The family Bullidae, and its sole genus Bulla, is the most well-known group of shell-bearing opisthobranch gastropods (order Cephalaspidea). It occurs worldwide in tropical areas, but also has some representatives in temperate latitudes. These molluscs can be found on intertidal flats, in tide-pools and to depths down to 70 m, in habitats of sand, mud, gravel, green algae and seagrass. The living animals can be seasonally common, but are best known from their distinctive shells that are frequently washed ashore.
During the 18th and 19th centuries the genus Bulla was used for the majority of bubble-shelled (‘bullomorph’) gastropods, including species now assigned to the opisthobranch genera Haminoea, Retusa, Cylichna, Roxania and many others (e.g. da Costa, 1778; Bruguière, 1792; Férussac, 1822; Lamarck, 1822; Philippi, 1836; Menke, 1853). Of the more than 400 names assigned to Bulla the vast majority have already been attributed to other genera (e.g.Thompson, 1988).
The modern concept of the genus Bulla was established during the second half of the 19th century by Adams (1850), Sowerby (1868) and Pilsbry (1895), who produced comprehensive monographs of worldwide species based on shell characters. Since then, very few authors have either included in Bulla any species other than those discussed here, or have recognized more than the single genus Bulla (or one of its synonyms) within the family. Dall (1908) described the new subgenus and species Bulla (Leucophysema) morgana from abyssal depths off the Pacific coast of Panama, but this is known only from its shell, which is unlike that of Bulla species. Nordsieck & García-Talavera (1979) listed species of Bulla under both Bulla and Cylindrobulla, but the latter is a genus of sacoglossan opisthobranchs.
Most systematic work on Bulla has focused on the description of species, but the external form and colour of the shell have been the main source of characters. However, in the majority of cases these characters are either highly variable within species or very similar among them, leading to continuing taxonomic confusion. Mikkelsen (1993, 2002) pointed out that the radula and particularly the reproductive system might be useful for discrimination between species of Bulla. However, these features have previously been investigated in only a few species, and the taxonomy of the genus is still far from resolved.
Thirty-one species names plus six varietal names have been used for species of Bulla in the Atlantic and eastern Pacific oceans. These taxa were mostly introduced by early malacologists, based on poor descriptions of a few shells from restricted geographical regions. There is yet no consensus on the number of valid species. For example, in the eastern Atlantic Nordsieck (1972) accepted six species, García-Talavera (1983) recognized one, Poppe & Goto (1991) recognized three, and in the Iberian Peninsula and adjacent islands Cervera et al. (2004) listed five. The relationship between similar forms on either side of the Atlantic has been a matter of enduring debate. Whereas some authors have accepted an amphi-Atlantic species (under the name B. striata or B. amygdala, e.g. Dall & Simpson, 1901; Nordsieck, 1972; Eisenberg, 1981; García-Talavera, 1983; Poppe & Goto, 1991; Rios, 1994; Macedo, Macedo & Borges, 1999), others have employed a different name for the American taxon (e.g. Vilas & Vilas, 1970; Damerval & Damerval, 1990; Redfern, 2001). One western Atlantic species, B. solida, has seldom been illustrated. In the eastern Pacific two species are generally recognized in the recent literature (e.g. Behrens & Hermosillo, 2005), but their geographical limits are not well established.
In the Indo-West Pacific there is similar confusion, with 41 specific and ten varietal names available. The reasons are similar, authors having attributed too much weight to minor shell differences without comparing material from different localities, while ignoring the study of anatomical traits. Thus, for example, B. quoyii from Australasia has received nine names with two authors naming it more than once (Gray, 1825, 1843; Adams, 1850) and the tropical B. vernicosa has 11 synonyms. Currently the number of accepted species in the Indian and western and central Pacific oceans varies, but three to four species are often referred to in the recent literature (B. ampulla, B. punctulata, B. quoyii and B. vernicosa, e.g. Burn & Thompson, 1998; Steyn & Lussi, 1998; Hori, 2000). The use of incorrect taxonomy and uncertainty about valid species still prevails (e.g. Willan, 1978; Hori, 2000; Thach, 2005).
Bullidae are usually considered to have a long palaeontological history, dating back to the Late Jurassic of Europe (Zilch, 1959; Tracey, Todd & Erwin, 1993). Nevertheless, a re-examination of the fossil record has shown that these older fossils, despite their bullomorph shape, probably belong to a different taxonomic group (M. A. E. Malaquias, unpublished data). The oldest fossil that can confidently be attributed to Bullidae is from the Miyaragawa Formation in the Ryukyu Islands, southern Japan, dated to the Middle–Late Eocene (33.9–48.6 Ma) (MacNeil, 1964; Nakamori, Kawano & Iryu, 1998).
Similar taxonomic uncertainty is common among the majority of shell-bearing cephalaspids, where shells remain in most cases the only source of systematic characters, and anatomical data when available tend to be limited to few species from a particular taxonomic group (e.g. Rudman, 1971, 1972a, b; Gosliner, 1979; Gibson & Chia, 1989; García, Pérez-Hurtado & García-Gómez, 1991; Álvarez, García & Villani, 1993; Martínez & Ortea, 1997; Burn & Thompson, 1998) or to specific geographical areas (e.g. Bouchet, 1975; Talavera, Murillo & Templado, 1987; Valdés & Camacho-García, 2004; Malaquias & Cervera, 2006). Molecular data now offer enormous potential to discriminate among cryptic species and to infer phylogenetic relationships (Dayrat et al., 2001; Lapègue et al., 2002; Meyer, 2003; Williams, Reid & Littlewood, 2003; Williams & Reid, 2004; Collin, 2005). Nevertheless, among cephalaspid molluscs molecular data have so far been used only to infer relationships at higher levels in the context of the opisthobranchs as a whole (Thollesson, 1999; Medina & Walsh, 2000; Dayrat et al., 2001; Grande et al., 2004a, b; Vonneman et al., 2005).
In this monograph the systematics of the worldwide species of the family Bullidae are revised. This study adopts an integrative approach (e.g. Dayrat, 2005, 2006), defining species through a combination of morphological and molecular characters. This is the first comparative anatomical study of worldwide Bullidae and the first use of molecular data for species delimitation in a global systematic revision in the Cephalaspidea.