to several gonophores (Pugh 1999). Overlying the cormidium is a bract that has a characteristic shape and is often useful in identifying species (Figure 3.31c). Cormidia within the physonects are larger and more complex, usually possessing several leaf‐like bracts, a gastrozooid, dactylozooids, and clusters of gonophores. The cystonects have simple cormidia consisting of a gastrozooid and clusters of gonophores radiating from a gonodendron or stalk. The cystonects have no bracts.
In the calycophorans, the terminal mature cormidia on the stem break away to form free‐swimming “eudoxids.” Until their true origins were resolved, eudoxids were thought to be a distinct group within the siphonophores (Hyman 1940). Limited propulsion within the eudoxids is achieved by the gastrozooids pinch‐hitting as nectophores. Shedding of cormidia for a brief free‐swimming existence is limited to the Calycophorae; neither the physonects nor the cystonects produce eudoxids.
Figure 3.31 Siphonophore cormidia. (a) Calycophoran Nectocarmen antonioi. (b) Diphyid Muggiaea sp. (c) Enlarged cormidium of Muggiaea.
Sources: (a) Adapted from Alvarino (1983), figure 1 (p. 342); (b) Kaestner (1967), figure 4‐34 (p. 74); (c) Adapted from Hyman (1940), figure 150 (p. 174).
Life Histories
Development proceeds from a planula larva and differs between groups. In the physonects, the planula becomes a siphonula larva as summarized in Figure 3.32 from Mackie (1986). The two budding zones produce the nectophores and the gastrozooids, gonophores, and bracts as the stem elongates during development. Development in the Calycophorae is somewhat different, with the initial production of a larval nectophore followed by development of the stem and gastrozooid. The larval nectophore is retained in some taxa (e.g. Sphaeronectes) and lost in others (e.g. Hippopodius). Early development in the cystonects is currently undescribed but defined budding zones beneath the float have been identified in both Physalia and Rhyzophysa.
The Siphonophore Conundrum
Is a siphonophore an individual or a colony? Siphonophores develop from a single egg, differentiating during development into a complex organism that is made up of parts that at some point in the distant past were probably free‐living polyps and medusae. How they came together and whether they are best considered as a colony or individual are questions that have piqued the interest of such luminaries as Ernst Haeckel, T. H. Huxley, E. O. Wilson, and S. J. Gould (cited in Mackie et al. 1987). However, for ease of understanding based on a lifetime of perspective, the observations of G. O. Mackie himself are the most cogent. In Mackie et al. (1987), he observed that “siphonophores … are linear in form with little branching, and are polarized, with a distinct anterior end. They are also bilaterally symmetrical. They grow by addition of modules at localized growth zones. The result is a high degree of determinacy of form.” Earlier (Mackie 1963), he offered this elegant observation.
Figure 3.32 Comparison of the life cycle of an agalmid siphonophore with that of an athecate hydroid.
Source: Reproduced from G. O. Mackie (1986), From Aggregates to Integrates: Physiological Aspects of Modularity in Colonial Animals, Philosophical Transactions of the Royal Society of London Series B, Biological Sciences, 1986, Vol 313, No. 1159, page 179, by permission of the Royal Society.
They have developed colonialism to the point where it has provided them with a means of escaping the diploblastic body plan. The higher animals escaped these limitations by becoming triploblastic using the new layer, the mesoderm, to form organs. The siphonophores have reached the organ grade of construction by a different method – that of converting whole individuals into organs.
It is important to understand the concept that siphonophores function as highly coordinated individual organisms and not as a loosely collected gaggle of different zooids. Natural selection acts on the whole individual.
Figure 3.33 Cyclic fishing behavior and optimization of feeding space. (a) The Calycophoran Muggiaea atlantica begins a feeding cycle by swimming upward with contracted stem and tentacles; (b) it releases the stem and curtain of tentacles and swims in a circle; (c) swimming stops, with fully extended stem and tentacles; the stem then sinks under its own weight creating a corkscrew shape that slowly moves downward through the water column until the stem is vertical again, where it contracts, and the cycle begins again. This deployment cycle results in the fishing array moving through a maximum volume of water.
Feeding
Fishing Behavior
Unlike many of the medusae, siphonophores feed while drifting motionless, creating a “curtain of death” with their extended tentacles. Prey must blunder into the “curtain” for siphonophores to feed. Most calycophorans and physonects spread their tentacles for short bouts of swimming (Figure 3.33) followed by longer periods of drift, usually several minutes in larger species. The frequency of swimming episodes for the purpose of tentacle‐reset scales roughly inversely with size, with the larger physonects swimming every few minutes and smaller calycophorans every minute or less. Species with large siphosomes, e.g. Apolemia at 20 m or greater, assume a horizontal position with the tentacles hanging down in a curtain. Swimming is less important for spreading the tentacles in the longer species (Mackie et al. 1987). Changes in the tentacle net happen gradually as structures sort themselves by density, sinking or rising with time. In some physonects, prey are tempted by periodic contractions of tentilla, mimicking the motion of zooplankters. In some cases, structures on the tentacle such as nematocyst batteries resemble “prey of the prey” and bring unwary prey species within the kill zone.
Digestion
When prey contacts a tentacle, only the tentacles attached to the prey contract to bring the prey to the mouth of the gastrozooid. The rest of the colony continues fishing. Gastrozooids work individually or in teams. If the prey is too large to be completely engulfed by a single zooid (e.g. a fish in the case of Physalia), more zooids will be recruited to finish the job. Presumably, the additional zooids are joining the job in response to chemical signals released by the captured prey. Digestion of small prey items such as copepods can be quite rapid: 1–2 minutes (Mackie et al. 1987).
As is the case with other cnidarian groups, digestion of prey is achieved using both extracellular and intracellular mechanisms. It takes place primarily in the gastrozooid, where breakdown of food is accompanied by rhythmic contractions. Particles created during the process of extracellular digestion are taken up via phagocytosis by the endodermal cells of the gastrozooid or, in some cases, can be passed along to palpons via the contractions noted above and taken up there. Further processing takes place in the endodermal cells of both structures. Eventually the
