Interannual Variability in Seasonal Sea Surface Temperature and Chlorophyll a in Priority Marine Regions of the Northwest of Mexico

Abstract

The northwest of Mexico has important zones for biodiversity conservation, denominated Priority Marine Regions (PMRs), and to study key oceanographic features related to ecological structure, it is necessary to understand environmental variability and observe climatic trends. Sea Surface Temperature (SST) is tightly associated with photosynthesis and serves as a control and driver for biological processes linked to the phytoplankton. Global climatic systems, like El Niño Southern Oscillation (ENSO), are responsible for the interannual and interdecadal variation in SST, since global circulation is modified by them. An important metric to assess phytoplanktonic biomass/photosynthesis is Chlorophyll a (Chl a), constituting the primary basis of the marine trophic web. The present study aims to examine the interannual oceanographic variability across 24 PMRs by employing monthly SST (°C) and Chl a (mg/m³) data derived from remote sensing instruments with spatial resolution of 4 km and 1 km from September 1997 to October 2018. We grouped the Priority Marine Regions into 18 main areas, based on a cluster analysis of Sea Surface Temperature. Significant differences were observed, showing higher SST levels during El Niño phase and higher Chl a concentration during La Niña phase, primarily in winter and spring, which will impact marine ecosystems.

1. Introduction

The South Californian Pacific and Gulf of California are some of the richest and most diverse marine habitats in the world, with unique oceanographic conditions due to circulation patterns of current systems and the effect of seasonal winds that influence the upwelling events determining high primary productivity levels [1,2,3,4]. However, as a consequence of current global shifts in climate conditions, key drivers for oceanic processes (such as temperature) are being modified [5,6,7], and potential effects of those changes on the region are considerably large [8,9,10]. Therefore, the monitoring and conservation of these regions is a priority to preserve and manage the natural and social capital they represent. Addressing the previous, and to protect and monitor valuable marine and coastal environments, the Mexican “National Biodiversity Commission” (CONABIO by its acronym in Spanish) characterized a set of “Priority Marine Regions” (PMRs), classified as important areas due to their high abundance of natural resources and biodiversity. To establish these areas, CONABIO analyzed environmental (abiotic and biotic factors) and economic (fishing areas, tourism) characteristics and factors, as well as the main ecosystem threats posed to each region (pollution, modification of the environment) [11].

Different variables have been used for analysis of the oceanographic variability describing physicochemical and biological variables, useful to describe changes in the coastal and marine ecosystems. One of them is the Sea Surface Temperature (SST), classified as an essential variable to describe climatic and oceanographic conditions that dictate physicochemical characteristics in the oceans [12] and an indicator of climatic change worldwide [13,14]. Another important variable is Chlorophyll a (Chl a) concentration, which determines phytoplankton biomass production in the water column [15,16]. Considered as the main component of the marine trophic webs, phytoplankton regulate the rate of carbon dioxide conversion into organic carbon through the photosynthesis process [17]. Not only are SST and Chl a analyses important to describe climatic changes associated with anthropogenic or natural factors [18,19,20,21], but they also constitute good proxies to understand modifications and trends in ecosystem function [22,23,24,25].

Global and regional climatic patterns are closely linked to driving large-scale atmospheric-oceanic phenomena, such as El Niño–Southern Oscillation (ENSO) [26,27,28]. Caused by temperature oscillation on the Southern Pacific Ocean, ENSO has been documented to have a diving effect on SST and Chl a along the Tropical Pacific Ocean, causing an irregular climatic and oceanographic variability characterized by its two different phases: El Niño (warm phase) and La Niña (cold phase) [29]. El Niño phase is described by an anomalous warming of surface waters along the Eastern Equatorial Pacific Ocean, alteration of the atmospheric pressure at tropical sea level between the Eastern and Western Pacific, causing advection of warm waters, and a depression of the water column thermocline [29,30,31]. Conversely, La Niña phase represents the opposite of El Niño, promotes the accumulation of warm water in the Western Equatorial Pacific, and has a cooling effect in the Eastern Tropical Pacific [29,32]. Previous work has discussed the ENSO effects on the SST levels and Chl a concentration, and the modifications to physical properties and ecosystem processes along the South Californian Pacific [33,34,35,36] and Gulf of California [37,38,39,40].

From this perspective, the use of satellite images obtained through remote sensing allows us to observe and quantify key oceanographic variables at different spatial and temporal scales [41]. Several SST and Chl a datasets have been developed to provide quality information regarding physical and biological characteristics of oceans [42,43]. These datasets have been used to study long-term biophysical trends of change, associated with climate shifts in marine and coastal environments [44] and Chl a [45,46] through different studies focused on ENSO productivity and climatic variability using satellite images in various regions of the world [47,48,49,50]. Therefore, the use of SST and Chl a analyses, as tools for regionalization of marine and coastal processes, could be useful to describe and understand the effect of climatological phenomena on primary productivity [51,52]. Moreover, regionalization supported by biophysical analysis would be useful to assess/model potential changes on marine and coastal ecosystems, under different climate change scenarios [52,53].

Previous studies have analyzed SST and Chl a variability and correlation, describing how climate variability and trends over large oceanographic regions of the Pacific Ocean pose a press-like driver for change over current conditions [54,55,56]. There have been numerous reports on the effect of SST changes on phytoplanktonic communities, and corresponding shifts in Chl a concentration [57,58,59]. However, there has been a general lack of approaches, trying to understand the seasonal variability posed by climatic systems, like El Niño Southern Oscillation (ENSO) on SST and Chl a concentration, even though these systems do not have a natural annual cycle [60,61,62]. The previous is extremely relevant, since climatic predictions under current climatic change conditions might not be even throughout the year, and the effect on seasonal oceanic productivity might reflect short-term seasonal changes, better than annual periods.

The effect of rapid climatic changes on conservation has to be addressed, since ecosystems might experience a shift in their environmental baseline. Therefore, to study areas intended for conservation (such as the PMR), it is of utmost importance to understand the effect their changes might pose on environmental drivers such as SST, on key response variables like Chl a, since these are a proxy for primary productivity, which will have a profound effect on ecosystem characteristics [38,63,64]. Specifically, ENSO events have been reported for the South Californian Pacific [34,35,36,59,65,66,67] and the Gulf of California [39,40,68,69,70,71,72], showing a profound effect on climatic variability, and, therefore, on SST levels and Chl a concentration; however, its seasonal effect is not addressed as extensively.

The present study analyses the interannual effects of ENSO phases on SST and Chl a variability and proposes a biophysical approach to the regionalization of the PMR in northwestern Mexico. Specifically, we use a remote sensing dataset time series (from 1997 to 2018) to (1) divide and regionalize PMR based on common biophysical characteristics using a cluster analysis, and (2) decompose the response of SST and Chl a by season, considering ENSO phases. This study seeks to understand the effects of a shift/trend in climatic conditions might have on marine environments.

2. Materials and Methods

2.1. Study Area

The study area comprises 24 PMRs (

Table 1. Geographic coordinates of the Priority Marine Regions of the northwest of Mexico, located in the South Californian Pacific and the Gulf of California.

Number	Marine Area	Priority Marine Region	Upper Left Latitude/Lower Right Longitude
1	South Californian Pacific	Ensenadan	32°31′48″ N/115°42′ W
2	South Californian Pacific	Vizcaíno	28°57′36″ N/113°43′48″ W
3	South Californian Pacific	San Ignacio	27°18′36″ N/112°46′48″ W
4	South Californian Pacific	Magdalena Bay	25°47′24″ N/111°21′36″ W
5	South Californian Pacific	Barra de Malva-Cabo Falso	25°47′24″ N/111°21′36″ W
6	South Californian Pacific	Guadalupe Island	29°22′12″ N/118°2′24″ W
7	South Californian Pacific	Alijos Rock	25°08′24″ N/115°32′24″ W
8	South Californian Pacific	Revillagigedo Islands	21°05′24″ N/109°30′00″ W
9	Gulf of California	Los Cabos	23°39′ N/109°21′36″ W
10	Gulf of California	Baja California Sur Island Complex	26°31′48″ N/109°47′24″ W
11	Gulf of California	Concepcion Bay	27°07′12″ N/111°33′ W
12	Gulf of California	Eastern Vizcaíno Coast	27°59′24″ N/112°18′36″ W
13	Gulf of California	Baja California Island Complex	29°57′36″ N/112°12′36″ W
14	Gulf of California	Upper Gulf	32°10′12″ N/114°11′24″ W
15	Gulf of California	Infiernillo Channel	29°22′12″ N/111°43′48″ W
16	Gulf of California	Cajón del Diablo	28°16′48″ N/111°09′36″ W
17	Gulf of California	Southern Sonora Lagoon System	27°34′12″ N/109°21′36″ W
18	Gulf of California	Santa María La Reforma Lagoons	25°26′24″ N/107°49′48″ W
19	Gulf of California	Chiricahueto Lagoon	24°29′24″ N/107°25′48″ W
20	Gulf of California	Piaxtla-Urías	23°48′ N/106°13′48″ W
21	Gulf of California	Marismas Nacionales	22°41′24″ N/105°9′36″ W
22	Gulf of California	Banderas Bay	21°27′36″ N/105°11′24″ W
23	Gulf of California	Entrance of the Gulf	22°51′ N/107°14′24″ W
24	Gulf of California	Guaymas	27°49′12″ N/110°54′36″ W

) distributed along the South Californian Pacific (Figure 1) and Gulf of California (Figure 2). The South Californian Pacific is mainly influenced by the California Current Systems, dominated by wind-driven upwelling, which provides nutrient-rich water to the euphotic section of the water column, supporting productivity levels in marine and coastal ecosystems [73]. This region is considered a transition between cold and fresh subarctic oceans, with a higher concentration of solutes, and warmer tropical and subtropical water [74]. The previous generates a seasonal variability, where subarctic characteristics dominate during winter and spring, while tropical and subtropical characteristics are present in summer and fall [3]. On the other hand, the Gulf of California is a dynamic sea rich in nutrients and high productivity levels [2]. These characteristics are associated with tidal mixing upwelling systems, seasonal winds, and high solar radiation [1,75]. The PMR in the region, where established/suggested by CONABIO, due to a combination of factors such as their high biodiversity, prominent social importance, or unique oceanographic conditions [11].

2.2. Sea Surface Temperature (SST) and Chlorophyll a (Chl a) Datasets

The present work uses a series of remotely sensing datasets, coupled with climatological indices and models, to explain ENSO effects on SST and Chl a on a seasonal basis (Figure 3).

To characterize oceanographic and environmental parameters of the 24 PMR, proxies from remote sensing products for SST and Chl a were obtained. Specifically, a period between September 1997 and October 2018 was selected to analyze two decades of data, and assess relationships and drivers to climatic phenomena. Specifically, for both variables, we obtained one monthly reading (temporal resolution), at 4 km (September 1997–January 2000 AVHRR datasets) and 1 km (February 2000–September 2018 WimSoft products) spatial resolution for SST, and at 1 km spatial resolution for Chl a (September 1997–September 2018 WimSoft).

SST datasets were downloaded from the Physical Oceanography Distributed Active Archive Center (PODAAC) from https://podaac.jpl.nasa.gov/ (accessed on 20 January 2025). Specifically, for the first period (September 1997–January 2000), we used monthly data from the sensor Advanced Very High Resolution Radiometer (AVHRR-Pathfinder version 5.0). The scale of the dataset is given in Celsius (°C), and it has a spatial resolution of 4 km. Monthly SST Level-3 products were mapped in uniformly distributed spatiotemporal grids. Each one of the pixels contains values, encompassing those potentially affected by clouds or other error sources. To exclude these errors, Overall Quality Flag (OQF) values were applied, obtaining an average of the highest-quality AVHRR Global Area Coverage (GAC) observations within approximately 4 km bins. Pathfinder algorithm uses the Reynolds Optimally Interpolated SST (OISST, Version 2) for quality control. Quality assessment is based on an Overall Quality Flag ranging from 0 (lowest) to 7 (highest), determined through a hierarchical set of tests. For studies requiring the highest precision, only level 7 data should be used [76,77,78]. To develop a continuous dataset (from September 1997 to October 2018), we complemented the previous period with monthly data, obtained from averaged composite images developed by the Scripps Institution of Oceanography at San Diego, California from http://www.wimsoft.com/Satellite_Projects.htm (accessed on 20 January 2025). from February 2000 to September 2018. After standardization of the AVHRR period, we used these datasets to complete 253 months of average SST database at each of our study sites.

Monthly Chl a datasets were also acquired from averaged composite images developed by the Scripps Institution of Oceanography at San Diego, California from http://www.wimsoft.com/Satellite_Projects.htm (accessed on 20 January 2025), from September 1997 to October 2018, which correspond to the same period of the SST dataset time series previously described (253-month period). The scale of the dataset is given in milligrams per cubic meter (mg/m³). Each composite image in the datasets was processed at Level 2 [79,80], data combining with unmapped datasets obtained by using daily images of SST and Chl a from unmapped data, and diverse sensors such as Sea-viewing Wide Field-of-view Sensor (SeaWiFS); Moderate Resolution Imaging Spectro-radiometer (MODIS) Terra and Aqua; Medium Resolution Imaging Spectrometer (MERIS); Visible Infrared Imaging Radiometer Suite (VIIRS); Visible Infrared Imaging Radiometer Suite-Joint Polar Satellite System (VIIRS-JPSS1); Ocean and Land Color Instrument-Water Reduced Resolution (OLCI-WRR) A and B, which were downloaded from the National Aeronautics and Space Administration (NASA) Ocean Color Web [80]. Individual satellite swaths were georeferenced to ensure a better spatial comparability among sensors, and only cloud-free pixels were merged to generate these monthly composite images by combining multiple satellite overpasses, and sensor data gaps caused by cloud cover were minimized, enhancing spatial and temporal coverage [80].

These monthly SST and Chl a datasets were accessed, analyzed, and extracted using Windows Image Manager Automation Module software (WIM/WAM), Wimsoft version 9.06 1991–2015, Copyright© Mati Karhu [81]. We used digital maps from CONABIO-PMR to locate the areas of interest 24 at the South Californian Pacific and Gulf of California [11]. We used the geographic limits of these areas to obtain monthly mean SST and Chl a, from the time period previously stated and conducted post-analysis of these variables, taking into account climatic drivers and seasonality.

2.3. Processing Calibration and Regionalization Through Sea Surface Temperature (SST) Datasets

To ensure consistency among SST datasets, a calibration to adjust AVHRR-Pathfinder values was necessary. Therefore, a period of seven years was used (February 2000 to February 2007) in which the data from AVHRR-Pathfinder version 5.0 coincide with composite images of multiple sensors from the WimSoft Web (February 2000–October 2018). As an analysis, we used a linear regression to observe how values of similar months correlate. Previous studies have applied a comparison of different sensors and their combinations [82,83], obtaining a similar distribution pattern and high coefficient correlation, which supports the application of long-term analysis through multiple sensors [82].

Once the calibration of the SST data was finished, a cluster analysis of the monthly SST data was performed to group the monthly average values of each PMR from both marine areas, the South Californian Pacific and the Gulf of California. The cluster analysis is a multivariate technique that combines objects into groups, based on similarity, applying fusion algorithms [84,85]. The present work uses cluster analysis on regional PMRs and establishes a maximum level of homogenization, considering the maximum differences between the groups, obtaining linkage distances to conform groups with similar SST levels (°C), to address Chl a concentration. Previous studies usually address climatic variation yearly, instead of seasonally, often miss the temporal variation posed by driving phenomena such as ENSO.

2.4. Processing and Statistical Analysis for Sea Surface Temperature (SST) and Chlorophyll a (Chl a)

As a key component of the present study, it was necessary to classify ENSO phases for particular seasons each year. For the previous, the Oceanic Niño Index (ONI) developed by the National Oceanic and Atmospheric Administration (NOAA) was used, which classifies SST anomalies of +0.5 higher as El Niño phases, and values of −0.5 lower as La Niña phases [86].

For each season, year, and PMR in the analysis, average values for SST and the Chl a value addition (of each season) were derived and related to a specific ENSO phase (present at the time). We used Permutation Tests (p-value < 0.05) to analyze the interannual differences among ENSO phases for each PMR studied. The permutation analysis determines statistical evidence against the null hypothesis by not finding differences [87,88], and all the analyses were completed in the Statistical Software version 7.5 and R Software version 3.4. In this last software, the libraries used to perform the Permutation Tests (p-value < 0.05) were wperm, lmperm, and FSA.

3. Results

3.1. Datasets and Regionalization

We obtained a total of 254 monthly measurements of SST and of Chl a for each of the PMR analyzed (Figure 1 and Figure 2). During this period, multiple ENSO phases were present, and, therefore, a representative number of each phase type (El Niño, La Niña, and neutral years) was ensured to assess variable behavior during those phases. Specifically, SST measurements were used to assess similarities among PMR and, therefore, achieve a consistent biophysical approach for regionalization.

A cluster analysis for SST suggests the presence of four complexes composed of two to four PMR, and 14 regions classified into individual areas. These analyses consider linkage distances of 13 units to establish conglomerates, based on the dissimilarity or similarity of the SST values [52,71]. A total of 18 regions (

Table 2. Regions obtained from the cluster analysis of monthly Sea Surface Temperature (SST) data of Priority Marine Regions of the northwest of Mexico.

Number	Marine Areas	Regions
1	Gulf of California	Group 1 (Concepción Bay, Cajón del Diablo, Southern Sonora Lagoon System and Guaymas)
2	Gulf of California	Group 2 (Eastern Vizcaíno Coast and Baja California Island Complex)
3	Gulf of California	Group 3 (Santa María La Reforma Lagoons and Chiricahueto Lagoon)
4	Gulf of California	Group 4 (Marismas Nacionales and Banderas Bay)
5	South Californian Pacific	Ensenadan
6	South Californian Pacific	Vizcaíno
7	South Californian Pacific	San Ignacio
8	South Californian Pacific	Magdalena Bay
9	South Californian Pacific	Barra de Malva-Cabo Falso
10	South Californian Pacific	Guadalupe Island
11	South Californian Pacific	Alijos Rock
12	South Californian Pacific	Revillagigedo Islands
13	Gulf of California	Los Cabos
14	Gulf of California	Baja California Sur Island Complex
15	Gulf of California	Upper Gulf
16	Gulf of California	Infiernillo Channel
17	Gulf of California	Piaxtla-Urías
18	Gulf of California	Entrance of the Gulf

, Figure 4) are suggested, of which four were gathered into groups (Group 1 to Group 4) recognized as a conjunction of areas, presenting similar SST behavior within each group. The rest of the regions were listed as individual sites by our analysis and were treated as such in our subsequent exercises. It is necessary to mention that the linkage distance was established to determine the whole conglomerates from the cluster analysis composed of PMR exclusively close to each other and located in the same marine area, which means that a conglomerate that represents a region is composed only of PMR of the South Californian Pacific or from the Gulf of California, but never a combination of both marine areas.

Results suggest that SST segregates PMRs as a consequence of the latitudinal gradient/location and SST patterns. This was expected, since this gradient poses changes in climatology controls, and, therefore, physical dynamics [4,33,89,90]. It is also worth noting that PMR, grouped within a cluster, tend to be proximate to each other, sharing location characteristics (coastal distance and latitudinal gradient). These results accentuate the influence of overarching controls on SST, which in turn affects biological activity in the ocean [91,92,93].

3.2. Interannual Analysis of ENSO: Sea Surface Temperature (SST)

To analyze the seasonal SST for each cluster, considering ENSO phases, a Pairwise Permutation Test was performed. For the previous, the means of the monthly SST data was used for the whole period of study, divided between Niño, Niña, and neutral phases of ENSO. The analysis of these phases was conducted for each season of the year (spring, summer, autumn, and winter) to compare behavior and the influence of ENSO on temperature. Pairwise Permutation Tests comparisons and results are depicted in the following tables (

Table 3. Pairwise Permutation Test for the interannual analysis of spring and summer monthly mean Sea Surface Temperature (SST) values during interannual events (El Niño and La Niña) and neutral conditions of the Regions of the northwest of Mexico. Similar groups represented with the same letter indicate no evidence of statistical differences within the same group, and vice versa, and sharing different letters shows evidence of significant differences.

Regions	Spring			Summer
	El Niño	La Niña	Neutral Conditions	El Niño	La Niña	Neutral Conditions
1	23.04 (a)	21.06 (b)	21.86 (c)	29.90	29.43	29.72
2	21.16 (a)	19.19 (b)	19.76 (c)	28.41	28.20	28.32
3	24.58 (a)	22.76 (b)	23.62 (a)	30.47	29.77	30.10
4	26.31 (a, b)	25.11 (a)	25.89 (b)	30.66 (a)	29.61 (b)	30.07 (a, b)
5	17.18 (a)	15.58 (b)	16.08 (b)	19.38 (a, b)	18.41 (a)	19.61 (b)
6	18.17 (a)	16 (b)	16.58 (c)	19.38	18.97	19.98
7	19.22 (a)	16.45 (b)	17.01 (b)	21.08	21.09	21.23
8	20.73 (a)	18.12 (b)	18.80 (c)	23.16	22.79	23.14
9	22.07 (a)	19.74 (b)	20.36 (b)	25.11	24.52	24.71
10	17.82	16.42	17.05	19.97 (a, b)	18.74 (a)	20.02 (b)
11	19.64 (a)	18.09 (b)	18.64 (b)	21.45	20.69	21.64
12	25.02 (a)	23.73 (b)	24.33 (a, b)	27.05 (a)	26.13 (b)	26.55 (a, b)
13	24.71 (a)	23.35 (b)	23.65 (b)	28.14	27.63	28.13
14	23.51 (a)	21.79 (b)	22.54 (c)	28.97	28.42	28.58
15	22.14 (a)	20.88 (b)	21.17 (b)	29.74 (a)	29.48 (a)	29.95 (a, b)
16	22.73 (a)	20.93 (a, b)	21.09 (b)	30.41	30.45	30.17
17	25.32 (a)	23.60 (b)	24.45 (a)	30.45 (a)	29.71 (b)	29.97 (a, b)
18	22.53 (a)	20.79 (b)	21.61 (c)	29.83	29.25	29.65

and

Table 4. Pairwise Permutation Test for the interannual analysis of the autumn and winter monthly mean Sea Surface Temperature (SST) values during interannual events (El Niño and La Niña) and neutral conditions of the Regions of the northwest of Mexico. Similar groups represented with the same letter indicate no evidence of statistical differences within the same group, and vice versa, sharing different letters shows evidence of significant differences.

Regions	Autumn			Winter
	El Niño	La Niña	Neutral Conditions	El Niño	La Niña	Neutral Conditions
1	28.52 (a)	27.81 (b)	27.94 (a, b)	20.46 (a)	18.28 (b)	19.01 (b)
2	27.34 (a)	26.54 (b)	26.90 (a, b)	19.26 (a)	17.27 (b)	17.83 (a, b)
3	29.65	29.12	29.06	23.05 (a)	20.57 (b)	22.02 (a)
4	30.18 (a)	29.37 (b)	29.81 (a, b)	25.99 (a)	24.41 (b)	25.57 (a)
5	20.73 (a)	19.10 (b)	19.57 (a, b)	17.03 (a)	15.69 (b)	16.35 (a, b)
6	22.62 (a)	20.06 (b)	21.41 (a)	18.92 (a)	17.06 (b)	17.61 (a, b)
7	25.21 (a)	22.47 (b)	23.98 (a)	20.61 (a)	18.65 (b)	19.18 (b)
8	26.77 (a)	24.77 (b)	25.77 (a, b)	22.24 (a)	20.59 (b)	21.11 (a, b)
9	27.95 (a)	26.74 (b)	27.06 (a, b)	23.57 (a)	22.14 (b)	22.54 (a, b)
10	21.30 (a)	19.89 (b)	20.37 (a, b)	17.82 (a)	16.67 (b)	17.20 (a, b)
11	23.67 (a)	21.60 (b)	22.66 (a)	20.39 (a)	18.68 (b)	19.36 (a, b)
12	27.45	26.90	26.86	25.16 (a)	24.27 (b)	24.80 (a, b)
13	29.39	28.94	29.12	23.97 (a)	22.68 (b)	23.17 (a, b)
14	28.97	28.43	28.46	22.15 (a)	20.20 (b)	20.90 (b)
15	27.54	26.96	27.41	17.60 (a)	16.68 (b)	16.96 (a, b)
16	27.76	27.25	27.20	18.52 (a)	16.81 (b)	17.17 (b)
17	29.96 (a)	29.16 (b)	29.39 (a, b)	24.71 (a)	22.26 (b)	23.79 (a)
18	28.39 (a)	27.50 (b)	27.75 (b)	20.37 (a)	18.06 (b)	18.83 (a)

), where the presence of the letters indicates that at least one of the groups (El Niño, La Niña, and neutral conditions) shows statistical differences (p-value < 0.05) on SST levels. It is important to indicate that this process was also performed with the interannual analysis of ENSO with the Chl a concentration. Additionally, to illustrate general trends, simple scatterplots showing the SST behavior for each phase per season are shown (Figure 5).

According to results, 17 of the 18 regions show higher temperatures during El Niño phase for spring, when compared to neutral and La Niña phases. As for summer, the trend is less clear, since only six regions present differences, of which three with higher temperatures occurred during El Niño, and the other three during neutral phases. Autumn presents a clear distinction of temperature readings, according to the ENSO phase in place, with 12 of the 18 regions presenting higher mean temperatures during El Niño events, and neither La Niña nor the neutral phases presenting higher mean temperatures for any of the regions. However, even more apparent than autumn was the winter season, where the whole 18 regions presented higher mean temperature during El Niño phases, suggesting ENSO strongly influences SST during this season. Monthly images of SST of March, August, November, and January 2011, 2013, and 2015 were taken to show the SST variability for each season of the year during the ENSO phases and neutral conditions, according to the results obtained from the Pairwise Permutation Test (Figure 6, Figure 7, Figure 8 and Figure 9).

3.3. Interannual Analysis of ENSO: Chlorophyll a (Chl a)

Seasonal analysis of Chl a concentration was also conducted via Pairwise Permutation Test, to assess differences among PMR clusters. For this analysis, the sum of the monthly Chl a data was used to assess differences in seasonality according to the ENSO phase (Niño, Niña, and neutral). Statistical differences (p-value < 0.05) among phases were observed in several regions, and during different seasons, suggesting significant ENSO influences on oceanic Chl a concentration (

Table 5. Pairwise Permutation Test for the interannual analysis of the spring and summer monthly mean Chlorophyll a (Chl a) values during interannual events (El Niño and La Niña) and neutral conditions of the Regions of the northwest of Mexico. Similar groups represented with the same letter indicate no evidence of statistical differences within the same group, and vice versa, and sharing different letters shows evidence of significant differences.

Regions	Spring			Summer
	El Niño	La Niña	Neutral Conditions	El Niño	La Niña	Neutral Conditions
1	3.60 (a)	7.82 (b)	6.50 (b)	1.66	2.46	1.90
2	7.89	8.21	7.61	3.58	3.94	3.78
3	9.17 (a)	19.26 (b)	12.18 (a)	7.83	6.60	6.91
4	2.48 (a, b)	14.62 (a)	4.53 (b)	2.97 (a, b)	3.30 (a)	2.04 (b)
5	1.90 (a)	3.79 (b)	3.26 (b)	1.66	1.90	1.92
6	1.94 (a)	4.66 (b)	3.47 (c)	3.31	2.75	2.80
7	4.12 (a)	10 (b)	7.39 (b)	10.53	8.81	8.76
8	2.31 (a)	5.67 (b)	4.11 (c)	4.39	4.21	4.43
9	1.34 (a)	4.35 (b)	2.90 (c)	2.74	2.69	3.30
10	0.35	0.39	0.37	0.43 (a)	0.35 (a, b)	0.33 (b)
11	0.51	0.32	0.37	0.29	0.32	0.31
12	0.32	0.34	0.31	0.30	0.30	0.28
13	0.83 (a)	2.35 (b)	1.65 (c)	1.82	1.90	1.87
14	1.67 (a)	3.74 (b)	3.35 (b)	1.87	2.03	1.97
15	7.77	8.29	9.62	7.28 (a)	7.71 (a, b)	6.58 (a)
16	13.53 (a)	19.20 (b)	16.90 (a, b)	8.04	10.60	10.27
17	3.06 (a)	13.69 (b)	6.20 (a)	2.28	3.41	2.15
18	0.52 (a)	1.07 (b)	0.72 (a)	0.73	0.61	0.62

and

Table 6. Pairwise Permutation Test for the interannual analysis of the autumn and winter monthly mean Chlorophyll a (Chl a) values during interannual events (El Niño and La Niña) and Neutral Conditions of the Regions of the northwest of Mexico. Similar groups represented with the same letter indicate no evidence of statistical differences within the same group, and vice versa, and sharing different letters shows evidence of significant differences.

Regions	Autumn			Winter
	El Niño	La Niña	Neutral Conditions	El Niño	La Niña	Neutral Conditions
1	3.02 (a)	4.11 (b)	3.98 (a, b)	4.84 (a)	7.14 (b)	5.95 (a, b)
2	4.67	5.50	5.10	4.65	5.05	4.79
3	10.86	11.74	10.98	9.23 (a)	15.08 (b)	13.38 (b)
4	2.12 (a)	4.12 (a, b)	3.40 (b)	2.31 (a)	6.13 (b)	3.31 (a, b)
5	0.99 (a)	1.44 (b)	1.32 (b)	1.28	1.86	1.45
6	1.34	1.53	1.38	1.35 (a)	2.06 (b)	1.63 (a, b)
7	1.86	2.27	1.94	1.74 (a)	2.68 (b)	2.14 (a, b)
8	1.14	1.39	1.53	1.13 (a)	1.79 (b)	1.52 (a, b)
9	0.52 (a)	0.65 (b)	0.63 (a, b)	0.79 (a)	1.50 (b)	1.08 (a, b)
10	0.45	0.38	0.37	0.53	0.53	0.46
11	0.42	0.36	0.35	0.47	0.43	0.49
12	0.38	0.32	0.34	0.34	0.40	0.36
13	0.58 (a)	0.77 (a, b)	0.79 (b)	1.14 (a)	2.58 (b)	2.19 (b)
14	1.11	1.35	1.18	2.61 (a)	3.51 (b)	3.28 (a, b)
15	7.28	8.04	7.68	7.12	8.46	7.63
16	11.67	13.46	14.18	12.81	17.43	16.70
17	3.35	4.33	3.98	3.92 (a)	11.14 (b)	5.24 (a)
18	0.40	0.46	0.46	0.62 (a)	0.94 (b)	0.73 (a, b)

). However, when using Pairwise Permutation as a post hoc analysis, winter and spring show more differences among phases than summer and autumn (Figure 10).

Our analysis for spring shows a higher concentration of Chl a in 13 of the regions during La Niña events, as opposed to none for El Niño or neutral conditions. For the summer, our results suggest that just 2 of the 18 regions show higher concentrations of Chl a during La Niña phases, and 1 during El Niño. On the other hand, when analyzing results for the autumn season, we found that 5 of the 18 regions showed higher concentrations of Chl a during La Niña events, as opposed to none for El Niño or neutral phases. Similar to spring, winter presents a high number of differences among the regions. Specifically, we found that 11 of the 18 regions present higher concentrations of Chl a during La Niña phases. The previous suggests a deep effect on productivity, as a consequence of ENSO driving forces, and a potential shift in oceanic ecological configurations/structure, if future conditions are dominated by one of those phases/regimes. Chl a interannual variability was also described using monthly images with the same months and years to represent a season of the year during neutral conditions and ENSO phases, based on the results obtained from the Pairwise Permutation Test (Figure 11, Figure 12, Figure 13 and Figure 14).

4. Discussion

4.1. Priority Marine Regions Aggregation

According to clustered SST analysis, this study identifies a total of 8 PMR in the South Californian Pacific, while the other 10 regions are grouped in the Gulf of California (Table 2, Figure 3). This agrees with previous studies, which have attempted to classify these regions using SST and/or Chl a, in an array of regional patterns addressing variability, trends, and dynamics [94,95,96]. Results from these works suggest that climate phenomena and characteristics, in association with dynamic oceanic processes such as wind/upwelling systems [4], constitute key drivers for Chl a [97,98], and SST [99,100] dynamics along the Baja California Peninsula. It has been reported that, for the South Californian Pacific Ocean, SST and Chl a variability is tightly linked to ocean-atmosphere interactions, as well as to the physical dynamics in the water column. Therefore, seasonal variability [3,4], and coastal upwelling events [15,101,102,103], contribute to variation in SST levels and Chl a concentration, as a consequence of fluctuation in phytoplankton biomass [33,98]. On the other hand, the Gulf of California is considered a transitional zone, with a climatology much more continental than oceanic [104], which generate a temperate-tropical climate, derived from the influence of the Intertropical Convergence Zone (ITCZ) movements [38,90] from the Tropical Pacific Ocean. And from these premises, the present study aggregates PMR using SST as a proxy to study variation and climatic trends. It is important to mention that the regionalization process obtained constitutes an improvement on spatial (1 km resolution from 2000 to 2018) and temporal resolutions used in previous works (monthly SST data), using a relatively extended time series (September 1997–October 2018).

4.2. Interannual Analysis of ENSO: Sea Surface Temperature (SST) and Chlorophyll a (Chl a) Variability in the South Californian Pacific and Gulf of California

The present work highlights the importance of interannual variability of ENSO phases, finding significant phase-dependent differences in SST levels (°C) and Chl a concentration (mg/m³), for all seasons (spring, summer, fall, and winter) for all the regions of the Southern Californian Pacific and Gulf of California, and primarily during winter and spring. For the South Californian Pacific, ENSO events were described through SST data, reporting higher values than the annual average in Magdalena Bay [105] and in Todos Los Santos Bay [34], effects that can be associated with weakening of the wind’s patterns and upwelling events in conjunction with the advection of tropical and subtropical water [3,106]. Similar effects were also described in the Gulf of California [83,89,107], reporting SST anomalies in the south and central regions of the Gulf in comparison with the Midriff Islands and the northern regions, these patterns are generally attributed to the link between the Gulf and the Pacific Ocean. Other studies suggests smaller effects of ENSO events in the SST, along the central and Midriff Islands of the Gulf of California, where major drivers of temperature variability are associated with other oceanic physical dynamics, such as tidal mixing effects and upwelling events [108,109,110] as well as to the modification of the seasonal wind patterns along the eastern coast of the gulf [111].

El Niño phase effects in Chl a variability have also been discussed for the South Californian Pacific in previous works [59,66], suggesting a decrease in phytoplankton biomass and important changes in phytoplankton composition. Furthermore, El Niño events are associated with the weakening of coastal upwelling and mixing processes, causing a lack of nutrients (nutricline sinking) and stratification of the water column (temperature thresholds). Other studies describe the presence of eventual anomalous subarctic water along the west coast of Baja California, which also influences productivity and concentration of Chl a [55,112]. Similar studies have been conducted across several sites of the Gulf of California [68,107,113], acknowledging a direct influence to the communication with the Pacific Ocean, which allows the entrance of equatorial surface water [114] from the Mexican Coastal Current [38], resulting in a more oligotrophic condition in comparison to the gulf, affecting the Chl a concentration [52,71,113,115]. Previous work suggests the effects of ENSO events (El Niño and La Niña) are evident at the entrance of the Gulf [107,116], associating interannual chlorophyll variability with this phenomenon on the Nayarit continental shelf. However, other studies suggest lesser ENSO effects at the central and Midriff Islands regions of the gulf, where oceanographic conditions (tidal mixing, constant upwellings) inhibit the influence of El Niño and La Niña events [110,117]. Similar results were reported in the northern Gulf of California [118,119], showing that phytoplankton biomass has a low response to interannual events due to the direct effect of northwestern wind patterns. Escalante et al. [107] analyzed SST anomalies and reported a delayed response of 3–6 months for the reestablishment of normal phytoplankton conditions once the ENSO warm phase ended. Hakspiel-Segura et al. [70] described Chl a variability as an indicator of resilience, reported a regulating process of Chl a value during warm and cold periods, indicating a mechanism of ecosystem resilience to climate variability at different spatiotemporal scales.

The present study suggests a strong ENSO influenced by driving SST and Chl a variability, which can only be observed when the phenomena are decomposed seasonally, rather than annually. In addition, the breakup of ENSO events in seasonal patterns uncovers environmental variability and oceanographic dynamics not fully addressed in previous studies. However, previously reported results support much of the analyses of SST levels and Chl a concentration for the South Californian Pacific and Gulf of California presented in this study, even though these studies were mainly conducted by characterizing entire years as a single phase of ENSO (El Niño, La Niña, or neutral events). However, a more specific and concise characterization of the effects of interannual ENSO events, for each season, consistently highlights higher SST levels during El Niño events, and higher Chl a concentration during La Niña events. These findings suggest that scenarios where SST rises due to climatic shift would result in significantly less productive oceans for the region studied.

Analysis of SST for the study areas in the present work shows significant differences between ENSO phases, in 17 of the 18 regions analyzed during spring, 6 regions differ during summer, 12 during autumn, and finally, all regions showed significant differences during winter, consistently finding higher SST levels during El Niño phases. Also, significant differences were found for Chl a concentration between ENSO phases, where concentrations rise during La Niña events for 13 of the 18 regions in spring, for 3 during summer, for 5 in the autumn, and finally for 11 of the 18 regions during winter. Previous studies have reported the effect on oceanographic dynamics and environmental variability, using SST and Chl a as a response variable, to describe possible scenarios, and trends of changes which might influence the structure and function of marine and coastal ecosystems [64,69,70,120,121]. López-Martínez et al. [71] analyze the increase in SST levels in the Gulf of California, and its linkage to the decreases in Chl a concentration and productivity, discussing the potential effect of warming trends on the decrease in productivity, and the subsequent potential effects on the habitat loss for many species, which could cause a decline in fisheries. Similar results were described by García-Morales et al. [122] for the Pacific region of Baja California, describing a tropicalization process in the southern area, indicating an increase in SST levels due to long-term events or a potential sign of global warming, which can alter the Chl a concentration as well as the biodiversity of marine ecosystems.

The results of this study show evidence that the spatiotemporal analyses of SST levels and Chl a concentration are important for the characterization of the oceanographic and environmental aspects of its variability, considered necessary to explain physical, climatological, and biological processes of the coastal and marine ecosystems, which can influence the distribution and abundance of marine resources. The process of developing oceanic indicators, tightly linked to ecological processes, to describe the interannual effect of climate phenomena, increases the understanding of current trends of change associated with environmental and oceanographic factors, which generate useful information to understand current and future stages of marine ecosystems.

5. Conclusions

Results from the present research show strong evidence, suggesting that SST and Chl a variability in the South Californian Pacific and Gulf of California is tightly related to ENSO seasonal events (El Niño and La Niña). Therefore, the assessment of oceanographic conditions within the year (adapting into a seasonal approach), and considering ENSO phases for those seasons, increases our understanding of the biological and climatological processes occurring in the ocean under current (and potentially future) conditions. The oceanographic variability using monthly SST data analysis performed an optimal ecological characterization of the regions analyzed (the South Californian Pacific and Gulf of California), and cluster analysis suggests that these areas could be separated into 18 (currently 24 PMR), with similar oceanographic and environmental aspects. Statistical analysis suggests interannual variability of the monthly SST and Chl a data in all the regions for all the seasons during interannual events (El Niño and La Niña) and neutral conditions were statistically different, showing higher SST levels during warm ENSO season and lower Chl a concentrations, while during cold ENSO season, higher Chl a concentrations and lower SST levels. If these trends drive oceanographic dynamics to a new baseline, the effects on the distribution and structure of marine and coastal ecosystems could be significant, affecting the distribution and abundance of marine organisms. This study provides a significant framework for the analysis and characterization of environmental and oceanographic patterns to describe possible trends of changes associated with global climate patterns that might result in a series of noble conditions for marine ecosystems. Fluctuations in oceanic temperature, which are tightly related to climate phenomena (such as ENSO), exert pressure on productivity. Results from this work further our understanding of trends of change associated with environmental and oceanographic factors. For the PMR analyzed, drastic inter-seasonal SST fluctuations might have profound effects on productivity, and, therefore, a significant effect on the biota present on site, due to resource availability [63,64,123,124,125]. For conservation and management, the analysis of potential present and future climate scenarios would contribute to a better-informed decision-making process. Future research, regarding phenological response from algae, might include addressing the effects of ENSO and Sea Surface Temperature on greening (Chlorophyll a production), using higher temporal resolution datasets.