OBSER V ATIONS ON PELAGIC lVIOLLUSKS ASSOCIATED WITH THE SIPHONOPHORES VELELLA AND PHYSALIA1 FREDERICK M. BAYER Institute of Marine Science, University of Miami ABSTRACT Specimens of three species of violet snail. Ianthina janthina. I. pallida and I. prolongata, and of a nudibranch, Fiona pinnata, all of which feed upon siphonophores, were collected during strandings of Velella and studied in the aquarium. Observations were made upon float-building and feeding behavior of Ianthina, and upon feeding, growth and reproduction of Fiona. INTRODUCTION During the winter months in southern Florida, strong easterly breezes often blow ashore various members of the pelagic community of the Florida Current. The most conspicuous of these are three siphonophores, Physalia physalis (Linnaeus), the Portuguese man-of-war; Velella velella (Lin- naeus), the By-the-wind sailor or Purple sail; and the less abundant Porpita umbella (Muller). Accompanying these in greater or lesser numbers are several moJlusks that prey upon them. The abundance of these animals varies from year to year, depending upon weather conditions. During the winter of 1961-62, Physalia was present, as it usually is, but there were no Velella or Porpita; in 1962-63, moderate numbers of Vdella and fewer Porpita accompanied heavy strandings of Physalia. Collections made along the ocean beach of Biscayne Key, the shore of Virginia Key facing the open ocean through Bear Cut, the beach adjacent to the Marine Laboratory on Virginia Key, and the south side of Rickenbacker Causeway just west of Virginia Key provided specimens of three species of Ianthina and an eolid nudibranch, Fiona pinnata. Most of these were maintained alive in running-water aquaria for considerable periods of time, thus affording the biological observations reported herein. Subclass PROSOBRANCHIA Genus Ianthina Roding The many nominal species of this genus have been reduced by Laursen (1953) to five, of which three were taken at Miami during February and March of 1963. In view of the fact that the names used in the popular conchologicalliterature are not entirely correct, the three species discussed here are briefly described and illustrated in order to facilitate their identi- IContribution No. 488 from The Marine Laboratory, Institute of Marine Science, University of Miami, Florida. 19631 Bayer: Pelagic Mollusks 455 fication by those who may have an opportunity to conduct further studies of their biology. During the time that specimens of Ianthina were maintained alive in the laboratory, observations were made upon float-building, feeding and reproduction. Taxonomy.-Three of the five known species were collected in the vicinity of Miami during February and March of 1963; the other two have been obtained in past years but appear to be less abundant and therefore not so commonly washed ashore. The nomenclature employed here follows that used by Laursen in his DANA Report. FIGURE I. Ianthina janthina (Linnaeus). Virginia Key, opposite Bear Cut, Miami, Florida. X 1.5. Ianthina janthina (Linnaeus) This is the most abundant species around southern Florida. The trochoid shape of its shell is characteristic (Fig. 1). Because of its variable color and height of spire, it has received a number of names of which the most familiar are communis, fragilis, and violacea, but it is correctly identified under the name janthina in the most recent conchological works (Abbott, 1954; Keen, 1958; Kira, 1961, 1962; Morris, 1947; Warmke and Abbott, 1961). A few authors still employ incorrect nomenclature or recognize variant forms (Kira, 1961, 1962: balteata Reeve; Rippingale and McMi- chael, 1961: violacea Roding). Janthina pallida Thompson This species was second in abundance during February and March of 1963. It is often misidentified as I. globosa because of its globose shell 456 Bulletin of Marine Science of the Gulf and Caribbean [13(3) FIGURE2. Ianthina pallida Thompson. Key Biscayne, Florida. X 1.5. (Fig. 2). The shell characters used by Laursen (1953) to separate this species from I. umbilicata appear to be variable and must be corroborated by reference to the shape of the radular teeth. The color of the shell is pale violet; the outer lip merges with the columella in a smooth arc; the sinus is shallow and leaves behind inconspicuous traces of sculpture called a "keel" by Laursen. FIGURE3. Ianthina prolongata Blainville. Key Biscayne, Florida. X 1.5. Ianthina prolongata Blainville This species was found only once in 1963, during strong easterly winds on February 12. It often is identified as 1. globosa and not infrequently is confounded with I. pallida. The color is usually much deeper purple than in I. paUMa; the outer lip forms a right angle with the columella and does not describe a regular arc (Fig. 3). The sinus is rather shallow and leaves behind an inconspicuous keel-like trace on the body whorl only (Laursen, 19631 Bayer: Pelagic Mollusks 457 1953:15,28, fig. 30). This species is called I. globosa Swainson 1823 (not globosa Blainville 1825 = umbilicata d'Orbigny 1840) in some recent works (Morris; Kira; Abbott; Warmke and Abbott). OBSERVATIONS Feeding.-As reviewed by Laursen (1953: 14), not much is known of the food and feeding of Ianthina. Although it sometimes is stated that the food of violet snails consists entirely of Velella, Laursen discovered from an examination of stomach contents that they will eat anything available. Specimens kept in the laboratory tend to confirm that observation. When undisturbed, I. janthina hangs from its float with head and tentacles extended from the shell and the proboscis more or less everted. When an extended tentacle was very gently touched with the edge of a Velella, the snail extended its head to the maximum extent and fully everted the proboscis, exposing the radular teeth. When the tentacle was again touched with the Velella, the whole anterior part of the snail made a swift, darting movement toward the source of stimulus. This usually brought the tip of the extended proboscis in contact with the fleshy margin of the siphonophore, which was firmly grasped by the "lips" of the slit-like mouth. Radular action was evident immediately, and the snail chewed rapidly along the edge of the coelenterate, ingesting most of the fleshy part as it moved along. When a supply of Velella was not available, the lanthinas were offered Physalia as a substitute. Upon first contact with the Physalia, the snails performed feeding responses identical to those elicited by Velella. The proboscis was thrust deep into the fringe of zooids and showed no evidence of being affected by the powerful nematocysts. Two violet snails of moderate size virtually devoured a Physalia with a float about four inches long in less than one day, leaving only some remnants of the pneumato- phore and a small mass of zooids. On one occasion when Velella was not available, a specimen of I. janthina caught and ate a large I. pallida. Laursen observed the radular apparatus of I. janthina in the stomach of another of the same species, indicating that cannibalism occurs. Laursen (p. 8) mentions the numerous published references to the fact that Janthina emits a purple fluid when irritated, but attributes no function to it other than defense. Hardy (1956: 113) recounts the observations of P. David aboard the RRS DISCOVERYII, who stated that Ianthina released the purple dye periodically while browsing on the zooids of Velella and concluded that it may have an anesthetic function to block the action of nematocysts. The specimens of I. janthina observed in aquaria usually emitted purple fluid when disturbed, especially in the first few hours after collection. They did not release the dye when feeding upon Velella, but did so when feeding 458 Bulletin of Marine Science of the Gulf and Caribbean l13(3) upon Physalia. It could not be determined whether this was due to the disturbance associated with placing the food organism near the snails, or to the nature of the coelenterate itself. Several examples of I. pallida were observed while attached to Velella and browsing upon its zooids, but they were not seen to release the colored fluid. Furthermore, they did not do so when disturbed, or when attached to their own floats and feeding upon Velella in the manner of I. janthina. A number of very small individuals of I. pallida were observed feeding upon Velella and it is evident that the coelenterate is large enough to maintain the young mollusks for a substantial period. The smallest Ianthina observed was situated in the fleshy edge of a large Velella, in which it had eaten a hole larger than itself. Slightly larger individuals were found crawling freely about on the siphonophores. It seems likely that I. pallida, and possibly the other ovigerous species, may have no need for a float when young and do not construct one until the original host has been eaten. Possibly the veligers metamorphose upon contact with the host coelenterate and undergo a substantial part of post-larval development there. The observations of Ganapati and Rao (1960) concerning the feeding of Ianthina upon Porpita undoubtedly deal with one of the ovigerous species rather than with I. janthina as reported. The photograph illustrating their note shows a globose shell which could belong to either I. pallida or I. umbilicata, less probably to I. prolongata, but almost certainly not to I. janthina. Float-building.-Construction of the float has been described by several authors, summarized by Laursen (1953). Photographs of the steps in bubble formation by I. janthina are given herewith. From its usual position at the end of the float (Fig. 4, a), the propodium advances along the surface of the water in a broad, fan shape (Fig. 4, b). The edge of the propodium then curves downward hollowing the sole (Fig. 4, c), quickly followed by an in-rolling (Fig.