, digestion, reproduction. The organisation of multicellular life, where cells specialised and cooperated within a single organism, was taken to a new level, in a transition engineered, it seems, to convey an expressly theistic message: the hydrozoan genome was able to reprogram itself in this way because it had been endowed with the code necessary to effect the reprogramming. It ‘knew’ where it was going from the outset.

Siphonophores come in three basic designs: physonects, cystonects and calycophorans, and the designs consist of three basic modules: a float, a nectosome and a stem. Physonects have the full complement of modules. Cystonects lack the nectosome and thus have no capacity for independent movement. Calycophorans lack the float. In the phylogeny (genealogy of species) determined by Casey Dunn et al (2005) cystonects are the most ancient of the three siphonophore groups, with calycophorans nesting within physonects.

The Portuguese man-o’-war is a cystonect and consists of four main types of polyp. A single polyp, known as a pneumatophore, forms the large gas-filled float that sits on the ocean surface and acts as a sail. Other polyp types specialise as feeding tentacles (gastrozooids, of which there are three types), as defensive or prey-capturing tentacles (dactylozooids) and as reproductive organs (gonozooids). Hanging to a depth of up to 40 meters, the tentacles deliver a powerful sting.

Although rhodaliids, in stark contrast to Physalia, live on the ocean floor on hydrothermal vents a thousand of metres or more beneath the surface, they too have highly inflated pneumatophores. In most species, however, the float is small. Attached to the float is the nectosome, studded with medusae whose pulsations propel the colony forwards and backwards. Below that the stem, which all siphonophores have, bears the other specialist zooids, arranged in a very precise pattern that is unique to each species. As the photographs further down on this page illustrate, the diversity of form produced by these variations is immense.

Different authors have attempted to deduce phylogenies from the number of zooid types in various species, with opposite results, depending on whether they thought siphonophore evolution proceeded by a reduction or an increase in functional specialisation. The most recent reconstruction suggests that, if anything, the trend has been one of zooid loss:

Our findings indicate that there has been a complex history of functional specialization in siphonophores. … We find no bias in favor of the gain of zooid types, and some evidence of a bias towards the loss of functionally specialized zooids.

C W Dunn et al. (2005), p 933.

There are two types of zooid that appear to have emerged at a later evolutionary stage. One is the medusa zooids that populate the nectosome, since cystonects lack a nectosome. On the other hand, there are no siphonophores intermediate between cystonects and physonects, where we might see a structure on the way to becoming a nectosome; the nectosome, together with its zooids, is therefore likely to have emerged abruptly. The second zooid to have emerged later is the bract, of which there are four types. They are absent in the cystonects but present in physonects and calycophorans. Bract types both increase and decrease in number across the phylogeny. The other main zooid types studied were gastrozooids and palpons. Cystonects have three types of gastrozooid; most species make do with only one, the other two types being lost. Palpons are present in cystonects but absent in physonects and calycophorans. The fact that calycophorans nest within physonects indicates that the float polyp was also present in the most ancient siphonophores and subsequently lost.

The origin of the siphonophore body plan is thus unexplained by a scenario where novelties are acquired little by little through mutations that happen to confer an advantage, each mutation taking its place within a continually functional, complex system of interdependent parts. The body plan appears to have emerged from its hydrozoan forbear with close to its full comple- ment of zooids. Past and present, it is a wonderful example of design.