Redescriptions of Nereis oligohalina (Rioja, 1946) and N. garwoodi González-Escalante & Salazar-Vallejo, 2003 and description of N. confusa sp. n. (Annelida, Nereididae)

Abstract Type material of several polychaete species described by Enrique Rioja from Mexican coasts are lost, and the current status of some species is doubtful. Nereis oligohalina (Rioja, 1946) was described from the Gulf of Mexico, but it has been considered a junior synonym of Nereis occidentalis Hartman, 1945, or regarded as a distinct species with an amphiamerican distribution. On the other hand, Nereis garwoodi González-Escalante & Salazar-Vallejo, 2003, described from Chetumal Bay, Caribbean coasts, could be confused with Nereis oligohalina. In order to clarify these uncertainties, Nereis oligohalina is redescribed based on specimens from the Mexican Gulf of Mexico, including a proposed neotype; further, Nereis garwoodi is redescribed including the selection of lectotype and paralectotypes, and Nereis confusa sp. n. is described with material from the Gulf of California. A key for the identification of similar species and some comments about speciation in nereidid polychaetes are also included.


Introduction
Among the non-marine polychaetes, the family Nereididae de Blainville, 1818 has the largest number of brackish and freshwater species (61), and 31 out of these species occur in estuaries and coastal lagoons ). Of the 40 species of nereidid species recorded from the Gulf of Mexico (Fauchald and Solís-Weiss 2009), seven are reported in brackish or freshwater areas, and among the 10 species belonging to Nereis, only N. oligohalina (Rioja, 1946) is reported from estuaries ).
Enrique Rioja documented extensively the Mexican polychaetes from Pacific or Atlantic coasts in a series of papers; unfortunately, his material is lost, and most species require designation of neotypes (Salazar-Vallejo 1989). Rioja (1946) dealt with three estuarine nereidids from Veracruz, Mexico; he regarded one as a known species, Neanthes succinea (Leuckart, 1847), and the two others were described as new: Neanthes oligohalina and Lycastopsis tecolutlensis. The former species is now regarded as belonging in Alitta, but it differs from the North Sea species (T.F. Villalobos-Guerrero, pers. comm.); L. tecolutlensis was regarded as a junior synonym of Namanereis amboinensis (Pflugfelder, 1933), nowadays a widespread species (Glasby 1999); and N. oligohalina has been regarded as amphiamerican (Dean 2001), or restricted to Atlantic coasts Lana 2003, Liñero-Arana andDíaz-Díaz 2007). However, other amphiamerican species have been shown to be restricted to one coast or the other, often resulting in description of new taxa (e.g. Carrera-Parra and Salazar-Vallejo 2011; Yáñez-Rivera and Carrera-Parra 2012). On the other hand, the Caribbean species N. garwoodi González-Escalante & Salazar-Vallejo, 2003 could be confused with N. oligohalina, and without an updated description of the latter, a synonymy can be anticipated.
However, Nereis oligohalina and N. garwoodi have morphological differences that separate them. In this contribution our objectives were first, to redescribe N. oligohalina based upon material collected from Veracruz, including topotypes, and to propose a neotype. Second, to redescribe N. garwoodi to clarify some doubtful features in the original description, and to select lectotype and paralectotypes specimens from the syntype series. Third, to recognize what has been regarded as N. oligohalina from the Mexican Pacific as a distinct species and describe it as N. confusa sp. n., based upon material from the Gulf of California. Further, a key to identify similar Nereis species and comments about species delimitation are also included.

Material and methods
Specimens studied are deposited in the Reference Collection of El Colegio de la Frontera Sur, Chetumal (ECOSUR) including ethanol-fixed specimens (ECOSUR-OH), and in the Polychaetological Collection of the Universidad Autónoma de Nuevo León (UANL).
Topotypes of N. oligohalina from Estero Casitas, Nautla were examined, including additional specimens that were recently collected along the coast of Veracruz, Mexico, Gulf of Mexico. Specimens of N. confusa sp. n. from Bahía La Paz and Bahía de Los Ángeles, Gulf of California were found in unidentified material in ECOSUR, now formally deposited. To assess variation in paragnath morphology, specimens of N. pelagica Linnaeus from England (ECOSUR P2840), and Pseudonereis sp. from the Caribbean Sea (ECOSUR P1170), were also examined.
The best preserved specimens were used for designation of type material. Some specimens were fixed and preserved in 96% ethanol directly, otherwise the specimens were fixed with formalin and later preserved in 70% ethanol.
For analysis of variation, type materials and a number of non-type specimens for each species were measured. Total body length (TL), length up to chaetiger 3 (L 3 ) or 10 (L 10 ), width at same chaetigers (W 3 and W 10 ), number of chaetigers (nC), and length of longest tentacular cirrus (rTC) were measured with a millimeter rule under the stereomicroscope. TL was measured from palp tips to the end of the pygidium, W 3 and W 10 were measured excluding parapodia. Also, paragnath numbers in all areas were counted; if the pharynx was not everted, a ventral dissection was made, and areas VII-VIII were described as if the pharynx was exposed. With these results simple descriptive statistics (mean, range and standard deviation) were performed. As different fixation methods were used, a Mann-Whitney U-test was used for evaluating if there were significant differences in body measures as a result of fixation method.
For the microscopical observation of parapodial features and chaetae, right-side parapodia along body were removed and mounted in semi-permanent slides; the photographs were made with a digital camera, and distal-view drawings of parapodia were included to depict spatial disposition of ligules and chaetae. Descriptions of pigmentation patterns were included, because they are consistent and useful for recognizing the three species; other authors have noted their utility for identifying cryptic nereidid species (Read 2007, Glasby et al. 2013. Bakken and Wilson's (2005) terminology was followed for describing parapodia, and Bakken et al. (2009) for paragnaths. Parapodia from both atokes and epitokes (if available) were illustrated to show parapodial changes along the body. For determining dorsal cirri length and position of attachment, we considered the beginning of the dorsal or notopodial ligule to be approximately at the same vertical position of attachment as the ventral cirri; therefore, the relative length of dorsal cirrus was measured from that position toward the distal end of dorsal or notopodial ligule. The dorsal cirrus was considered basally attached if placed at, or near such a position, or medially attached if it was displaced more distally from that position.
Also, the reach of the dorsal cirrus and its relative length in respect of the dorsal or notopodial ligules were considered as separate attributes. For determining the reach of the dorsal cirrus, the tips of the both dorsal cirrus and dorsal or notopodial ligules were taken into account; if the dorsal cirrus tip extended beyond the tip of the dorsal or notopodial ligule, then we report "dorsal cirrus extended beyond dorsal/notopodial ligule" rather than "dorsal cirrus longer than dorsal/notopodial ligule". On the contrary, if dorsal cirrus is shorter or not exceeding the dorsal or notopodial ligules, then we report "dorsal cirrus not extended beyond dorsal/notopodial ligule". In the species herein treated, length of dorsal cirri and length of dorsal or notopodial ligules were generally subequal, and the dorsal cirri change their attachment along the body, but not their length necessarily.
For the designation of lectotype and neotype, the International Code of Zoological Nomenclature (ICZN 1999) was followed. The designation of a neotype for N. oligohalina follows Article 75, and the designation of a lectotype for N. garwoodi follows Article 74 (ICZN 1999). The non-formal term 'paraneotypes' is used for figured topotypic specimens (Evenhuis 2008), and their utility has been pointed out elsewhere (Salazar-Vallejo 2011, Sendall andSalazar-Vallejo 2013).
Although the species described here have more attributes than those included in the current diagnosis of the genus (Bakken and Wilson 2005), the generic diagnosis was not modified because it first requires a redescription of the type-species, and a phylogenetic analysis with subsequent delimitation of the genus. Among the traditionally used features for descriptions and delimitation of Nereis species are some that are highly variable, especially paragnath number. Bakken et al. (2009) made a useful revision of paragnath morphology and introduced new terminology to standardize descriptions. In addition to conical paragnaths, the species described here present other types of paragnaths that are not currently included in the diagnosis of the genus. Conical paragnaths are pointed to various degrees, being more acute in the maxillary ring, especially on area II. The pyramidal paragnaths in N. oligohalina and N. garwoodi have quadrilateral bases but they can also be polygonal, having more defined surfaces in the latter species. Further, N. confusa sp. n. apparently has smooth bars on area IV, but a closer inspection confirms that this is an artifact because the bars are formed by lateral and basal fusions of some small conical paragnaths (Fig. 6H); these modified structures were regarded as melted paragnaths (Bakken et al. 2009). However, Glasby et al. (2011) suggested limiting use of the term for conical paragnaths mounted on a platelike basement as occurs in Neanthes pachychaeta (Fauvel, 1918), and Villalobos-Guer-rero and Carrera-Parra (2015) found paragnaths on a soft basement in A. acutifolia (Ehlers, 1901). Because neither a basement is present in N. confusa sp. n., we suggest the term 'merged' for paragnaths fused at the base but without formation of a plate.
In his revision of Pseudonereis Kinberg, 1865, Bakken (2007) introduced the term 'P-bar', which was later defined by Bakken et al. (2009) as "small bars having a protruding apex in one end of the bar"; and they can appear in areas II, III, IV and VII-VIII, often accompanied by conical paragnaths. The monophyly of Pseudonereis was supported by, among other characters, the presence of both P-bars and paragnaths in comb-like rows (Bakken 2007). The Nereis species studied herein have two main rows, each one with other two sub-rows; the anterior-most sub-rows are often aligned horizontally, while the posterior-most ones form a jagged line. The anterior-most rows have P-bars alternating with conical or pyramidal paragnaths in a similar way as in Pseudonereis (Fig. 6D-F), which has been also reported for Alitta (Villalobos-Guerrero and Carrera-Parra 2015). Therefore, P-bars are not an exclusive feature of Pseudonereis as Bakken et al. (2009) concluded.
Fully transformed male with pharynx everted, jaws amber with 10 teeth, inner edge toothed throughout. Maxillary ring: I = 8 cones in triangle, II = 30-32 cones in arc, III: 40 cones in rectangle, IV: 28-28 pointed cones in arc. Oral ring: V = 1 cone, VI: 4-4 pyramids in diamond, VII-VIII: 46 in two irregular rows, pyramids alternating with small cones in most-anterior row, pyramids with similar size alternating in most-posterior row.
In fully transformed males pygidium with anus surrounded by rosette of papillae ( Variation. The results of the analysis of body variation and paragnath numbers are summarized in Tables 1 and 2. The effect of fixation techniques on the shape of specimens have been recently evaluated by Oliveira et al. (2010) for Laeonereis acuta (Treadwell, 1923). The authors concluded that techniques of fixation can influence the shape and body proportions, especially if specimens were not previously relaxed, leading to erroneous identifications.
In the case of N. oligohalina, fixation with 96% ethanol clearly affected the anterior portion of specimens with strong contraction of first segments, hence the tentacular cirri can reach more posterior chaetigers but without modifying their lengths, reaching up to chaetiger 14 (Table 1). Differences are not significant for L 10 , W 3 and W 10 (P = >0.05), but significant for L 3 and rTC (P = <0.001, P = <0.0001). Maximum rTC for formalin specimens was preferred for the identification key because the data were less variable (Table 1). Nevertheless, these differences would not cause misidentification, because parapodial topology is not affected appreciably. Also pigmentation is very useful for recognizing the species.
In the maxillary ring, area I showed the least variation (Fig. 6J), and in oral ring areas V and VI rarely vary in one paragnath only (Fig. 6K), such that these areas can be regarded as the most stable ones. The fingerprint-like pattern starts in chaetigers 10-11 (Fig. 6L); it is size-independent, but in smaller specimens this pattern is faint; however, as shown below it is absent in the two other species. The divergence between parapodial rami reported by Rioja (1946) is evident in posterior chaetigers but only in some specimens, forming a furrow (Fig. 6O). Glandular masses appear more visible and also on neuropodial ligules (Fig. 6Q); perhaps these glands fade in specimens fixed with formalin, such as the type material. In mature specimens, the natatory region starts in chaetiger 15-17 in males, and 24-25 in females. (Rioja, 1946) is considered as a widespread species and even amphiamerican, but this stems from taxonomic confusion and the lack of type material. Designation of a neotype for N. oligohalina was considered necessary because there are no type specimens and being a problematic species, there must be an objective definition for it (ICZN 1999, Art. 75.1). Consequently, a neotype has been selected, described and illustrated (ICZN 1999, Art. 75.3.3); this neotype fits the original description by Rioja (1946) (ICZN 1999 Art. 75.3.5). Because Rioja did not designate holotype, his material became syntypes and the species had two type localities (ICZN 1999 Art. 73.2.3, 76.1): Estero de Larios, Tecolutla, and El Cocal, Estero Casitas, both in Veracruz, Mexico. Although topotypic specimens from Estero Casitas are available, they are in poor condition, and therefore better specimens collected from nearby Actopan River were preferred once they were shown to conform to the same species (ICZN 1999, Recomm. 75A). The proposed neotype was collected in a similar environment and on oysters, as the original specimens (ICZN 1999, Art. 75.3.6); but the neotype locality is modified accordingly (ICZN 1999, Art. 76.3). The neotype was deposited in ECOSUR (ICZN 1999, Art. 75.3.7), including 'paraneotypes' and part of the additional material.

Nereis oligohalina
The first synonymy involving these species was made by Pettibone (1956); she considered N. p. occidentalis different from N. pelagica and raised it to species level as Nereis (Nereis) occidentalis Hartman. Also, she regarded Neanthes oligohalina as a junior synonym of N. occidentalis being regarded as a variety. The detailed description provided by Pettibone allowed us to recognize differences in comparison to the Laguna Madre, Texas specimens. She recognized slight but important differences among these variants, mainly in paragnath number in areas V and VI; adding the relative size of neuropodial ligules in middle and posterior chaetigers. In the same work, she determined that Nereis largoensis Treadwell, 1931 was a junior synonym of N. pelagica, and that other material identified as N. largoensis based upon material examined by Treadwell corresponds to N. occidentalis (Pettibone 1956). We follow, however, González-Escalante and Salazar-Vallejo (2003), who concluded that these three species are not synonyms.
Nereis oligohalina differs from N. occidentalis in some diagnostic features. In N. oligohalina there are 8-15 paragnaths on area I and 1-2 on area V, whereas in N. occidentalis there are 2-3 paragnaths on area I and no paragnaths on area V. Further, in N. oligohalina neuropodial ventral ligules are 2-3 times longer than neuroacicular ligules, but in N. occidentalis neuropodial ventral ligules from posterior chaetigers are shorter than neuroacicular ones. Regarding chaetae, and as an additional difference, in N. oligohalina the notopodial homogomph falciger has the distal tooth less developed than in N. occidentalis.
Another synonymy was made by Day (1973), who regarded N. pelagica occidentalis Hartman and N. occidentalis fide McCloskey as junior synonyms of Nereis falsa de Quatrefages, 1865; however, in the list of synonyms of these two species he did not include Pettibone (1956), therefore N. oligohalina was not considered by him. Nereis falsa, has a rather complex or confusing delineation; according to Fauvel (1923) it differs from similar species by having different numbers of paragnaths in areas I and V, and different proportions in parapodial ligules. In fact, N. falsa is another species regarded as widely distributed and requires a critical revision and we could anticipate a restriction of its distribution to the Mediterranean region because its type locality is the Black Sea.
On the other hand, N. oligohalina has been recorded along American Atlantic coasts from northeastern Brazil, chiefly in ecological (community assemblages on Spartina alterniflora and mangroves), or population studies (secondary production and population dynamics), as well as part of taxonomic or genetic studies (Amaral et al. 2012). A detailed record was made by Santos and Lana (2003); unfortunately, a commentary and one plate was based upon specimens collected in Todos Los Santos Bay, and indicated that their material agrees with the original description regarding neuropodial ventral ligules in posterior parapodia, and the feature was less developed in specimens from other localities but were regarded as the same because the prostomial pigmentation and paragnath number remained constant (Santos and Lana 2003). Lana et al. (2006) however, considered their previous record as a probable misidentification requiring a revision. Liñero-Arana and Díaz (2007) recorded N. oligohalina from Venezuela in La Restinga Lagoon, Margarita Island, associated with Crassostrea rhizophorae (Guilding), and recognized that their specimens resembled Brazilian ones and differed from those described by Rioja, mainly in parapodial morphology. These two publications pointed out the need for a revisionary work, recognizing N. oligohalina as valid species, but that their specimens were probably not the same as those described from Mexico because of parapodial features; at least the specimens from South America were regarded as a different species that should be clarified elsewhere.
Records of N. oligohalina from the Eastern Tropical Pacific Berkeley 1958, 1960) belong to a new species described below.
Habitat. The species is associated with red mangrove Rhizophora mangle and with oysters. It has been found in Gulf of Mexico estuaries, including Tecolutla, Casitas-Nautla and Actopan (these estuaries have sand bars in their respective mouth rivers), and from coastal lagoons such as Mandinga and La Mancha. These systems have direct connection with the sea, some with seasonal closure of their mouths, with polyhaline to mesohaline waters (Lara-Domínguez et al. 2011).
The neotype and associated specimens were found in Crassostrea virginica (Gmelin) reef, in the Actopan river mouth. The specimens studied by Rioja (1946) from the Tecolutla estuary were found on mangrove roots covered by cirripedians, whereas specimens from the Casitas-Nautla estuary were collected between oysters (possibly C. virginica) and mytilids as Ischadium recurvum (Rafinesque) (reported as Mytilus recurvatus (sic) by Rioja); as Rioja indicated, N. oligohalina specimens cohabit with an Alitta species in the Actopan river mouth. This species has been reported as N. occidentalis Hartman, together with Polydora websteri Hartman, as epifauna of C. virginica (Ruiz-Guerrero and López-Portillo Guzmán 2006), and on Rhizophora mangle roots (Ruiz and López-Portillo 2014), from La Mancha.
Distribution. Restricted to the southwestern Gulf of Mexico. Description. Lectotype complete (ECOSUR 0065), atokous female, damaged with incisions at level of chaetigers 1, 15, and 30. Body tapering, 36 mm long, 1.7 mm wide, 95 chaetigers. Body pale, pigmentation faint, brown rectangle present dorsally on middle of anterior chaetigers, striated, discoloring toward end of body, lateral pale lines in anterior chaetigers only, oocytes present. Prostomium with brown pigment along inner margins of palps, two lines extending from antennae toward anterior pair of eyes separated by a longitudinal pale area, and two oval lateral patches; peristomium slightly pigmented, pale lines present (Fig. 3A).
Parapodial cirri pattern: Dorsal cirri longer than upper dorsal ligules throughout body; basally inserted on anterior region, displaced medially in midbody region, becomes subdistal in posterior chaetigers. Ventral cirri longer than neuropodial ligules in a few anterior chaetigers, progressively reduced throughout body; basally inserted on anterior region, barely migrating ventrally throughout body.
Male with pharynx everted, jaws amber with 9 teeth. Maxillary ring: I = 6 pointed cones in triangle, II = 19-20 pointed cones in arc, III = 28 pointed cones in rectangle, IV = 22-19 pointed cones in arc. Oral ring: V = 1 pointed cone, VI = 4-3 pyramids in diamond, VII-VIII = 42 in two irregular rows, P-bars alternating with small pyramids in most-anterior row, pyramids alternating with cones with similar size in most-posterior.
Notopodial dorsal ligules subconical, longer than ventral ones in male, subequal in female; notopodial ventral ligules subconical, developing a large ventral lamella in males only, with a round projection. Neuroacicular ligules subconical, shorter than notopodial ventral ones; postchaetal lobes developing into flabellate lamellae with a round projection in dorsal edge in males, small lamellae in females, progressively in-  creasing in size and decreasing in far posterior segments; neuropodial ventral ligules digitate, basally attached to neuroacicular ones. Ventral cirri subulate, slightly longer than neuroacicular ligules, with two basal lamellae of different sizes; dorsal cirri wider than ventral ones.
Variation. The results of the analysis of body variation and analysis of paragnath numbers are summarized in Tables 1 and 2. The width measurements reported here differ from the original description because chaetiger width without parapodia was used, instead of measuring them including parapodia. The arrangement and number of paragnaths have similar ranges as those reported for N. oligohalina (Fig. 6B, E; Table 2); however, area I has a larger range, and the arrangement is somewhat variable, often in a triangle (Fig. 6I). Also, paragnaths are more robust than in N. oligohalina and N. confusa sp. n.
Regarding pigmentation, the striated rectangle seen in lectotype is more conspicuous in some specimens (Fig. 6M), which is also sometimes present in N. oligohalina, but the color is much more intense whereas the fingerprint-like pattern of the latter species was not observed. In mature specimens, the natatory region starts from chaetiger 17 only in males and 25-27 in females, which differs from the original description (22 in males and 21 in females). One specimen presented a duplicated ventral cirrus, but it was regarded as abnormal (Fig. 6R).
Remarks. González-Escalante and Salazar-Vallejo (2003) indicated that they had six atokes and two epitokes as syntypes. Five atokous syntypes were expected to be sent to four foreign museums, but were never dispatched. Further, these syntypes were not formally deposited and labeled, and parts of the descriptions and illustrations are too imprecise to enable separation of N. garwoodi from N. oligohalina.
In an attempt to redefine the species, a lectotype has been selected (ICZN 1999, Art. 74.1) to avoid future confusion; although the syntype series has better preserved specimens, the lectotype matches the original description and illustration, and was therefore preferred (ICZN 1999, Recomm. 74B). In order to ensure their validity, the term has been introduced in the material section and in the description (ICZN 1999, Art. 74.7.1, 74.7.3), and the lectotype has been described, illustrated and their data updated for its recognition (ICZN 1999, Art. 74.7.2, Recomm. 74C, 74E); the remaining syntypes are regarded as paralectotypes (ICZN 1999, Recomm. 74F). These specimens are deposited in ECOSUR.
Nereis garwoodi is closely allied with N. oligohalina, but they differ in some features in both atokous and epitokous forms, and in their habitats. In atokes, N. garwoodi never shows the dark brown coloration nor the fingerprint-like pattern found in N. oligohalina. The ranges of paragnath numbers of both species overlap and therefore are not useful to separate them, and the relative length of tentacular cirri would be useful if fixation method is the same (Table 1).
In N. garwoodi, both dorsal and neuropodial ventral ligules are twice as long as neuroacicular ligules in uniramous chaetigers, whereas in N. oligohalina these are subequal and slightly longer, respectively. Also, in N. garwoodi the neuropodial postchaetal lobes are visible in the anterior and midbody only, whereas in N. oligohalina they are visible throughout body.
Further, N. garwoodi has notopodial ventral ligules twice as long as neuropodial ventral ones in posterior chaetigers, whereas in N. oligohalina these ligules are subequal to each other; further, in N. oligohalina neuropodial ventral ligules are medially attached in posterior chaetigers, whereas in N. garwoodi they are basally attached throughout body. Moreover, in N. garwoodi notopodial homogomph falcigers have more teeth and they are narrower than in N. oligohalina; also, in N. garwoodi the blades of supra-acicular heterogomph falcigers become broader and shorter in posterior chaetigers, but this modification is not present in N. oligohalina.
In epitokes, N. garwoodi has modified, cattail-like dorsal cirri present in biramous chaetigers with the basal sections as long as distal ones, whereas in N. oligohalina basal sections are longer. Also, in general N. garwoodi have better developed lamellae in natatory chaetigers than N. oligohalina, especially the basal lamellae of the dorsal cirri, the lamellae of both notopodial ventral and neuroacicular ligules. Moreover, epitokal transformation is more pronounced in females of N. garwoodi than in females of N. oligohalina. On the other hand, N. garwoodi is associated with calcareous rocks, while N. oligohalina is associated with reef-building bivalves and the mangrove R. mangle. Table 2. Ranges, means and standard deviations (SD) in number of paragnaths in three Nereis species (r: right, l: left). Habitat. Chetumal Bay is a semi-closed, dynamic system linked to the Caribbean Sea by several freshwater tributaries, having a salinity gradient ranging 7-18 practical salinity units (psu) (Carrillo et al. 2009). The species bores into calcareous sedimentary rocks, building mucous tubes, and has been regarded as a sedentary herbivore (González-Escalante and Salazar-Vallejo 2003); to obtain the specimens, rocks must be broken. Although the Bay has extensive zones of mangroves, N. garwoodi has never been found among them.

Pharynx areas N. oligohalina (n=27) I II-r II-l III IV-r IV-l V VI-r VI-l VII-VIII
Distribution. Apparently restricted to Chetumal Bay. González-Escalante and Salazar-Vallejo (2003) report a gradient of decreasing abundance from the southern to the northern regions of the bay, probably related to organic matter load.
Parapodial cirri pattern: Dorsal cirri longer than upper dorsal ligules throughout body; basally inserted on anterior region, displaced medially on midbody and posterior regions. Ventral cirri as long as neuropodial ligules in a few anterior chaetigers, progressively reduced throughout body; basally inserted in anterior region, migrating ventrally throughout body.
Variation. The results of the analysis of body variation and paragnath numbers are summarized in Table 1 and 2. The arrangement and number of paragnaths is similar to that in N. oligohalina and N. garwoodi, but in N. confusa sp. n. paragnaths in areas III and IV are more numerous than in the other two species (Fig. 6C, F; Table 2), and cones have rounded tips; further, N. confusa sp. n. is the only species of the three with merged paragnaths.
All specimens examined show the same dorsal spotted pigmentation, but in some, especially the largest specimen, the middorsal spot disappears and only two discontinuous lines are visible along chaetigers 1-10; fingerprint-like or striated patterns were not observed. In mature specimens, the transformation starts in chaetiger 18 in males, 22 in females as previously noticed by Berkeley (1958, 1960).
Remarks. Nereis confusa sp. n. has been recorded as N. oligohalina; however, there are several differences between these two species. In N. confusa sp. n. the spotted pigmentation pattern extends up to chaetigers 10-14, and the jaws have 8 teeth restricted to the distal part of its inner edge, whereas in N. oligohalina the pale areas are replaced by fingerprint-like patterns from chaetiger 11, and its jaws have 11 teeth along its inner edge. Further, in N. confusa sp. n., both notopodial ligules and neuroacicular ligules are subequal to, or slightly longer than, neuropodial ventral ligules in midbody and posterior parapodia, whereas in N. oligohalina, they are twice as long as the neuroacicular ligules in midbody and posterior parapodia. On the other hand, N. confusa sp. n. has falcigers with broad blades, whereas in N. oligohalina they are narrower. Furthermore, in N. confusa sp. n. distal tooth of notopodial homogomph falciger is short and well developed, whereas in N. oligohalina it is longer and weakly developed.
The first records for N. confusa sp. n. (as N. oligohalina) from the Mexican Pacific were made by Berkeley and Berkeley; first, they reported Nereis (Neanthes) oligohalina males from Hipolito Bay (Berkeley and Berkeley 1958), and males and females from La Paz (Berkeley and Berkeley 1960). They argued that prostomium, anterior chaetigers and arrangement of paragnaths all matched Rioja's descriptions. In their brief comments, they indicated the start of the modified region or first epitokous parapodium (17 in males, 22 in females), and a spotted pattern of pigmentation in males. These features match with N. confusa sp. n. rather than N. oligohalina. Rioja (1962) cited Berkeley and Berkeley (1958), and recorded N. confusa sp. n. (as N. oligohalina) from El Mogote, Ensenada de La Paz, Baja California Sur; he mentioned a slight discrepancy in number of paragnaths in area I, and that paragnaths in the periphery of area III were larger than the rest of the group forming a borderline; also, he regarded the glandular parapodial masses as typical. Despite the fact that he did not provide more information, we regard his specimens as belonging to N. confusa sp. n. Other Mexican Pacific reports of N. pelagica occidentalis by Bastida-Zavala (1993, 1995 from nearby localities might also be conspecific. Dean (2001) reported N. oligohalina from Pacific Costa Rican coasts, noticed the problems in the taxonomic history of the species, and regarded it as different from N. oc-cidentalis. According to his description Costa Rican specimens differ from N. oligohalina in the number of paragnaths, mainly in areas I, III and IV. Also, in his specimens the longest tentacular cirri reached chaetiger 3, and the notopodial dorsal and neuropodial ventral ligules were subequal to or shorter than notopodial ventral and neuroacicular ligules throughout body, whereas in N. oligohalina the longest tentacular cirri reaches chaetiger 7, and their ligules are larger in midbody and posterior chaetigers. Likewise, Costa Rican specimens resemble N. confusa sp. n. and probably belong to the same species and this might also include the record from Cocos Island (Dean et al. 2012). Nevertheless, these records cannot be assigned to N. oligohalina unequivocally until specimens are evaluated.
Habitat. Holotype found on wrinkled penshell Pinna rugosa Sowerby, 1835, sponges and filamentous green algae; other specimens were found in sponges and green algae near the type locality. Specimens from Bahía de Los Ángeles were associated with another penshell, Atrina maura (Sowerby, 1835), and specimens from Estero El Yugo were found on filamentous green algae on mangrove roots. Bastida-Zavala (1995) found specimens on corals.
Distribution. Gulf of California, Eastern Pacific coasts of Mexico. Probably extends to Costa Rica, in shallow water.
Reproduction patterns and dispersal in Nereis species studied. Some authors have emphasized the utility of reproductive patterns in taxonomy for species discrimination in closely related taxa (Smith 1958, Clark 1977. Also, strategies for larvae survival would be relevant, especially because they determine larval transport and its dispersal potential. An interesting strategy is when only males form epotikes and females remain atokous or are barely modified, as in Alitta virens (Sars, 1835) and Websterinereis glauca (Claparède, 1870). In A. virens, males form epitokes but females remain atokous or present very slight changes, also females spawn into or in the opening of their burrows (Bass and Brafield 1972). In W. glauca, female transformation is reduced, and females produce mucous tubes to deposit and incubate their eggs, while males can swarm (Pettibone 1971).
Early studies considered that N. garwoodi presented a similar reproductive mode as A. virens or W. glauca, because the paralectotype female of N. garwoodi has a slight transformation; however, after further revision of additional material, a fully transformed female was found. We have no further details about its capacity for building mucous tubes or if females do not emerge to the water column or if gametes are retained in tubes.
In the material available of N. oligohalina, no completely transformed females were encountered, but perhaps with further sampling efforts they may appear. Another important consideration is the reduced size of males compared to females, even in the same sample; this has been noted for H. diversicolor (Bartels-Hardege and Zeeck 1990). In this case, N. oligohalina males can swarm while females remain inside cavities, as in A. virens and W. glauca; the fact that there are many large females filled with oocytes points toward this direction.
Reproduction modes in estuarine species play a crucial role in their dispersal because the formation of planktonic larvae can determine their distribution range. Bilton et al. (2002) proposed two life-cycle models for estuarine species that have larvae: export vs retention strategies. In the former, the adults respond to physical or biological factors by releasing gametes or larvae in the lower estuary; larvae are driven out from the estuary, mainly by tides, and later juveniles or adults return to the estuary. In the retention strategy, adults release their gametes or larvae in the upper estuarine areas, then they undergo early development in middle estuary; there, larvae have vertical migrations during circadian ebb-flood tidal regimes such that larvae are not exported but retained within the estuary (Bilton et al. 2002).
Based on the above, we hypothesize that N. oligohalina has an export strategy; it could disperse thanks to surface currents running parallel to the continental margin, and this would explain its presence along Gulf of Mexico estuaries. Similarly, distribution of N. confusa sp. n., with mainly marine habitats, could be due to current patterns along the Gulf of California.
However, sometimes the distribution patterns cannot be explained by currents and tidal dynamics. For example, De Jesús-Flores et al. (2015) determined that Laeonereis nota (Treadwell, 1941), described for Galveston, Texas, is also present in Chetumal Bay; L. nota spawn into their burrows, limiting their dispersal by currents. The explanation for this discontinuous distribution lies in passive dispersal through migratory birds, because they use nereidids as food (De Jesús-Flores et al. 2015). Similarly N. garwoodi could have a classical retention strategy, but a wide, fragmented distribution caused by migratory birds.
Further considerations. The present study demonstrates the need to encourage redescriptions of closely related and widely distributed species and, should it be necessary, the establishment of new species if there are conspicuous morphological differences.
Further, clarifying species delineation and distribution are urgent because they are essential for biogeography and phylogenetics. Reuscher and Shirley (2014) studied the distribution patterns of polychaetes from the Gulf of Mexico; a recent species list was used (Fauchald and Solís-Weiss 2009) and current taxonomy verified in WoRMS (Read and Fauchald 2015). They found that among all species recorded, 32% were cosmopolitan, 15% Pan-American and 9% Pan-Atlantic (N. oligohalina was regarded as Pan-American, which is incorrect as shown above). They concluded: "Most polychaete families are in need of global and regional revisions. Clear species boundaries have to be established by means of taxonomic research based on morphology and genetic analyses. Geographical ranges of species should be revised in order to eliminate false conclusions about distributions of species."   This species has not been found recently (Báez and Ardila 2003). 2 It has a doubtful amphiamerican status and there is no recent redescription. Here, descriptions by Renaud (1956) with specimens from Miami, and by Hoagland (1919) for N. glandulata from Porto Rico, were used. Hartman (1956) considered N. glandulata as a junior synonym of N. arroyensis Treadwell, 1901, but they differ in number of paragnaths on area VII-VIII.