First records and two new species of sipunculans (Sipuncula) from the Southern Mexican Pacific

Sipunculans are a poorly studied group in the Tropical Eastern Pacific. For the Southern Mexican Pacific (SMP) there is only one record of a sipunculan species. The main objective of this work was to determine the species composition of the phylum Sipuncula present in the SMP. The study area covered three Mexican states: Guerrero, Oaxaca and Chiapas; specimens from 28 localities were examined from both intertidal and subtidal zones. A total of 551 specimens were reviewed, from which 11 species were identified. Five of them have previously been recorded in the Tropical Eastern Pacific (TEP): Apionsoma (A.) hespera comb. nov., A. (Edmondsius) pectinatum, Aspidosiphon (A.) elegans, Phascolosoma (P.) puntarenae and Themiste (T.) hennahi; four species are similar to nominal species: Sipunculus (S.) cf. polymyotus, Siphonosoma cf. vastum, Siphonosoma cf. cumanense and Phascolosoma (P.) cf. perlucens; and two new species are described: Aspidosiphon (Paraspidosiphon) cutleri sp. nov. and Aspidosiphon (Paraspidosiphon) pastori sp. nov. A checklist and an identification key for all sipunculan species from the TEP are presented. The checklist includes 53 taxa, 25 of which are questionable records. This work generated 11 new records of sipunculans in the SMP and five new records in the TEP.


Introduction
Sipunculans are a small group of unsegmented vermiform, coelomate and protostome marine worms (Cutler 1994;Murina 1984) and are considered close to or even within the annelid group (Staton 2003;Struck et al. 2007;Dordel et al. 2010). They are commonly known as 'peanut worms' or 'star worms' (Schulze et al. 2019), and they have a body divided in a retractable introvert and trunk. Sipuncula earlier comprised 320 species (sensu ; however, the revisions of Cutler & Cutler (Cutler 1979;Cutler & Cutler 1982, 1985a, 1985b, 1986, 1988, 1989, 1990 reduced the number to 150 valid species (Cutler 1994). Traditionally, the phylum has included two classes, four orders, six families and 17 genera (Cutler & Gibbs 1985; but Kawauchi et al. (2012) proposed a new classifi cation based on a molecular phylogenetic analysis. Subsequently, Lemer et al. (2015) corroborated these results. The new classifi cation now includes six families and 16 genera, leaving out the classes and orders.
In the Tropical Eastern Pacifi c (TEP), most sipunculan species are reported from northwestern Mexico and Costa Rica, as well as a few other few localities from Colombia, Panama and the Galapagos Islands (Gray 1828;Grube 1858;Keferstein 1866Keferstein , 1867Fischer 1895;Steinbeck & Ricketts 1941;Fisher 1952;Cutler & Cutler 1980;Salazar-Vallejo 1983;Fonseca & Cortés 1988;Cutler et al. 1992;Dean 2001;Cantera et al. 2003;Fonseca et al. 2005;Spongberg 2006;Yupanqui et al. 2007;Melwani & Kim 2008;Dean et al. 2010;Hermoso-Salazar et al. 2013;Morales-Zárate et al. 2016). Also, no identifi cation key or updated list of the sipunculans from TEP exist. The presence of this group in the Southern Mexican Pacifi c (SMP) is completely unknown (Bastida-Zavala et al. 2013); a single record of one new species in the SMP was made by Silva-Morales et al. (2019). Therefore, the main objective of this work was to determine the composition of species of the phylum Sipuncula present in the study area.

Material and methods
The revised specimens are from the Laboratorio de Sistemática de Invertebrados Marinos (LABSIM), Universidad del Mar, campus Puerto Ángel, Oaxaca, Mexico. The totality of specimens is derived from 28 localities in three coastal Mexican states: Guerrero, Oaxaca and Chiapas (Table 1). Most of the material is from intertidal and subtidal zones and some was collected during free diving on different substrates such as sand and rock, mainly coralline rocks. Collected specimens were fi xed in 10% formalin and preserved in 70% ethanol. To identify the species, the identifi cation keys of Cutler (1994) and Fisher (1952) were used. Internal and external anatomy was observed and described using a Zeiss stereo microscope; to examine the hooks and papillae, semi-permanent slides were prepared and visualized in a Zeiss compound microscope. The specimens were deposited in the Sipunculan Section of the Scientifi c Collection with the accession number OAX-CC-249-11 of the Laboratorio de Sistemática de Invertebrados Marinos, Universidad del Mar (UMAR), campus Puerto Ángel, Oaxaca, Mexico. Additional specimens were examined from the Polychaeta Collection of the Laboratorio de Poliquetos (ECOSUR), Chetumal, Quintana Roo, Mexico with the number QNR.IN.021.0497. The checklist and identifi cation keys were assembled following an exhaustive review of the literature.  Sipunculus (Sipunculus) cf. polymyotus Fisher, 1947 Figs 1-2

Description
Trunk 80 mm in length (Fig. 1A). Introvert 10% of total length; with triangular papillae and associated microbivalves ( Fig. 1B-D). Transparent and iridescent body. Tentacular membrane with two large ventral lobes and two smaller dorsal lobes with margin greatly subdivided (Fig. 1E). Longitudinal and circular musculature in bands, with 49 longitudinal muscle bands (LMBs). Gut with post-esophageal loop. Two pairs of retractor muscles, each ventral retractor muscle attached to six LMBs starting from second band after ventral nerve cord; each dorsal retractor muscle attached to fi ve LMBs starting from 15 th band after ventral nerve cord ( Fig. 2A). Most LMBs split and double in posterior end. Brain with approximately ten digitate processes, large, conspicuous, leaf-like fl aps and long strings (Fig. 2B). A pair of brownish nephridia with irregular knobby surfaces, about 10% of trunk length (Fig. 2C). Anus with small prominences surrounding its boundary (Fig. 2D). Nephridiopore anterior to anus. Spindle muscle not attached to body wall posteriorly.

Morphological remarks
The prominences of the anus are conspicuous. This morphological feature was not described by Cutler (1994), or even by Fisher (1947). The prominences are not related with the elevation of the anus or the fi xation process.
Once more specimens of Sipunculus (S.) cf. polymyotus become available, the name S. natans could be revalidated. This hypothesis is based on some recent papers where the use of molecular data has rejected the supposed wide distribution of some species (Staton & Rice 1999;Kawauchi & Giribet 2010, 2014Schulze et al. 2012;Johnson et al. 2016;Silva-Morales et al. 2019).
Approximately 80 peripheral tentacles of equal length (Fig. 3C). Without hooks. Musculature divided in 20 anastomosing longitudinal muscle bands (Fig. 3B). Two pairs of retractor muscles originating at same level along anterior-posterior axis. Dark brown nephridia occupying less than 10% of trunk. Villi present in contractile vessel. Spindle muscle attached to body wall posteriorly.

Remarks
This species is similar to Siphonosoma cumanense (Keferstein, 1867), but the widespread distribution is unlikely.

Distribution
Santa Cruz Beach, Oaxaca, Mexico.

Description
Trunk 150 mm in length in preserved specimens (Fig. 4A), 300 mm in living specimens. More than 60 fi liform tentacles. Introvert with dispersed orange papillae (Fig. 4C) and reddish region with rectangular

Remarks
The species is similar to Siphonosoma australe australe (Keferstein, 1865), described from Sydney, Australia, and to S. vastum (Selenka & Bülow in Selenka, 1883), described from Jaluit, Marshall Islands (Table 3). Siphonosoma cf. vastum differs from S. australe australe in that the former species has multiple rectal caeca in 90% of the rectum, which are absent in the latter. According to , S. vastum is able to disperse across the entire Tropical Pacifi c Ocean. However, the nominal species is different from S. cf. vastum from Mexico in many diagnostic characters, mainly the coverage of the multiple rectal caeca in the rectum and the attachment of the muscles in the body wall (Table 2). This species is a possibly a new species; however, more specimens are needed to describe it.

Description
Specimen damaged. Trunk white in color, 15 mm in length (Fig. 5A). Introvert 5 mm in length, without hooks or pigmented collar. Tentacular crown with four asymmetrical stems, giving appearance of six primary stems (Fig. 5B). Peripheral tentacles with pigment spots. Longitudinal muscle of body wall gathered in uniform continuous layer. Nephridiopores opening at level of anus. A pair of retractor muscles attached to body wall at 75% of trunk length (Fig. 5C). Five tubular extensions, very fragile (not illustrated but observed).

Remarks
The type locality from Themiste hennahi is Peru. Unfortunately, Gray (1828) did not specify the locality. The species has also been recorded from Laguna Beach, Southern California, and Ensenada, Baja California (Fisher 1952), under different names (Cutler 1994). We advise the revision of the specimens from the Subtropical Eastern Pacifi c to confi rm the widespread distribution of T. hennahi.

Remarks
Although there are no morphological characters to support that the populations of Phascolosoma perlucens from the Caribbean and the Eastern Pacifi c are different species, the molecular data provides strong evidence of a differentiation between both populations (Kawauchi & Giribet 2010); therefore, we propose to consider this species as being near to the nominal species. It is possible that the specimens recorded here and other records from the Tropical Eastern Pacifi c (Fisher 1952;Dean 2001;Dean et al. 2010) are a new species.

Habitat
Intertidal and subtidal (20 m); as epibionts of Pinctada mazatlantica, in dead coral, between the fouling of pier piles, in macroalgae, in a sabellariid reef, in mangrove roots and in sand.

Distribution
Southern Mexican Pacifi c, from Ixtapa, Guerrero to Puerto Chiapas.

Description
Trunk 10 mm in length (Fig. 7A). Light brown trunk with uniform dome-shaped papillae, some with black pigment randomly distributed ( Fig. 7D-E). Introvert with bands of dark pigmentation and more than 100 complete and incomplete rings of hooks. Hooks with clear streak expanded near midpoint of vertical and middle of horizontal portions ( Fig. 7F-G). Secondary tooth almost indistinct, protuberance of streak short. Fourteen tentacles encircling nuchal organ (Fig. 7C). Longitudinal muscles of body wall gathered into anastomosing bands. Four retractor muscles. Nephridia about 40% of trunk length, open at same level as anus (Fig. 7B). Spindle muscle attached posteriorly.

Remarks
Cutler & Cutler (1990) considered Phascolosoma (P.) puntarenae from Puntarenas, Costa Rica, as a synonym of P. (P.) nigrescens from Fiji. Nevertheless, we found a difference between the hooks of the type material of P. nigrescens illustrated by Selenka (1883) and those in the illustrations of Fisher (1952) of P. puntarenae from Baja California, Costa Rica and Panama. The hooks of P. nigrescens have a conspicuous secondary tooth and the protuberance of the streak is a sharp point, whereas the hooks of P. puntarenae do not have a well-developed secondary tooth and the protuberance of the streak is a fl attened point (Fig. 8F-G).

Habitat
Intertidal to subtidal (15 m); in sabellariid tubes, as epibionts of Pinctada mazatlanica, in dead coral and in Porites.

Description
Trunk 8 mm in length, spindle-shaped (Fig. 9A). Introvert seven times trunk length. Posterior end of trunk with numerous distinctive papillae (Fig. 9C-D). More than 40 rings of hooks, with 7-8 basal spinelets. Spinelets longer than principal tooth (Fig. 9E). Body wall with continuous muscle layers. Four retractor muscles equidistant from ventral nerve cord near middle of trunk. Nephridia bilobed, with .

Remarks
We reinstate the name Phascolosoma hespera in the correct genus as the new combination Apionsoma (Apionsoma) hespera. The most similar species to A. (A.) hespera comb. nov. is A. (A.) misakianum (Ikeda, 1904) from Misaki, Japan (Table 4). Fisher (1952) illustrated type material and he showed the morphological and ecological differences between the two species. Cutler (1979) Fig. 8D) and other localities of Japan (Cutler et al. 1984: 300-301; Fig. 8E), with the hooks of the specimens revised in this study (Fig. 8A) and those of specimens recorded from Peru as Golfi ngia (Mitosiphon) hespera (Amor 1975; Fig. 8B-C). Staton & Rice (1999) stated that there is strong reproductive isolation between the northern and southern populations of the A. misakianum species complex, another argument to consider that A. hespera is a valid species and likely could be part of this species complex.

Habitat
Intertidal, in Cerianthus tubes, in rocks with Phragmatopoma and Perumytilus.

Description
Trunk 20 mm in length (Fig. 10A). Trunk light yellowish to dark brown in color, with dome-shaped papillae around trunk and base of introvert, scattered on body. Body wall wrinkled, giving a rough appearance (Fig. 10B). Introvert 2-4 times as long as trunk. Tentacles digitiform, encircling nuchal organ. Thirty rings of hooks in anterior region of introvert, with seven basal spinelets (Fig. 10C). Two pairs of retractor muscles. A pair of bilobed nephridia. Longitudinal bands of body wall show slight anastomosis. Nephridia with 40% of trunk length. Spindle muscle not attached to posterior end of trunk (Fig. 10A).

Remarks
With only a few records at the time, Cutler (1994) concluded that the distribution of Apionsoma (Edmondsius) pectinatum in the Eastern Pacifi c went from Baja California to Panama, so its distribution in the rest of TEP could only be inferred. This new record from the Southern Mexican Pacifi c breaks the disjunctive distribution in the Tropical Eastern Pacifi c for this species.

Description
Trunk 8 mm in length, white (Fig. 11A). Introvert three times the trunk length. Anal shield without grooves, caudal shield absent. Fourteen rings of bidentate compressed hooks (Fig. 11C-D) followed by scattered unidentate, dark, conical hooks (Fig. 11E-F). Longitudinal muscle of body wall in uniform continuous layer. A pair of retractor muscles originate at about 50% of trunk length. A pair of nephridia, unilobed, at 60% of trunk length. Spindle muscle attached posteriorly (Fig. 11B).

Remarks
The original description by Chamisso & Eysenhardt (1821) did not include a description of the hooks; thus, we cannot compare our specimens with the original description. Nevertheless, a revision is necessary because it is likely that the population of the Marshall Islands is different from the one of the SMP. This species has been recorded from Cocos Island, Costa Rica ). We did not fi nd differences between the species in the SMP and the Mexican Caribbean (Fig. 11G-J).

Distribution
Widespread and common in the Indian and western Pacifi c Oceans, from south-central Japan to northern Australia to Hawaii, the Red Sea, and Israel. In the Caribbean from northern Brazil to the Florida Keys and Bermuda (Cutler 1994). In the Eastern Pacifi c from the South Mexican Pacifi c to Costa Rica (Fonseca & Cortés 1998;Dean et al. 2010).

Etymology
In memory of Edward Cutler, expert in sipunculans and the principal source of inspiration for this work.

Description
Trunk 30 mm in length, white (Fig. 12A). Anal and caudal shield brown. Sixteen longitudinal grooves in anal shield. Caudal shield also with grooves ( Fig. 12C-D). Units of anal region distributed in longitudinal lines (Fig. 12E). More than 100 rings of unidentate proximal hooks, smaller than distal hooks (Fig. 12F) followed by scattered pyramidal hooks. Longitudinal muscles of bo dy wall gathered into anastomosing bands. A pair of retractor muscles, fused for most of their length. Nephridia unilobed, occupying 50% of the trunk. Spindle muscle bifurcate d near its anterior end (Fig. 12B).

Remarks
The species that are most similar to Aspidosiphon (Paraspidosiphon) cutleri sp. nov. are A. (P.) coyi de Quatrefages, 1865, described from the Indian Ocean, and A. (P.) laevis de Quatrefages, 1865. The type material of these two species was described by Saiz-Salinas (1983). The main differences between these species are the following: Aspidosiphon (P.) coyi has bidentate hooks and disperse units in the anterior region of the trunk; Aspidosiphon laevis lacks bidentate hooks and the units on the anterior region of the trunk are dispersed; nephridiopores are at the same level as the anus in both these species; on the other hand, A. (P.) cutleri sp. nov. only has unidentate hooks and the units are distributed in lines on the anterior region of the trunk; the nephridiopores are posterior to the anus. It is likely that the record of A. (P.) laevis from Costa Rica  corresponds to this new species; however, it is necessary to review the specimens.

Etymology
This new species was named as "pastori" after Mr Pastor Silva Cruz, for his invaluable support and inspiration to work in the fi eld of marine science.

Remarks
The species that is most similar to Aspidosiphon (Paraspidosiphon) pastori sp. nov. is A. (P.) fi scheri (ten Broeke, 1925) from Caracas Bay, Venezuela; however, they are different in the shape of the hooks and the longitudinal muscle bands. Aspidosiphon (Paraspidosiphon) pastori sp. nov. has hooks with an oblique streak, while A. fi scheri has hooks with a straight streak (Fig. 13G). Aspidosiphon (Paraspidosiphon) pastori sp. nov. has 14-16 longitudinal muscle bands, while A. fi scheri has 18-19 LMB.

Distribution
South Mexican Pacifi c, from Ixtapa Island, Guerrero, to La Entrega Beach, Oaxaca.

Discussion
The taxonomy of Sipuncula remains complicated. There are too few morphological characters to easily separate species. The study of sipunculans requires caution when reviewing records of species from any locality. Cutler (1994) produced a list of 150 species, most of those cosmopolitan or of wide distribution, and a long list of synonyms. This led to cases of specimens from far distant populations being assigned the same name. For example, Dean (2001), Dean et al. (2010) and Hermoso-Salazar et al. (2013) recorded some species from Pacifi c Costa Rica and Mexico with names from Europe, the Indian Ocean, Western Pacifi c and Caribbean. The idea that most marine invertebrates have wide distributions has been questioned recently (Hutchings & Kupriyanova 2018). Regarding sipunculans, there are many studies rejecting these wide distributions, supported by careful morphological and molecular investigations (Staton & Rice 1999;Kawauchi & Giribet 2010, 2014Schulze et al. 2012;Johnson et al. 2016;Silva-Morales et al. 2019; Silva-Morales 2020).
The importance of being cautious implies giving due importance to the type locality in the records of species. In this study, the records of Themiste (T.) hennahi and Apionsoma (Edmonsius) pectinatum were based on the correspondence of the diagnostic characters with the original description and the proximity of the type locality with the sample locality. In the case of Aspidosiphon (A.) elegans we recorded the nominal species; nevertheless, we suggest other types of studies to confi rm or reject this wide distribution because we did not fi nd any morphological distinction. With Siphonosoma cf. vastum, Sipunculus (S.) cf. polymyotus and Phascolosoma (P.) cf. perlucens we found morphological differences or molecular data to consider them "confer" and not the nominal species; the possibility remains that these could be new species for the tropical eastern Pacifi c. On the other hand, two names were reestablished: Apionsoma (A.) hespera comb. nov. and Phascolosoma (P.) puntarenae. For similar cases we suggest reviewing the 320 previously existing species names (e.g., , because other valid names could be buried in the synonyms of presumed widely distributed species. The new species, Aspidosiphon (P.) cutleri sp. nov. and A. (P.) pastori sp. nov. can be easily separated from the others of the same genus and subgenus by observing the internal characteristics of the hooks and the longitudinal muscle bands.
This study is the fi rst work about sipunculans in the Southern Mexican Pacifi c, more than 1000 km of coastline for which no records of this phylum existed (Bastida-Zavala et al. 2013). This work will improve our knowledge of the sipunculans in the Tropical Eastern Pacifi c because it gathers information that was scattered, with a commented list and an identifi cation key to help future studies.