Introduction
Cephalopods are mainly nektonic molluscs ecologically and commercially important in the world ocean (Boyle & Rodhouse, Reference Boyle and Rodhouse2005). Hatchings of squids and several octopods are represented by planktonic stages known as paralarvae (PL). They are relatively rare in plankton communities because of their ephemeral existence due to fast growth (Roura et al., Reference Roura, Álvarez-Salgado, González, Gregori, Rosón, Otero and Guerra2016). Knowledge of cephalopod paralarvae is still limited, especially in their horizontal and vertical distribution and abundance in tropical waters.
Paralarval abundance in the water column is influenced by environmental changes caused by ocean dynamics. Frontal areas at the edges of water masses segregate different PL assemblages (De Silva-Dávila et al., Reference De Silva-Dávila, Franco-Gordo, Hochberg, Godínez-Domínguez, Avendaño-Ibarra, Gómez-Gutiérrez and Robinson2015). Thermocline, halocline and food availability related to mesoscale eddies and upwelling fronts play important roles in distribution, growth and survival of PL (Ichii et al., Reference Ichii, Mahapatra, Sakai, Inagake and Okada2004; Aceves-Medina et al., Reference Aceves-Medina, De Silva-Dávila, Cruz-Estudillo, Durazo and Avendaño-Ibarra2017). Ocean currents are also a key factor in transporting eggs or egg masses and PL from the spawning areas to the hatching, nursery or feeding grounds (Díaz et al., Reference Díaz, Ardila and Gracia2000; Downey-Breedt et al., Reference Downey-Breedt, Roberts, Sauer and Chang2016). They influence the dispersion of PL of different biogeographic affinities (Diekmann & Piatkowski, Reference Diekmann and Piatkowski2002; Haimovici et al., Reference Haimovici, Piatkowski and Aguiar dos Santos2002), and may keep the PL inside a macroscale circulation cell that results in an essential strategy to complete their life cycles (Roberts & van den Berg, Reference Roberts and van den Berg2002). Thus cephalopods are likely to be affected by fluctuations in oceanic circulation due to global change with unknown effects (Xavier et al., Reference Xavier, Allcock A, Cherel, Lipinski, Pierce, Rodhouse, Rosa, Shea, Strugnell, Vidal, Villanueva and Ziegleret2015).
Cephalopod fauna from the Caribbean has only been considered in one study (Nesis, Reference Nesis1975), resulting in an apparently smaller cephalopod diversity than adjacent Florida waters (Díaz et al., Reference Díaz, Ardila and Gracia2000; Gracia et al., Reference Gracia, Ardila and Díaz2002; Judkins et al., Reference Judkins, Vecchione, Roper and Torres2010). Nesis (Reference Nesis1975) did not consider the Mesoamerican Barrier Reef System (MBRS) of the western Caribbean, which represents one of the largest marine biodiversity hotspots worldwide (Roberts et al., Reference Roberts, McClean, Veron, Hawkins, Allen, McAllister, Mittermeier, Schueler, Spalding, Wells, Vynne and Werner2002). This area shows a highly dynamic oceanography with physical processes driving the dispersion of planktonic larval stages. During 2006 and 2007 the oceanographic background and the distribution of larval fish, lobster phylosomas and pteropods were studied in the MBRS (Parra-Flores & Gasca, Reference Parra-Flores and Gasca2009; Muhling et al., Reference Muhling, Smith, Vásquez-Yeomans, Lamkin, Johns, Carrillo, Sosa-Cordero and Malca2013; Carrillo et al., Reference Carrillo, Johns, Smith, Lamkin and Largier2015, Reference Carrillo, Johns, Smith, Lamkin and Largier2016, Reference Carrillo, Lamkin, Johns, Vásquez-Yeomans, Sosa-Cordero, Malca, Smith and Gerard2017; Canto-García et al., Reference Canto-García, Goldstein, Sosa-Cordero and Carrillo2016). The aim of this study is to show for the first time the abundance, diversity and horizontal and vertical distribution patterns of cephalopod paralarvae from the surface waters of the MBRS and to discuss the influence of oceanographic features on these organisms.
Materials and methods
Study area
The Mesoamerican Barrier Reef System (MBRS) runs along nearly 1000 km from the north-eastern tip of the Yucatan Peninsula to the Gulf of Honduras (Figure 1). Physical parameters are defined by the latitude of impingement of the Cayman Current (CC) on the MBRS separating into three main dynamic environments: a northern region with a well-defined Yucatan Current (YC), a southern region with weak southward and/or variable flow, and a central, transitional region where the CC impinges upon the coast (Carrillo et al., Reference Carrillo, Johns, Smith, Lamkin and Largier2015). A strong north-westerly current with a uniform speed (~1.1 m s−1) in the north of the MBRS approaches the Yucatan Channel where YC reaches maximum speeds of 2.0 m s−1. In the transitional region most of the CC turns west with onshore flow at the Belize–Mexico border, and the bifurcation of this current with a divergence zone results in a southerly flow along the coast of Belize and a northward flow along the coast of Mexico. The southern MBRS region is characterized by the large cyclonic Honduras gyre (Carrillo et al., Reference Carrillo, Johns, Smith, Lamkin and Largier2015).
The Mesoamerican Barrier Reef System in the western Caribbean, Mexico, showing oceanographic lines A, C, F, G, I, N and P where sampling was conducted. White dots represent stations occupied during March 2006, black dots during January 2007, and grey dots during both years. Dashed line represents the 200 m depth isobaths. Main currents are in grey.
The MBRS receives water masses that are imported by the CC and the Caribbean currents entering from the Atlantic Ocean through the multiple passages between the northern and eastern Caribbean islands. The Caribbean Surface Water (CSW) represents the surface mixed layer with a mean thickness of 85 m depth, a mean temperature of 27°C and a low salinity (<36 at surface). CSW occupied almost all the sampling stations. Below 100 m depth, the high-salinity (>36.75) North Atlantic Subtropical Underwater (SUW), centred at about 150 m depth, lies beneath the CSW. Dissolved oxygen in the surface mixed layer is >4 ml l−1. The northernmost region of the MBRS was characterized by an upward displacement of water close to the shelf due to the upwelling associated with the YC forming a marked thermal front. At the shelf edge the 25°C isotherm was displaced by 100 m and the mixed layer was ~50 m thick (Carrillo et al., Reference Carrillo, Johns, Smith, Lamkin and Largier2016).
Sampling
Two oceanographic campaigns during 18 March–1 April 2006 and 14–30 January 2007 aboard the NOAA ship ‘Gordon Gunter’ surveyed the MBRS, with 23 and 27 sampling stations visited respectively, along seven lines (Figure 1): Line A, the northernmost line offshore Contoy Island to ~115 km offshore; Line C, located 60 km north of Ascension Bay and south of Cozumel Island; Line F south of Ascension Bay; Lines G and I, north and south of Chinchorro Bank respectively, in the Cayman Current impingement zone (Carrillo et al., Reference Carrillo, Johns, Smith, Lamkin and Largier2015); and Lines N and P, which are representative in Belizean coastal waters, with Line P just south of Glovers Reef in the Gulf of Honduras. Lines A, C, G and I were occupied both years. Line F was occupied in 2006 while southernmost lines N and P were visited only during 2007.
At each station temperature, salinity and dissolved oxygen were recorded with calibrated CTD casts and published elsewhere (Carrillo et al., Reference Carrillo, Johns, Smith, Lamkin and Largier2016). Current velocity measurements at 35 m depth were obtained from a hull-mounted Acoustic Doppler Current Profiler (Carrillo et al., Reference Carrillo, Johns, Smith, Lamkin and Largier2015). Vertical oceanographic features that could be related to PL abundance were plotted at lines A, C, F, G and I for 2006 and at lines A, C, G, I, N and P for 2007.
Depth-stratified plankton sampling was performed using a multiple opening and closing net environmental sensing system (MOCNESS) with 1 m2 mouth and 333 μm mesh. It sampled four depth strata: 0–25 m, 25–50 m, 50–75 m and 75–100 m (Muhling et al., Reference Muhling, Smith, Vásquez-Yeomans, Lamkin, Johns, Carrillo, Sosa-Cordero and Malca2013; Canto-García et al., Reference Canto-García, Goldstein, Sosa-Cordero and Carrillo2016). Flowmeters were fitted to the centre of the mouth of each net to estimate the water volume filtered in each tow. Plankton samples were preserved in 95% ethanol, which was replaced after 24 h to ensure proper preservation. All stations were sampled once. Sampling local time was classified as during the day (0600–1800 h) or at night (1801–0559 h) (Canto-García et al., Reference Canto-García, Goldstein, Sosa-Cordero and Carrillo2016).
Data analysis
All cephalopod paralarvae were sorted from the samples and identified to the most precise taxonomic level possible according to Nesis (Reference Nesis1979), Young & Harman (Reference Young and Harman1987), Sweeney et al. (Reference Sweeney, Roper, Mangold, Clarke and Boletzky1992), Vecchione et al. (Reference Vecchione, Roper, Sweeney and Lu2001), Bolstad (Reference Bolstad2010), De Silva-Dávila (Reference De Silva-Dávila2013), Lenz et al. (Reference Lenz, Elias, Leite and Vidal2015) and Roper et al. (Reference Roper, Gutierrez and Vecchione2015). Abundance of PL was standardized for each tow to number of paralarvae (PL) in 1000 m3 of filtered water (PL 1000 m−3) (Diekmann et al., Reference Diekmann, Nellen and Piatkowski2006). Paralarvae were measured for mantle length (ML) to the nearest 0.1 mm. Those smaller than 2.0–3.0 mm ML are seldom described in oceanic squids (Sweeney et al., Reference Sweeney, Roper, Mangold, Clarke and Boletzky1992). Therefore, paralarvae at these sizes not matching published descriptions, but with similar chromatophore patterns, were classified as morphotypes of a given genus, e.g. Enoploteuthis sp. 1, Octopus sp. 1. Unpigmented paralarvae without chromatophores within these sizes were identified only to genus or family level, e.g. Onychoteuthis spp., Onychoteuthidae. Unidentified paralarvae included deteriorated, strongly twisted specimens lacking arms, eyes, heads and/or chromatophores.
DNA barcoding was performed for the taxonomic identification of the three dominant taxa, for the smallest paralarvae of the Ommastrephidae family (≤2.0–3.0 mm ML), and for paralarvae of species of commercial importance such as Octopus insularis Leite & Haimovici, 2008. Representatives of these taxa were photographed and DNA extraction, amplification and sequencing followed Elías-Gutiérrez et al. (Reference Elías-Gutiérrez, Valdez-Moreno, Topan, Young and Cohuo-Colli2018). DNA strands were sequenced in both directions and were assembled and edited manually, using the Geneious premium 2019.2.1. software (Kearse et al., Reference Kearse, Moir, Wilson, Stone-Havas, Cheung, Sturrock, Buxton, Cooper, Markowitz, Duran, Thierer, Ashton, Meintjes and Drummond2012). A total of 22 sequences (654 bp) of the mitochondrial partial gene COI from individuals with different sizes were analysed. The basic local alignment search tool (BLAST) was applied to each sequence to corroborate the specific identification using the NCBI-BLAST also performed in Geneious software. These sequences were aligned with Clustal W (Thompson et al., Reference Thompson, Higgins and Gibson1994), with default values included in the MEGAX v10.0.5 software (Kumar et al., Reference Kumar, Stecher and Tamura2016). Observed alignment was clean, without gaps, and was translated into amino acid sequences as an additional check of alignment.
These sequences were deposited at BOLD system (available in the dataset Mexican Caribbean molluscs; DS-MOCAR) and at NCBI with the registration GenBank accession numbers. The voucher specimens were deposited in the reference collection at El Colegio de la Frontera Sur, Campus Chetumal.
Horizontal and vertical distribution of total PL abundance of the three dominant taxa, and the environmental variables in the MBRS were plotted using Surfer 14 from Golden Software. Total PL numbers were summed for each station and divided by summed total volumes of seawater filtered for the four sampled strata. This resulted in one depth-aggregated assemblage per station (Muhling et al., Reference Muhling, Smith, Vásquez-Yeomans, Lamkin, Johns, Carrillo, Sosa-Cordero and Malca2013). PL abundances were standardized to 1000 m3, then square-root-transformed prior to further analyses as follows. Differences of PL abundance among lines were assessed through an ANOVA or through a Kruskal–Wallis test when normality assumptions were not satisfied. The paired t test or the Mann–Whitney U test were used to assess the statistical significance of differences in PL abundances between daytime and night-time stations.
Abundances of PL for each depth stratum were calculated in each station, standardized to 1000 m3 and square-root-transformed. Differences among strata were tested using either ANOVA or a Kruskal–Wallis test when normality assumptions were not satisfied. Abundances for the most common taxa were calculated also by depth strata for day and night. Their abundance-weighted mean depths were calculated according to Muhling et al. (Reference Muhling, Smith, Vásquez-Yeomans, Lamkin, Johns, Carrillo, Sosa-Cordero and Malca2013).
Results
In March 2006, 603 cephalopod paralarvae (PL) (8.6 PL 1000 m−3) were collected and 431 PL (6.7 PL 1000 m−3) were collected in January 2007. All were found in the Caribbean Surface Water, defined as being >25°C, oxygen values of 4.2–4.5 ml l−1, and low salinities <36.6 (Figures 2 & 3).
Environmental variable profiles of temperature (T, °C), salinity (S, psu) and dissolved oxygen (DO, ml l−1) along lines A, C, F, G and I from the Mesoamerican Barrier Reef System during March 2006. Isolines indicate the limits of the Caribbean Surface Water (≥25°C, 34.5–36.6 psu) according to Carrillo et al. (Reference Carrillo, Johns, Smith, Lamkin and Largier2016). Paralarval abundances (PL 1000 m−3) are shown in black circles.
Environmental variable profiles of temperature (T, °C), salinity (S, psu) and dissolved oxygen (DO, ml l−1) along lines A, C, G, I, N and P from the Mesoamerican Barrier Reef System during January 2007. Isolines indicate the limits of the Caribbean Surface Water (≥25°C, 34.5–36.6 psu) according to Carrillo et al. (Reference Carrillo, Johns, Smith, Lamkin and Largier2016). Paralarval abundances (PL 1000 m−3) are shown in black circles.
In 2006, larger mean PL abundances were found in the central lines F, G and I (Figure 4A) where relatively weak flows (≤0.25 m s−1, Carrillo et al., Reference Carrillo, Johns, Smith, Lamkin and Largier2015) of the Caiman Current were present. A maximum abundance of 76 PL 1000 m−3 occurred in a surface tow of line I (Figure 2). In 2007 larger PL abundances occurred in surface waters (<25 m depth) of southern Cozumel (line C, characterized by moderate velocity currents ranging 0.2–0.8 m s−1; Carrillo et al., Reference Carrillo, Johns, Smith, Lamkin and Largier2015) and the Gulf of Honduras (Line P, with southward flow and speeds ≤0.25 m s−1). Central lines G and I yielded lower PL abundances than in 2006, as the Yucatan Current (YC) formed south of Chinchorro and a stronger flow exceeding 1.0 m s−1 builds north of that bank (Carrillo et al., Reference Carrillo, Johns, Smith, Lamkin and Largier2015). Lower mean abundances were consistently found in the northernmost line A in both years (Figure 4A), where YC attains the highest speeds >1.25 m s−1 (Figure 5A, E). In this region, coastal upwelling associated with the YC favoured the inshore shallowing of the SUW (Figures 2A & 3A). In this line, the maximum (11 PL 1000 m−3) and the minimum (1 PL 1000 m−3) abundances were found at surface and at 100 m depth (Figure 2).
Mean and standard deviation of paralarval abundance (PL 1000 m−3) at the Mesoamerican Barrier Reef System for 2006 and 2007 by (A) lines from depth-aggregated stations; (B) depth strata from stations; and (C) daytime vs night-time depth-aggregated stations.
Horizontal distribution of depth-aggregated paralarval abundances (PL 1000 m−3) by station in the Mesoamerican Barrier Reef System (black circles) for 2006 (A–D) and 2007 (E–H). (A, E) all paralarvae coupled to surface geostrophic velocities (vectors, m s−1) (modified from Carrillo et al., Reference Carrillo, Johns, Smith, Lamkin and Largier2015); (B, F) Abralia redfieldi; (C, G) Onychoteuthis banksii; and (D, H) Ornithoteuthis antillarum. Octopus insularis single occurrences are denoted with a white × in (A).
Horizontal mean PL abundances (considering stations depth-aggregated) did not show significant differences among lines for 2006 (Kruskal–Wallis test, H = 9.34, df = 4.23, P > 0.05), nor for 2007 (ANOVA, F(4,15) = 1.2, P > 0.05) (Figure 4A). Regarding vertical distribution, PL largest mean abundances (stations-aggregated) were found in the upper 25 m (Figure 4B). Nevertheless, differences in abundances among strata were not statistically significant, for 2006 (ANOVA, F(3,85) = 0.16, P > 0.05) or for 2007 (Kruskal–Wallis test, H = 6.13, df = 3107, P > 0.05), probably because of large variances. PL mean abundance for stations (depth-aggregated) visited during daytime were larger than those occupied during night-time (Figure 4C). However, these differences were not significant for 2006 (Mann–Whitney U test, Z = 1.67, N = 11, P > 0.05) or 2007 (Paired t test, t25 = 0.21, P > 0.05).
Morphological identification of paralarvae found in the MBRS indicated the presence of 12 families, 22 genera, 24 species, 5 morphotypes and a species complex (Table 1). The number of species plus morphotypes was very similar between years, with 24 species in 2006 and 19 in 2007. However, 10 species (Onykia carriboea, Ommastrephes bartrammii, Helicocranchia pfefferi, Lycoteuthis lorigera, Selenoteuthis scintillans, Chiroteuthis veranii, Amphioctopus burryi, Octopus insularis, Argonauta argo and A. hians) occurred only in 2006, while five taxa at species level (Bathothauma lyromma, Helicocranchia papillata, Liocranchia reinhardti, Lycoteuthis springeri and Octopus sp. 2), and a paralarva of the Loliginidae family, were collected only in 2007.
Cephalopod paralarvae (PL) collected in the Mesoamerican Barrier Reef System during March 2006 and January 2007
N, Total catch by numbers; Ab, total abundance expressed as PL 1000 m−3; %, percentage of abundance.Bold numbers indicate total abundance by family.
Three squid families accounted for 85–91% of the total cephalopod abundance by year: Enoploteuthidae (49–61%) represented mainly by Abralia redfieldi, Onychoteuthidae (20–24%) represented by Onychoteuthis banksii, and Ommastrephidae (10–12%) by Ornithoteuthis antillarum. The remaining 9–15% was represented by nine families: Lycoteuthidae, Cranchiidae, Ancistrocheiridae, Thysanoteuthidae, Pyroteuthidae, Chiroteuthidae, Loliginidae, Argonautidae and Octopodidae. Small PL (≤3.0 mm ML) dominated the samples (81% of all). They ranged from 90–93% in Enoploteuthidae, 73–76% in Onychoteuthidae, and 62–80% in Ommastrephidae. Octopus insularis PL were found for the first time in the wild. They were represented by three specimens 1.4, 1.4 and 3.6 mm ML.
DNA barcoding of dominant species corroborated the taxonomic identification for A. redfieldi, O. banksii and O. antillarum, as well as the octopus O. insularis (Table 2). In the BLAST analysis, three out of 22 sequences (654 bp) revealed high similarity with A. redfieldi (99.5–99.8% and E Value 0.0), three sequences with O. banksii (98.7–100% and E Value 0.0), six sequences with O. antillarum (98.2–100% and E Value 0.0) and three with O. insularis (99.8% and E Value 0.0). This analysis (seven sequences) also allowed us to identify smaller ommastrephid PL (≤2.0–3.0 mm ML), which could not have been identified using morphological features, as O. antillarum (Table 2).
Partial COI-gene sequences of the three dominant taxa of cephalopod paralarvae from the Mesoamerican Barrier Reef System
ID, identification; ML, mantle length (mm).
The most abundant and frequent species distributed along the MBRS was Abralia redfieldi. In 2006 larger abundance occurred in nearshore tows of lines C to I. In 2007 it was very abundant in line P where two surface tows yielded 33 and 41 PL 1000 m−3. Moderate abundances were found south of Cozumel Island and north of Chinchorro Bank (Figure 5B, F). Onychoteuthis banksii, the second most abundant taxon, occurred in all lines with a maximum of 25 PL 1000 m−3 in a shallow tow south of Cozumel during 2006 (Figure 5C, G). The third most abundant species, Ornithoteuthis antillarum, showed a similar horizontal distribution as Onychoteuthis banksii. A maximum of 9 PL 1000 m−3 occurred in a shallow tow of line G in 2006 (Figure 5D, H). Octopus insularis PL were found in the shallower layers of shoreward stations of lines C, F and G during 2006 (Figure 5A).
Abralia redfieldi showed evidence of vertical migration with highest abundance of 4 PL 1000 m−3 at the surface at night, while maximum values during daytime occurred below 50 m depth in 2006 (Figure 6A). Onychoteuthis banksii was almost absent from the deepest strata >75 m depth, with higher abundances at the surface at night (3 PL 1000 m−3) than during the day (Figure 6B). Ornithoteuthis antillarum occupied all the strata with abundances decreasing with depth at night and increasing during daytime (Figure 6C). Abundance-weighted mean depths showed slight diel differences of 12 m depth for these three species. Abralia redfieldi and Ornithoteuthis antillarum distributed around 50 m depth day and night, while Onychoteuthis banksii showed the shallowest mean depths at 25 m depth (Figure 6D).
Day (white) and night (black) paralarval abundance vertical distribution of dominant species in the Mesoamerican Barrier Reef System, station-aggregated for 2006 (A–D) and 2007 (E–H). (A, E) Abralia redfieldi; (B, F) Onychoteuthis banksii; and (C, G) Ornithoteuthis antillarum. (D, H) Abundance-weighted mean depths of these species, for both day and night samples.
Depth distributions for 2007 were contrasting with those from the previous year. Maximum day abundances occurred at the surface for the three species (Figure 6E–G). Mean depths during daylight were shallower than during night for Abralia redfieldi and Onychoteuthis banksii (Figure 6H).
Discussion
We report, for the first time, a systematic cephalopod paralarvae study based on physical parameters in the Caribbean Sea, while identifying and corroborating genetically the identity of the three dominant taxa and of Octopus insularis. Abundance and horizontal and vertical distribution of PL from the epipelagic waters of the Mesoamerican Barrier Reef System (MBRS) during March 2006 and January 2007 showed differences in the distribution patterns of the three dominant taxa.
Paralarvae were collected in the Caribbean Surface Water, the relatively well-mixed warm surface layer of the MBRS. Major abundances occurred mainly in the first 25 m depth with no significant differences between day-night collections and in moderate velocity currents ranging 0.2–0.8 m s−1. The PL were encountered at relatively high temperatures of 25–27°C, high oxygen concentrations (4.2–4.5 ml l−1) and salinities (<36.0). In these conditions, the horizontal pattern in PL abundance showed a similar pattern of horizontal distribution to ichthyoplankton (particularly myctophids) and lobster phyllosoma larvae studied in the same campaigns (Muhling et al., Reference Muhling, Smith, Vásquez-Yeomans, Lamkin, Johns, Carrillo, Sosa-Cordero and Malca2013; Canto-García et al., Reference Canto-García, Goldstein, Sosa-Cordero and Carrillo2016). Oceanographic processes can therefore explain their similar distributions.
In March 2006 the central lines (F, G, I) around Chinchorro Bank accounted for larger abundances. The region north-east of this bank (line G) with an average abundance of PL has been identified as another possible area of larval retention because the island-wake effect may produce small eddies probably responsible for accumulation of the plankton (Canto-García et al., Reference Canto-García, Goldstein, Sosa-Cordero and Carrillo2016; Carrillo et al., Reference Carrillo, Lamkin, Johns, Vásquez-Yeomans, Sosa-Cordero, Malca, Smith and Gerard2017).
In January 2007 the largest PL abundance was found in the mesoscale Honduras Gyre at line P. This gyre has also been associated with larval retention, particularly for myctophid larvae and pelagic phyllosomas (Muhling et al., Reference Muhling, Smith, Vásquez-Yeomans, Lamkin, Johns, Carrillo, Sosa-Cordero and Malca2013; Canto-García et al., Reference Canto-García, Goldstein, Sosa-Cordero and Carrillo2016; Carrillo et al., Reference Carrillo, Lamkin, Johns, Vásquez-Yeomans, Sosa-Cordero, Malca, Smith and Gerard2017). Moreover, virtual fish larvae tracking from numerical models in the MBRS showed that the Honduras Gyre potentially could retain particles for over 40 days (Martínez et al., Reference Martínez, Carrillo and Marinone2019, Reference Martínez, Carrillo, Sosa-Cordero, Vásquez-Yeomans, Marinone and Gasca2020). Another region with large abundance of PL was located south of Cozumel Island (line C). It might be associated with the presence of the sub-mesoscale cyclonic eddy, the ‘Cozumel eddy’, which presumably is generated by the separation of the YC from the coast (Carrillo et al., Reference Carrillo, Johns, Smith, Lamkin and Largier2015, Reference Carrillo, Lamkin, Johns, Vásquez-Yeomans, Sosa-Cordero, Malca, Smith and Gerard2017). Sub-mesoscale cyclonic eddies like this one tend to retain and concentrate larvae at their edges. In this area, higher abundances of tuna, myctophid and phyllosoma larvae and pteropods were also observed in the same cruise (Parra-Flores & Gasca, Reference Parra-Flores and Gasca2009; Muhling et al., Reference Muhling, Smith, Vásquez-Yeomans, Lamkin, Johns, Carrillo, Sosa-Cordero and Malca2013; Canto-García et al., Reference Canto-García, Goldstein, Sosa-Cordero and Carrillo2016; Carrillo et al., Reference Carrillo, Lamkin, Johns, Vásquez-Yeomans, Sosa-Cordero, Malca, Smith and Gerard2017). The planktonic environment along much of the MBRS is characterized by low retention conditions (Martínez et al., Reference Martínez, Carrillo and Marinone2019, Reference Martínez, Carrillo, Sosa-Cordero, Vásquez-Yeomans, Marinone and Gasca2020) and with northward transport of PL, largely influenced by the YC and coastal upwelling in the northernmost MBRS.
In sharp contrast with a tendency for larval retention and local recruitment in the southern and impingement regions of the MBRS, the northern region shows northward advection and connectivity (Martínez et al., Reference Martínez, Carrillo and Marinone2019, Reference Martínez, Carrillo, Sosa-Cordero, Vásquez-Yeomans, Marinone and Gasca2020), which yielded the lowest paralarval abundances. In this northernmost region of the MBRS strong YC reaches up to 2.0 m s−1 (Carrillo et al., Reference Carrillo, Johns, Smith, Lamkin and Largier2015, Reference Carrillo, Lamkin, Johns, Vásquez-Yeomans, Sosa-Cordero, Malca, Smith and Gerard2017) and could potentially transport paralarvae from there to the Gulf of Mexico and as far as distant areas such as off Florida Keys in a few days (Martínez et al., Reference Martínez, Carrillo and Marinone2019). The most abundant paralarvae Abralia redfieldi, Onychoteuthis banksii and Ornithoteuthis antillarum were also among the most common adult and juvenile squid collected in the north-eastern Gulf of Mexico (Passarella, Reference Passarella1990; Judkins et al., Reference Judkins, Vecchione, Cook and Sutton2016) as well as in the Straits of Florida (Cairns, Reference Cairns1976). They could most likely come from an upstream source such as the MBRS. However some marine taxa such as reef fishes showed little genetic evidence of connectivity between MBRS and the Florida Keys (Muhling et al., Reference Muhling, Smith, Vásquez-Yeomans, Lamkin, Johns, Carrillo, Sosa-Cordero and Malca2013). Morphological, genetic and growth studies on PL together with regional oceanographic numerical models could lead to a better understanding of the role of the MBRS in exporting marine resources such as cephalopods to other systems.
A rich cephalopod paralarval assemblage comprising at least 24 species, was identified from 1034 specimens collected in only two cruises. The same number of species was identified from 3731 PL in one of the most comprehensive studies on cephalopod paralarvae performed in the western Atlantic Ocean that included 21 oceanographic cruises covering a vast area from north of the Greater Antilles to the Scotian Shelf (Vecchione et al., Reference Vecchione, Roper, Sweeney and Lu2001). However, only seven species coincided in both the western Atlantic and the MBRS. These results indicate that the MBRS supports a paralarval community with a high species richness. A wider spatial and temporal monitoring of PL in this region could have revealed additional species. Our study suggests a greater species richness for the Caribbean than the maximum of 20 species of adult cephalopods from selected regions in this sea (Judkins et al., Reference Judkins, Vecchione, Roper and Torres2010).
Absence of a wide continental shelf on the MBRS accounted for the dominance of the oceanic families Enoploteuthidae, Onychoteuthidae and Ommastrephidae, also dominant in the paralarval assemblages of other Atlantic oceanic waters (Goldman, Reference Goldman1993; Haimovici et al., Reference Haimovici, Piatkowski and Aguiar dos Santos2002; Diekmann et al., Reference Diekmann, Nellen and Piatkowski2006).
Three species, Abralia grimpei (Voss, 1959), Abralia redfieldi and Abralia veranyi (Rüppell, 1844) in adult stage are known to occur in the tropical western Atlantic (Jereb & Roper, Reference Jereb and Roper2010). Adults of the last two species are found in the Florida Straits and in the north-eastern Gulf of Mexico, where A. redfieldi is the most common (Cairns, Reference Cairns1976; Passarella, Reference Passarella1990; Judkins et al., Reference Judkins, Vecchione, Cook and Sutton2016). Paralarvae of these three species currently are not morphologically distinguishable. Nesis (Reference Nesis1975) described A. redfieldi PL from the eastern Caribbean, where it is one of the most numerous species although it avoids the continental coasts. Unfortunately, his description lacks information on the chromatophore patterns that are important for the identification of this species. Otherwise, Abralia veranyi is the most abundant paralarvae found in east Florida (Adams, Reference Adams1997; Erickson et al., Reference Erickson, Roper and Vecchione2017). Paralarvae of A. cf. veranyi have been described by Vecchione et al. (Reference Vecchione, Roper, Sweeney and Lu2001) who considered it the most common species of the genus north of the Caribbean, associated with continental shelf and slope waters. However, these same authors mention that some of the specimens they refer to as A. cf. veranyi might include other Abralia species in the region. Information on the distribution of A. grimpei is scarce, being only reported for the West Indies, North Atlantic and from the northern Sargasso Sea (Tsuchiya, Reference Tsuchiya2009). The COI analysis concluded that A. redfieldi is the only species in the genus found in the MBRS.
At least three species of the Onychoteuthis genus, Onychoteuthis banksii, O. compacta and O. prolata occur in the tropical western Atlantic (Bolstad, Reference Bolstad2010). Descriptions of paralarvae of the latter two (Young & Harman, Reference Young and Harman1987; Sweeney et al., Reference Sweeney, Roper, Mangold, Clarke and Boletzky1992) allowed the identification of some of our specimens. A very abundant morphotype different from O. compacta and O. prolata lead us to identify them as O. banksii, even when the chromatophore pattern in this species is still undescribed. Onychoteuthis banksii was corroborated with the DNA barcoding with a similarity of 98.7–100%. Paralarvae of this species have been collected from February through August in the western North Atlantic (Vecchione et al., Reference Vecchione, Roper, Sweeney and Lu2001). In the MBRS higher abundance of this species in March 2006 compared with January 2007 may be related to the beginning of the reproductive season of the species as these are abundant paralarvae in the Caribbean (Nesis, Reference Nesis1975) and the adjacent Atlantic (Vecchione et al., Reference Vecchione, Roper, Sweeney and Lu2001). The absence of reliable descriptions of small paralarvae (≤3.0 mm ML) precluded a more detailed identification of Onychoteuthis species and 44% of them were left at genus level.
Two forms of ommastrephid PL, rynchoteuthion types ‘A’ and ‘B’, have been described from the North-west Atlantic, and Ornithoteuthis antillarum have been assigned to both in the literature (Roper & Lu, Reference Roper and Lu1979; Goldman, Reference Goldman1993; Vecchione et al., Reference Vecchione, Roper, Sweeney and Lu2001). Rynchoteuthions from the Caribbean described by Nesis (Reference Nesis1975) as Sthenoteuthis pteropus were in fact O. antillarum (Nesis, Reference Nesis1979). We identified O. antillarum by the presence of both ocular and visceral photophores and enlarged lateral proboscis suckers (Nesis, Reference Nesis1979; Sweeney et al., Reference Sweeney, Roper, Mangold, Clarke and Boletzky1992). Spawning peak in the Caribbean takes place in February and March (Nesis, Reference Nesis1975), and our PL abundances for this species were slightly higher in March than in January. Roper & Lu (Reference Roper and Lu1979) considered that their Rynchoteuthion Type ‘B’ belongs to O. antillarum. Thus, this species probably is not so rare in the western North Atlantic (Vecchione et al., Reference Vecchione, Roper, Sweeney and Lu2001). Other authors found that Type ‘A’ are the most abundant paralarvae off southern and eastern Florida, identifying this type solely as O. antillarum (Goldman, Reference Goldman1993) or together with Ommastrephes bartramii (Adams, Reference Adams1997; Erickson et al., Reference Erickson, Roper and Vecchione2017). The DNA barcoding revealed that the smallest paralarvae ≤2.0–3.0 mm ML of this family collected in the MBRS were also O. antillarum, suggesting that this species is the most abundant ommastrephid in the Caribbean. Paralarvae of other ommastrephid species, some with commercial potential in the wider Caribbean such as Sthenoteuthis pteropus, were particularly scarce in this study.
Our study reports the first three paralarvae of Octopus insularis found in the wild as they have only been described in captivity (Lenz et al., Reference Lenz, Elias, Leite and Vidal2015). Recent studies have increased its known distribution in the western tropical Atlantic, with a single adult reported in the MBRS at Isla Mujeres (Lima et al., Reference Lima, Berbel-Filho, Leite, Rosas and Lima2017). This octopus supports commercial fisheries in distant waters such as Veracruz reefs and central and northern Brazil (Sauer et al., Reference Sauer, Gleadall, Downey-Breedt, Doubleday, Gillespie, Haimovici, Ibáñez, Katugin, Leporati, Lipinski, Markaida, Ramos, Rosa, Villanueva, Arguelles, Briceño, Carrasco, Che, Chen, Cisneros, Conners, Crespi-Abril, Kulik, Drobyazin, Emery, Fernández-Álvarez, Furuya, González, Gough, Krishnan, Kumar, Leite, Lu, Mohamed, Nabhitabhata, Noro, Petchkamnerd, Putra, Rocliffe, Sajikumar, Sakaguchi, Samuel, Sasikumar, Wada, Zheng, Tian, Pang, Yamrungrueng and Pecl2019), so pelagic paralarvae may interconnect different populations along this geographic range. Remarkably, no PL of Octopus vulgaris which supports a fishery downstream in the neighbouring Campeche Bank (Sauer et al., Reference Sauer, Gleadall, Downey-Breedt, Doubleday, Gillespie, Haimovici, Ibáñez, Katugin, Leporati, Lipinski, Markaida, Ramos, Rosa, Villanueva, Arguelles, Briceño, Carrasco, Che, Chen, Cisneros, Conners, Crespi-Abril, Kulik, Drobyazin, Emery, Fernández-Álvarez, Furuya, González, Gough, Krishnan, Kumar, Leite, Lu, Mohamed, Nabhitabhata, Noro, Petchkamnerd, Putra, Rocliffe, Sajikumar, Sakaguchi, Samuel, Sasikumar, Wada, Zheng, Tian, Pang, Yamrungrueng and Pecl2019) was found.
In 2006 only Abralia redfieldi PL showed clear evidence of performing a diel vertical migration (DVM), while there was no evidence of DVM for Onychoteuthis banksii and Ornithoteuthis antillarum. For 2007 no clear evidence of DVM was shown for any species at all. In the eastern Atlantic, paralarval enoploteuthids did show vertical migrations while onychoteuthids did not (Diekmann et al., Reference Diekmann, Nellen and Piatkowski2006). On the contrary, Young & Harman (Reference Young and Harman1987) reported significant differences between day–night catches of PL of two Onychoteuthis species in Hawaii, ranging mainly 3.0–5.9 mm ML. Juveniles of Abralia and Onychoteuthis are known to perform diel migrations in the Straits of Florida (Cairns, Reference Cairns1976; Roper & Young, Reference Roper and Young1975). In contrast O. antillarum PL occur mainly in the thermocline water layer, avoiding uppermost waters, and do not perform vertical migrations (Nesis, Reference Nesis1979; Arkhipkin et al., Reference Arkhipkin, Laptikhovsky, Nigmatullin, Bespyatykh and Murzov1998), as we found in the MBRS. Vertical distribution of O. antillarum recorded here in the MBRS, however, closely resembles that of ommastrephid Type ‘A’ off the Florida Keys (Goldman & McGowan, Reference Goldman and McGowan1991) by night in 2007, but by day in 2006.
Vecchione et al. (Reference Vecchione, Roper, Sweeney and Lu2001) did not find consistent diel patterns in paralarval vertical distribution. They proposed that diel variability in abundance observed in surface samples probably was a result of changes in the ability of young squids to visually avoid the sampler in different light conditions. Vertical distribution patterns of squid paralarvae differ among squid taxa and at least in some cases DVM behaviour appears to develop in post-paralarval stages (Vecchione, Reference Vecchione1987; Bower & Takagi, Reference Bower and Takagi2004; Shea & Vecchione, Reference Shea and Vecchione2010). In the MBRS dominance of small PL sizes at which morphology, swimming and feeding abilities may not be fully developed could account for an absent or very limited DVM. Prevalence of small PL also could indicate reproductive activity during the winter.
Acknowledgements
The authors thank the captain and crew of the NOAA ship ‘Gordon Gunter’. Ivan Méndez improved some drawings. Alma E. Morales supported during the genetic analysis. This contribution is part of G.C.E. and L.A.D. MSc thesis requirements in the postgraduate programmes of ECOSUR and Ciencias del Mar y Limnología, UNAM, respectively, for which they received a CONACYT scholarship grant. R.S.D. is an EDI-IPN and COFAA-IPN fellow. Two anonymous reviewers greatly improved this work through their criticism.
Financial support
This work was partially funded by NOAA's Coral Reef Conservation Program #1244 led by J. Lamkin. G.C.E. and L.A.D. performed a research stay at Instituto Politécnico Nacional (CICIMAR-IPN) supported by the research project ‘Colección biológica de referencia de paralarvas de cefalópodos’ SIP20161524. This publication is a contribution to the Mexican Barcode of Life (MEXBOL) Nodo Chetumal network supported by CONACYT (251085).
