The Gorgona National Natural Park (GNNP; 2° 55’45” - 3° 00’ 55” N, 78° 09’ 00″ - 78° 14’ 30” W) is located on the continental shelf of the Colombian Pacific basin ( Figure 1 ). This region is characterized by warm and low-salinity surface waters (Giraldo López, 2008; Giraldo et al., 2008, 2011, 2014), and constitutes one of the rainiest regions of the world, being also the rainiest area of Colombia with annual values ranging from 2500 mm to 8000 mm (Rangel and Rudas, 1990; Blanco, 2009). This condition is mainly associated with the latitudinal migration of the Intertropical Convergence Zone (ITCZ; Díaz Guevara et al., 2008) and the Choco Jet Stream, which increases cloudiness and precipitation due to the convergence of trade winds from the northern and southern hemispheres (Prahl et al., 1990; Díaz Guevara et al., 2008; Guzmán et al., 2014; Serna et al., 2018). During the second half of the year, the ITCZ moves north, altering the wind direction near 3°N latitude, which becomes dominant in the southwesterlies and forms the Choco Jet off western Colombia (Poveda and Mesa, 2000; Amador et al., 2006; Guerrero Gallego et al., 2012). During the study period (September 13 to November 7, 2021), the area was affected by the strong La Niña 2020–2023 event (thermal anomalies of -1.4 ± 0.1°C; Supplementary Table 1 , IDEAM, 2021) that increased total precipitation in the Colombian Pacific.

Gorgona Park is in front of the Patia-Sanquianga delta complex ( Figure 1B ), the largest in the country, which contributes approximately 23% of the total freshwater discharged to the Colombian Pacific (2045 m³ s^-1; Díaz, 2007). This delta comprises several rivers, including the Guapi, Patia, Iscuande, Tapaje, and Sanquianga (Díaz, 2007, Figure 1B ). Likewise, the tidal regime around GNNP is semi-diurnal, repeating twice in 24 hours, with two alternating high tides (5.82 m) and two low tides (-0.78 m; Flanders Marine Institute [VLIZ] and Intergovernmental Oceanographic Commission [IOC], 2025).

The Pacific coast has an average daily solar radiation of 3500 to 4000 Wh m^-2 and an annual availability of solar radiation that varies between 1,080,000 and 1,440,000 Wh m^-2 (UPME, 2005). During the year’s second half, solar radiation decreases until it reaches its minimum in December (Rangel and Rudas, 1990). Regarding its biodiversity, GNNP hosts one of the most critical coral reef areas in the Eastern Tropical Pacific (ETP) region in terms of structure and species diversity (Glynn et al., 1982; Zapata, 2001; Díaz et al., 2001) and connects the ETP with other Pacific regions (e.g., Hawaii, Polynesia) via larval potential dispersion (Romero-Torres et al., 2018). Remarkably, these reefs persist despite being regularly exposed to multiple environmental stressors. At the beginning of the year, they experience sub-aerial exposure during extreme low tides (Zapata, 2001; Zapata et al., 2010), low pH levels reaching as low as 7.4 (Giraldo and Valencia, 2013), cooler temperatures (16.69 °C; Giraldo et al., 2008), and hypoxic conditions (< 3.7 mg L^-1) associated with seasonal upwelling (Castrillón-Cifuentes et al., 2023a, b). During the second half of the year, these reefs are subjected to low salinity levels during the rainy season (as low as 25; Giraldo et al., 2008), warmer temperatures (27 °C; Giraldo et al., 2008), and increased sedimentation rates (23.30 ± 4.34 g m^-2 d^-1; Lozano-Cortés et al., 2014). Coral species such as Pocillopora spp. and Pavona spp. dominate the eastern (leeward) and southern areas of Gorgona Island, while occurring more sparsely in the western and northern areas ( Figure 1A , Zapata, 2001). The sampling point ( Figure 1A ) was located at the north end of El Muelle reef (northward surface current), where the genus Pocillopora spp. predominantly dominates this and the major eastern coral patch of Azufrada (Zapata, 2001; Romero-Torres and Acosta, 2010).

Continuous measurements of pH_T, conductivity (mS cm^-1), and temperature (°C) were made at a fixed depth of around 2 meters in El Muelle reef, GNNP (2° 57’ 39” N, 78° 10’ 25” W; Figure 1A ) from 13 September to 07 November 2021, every three hours, using a buoy anchoring system designed to compensate the tidal changes and to keep the equipment at a similar depth. The iSAMI-pH equipment (programmed with SAMI Client v2.5 software) was used to record pH_T (accuracy: 0.0004 and precision: ± 0.0024) and temperature (accuracy: ± 0.14°C; at 24.55 °C Tris bottle number 10 Sunburst Sensors). The CTD Diver equipment (programmed with Diver Field software, Van Essen Instruments) was used to take conductivity measurements (accuracy: ± 1% mS cm^-1). Conductivity data were also downloaded using Diver Field software. The salinity was calculated using the practical scaling relationship proposed by Aminot and Kérouel (2004). The iSAMI-pH data were downloaded with the SAMI program (QC pH v4.4). Because the QC pH v4.4 program handles a constant salinity of 35 units by default, the pH_T data were corrected for salinity. To do the correction, the program was set to the corresponding average salinity (29.74 ± 0.94, 31.29; mean ± SD, maximum, n = 403) of the sampling point (2° 57’ 39” N, 78° 10’ 25” W). With the correction, differences in pH_T of ± 0.01 units were observed. We used the average salinity value to ensure consistency in salinity correction, as the CTD Diver recorded salinity 30 minutes after iSAMI-pH measurements from October 10 to November 7. When both devices recorded simultaneously between September 13 and October 10, the difference between pH corrected with average and measured salinity was insignificant (± 0.0001). iSAMI recorded 49 outliers out of 441 data points, which were excluded from the analysis as iSAMI-pH identified them as measurement anomalies (error codes 100, 1001, 1010, and 1000) associated with issues related to pumping, dye supply, or blank measurement. Additionally, 19 temperature and 38 salinity data points were identified as outliers when plotting variables and their relationship to total pH_T. These points were excluded from the analysis after applying a criterion based on standard deviation to consecutive measurements. Any change greater than 0.030 in pH, 0.30°C in temperature, or 0.94°C in salinity between measurements was considered a potential outlier. Flagged data points underwent contextual evaluation, considering environmental consistency (e.g., salinity-TA relationships, diel cycles, water mass characteristics), sensor artifacts, and physical plausibility. Only measurements inconsistent both statistically and contextually were excluded.

Total alkalinity (TA; μmol kg^-1) was calculated using a regression line between salinity (in situ data taken with CastAway-CTD equipment) and TA. Preciado (2023) built the regression (y = 57.479x+249.08; R2 = 0.94; n=268) using measured local discrete TA data obtained from 8 sampling stations around Gorgona, one of which included the iSAMI-pH deployment site. The data were collected over 13 sampling months, from September 2021 to October 2022, at depths ranging from 2 to 80 meters, following the SOP-3b protocol (Dickson et al., 2007). Discrete samples of TA (± 13 μmol kg^-1, Batch #182 2230.91μmol kg^-1) were collected monthly at the iSAMI-pH site from November 2021 to July 2022. Comparing the measured TA values ​​with those estimated from salinity, a mean difference of 11 ± 5 μmol kg^-1 and a regression coefficient of R² = 0.91 were observed ( Supplementary Figure 1 ). Specifically, in November 2021, the difference was 10 ± 3 μmol kg^-1. This regression approach for estimating TA through salinity has been validated in previous studies (Lee et al., 2006; Takatani et al., 2014; Carter et al., 2016; Fassbender et al., 2017; Metzl et al., 2024).

For the derivation of the carbonate system pCO_2w (μatm), dissolved inorganic carbon (DIC; μmol kg^-1), and omega aragonite (Ω_a), we used the Carbonate System equation solution (CO2sys v3.0, Pierrot et al., 2021), with the pair pH_T (n = 392) from iSAMI-pH and TA derived from the former regression, as well as salinity from CTD (Aminot and Kérouel, 2004) and temperature (iSAMI-pH). Propagation error analysis using CO2sys v3.0 estimated uncertainties of ±52 µmol·kg^-1 for DIC, ± 33 μatm for pCO₂, and ±0.16 for Ω_a. In the carbonate system equation solution, the following constants were used: (i) Dissociation constants for K₁ and K₂ from Millero, 2010) for waters ranging from 0 to 40, given that the study area the salinities variation are between 27.03 to 31.29, (ii) KHSO₄ dissociation constant from Dickson (1990), (iii) KHF from Perez and Fraga (1987), (iv) Total pH scale (mol-kg SW), (v) [B]T value from Lee et al. (2010), and (vi) EOS-80 standard.

To evaluate the influence of temperature and salinity, pH_T was normalized using the mean temperature (27.39°C) and salinity (29.74), as recommended by Terlouw et al. (2019). Additionally, normalization was performed using constant values representing the minimum observed temperature (26.72°C) and salinity (27.03), to account for the most extreme conditions in the dataset ( Supplementary Figure 2 ). This normalization process isolates the potential effect of temperature and salinity on pH_T, following the methodology outlined by Sarmiento and Gruber (2006). The influence of temperature and salinity corresponds to the difference between the normalized pH_T and the in situ measured pH_T.

No significant differences were found between high and low tides for the dependent variables of pH_T, salinity, temperature, pCO_2w, DIC, and TA ( Supplementary Table 3 ). However, pH_T during flood tide (8.010 ± 0.031) was slightly higher than during ebb tide (8.008 ± 0.029), with a lower positive slope on the regression line ( Figure 3A ). pCO_2w and DIC showed similar values during the flood tide (396 ± 38 μatm and 1726 ± 52 µmol kg^-¹, respectively) and the ebb tide (397 ± 35 μatm and 1730 ± 46 µmol kg^-¹, respectively), with a slight negative slope observed in the regression trend ( Figures 3D, E , respectively). Variables such as temperature, salinity, and TA exhibited slopes close to zero, indicating that this area’s seawater and river water are well-mixed.

The independent atmospheric variables (solar radiation and total precipitation) were correlated with water type (sea surface temperature and salinity) and carbon chemistry data (pH_T, pCO_2w, DIC, and TA). Total daily solar radiation showed the strongest correlations with all dependent variables; suggesting that average radiation values of 3078 ± 1098 Wh m^-² (n= 56), as well as radiation values in a range of 1378 to 5968 Wh m^-² displayed positively correlations with mean daily temperature and pH_T, and negative correlations with salinity, pCO_2w, DIC, and TA ( Table 2 ). In contrast, total daily precipitation (mean average of 31 ± 27 mm; n= 56) within the 0 to 115 mm range only correlated negatively with mean daily salinity ( Table 2 ). No significant correlation was observed between daily precipitation and pH_T, temperature, pCO_2W, DIC, and TA ( Table 2 ). In addition, rainwater samples collected during the high rainfall season (May 2022) in GNNP showed the lowest salinity values (0.26).

pH_T, temperature, pCO_2W, and DIC followed a pronounced diurnal cycle with significant differences (p < 0.05; Supplementary Table 4 ) between early morning and late evening hours ( Figure 4 ). During early morning hours, after sunrise (5:00 - 6:00), at 8:00 the lowest values of pH_T (7.993 ± 0.040; Figure 4A ) and temperature (27.23 ± 0.18; Figure 4B ) were recorded, as well as the highest values of pCO_2W (413 ± 40 μatm; Figure 4C ) and DIC (1739 ± 46 μmol kg^-¹; Figure D). In the late afternoon (17:00 - 18:00), during sunset, the highest values of pH_T (8.030 ± 0.025; Figure 4A ) and temperature (27.61 ± 0.36; Figure 4B ) were found, together with the lowest pCO_2W (374 ± 25 μatm; Figure 4C ) and DIC (1718 ± 50 μmol kg^-¹; Figure 4D ). In contrast, salinity ( Figure 4E ), TA ( Figure 4F ), and carbonate alkalinity did not show significant differences (p > 0.05, Supplementary Table 4 ) between early morning and late evening hours.

The high variability found in variables pH_T, salinity, temperature, pCO_2w, DIC, and TA at El Muelle reef highlights the influence of La Niña 2020–2023 event. La Niña event strengthened the Choco Jet (Hoyos et al., 2013; Arias et al., 2015; Serna et al., 2018), which caused an increase in precipitation in the Colombian Pacific basin and, consequently, an increase in the river flow into the sea (Restrepo and Kjerfve, 2000; Blanco, 2009; Serna et al., 2018). During the sampling period from September to November 2021, the station “57025020 Gorgona Guapi” recorded a total precipitation of 2878 mm, exceeding the total multiannual monthly average by more than 59% to 1815 mm.

Likewise, the station “53047010 Sangaral” recorded a monthly average flow of the Guapi River of 447 m³ s^-1, exceeding by 44% the total multiannual monthly average of 309 m³ s^-1. Thus, the sampling period was characterized by a considerable inflow of freshwater from river discharges and high rainfall that explained the reduced recorded salinity values and part of the pH_T variability ( Figure 2 ). The salinity normalization and correlation ( Table 2 ) suggest that salinity plays a significant role in pH_T variability. The interplay between low salinity and reduced alkalinity (directly correlated) and pH variability underscores its critical role in coastal biogeochemical cycles. Reduced salinity not only dilutes carbonate and bicarbonate ions but also disrupts the equilibrium of the CO₂ system, as observed in estuarine zones by Feely et al. (2010) and Cai et al. (2021), where freshwater influx exacerbates anthropogenic CO₂ uptake. These dynamics challenge the buffering capacity and resilience, particularly in reefs, where a range of water pH is essential for calcification.

Additionally, high rainfall and the consequent excess of freshwater (characteristic of a RioMar system) acted as a surface layer dilution factor (Zeebe and Wolf-Gladrow, 2001; Turk et al., 2010; Trujillo and Thurman, 2016; Ho and Schanze, 2020), contributing to reducing, even more, the salinity over time. For example, Ho and Schanze (2020) reported a rapid decrease in salinity of 4.4 at depths of 2 to 3 cm in response to heavy rainfall in the Eastern Equatorial Pacific, highlighting the impact of precipitation on coastal salinity dynamics. Salinity gradients and alkalinity depletion may act as synergistic stressors, which align with broader concerns about coastal acidification (Doney et al., 2009) and the ecological risks associated with climate change scenarios.

The tides exert a continuous physical influence, due considerable tidal variation, having maximum amplitudes of 5.82 m and minimum of -0.78 m in 2021 (3.39 ± 1.20 m; Flanders Marine Institute [VLIZ] and Intergovernmental Oceanographic Commission [IOC], 2025), implying millions of cubic meters of freshwater entering and leaving the coastal system, mixing oceanic water with freshwater coming from the Patia-Sanquianga complex plus the rainfall. The well-mixed RioMar water has minimal salinity variations throughout a full ebb and flow cycle ( Figure 3 ). This high mixing condition persists up to 55 km downstream of the river mouth. As indicated by Palacios Moreno and Pinto Tovar (1992), due to the high tide, mixing begins at the river mouth, where seawater enters through the bottom of the Guapi River, resulting in waters with a salinity of 25.60 at 10 m, compared to 17 at 1 m. Furthermore, tides and the southern trade winds during La Niña events generate coastal currents and a strong swell (Osorio et al., 2014) that contribute to mixing these two water types, creating an estuary environment. Consequently, this coastal zone can be classified as a Type 2 RiOMar regime (Dai et al., 2022a), characterized by a high-energy tidal system and significant mixing near the river mouths (Lamarque et al., 2022).

In this scenario, the riverine influence reaching GNNP may explain the observed decrease in pH_T over the study period. This trend is likely driven by the input of CO₂ derived from riverine DIC, potentially due to the input and decomposition of organic matter, which increases H⁺ ion concentrations, as observed in other coastal regions (Vargas et al., 2016; Cai et al., 2021). In fact, in the Eastern Colombian Pacific, the observed decrease and low values of pH_T were strongly correlated with salinity, pCO_2w, and DIC ( Table 2 ). Between 2018 and 2021, data collected at 13 river stations near the GNNP showed pH_NBS values ranging from 6.3 to 8.3. Suspended solids ranged from 11 to 407 mg L^-1, while nitrate (NO₃ ^-) concentrations ranged from 2 to 90.9 µg L^-1. Phosphate (P-PO₄ ^3-) levels fluctuated between 2 and 3.3 µg L^-1, and chlorophyll-a concentrations ranged from 0.07 to 9.8 µg L^-1. The highest values ​​for these parameters were recorded during the year’s second half, coinciding with the rainy season (INVEMAR, 2020b; INVEMAR, 2022), which could contribute to acidification. This phenomenon aligns with RiOMar systems, where high DIC and nutrient inputs from rivers reach nearshore waters, and organic matter is subsequently exported to the continental shelf (Gan et al., 2009; Dai et al., 2022a). Decreases in pH_T in other estuaries have been indirectly linked to fluvial discharges. For example, along the central-southern Chilean Pacific coast (Biobío River basin), the expansion of the river plume seaward led to reduced pH_T conditions (7.6) during periods of maximum higher river discharge, compared to more oceanic stations where pH_T ranged from 7.95 to 8.15 (Vargas et al., 2016). Likewise, in the South Pacific Ocean, in Coral Bay, fluvial discharges from low-pH rivers (7.739 ± 0.022) cause coastal areas with low pH_T (8.014 ± 0.015) (Aguilera et al., 2013), similar to low pH observed across the GNNP Colombian Pacific Region. Indeed, the surface values (2m) of pH_T (8.01 ± 0.03), salinity (29.74 ± 0.94), TA (1959 ± 54 µmol kg^-1), DIC (1728 ± 49 µmol kg^-1) and Ω_a (2.71 ± 0.17) at El Muelle reef (GNNP) are below the surface averages reported for the North Pacific (8. 105 ± 0.06, 34.05 ± 0.86, 2255 ± 34 µmol kg^-1, 1959 ± 42 µmol kg^-1 and 3.30 ± 0.7; respectively) and the South Pacific (8.079 ± 0.03, 35.29 ± 0.55, 2319 ± 35 µmol kg^-1, 2003 ± 39 µmol kg^-1 and 3.52 ± 0.4) (Feely et al., 2009).

On the contrary, oceanic areas without river input, such as North Pacific Monterey Bay, have recorded relatively high pH_T (Gray et al., 2011). However, at higher latitudes on the Pacific coast of Costa Rica, low pH_T levels were associated with the rainy season (Sánchez-Noguera et al., 2018). A drop in pH_T is also known to occur in the estuaries of Swartkops and Sundays in South Africa, where, due to strong inflows of freshwater, a high pH_T decrease of 0,46 pH units was produced (Edworthy et al., 2022). The relatively low pH_T of the Pacific Colombian Coast (El Muelle, GNNP, Table 3 ) is similar to studies in the East River China Sea, where the low pH_T (7.98 to 8.07) was attributed to the high river inputs from the Changjiang plume, known to carry organic carbon, nutrients, and CO₂ supersaturated water (Wu et al., 2021). We already know that freshwater inputs from rivers play a critical role in shaping carbonate chemistry in coastal systems, particularly in the continental shelf. As well as in our study site, Rérolle et al. (2018) have identified riverine discharges, characterized by high DIC and TA, as key drivers of reduced pH_T in the Irish Sea, southern North Sea, and Skagerrak region. Further studies on carbon chemistry in the region are needed, including direct measurements of DIC and dissolved organic matter in the Patia-Sanquianga complex during wet and dry seasons. These and other chemical parameters, such as nutrients, are crucial to evaluating rivers’ contribution to coastal carbonate chemistry, particularly near reef systems, and the impact on ocean acidification (Borges et al., 2005; Cai et al., 2021). The main knowledge gaps include the effects of remineralization on pH in tropical estuaries (Dai et al., 2022b; Vargas et al., 2016) to understand the vulnerability of ecosystems to OA.

During the study period, the ITCZ was located close to the equator in the Colombian Pacific Ocean (Díaz Guevara et al., 2008), which intensified cloudiness and precipitation due to the convergence of trade winds from the northern and southern hemispheres (Díaz Guevara et al., 2008). The increase in cloudiness likely contributed to the decline in solar radiation, which diminished from 3,709 ± 1,205 Wh m^-2 in September to 2,501 ± 789 Wh m^-2 in November, with a marked reduction between October 1 and 25 (2,488 ± 590 Wh m^-2; Figure 5A ). Given the positive correlation between solar radiation and temperature (Trujillo and Thurman, 2016; Webster, 2020), the decrease in solar radiation may explain the low SST observed from September 13 to November 7. According to Millero (2007, 2013) lower temperatures increase CO₂ solubility in seawater, suggesting that the temperature decline likely improved oceanic CO₂ uptake by enabling its transfer from the atmosphere to the ocean during the study period. The resulting CO₂ increase could explain pCO_2w and DIC’s highest values ( Figure 2 ), particularly between October 1 and 25, when solar radiation was at its lowest ( Figure 5 ). Thus, the solar radiation and temperature reduction may partially explain the observed decline in pH_T during this period. The inverse correlation between pH_T and pCO_2w also suggests that fluctuations in pCO_2w (indirectly DIC) could help explain pH_T variations over September-November ( Table 2 ; Zeebe and Wolf-Gladrow, 2001; Millero, 2013; Cai et al., 2021). After October 25, the rise in solar radiation and temperature may have led to a pH_T increment by approximately 0.028 units, along with a decrease in pCO_2w by 41 μatm and DIC by 100 μmol kg^-1 ( Figures 2D, E ).

Between October 3 and 16, total precipitation amounted to 22 mm ( Figure 5B ), with most days recording values below the mean daily precipitation of 31 mm. The observed reduction in precipitation likely diminished freshwater inflow relative to seawater, thereby enhancing salinity levels and elevating TA and DIC concentrations during this period. During the sampling period (September 13 to November 7), evaporation (Copernicus Climate Change Service, 2024) was highest in October (-62 mm), followed by September (-36 mm) and November (-15 mm). However, in all cases, the monthly accumulated evaporation remained well below the corresponding monthly precipitation totals of 680 mm, 727 mm, and 261 mm, respectively (DHIME, 2025).

In addition to previously discussed mechanisms, the elevated levels of pCO_2w, DIC, and salinity observed between October 1 and 25 may also be influenced by a coastal upwelling event. This hypothesis is supported by the positive values of zonal Ekman transport (ZET) during this period, which indicates eastward surface water movement, given the regional coastline orientation and prevailing wind patterns (Corredor-Acosta et al., 2020). The eastward Ekman transport suggests offshore water displacement and the potential for deeper, carbon- and nutrient-rich water upwelling ( Supplementary Figure 4 ).

During early morning hours (5:00 - 6:00 am), we registered the lowest pH_T and temperature values, along with the highest concentration of pCO_2w and DIC, suggesting the influence of night-time respiration processes before daylight (Millero, 2013; Albright et al., 2013; Sánchez-Noguera et al., 2018). On the other hand, the highest values of pH_T and temperature and the lowest values of pCO_2w and DIC recorded during the late afternoon (17:00 - 18:00) suggest photosynthetic processes during daylight driven by increased solar radiation (Cyronak et al., 2013). The mean diel change of 0.037 units in pH_T and 0.38 °C in temperature between early morning and afternoon hours fall within the expected diel variability for shallow environments (0.7–17 m), where biological and physical processes significantly influence pH and temperature dynamics (Cyronak et al., 2013).

The absence of a significant difference in TA and carbonate alkalinity between early morning and late afternoon hours may be due to the derivation of TA using a TA-salinity relationship, which primarily accounts for processes such as evaporation, precipitation, and mixing tides (Spaulding et al., 2014). In contrast, calcification and dissolution, more likely to exhibit variations over a 24-hour cycle, may not be adequately reflected (Zeebe and Wolf-Gladrow, 2001; Millero, 2013; Cai et al., 2021). However, the data indicate that the El Muelle reef, close to the sampling point, is composed predominantly of Pocillopora spp. ( Figure 1 ; Zapata, 2001; Acosta et al., 2007), was exposed during the sampling period to low pH_T (8.01 ± 0.03), salinity (29.32 ± 1.01), and Ω_a (2.71 ± 0.17). These environmental conditions are considered challenging for Pocillopora spp., which are known to be highly sensitive to reduced pH and lowered salinity associated with river discharge (Alvarado et al., 2005; Lizcano-Sandoval et al., 2018; Sánchez-Noguera, 2019). Notably, the global, annually averaged tolerance limit for coral reefs is an Ω_a of 2.82, indicating that corals at this site live below the threshold commonly reported in the literature (Guan et al., 2015).

Despite these adverse conditions and frequent riverine influence, especially during the rainy season in the latter half of the year (Giraldo et al., 2008, 2011), Pocillopora spp. persist and even dominate the area (Zapata, 2001; Acosta et al., 2007; Lizcano-Sandoval et al., 2018). Previous studies have shown that Pocillopora spp. in GNNP can maintain growth rates (Lizcano-Sandoval et al., 2018) and tolerate hypoxia, low salinity, and temperature fluctuations by reducing reproductive output rather than growth (Castrillón-Cifuentes et al., 2023a, 2023). This suggests that these corals are either locally adapted or possess a degree of physiological tolerance to low pH, salinity, and Ω_a. Similar findings from the Pearl River Estuary in Southeast China indicate that long-term hypo-salinity acclimation can enhance the tolerance of Pocillopora spp. to low salinity by reducing energy consumption, slowing metabolism, improving the energy metabolism of their symbiotic algae (Symbiodiniaceae), and altering their symbiotic bacterial communities to avoid bleaching (Chen et al., 2024). Nevertheless, Pocillopora spp. corals at Gorgona Island exhibit lower calcification rates (3.16 g CaCO₃ cm^-2 yr^-1; Lizcano-Sandoval et al., 2018) compared to those reported in other reef systems, such as Panama (5–6 g CaCO₃ cm^-2 yr^-1; Manzello, 2010) and Mexico (2.99–6.02 g CaCO₃ cm^-2 yr^-1; Medellín-Maldonado et al., 2016; González-Pabón et al., 2021; Tortolero-Langarica et al., 2022). To establish a more robust correlation between the carbonate system and calcification rates in Pocillopora spp. at GNNP, future research efforts should prioritize in situ measurements of TA, dissolved oxygen, and direct calcification assessments (e.g., via buoyant weight or alkalinity anomaly techniques) on a 24-h scale during both rainy and dry seasons, to better quantify metabolic dynamics and clarify the adaptive capacity of this species under adverse environmental conditions.

In addition, future research should prioritize high-frequency measurements of pH, nutrient concentrations, and stable isotopes across diurnal, seasonal, and interannual timescales. These efforts should encompass both rainy and dry seasons and different ENSO phases, such as El Niño and neutral conditions, to better distinguish water sources in the region and assess the influence of these temporal scales on carbonate chemistry variability.