Cribrilinidae Hincks, 1879

Rosso, A., Beuck, L., Vertino, A., Sanfilippo, R. & Freiwald, A., 2018, Cribrilinids (Bryozoa, Cheilostomata) associated with deep-water coral habitats at the Great Bahama Bank slope (NW Atlantic), with description of new taxa, Zootaxa 4524 (4), pp. 401-439 : 405-407

Family Cribrilinidae Hincks, 1879

The generic allocation of a species first included in the genus Puellina and of two new species in the material collected at the Great Bahama Bank (see below) prompted us to reconsider the ranking of Puellina Jullien, 1886 (type species Lepralia gattyae Landsborough, 1852 ), Cribrilaria Canu & Bassler, 1929 (type species Eschara radiata Moll, 1803 ) and Glabrilaria Bishop & Househam, 1987 (type species Puellina pedunculata Gautier, 1956 ). Glabrilaria was formally introduced as a subgenus by Bishop & Househam (1987), whereas the two other taxa were first erected as separate genera. The lowering of Cribrilaria to subgenus level and its inclusion within Puellina was suggested by Gordon (1984: 63) based on the sharing of the pinnate costae, setiform papillae, and a suboral lacuna simulating an ascopore. However, the inclusion of Cribrilaria species within Puellina forced to change the former description by Jullien (1886) to include a wide range of variability in relation to several characters, especially the extent of the frontal shield and gymnocyst development, as well as the presence/absence of interzooidal avicularia and, particularly, of ooecial pseudopores. Following Gordon (1984), Bishop & Househam (1987: 5) formally amended Puellina and included further characters in its description, which were either excluded (presence of pelmatidia) or not considered by Gordon (1984), namely the presence of apophyses on oral spines and the ooecial development by 1) the autozooid distal to the maternal one (Type A of Bishop & Househam 1987), 2) a distal kenozooid (Type B of Bishop & Househam 1987), or 3) the maternal autozooid (type C of Bishop & Househam 1987). In addition, they introduced the new subgenus Glabrilaria ( Bishop & Househam 1987: 7) , which has, among other features, pedunculate, erect or semi-erect avicularia budded from a single porechamber along zooidal lateral walls, and that are, therefore, not in direct contact with the substratum. Consequently, they just mentioned avicularia in their diagnosis of the genus Puellina , which then became inclusive of even higher variability. Bishop & Househam (1987: 5) also gave a diagnosis for the nominal subgenus Puellina , attributing to it the main features that Jullien (1886: 607) had first assigned to his genus, and adding a modern description of the ovicell as well as a reference to setiform papillae from the distalmost papilla pores, possibly corresponding to what Jullien (1886) described as “vibracellaires”. Unfortunately, Bishop & Househam (1987) inconsistently used subgenera in their paper including in them the names of some but not all species, and in the text but not in the figure captions.

In the following years, some authors retained these intrageneric subdivisions (Harmelin 1988; Harmelin & Aristegui 1988; Reverter & Fernandez 1996; D’Onghia et al. 2015; Rosso & Di Martino 2016; Rosso et al. 2018). In particular, Harmelin & Aristegui (1988) followed Bishop & Househam’s (1987) unifying approach, noting that some characters (oral spines, avicularia and ovicell types A-B-C) seem to be intergradational between different subgenera.

In more recent times, Puellina has often been used in its wide sense (including species with unperforated ooecium, avicularia and very wide costal shield and narrow gymnocyst that are lacking in the Puellina type species, see below), but without any indication of subgenera (e.g. Harmelin & d’Hondt 1992; Hayward & Ryland 1998; Harmelin 2001, 2006; Winston 2005; Di Martino & Taylor 2015; Dick & Grischenko 2016; Harmelin et al. 2016). The twofold names (which often result in double entries) lead to a certain confusion that is also obvious when using WoRMS (compiled by Bock & Gordon 2018, accessed 16 March 2018). In our opinion the widespread inclusive usage of the genus Puellina without indication of the subgenus is uninformative and even misleading when considering the characters of the type species of Puellina , i.e. L. gattyae . This species has a distinctive pseudoporous ooecium, and also a small subcircular coastal shield composed of few costae, a large gymnocyst, and no avicularia. All these characters co-occur in a very restricted number of species presently included in Puellina . In contrast, a high number of species possess an imperforate ooecium, plus either interzooidal or adventitious (often pedunculate) avicularia, and lack a wide gymnocyst. Although some of these characters are quantitative and can sometimes show gradational changes (see Harmelin & Aristegui 1988), they seem to be distinctive and to persist for long time within the cribrilinid long-lasting group. Consequently, we suggest to use the set of characters indicated by Bishop & Househam (1987: 5, 7) for the subgenera Puellina , Cribrilaria, and Glabrilaria at the genus level rather than at subgenus level. At the same time, we suggest to elevate Glabrilaria from subgenus to genus level. The distinctive characters for the three genera, largely based on Bishop & Househam (1987), are reported below.

The genus Puellina Jullien, 1886 , with type species Lepralia gattyae Landsborough, 1852 , is mostly characterised by the perforate ooecium, produced by the maternal zooid (C type). Additional characters include: the absence of avicularia; the comparably large gymnocyst and the small coastal shield (always formed by a few number of costae); the presence of setiform papillae from the two distalmost papilla pores and the presence of articulated oral spines, which lack apophyses and are reduced to two on ovicellate zooids; lateral walls exposing relatively small uncalcified windows of pore-chambers; the presence of a median proximal spine in the simple tatiform ancestrula; ancestrula regenerated as a miniature zooid or a kenozooid.

The genus Cribrilaria Canu & Bassler, 1929, with Eschara radiata Moll, 1803 as type species designated by Canu & Bassler (1929: 33), is characterised by the co-occurrence of an imperforate ooecium (usually of the A or B type) and interzooidal avicularia, which can be more or less frequent and variable in size and shape. Both characters occur in the type species and are among those reported by Canu & Bassler (1929), although these authors reported an “ovicelle hyperstomiale fermée par l’opercule” not mentioning the nature of its frontal surface. In addition, the gymnocyst is narrow and the frontal shield is extensive, often formed by numerous costae (although few costae occur in certain species); the oral spines may have lateral pointed apophyses; they are 4–8 (usually 5) and decrease to four in ovicellate zooids; the uncalcified windows of pore-chambers are relatively large; the ancestrula has a median proximal spine and is regenerated as a kenozooid.

The genus Glabrilaria Bishop & Househam, 1987 , with type species Puellina pedunculata Gautier, 1956 , is characterised mostly by the presence of pedunculate or semi-erect avicularia that are born from a single porechamber on zooidal lateral walls, and are consequently not in contact with the substratum. Additional characters include: an imperforate ooecium (more often of the B type but also of the A type; both types co-occurr in particular species: see below); absence of papilla pores; 6–7 (rarely 5) oral spines reduced to four in ovicellate zooids; a relatively narrow gymnocyst and a relatively large frontal shield; tiny to moderately sized uncalcified windows of pore-chambers; a pointed median proximal spine in the ancestrula; ancestrula regenerated as a kenozooid.

When applied, the present suggestions will also contribute to solve problems about the double entries and/or the twofold naming of these taxa, some of which are reported as “in alternate representation” in Bock & Gordon, 2018 (see WoRMS: http://www.marinespecies.org/aphia.php?p=taxdetails&id=467129, accessed 16 March 2018 for Glabrilaria ; http://www.marinespecies.org/aphia.php?p=taxdetails&id= 111339 and http:// www.marinespecies.org/aphia.php?p=taxdetails&id=471844 for Cribrilaria; and also http://marinespecies.org/ aphia.php?p=taxdetails&id= 110742 &allchildren=1 for all three genera, accessed 30 July 2018).

An extensive revision of the wide range of cribrilinids, also including DNA analyses, is necessary to state the significance of the observed intergradation in some characters, to more confidently support the ranking of Puellina , Cribrilaria, and Glabrilaria and allocate all species within these genera more suitably. Although this is beyond the scope of the present paper, we provisionally suggest some combinations, most of which are new, except for a few ones which accept the first allocation of the species when they were erected. These are: Cribrilaria radiata ( Moll, 1803) ; C. bathyalis ( Harmelin & Aristegui, 1988) ; C. bifida (d’Hondt, 1970) ; C. cassidainsis Harmelin, 1984 ; C. denticulata ( Harmelin & Aristegui, 1988) ; C. hexaspinosa ( Harmelin, 1978) ; C. hinchsi ( Riedl, 1917) ; C. innominata ( Couch, 1844) ; C. scripta ( Reuss, 1848) ; C. setiformis setiformis Harmelin & Aristegui, 1988 and C. setiformis romana Harmelin & Aristegui, 1988 ; C. venusta Canu & Bassler, 1925 ; Glabrilaria pedunculata ( Gautier, 1956) ; G. africana ( Hayward & Cook, 1983) ; G. corbula ( Bishop & Househam, 1987) ; G. cristata ( Harmelin, 1978) , G. orientalis (Harmelin 1988) including the three subspecies, namely G. orientalis orientalis (Harmelin, 1988) , G. orientalis lusitanica (Harmelin, 1988) and G. orientalis azorensis (Harmelin, 1988) ; Puellina gattyae ( Landsborough, 1852) ; P. modica Bishop & Househam, 1987 and P. setosa ( Waters, 1899) . We also designated further new combinations for some species described or mentioned below.

It must be remarked that the assignment of some species to one of these three genera could be problematic. This is the case with Cribrilaria octospinosa Harmelin, 1978 and Puellina (Cribrilaria) pseudoradiata Harmelin & Aristegui, 1988 , both possessing characters pointing to their allocation in Cribrilaria, but apparently missing avicularia ( Harmelin 1978; Harmelin & Aristegui 1988). Examination of further material could help to solve this problem because these species could parallel C. bifida , a species lacking avicularia in the type material, but having rare avicularia in some colonies ( Bishop & Househam 1987: 39).

It could also be suggested that interzooidal avicularia represent a plesiomorphic character in Cribrilaria that could be reduced and ultimately lost in some species during evolution.