Taxonomy of reproductive Nereididae (Annelida) in multispecies swarms at Ambon Island, Indonesia

Abstract Multispecies, or mass, spawning of different invertebrate species is well known for coral reef systems; however, incidences involving polychaetes are poorly documented. In this study we report on mass swarming, prior to spawning, of Nereididae at Ambon Island, Maluku, on three occasions: in 1866, inferred from an historical sample deposited in Naturalis, Leiden, and in March, 2009 and 2014, based on newly collected samples. The 2009 and 2014 events co-occurred with spawning of other polychaetes, known locally as wawo and including the widespread Indo-Pacific eunicid, Palola viridis (Gray in Stair). Ten species of reproductive Nereididae are described, including Composetia marmorata (Horst) new combination, formerly Ceratonereis marmorata; epitokous modifications are described for both sexes of each species including taxonomically important features such as body colour and number of pre-natatory chaetigers. Three distinct types of natatory region morphologies are recognized, which appear to characterise groups of genera. The ten new records brings to 13 the total number of nereidid species known to undergo mass swarming at Ambon Island; a key to the 13 species is provided. Species composition varies slightly between the three time periods: four species were common between all three periods, five species were in common between 1866 and 2014, and four species were in common between 1995 and 2009/14. Two species of Neanthes and one of Nereis are identified as potentially new and will be described in subsequent papers.


Introduction
Synchronised swarming of polychaetes at the surface of the sea for the purpose of breeding is perhaps best known among the family Eunicidae, in particular the Indo-Pacific species Palola viridis Gray in Stair, 1847. However, swarming is also common among Nereididae and Syllidae and occurs also in at least 15 other families (Clark 1961). Two main strategies are employed: either the adult worm metamorphoses into a swimming form called an epitoke or heteronereid (many Nereididae, some Syllidae), or separate, independent reproductive individuals, called stolons, are formed (Syllidae, some Eunicidae such as Palola species); the epitokes and stolons then congregate (swarm) at the water surface to spawn, usually by rupture of the body wall. The epitokous strategy of Nereididae causes non-reversible modifications in adults that usually results in the death of the adult shortly after spawning. The timing of swarming in both strategies is regulated by both environmental as well as endogenous stimuli (Clark 1961;Watson et al. 2003).
In Ambon and surrounding islands of Maluku, mass swarming of polychaetes occurs every year from February to April. Locals refer to the worms as 'wawo' or, more popularly, as 'laor'. The swarming has been known for a very long time, as occurring either in February and March (Rumphius 1705) or March and April (Radjawane 1982). Animals emerge from their coral substrate habitat and enter the water column two times each year, usually on the second to fourth night after a full moon immediately after sunset and continue for about two hours. In that two hour period, wawo are caught by the islanders in fishing nets (Ambonese: siru-siru) then cooked and eaten. The activity of catching the worms is known as 'timba laor' by the locals (Pamungkas 2011). The entire species composition of wawo from the present collections (2009,2014), comprising 5 families and 25 species, has been listed by Pamungkas (2015); there are similar numbers of species of Eunicidae and Nereididae, but wawo is dominated in mass by the larger-bodied eunicids.
The only previous taxonomic studies of Ambonese wawo are those of Horst (1904in English, 1905 in Dutch) and Martens et al. (1995). A small collection of epitokous worms from Ambon Island by Dutch biologist D.S. Hoedt in 1866, now at Naturalis (Leiden), was never published on, and the identity of these worms -all Nereididae -is reported in this study. The two investigations by Pamungkas (2009Pamungkas ( , 2011 were more concerned with, respectively, sexual behaviour and anthropological aspects of the worms rather than their taxonomic aspects. Since none of the above-mentioned taxonomic research was published in Indonesian (and because of their limited availability in Indonesian libraries), the natives of Ambon Island have been more influenced by Radjawane (1982), an Ambonese researcher who identified the eunicid Lysidice oele Horst, 1905 as the only wawo species in Maluku waters, and popularized the mass spawning phenomenon of wawo among locals. Reports of swarming Nereididae in other tropical regions are equally scarce: Hutchings and Howitt (1988) found 16 nereidid species in multiple swarming events at Lizard Island, northern Great Barrier Reeef, Australia, between October and January, 1985/86 -none of the species were identified beyond family, but a later review of the nereidids of this island (Glasby 2015) indicated that at least some species are in common with those at Ambon Island.
In this study we describe the taxonomy of the nereidid component of wawo collected in 2009/14 and nereidids in Hoedt's collection (1866), in particular the epitokous modifications of each sex and species. Because all samples contained a small fraction of spent worms (i.e., lacking coelomic gametes), we have assumed that spawning occurred about the time the swarming worms were collected. The species composition of the two collections, almost 150 years apart, is compared with that reported by Martens et al. (1995).

Material and methods
Sexually mature polychaetes (wawo) were collected during the swarming period on 14 March, 2009 and18-19 March, 2014, from Ambon coastal waters (Table 1, Fig. 1). At each station, the worms were collected using a small, fine-mesh hand net and were immediately fixed with 10% formaldehyde solution for at least 24 hours. A lantern or torch was used to illuminate the surroundings and attract the worms. Collected worms were rinsed with tap water to remove both the fixing agent and salt crystals, and were then preserved in 70% ethanol.
In the laboratory, the epitokous nereidids were separated from other species of wawo by gross differences in the body form. The primary characteristics used to identify the different types of nereidid wawo were form of the parapodia and head morphology including eyes, antennae number and proboscis form. Voucher specimens were deposited at the Museum and Art Gallery of the Northern Territory, formerly Northern Territory Museum (NTM), Darwin, Australia, the Museum Zoologicum Bogoriense (MZB), Bogor, Indonesia, and the Reference Collection LIPI Ambon (RCLA), Indonesia (RCLA belongs to the Research Center for Deep Sea, Indonesian Institute of Sciences). Comparative material was sourced from the former Zoological Museum Amsterdam (ZMA) and Naturalis (formerly Rijksmuseum van Natuurlijke Historie, RMNH), Leiden; the ZMA collection is now integrated into the RMNH collection.
Nereidid samples collected by D.S. Hoedt and V.D. Velde were studied during a visit to Naturalis (RMNH), Leiden, in 2009 by CJG. The former comprises hundreds of epitokous Nereididae in one jar (Fig. 2), which was at the time uncatalogued. The sample was received by Naturalis (formerly the Leiden Museum) in 1867 according to the register. The entire sample was studied, and selected male and female representatives of each putative species separated. A voucher collection was registered separately with the NTM.
Preserved specimens were examined using stereo (Nikon SMZ 1500 and Nikon SMZ 645) and compound (Nikon ECLIPSE 80i and Nikon ECLIPSE 50i) light microscopes. Macrophotographs of preserved animals were taken with a Canon 5D Mark II with a Canon MPE-65 Macro Lens.
Potential new species were flagged and identified using informal names or 'cf.'; for these taxa there is currently insufficient information on morphological varia-  Table 1 for co-ordinates and collection times.  Bakken and Wilson (2005) and for metamorphosed individuals (Read 2007), and for the paragnath form follows Bakken et al. (2009). Each morphospecies is diagnosed using a combination of paragnath form and pattern, epitokous modifications, and ethanol-preserved colour patterns of the body and eggs. In the absence of a robust phylogeny for Nereididae -and therefore the possibility of using a phylogenetic-based species concept -we use the morphospecies species concept as defined by Cronquist (1978), i.e., species are the smallest groups that are consistently and persistently distinct, and distinguish- able by ordinary means. Our proposed species are therefore hypotheses which are falsifiable when independent data, for example morphological synapomorphies and DNA sequences, become available. Size range. Female: length (11-15 mm), maximum width (2.9-3.0 mm). Male: length (25 mm), maximum width (3.1 mm).
Remarks. The specimens examined agree well with Hartmann-Schröder's (1985) type description especially the characteristic pigmentation pattern that we observed in female specimens, and paragnath count and form of the unmodified parapodia. It is the same as the material identified by Horst (1924) as Nereis (Ceratonereis) tentaculata from eastern Indonesia and as Ceratonereis (Ceratonereis) cf. singularis australis from Ambon Island by Martens et al. (1995). Specimens of C. australis from Lizard Island, northern Great Barrier Reef described by Glasby (2015) are also likely to be conspecific. Ceratonereis (Ceratonereis) cf. perkinsi by Martens et al. (1995) appear to correspond to specimens observed by CJG from northern Australia, near the type locality (Broome, Western Australia) in terms of parapodial form and pigmentation pattern. Specimens reported by these authors as Ceratonereis (Ceratonereis) sp., do indeed belong to the Ceratonereis sensu stricto as described by Hartmann-Schröder (1985), as indicated by the illustrated deeply cleft prostomium ( fig. 13); however, the suggestion that they are similar to C. japonica is incorrect because the latter species has a very different type and arrangement of paragnaths, more in line with a species of Solomononereis (see Glasby 2015).
Distribution. Eastern Indonesia and Australia (widespread).
Remarks. The distinctive marmorated (=veined) stripes on this species of Composetia facilitate identification. The markings are as described by Horst (1924: 177) for Nereis (Ceratonereis) marmorata, and paragnath counts match well, although Horst reports considerable variation. Examination of paragnath numbers on the syntype material confirmed the identification. Hartmann-Schröder's (1985) tentative allocation of the species to Composetia is confirmed, and formalized here. Specimens from Lizard Island, northern Great Barrier Reef described by Glasby (2015) are likely to be conspecific.
Remarks. The specimens examined in this study agree well the description of Willey (1905), which however is not very detailed. This species of Neanthes may be distinguished from other wawo Neanthes by the brown band on the dorsal side of chaetiger 2.
Remarks. The specimens examined in this study were compared with Horst's, syntypes, which they resemble closely, especially in the paragnath count and pattern; however, the present specimens are much larger than those of Horst's, the modified chaetigers start later in the present specimens, and Horst does not mention the presence of a brown band which we observed on chaetiger 4. Therefore our material possibly represents a new species; it is very similar to specimens also referred to as Neanthes sp. cf N. gisserana by Glasby (2015). Diagnosis. Neanthes species having uniform brown pigmentation on dorsal surface, apparently darker in male compared to female (Fig. 5A, B). Paragnaths conical arranged as follows: Areas I: 0; II: 10-13; III: 19-25; IV: 14-20; V: 0; VI: 1-2: VII-VII: 2-5 in line. Male and female epitokes both with natatory region extending to pygidium, but differ in number of pre-natatory chaetigers and number modified dorsal and ventral cirri. Male has pre-natatory region comprising 28-30 chaetigers; basally swollen dorsal cirri on chaetigers 1 to 7; basally swollen ventral cirri on chaetigers 1 to 6. Female has pre-natatory region comprising 32 chaetigers; basally swollen dorsal cirri on chaetigers 1 to 5; basally swollen ventral cirri on chaetigers 1 to 5. Also male has pygidial rosette and sub-distally swollen, scalloped dorsal cirri in anterior part of natatory region (dorsal cirri unmodified posteriorly), lacking in female. Female with brown pygidium.
Remarks. The present material bears a general resemblance to Grube's species, which was described from Masalac, Philippines, in terms of pigmentation pattern and paragnath arrangement and counts. Our specimens differed in lacking paragnaths in Area I (Grube illustrates 2 paragnaths) and in the relatively longer dorsal cirri (up to 3 times length of parapodia in anterior chaetigers), whereas in Grube's specimens the dorsal cirri appear to be slightly longer than the parapodia, but the angle of the illustration makes it difficult to ascertain how much longer. It probably represents a new species. Hartman (1959) referred Neanthes masalacensis to Pseudonereis based on the comparisons of earlier workers (e.g., Fauvel, Gravier) with other species of Pseudonereis. However, examination of the holotype of Nereis (Lycoris) masalacensis by Hutchings and Glasby (1985), showed the absence of elongated dorsal notopodial lobes in posterior chaetigers, and the original figure by Grube (1868) showed the paragnaths in Area II to be conical shaped rather than p-bars (sensu Bakken). Both features strongly suggest that the species does not belong to Pseudonereis. Hutchings and Glasby (1985) consider it to be indeterminable unless other material from the type locality (Masalac, Philippines) can be examined (and the distinctive pigmentation pattern matched). With this new material, from the same general biogeographic region (Western Coral Triangle) that resembles closely the type description of N. masalacensis, we suggest the species be considered a member of Neanthes, as listed in Pamungkas (2015).  Diagnosis. Nereis species lacking pigmentation (Fig. 5C). Paragnaths include cones and bars arranged as follows: Areas I: 0-1 cone; II: 15-20 cones; III: 19-26 cones; IV: about 45 in large crescentic patch includes cones and bars; V: 0; VI: 15-20 small patch of minute cones; VII-VIII: one narrow band of 50-60 cones with several additional rows at right angles. Male and female epitokes both with natatory region extending to pygidium, but differ in number of pre-natatory chaetigers and number modified dorsal and ventral cirri. Male has pre-natatory region comprising 22-23 chaetigers; basally swollen dorsal cirri on chaetigers 1 to 7; basally swollen ventral cirri on chaetigers 1 to 5 or 6. Female has pre-natatory region comprising 26 chaetigers; basally swollen dorsal cirri on chaetigers 1 to 5; basally swollen ventral cirri on chaetigers 1 to 4. Also male has pygidial rosette and scalloped dorsal cirri in natatory region (except near pygidium where dorsal cirri become smooth and tapered again), rosette and scalloped dorsal cirri lacking in female.

Nereis 'sp_Ambon_NTMW19037' Glasby
Remarks. This appears to be a new species, and will be described when further specimens, including atokous individuals become available. It was listed by Pamungkas (2015) as Nereis sp.
Remarks. The specimens examined in this study agree well the re-description of  of this species. This is one of the largest of the wawo nereidids. The specimens reported as Perineris cultrifera by Martens et al. (1995) are most likely P. helleri, as they fit well the paragnath number and pattern described here for this species (see also , although there is a discrepancy in the number of anterior unmodified chaetigers: Martens et al. (1995) report this as 14 and 18 (male and female respectively), but our observations for this species are 17 and 19-20 (ditto). We believe Martens et al. may have confused P. helleri with P. nigropunctata, which were also found in their samples, as the latter has about 14 and 17 anterior unmodified chaetigers for male and female respectively (see following account).

Perinereis nigropunctata
Remarks. The specimens examined in this study agree well the description of  of this species. This is the smaller of the two wawo Perinereis.
Remarks. The present specimen fits the type description closely. Distribution. Indonesia, Northern Australia, Solomon Islands.
Remarks. The specimens reported as Nicon sp. by Martens et al (1995) are almost certainly Websterinereis foli. We reached this conclusion based on a comparison of their figures of parapodia and chaetae with those of Pettibone (1971: figs 10, 11) and their description of the pigmentation pattern compared to that described for this species by Pettibone (1971: fig. 10a) and Glasby (2015: figs 42, 43). Martens et al. (1995) are likely to have overlooked the pharyngeal (oral ring) papillae in their specimens as they are very small and only observable when the pharynx is fully everted: this explains the incorrect generic assignment.

Key to reproductive Nereididae (wawo) of Ambon Island
The following key provides a means of identifying male and females of each species of Nereididae known to swarm at Ambon Island. It includes species reported both in the present study and in Martens et al. (1995).

Discussion
Taxonomic utility of epitokal modifications. Sexual dimorphism among reproductive Nereididae is well known -it includes differences in the number of basally-swollen anterior dorsal and ventral cirri, the number of pre-natatory region chaetigers, and the presence of undersurface scalloping on the natatory dorsal cirri and pygidial rosettes in males only (Schroeder and Hermans 1975 and references therein). All are useful taxonomic features at the species level, but as noted by Read (2007) they are infrequently known. The present study adds to the knowledge of these species-specific epitokal modifications for a number of different species. In addition, we found that males as a rule appear to have an equal number, or fewer, pre-natatory chaetigers compared to females. Further, possibly there exists colour sexual dimorphism in Ceratonereis singularis australis, as the presence of dorsal banding was absent in the only confirmed male specimen examined, but present in all females examined. Less well known is that Nereididae appears to exhibit epitokal modification patterns at the generic level. Combining data from this study with literature information on Platynereis (Horst 1924) and on Websterinereis (Martens et al. 1995), we can recognize three types of natatory region morphologies: natatory regions extending over the mid-and posterior body to the pygidium (Composetia, Neanthes, Nereis, Perinereis, Platynereis; Figs 3C, D, 4A-D, 5A-C, 6A-D); natatory regions restricted to the mid-body and followed by an unmodified region extending to the pygidium (Ceratonereis, Websterinereis; Fig. 3A); and natatory regions of two types (mid-body part with typically modified parapodia; and posterior oval-shaped part with extremely reduced parapodia extending to pygidium (Solomononereis; Fig. 5D). Further, anterior dorsal and ventral cirri may be modified, i.e., basally swollen (Composetia, Neanthes, Nereis, Perinereis, Platynereis; Fig. 4A, B) or unmodified (Ceratonereis, Solomononereis). Finally, the dorsal cirri in the natatory region in males may be scalloped (some Neanthes, Nereis, Perinereis, Platynereis; Fig. 6A); or smooth (Ceratonereis, Composetia, Neanthes cf. gisserana). On the last point, our observations indicate that Neanthes is polymorphic for dorsal cirri scalloping; Schroeder and Hermans (1975) also noted variability in scalloping among Nereis species. This is further indication that the large nereidid genera such as Neanthes and Nereis are non-momophyletic, as suggested by Bakken and Wilson (2005).
Finally, some caution must be attributed to the generality of these patterns of epitokous modifications as possibly they are dependent on the state of sexual maturity of the individual and, of course, they are based on observations of very few species per genus.
How many species? The number of species recognised as comprising wawo in Maluku is increasing as we collect more intensively and overcome taxonomic problems. In the first publication on wawo, Rumphius (1705) considered that wawo was a single species, which he called Vermiculi Marini; however, it is clear that he recognised several different forms under this name. As this was more than a century before any formalised evolutionary thought, presumably he attributed the variation to different forms of a creature specially created. As the name Vermiculi marini was suggested about 50 years before Linnaeus's, binomial taxonomy system was in place, it is not accepted by the International Code of Zoological Nomenclature. Rumphius (1999) describes one form that resembles a millipede: … about the thickness of an oaten pipe, quite like young Millipedes, of a mixed green, brown, and white, and look indeed somewhat disgusting, but these have a special name, and are not considered the true Wawo.
And further, that the millipede form: … illumes at night, giving off a clear light, which makes people avoid them even more, since they share this attribute with the Millipedes.
The millipede form may represent the shorter-bodied nereidids with their prominent parapodia; however, luminescence has, as far as we know, not been reported previously in swarming members of this family, so doubt must remain over the identity of Rumphius's, 'millipedes'. The more abundant true wawo referred to by Rumphius (1999) as 'like silken Floss, all entangled in small clumps' are most likely Palola viridis and possibly other eunicid species (see Pamungkas 2015). Horst (1904; also claimed that wawo is a single species, namely Lysidice oele (Eunicidae) based on collections from Banda waters during the Dutch Siboga Expedition (1899)(1900). However, by the end of the twentieth century, Martens et al. (1995) had clarified the multispecies nature of wawo. He recognised 13 different species (five families) from Ambon Island waters with the eunicid Palola viridis as the dominant species; 7 out of 13 species of wawo were nereidids. Three of these species were the same as reported in this study, viz. Ceratonereis singularis australis, Neanthes unifasciata and Perinereis nigropuncata (Table 2). In the present study, six species of Nereididae were identified in the collections of Hoedt and nine species in the recent collections (Table 2; also Pamungkas (2015)). The Hoedt nereidids are all now in one jar (Fig. 2), so we do not know for sure whether they were collected from more than one site and/or on more than one occasion. However, assuming samples from all different periods each represented a single spawning event, the differences in species diversity between the three periods -1866 (6 species), 1995 (7 species), 2009 and 2014 (9 species) -is entirely consistent with spatial or short term temporal variation and does probably not indicate long term change.
Finally, it is worth noting that wawo have been reported to swarm in islands nearby Ambon, such as Haruku and Nusalaut (JP pers. obs.) and therefore are probably widespread in Maluku Province. In the Banda Islands (south-east of Ambon Island), the natives call them 'oele'. Further, studies of Nereididae by Horst (1889Horst ( , 1924 from the 'Malay Archipelago' (includes present-day eastern Indonesia, Malaysia, Brunei, Timor Leste) yielded 44 nereidid species, many of which were reproductive; four species were the same as found in our recent collections (i.e., Ceratonereis singularis australis, Composetia marmorata, Nereis gisserana and Perinereis nigropunctata ), with the latter two having widespread distributions throughout the western Coral Triangle (the actual number of species widespread in the region is likely to be much higher). Therefore, it is highly likely that polychaete swarming events throughout the western Coral Triangle region will be found to contain not only Palola viridis and other eunicids, but a range of nereidid species, including those found in this study and other species.  (1866), reported by Martens et al. (1995), and found in this study (2009/2014 (1995)) Absent Present Absent