**4. Discussion**

It is generally accepted that the morphology of the gastropod protoconch determines the mode of development and the duration of larval stage in the ontogenesis [3, 6, 11]. We are of the opinion that such a change in the larval morphology in members of different genetically isolated molluscan lower taxa must have taken place independently and must be due to a simple genetic change involved in the cell's contingency. Otherwise, it would be rather improbable for the same, yet complex, genetic changes to have taken place simultaneously in different individuals belonging to different taxa. Point mutations or even reverse mutations could easily occur leading also to switches. Such multiple switches in the mode of protoconch development are shown to have occurred in the evolutionary history of the Indo-Pacific *Kermia*–*Pseudodaphnella* complex and the diversity of protoconch morphologies exhibited in this group points to a high developmental and evolutionary plasticity [10].

Both, planktotrophy and lecithotrophy commence with an initial short lecithotrophic stage equipped with the larval shell I. The initial difference between the two types of larval shells I lies in their size which, in turn, is associated with the size of the egg. Larval shell I, leading to multispiral protoconch II, is smaller in width and accounts for the planktotrophic developmental mode, whereas larval shell I, directly linked to lecithotrophy (paucispiral), is larger. In the case of the Mediterranean pair of the so-called sibling species, *R. spadiana* and *R. contigua* (**Figure 1**) were examined in this work; the first bears a larger lecithotrophic larval shell (protoconch I), while the second bears a smaller protoconch I (of its planktotrophic larval shell) of a diameter that corresponds to 3 additional cell cycles (**Table 1**) of the stem cells. In the case of *R. smriglioi* and *R. lineolata* (**Figure 2**), the size of the protoconch I of the latter corresponds to 2 additional cell cycles (**Table 1**), while in that of *R. philberti* and *R. locardi* (**Figure 3**) also to 2 (**Table 1**). This kind of difference in the size of protoconch I could be attributed simply to a single gene intervening in the control of the germ line cell cycle in the gonads, functioning either in favor of a few additional mitoses and thus to a larger number and smaller germ cells and eggs (reduced parental investment per offspring) and eventually smaller in diameter (by a factor of approximately 0.7937 per cell cycle) stage I embryos, or in favor of no further mitoses and thus to a smaller number of larger germ cells, eggs and embryos (increased parental investment per offspring). At the same time, the reproductive output per adult biomass (total energy expenditure or reproductive effort) in those alternative gene actions would not constrain any of the two types of larval development; otherwise, all mollusks should tend to have the same mode of development if that mode gave higher returns per effort [50].

almost 1.35 irregularly cancellated and convex whorl, the first of which is decorated only with fine spiral striae. The teleoconch consists of five convex whorls separated by a deep suture. The body whorl occupies almost 65% of the total length and bears 16–17 orthocline axial ribs with interspaces approximately two times wider than the ribs themselves and 19–20 spiral cords thinner than the ribs, six of which are situated above the aperture and the rest 14 below the aperture. The spiral cords in their intersections with the axial ribs form erasures in the form of small elongated rectangular tubercles. The tubercles on the first two adapical cords are spiky and close to each other. The shell's inner wall viewed through the aperture exhibits a transparency. A narrow ramp is evident immediately below the suture, formed by the vestigial first two spiral cords and the much prominent and spiky third cord. The aperture occupies approximately 45% of the shells length and exhibits a smooth and slightly sinuous columella in its lower part, angled at its upper part. The anterior siphonal canal is short and wide, while the posterior one is deep and narrow. The outer lip bears 11 strong teeth with the first one delimiting the posterior canal and the last the anterior. The shells are of yellow-beige background color, while the tubercles and some irregularly situated areas or isolated tubercles are of lighter color. The body whorl usually bears at its middle a lighter color spiral band as a prolongation of the suture.

*Similar species*: *R*. *spadiana* is different from: *R. alternans* (Monterosato, 1884), in which the latter is slender and with a different color pattern; *R. atropurpurea* in its color which is light redbrown instead of purple-brown in *R. atropurpurea* and in its more inflated spire; *R. contigua*, in which the latter bears a multispiral protoconch, is larger, of lighter color, and more robust; *R. densa* (Monterosato, 1884) in the color and the less dense sculpture; *R. lineolata* (Bucquoy, Dautzenberg and Dollfus, 1883), in which *R. lineolata* has a less robust shell, a more narrow aperture, and a more narrow subsutural ramp; *R. oblonga* (Jeffreys, 1867) in its wider aperture

It is generally accepted that the morphology of the gastropod protoconch determines the mode of development and the duration of larval stage in the ontogenesis [3, 6, 11]. We are of the opinion that such a change in the larval morphology in members of different genetically isolated molluscan lower taxa must have taken place independently and must be due to a simple genetic change involved in the cell's contingency. Otherwise, it would be rather improbable for the same, yet complex, genetic changes to have taken place simultaneously in different individuals belonging to different taxa. Point mutations or even reverse mutations could easily occur leading also to switches. Such multiple switches in the mode of protoconch development are shown to have occurred in the evolutionary history of the Indo-Pacific *Kermia*–*Pseudodaphnella* complex and the diversity of protoconch morphologies exhibited in

this group points to a high developmental and evolutionary plasticity [10].

and the different color pattern [17, 18].

34 Organismal and Molecular Malacology

**4. Discussion**

*Habitat and distribution*: Whole Mediterranean Sea [17]. *Status*: Uncommon [17]. First record for the Greek seas. Larger embryos (lecithotrophic) leave the water column early, while smaller ones (planktotrophic) later and thus continue with the development of protoconch II. In spite that time latency, both types of larvae eventually lose their buoyancy and sink. There seems to be no


**Table 1.** Expected and measured maximum diameter of protoconch I in planktotrophic sibling species after hypothetical additional cell cycles. Bold characters indicate concurrence between expected and measured protoconch I maximum diameter.

reason why, at least some of those two different types of larvae, not to find themselves in the same locality and, as the initial mutation responsible for the differentiation of the larval mode of life has not led to the establishment of a genetic barrier, when maturation is reached, to interbreed. There are no publications on interbreeding of Mediterranean 'sibling' species so that one can draw conclusions on the existence of a genetic barrier and thus to a confirmation that loss of either planktotrophy or lecithotrophy in the past has eventually led to speciation.

If there is no genetic barrier, then a rising question is associated with the type of inheritance imposed by the initial mutation on that gene controlling the germ cell cycle prior to meiosis. If it displayed a Mendelian inheritance, we would expect also the production of heterozygotes exhibiting a kind of semi-planktotrophic mode of life of shorter duration and presumably protoconches with fewer whorls. In conclusion, someone would expect to find in the same morphological species all three types of protoconches, e.g., paucispiral, multispiral, and intermediate. As this is not the case, at least in Mediterranean *Raphitoma* species [17, 18], we are inclined to propose that the mutated gene cooperates in conjunction with other genes and environmental factors in a discontinuous multifactorial inheritance in which environmental or even population factors also effect a threshold. *Ceteris paribus*, below that threshold, the animals would produce fewer and larger germ cells giving rise through meiosis to fewer and larger eggs that after fertilization produce large lecithotrophic embryos with large paucispiral protoconch I, while, above that threshold, more and smaller germ cells would be produced leading after meiosis to smaller eggs which eventually give rise to planktotrophic larvae with small protoconch I and large multispiral protoconch II. Those two possibilities would jointly constitute protoconchrelated poecilogony, a phenomenon already known in some sacoglossan mollusks [51].

In support to the above hypothesis, there are at least 10 pairs of similar, most probably closely related species of *Raphitoma*, one of which bears planktotrophic protoconch and the other lecithotrophic, often sampled in the same localities [17, 18]. That implies the same mode of adulthood life supported by a common gene pool maintained by free gene exchange. Populations of such pairs would employ simultaneously (under different environmental conditions) different dispersal strategies that might reduce interspecific competition. Apparently, long-living planktotrophic larvae maintain a wide geographic range of a species and high genetic integrity between distant subpopulations [11]. Comprehensive accounts on the benefits of these strategies are already given [10, 50]. At the same time, it is generally accepted that, in shelled molluscs, the presence or the absence of any nutritional resource during development affects egg size, which, in turn, affects the size, the number of whorls, and the morphology of the protoconch [1, 52].

Apart from the loss of planktotrophy (in our view, in some members of the same population, as mutations are random phenomena) in Raphitomidae, there is also a well-documented tendency for repeated loss of other conoidean important foregut structures such as radula, proboscis, and venom gland without alteration of the teleoconch morphology [12, 53–55]. Nevertheless, it is worth noting that the loss of planktotrophy in some turrids, like *Raphitoma*, is not necessarily related to simplification in shell morphology [56] which means that the teleoconch morphology could remain unaltered in a species population consisted of individuals with either lecithotrophic or planktotrophic protoconch. We are of the opinion that at least some of the Mediterranean *Haedropleura*, *Mangelia*, and *Bela* species also fall in the same case as *Raphitoma*, but we conservatively refrain from formally proposing that possibility pending the mitochondrial DNA markers analysis and employment of interbreeding experiments that could answer the questions raised and solve existing paradoxes.

Finally, through the present work the Raphitomid fauna of the Hellenic sea has been enriched by five new members as the search targeted appropriate environments such as biogenic backgrounds, maerl beds, and deeper waters.
