Recognizing two new Hippolyte species (Decapoda, Caridea, Hippolytidae) from the South China Sea based on integrative taxonomy

Hippolyte shrimps exhibit abundant biological diversity and display great ecological significance in seaweed bed ecosystems. Dozens of Hippolyte specimens were collected from Hainan Island and the Xisha Islands in the South China Sea. Detailed examination indicates that some of these specimens represent new Hippolyte species. Based on morphological, genetic, and ecological data, Hippolyte chacei sp. nov. and H. nanhaiensis sp. nov. are described. H. chacei sp. nov. was collected from the Sargassum sp. biotope in Hainan Island and is distinguished from congeners by its unique mandible and the dactylus of the third to fifth pereiopods; this species has a basal position in the Indo-West Pacific species clade in the phylogenetic tree which is reconstructed based on 16S rRNA gene. H. nanhaiensis sp. nov. was collected from the biotopes of Galaxaura sp. or Halimeda sp. in the Xisha Islands, and it differs from congeners in a series of characters associated with rostrum, scaphocerite, antennular peduncle, and spines on the dactylus of the third to fifth pereiopods. Additionally, it is sister to H. australiensis in the phylogenetic tree. A key to identifying mature female Hippolyte species of the Indo-West Pacific and neighboring seas is provided.


INTRODUCTION
Shrimps of the genus Hippolyte Leach, 1814 display high diversity in morphology, coloration, and ecological habits. They occur mainly in tropical and temperate oceans, although some species, such as Hippolyte varians Leach, 1814, are known from the Arctic Circle (D'Udekem D'Acoz, 2007). Most Hippolyte species inhabit seaweed beds, but some are obligate or facultative symbionts of other organisms, such as gorgonians and crinoids (D'Udekem D'Acoz, 2007;Marin, Okuno & Chan, 2011). The taxonomy, phylogeny, and biology of Hippolyte taxa have attracted considerable attention in recent collected using handheld nets while snorkeling. After being photographed, specimens were preserved in 95% ethanol. Dissection and illustrations were carried out using Nikon stereo-and compound microscopes (SMZ 1500 and AZ100). Measurements and length ratios were calculated following D'Udekem D'Acoz (1996). All specimens are deposited in the Marine Biological Museum of the Chinese Academy of Sciences (MBM), in the Institute of Oceanology of Chinese Academy of Sciences, Qingdao, China.

Molecular data and analysis
Total genomic DNA was extracted from pleopods of specimens using a QIAamp DNA Mini Kit (QIAGEN, Hilden, Germany), following manufacturer instructions. Extracted DNA was eluted in 100 ml of double-distilled H 2 O (ddH 2 O). Partial sequences of 16S rRNA genes were amplified from the diluted DNA via polymerase chain reaction (PCR). Reactions were carried out in a 50-ml volume containing: 25 ml Premix Taq (TaKaRa Taq TM Version 2.0 plus dye; TaKaRa, Kusatsu, Japan), one ml each of forward and reverse primers (10 mM), respectively, three ml DNA template, and 20 ml ddH 2 O. Primers 16S-AR/1472 (5′-CGCCTGTTTATCAAAAACAT-3′/5′-AGATAGAAACC AACCTGG-3′) was used (Crandall & Fitzpatrick, 1996). The PCR profile involved: 3 min at 94 C for initial denaturation, 35 cycles of denaturation at 94 C for 30 s, annealing at 52 C for 40 s, elongation at 72 C for 50 s, and final extension at 72 C for 10 min. PCR products were purified using a QIAquick Gel Extraction Kit (QIAGEN, Hilden, Germany), and bidirectionally sequenced using the same primers with an ABI 3730xl Analyzer (Applied Biosystems, Foster City, CA, USA). Sequences were checked and proofread by ContigExpess 6.0 (a component of the Vector NTI Suite 6.0).
Molecular data (Dataset S2), including 37 sequences of 16S rRNA genes, were aligned using MUSCLE 3.8 (Edgar, 2004). Highly divergent and poorly aligned sites were omitted from alignment according to Gblocks 0.91b (Castresana, 2000). The best-fitting nucleotide base substitution model (GTR+I+G) for the alignment data was determined by Modeltest 3.7 (Posada & Crandall, 1998). A maximum likelihood tree was constructed using PhyML 3.1 (Guindon & Gascuel, 2003) with 1,000 bootstrap reiterations. A Bayesian inference tree was constructed using MrBayes 3.2 (Huelsenbeck & Ronquist, 2001). Markov chains were run for 2,000,000 generations, sampled every 2,000 generations; the first 25% trees were discarded as burn-in, after which remaining trees were used to construct the 50% majority-rule consensus tree and to estimate posterior probabilities. Genetic distances were calculated using the Kimura 2-parameter model in MEGA 7.0 (Kumar, Stecher & Tamura, 2016).

Ecological data
The biotope (mainly the algal colony) in which a shrimp lived was recorded on capture.
The following abbreviations are used: CL, carapace length, the length from the posterior orbital margin to the posterior dorsal border of the carapace; Coll., collector (s).
The electronic version of this article in portable document format will represent a published work according to the International Commission on Zoological Nomenclature (ICZN), and hence the new names contained in the electronic version are effectively published under that Code from the electronic edition alone. This published work and the nomenclatural acts it contains have been registered in ZooBank, the online registration system for the ICZN. The ZooBank Life Science Identifiers (LSIDs) can be resolved and the associated information viewed through any standard web browser by appending the LSID to the prefix http://zoobank.org/. The LSID for this publication is urn:lsid: zoobank.org:pub:1186ACB4-410C-4061-BE93-97CE040F0702. The online version of this work is archived and available from the following digital repositories: PeerJ, PubMed Central, and CLOCKSS. Description. Outline robust (Fig. 1). Ratio lateral length/height of carapace 1.56-1.72. Rostrum long, slightly shorter than carapace, distinctly overreaching antennular peduncle, nearly reaching to the end of scaphocerite. Rostrum without lateral carina, superior border slightly concaved, unarmed in the female specimens (Figs. 2A and 2B) and armed with only one proximal tooth in the male specimen (Fig. 2C); inferior border slightly convex, armed with four teeth in the distal half length. Carapace smooth and glabrous, location of supraorbital spine behind of the posterior orbital margin; tip of antennal spine slightly overreaching inferior orbital angle; tip of hepatic spine falling short of anterior edge of carapace. Inferior orbital angle strongly produced, knob-like (Figs. 2B and 2D). Branchiostegal margin with a distinct notch. Pterygostomian region rounded, strongly produced (Figs. 1 and 2B).

Taxonomy
Abdominal segments smooth ( Fig. 1). Third abdominal segment geniculately curved. Ratio dorsal length/height of the sixth abdominal segment 1.95-2.10. Telson (Fig. 2E) longer than the sixth abdominal segment, posterior margin rounded, armed with eight strong spines, outer spines smallest, medial two longest, without intermediate spinule or seta; dorsal surface armed with two pairs of spines situated on distal 0.31-0.35 and 0.59-0.63 telson length. Eye ( Fig. 2A) well developed, tip of cornea nearly reaching to the end of first segment of antennular peduncle when extended forward; unpigmented part of eyestalk longer than broad; cornea semispherical, distinctly shorter than unpigmented part of eyestalk.
Second pereiopod (Fig. 4C) slightly overreaching the end of third maxilliped when extended forward. Carpus with three subsegments, first subsegment 1.70-1.85 times as long as second subsegment, third subsegment slightly longer than or subequal to first subsegment; first subsegment 2.45-2.56 times as long as wide, second subsegment 1.08-1.12 times as long as wide, third subsegment 2.06-2.12 times as long as wide. Cutting edges of chela not denticulate, outer margin of fingers with long simple setae, tip of fixed finger and dactylus armed with three acute spines, respectively (Fig. 4D).
Third to fifth pereiopods long and robust. Third pereiopod (Fig. 4E) reaching to the distolateral spine of scaphocerite when extended forward; dactylus with 13-16 spines, the last two to three subdorsal spines distinctly shorter than the neighboring ones (Fig. 4F); propodus 5.56-5.62 times as long as wide, armed with six to seven pairs of spines on ventral margin; carpus 2.66-2.73 times as long as wide, armed with one proximal lateral spine; merus 5.58-5.62 times as long as wide, armed with three lateral spines. Ratio length of third pereiopod dactylus with longest apical spine/length of propodus 0.45-0.49; Coloration. Generally light brown over body (Fig. 5A), with few tawny stripes on carapace and faint tawny spots on abdomen.
Biotope. All specimens were captured among gulfweed (Sargassum sp.) at depths of one to three m. Numerous Hippolyte cf. ventricosa were captured simultaneously.  Description. Outline stout (Fig. 6). Ratio lateral length/height of carapace 1.49-1.58. Rostrum distinctly shorter than carapace, reaching to or slightly overreaching the end of antennular peduncle. Rostrum without lateral carina, superior border straight, armed with one or two teeth in proximal position (Figs. 7A-7D); inferior border armed with one subdistal tooth in female specimens (Fig. 7C), unarmed or only with a tiny distal notch in male specimens (Fig. 7D). Carapace smooth and glabrous. Location of supraorbital spine behind of the posterior orbital margin. Antennal spine small, slightly overreaching inferior orbital angle. Hepatic spine reaching to or slightly overreaching anterior edge of carapace. Inferior orbital angle produced, knob-like (Figs. 7B and 7C). Branchiostegal margin sinuous. Pterygostomian region rounded, strongly produced (Fig. 7C).
Abdominal segments smooth (Fig. 6), without or with few long plumose setae on tergum. Third abdominal segment geniculately curved. Ratio dorsal length/height of the sixth abdominal segment 1.91-2.08. Telson (Fig. 7E) longer than sixth abdominal segment, posterior margin rounded, armed with eight strong spines, outer spines smallest,  propodal segment bearing few long plumose setae; carpus longer than broad, shorter than merus. Third maxilliped (Fig. 8F) reaching to mid-length of scaphocerite when extended forward; exopod reaching to 0.72-0.79 of antepenultimate segment; ultimate segment (excluding apical spine) of endopod 1.61-1.78 times as long as penultimate segment, distal half armed with six to nine strong spines; antepenultimate segment slightly shorter than the last two segments combined. First pereiopod (Fig. 9A) shortest among pereiopods, oblique, nearly reaching to mid-length of the scaphocerite when extended forward. Ventral margin of basis, ischium, and merus furnished with long plumose setae. Terminal margin of carpus cotyloid. Cutting edges of chela non-denticulate, outer margin of fingers with long simple setae, tip of fixed finger with three acute spines, tip of dactylus with four acute spines (Fig. 9B).
Second pereiopod (Fig. 9C) slightly overreaching the distolateral spine of scaphocerite when extended forward. Carpus with three subsegments, first subsegment 2.13-2.26 times as long as second subsegment, third subsegment slightly shorter than first subsegment; first subsegment 2.65-2.76 times as long as wide, second subsegment 1.08-1.16 times as long as wide, third subsegment 1.76-1.83 times as long as wide. Cutting edges of chela not denticulate, outer margin of fingers with long simple setae, tip of fixed finger with three acute spines, tip of dactylus with four acute spines (Fig. 9D).
Third to fifth pereiopods long and robust. Third pereiopod (Fig. 9E) reaching beyond terminal blade of scaphocerite by dactylus when extended forward; dactylus with 8-10 spines, all spines in ventral and apical position (none in dorsal or subdorsal position), with two apical spines larger than others (the ultimate one longer but thinner than the penultimate one) (Fig. 9F); propodus 6.98-7.12 times as long as wide, armed with four to six pairs of spines on ventral margin; carpus 2.96-3.14 times as long as wide, armed with one proximal lateral spine; merus 6.45-6.63 times as long as wide, armed with two lateral spines. Ratio length of third pereiopod dactylus with longest apical spine/length of propodus 0.42-0.46; ratio length of third pereiopod dactylus with longest apical  0.25-0.29 times as long as exopod. Second pleopod (Fig. 9M) of male specimen with endopod about 0.79-0.86 times as long as exopod; appendix masculina with eight apical setae, about 0.41-0.47 times as long as appendix interna (Fig. 9N). Remarks. These specimens had the following features: (1) first article of the antennular peduncle with one distolateral tooth, and fifth pleonite no dorsolateral tooth; (2) carapace length of mature females among 1.8-3.3 mm, and total length among 13-24 mm; (3) rostrum distinctly overreaching the end of the antennular peduncle but falling short of scaphocerite apex, superior border with one to two teeth and inferior border with one to five teeth; (4) incisor process of mandible with five to six teeth; (5) scaphocerite 2.79-3.38 times as long as wide; (6) dactyli of the third to fifth pereiopods with two large apical spines, but the longest apical spine never exceeding the half-length of dactyli properly, the ratio of the longest spine of dactylus/length of dactylus without spines among 0.33-0.41; (7) specimens displaying various colorations (Figs. 5F-5J). These features differ from those described for H. acuta, H. australiensis, H. ngi, H. singaporensis, and H. nanhaiensis sp. nov., but are similar to the morphological characters of H. ventricosa (referring to the redescription of D'Udekem D'Acoz, 1999). More than four cryptic or pseudocryptic species were recently detected using molecular markers, which were also morphologically very similar to H. ventricosa (De Grave et al., 2014;Terossi, De Grave & Mantelatto, 2017). Therefore, it is not clear which specimens represent the true H. ventricosa; 16S rRNA or other genetic data derived from the H. ventricosa topotype are expected to resolve this issue.
Coloration and Biotopes. Specimens captured among Thalassia sp. were generally bright green over body (Fig. 5F), or green over body with pink or brown stains on carapace, abdomen, and telson (Fig. 5J); specimens (Figs. 5G-5I) captured among Sargassum sp. are generally sandy brown or reddish brown over body, with or without white stains on carapace, abdomen, and telson. All specimens were captured at depths of one to three m.

DISCUSSION
Hippolyte chacei sp. nov. is distinguished from all other valid Hippolyte species by the unique dactylus of the third to fifth pereiopods. This type of dactylus has previously reported only for specimens attributed to H. ventricosa, such as those reported from the Malayan Archipelago (Holthuis, 1947), Madagascar (Ledoyer, 1970), and Hawai (Hayashi, 1981), which D'Udekem D'Acoz (1996) considered represented undescribed species. Our work, based on molecular data, confirms this suspicion. In the 16S rRNA phylogenetic tree (Fig. 10), H. chacei sp. nov. (two specimens) form an isolated branch clustered in the subbasal position of the Indo-West Pacific clade (Terossi, De Grave & Mantelatto, 2017). Additionally, the average genetic divergence between H. chacei sp. nov. and other Hippolyte species is 20.8%, which is slightly greater than the average interspecific genetic divergence of 20.5% (calculated from the 30 Hippolyte species in our study).
Specimens attributed to H. ventricosa from the Malayan Archipelago and Madagascar by Holthuis (1947) and Ledoyer (1970), respectively, are very similar to H. chacei sp. nov. in morphology. We speculate that they are conspecific, but this speculation requires a detailed examination of their specimens. Hayashi (1981) stated that the mouthparts of Hawaiian specimens were similar to those of H. edmondsoni and H. jarvisensis, of which distinctly differ from those of H. chacei sp. nov.; moreover, difference is also apparent in the position of hepatic spine. Those specimens reported by Hayashi (1981) Acoz, 1999).
They all have the first article of the antennular peduncle with one distolateral tooth, fifth pleonite without dorsolateral teeth, and third to fifth pereiopods with normal dactyli. H. nanhaiensis sp. nov. differs from H. acuta, H. australiensis, and H. ngi by its shorter rostrum (reaching to or only slightly overreaching the end of the antennular peduncle vs. distinctly overreaching the end of the antennular peduncle). H. acuta is further distinguished from H. nanhaiensis sp. nov. by its particularly long eyestalk (Stimpson, 1860;Hayashi & Miyake, 1968;Yanagawa & Watanabe, 1988;D'Udekem D'Acoz, 1996). H. australiensis is further distinguished from H. nanhaiensis sp. nov. by its rostrum, which has a sharp lateral carina, and also by the dactylus of the third to fifth pereiopods, which have four large apical spines (D'Udekem D'Acoz, 2001). H. ngi differs from H. nanhaiensis sp. nov. by its hepatic, which overreaches the anterior edge of carapace by distal half length, and also by the dactylus of the third to fifth pereiopods, which have three large apical spines (Gan & Li, 2017b).
According to D'Udekem D'Acoz (1999), H. ventricosa also has two large apical spines on the dactylus of the third to fifth pereiopods, although the apical spines of H. nanhaiensis sp. nov. are proportionally longer. The ratio of the longest spine of the dactylus/length of the dactylus without spines is 0.53-0.58 in H. nanhaiensis sp. nov., but it is only 0.35 in H. ventricosa. The rostrum of H. ventricosa distinctly overreaches the end of the antennular peduncle, but it only reaches to or slightly overreaches the end of the antennular peduncle in H. nanhaiensis sp. nov. The scaphocerite of H. ventricosa is 3.10 times as long as wide, but it is 2.19-2.38 times as long as wide in H. nanhaiensis sp. nov. The total length of the H. ventricosa syntypes reaches 17 mm (D'Udekem D'Acoz, 1999), nearly two times longer than the largest H. nanhaiensis sp. nov. Furthermore, the two species inhabit different ecological niches. H. ventricosa lives among Zostera sp. or Padina sp., and may also be found among Sargassum sp., but H. nanhaiensis sp. nov. was found only among Galaxaura sp. or Halimeda sp., and no other congeners were found in these biotopes.
In the 16S rRNA phylogenetic tree (Fig. 10), H. nanhaiensis sp. nov. (two specimens) form a clade with H. ventricosa group-sp. 4 (Terossi, De Grave & Mantelatto, 2017), with this clade being a sister to H. australiensis. The average genetic divergence between H. nanhaiensis sp. nov. and other Hippolyte species is 22.5%, which is greater than the average interspecific genetic divergence (20.5%). 16S rRNA sequence alignment reveals H. nanhaiensis sp. nov. to be identical to, or to has a single nucleotide base difference from specimen of H. ventricosa group-sp. 4 (KX588916). Therefore, the specimen attributed to H. ventricosa group-sp. 4 and H. nanhaiensis sp. nov. might ultimately prove to be conspecific.

CONCLUSIONS
As noted by D'Udekem D' Acoz (1999 and2001), the systematics of Indo-West Pacific Hippolyte is extremely complicated, even though this region is considered the origin center of the genus (Terossi, De Grave & Mantelatto, 2017). Much of this taxonomic confusion stems from a lack of knowledge of several species, such as H. proteus, H. kraussiana, and H. acuta, and the plasticity in morphological characters of deemed taxonomic importance. Our study demonstrate the length of the rostrum relative to the antennular peduncle, the ratio of width to height of the scaphocerite, the position of the hepatic spine, and the features of the dactylus of the third to fifth pereiopods to be taxonomic value. A preliminary key for the indentification of mature female of the genus Hippolyte occurring in the Indo-West Pacific and neighboring seas is provided. This key only contains valid species listed in WoRMS (http://www.marinespecies.org); the cryptic or pseudocryptic H. ventricosa species are temporarily pooled as "H. ventricosa" sensu lato.
Key to mature female of Hippolyte for the Indo-West Pacific and neighboring seas