Introduction

1Over one-half (32,000 to 33,300 species and counting) of the world’s living vertebrates are fish (Nelson et al., 2016; Froese & Pauly, 2018) and from these, about 40 % occur in shallow tropical waters (Moyle & Cech, 2004). In recent surveys, a total of 1,541 fish species have been recorded from the Gulf of Mexico (from Florida Keys to northwestern Cuba). This represents 64.3 % of the species occurring in the Western Central Atlantic Region (McEachran, 2009). The Gulf of Mexico can be divided in three biogeographical subregions (McEachran, 2009; McEachran & Fechhelm, 2005). The eastern subregion extends from the Florida Keys and western Cuba to Pensacola, Florida, which is the most species-rich subregion, with about 1,300 species. The northwestern subregion extends from Pensacola, Florida, to Cabo Rojo (Tamiahua Lagoon), Veracruz, and is second in species richness, with about 1,100 species. The southern subregion extends from Cabo Rojo, to Cabo Catoche, Yucatan, and is third place with about 900 species.

2Also, according to McEachran & Fechhelm (2005) and Briggs & Bowen (2012), two biogeographical provinces of the Warm Western Atlantic Region converge in the Gulf, the Carolina Province with warm-temperate biota, that occupies the area north of the tropical boundaries between Cape Romano, Florida, and Cape Rojo (similar to the limits of the eastern and northwestern subregions), and the Caribbean Province with tropical and subtropical biota, which extends from Cabo Rojo to Cabo Catoche, in Yucatan (same boundaries of the southern subregion, Figure 1).

3Many coastal fish species (which can vary between 50-95 %) use estuaries permanently or during some period of their lifecycle (Able, 2005), so the study of these ecosystems is vital for knowledge about fish diversity. In the southwestern Gulf of Mexico (Caribbean province), there are about 30 major estuarine systems, of which 19 are located in Veracruz, 4 in Tabasco, 3 in Campeche and 4 in Yucatan (Contreras, 2010). In most of them, fish species compositions have been studied, thus for the state of Veracruz 82 species have been recorded in the Pueblo Viejo lagoon (the northernmost case) (Castillo-Rivera et al., 2002, 2010, 2011), 136 in the Tamiahua lagoon (Franco-López & Chávez-López, 1992; Díaz-Ruiz et al., 2003), 179 in the Tuxpan-Tampamachoco estuarine system (Kobelkowsky, 1985; Pérez-Hernández & Torres-Orozco, 2000), 22 in the Chica-Grande lagoon system (Aguirre-León et al., 2014), 107 in the Camaronera-Alvarado system (Chávez-López et al., 2005) and 36 in the Ostión lagoon (Bozada & Chávez, 1986). For the State of Tabasco, there are records in the Carmen-Machona-Redonda lagoon system with 81 species (Reséndez-Medina & Kobelkowsky, 1991) and at the lower basin of the Grijalva-Usumacinta riverine system with 46 species (Arévalo-Frías & Mendoza-Carranza, 2015). In the State of Campeche, for the Sabancuy estuary, 33 species have been recorded (González-Solís & Torruco, 2013), and at Términos lagoon up to 154, although in recent years it has been noted a reduction in species richness to about 105 species (Ramos-Miranda et al., 2005). Finally, for the State of Yucatan 56 fish species have been reported in the Celestun lagoon (Arceo-Carranza et al., 2010), 81 for the “Bocas de Dzilam” lagoon (Arceo-Carranza & Vega-Cendejas, 2009) and 81 species for the Rio Lagartos (Vega-Cendejas & Hernández, 2004) (Fig. 1).

Figure 1. Geographic location of the main estuarine systems in the southern Gulf of México (Caribbean Province). The approximate surface (km²) of each one is also indicated in parentheses.

4Particularly for Sontecomapan Lagoon there are historical reports of its fish fauna, with sampling periods during 1975-1979 (Reséndez-Medina, 1983), 1984-1987 (Fuentes-Mata et al., 1989), 1985-1995 (Fuentes-Mata & Espinosa-Perez, 1997) and more recently during 2005-2006 (Rodriguez-Varela et al., 2010). In addition, there are studies comparing the fish fauna of several estuarine systems in the state of Veracruz which also provide information on this lagoon fish species richness (Lara-Domínguez et al., 2011; Abarca-Arenas et al., 2012).1

5In all these studies, the species richness reported for the system did not exceed 92. Therefore, the objective of the present study is focused on a comprehensive literature review of the fish fauna of the Sontecomapan Lagoon, in order to have a better estimate of its species richness.

Methodological procedures

6The present study was conducted through a comprehensive literature review, mainly based on field studies that were directly carried out in this system (Reséndez-Medina, 1983; Fuentes-Mata et al., 1989; Fuentes-Mata & Espinosa-Pérez, 1997; Rodríguez-Varela et al., 2010), as well as on two comparative checklists (Lara-Domínguez et al., 2011; Abarca-Arenas et al., 2012). The name of every species recorded in each of these studies is presented as originally written (Appendix) and its consistency was checked according to Froese & Pauly (2018), to detect problems of synonyms (each of two or more scientific names of the same rank used to denote the same taxon), probable misidentifications (the incorrect assignment of a specimen to a taxon) and misspellings (spelling incorrectly the scientific name of a species). Based on the above, the total number of species that has been historically reported for the Sontecomapan Lagoon was determined. The systematic sequence of orders and families follows the new fish classification of Nelson et al. (2016), which implies major changes in some taxa, principally at order level (in the present case, nine new orders, previously covered by only two).

7The Jaccard index (for presence/absence data) was used to assess the similarity degree of species among the different studies (Magurran, 2004). Then, a similarity matrix was generated and subjected to an UPGMA (unweighted pair-group method with arithmetic mean) in order to generate a similarity dendrogram (McGarigal et al., 2000) using MVSp. (Multivariate Statistical Package, version 3.1; Kovach, 2005).

8Since the number of species counted in a biodiversity study is usually a biased underestimate of total species richness (because many rare species are not detected), a species accumulation curve was employed to provide an estimate of the total fish richness in the system (Magurran, 2004) using data from each study as samples. For this purpose, the nonparametric estimators of Mao Tau, Chao 2, Jackknife 1, Jackknife 2, and Bootstrap were calculated using EstimateS software (Colwell, 2013).

General description of previous studies in the system

9The first formal study on the fish fauna in the system was developed by Reséndez-Medina (1983), and was conducted from 1975 to 1980, with 15 collection sites using a wide range of fishing gear (seine net, dip net, cast net and gill net), as well as a Chem Fish collector. In this study 26 families, 45 genera and 52 species were recorded (Appendix).

10Fuentes-Mata et al. (1989), with surveys conducted between 1984 and 1987, reported new records in accordance with previous studies, using a beach seine and recording 18 families, 30 genera and 31 species. In a similar survey, between 1985 and 1995, Fuentes-Mata & Espinosa-Pérez (1997), a study of the natural history of Los Tuxtlas, reported 32 families, 60 genera and 83 species (Appendix).

11More recently, during 2005-2006, Rodríguez-Varela et al. (2010) selected 17 sampling sites using stratified sampling. Fish larvae and juveniles were caught using a Renfro net and adults using a “rowing-boat”, with a mesh aperture of ¼ inch. These authors recorded 24 families, 41 genera and 52 species (Appendix).

12Finally, Lara-Domínguez et al. (2011) and Abarca-Arenas et al. (2012) surveys, recorded the fish species composition for many estuarine systems of Veracruz, which include extensive information from theses, symposia and institutional internal reports. Thus, for the Sontecomapan Lagoon Lara-Domínguez et al. (2011) recorded 33 families, 65 genera and 91 species, whereas Abarca-Arenas et al. (2012) recorded 34 families, 66 genera and 92 species (Appendix).

Taxonomic changes in recorded species

13The taxonomic status of some species previously recorded in the Sontecomapan Lagoon have changed to current date. These changes are listed below, following the sequence of orders and families proposed by Nelson et al. (2016) and, when necessary, species are sorted alphabetically within each family.

14Dasyatis sabina. Reséndez-Medina (1983), Lara-Domínguez et al. (2011), and Abarca-Arenas et al. (2012) recorded this species for the system, which is a name accepted by many other authors. However, according to recent studies on molecular analyses, combined with comprehensive morphological investigations of the Dasyatidae family (Last et al., 2016a; 2016b), the valid name for this taxon is Hypanus sabinus.

15Astyanax fasciatus. Fuentes-Mata & Espinosa-Pérez (1997) recorded this species, however Miller et al. (2009: 172) note that is no longer acceptable to believe that Astyanax fasciatus (restricted to Rio Sao Francisco, Brazil) occurs as far north as Mexico. Based on molecular analyses, morphological characters and patterns of geographic distribution of the genus Astyanax species (Strecker et al., 2004; Nelson, 2006; Ornelas-García et al., 2008; Miller et al., 2009; Hausdorf et al., 2011; Schmitter-Soto, 2016.), it is considered in the present study that this record could correspond to Astyanax aeneus (Günther, 1860), which is spread south of the Trans-Mexican Volcanic Belt, from Papaloapan river, southward to at least Costa Rica.

16Arius aguadulce. Although Reséndez-Medina (1983) recorded this species, it has been widely recognized that this name is a synonym of the valid name Cathorops aguadulce (Marceniuk & Menezes, 2007; Froese & Pauly, 2018).

17Arius felis. This species is recorded by Reséndez-Medina (1983), but it has been long recognized as a synonym of Ariopsis felis (Linnaeus, 1766) (Acero, 2002; Froese & Pauly, 2018) and with this name it is reported in Sontecomapan by Fuentes-Mata & Espinosa-Pérez (1997), Rodríguez-Varela et al. (2010), Lara-Domínguez et al. (2011), and Abarca-Arenas et al. (2012).

18Arius melanopus = Cathorops melanopus. Fuentes-Mata & Espinosa-Pérez (1997) recorded the Arius melanopus species, whereas Rodríguez-Varela et al. (2010) and Lara-Domínguez et al. (2011) recorded the Cathorops melanopus species. In this sense, it has been noted that Arius (or Cathorops) melanopus does not occur in Mexico and their records frequently are misidentifications of Cathorops aguadulce (Meek, 1904) (Marceniuk & Betancur-R, 2008; Miller et al., 2009; Froese & Pauly, 2018). Cathorops aguadulce is distinguished from all congeners by the fleshy papillae intercalated with gill rakers on the first two gill arches (vs. papillae absent in the remaining species, except in C. kailolae) (Marceniuk & Betancur-R, 2008; Marceniuk et al., 2012). Moreover, the recognized current distribution for C. aguadulce is from Panuco river basin, to Izabal Lake, in Guatemala, whereas C. melanopus is restricted to the freshwaters of Motagua river (in Guatemala and probably Honduras), and is commonly recognized as the only species of Cathorops described for the entire Caribbean portion of Central America (Ferraris, 2007; Marceniuk & Betancur-R, 2008; Miller et al., 2009). Although there has been an apparent resistance in Mexico to recognize these facts, especially by the ecologists, the species C. aguadulce is properly recorded in the Sontecomapan Lagoon by Fuentes-Mata et al. (1989), Fuentes-Mata & Espinosa-Pérez (1997), Tenorio-Colín et al. (2010), and Abarca-Arenas et al. (2012).

19Rhamdia quelen. Recorded by Abarca-Arenas et al. (2012) as a member of the Pimelodidae family, however, for a long time the Rhamdiinae and Heptapterinae subfamilies have been recognized within the Heptapteridae family (Bockmann & Guazzelli, 2003; Nelson 2006; Nelson et al., 2016).

20Gobionellus boleosoma. Recorded by Reséndez-Medina (1983), but based on phylogenetic analysis of osteological and cephalic lateralis characters, Pezold (2004a; 2004b) formally reassigns the genus to Ctenogobius. Thus, the current valid name for this species is Ctenogobius boleosoma (Jordan & Gilbert, 1882). Despite this, recent records for the Sontecomapan Lagoon still retain the incorrect name of G. boleosoma (Rodríguez-Varela et al., 2010; Lara-Domínguez et al., 2011; Abarca-Arenas et al., 2012). Moreover, based on phylogenetic works (Thacker, 2009; Betancur-R et al., 2013; Nelson et al., 2016), this species, along with other recorded in the Sontecomapan Lagoon, as Evorthodus lyricus, Gobioides broussonnetii, and Gobionellus oceanicus (formerly included in Gobionellinae subfamily), are now included within Oxudercidae (= Gobionellidae from some authors), in the Gobiiformes order.

21Gobionellus hastatus. Fuentes-Mata et al. (1989) recorded this species. However, based on morphological and allozymic analysis, Pezold & Grady (1989) placed Gobionellus hastatus, Girard, 1858 and Gobionellus gracillimus Ginsburg, 1953, in the synonymy of Gobionellus oceanicus (Pallas, 1770) (Pezold, 2004b). In the same way as in the previous case, although G. oceanicus has been recognized as valid for a long time, for the Sontecomapan Lagoon, Rodríguez-Varela et al. (2010), Lara-Domínguez et al. (2011) and Abarca-Arenas et al. (2012) persisted in the use of the invalid name of G. hastatus.

22Cichlasoma fenestratum. Reséndez-Medina (1983) and Fuentes-Mata & Espinosa-Pérez (1997) recorded Cichlasoma fenestratum (Günther, 1860), which corresponds to the original description of Chromis fenestrata Gunther, 1860. Later, this was recognized as Vieja fenestrata (Günther, 1860) by some authors (Kullander, 2003; Miller et al., 2009; Mercado-Silva et al., 2011) and recorded in the system with this name by Lara-Domínguez et al. (2011) and Abarca-Arenas et al. (2012). Although studies on molecular systematics of the genus Vieja (López-Fernández et al., 2010; McMahan et al., 2010) proposed the name Paraneetroplus fenestratus for this taxon, recent studies rediagnose it as Vieja fenestrata (McMahan et al., 2015). This species, along with other recorded in the Sontecomapan Lagoon, as Cichlasoma urophthalmum, Cichlasoma geddesi, Rocio octofasciata, Theraps irregularis, and Oreochromis sp. (Cichlidae family), are now included within the order Cichliformes (Nelson et al., 2016).

23Cichlasoma urophthalmus. There are numerous problems of species discrimination within the Cichlidae family and some of the most enigmatic cases include ‘Cichlasoma’ urophthalmus. Indeed, the generic allocation of this species has been uncertain for some time (Kullander, 2003; Concheiro-Pérez et al., 2007; Říčan et al., 2008; López-Fernández et al., 2010). Although this name has been recognized by many authors (Kullander, 2003; Miller et al., 2009; Page et al., 2013) and reported by Rodríguez-Varela et al. (2010) for the system, according to recent studies on the Middle American cichlid fishes (Říčan et al., 2016), it has been proposed that Mayaheros urophthalmus be the valid name of the species.

24Rocio octofasciatus. The species R. octofasciatus was recorded by Lara-Domínguez et al. (2011) and Abarca-Arenas et al. (2012). However, since Schmitter-Soto (2007), the current valid name for this species is Rocio octofasciata (Regan, 1903).

25Tilapia sp. An unidentified species of the Tilapia genus was recorded by Lara-Domínguez et al. (2011) and Abarca-Arenas et al. (2012), but this record can be very controversial. According to Schmitter-Soto (Miller et al., 2009), there has been a nomenclatural decision to assign species of the genus Oreochromis to species of the Tilapia genus, which goes against the recent use and both genera can be consistently separated (Sodsuk & McAndrew, 1991). Thus, in this case, the first problem is to know to which genus corresponds this record. Amador-del Ángel & Wakida-Kusunoki (2014) consider that in southwestern Mexico there are only four introduced species from the Cichlidae family (all of African origin), three of them within the Oreochromis genus (O. aureus, O. niloticus niloticus, and O. mossambicus), and one within the Tilapia genus (T. rendalli) [According to Froese & Pauly (2018) the current valid name for the latter species is Coptodon rendalli (Boulenger, 1897)]. Page et al. (2013) recognize in Mexico the same three introduced species for the genus Oreochromis, but another for the genus Tilapia (T. zillii) [Also according to Froese & Pauly (2018) the current valid name for this species is Coptodon zillii (Gervais, 1848)]. In the same way, in the checklist of freshwater fish from Veracruz reported by Mercado-Silva et al. (2011: Appendix VIII.40), only O. aureus, O. niloticus niloticus and O. mossambicus are recorded (species of the Tilapia genus are not reported). Finally, Fuentes-Mata & Espinosa-Pérez (1997) for Catemaco lake and Escondida lagoon (both very near to Sontecomapan Lagoon) recorded Oreochromis niloticus. According to the above, in the present study the record of Tilapia sp. is considered as Oreochromis sp., which most likely corresponds to Oreochromis niloticus (Linnaeus, 1758).

26Blennius nicholsi. Reséndez-Medina (1983) recorded the species Blennius nicholsi, but for a long time, this has been recognized as a synonym for the current valid name Lupinoblennius nicholsi (Tavolga, 1954) (Froese & Pauly, 2018). With this valid name, Fuentes-Mata & Espinosa-Pérez (1997), Lara-Domínguez et al. (2011) and Abarca-Arenas et al. (2012) recorded this species for the system. The family of this species (Blenniidae) is currently included in the Blenniiformes order (Nelson et al., 2016).

27Archomenidia sallei. This species was recorded by Reséndez-Medina (1983) within the Archomenidia genus, but currently this species is recognized within the Atherinella genus (Chernoff, 1986; Miller et al., 2009), and its valid name is Atherinella sallei (Regan, 1903). Reséndez-Medina (1983) also assigns this species to the Atherinidae family. However, the New World silversides that had formerly been recognized as a subfamily of Atherinidae, were formally removed from this family and were reassigned in Atherinopsidae (Dyer & Chernoff, 1996; Dyer, 2006; Nelson et al., 2016). Lara-Domínguez et al. (2011) and Abarca-Arenas et al. (2012) recorded this species for the system with the correct family and species names.

28Membras vagrans. Rodríguez-Varela et al. (2010) recorded this species, but it can also be controversial. Despite that it is a valid name (Castro-Aguirre & Espinosa-Pérez, 2006; Castro-Aguirre et al., 1999; Froese & Pauly, 2018), according to Hoese & Moore (1998), reports of Membras vagrans (Goode & Bean, 1879) refer to Membras martinica (Valenciennes, 1835). Membras vagrans is generally found in the deeper; more saline portions of the bays and out into the Gulf as far as 15 miles offshore. Similarly, Chernoff (1986) consider this species as M. martinica. In this way, in many studies, M. martinica is recorded in the Gulf of Mexico, but not M. vagrans (Robins & Ray, 1986; Hoese & Moore, 1998; Chernoff, 2002; McEachran & Fechhelm, 2005; McEachran, 2009; Mercado-Silva et al., 2011; Page et al., 2013). On the other hand, in the lists of fish species in some coastal lagoons of Veracruz, as Tamiahua (Franco-López & Chávez-López, 1992; Díaz-Ruiz et al., 2003), Mandinga (Lara-Domínguez et al., 2011), and Camaronera-Alvarado (Chávez-López et al., 2005), M. vagrans is recorded, but not M. martinica. This probably involves an identification problem, mainly related to the number of anal soft rays, which according to Castro-Aguirre et al. (1999) and Castro-Aguirre & Espinosa-Pérez (2006), it is the main characteristic to differentiate both species. However, meristic and morphometric variation within a single species may be associated with environmental variation, and differences between size classes (Chernoff, 1982), in part, the problem is also that Rodriguez-Varela et al. (2010) recorded both species. Indeed, this seems to be the only report from southwestern Gulf of Mexico (Caribbean province) which records both species in sympatry. Anyway, in the present study it was decided to recognize the record of M. vagrans for the system, but it is clear that further analysis is necessary not only to determine the presence of M. vagrans in this system, but also to validate all records reported for coastal systems in the southwestern Gulf of Mexico.

29Hyporhamphus roberti. This species is recorded by Fuentes-Mata et al. (1989), which is a name accepted by other authors (Cervigón et al., 1992; Collette, 2002; Page et al., 2013). Within this species, two subspecies are recognized (Collette, 2002), Hyporhamphus roberti hildebrandi Jordan & Evermann, 1927 and Hyporhamphus roberti roberti (Valenciennes, 1847). The latter taxon is recognized as a valid name by Froese & Pauly (2018) and with this name, the species is recorded for the system by Lara-Domínguez et al. (2011) and Abarca-Arenas et al. (2012).

30Strongylura notata. Fuentes-Mata et al. (1989), Fuentes-Mata & Espinosa-Pérez (1997), Rodríguez-Varela et al. (2010), Lara-Domínguez et al. (2011), and Abarca-Arenas et al. (2012) recorded this species for the system which is a name accepted by other authors (Robins & Ray, 1986; Castro-Aguirre et al., 1999; Collette, 2002; McEachran, 2009; Page et al., 2013). As with the previous species, two subspecies are recognized (Collette, 2002): Strongylura notata forsythia Breder, 1932 and Strongylura notata notata (Poey, 1860). The latter is recognized as the valid name by Froese & Pauly (2018).

31Tylosurus crocodilus crocodilus. Fuentes-Mata & Espinosa-Pérez (1997) recorded this species and the name is accepted by other authors (Cervigón et al., 1992; Collette, 2002). However, Froese & Pauly (2018) consider that this name is a synonym of Tylosurus crocodilus (Péron & Lesueur, 1821), which is also used by other authors (Robins & Ray, 1986; Castro-Aguirre et al., 1999; McEachran, 2009; Page et al., 2013), including the original record for the system by Fuentes-Mata et al. (1989).

32Heterandria bimaculata. Reséndez-Medina (1983), Fuentes-Mata & Espinosa-Pérez (1997), Lara-Domínguez et al. (2011), and Abarca-Arenas et al. (2012) recorded this species for the system, however, based on available phylogenetic information (Morales-Cazan & Albert, 2012), species of “Heterandria” were reassigned to the genus Pseudoxiphophorus, and the valid name proposed for this species is Pseudoxiphophorus bimaculatus.

33Poeciliopsis sp. An unidentified species of the Poeciliopsis genus was recorded by Lara-Domínguez et al. (2011) & Abarca-Arenas et al. (2012). According to Miller (1975), Miller et al. (2009), and Mercado-Silva et al. (2011), this record may correspond to two species, Poeciliopsis catemaco Miller, 1975, with distribution range in the Atlantic slope, in Catemaco lake and its outlet near the lake, the Papalopan river basin, Veracruz; and Poeciliopsis gracilis (Heckel, 1848), with a distribution also in the Atlantic slope, from a 20 km stream, north of Cardel city, southward into the Coatzacoalcos river basin, Veracruz (Miller et al., 2009). Given this uncertainty, in the present study, this record is considered as was originally proposed.

34Oostethus brachyurus lineatus = Oostethus (B.) lineatus = Microphis brachyurus lineatus. This species is recorded as Oostethus brachyurus lineatus by Reséndez-Medina (1983) and as Oostethus (B.) lineatus by Fuentes-Mata & Espinosa-Pérez (1997). However, Dawson (1984) proposed Oostethus as a subgenus of Microphis, and it is currently considered as the genus accepted for this species. Meanwhile, Rodríguez-Varela et al. (2010), Lara-Domínguez et al. (2011) and Abarca-Arenas et al. (2012) recorded Microphis brachyurus lineatus for the system. But, according to Froese & Pauly (2018), all of these names are synonyms of the valid name Microphis lineatus (Kaup, 1856); although this species is still referred in recent literature as Oostethus lineatus, Oostethus brachyurus lineatus or Microphis brachyurus lineatus (Fritzsche & Vincent, 2002; Ferraris, 2003).

35Eucinostomus lefroyi. In comparative studies of several estuarine systems from Veracruz, Lara-Domínguez et al. (2011) and Abarca-Arenas et al. (2012) recorded this species for the system. Although the name is accepted by many authors (Robins & Ray, 1986; Hoese & Moore, 1998; Gilmore & Greenfield, 2002; McEachran & Fechhelm, 2005; McEachran, 2009; Page et al., 2013), the genus of this species has not been fully defined (Castro-Aguirre et al., 1999). According to Böhlke & Chaplin (1993) this species should be recognized within the genus Ulaema, and Froese & Pauly (2018) consider that Eucinostomus lefroyi is synonym of Ulaema lefroyi (Goode, 1874). It is with this name that is recognized in the present study, which was also registered in the system by Fuentes-Mata et al. (1989), Fuentes-Mata & Espinosa-Pérez (1997), and Rodríguez-Varela et al. (2010).

36Aluterus schoepfi. This species is recorded by Fuentes-Mata et al. (1989) within the Balistidae family, but it is currently included within the Monacanthidae family. The separation of Monacanthidae as a family distinct from Balistidae (formerly as a subfamily of the latter), has been long recognized (Berry & Vogele, 1961). Moreover, the original report has a misspelling problem, the correct name for this species is Aluterus schoepfii.

List of species historically recorded in the Sontecomapan Lagoon

37According to previous analyzes, the total number of species historically recorded for the Sontecomapan Lagoon is 115, although two of them could correspond to misidentifications (Cathorops melanopus for Cathorops aguadulce, and Membras vagrans for Membras martinica). All species listed correspond to 26 orders, 41 families and 85 genera. Of these 115 species, 37 were reported in five studies, 30 in four, 14 in only three, 20 were exclusive for two, and 14 were recorded in only one study (Table 1). From these 14 species, six (C. urophthalmus, M. vagrans, M. beryllina, S. guachancho, L. rhomboides, and S. testudineus) were recorded only by Rodríguez-Varela et al. (2010), five (A. fasciatus, A. alvarezi, S. plumieri, C. arenarius, and B. capriscus) were recorded exclusively by Fuentes-Mata & Espinosa-Pérez (1997), two (O. gomesii and A. schoepfii) were recorded only by Fuentes-Mata et al. (1989), and only one (R. quelen) was recorded exclusively by Abarca-Arenas et al. (2012). All species recorded by Reséndez-Medina (1983) and Lara-Domínguez et al. (2011) were considered in at least one of the remaining four studies.

Classification changes at higher levels

38In the present study, systematic sequence of orders and families of the new fish classification of Fishes of the World (Nelson et al. 2016) was adopted, which differs in many respects from the previous editions. For the present case, at family level, the main change was the recognition of Oxudercidae (= Gobionellidae from some authors), based on phylogenetic studies (Thacker, 2009; Nelson et al, 2016), which includes the members of several former subfamilies of the Gobiidae. At order level, Gobiiformes were formerly treated as suborder Gobioidei within the Perciformes order (Nelson, 2006), but according to molecular evidence and some morphological characters (Betancur-R. et al., 2013; Nelson et al., 2016), it is now recognized as a new order, including Eleotridae, Oxudercidae and Gobiidae families. The Cichlidae family (formerly in Labroidei suborder, Perciformes order) is currently in Cichliformes. The family Blenniidae, formerly treated in the Blennioidei suborder (Perciformes order), is placed in Blenniiformes. The Sphyraenidae and Trichiuridae families had been placed in the Scombroidei suborder (Perciformes order) by Nelson (2006), but are now placed in Istiophoriformes and Scombriformes, respectively. The seven families formerly treated as Syngnathoidei suborder (included Syngnathidae) within the Gasterosteiformes order by Nelson (2006) are now placed in the Syngnathiformes order. The Carangidae, Sciaenidae, and Sparidae families were formerly treated as Percoidei suborder within the Perciformes order, and now placed in the new Carangiformes, Acanthuriformes and Spariformes orders respectively. Finally, the phylogenetic position of some orders like Pleuronectiformes and Lophiiformes have drastically changed.

Species similarity among studies

39According to the analysis of similarity between the species’ composition of each study (Table 2), it can be observed that similarity values vary widely from 0.012 to 0.968. In part, this variability may be due to the different fishing gear used, as well as to changes in species composition, or at least, to changes abundance. In addition, this variability can be influenced by the characteristics of each report. Thus, the lower similarities (≤0.295) are among Fuentes-Mata et al. (1989) and the remaining five studies (Figure 2). This is due to two reasons: the first one is that Fuentes-Mata et al. (1989) did not list all species caught, reporting only those considered as new records (31 in total); the second, is that comparative studies of Lara-Domínguez et al. (2011) and Abarca-Arenas et al. (2012) ignored the Fuentes-Mata report. In this sense, it is particularly interesting that studies of Fuentes-Mata et al. (1989) and Fuentes-Mata & Espinosa-Pérez (1997) show a low similarity (0.295), with five species (O. gomesii, E. pisonis, G. hastatus [= G. oceanicus], H. roberti, and A. schoepfii) recorded in the first study, but not in the second, although both sampling periods concurred during 1985-1987.

Table 2. Species similarity (above the diagonal) and shared species (below the diagonal) among the different lists of species reported by: Reséndez-Medina (1983), Fuentes-Mata et al. (1989), Fuentes-Mata & Espinosa-Pérez (1997), Rodríguez-Varela et al. (2010), Lara-Domínguez et al. (2011), and Abarca-Arenas et al. (2012).

40The species composition reported by Rodríguez-Varela et al. (2010) also tends to show low similarities with the other studies (Fig. 2), with a particularly low similarity (0.351) between this study and that of Reséndez-Medina (1983). Although both studies recorded a total of 52 species, these only shared 27. In part, this may be because the sampling periods have the greatest separation in years (1980-2005), which can mean drastic changes in the composition or relative abundance of species in 25 years.

41In contrast, the checklists of Lara-Domínguez et al. (2011) and Abarca-Arenas et al. (2012) tend to show higher similarities. This is mainly because they are comparative studies, which take into account previous studies on the lagoon.

Figure 2. Similarity dendrogram (Jaccard, UPGMA) from the species lists reported by Reséndez-Medina (1983), Fuentes-Mata et al. (1989), Fuentes-Mata & Espinosa-Pérez (1997), Rodríguez-Varela et al. (2010), Lara-Domínguez et al. (2011) and Abarca-Arenas et al. (2012).

Expected number of fish species for the Sontecomapan Lagoon

42The number of species reported in the studies carried out in situ varies from 31 to 83, with an accumulated of 103 species recorded, from which at least, 38 were common in all four studies (Reséndez-Medina, 1983; Fuentes-Mata et al., 1989; Fuentes-Mata & Espinosa-Pérez, 1997; Rodríguez-Varela et al., 2010). The two comparative checklists report between 91 and 92 species, from which 90 are common (Lara-Domínguez et al., 2011; Abarca-Arenas et al., 2012). However, the number of species counted in these studies may be an underestimation of the total species richness (due to fishing gear limitations and the fact that many rare species are not detected).

43Thus, considering the total of species recorded in approximately 25 years, the total number of expected species according to each of the nonparametric estimators used was: Mao Tau = 115, Chao 2 = 118.61, Jackknife 1 = 126.67, Jackknife 2 = 125.33, and Bootstrap = 121.71. The species accumulation curves for the six studies considered using these nonparametric estimators are shown in Figure 3.

Figure 3. Species accumulation curves considering the species lists (as samples) reported by Reséndez-Medina (1983), Fuentes-Mata et al. (1989), Fuentes-Mata & Espinosa-Pérez (1997), Rodríguez-Varela et al. (2010), Lara-Domínguez et al. (2011), and Abarca-Arenas et al. (2012). Nonparametric estimators used were: Mao Tau (—•—), Chao 2 (—▲—), Jackknife 1 (—■—), Jackknife 2 (–▼–), and Bootstrap (–♦–).

44The main ecological interpretation of these results is that the estimates of nonparametric methods are not very different from the total number of species recorded for the system (maximum difference of roughly 12 species) thus, implying that this historic review results in a good estimate of the richness of fish species for this system. Largely, this is because different studies in the system have covered a wide range of developmental stages, including larvae and juveniles (e.g. Rodríguez-Varela et al. 2010), as well as a wide range of fishing methods (e.g. Reséndez-Medina 1983).

Concluding remarks

45Considering the previous information for the Sontecomapan Lagoon, 115 species have been recorded, two of which may correspond to misidentifications (Cathorops melanopus and Membras vagrans). At least for 17 cases the taxonomic status has changed, with Hypanus sabinus (recorded as Dasyatis sabina), Ctenogobius boleosoma (recorded as Gobionellus boleosoma), Gobionellus oceanicus (recorded as Gobionellus hastatus), Mayaheros urophthalmus (recorded as Cichlasoma urophthalmus), Pseudoxiphophorus bimaculatus (recorded as Heterandria bimaculata) and Microphis lineatus (recorded as Oostethus brachyurus lineatus, Oostethus (B.) lineatus and Microphis brachyurus lineatus) being the most relevant. Although the genus level is known, the species level is uncertain for two cases, Tilapia sp. (in the present study considered as Oreochromis sp.) and Poeciliopsis sp. For both, the species to which these taxa can correspond is proposed. The generic allocation of one species (Eucinostomus lefroyi or Ulaema lefroyi) seems to be a case in which taxonomists did not end up agreeing. Considering the new fish classification by Nelson et al. (2016), the Ctenogobius boleosoma, Evorthodus lyricus, Gobioides broussonnetii and Gobionellus oceanicus species are now included in the new family Oxudercidae (= Gobionellidae), and 11 of the 41 recorded families are now included in nine new orders.

46Despite having an approximately 8.9 km² surface, the lagoon is one of the most species-rich estuarine ecosystems in the southern Gulf of Mexico (Caribbean province), with approximately 13 % of the total fish species recorded in this area. In relation to the similarity analysis, fish composition showed lower values (<0.6) among most studies in the Sontecomapan Lagoon, even in those studies involving the same authors and similar collection dates. Finally, according to nonparametric methods for species richness estimation, the total number of species expected for the lagoon could be between 119 and 127 species.

Acknowledgements

47We would like to thank G. Morgado-Dueñas for her assistance in the elaboration of Figure 1. Authors also thank three anonymous reviewers and, editors from Mexico and France for their valuable comments.