Plakobranchus papua Meyers-Muñoz & van der Velde, sp. n. from West Papua (Papua Barat province, Indonesia), is described based on its external morphology, colour pattern, internal anatomy, radula and reproductive system. In a molecular phylogenetic study specimens of this new species were compared with those of ten candidate taxa under the name Plakobranchus ocellatus van Hasselt, 1824. DNA analyses of COI mtDNA showed a clear distinction between P. papua sp. n. and “P. ocellatus”. Plakobranchus papua, sp. n. also differed from all taxa that have been synonymised with P. ocellatus. The genus is in dire need of taxonomic revision, preferably based on an integrative analysis involving morphology and DNA of all known Plakobranchus varieties.

COI, phylogeny, Plakobranchus ocellatus , reproductive system, Sacoglossa , taxonomy

Sea slugs of the genus Plakobranchus van Hasselt, 1824 (Order Sacoglossa, Suborder Plakobranchacea) have an elongated body and dorsoventrally flattened, lateral parapodia, which are folded up on the dorsal surface (van Hasselt 1824; Jensen 1992). According to Jensen (1997a) this genus possesses a number of plesiomorphic characters such as an anterodorsal anus, a pharyngeal pouch, triangular, denticulate teeth, a long, curved penial stylet, and the absence of dorsal vessels. The genus also possesses a number of autapomorphies: a broad and flat head, rhinophores located at the anterior corners, mediodorsal eyes on a small papilla, numerous longitudinal dorsal lamellae containing branches of the digestive gland, and a truncate tail; the hermaphrodite ampulla has apparently been lost.

These animals can be found in shallow sandy habitats, crawling over it or half-buried (Gosliner et al. 2008; Mehrotra et al. 2015), or on coral rubble and in rock pools (Yonow 2008) where they consume green macroalgae (Jensen 1993).

Plakobranchus ocellatus van Hasselt, 1824, feeds on a wide variety of marine green algae (Chlorophyta), including at least five species of Ulvophyceae (Wägele et al. 2011). Many studies on Plakobranchus deal with their kleptoplasty, the ability to retain functional chloroplasts from their green algae in their digestive gland cells (Clark et al. 1990; Jensen 1996, 1997a). Species of this genus as well as other sacoglossans belong to the few known animal species with the ability of photosynthesis (Trench 1969; Hirose 2005; Bass 2006; Händeler et al. 2009; Maeda et al. 2010; Wägele et al. 2011; Christa et al. 2013; Yamamoto et al. 2013).

Plakobranchus species are simultaneous hermaphrodites, which possess a penial stylet used in hypodermic insemination. Penial stylets and hypodermic insemination are commonly found within the Sacoglossa (Schmitt et al. 2007; Smolensky et al. 2009). Jensen (1992) also observed extensively branched prostate and albumen glands and a pair of secondary copulatory bursae in Plakobranchus, which are unique for the genus.

During the last decades only Plakobranchus ocellatus has been considered a valid species within the genus (Jensen 1992). This species was described from shallow waters in the Sunda Strait near Anyer, northwest Java, Indonesia. van Hasselt (1824) gave this name because of the blue- and yellow-centred ocellated spots covering the dorsal side and flanks of the body (Figure 1). The original description is based on the species’ phenotype and some characteristics of the parapodial lamellae, heart and reproductive system as shown in the original illustrations supplied by van Hasselt (1824). The name Plakobranchus has occasionally been misspelled as Placobranchus, which started when the original description was translated from Dutch to French in 1824 (Bergh 1887; Jensen 1997a, 1997c).

Figure 1. 

a–c Plakobranchus ocellatus, drawings by van Hasselt (1824): a dorsal view with parapodia folded up on the dorsal body surface b dorsal view with open parapodia, showing longitudinal lamellae c internal anatomy d–f P. ianthobaptus, drawings by Gould (1852): d dorsal view with parapodia folded up on the dorsal body surface e ventral view f dorsal view with open parapodia, showing longitudinal lamellae g–j two Plakobranchus species illustrated by Pease (1871)g–h P. gracilis: g ventral view h dorsal view i–j P. variegatus: i ventral view j dorsal view k–l drawings of P. chlorophacus by Bergh (1873): k dorsal view with parapodia folded up on the dorsal body surface l ventral view with ocellated spots m–n drawings of P. ocellatus by Marcus (1982): m dorsal view with parapodia folded up on the dorsal body surface n dorsal view with open parapodia, showing the longitudinal lamellae (en = pericardium; re = renal prominence).

Jensen (1992) studied the anatomy of several Plakobranchus specimens from various Indo-West Pacific locations (Red Sea, Thailand, Guam) and synonymised the previously described species. The descriptions were mainly based on external anatomy, colour pattern and geographic distribution (Yonow 1990, 2008; Debelius 1996; Jensen 1997a; Marshall and Willan 1999; Gosliner et al. 2008) and hardly included descriptions of the internal anatomy as presented by van Hasselt (1824) and Jensen (1992). Plakobranchus ocellatus is now considered the only known valid species of the genus, with ten synonyms and a wide Indo-Pacific distribution (Jensen 2007). This may have been premature (Jensen 1992, 1997a, 1997c), because Gosliner et al. (2008) distinguished two undescribed Plakobranchus species in addition to P. ocellatus. Subsequently, Krug et al. (2013), who used the mitochondrial barcoding gene COI and the nuclear histone 3 gene, found ten distinct phylogenetic lineages in Plakobranchus. This suggests that the taxonomy of Plakobranchus still deserves further study. In the present study a new Plakobranchus species from West Papua, Indonesia, is described and a phylogenetic reconstruction based on the mitochondrial barcoding COI gene is included to show its position within the genus Plakobranchus.

Twenty specimens were collected by Gerard van der Velde in Indonesia during the 2007 Raja Ampat Expedition (Figure 2; Hoeksema and van der Meij 2008). The specimens were observed alive, photographed, and subsequently preserved in 96% ethanol. Material analysed in this study was deposited in the mollusc collection of Naturalis Biodiversity Center, Leiden, The Netherlands, and catalogued as RMNH.MOL. One specimen of P. ocellatus (RMNH.MOL.336426), collected in the Philippines, 4 November 1999 (Sta. CEB.01, Cebu Strait, E side of Olango Island 10°15'54"N 124°04'17"E, coll. BWH) was used for comparison of external characters with the new species in absence of the holotype of P. ocellatus (Figures 7b–d). The original drawings of van Hasselt (1824) are available in scientific archives of Naturalis Biodiversity Center and reprinted here (Figures 1a–c).

Figure 2. 

Map of West Papua and Papua, Indonesia, showing the area and locality where Plakobranchus papua specimens were collected: south Gam Island, Mangrove Creek.

To study the radula, the buccal masses of two specimens were dissected and immersed in 10% NaOH until the tissue surrounding the radulae was dissolved. The radulae were rinsed in distilled water and transferred to 70% ethanol. They were subsequently examined by means of a light microscope, photographed, mounted on stubs, and gold-sputter-coated for scanning electron microscopy (SEM). Eight specimens were dissected for anatomical studies. One specimen (Table 1) was dehydrated in a graded ethanol series, embedded in paraffin, and cut into 7 μm serial sections with a manual microtome for histological observations. Sections were stained with toluidine blue to identify structures on the basis of metachromasia.

The holotype (RMNH.MOL.336417) and two paratypes (RMNH.MOL.336418–336419) of Plakobranchus papua sp. n. were used for DNA analysis. A piece of foot tissue was extracted and treated according the DNeasy Blood and Tissue, spin-column protocol for the purification of total DNA. Primers (LCOI490–HCO2198) were used for the amplification of the mitochondrial gene COI (cytochrome c oxidase subunit, partial) region (Folmer et al. 1994).

PCR reactions were performed with Thermocycle Biometra T professional. All PCR reactions were carried out in 50 μl reaction volume, including 25 μl Q PerfeCTa®SYBR® Green FastMix® from Quanta BioScience Inc, (Gaithersburg, USA), 2 μl of each primer, 19μl of DPEC treated water and 2 μl of genomic DNA. Thermal cycling started with an initial melting step for 3 min at 94 °C, followed by 39 cycles at 94 °C for 15 sec, 50 °C for 30 sec, annealing using a temperature gradient from 50–60 °C for 1 min, 72 °C for 1 min, and a final elongation step at 72 °C for 5 min.

The PCR products were purified with a Gene JetTM PCR Purification kit (Fermentas Life Sciences Lithuania) and sequenced directly, using the same primers as for PCR. DNA sequences were obtained using the Big Dye terminator ver. 3.1 Cycle Sequencing kit (Applied Biosystems) and analysed with the automated sequencer ABI 3130 Genetic analyzer at the sequencing facility of the Department of Anthropogenetics at the University Medical Centre, Radboud University, Nijmegen. The sequences were analysed with the program Chromas Lite (Technelysium Pty Ltd.) and deposited in Genbank (KU934191–KU934193).

Phylogenetic analyses were carried out on a dataset of 81 sequences including two outgroup species (Table 1). All COI sequences of Plakobranchus available on GenBank were used, and three newly obtained sequences of P. papua sp. n. were added. Elysia ornata (Swainson, 1840) (AB758962) and Thuridilla carlsoni Gosliner, 1995 (GQ996681) were selected as outgroups (Bass and Karl 2006). Sequences were aligned using the Guidance server (ClustalW), resulting in an alignment score of 1.00 for the dataset (Penn et al. 2010). A model selection analysis was carried out in jModeltest (Posada 2008) to select the best-fit model based on AICc (corrected Akaike Information Criterion), rendering TrN + G as the best-fit model. A maximum likelihood analysis using the GTR + G model (1000 bootstraps) was carried out in Phyml 3.1 (Guindon et al. 2010) in the Seaview platform (Gouy et al. 2010) and a majority rule consensus tree was constructed. Bayesian inferences (3,000,000 million generations) were estimated in MrBayes 3.1.2 (Ronquist and Huelsenbeck 2003) using the GTR + G model (because of unavailability of the TrN + G model). Average standard deviation of split frequencies was 0.09851. A majority-rule consensus tree was constructed in MrBayes with a burnin of 25%, and visualised in FigTree 1.3.1 (http://tree.bio.ed.ac.uk/software/figtree/).

The web version of ABGD (Automatic Barcode Gap Discovery, Puillandre et al. 2012) was used to estimate the genetic distance corresponding to the difference between a speciation process versus intraspecific variation. Runs were performed using the default range of priors (pmin = 0.001, pmax = 0.10) using the JC69 Jukes-Cantor measure of distance. The analysis involved 25 nucleotide sequences. All ambiguous positions were removed for each sequence pair; there was a total of 657 positions in the final dataset.

Plakobranchus papua sp. n. differs not only from P. ocellatus as illustrated by van Hasselt (1824), but also from several other Plakobranchus colour varieties, which can be found at SeaSlugForum and NudiPixel. Colour variations of Plakobranchus and geographical distributions are presented in Table 2. Based on this data several species of Plakobranchus are to be expected. The new species differs from all other descriptions and illustrations of individuals ascribed to P. ocellatus by a clearly different colour pattern, with black rhinophores and tail, non-ocellated spots, and a foot sole without spots. Plakobranchus papua sp. n. also differs externally from all taxa that have been synonymised with P. ocellatus in the colouration, distribution of spots, and the absence of real ocellated spots. DNA analyses of COI mtDNA show a clear distinction between P. papua sp. n. and P. ocellatus s.l.

The original drawings of P. ocellatus made by van Hasselt (1824) do not show many details (Figure 1), but enough to separate it from P. papua sp. n. The dorsum and flanks of P. papua sp. n. have white spots, while P. ocellatus possesses ocellate spots with dark pigment surrounded by a yellow ring or surrounded by a black ring (Figures 1, 7). These large ocellate spots of P. ocellatus are also present all over the foot sole, while in P. papua sp. n. the foot sole is white without any spots. Furthermore, the ocellate spots of P. ocellatus are very abundant all over the dorsum in combination with broad transverse stripes of darker or lighter pigment. These ocellate spots are abundant on the dorsum of the preserved specimen (Figure 7c, d) and less abundant on the ventral part of the body (Figure 7e). The tail of P. papua is black, while in P. ocellatus only the margin of the tail has dark pigment. Bergh (1872) described the radula of P. ocellatus as consisting of 19 teeth, eight ascendant teeth, one not developed, one tooth at tip and nine descendant teeth, and between 90 to 100 teeth inside the ascus. The teeth possessed up to 14 strong denticles at each side. In P. papua the shape of the teeth appears more arched than the teeth of P. ocellatus described and figured by Jensen (1997a). In our specimens of P. papua the number of teeth is similar, and the lateral denticles ranged between 10 and 14 in number. In the SEM photos of P. ocellatus by Jensen (1992) at least 10 denticles can be counted. The illustrations presented by Bergh (1873: Plate XI figs 5, 6) for P. chlorophacus show 10–11 denticles. Within Sacoglossa, species like Elysia viridis (Montagu, 1804) are able to modify the teeth size and shape in response to changes in their diet (Jensen 1993) but it is not known if this ability also exists in Plakobranchus.

Wägele et al. (2010) discovered special glandular structures on the dorsoanterior to lateral parts of the pharynx in P. ocellatus. Similar glands were found surrounding the pharynx in P. papua, and stained dark blue when treated with toluidine blue.

Jensen (1997a) described a pair of copulatory bursae in the reproductive system of P. ocellatus, which were also found in P. papua sp. n. (Figure 5: gv).

When comparing the drawings of the P. ocellatus holotype (Figure 1a–c) and those by Marcus (1982) of specimens collected in the Farasan Islands, Red Sea (Figure 1m, n), differences can be observed between both sets of illustrations. The schematic drawing made by Marcus (1982) is based on one preserved specimen and shows some scattered large ocellated dorsal spots and a line of ocelli along the head (Figure 1m). In the P. ocellatus described by van Hasselt (1824) the renopericardial area is fused together and appears as a nearly rounded prominence, while in the specimen illustrated by Marcus (1982) the pericardium is separated with respect to the renal prominence and this renal prominence is very elongated (Figure 1n), which is not confirmed in the text (Marcus 1928). Since Red Sea specimens externally resemble P. ocellatuss.s. (Yonow 1990, 2008), these differences need more study before they can be considered reliable. In this regard, it is noteworthy that Plakobranchus papua possesses a shorter renopericardial area as that illustrated by van Hasselt for P. ocellatus.

Marcus (1982) used the absence of dorsal vessels in Plakobranchus to separate this genus from Pattyclaya (which possesses clear dorsal vessels) and Jensen (1992) also used this same absence to separate Plakobranchus from Elysia. Plakobranchus species have tiny anastomosing lamellae connecting the external lamellae to the most internalones. In P. papua fine short dorsal anastomosing lamellae connect the anteriormost external lamellae with the most internal ones, which are in direct connection with the renopericardial area. This is endorsed by Wägele et al. (2010), who stated that the Plakobranchidae are characterized as follows: “a number of pericardial vessels are found branching from the pericardium along the dorsal surface of the body and parapodia”.

The new COI sequences were used in a phylogenetic analysis together with the 76 sequences available in GenBank under the name P. ocellatus from various Indo-West Pacific localities (Figure 8). The phylogenetic analysis shows that P. papua with Plakobranchus sp. 1 from Krug et al. (2013) are in the same cluster, and separated from all other P. ocellatus sequences. Solely based on COI, P. papua could be considered the same species as Plakobranchus sp. 1 of Krug et al.(2013). This would imply that P. papua is not restricted to West Papua but also occurs at Panglao Island (Philippines) and Sulawesi (Indonesia).

The ten groupings retrieved in our ABGD analysis and phylogeny reconstruction (Figure 8) agree with the results of Krug et al. (2013). Thus, although morphological analyses indicate a single highly polymorphic P. ocellatus (Jensen 1992, 1993, 1996, 1997a, 1997b, 2007), molecular analyses indicate a species complex containing ten candidate species. Christa et al. (2013: fig. 1b) identified a specimen from Guam as P. ocellatus, but according to our Plakobranchus colour pattern and distribution review (Table 2), this specimen resembles P. ocellatus var. A more closely. Unfortunately the GenBank accession numbers cannot be linked to their illustrated specimens. Krug et al. (2013) stated that some differently coloured Japanese specimens could also be considered separate species. The eleven varieties of P. ocellatus found in the web and publications (one of which is P. papua sp. n.) could not be linked because Krug et al. (2013)did not provide enough information on the morphology and colouration pattern of their specimens, nor did they include illustrations. Based only on phenotypes of images retrieved from the internet, eleven varieties of Plakobranchus were distinguished, including P. ocellatus (sensu stricto) and P. papua sp. n. (Table 2). All Plakobranchus species and varieties were recorded from the Indo-West Pacific, including the Red Sea.

Further research on the genus Plakobranchus based on morphology and DNA is necessary to unravel the phylogenetic and taxonomic relationships within this genus, with clear illustrations included of the external morphology. The present study shows that in situ photographs are indispensable to morphologically differentiate between Plakobranchus species because of their highly variable colour patterns, since ethanol-preserved specimens tend to lose much of their original colouration (Figure 7). Museum collections as a source for taxonomic work, studies on historical ecology, or other kinds of biodiversity studies (Rainbow 2009, Hoeksema et al. 2011, Rocha et al. 2014), may not be sufficient without photographic documentation of the specimens in situ. It is suggested that illustrations be included in DNA research as well as included in GenBank data.

The expeditions were part of the research programme “Ekspedisi Widya Nusantara (E-Win)” of PPO-LIPI. The research permit application was sponsored by Prof. Dr. Suharsono of PPO-LIPI. LIPI granted research permit 6559/SU/KS/2007 for the fieldwork in the Raja Ampat Islands, West Papua. We want to thank Max Ammer and staff of Papua Diving at Kri Eco Resort and Raja Ampat Research and Conservation Center (RARCC) for logistic support at Kri Island, Raja Ampat. Dr. Mark Erdmann (Conservation International, Sorong, West Papua) provided useful advice and encouragement. Financial support was given by the van Tienhoven Stichting, the Schure-Beijerinck Popping Fund (KNAW), the Leiden University Fund, the Jan Joost ter Pelkwijk Fund (Naturalis), and the Alida M. Buitendijk Fund (Naturalis). We thank Erik-Jan Bosch (Naturalis) for making the maps. The authors thank three anonymous reviewers for critical remarks and the editor Nathalie Yonow for their help in improving the paper.