Hydrographic and fish larvae distribution during the “Godzilla El Niño 2015–2016” in the northern end of the shallow oxygen minimum zone of the Eastern Tropical Pacific Ocean

Authors


Abstract

Based on hydrographic data and vertical distributions of tropical species of fish larvae (Diogenichthys laternatus, Vinciguerria lucetia, Bregmaceros bathymaster, and Auxis spp.), effects of “Godzilla El Niño 2015–2016” in the shallow oxygen minimum zone off Mexico were analyzed. Zooplankton samples were collected during four cruises, before (February 2010 and April 2012) and during (June 2015 and March 2016) the warm event. Temporal series of sea surface temperature revealed that June 2015 was the warmest June of the last years. Conservative temperature was >2°C higher than normal in the surface mixed layer, and the suboxic layer (4.4 µmol/kg) reached as shallow as 100 m depth. Unexpected results were that larval abundances were relatively high during the warm event, unlike zooplankton volumes, which declined. Before the warm event, V. lucetia and Auxis spp. were more abundant in the surface mixed layer, while B. bathymaster and D. laternatus dominated in the thermocline and shallow hypoxic layer (44 µmol/kg). However, during the event in June 2015, all species were most abundant in the surface mixed layer, which implied that the species adapted to hypoxia had inverted their normal pattern of distribution, possibly as consequence of the rise of the suboxic layer; however, further observations are required to confirm this generality. Results showed no dramatic change in the total larval abundance during the warm event. Nevertheless, a differential response in their vertical distribution was evident in association with changes in the depth of the shallow hypoxic and suboxic layers. This might indicate adaptability of tropical species to prolonged periods of warming in the oceans.

1 Introduction

Strong interannual changes indicated by intense positive anomalies in sea surface temperature (SST) and wind anomalies have been reported throughout the Eastern Pacific Ocean during recent years [Bond et al., 2015; Hartmann, 2015; Schiermeier, 2015; Whitney, 2015; Stramma et al., 2016; Di Lorenzo and Mantua, 2016]. The highest positive anomalies of SST since at least the 1980s over an extended area of the Eastern North Pacific were recorded in early winter 2014 and found to be independent of the remote phenomenon of El Niño-Southern Oscillation (ENSO) [Bond et al., 2015; Whitney, 2015; Peterson et al., 2015; Cavole et al., 2016; Di Lorenzo and Mantua, 2016]. Because of its great extent, strength, and environmental impact, the anomaly was referred to as “The Blob” [Bond et al., 2015]. Similar anomalies unrelated to the ENSO were also detected in the northeastern Tropical Pacific off Mexico, although they were weaker than further north [Cavole et al., 2016; Di Lorenzo and Mantua, 2016].

Consecutive to the establishment of “The Blob,” strong signals of developing warm ENSO event were recorded in the Tropical Pacific from late spring 2015 to early summer 2016 (http://www.esrl.noaa.gov/psd/enso/mei/table.htm). Although it is not known if “The Blob” influenced the development of this event, the warming in some regions of the Eastern Tropical Pacific was unusually prolonged [Di Lorenzo and Mantua, 2016; Jacox et al., 2016]. The strong start of the event generated the expectation of its being one of the warmest ENSO events in the last 50 years [Whitney, 2015], and it was popularly nick-named the “Godzilla El Niño” [Schiermeier, 2015]. Leising et al. [2015] and Stramma et al. [2016] were among the first to describe its development in the northern and southeastern subtropical Pacific, respectively. They detected a clear increase in SST and decrease in nutrient concentration in the upper 200 m layer, as a response to strong wind variations. To date, little is known of the biological consequences arising from the changes in the water column properties related to this last warm ENSO event, despite the importance of understanding the ocean's ecological sensitivity to these recent and ongoing changes.

One of the largest naturally shallow Oxygen Minimum Zones (OMZ) of the world is located in the Eastern Tropical Pacific off Mexico [Fiedler and Talley, 2006; Prince and Goodyear, 2006; Stramma et al., 2010]. There is no consensus on the oxygen threshold defining an OMZ [Hofmann et al., 2011]. Moreover, the critical level of oxygen uptake in marine organisms varies widely because it is linked to pressure and the hydrographic conditions, as well as to metabolic changes during the life cycle of each species [Hofmann et al., 2011]. However, it has been reported that when oxygen concentrations drop below ∼44 µmol/kg or 1 mL/L (denoted here as hypoxic conditions) epipelagic species may be stressed. When the oxygen concentrations fall below ∼4.4 µmol/kg (denoted here as suboxic conditions) some epipelagic and mesopelagic fish species organisms may die [Gray et al., 2002; Diaz and Rosenberg, 2008; Stramma et al., 2008; Sánchez-Velasco et al., 2016].

The OMZs, according to predictive models, are presently undergoing vertical and horizontal expansion and concomitant reduction in the thickness of the near-surface oxygenated layer (>44 µmol/kg) [Prince and Goodyear, 2006; Diaz and Rosenberg, 2008]. This upper layer is the habitat of the planktonic phases of most pelagic species, including those of fish, squid, and commercially important crustaceans [Fernández-Alamo and Färber-Lorda, 2006; Davies et al., 2015; Sánchez-Velasco et al., 2016]. While this reduction of the oxygenated habitat must be generating stress in the planktonic organisms, the impact of an interannual event like the “Godzilla El Niño 2015–2016” preceded by “The Blob” might increase significantly the thermal and trophic (food availability) stresses, and thereby might affect the vertical distribution and survival of the planktonic organisms.

Fish larvae of tropical species that inhabit a range from the southern California Current system to northern Chile in the Tropical and Subtropical Oriental Pacific, with planktonic phases of ∼8–12 days [Moser, 1996], may provide a good biological indicator of interannual changes because of their wide distribution in the ocean and their differential response to the structure of the water column [Moser and Smith, 1993; Apango-Figueroa et al., 2015; Davies et al., 2015; Contreras-Catala et al., 2016]. Good indicators would include (i) larvae of Vinciguerria lucetia, which have high abundance and frequency in the oceanic domain and have showed strong affinity to the surface mixed layer and thermocline; (ii) larvae of Diogenichthys laternatus and Bregmaceros bathymaster, which have been recorded predominantly in the thermocline and shallow hypoxic layer; and (iii) larvae of Auxis spp., which have been recorded mainly in the surface mixed layer. In this context, it might be expected that unpredictable changes in the vertical distribution and survival rate of fish larvae could occur in response to changes in the water column structure and hypoxic and suboxic layer depth generated by interannual events.

As a contribution to the understanding of the biological response to these interannual events, this study describes changes in the hydrographic structure of the water column and biological indicators (fish larvae of D. laternatus, V. lucetia, B. bathymaster, and Auxis spp. larvae) in the pelagic ecosystem in the OMZ of the Eastern Tropical Pacific off Mexico before establishment, and during the development, of the warm “Godzilla El Niño event of 2015–2016.” Comparison of the hydrography and the vertical distribution of fish larvae among the oceanographic cruises that support this study, framed in a time series of satellite data (1981–2016), will provide insight into the evolution of the system during the period of anomalous warming.

2 Methodology

2.1 Interannual Events by Satellite Data

To obtain a synoptic vision of the study region and provide the context of the field work, multisensor SST and CHL composites (Modis Aqua, Modis Terra, and VIIRS sensor) were built at level 3 Local Area Cover (LAC) (1 km resolution). The level 1b data were obtained from the NASA Ocean Color Data web page (http://oceandata.sci.gsfc.nasa.gov/). These multisensory composites (Figure 1) represent the multisensory imagery in the cruise periods. The level 1b to level 2 conversion was carried out using SEADAS 6 and the level 2 to level 3 multisensory imagery processing was made using WIM software (http://wimsoft.com/) with the WIM automation Module (WAM) ver. 9.10.

Figure 1.

SST and CHL satellite images at 1 km resolution. Cruises made before “Godzilla El Niño 2015–2016”: (a) SST and (b) CHL from 14 to 27 February 2010; (c) SST and (d) CHL from 26 April 2012 to 4 May 2012. Cruises made during “Godzilla El Niño 2015–2016”: (e) SST and (f) CHL from 3 to 16 June 2015; (g) SST and (h) CHL from 2 to 8 March 2016. White circles, zooplankton and CTD sampling stations; and black circles, CTD-only sampling stations. Red lines indicate the limits of the area used to calculate SST anomalies in the Figures 2 and 3. LP, Bahía de La Paz; SJ, San José del Cabo; IM, Islas Marias; CC, Cabo Corrientes.

The temporal context of the four cruises within recent interannual activity is provided by monthly composites of multisensor SST (AVHRR, MODIS Aqua, MODIS Terra, VIIRS) at Global Area Cover (4 km resolution) over the years 1981–2016 for the region of the southern Gulf of California and the adjacent Pacific enclosed by red lines in Figure 1a, encompassing the area of field work. These data were processed using the imagery bank in 1 km by pixel size. From these time series, standardized anomalies were calculated following Santamaría-del-Angel et al. [2011]. Negative anomalies represent SST values below average, and positive anomalies those above average.

For illustration of recent interannual activity (from 2003 to 2016), the data were also represented as “area anomalies,” the relative proportion (%) of the area occupied by waters with different ranges of SST standardized anomalies from the southern Gulf of California to adjacent Pacific off Cabo Corrientes (Figure 2). Standardized anomalies were extracted from the overall series at the sampling station positions in the alongshore transect for comparison with the in situ data.

Figure 2.

Series monthly of SST area anomalies (see text for definition) for the years 2002–2016 (see Figure 1 for area). The data are represented as the relative proportion (%) of the area occupied by waters with different ranges of SST standardized anomalies. Yellow line represents the Southern Oscillation Index (SOI) over the same period.

Finally, to compare the SST from June of different years, monthly values of SST multisensor 1 km pixel size (2010, 2012, 2015, and 2016) for the sampling station locations in the alongshore transect were extracted.

The Southern Oscillation Index (SOI) (http://www.bom.gov.au/climate/current/soihtm1.shtml) monthly anomalies were plotted over the monthly % areas of the SST anomalies. Low SOI anomalies values (<−1.6) represent strong ENSO events.

2.2 Field Methods

Details of the four oceanographic cruises are provided in Table 1. The oceanographic cruises in February 2010 and April 2012, onboard the R/V “Francisco de Ulloa” (CICESE), were made before the warm ENSO event, while the cruises made in June 2015 and March 2016, onboard the R/V “Alpha Helix” (CICESE), were during the event according to NOAA reports (https://www.ncdc.noaa.gov/teleconnections/enso/indicators/soi/). During the four oceanographic cruises, vertical profiles were obtained at each station using a SeaBird 911plus CTD probe equipped with dissolved oxygen and fluorescence sensors. Conservative temperature Θ (°C) and absolute salinity (SA, g/Kg) were calculated from in situ temperature and practical salinity with the TEOS-10 (Thermodynamic Equation of Seawater-2010) software, which was downloaded from http://www.TEOS-10.org [IOC, SCOR, and IAPSO, 2010; Pawlowicz et al., 2010].

Table 1. Details of the Four Oceanographic Cruises Made From Southern Gulf of California to off Cabo Corrientes Area, Before (February 2010 and April 2012) and During (June 2015 and March 2016) “Godzilla El Niño 2015–2016”
Cruise NameMARIAS1002MARIAS1204MARIAS1506MARIAS1603
Cruise dataFrom 14 to 27 February 201026 April to 4 May 2012From 3 to 16 June 2015From 2 to 8 March 2016
Physical stations33413628
Zooplankton hauls15191720
Zooplankton samples45574457

The ocean surface mixed-layer depth was calculated following the methodology of Kara et al. [2000] which consists of a gradient-based criterion having a fixed temperature difference of 0.8°C and variable salinity. The thermocline was defined as the temperature band ±1°C centered on the depth of the maximum temperature gradient (δT/δz), which varied between cruises.

Based on prior knowledge of the vertical distribution of fish larvae [e.g., Danell-Jiménez et al., 2009; Davies et al., 2015; Sánchez-Velasco et al., 2016], three depth strata were selected at each sampling station. The first stratum was essentially the surface mixed layer (from the upper limit of the thermocline to surface). The second covered the thermocline and the chlorophyll a maximum layer, coinciding in most cases with the 44 µmol/kg oxypleth, and the third was from the bottom limit of the shallow hypoxic layer (4.4 µmol/kg oxypleth) to the base of the thermocline/chlorophyll a maximum layer, coinciding in most cases with the 44 µmol/kg oxypleth. The depth of each net haul was decided after a visual inspection of the CTD profile that preceded each zooplankton tow. The profiles of temperature, chlorophyll a, and dissolved oxygen of each station can be found in Godínez et al. [2010, 2012, 2015, 2016].

The hauls were performed day and night using opening-closing conical zooplankton nets with a 60 cm mouth diameter, 250 cm net length, and 505 μm mesh size (http://www.generaloceanics.com). To estimate the true depth of each zooplankton tow, the depth of the net was calculated by the cosine of the wire angle method, following the standard specifications of Smith and Richardson [1979]. The volume of filtered water was calculated using calibrated flowmeters placed in the mouth of each net. Samples were fixed with 5% formalin buffered with sodium borate.

Zooplankton displacement volume [Beers, 1976; Smith, 1971; Kramer et al., 1972] was standardized to mL/1000 m3. Fish larvae of V. lucetia and D. laternatus (mesopelagic species), B. bathymaster (neritic pelagic), and Auxis spp. (epipelagic) were removed from all samples. These were identified according to the descriptions in Moser [1996], and abundance was standardized to number of larvae per 10 m2 [Smith and Richardson, 1979; Moser and Smith, 1993]. Early juveniles were not considered in the study.

2.3 Statistical Analysis

For each cruise, the nonparametric Kruskal-Wallis test [Sokal and Rohlf, 1985; Siegel and Castellón, 1988] was used to assess the statistical significance of equals in zooplankton displacement volume and total larval abundance between daytime and nighttime.

In addition, this test was used to assess the statistical significance of differences in zooplankton displacement volume and total larval abundance between the cruises made before and during the ENSO event. When the null hypothesis was rejected, a Mann-Witney test was used to establish whether significant differences occurred between cruises [Daniel, 2008].

A canonical correspondence analysis [Ter Brack, 1986] was run to define the relation between environmental variables and larval fish distribution for the four cruises, after fourth-root transformation of the standardized biological data. This matrix contained the zooplankton displacement volume (mL/1000 m3) of each stratum and the stratum-average values of conservative temperature (Θ °C), absolute salinity (SA, g/kg), chlorophyll a (mg/m3), and dissolved oxygen (µmol/kg).

3 Results

3.1 Interannual Events by Satellite Data

The monthly areas anomalies from the multisensor SST data over the last 14 years of the southern Gulf of California and the adjacent Pacific (Figure 2) showed an alternation of positive and negative values ≤1 SD throughout the series. The strongest and longest-lasting positive anomalies (>2 SD) occurred in the years 2003 and 2010, and from 2014 to 2016. This last warm period appears to be the most intense and prolonged of the last decades. The signs of the developing “The Blob” in the study region were evident in 2014 and the “Godzilla” ENSO was identifiable from early 2015 to summer 2016. The SOI Index monthly anomalies (see yellow line in Figure 2) showed a general negative relation with the area anomalies, in particular, the longest negative (<−1.6 SD) period was between 2014 and 2016, coinciding with the prolonged warning.

Standard SST anomalies in the sampling station positions of the alongshore transect of each cruise (Figures 3a and 3b) showed that June 2015 was the warmest with positive anomalies >1 SD in the entire transect, during the peak of the warm ENSO event. In contrast March 2016 had anomalies between −0.1 and −0.5 SD throughout the transect, which suggests the warm event was already relaxing in the area. The other two cruises (before the event) showed larger negative anomalies ranging between −0.5 and −1.0 SD. A similar pattern was seen using the in situ temperature anomalies from the four cruises, although the deviations were larger because of the instantaneous nature of the ship sampling (not shown). The June comparisons, corresponding to the years when the cruises took place, indicated that June 2015 was the warmest, 2–3°C warmer than the others, reaching up to 30°C off Cabo Corrientes (Figure 3c).

Figure 3.

(a) Sampling station locations in the alongshore transect made from the southern Gulf of California to off Cabo Corrientes area. (b) Standardized anomalies of February 2010 (black line), March 2016 (black red), April 2012 (green line), and June 2015 (blue line) based on monthly time series from August 1981 to July 2016 constructed from SST multisensor (AVHRR, MODIS Aqua, MODIS Terra, VIIRS) (4 km resolution). (c) Comparison of the SST from June of different years from the extracted monthly values of SST multisensor (2010, 2012, 2015, and 2016) in the sampling station locations in the alongshore transect (see Figure 3a).

3.2 Synoptic Vision and Hydrographic Conditions

3.2.1 Surface Patterns

SST and CHL distribution patterns shown by satellite data (Figure 1) reveal environmental differences between the cruises made before and during the ENSO event. Before the event (February 2010 and April 2012), the maximum values of SST were ∼28°C (Figures 1a and 1c) while during the ENSO event (June 2015 and March 2016), the highest values reached almost 30°C (Figures 1e and 1g).

The study region was characterized by oligotrophic conditions (Figures 1b, 1d, 1f, and 1h), although the CHL distribution showed changes between the cruises made before and during the ENSO event. Before the event, the CHL values were >0.5 mg/m3 (Figures 1b and 1d), with the highest values of the study (>2 mg/m3) found in April 2012 off Cabo Corrientes, during the seasonal upwelling. During the cruises made during the ENSO event, the CHL decreased in all region (Figures 1f and 1h), with values <0.5 mg/m3.

3.2.2 Hydrographic Structure of the Water Column

The distributions of the conservative temperature revealed clear changes in the upper 300 m between the cruises made before and during the ENSO even, particularly in the first 100 m depth. Before the event, the vertical distribution of the conservative temperature revealed a range from 12 to 25°C (Figures 4a and 4b), while during the event it ranged from 12 to ∼28°C (Figures 4c and 4d).

Figure 4.

Vertical distribution of physical and chemical parameters in the alongshore transect made from the southern Gulf of California to off Cabo Corrientes area. Cruises made before “Godzilla El Niño 2015–2016”: February 2010 and April 2012; and cruises made during “Godzilla El Niño 2015–2016”: June 2015 and March 2016. (a–d) Conservative temperature (°C); (e–h) dissolved oxygen concentration (µmol/kg); and (i–l) absolute salinity (g/kg). Blue line is surface mixed-layer depth; and red line is thermocline depth.

In February 2010 before the event (Figure 4a), the surface mixed layer (marked in blue in Figure 4) reached up to 60 m depth (24°C) near Islas Marias, and the thermocline (∼18 and 20°C isotherms, marked in red in Figure 4) fluctuated between ∼60 and 80 m depth along the transect. Similar temperature values were observed in April 2012 (Figure 4b), but the surface mixed layer and the thermocline were the shallowest of the study, fluctuating between ∼30 and 40 m depth. In contrast, in June 2015 during the ENSO event (Figure 4c), the surface mixed layer extended to ∼80 m depth near the Islas Marias (28°C), although its depth reduced gradually northward, as did the depth of the thermocline (lying here between ∼22 and 24°C isotherms). In March 2016 during the ENSO event (Figure 4d), the surface mixed layer and thermocline exhibited high temperature, but with similar distribution gradients to the cruises before the event.

Concentrations in the highly oxygenated (>200 µmol/kg) near-surface layers were lower in June 2015 during the warming event (Figures 4e–4h). The hypoxic water (<44 µmol/kg) was detected below the thermocline in all cruises. This hypoxic layer shoaled from about 100 m in the southern Gulf of California to Cabo Corrientes in February 2010 and April 2012 before the event, while in June 2015 during the ENSO event, the pattern was reversed (Figure 4g). However, the most evident effect during June 2015 was the thinning of the hypoxic layer due to the rise to <100 m depth of the suboxic layer (4.4 µmol/kg oxypleth) in great part of the alongshore transect. In March 2016 during the ENSO event, the oxygen distribution was similar to the cruises made in 2010 and 2012.

Before the ENSO event, the absolute salinity in the upper 300 m ranged between 34.6 and 35 g/kg in February 2010, and between 34.8 and 35.4 g/kg in April 2012 (Figures 4i and 4j), while during the ENSO event the salinity values fluctuated between 34.4 and 35 g/kg in both March 2016 and June 2015 cruises (Figures 4k and 4l). The area of lowest salinity was found off Cabo Corrientes, where its major vertical and horizontal extension occurred during the ENSO event cruises (Figures 4k and 4l). The highest values were detected in the entrance of the Gulf, covering a major area before the event (Figures 4i and 4j).

The chlorophyll a concentrations fluctuated between 0.1 and 2 mg/m3, typical of oligotrophic water conditions in the region. Maximum chlorophyll a values occurred near the thermocline in all cases, but these were higher before the event (>1 mg/m3), than those observed in March during the ENSO event (<1 mg/m3). In the June 2015 cruise, no chlorophyll a data were available because of technical problems with the equipment.

3.3 Zooplankton Displacement Volume and Larvae of Fish Species

There were no statistically significant differences in the zooplankton displacement volume and in the total larval fish, between day and night hours (P >0.05) in the four cruises, and so the data were combined for the analysis.

Zooplankton displacement volume differed significantly between the cruises before and during the ENSO event (P < 0.05), with greater volumes in February 2010 and April 2012 before the event. To summarize the variability, the average zooplankton displacement volume and total larval fish abundance per sampling stratum (transformed to fourth root) and average vertical profiles of hydrographic parameters were calculated over all stations in each transect (Figure 5).The highest zooplankton displacement volumes (>1000 mL/1000 m3) were observed in April 2012 before the ENSO event (Table 2 and Figure 5b), and the lowest (≤100 mL/1000 m3) in June 2015 during the ENSO event (Table 2 and Figure 5c). The zooplankton displacement volumes were mostly highest in the oxygenated and warm surface mixed layer, and decreased with depth, except in February, when they were also high in the hypoxic layer.

Figure 5.

Vertical distribution of the average of the zooplankton displacement volume (mL/1000 m3) and total larval fish (larvae/10 m2) per sampling stratum fourth root transformed and the average vertical profiles of hydrographic parameters in the alongshore transect from the southern Gulf of California to off Cabo Corrientes area. Cruises made before “Godzilla El Niño 2015–2016”: February 2010 and April 2012; and cruises made during “Godzilla El Niño 2015–2016”: June 2015 and March 2016. (a–d) Zooplankton displacement volume (mL/1000 m3) and conservative temperature (°C); (e–h) total larval fish (larvae/10 m2) and dissolved oxygen concentration (µmol/kg).

Table 2. Average Zooplankton Displacement Volume (mL/1000 m3) and Average Larval Total Abundance (Larvae/10 m2) Collected in the Alongshore Transect From Southern Gulf of California to off Cabo Corrientes Area, Before (February 2010 and April 2012) and During (June 2015 and March 2016) “Godzilla El Niño 2015–2016”a
 February 2010April 2012June 2015March 2016
 MLTHLMLTHLMLTHLMLTHL
  1. a

    ML, surface mixed-layer stratum; T, thermocline layer stratum; HL, hypoxic layer stratum.

Total larval abundance41.347.529.3795.72021.6957.4556.859.535.685.911.223.3
Zooplankton displacement volume500.7223.0469.01511.3714.3517.055.99.217.8252.932.162.0

The fish larvae composition showed higher number of taxa during the warm ENSO event, with 83 taxa in June 2015, and 88 taxa in March 2016. A relatively low number of taxa (≤60 taxa) was detected before the ENSO event (Table 3).

Table 3. Taxonomic Composition of the Total Larval Fish Obtained in the Four Oceanographic Cruises Made From Southern Gulf of California to off Cabo Corrientes Area, Before (February 2010 and April 2012) and During (June 2015 and March 2016) “Godzilla El Niño 2015–2016”
 2010201220152016
Families32334545
Genera53566973
Species51535651
Taxa58608388

The total larval fish abundance had a different distribution pattern to the zooplankton displacement volume. In this case, there were no statistically significant differences in the total larval fish between the cruises made before and during the ENSO event (P > 0.05). The highest (>1000 larvae/10 m2 in April 2012) and lowest (≤100 larvae/10 m2 in February 2010) total larval fish abundances were observed before the ENSO event (Table 2 and Figures 5e and 5f). Relatively high values (∼between 100 and 500 larvae/10 m2) were presented in the cruises made during the event (Table 2 and Figures 5g and 5h), where the maximum values were observed in the oxygen and warm surface layer, decreasing with depth.

To summarize the variability, the average abundance of the fish larvae of each species per sampling stratum (transformed to fourth root) and average vertical profiles of hydrographic parameters were calculated over all stations in each transect (Figure 6). The highest average larval abundance of Auxis spp. was recorded in the surface mixed layer in June 2015 and March 2016, both during the warm ENSO event (Table 4 and Figures 6c and 6d). In all cases (Figure 6), the larval abundance decreased with depth, except in February 2010 (before the ENSO), where these were almost absents from the water column.

Figure 6.

Vertical distribution of the average abundance of the fish larvae (larvae/10 m2) per sampling stratum fourth root transformed and the average vertical profiles of hydrographic parameters in the alongshore transect from the southern Gulf of California to off Cabo Corrientes area. Cruises made before “Godzilla El Niño 2015–2016”: February 2010 and April 2012; and cruises made during “Godzilla El Niño 2015–2016”: June 2015 and March 2016. (a–d) Larvae of Auxis spp. (larvae/10 m2) and conservative temperature (°C). (e–h) Larvae of V. lucetia (larvae/10 m2) and absolute salinity (g/kg). (i–l) Larvae of B. bathymaster (larvae/10 m2) and density (kg/m3). (m–p) Larvae of D. laternatus (larvae/10 m2) and dissolved oxygen (µmol/kg).

Table 4. Average Larval Abundance of the D. laternatus, V. lucetia, B. bathymaster, and Auxis spp. (Larvae/10 m2) Collected in the Alongshore Transect From Southern Gulf of California to off Cabo Corrientes Area, Before (February 2010 and April 2012) and During (June 2015 and March 2016) “Godzilla El Niño 2015–2016”a
 February 2010April 2012June 2015March 2016
 MLTHLMLTHLMLTHLMLTHL
  1. a

    ML, surface mixed-layer stratum; T, thermocline layer stratum; HL, hypoxic layer stratum.

Auxis spp.0.30.20.120.223.47.049.00.42.819.80.40.2
V. lucetia16.75.01.7551.91166.1191.3217.15.436.161.35.14.6
B. bathymaster11.923.217.836.5192.7330.172.014.59.619.815.933.3
D. laternatus0.43.02.111.184.0115.153.610.118.91.32.722.5

V. lucetia larvae also showed their highest average abundance in the surface mixed layer and decreased with depth in all cases (Table 4 and Figures 6e–6h). In April 2012 before the event, the greatest larval concentrations were also recorded in the thermocline. Relatively high abundance occurred in the cruises made during the ENSO event (Figures 6g and 6h).

Unlike the distribution pattern of V. lucetia larvae, B. bathymaster, and D. laternatus larvae displayed their greatest larval concentrations in the thermocline and hypoxic layer (Table 4 and Figures 6i–6p), although in June 2015 during the ENSO event the highest larval abundance of both species were detected in the surface mixed layer (Figures 6k and 6o), i.e., reversed their normal distribution.

The canonical correspondence analysis (Figure 7) detected correlation between the environmental variables from the four cruises and the larval fish distribution. The first axis had a Pearson correlation of 0.80 with an eigenvalue of 0.48 (Table 5). Axis 2 had a correlation of 0.53 with an eigenvalue of 0.1 (Table 5). The dissolved oxygen was the best correlated with the Axis 1 (−0.53) and the conservative temperature with the Axis 2 (0.52) (Table 5). Most of the samples of the two first sampling strata (marked with 1 and 2) from cruises made during the ENSO event, June 2015 (white square) and March 2016 (black square), were correlated with the highest temperature, as is observed in the right upper quadrant (Figure 7) where the centroid of Auxis spp. larvae is located. On the other hand, the first sampling stratum of the cruise made in April 2012 before the ENSO event (gray circle) was correlated with high temperature and the highest dissolved oxygen and zooplankton displacement volume values as seen in the left upper quadrant. Under this last quadrant, samples of the second sampling stratum of this cruise were correlated with the highest salinity values. The centroid of V. lucetia larvae was observed near to these last two quadrants. Most of the samples of the two first strata from the February 2010 (white circle) were correlated with intermediate temperature and low zooplankton displacement volume and low dissolved oxygen concentrations, where B. bathymaster and D. laternatus larvae centroids were displayed. Finally, most of the samples of the third sampling stratum of all cruises (marked with 3) were inversely correlated with the four variables, as observed in the lower right quadrant.

Figure 7.

The Canonical Correspondence Analysis (CCA) results show fish larvae abundance (D. laternatus, V. lucetia, B. bathymaster, and Auxis spp. centroids) and hydrographic conditions from of sampling stations alongshore transect from southern Gulf of California to off Cabo Corrientes area. Cruises made before “Godzilla El Niño 2015–2016”: February 2010 (white circle) and April 2012 (gray circle); and cruises made during “Godzilla El Niño 2015–2016”: June 2015 (white square) and March 2016 (black square). Number 1, surface mixed-layer stratum; number 2, thermocline stratum; number 3, hypoxic layer stratum. Tem, conservative temperature; Oxi, dissolved oxygen; Bio, zooplankton displacement biomass; Sal, absolute salinity.

Table 5. Eigenvalues and Contribution Percentage of the Explanatory Variables; and Correlations Among Environmental Variables and Ordination Axes by the Canonical Correspondence Analysis
 Axis 1Axis 2Axis 3
Eigenvalues0.480.100.03
Variance in species data   
% of variance explained32.096.81.59
Cumulative % explained32.0938.8940.48
Pearson correlation0.800.530.25
Conservative temperature (°C)−0.070.520.005
Absolute salinity (g/kg)−0.51−0.020.13
Dissolved oxygen (µmol/kg)−0.530.39−0.03
Zooplankton displacement volume (mL/1000 m3)−0.440.080.20

4 Discussion

The effects “Godzilla El Niño 2015–2016” in the pelagic ecosystem of the northern end of the shallow oxygen minimum zone off Mexico are described by comparison of the hydrographic structure and the vertical distribution of the dissolved oxygen and fish larvae of abundant tropical species (V. lucetia, D. laternatus, B. bathymaster, and Auxis spp.) during four oceanographic cruises. Two were made before the warm ENSO event (February 2010 and April 2012) and two during the ENSO event (June 2015 and March 2016) according to the SOI Index. The study was supported by SST series that showed the interannual variations attributable to the warn ENSO event preceded by “The Blob” effects in the pelagic ecosystem.

The series of SST anomalies over the last 14 years in the Gulf of California and adjacent Pacific (Figure 2) showed the highest positive anomalies (>2 SD) from 2014 to 2016. The strong positive anomaly detected in 2014 may correspond with the local manifestation of “The Blob,” which was strongly evident in the northeastern Pacific Ocean [Bond et al., 2015; Whitney, 2015; Peterson et al., 2015; Di Lorenzo and Mantua, 2016]. “The Blob,” whose origin is still under discussion [Di Lorenzo and Mantua, 2016], preceded “Godzilla El Niño 2015–2016,” the subject of study in this paper. The signs of this warm ENSO event developing were detected in the shallow oxygen minimum zone off Mexico from early 2015 to summer 2016 (Figure 2), but the satellite and hydrographic data from the sampling stations of the alongshore transect (Figures 3 and 4) suggest that the warm ENSO event started to relax from March 2016, earlier than in other Pacific regions.

The studied area, on the fringe between tropical and subtropical regions, provides an imperfect but informative glimpse, as through a key hole, of local climate variability and changes of large-scale processes in the Pacific. This is made possible by the area's unique environmental conditions. As a part of the shallow OMZ that is affected by seasonal upwelling, it also represents a confluence of tropical, subtropical, and temperate water masses typified by significant mesoscale activity and mixing of species of different affinities that vary in accordance with local and large-scale processes.

If it considered that the “The Blob” is mainly a large-scale thermal signal, while the El Niño event involve advection processes, it could expected that these events impact transport/advection of larval fish differently. However, it is difficult to identify the signal of each event in organisms with so short a planktonic life (∼10 days), including fish larvae and their zooplankton prey populations. Nevertheless, the almost consecutive presence of these two phenomena of large-scale ocean warming has generated an unprecedented sustained warming in the northern end of the shallow oxygen minimum zone off Mexico, where the marine biological effects have not yet been reported. The later contrasts with the detailed knowledge of the drastic changes recorded in the northeastern Pacific marine ecosystem in response to the same sustained warming event [Cavole et al., 2016; McClatchie et al., 2016].

Decrease of the zooplankton displacement volume during the warm ENSO event is an indicator of the changes in composition and abundance of the zooplankton components, as has been recorded during previous warm ENSO events [Sánchez-Velasco et al., 2000; Lavaniegos et al., 2002; Franco-Gordo et al., 2004]. This decrease happens because most zooplankton components reduce or inhibit their reproduction rates under thermal stress and changing trophic conditions (e.g., food availability), affecting the marine trophic webs [Chávez et al., 1999; Fiedler, 2002; Lavaniegos et al., 2002]. In this context, an unexpected result is that the larval fish abundances were relatively high during the influence of the warm ENSO event, contrasting with the decrease of the zooplankton displacement volume (Figure 5). The relatively high abundance of fish larvae during the ENSO event might be because species with tropical affinity such as V. lucetia, B. bathymaster, and D. laternatus, characterized by their domain in the Eastern Tropical Pacific [Loeb, 1979; Davies et al., 2015; León-Chavez et al., 2015], have a high tolerance of sustained warming. Moreover, an increase in the abundance of V. lucetia and D. laternatus larvae was also reported in the southern part of the California Current System, indicating the same signal to the north of this study area [McClatchie et al., 2016].

Additionally, important changes in their vertical distribution were observed (Figure 6). Before the warm ENSO event, V. lucetia and Auxis spp. larvae were more abundant in the surface mixed layer, in association with high temperature values (Figure 7), while B. bathymaster and D. laternatus larvae dominated in the thermocline and shallow hypoxic layer (<44 µmol/kg), in association with low dissolved oxygen concentrations and temperature values (Figure 7). These findings correspond with previous studies in adjacent areas like the southern Gulf of California [Apango-Figueroa et al., 2015; Contreras-Catala et al., 2016] and the southern California Current System [Koslow et al., 2011; Davies, 2015]. However, during the ENSO event in June 2015, all four species were most abundant in the surface mixed layer, implying that the species adapted to hypoxia, inverted their normal pattern of distribution, possibly because the elevation of the suboxic water (<4.4 µmol/kg) reduced the thickness of the hypoxic layer (Figures 4 and 6). This condition, i.e., development in the warm and oxygenated mixed layer, may provide a better survival opportunity for larvae of these species than development under the thermocline, near to suboxic water, where few species may survival [Gray et al., 2002; Diaz and Rosenberg, 2008; Stramma et al., 2008]. Further observations are required to confirm the generality of this adaptation and inversion of distribution.

As Hofmann et al. [2011] mentioned the critical levels of oxygen uptake in marine organisms are strongly linked to pressure and hydrographic conditions; moreover, the critical level may change in accordance with the phase of the life cycle. Comparing previous adult information (http://eol.org/and http://iobis.org/) and our results, three different patterns were found: (i) Auxis spp., both larvae and adults preferentially inhabit zones with high dissolved oxygen concentrations (∼> 100 µmol/kg); (ii) larvae of the V. lucetia and D. laternatus, have a more restricted oxygen range than the adults, but with inverse preference, i.e., V. lucetia larvae exhibits affinity to the oxygenated layer (∼> 100 µmol/kg), and D. laternatus larvae to the hypoxic layer (∼< 100 µmol/kg); and (iii) B. bathymaster larvae have more tolerance than the adults, but also with preference to hypoxic layer. This summary showed the complexity of the physiological response of each species, generating the need to increase the efforts toward these topics, mainly in OMZs.

Results discussed in this paper showed that “Godzilla El Niño 2015–2016,” preceded by “The Blob,” generated a prolonged warm period in the pelagic ecosystem of the northern end of the shallow oxygen minimum zone off Mexico, with temperature increases of up to 3°C in the surface mixed layer, associated with the shallow presence of hypoxic and anoxic water. Even though the zooplankton displacement volume decreased during the warm event, the total larval abundance did not fall. Nevertheless, a differential response in their vertical distribution was evident, associated with changes in the hydrographic structure of the water column and the depth of the shallow hypoxic and suboxic layers. This might indicate adaptability of the tropical species to prolonged periods of warming in the oceans.

4.1 Final Considerations

The observations presented in this study are a contribution to the understanding of the biological response to the interannual events such as the development of the “Godzilla El Niño 2015–2016” in the pelagic ecosystem in the OMZ of the Eastern Tropical Pacific off Mexico. However, because the ENSO events are products of ocean-atmosphere interconnections throughout the Pacific, a study made in a specific region of the ocean with sparse observations may only generate working hypotheses that must be tested in the surrounding region in order to obtain a wider view of the effects. Ongoing work on indicators of change in marine organisms, including studies from the base of the trophic chain to the top predators, will allow visualization of possible scenarios in future ENSO events. It may be considered that in the coming decades the conditions described in this paper could present in zones like the southern California Current and the central Gulf of California, if the OMZ continues its expansion toward the north and the temperature increases as result of global climatic change.

Acknowledgments

This work was made possible thanks to the financial support of SEP-CONACyT (contracts 2014-236864, L. Sanchez-Velasco and 2011–168034-T, E. Beier) and Fronteras de la Ciencia-CONACyT (contracts 2015-2-280, L. Sanchez-Velasco). The hydrographic and biological data used for this paper are available on request by writing to L. Sanchez-Velasco (lsvelasc@gmail.com) and E. Beier (ebeier@cicese.mx). The satellite data used for this paper are also available on request by writing to E. Santamaría-del-Angel (santamaria@uabc.edu.mx). We thank the scientific and technical staff who took part in the cruises aboard the R/V “Francisco de Ulloa” and “Alpha Helix.” Satellite data were processed in the Ocean color laboratory in the Facultad de Ciencias Marinas (Universidad Autonoma the Baja California). Special thanks are due to the two anonymous referees for helping improve this article.

Ancillary