Fish larvae from the Gulf of California to Colima, Mexico: An update

: An updated taxonomic list of marine fish larvae from the Gulf of California to Colima, Mexico is presented. A total of 579 taxa belonging to 119 families, 256 genera, and 423 species were recorded. The list was compiled using 14 publications on fish larvae research (1974-2012), and the fish larvae identified from 315 samples collected with Bongo nets during 10 oceanographic cruises made from the Gulf of California to Bahía de Banderas, Mexico, from 2003 to 2007 (this study). The most important families in this study were the Myctophidae (28.3%), Engraulidae (25.0%), and Clupeidae (15.4%). The most abundant species were Cetengraulis mysticetus (18.2%), Benthosema panamense (13.9%), and Opisthonema libertate (12.7%). The compiled taxonomic list shows the addition of 296 new taxa to the previous list published 10 years ago, and also the need of an increase in the effort on the taxonomy of fish larvae forms not identified to species level.


Introduction
Ichthyoplankton studies in the Gulf of California began almost 40 years ago when Moser et al. (1974) collected plankton during six oceanographic cruises made during 1956-1957 and published the first list of the occurrence and abundance of marine fish larvae. Moser (1996) provided the most important taxonomic tool for the identification of the early stages of fish larvae in the California Current, which account for an important number of species in the Gulf of California, while Aceves-Medina et al. (2003) contributed with the first and most extensive systematic list to that date. Many other early ichthyoplanktonic studies targeted only the main commercial species such as the Pacific sardine and the northern anchovy (Hamman et al. 1988;Green-Ruiz and Hinojosa-Corona 1997).
More recently, ichthyoplanktonic research related to the study of fish larvae abundance, distribution, and their relationship to environmental factors and mesoscale processes was conducted throughout several areas from the Gulf of California to the oceanic area in front of Colima (Franco-Gordo et al. 1999;Ávalos-García et al. 2003;Sala et al. 2003;Sánchez-Velasco et al. 2004;González-Armas et al. 2008;Peguero-Icaza et al. 2008;Silva-Segundo et al. 2008;Avendaño-Ibarra et al. 2009;León-Chávez et al. 2010;Contreras-Catala et al. 2012;Avendaño-Ibarra et al. 2013). Some of these included taxonomic lists of the fish larvae identified. At the same time, a strong sampling effort (10 oceanographic cruises from 2003 to 2007) directed to identify fish larvae communities and other zooplankton components and their relationship to the pelagic environment, provided the clues to determine a change in the status of our knowledge of the diversity in the Mexican Pacific area.
Abstract: An updated taxonomic list of marine fish larvae from the Gulf of California to Colima, Mexico is presented. A total of 579 taxa belonging to 119 families, 256 genera, and 423 species were recorded. The list was compiled using 14 publications on fish larvae research , and the fish larvae identified from 315 samples collected with Bongo nets during 10 oceanographic cruises made from the Gulf of California to Bahía de Banderas, Mexico, from 2003. The most important families in this study were the Myctophidae (28.3%), Engraulidae (25.0%), and Clupeidae (15.4%). The most abundant species were Cetengraulis mysticetus (18.2%), Benthosema panamense (13.9%), and Opisthonema libertate (12.7%). The compiled taxonomic list shows the addition of 296 new taxa to the previous list published 10 years ago, and also the need of an increase in the effort on the taxonomy of fish larvae forms not identified to species level. the northern and southern gulf (Figure 1), with thermal fronts induced by vertical mixing at the sills (≈400─600 m depth) of the archipelago that maintain low temperatures (Lavín et al. 1997;Soto-Mardones et al. 1999) and high chlorophyll concentrations throughout the year (Pegau et al. 2002;Espinosa-Carreón and Valdéz-Holguín 2007). The Southern Gulf of California (SGC) is located from the lower limit of the archipelago through Cabo San Lucas at the tip of the Baja California peninsula, and it is characterized by, among other features, strong upwelling and oceanic fronts particularly during winter-spring, deep basins with depths that range from 1000─3000 m, and warm sea surface temperature with small variations throughout the year (Soto-Mardones et al. 1999;Lavín and Marinone 2003). This area presents a mixture of tropical and temperate species. The continental coast of the gulf is characterized by a broad continental shelf and mostly sandy beaches, while the peninsular side has a narrow continental shelf and rocky beaches. The Entrance Zone of the gulf (from the tip of the Baja California peninsula and El Dorado, Sinaloa to Cabo Corrientes, Jalisco) is located in the Eastern Tropical Pacific (ETP) in front of Mexico. It is a transitional ocean region associated with the seasonal confluence of tropical water brought to the area by the poleward Mexican Coastal Current (Kessler 2006;Lavín et al. 2006), and subarctic water from a branch of the California Current System which, instead of flowing westward to join the North Equatorial Current, flows equatorward parallel to the coast of Mexico (Kessler 2006). In this area, submarine canyons, a narrow continental coast, an intense mesoscale activity consisting of fronts between the California Current, Gulf of California, and Tropical Surface Water masses, and eddies and filaments (related to wind originated coastal upwelling) are observed year-round (Torres-Orozco et al. 2005;Zamudio et al. 2007). The presence of all these features, and also of a number of islands, may promote and enhance a wide variety of habitats linked to a complex and diverse ichthyoplanktonic community. The oceanographic region from Jalisco to Colima presents a narrow continental shelf (Filonov et al. 2000). It is influenced by the poleward flow of the Costa Rica Coastal Current (CRCC) almost all the year, which is more or less intense according to the change of position of the intertropical convergence zone, and plays an important role carrying tropical and subtropical biota to northern latitudes. The variability in this region is influenced by interannual, interseasonal (equatorially generated), and higher frequency (storm-induced) signals (e.g. coastally trapped Kelvin waves) (Zamudio et al. 2001;.
A total of 315 plankton samples were collected during ten oceanographic cruises from November 2003 to August 2007 (this study). These included three transects made in the gulf from Bahía de La Paz, Baja California Sur to Bahía de Banderas (Figure 1). The cruises included the following months: November 2003, 2004, 2005March, May, andSeptember 2005, andMarch 2006 (CGC0503, CGC0505, CGC0509, andCGC0603); and January and July 2007 (GOLCA0701 and GOLCA0707). Cruises were made on board the B/O H05 Altair and B/O H03 Alejandro de Humboldt of the Secretaría de Marina, Armada de México; on the B/O El Puma of the Universidad Nacional Autónoma de México; and on the sailing vessel SSV Robert C. Seamans of the Sea Education Association, Woods Hole Laboratory, Massachusetts, U.S.A. For the six cruises (CGC0503, CGC0505, GOLCA0511, CGC0603, GOLCA0701, GOLCA0707), plankton samples were obtained from oblique tows by means of standard Bongo net tows (Smith and Richardson 1979), or with a conical 1 m diameter net (S-195 and S-207 cruises, 505 µm pore size) that was towed using the same Bongo nets methodology. During the September 2005 cruises (CGC0509), plankton samples were collected using only a surface conical net. Additional Neuston surface tows (345 µm pore size) were made during the cruises S-189, S-195, and S-207. In all cases the nets were fitted with calibrated flow meters. After collection, samples were drained and immediately fixed in 96% ethyl alcohol, followed by a full change of preservation fluids 24 hours later. In the laboratory, the ichthyoplankton was sorted from the entire zooplankton samples. In the case of bongo samples, only the larvae from the 505-μm mesh net were analyzed. Fish larvae were identified to the lowest possible taxonomic level, mainly using Moser et al. (1984), Moser (1996), Beltrán-León and Ríos (2000a, b), and Richards (2006a, b), and preserved in 96% alcohol. The number of larvae was standardized to 10 m 2 of sea surface (Smith and Richardson 1979). No samples were taken for DNA analysis. The taxa list follows the Eschmeyer, W. N. Catalog of Fishes electronic version (updated 04/01/13) (Eschmeyer 2013), and the updated species names, faunistic affinity, and habitat, follow the Fish Base (updated 12/2012) (Froese and Pauly 2012) web pages. The genera and their respective species are presented in alphabetical order. Voucher specimens were cataloged and deposited in the Ichthyoplankton Collection of the Mexican North Pacific (acronym ICTIOPLANCTON) at CICIMAR in La Paz, BCS, Mexico (catalog number SEMARNAT B.C.S. . Some larvae were not identified to species level because of the lack of larval descriptions of species inhabiting the area or due to damaged larvae. These were distinguished with the genera name plus the notation "sp." followed by a number to denote the number of different morphological forms in each recognized taxa. The list of species presented in this paper, comprises not only the results of the identified larvae from the ten oceanographic cruises made in this study, but also the compilation of 14 species lists of fish larvae of several areas located from the Gulf of California to the coasts of Jalisco-Colima published in oceanographic or ecological studies between 1974 and 2012. When compiling the list, we noted that different formats for species level were used by different authors in their published lists, e. g. ("Diplectrum type 1, Diplectrum T1, Diplectrum sp. A, or Diplectrum sp."). As a result, we adjusted them to fit the "Diplectrum sp. n" format in which "n" represents a number to denote the number of morphological forms reported per author. The same was observed at family level. In this case, we only mentioned the family in the list and included a table where the number of forms registered per family, per author is presented. Lists of species from bays or coastal lagoons in the study area were not included.
The species list presented by Aceves-Medina et al. (2003) was used as baseline to determine the historical addition of new taxa in each of the 14 lists and in this study. Our study represents the most recent addition to the list, and to our knowledge, the additional taxa we found were never registered before by the former authors.
In this study, the Myctophidae, Phosichthyidae, Engraulidae, and Clupeidae families were the most important in the Gulf of California accounting for 79.0% of total larval abundance. Similar results were found by Moser et al. (1974) and Aceves-Medina et al. (2003) that recorded 71.8% and 82.4% respectively ( Table 2). The slight differences in the relative abundance by family were the result of the collection of less myctophyds, but more Engraulidae and Clupeidae fish larvae in our study, compared to the relative abundance of these families in the collections of Moser et al. (1974) and Aceves-Medina et al. (2003) (Table 2). These authors also recorded that the main characteristic of the larval fish community of the gulf was the dominance of the mesopelagic and coastal pelagic taxa. In this study, mesopelagic and coastal pelagic taxa contributed almost in the same proportion to the total larval abundance (43.3 and 43.9% respectively), very similar to the records of Moser et al. (1974) (Table 2).
California to Bahía de La Paz. The taxa increment reported in our compiled list resulted from a continuous sampling effort through time and on the entire extension of the sampling area. In the Gulf of California, Moser et al. (1974) reported 27 different taxa not included by Aceves-Medina et al. (2003) (Table 4). Several studies conducted inside the gulf added 95 additional taxa. These studies covered from the north to the center of the gulf, and the area off Bahía de La Paz (Ávalos-García et al. 2003;Sánchez-Velasco et al. 2004;González-Armas et al. 2008;Peguero-Icaza et al. 2008;Avendaño-Ibarra et al. 2009;Contreras-Catala et al. 2012). Extensive sampling directed to a few commercial species carried out in the rocky reefs along the entire Gulf of California added only two new taxa in which adult fishes in spawning aggregations with eggs observed by diving were reported (Sala et al. 2003 (2003). Some discrepancies between the species names and also important changes in family names were found when compiling the list. These were due to the continuous change in the systematics of fishes to date. Fourteen species found in the analyzed lists changed: Antennarius avalonis to Fowlerichthys avalonis (Jordan & Starks, 1907), Antennarius sanguineus to Antennatus sanguineus (Gill, 1863), Bathylagus nigrigenys to Bathylagoides nigrigenys (Parr, 1931)

Discussion
In our study, we found that extremely high abundances of Cetengraulis mysticetus (40,149 Larvae/10 m 2 ) not recorded by Moser et al. (1974) or Aceves-Medina et al. (2003), and of Opisthonema libertate (24,252 Larvae/10 m 2 ) collected in two sampling stations very close to each other located in the northern gulf, indicated an spawning event produced by the coupling of reproductive strategies of the species to mesoscale processes in that area (Avendaño-Ibarra et al., 2013). Consequently, these high abundances influenced the relative abundance of the entire community. Excluding the abundance of these two stations from the analysis, the aforementioned mesopelagic families accounted for 60.2%; coastal pelagic families reached only 19.8%, while a decrease in the shallow demersal (8.1%) was observed.
Sampling directed to reef fish larvae communities inside the gulf only showed presence of coastal pelagic (27.0%), and dominance (64.6%) of shallow demersal taxa over all previous reports (Brogan 1994) (Table 2). To the southeast of our study area, in the Mexican Central Pacific, Acal (1991) reported also a different species community, with dominant coastal pelagic (72.4%), followed by mesopelagic species (18.5%). In contrast, in the Eastern Tropical Pacific, Ahlstrom (1971) reported dominance of mesopelagic species.
The compiled taxonomic list of taxa registered from the Gulf of California to the Colima area since 1974 to 2012 (including the results of this study) indicates a total of 579 taxa. From these, 306 were identified to species, 173 to genera, 94 to family, four to subfamily, and two taxa to order level (Table 3). These numbers indicate an addition of 296 new taxa since the most complete systematic list published ten years ago by Aceves-Medina et al. (2003) (283 taxa and 173 species), sampled from the north area of the Gulf of Mexico and offshore to ca. 400 nm (early CalCOFI sampling program). In addition to this area, they had an important source of fish larvae for early life history studies from different expeditions, programs, and cruises such as NORPAC (Northeast Pacific, 20°-48° N, 111°-154° W), the CalCOFI cruises in the Gulf of California (23°-32° N, 107°-115° W), and EASTROPAC I and II (Eastern Tropical Pacific, 20° N-20° S, 76°-126° W), among others. They reported 141 families and 467 species in this large region. From the 306 taxa they reported to be distributed only in the Gulf of California or in the Mexican Tropical Pacific, 72 species have never been collected again. Our compiled taxonomic list included 423 species that represented approximately 48.3% of the 875 species recorded as adults in the Gulf of California (Thomson et al. 2000). All this indicates that our species records are still low. The high number of "forms" registered in some taxa: eight in the Gobiidae family (Table 4, Ávalos-García et al. 2003), or in the Anguilliformes order with ten forms (Table 4, this study), indicate that these forms may be an important source of species increment if they do represent species. The presence of several forms in our compiled list occurred in 14 families and one order: Bothidae, Chiasmodontidae, Clinidae, Congridae, Cyematidae, Eleotridae, Gobiesocidae, Gobiidae, Haemulidae, Labrisomidae, Macrouridae, Paralichthyidae, Pomacentridae, Sciaenidae families, and the Anguilliformes order (Table 4). Taking into account only the maximum number of forms reported in each one of them, around 65 more could be representing species. These results show that a strong taxonomic effort to describe fish larvae still needs to be done, particularly in the mentioned taxa that represent an opportunity to eliminate the gap in our knowledge of the fish larvae diversity in the Gulf of California.
There are more fish larvae studies inside the Gulf of California, but they did not include any different taxa to the date they were published and we did not use them. Taxonomic lists from the coastal lagoon and bays were also not included in our compiled list. Although coastal ecosystems in the Gulf of California are reservoirs of great biological diversity (Martínez-López et al. 2007), the exportation and contribution of shallow fish larvae to neritic or oceanic adjacent waters depends on several environmental variables and still needs to be investigated. A coastal sampling program covering the first 20 nm from the coast could be a first step to understand, not only the close to the coast fish larvae diversity, but several other oceanographic processes linking coastal to oceanic waters.
In terms of sampling, the most common sampling technique for fish larvae collection used by the authors in the compiled list was the oblique Bongo net tow; only two used opening-closing conical zooplankton nets (Sánchez-Velasco et al. 2007;Contreras-Catala et al. 2012), one did surface trawls with a conical net (González-Armas et al. 2008), and only Silva-Segundo et al. (2008) used oblique shallow trawls at very coastal sampling stations. Trawl selectivity of these sampling techniques, depth of tow, distance to the coast, mesoscale oceanographic processes such as eddies and fronts, intra-and interannual variability, locality, bathymetric differences along the study area, sampling frequency, and reproductive strategies of adult fishes, among other variables, could determine different species composition in the samples collected (Aceves-Medina et al. 2003;Franco-Gordo et al. 2008;Inda-Díaz 2010;Jiménez-Rosenberg et al. 2010).
The taxonomic list of fish larvae species from the Gulf of California to Colima compiled here represents the most extensive and comprehensive list presented to date. This list added more than twice the number of taxa previously reported by Aceves-Medina et al. (2003) but it still remains conservative in relation to the total number of adult fishes reported. The numerous "forms" pertaining to the family level reported in 14 families and one order showed the need to emphasize the taxonomic work as a primary tool to identify and understand the alpha diversity of the Gulf of California.   -Gordo et al. (1999;, Silva-Segundo et al. (2008), andLeón-Chávez et al. (2010). This study =Gulf of California to Jalisco, 2003Jalisco, -2007