Dodecaceria carolinae n . sp . ( Polychaeta : Cirratulidae ) , a shallow-water species from the northwestern Caribbean Sea

The cirratulid genus Dodecaceria Örsted, 1843 is well characterized by having palps inserted dorsally or laterally and branchial filaments restricted to some anterior chaetigers. The genus has species recorded from many marine environments worldwide and species are differentiated mainly by the number of branchial filaments. In this contribution, D. carolinae n. sp. is described based upon extensive materials collected in the northwestern Caribbean Sea. This species is distinguished by having 12 pairs of branchiae in two distinctive sizes, the first three four times longer than the remaining ones, and by having spoon-shaped hooks from notopodia 14-19 and from neuropodia 13-18. An analysis of the intra-specific variability, together with a table of diagnostic features and a key to all known species, is also included.


INTRODUCTION
The Cirratulidae is a well-characterized polychaete family with a conical or anteriorly rounded prostomium lacking appendages and a peristomium fused with at least two segments.The parapodia are biramous, with papillar chaetal lobes and simple chaetae including capillaries and spines or hooks.Branchiae are paired, slender filaments arising from the dorsal surface of each segment and usually extending over most of the body (Fauchald, 1977).Species belonging to the rock-boring genus Dodecaceria Örsted, 1843 have been documented from all over the world.This genus was defined by the presence of one pair of grooved palps, arising dorsally or laterally, with branchiae roughly of the same size and restricted to a few or several anterior chaetigers, and chaetae always including spoon-shaped hooks.In view of our results, however, this definition must be emended to include also the possible presence of branchiae of two distinctive sizes (see below).
The delineation of Dodecaceria species has been rather problematic, especially since the brief, confus-ing description of the type species, D. concharum, based on specimens collected between Fredrikshavn and Skagen, Denmark.The name Dodecaceria refers to its 12 anterior filaments, which corresponds to the 5-6 pairs of branchiae (and palps) but, other than indicating the presence of ventral hooks, no further details were provided.Based upon this vague description, D. concharum has been reported as a widely distributed species.
One of the explanations for this now surprising distribution might stem from the detailed studies by Caullery and Mesnil (1898).They found five reproductive forms divided into three different stages, which were regarded as belonging to the same species: A) Atoky sedentary, B) Epitoky black, and C) Epitoky sedentary.Fauvel (1927) accepted this apparent reproductive polymorphism and set the synonymy for all the NW European species under a single name.Interestingly, Martin (1933) regarded this reproductive variability as an indication of the existence of a species complex, rather than a single polymorphic one.Dehorne (1933) went a step further by naming form B as D. caulleryi (George and Petersen, 1991).The subsequent development of the studies was reviewed by Petersen (1999).Thus, D. concharum corresponds to form B and includes D. caulleryi, while D. ater (de Quatrefagues, 1866) corresponds to forms A and C.
The prevalent taxonomic scenario for Dodecaceria involves using the number of branchial pairs, together with the starting chaetiger for the presence of the spoon-shaped hooks in both neuropodia and notopodia (Knox, 1971).However, no evaluation of the intraspecific variability for any of the known species has been reported to date.
Two species have been described for the Grand Caribbean region: Dodecaceria inhamata (Hoagland, 1919) from Bermuda and D. diceria Hartman, 1951 from off southwestern Florida.Dodecaceria con-charum Örsted, 1843, D. coralii (Leidy, 1855), D. laddi Hartman, 1954 andD. pulchra Day, 1955 have also been reported for the region, but these reports may correspond to misidentifications and the list deserves more detailed work.
In this paper, a new species is described from the NW Caribbean region and the generic diagnosis is consequently emended.Moreover, an evaluation of the intra-specific variability of morphologic or diagnostic features (including the shape and relative size of branchial filaments, the number of chaetigers and the starting chaetiger-and range of presence-for the spoon-shaped hooks), together with a comparative table of diagnostic features and a key to all known species of the genus, are also included.

MATERIALS AND METHODS
The samples were collected along the northern coast of the Yucatan Peninsula, in areas with many different types of calcareous substrates including coralline rocks.Rocks were broken with a hammer and chisel, and polychaetes were removed with forceps.The worms were anesthetized using osmotic-shock or by placing them in an ice chest with some ice.Specimens were then fixed in a 10% formalin-sea water solution, soaked in tap water for 24 h, and preserved in 70% ethanol.
All specimens were studied under light microscopy.For each location, 30 worms were selected to measure body length (both as relative length up to chaetiger 10, L 10 , and total body length) and width with a micrometer.The possible existence of significant relationships between total body length and L 10 or total number of chaetigers was assessed.Diagnostic features include the number of branchial filaments, together with their length relative to body width, the total number of chaetigers, and the range of chaetigers from that of first appearance of hooks to that where capillaries reappear in both noto-and neuropodia.The photographs of palps, branchiae, parapodia and whole specimens were taken with a digital camera.
Holotype and additional materials were deposited in ECOSUR, and paratypes were deposited in the fol- Diagnosis (emended).Prostomium conical, blunt.Peristomium long, achaetous.Palps between peristomium and first chaetiger, inserted dorsally or laterally; thicker and subsequally longer than branchiae, with longitudinal ciliate borrow.Branchiae up to 22 pairs, restricted to anterior chaetigers, either monomorphic (gradually decreasing in length) or dimorphic (the anteriormost larger than the following ones).Chaetae including capillaries in both parapodial rami and stout, acicular, always spoon-shaped hooks in noto-and neuropodia.
Remarks.The number of branchial filaments, together with their relative size and the number of filaments in the first branchial segment, have been widely employed to sort out the Dodecaceria species (Table 1).According to our observations, there are two distinctive patterns regarding the relative size of branchial filaments, which may either be all of about the same size, often slightly decreasing posteriorly, or the anterior pairs distinctly longer than the posterior ones (Fig. 1A, B).Accordingly, they are here regarded as monomorphic or dimorphic, respectively (Table 1), forcing us to emend the generic diagnosis.In the new species, the presence of some short, regenerating branchial filaments in the first few chaetigers seems to occur in specimens also regenerating the whole anterior end, which are exposed to predation when protruding from the galleries excavated in calcareous substrates for feeding (Fig. 1C).
The relative fusion of post-peristomial segments in Dodecaceria (de Quatrefages, 1866;McIntosh, 1911) and also in Cirriformia (Blake, 1975;Wilson, 1936) may result in two distinctive patterns.Most cirratulid genera (including several Dodecaceria species) have a single pair of branchiae per segment (Day, 1967), including the first branchial one.In turn, four species of Dodecaceria (D. capensis Day, 1961;D. choromytilicola Carrasco, 1977a;D. gallardoi Carrasco, 1977b andD. opulens Gravier, 1908) appear to have two pairs of branchiae on in the first branchial segment, which also carries the palps.This presumably occurs by fusion of the first two segments, which may be so complete that they cannot be distinguished in dorsal view, as, for example, in D. carolinae n. sp., where the branchiae are inserted in slightly different planes than palps and are better observed in lateral view (Fig. 1D).Moreover, these four species apparently have a more pronounced parapodial progression.Description.Holotype complete.Body dark brown, cylindrical, tapering towards both body ends, posterior region narrower, depressed; 40 mm long, 1 mm wide, 126 chaetigers (Fig. 1A).Prostomium rounded, expanded, smooth.Nuchal organs as two dorsal, short slits, placed towards posterior margin.Peristomium multi-annulated, with 6-7 dorsal rings.Palps laterally inserted (Fig 1B , D), 5 mm long.First segment achaetous, completely fused to peristomium.Branchiae dimorphic, from "peristomium" to chaetiger 11, one pair per segment; chaetigers 1-3 with filaments four times longer than posterior ones.Branchial insertion gradually displacing dorsally from first to last branchial chaetigers.Chaetigers annulated dorsally, 3-4 rings per segment.
Intra-specific variability (Table 2).The specimens had an average length of 28.31 mm (SD±7.36mm) and a maximum width of 1 mm, while the average number of chaetigers was 106.45 (SD±15.30).Body shape is variable, but usually the anterior chaetigers are wider than the rest of the body, often tapered although in some specimens the body is rather cylindrical.The colour in vivo was dark green, becoming dark, even black with some yellow chaetigers once preserved.Some specimens have a pale anterior region and a dark posterior one.The posterior region is often depressed.
All specimens had 12 pairs of branchiae, although in some specimens they were only noticed thanks to the presence of scars.When the whole set was present, a variability in relative length and proportion was observed, but 75% of the specimens had the first three branchial pairs longer than the remaining ones, while the others had up to six longer pairs.This variation was not size-dependent, since the latter measured 21 to 66 mm in length, but rather corresponds to differences in regeneration after predation or cropping of the long branchial filaments (see diagnosis remarks).
The correlation between total body length and L 10 (R²=0.2148)and number of chaetigers (R²=0.4694) is low and non-significant, which might be due to the presence of specimens with regenerating anterior ends or to differences in the intensity of body contraction during preservation.
Females: Usually longer than males (average length 35 mm), having a paler pigmentation and being easily recognized by the widening of the mid-body chaetigers.They have 1-2 rounded pits along cha-  etigers 55-86, and the coelom is completely full of oocytes, which apparently belong to the solitary type (Eckelbarger, 1983) and measured 45-55 μm in diameter.No epitokal transformation or chaetal modification was observed.
Habitat.This species was found along the Yucatan Peninsula, in subtidal limestone and corals (Porites and Pavona).The worms bore galleries and build tubes, with both ends protruding from a single opening (Fig. 1C) and no clear gallery pattern.In limestone, the galleries have an equal diameter and form parallel homogenous tubes.In living coral, the patterns differ somewhat, being U-shaped or curved, but also non-homogenous.Each tube may harbor 1-3 worms.

Methyl green straining pattern.
There is no evident pattern, staining is solid and homogeneous throughout the body.
Etymology.The specific name honours the unrestricted care of Mrs. Carolina Camacho, mother of JMAC.Without her support and courage he could not have reached any of his goals.
Remarks.Dodecaceria carolinae n. sp.resembles other species having 10 or more pairs of branchiae, such as D. choromytilicola Carrasco, 1977a, D. opulens Gravier, 1908and D. meridiana Elías and Rivero, 2009, but differs from the first two in having a single pair of branchiae in the first segment (two in these species).As for D. meridiana, the type specimen was *This species looks like an epitoke of D. coralli due to the presence of dark tips in the branchiae, as stated in the original description.However it was synonymized with D. concharum (George and Petersen, 1991).This difference might alter any comparisons, such as those based upon the number of chaetae, since a higher chaetal number might be expected for larger specimens.However, D. meridiana has 7-8 notochaetae and 6-7 neurochaetae in the anterior chaetigers (5 and 3 in D. carolinae n. sp., respectively), so the smaller species is in fact the most chaetose.Moreover, D. meridiana has the first 8-10 branchial pairs markedly longer than the remaining ones (first 3 to 6 in D. carolinae n. sp.).Dodecaceria carolinae n. sp. also differs from other species registered for the Grand Caribbean region.Thus, D. inhamata (Hoagland, 1919) from Bermuda has 5 branchial pairs, and D. diceria Hartman, 1951 from off Florida (117 fathoms deep) has a single branchial pair.
Furthermore, Hartman (1951) reported an incomplete specimen (8 mm long, 1.3 mm wide, with 29 chaetigers) collected in Alligator Harbor, Franklin, Florida as D. near concharum Örsted, 1843.The specimen resembled D. carolinae n.sp. in being dark green and having eight long and four short branchial filaments.Moreover, the hooks appeared three chaetigers after the branchial filaments (i.e. by chaetiger 15) and, consequently, it is here regarded as conspecific.Vinn (2009) studied the fine structure of tubes of some specimens recorded in Chicxulub Puerto, Yucatan, a place close to the type locality of D. carolinae n. sp.Since no morphological details for the specimens were provided, it might be included as belonging to this new species; however, these materials should be evaluated.
Further studies are needed to specify the distribution of the new species, which includes the Yucatan Peninsula and might extend to Florida, and to assess its reproduction and the substrate-drilling mechanism.1; specimens relative length in millimetres)