Introduction
Coastal ecosystems such as mangroves play a crucial role in coastal protection, biodiversity and provide a nutrient-rich environment for the growth of aquatic and terrestrial organisms. Also, the increase in anthropogenic CO2 emissions into the atmosphere has awakened greater interest in the study and conservation of mangroves, due to their capacity to store large amounts of carbon in the sediments (Alongi Reference Alongi2012). This helps to mitigate climate change by preventing carbon from being released into the atmosphere in the form of carbon dioxide (Jupin et al. Reference Jupin, Boussafir, Sifeddine, Ruiz-Fernández, Sanchez-Cabeza and Pérez-Bernal2024; Osland et al. Reference Osland, Feher, López-Portillo, Day, Suman, Guzmán Menéndez and Rivera-Monroy2018). The study of mangrove sediments is essential to better understand this process. Another aspect that makes the study of mangroves interesting is that their formation is proportional to the rate of sea level rise (Saintilan et al. Reference Saintilan, Khan, Ashe, Kelleway, Rogers, Woodroffe and Horton2020; Sasmito et al. Reference Sasmito, Murdiyarso, Friess and Kurnianto2015; C. D. Woodroffe Reference Woodroffe1990; S. A. Woodroffe et al. Reference Woodroffe, Long, Punwong, Selby, Bryant and Marchant2015). To understand these processes, the chronology of sediment accretion can provide information on the evolution over time of the coastal zone, through changes in their physical and chemical composition.
To understand processes such as carbon sequestration rates and mangrove accretion rates, isotopic techniques with nuclides such as 14C have proven to be useful in this kind of study (McKee et al. Reference McKee, Cohen, Dettman, Palacios-Fest, Alin and Ntungumburanye2005; Sefton et al. Reference Sefton, Woodroffe, Ascough and Khan2022). However, coastal ecosystems are regions with complex interaction between aquifer and the marine environment, representing significant challenges for dating using these nuclides. The temporal evolution of these systems is sometimes hindered by bioturbation (Martinetto et al. Reference Martinetto, Montemayor, Alberti, Costa and Iribarne2016) and because of the contribution from different sources to these radionuclides, which alter the ages of the stratigraphic layers.
The Yucatán Peninsula has many shallow coastal lagoons and concentrates about 60% of the mangroves in all of Mexico. It is also important to mention that Mexico has 6% of all the mangroves in the world (CONABIO 2024).
In this study, we present the ages of different sediment fractions in cores from two sites in the Yucatán Peninsula, Mexico: Jaina (JN) in Campeche and Ría Lagartos (RL) in Yucatán (Figure 1). Both mangrove-dominated areas are located in a region characterized by karst systems with groundwater discharge but have differences in geological and environmental conditions. Jaina, a reef island off the coast is characterized by the presence of hammock mangroves (known as Petenes) where fissures allow freshwater upwelling, while Ría Lagartos is a coastal estuary exposed to hydrometeorological impacts, such as hurricanes and storms (Herrera et al. Reference Herrera-Silveira, Pech-Cardenas, Morales-Ojeda, Cinco-Castro, Camacho-Rico, Caamal Sosa, Mendoza-Martinez, Pech-Poot, Montero and Teutli-Hernandez2020). We focused on dating fractions where the carbon originates from various sources, to reach an adequate way to determine the most probable age of sediment deposition. To establish mangrove chronologies, macrofossils and pollen are some of the preferred fractions, as these have very short transit time and their radiocarbon age best reflects the time of deposition. Given that macrofossils are often absent in sediments, and pollen rather provides information on the adaptation or mortality of mangroves with environmental changes (Ellison Reference Ellison2008), the criterion followed was to find the fraction whose age is closest to that of the macrofossils (Sefton et al. Reference Sefton, Woodroffe, Ascough and Khan2022; Strunk et al. Reference Strunk, Olsen, Sanei, Rudra and Larsen2020).
a) View of the Yucatán Peninsula, located in the southeast of Mexico. b) View of the Yucatán Peninsula with the two sites of interest. c) Location of Jaina, Campeche, Mexico (20°51’34.96” N, 90°19’59.59” W) (in the Gulf of Mexico) d) Location of Ría Lagartos, Yucatán, Mexico (21°33’28’’ N, 87°50’35’’ W) (in the intersection of the Gulf of Mexico and the Caribbean Sea).
Figure 1. Long description
The image consists of four elements: one map and three detailed views. The first element is a map showing the Yucatan Peninsula located in the southeast of Mexico. The second element is a closer view of the Yucatan Peninsula with two specific sites of interest marked: Jaina in Campeche and Ria Lagartos in Yucatan. The third element is a detailed view of the location of Jaina, Campeche, Mexico, situated in the Gulf of Mexico. The fourth element is a detailed view of the location of Ria Lagartos, Yucatan, Mexico, situated at the intersection of the Gulf of Mexico and the Caribbean Sea. The maps highlight the geographical context and specific locations of interest for studying mangrove ecosystems.
The radiocarbon ages from different organic matter fractions of sediment cores were extracted from each layer to define all carbon sources contributing to the total organic carbon (TOC), and to choose the best method to establish the chronology. The results of cores from the two areas were compared to observe whether, in two completely different sites, the fractions showed a similar pattern or had a different behavior. In case that such a pattern of behavior is found, it would be possible to determine the best fraction for dating this type of paleosols as closely as possible to the exact depositional age. This would be useful to better understand issues such as mean sea level rise, carbon sequestration and compositional changes due to storms or extreme biogenic events.
Sampling and methodology
Two sediment cores from Jaina (JN) (20°51’34.96”N, 90°19’59.59”W), Campeche were collected in 2017 (Figure 2a). Core JN02 was obtained from a “Petén” (known as hammock), consisting of patches of circular formations with mangrove tree vegetation where karst fissures allow freshwater upwelling (Barrera Reference Barrera1982; Duran Garcia Reference Durán García1995; CONANP 2006) and core JN03 was obtained from a pristine mangrove forest. One core from Ría Lagartos (RL03), Yucatán (21°33’28’’N, 87°50’35’’W) was collected in 2018 from a shallow submerge zone. The material accumulated in the bottom of the core was peat, likely associated with an ancient mangrove phase, with a recent carbonate sediment layer at the top (Figure 3).
a) Cores JN02 and JN03 collected in Jaina. It can be observed that the cores used in this work reached a length of between 50 and 53 cm of material. b) The Bayesian model corresponds to Macros from core JN02. For the age-depth model, the P_sequence model of OxCal was applied to dated samples using the curve Bomb21NH2 (14C calibrated age ranges correspond to 95% confidence interval). The carbon ages of TOM, Humin and HA fractions were obtained using the OxCal Mix Curves function, with IntCal20 and Marine20 calibration curves. The local ΔR for the closest zone was –111 ± 25 taken from the Marine Reservoir Correction Database (https://calib.org/marine/).
Figure 2. Long description
Two sediment cores labeled JN02 and JN03 are shown side by side. The core labeled JN02 has a length of 47 centimeters, while the core labeled JN03 has a length of 50 centimeters. Next to the cores is a depth chart that models dates using a Bayesian approach. The chart includes various fractions such as TOM, Humin, and HA, with carbon ages calibrated using IntCal20 and Marine20 curves. The local delta R value for the closest zone is -111 plus or minus 25, taken from the Marine Reservoir Correction Database. The depth chart shows modeled dates ranging from 1400 to 200 calibrated years before present (calBP), with different symbols representing TOM, AH, Humin, and Macros. The depth values range from 32 centimeters to 46 centimeters. All values are approximated.
a) The RL03 core collected in Ria Lagartos is shown. b) The Bayesian age depth model for the middle part from 21 to 45 cm of core RL03. Age model was made based on the dating results of Macros fractions.
Figure 3. Long description
The image contains two sediment cores from Jaina (JN). On the left, the RL03 core collected in Ria Lagartos is shown. The core is divided into sections marked by depth in centimeters, ranging from 0 to 45 centimeters. On the right, the Bayesian age depth model for the middle part from 21 to 45 centimeters of core RL03 is displayed. The age model is based on the dating results of Macros fractions. The graph includes multiple colored lines representing different data sets: MACROS in black, TOM in green, HUMIN in red, and HA in purple. The x-axis represents the modeled date in calibrated years before present (calBP), ranging from 2600 to 4400 calBP. The y-axis represents the depth in centimeters, ranging from 20 to 45 centimeters. The graph shows the relationship between depth and modeled dates, with the blue shaded area indicating the confidence interval. The graph also includes specific dating points labeled as R_Date LEMA with corresponding numbers. The trends, values, and confidence intervals are visually represented to show the age-depth relationship in the sediment core.
The cores were collected with a push corer. Slices of 1 cm or 5 cm were cut and freeze-dried. The sediments were sieved through a 1 mm sieve, powdered and TOC, TN (total nitrogen), C/N ratio and 14C were analyzed. Samples from Jaina showed a sediment texture without apparent changes. In Ría Lagartos, the core was 7.6 cm thick and 60 cm long. The stratigraphy of the core was associated with an old mangrove forest that was flooded.
For radiocarbon measurements, samples from each layer were pretreated to extract different organic fractions using a modified procedure of Abbott and Stafford (Reference Abbott and Stafford1996). The macrofossils (Macros) comprising seeds, stems and leaves of terrestrial species, retained in the 1 mm sieve, were cleaned, dried and pretreated with an acid-base-acid procedure, using 0.5M HCl at 60 °C to get rid of carbonates.
When the Macros stop bubbling, it was rinsed with deionized water several times until it reached a neutral pH. The sample was then treated with 0.1M NaOH for 12 hr at 60 °C (if necessary), rinsed to neutral pH and acid washed with 0.5M HCl for 12 h at 60 °C. Finally, the sample was rinsed to neutral pH, and dried at 60 °C.
For total organic matter (TOM) analysis, the collected bulk sediment sample smaller than 1 mm, was treated with 0.5M HCl, several times until carbonates were removed, indicated by the cessation of bubbling. The sample was then rinsed to neutral pH and dried at 60 °C for 48 hr, yielding the TOM fraction.
One portion of the TOM was treated with 0.05M KOH, and the dark supernatant containing humic acids was separated to continue later with their extraction. The remaining solids were rinsed to neutral pH, treated with 0.5M HCl at 60 °C for 12 hr, rinsed again, and dried at 60 °C, yielding the base residue fraction (Humin).
Humic acids (HA), representing the base-soluble fraction, were recovered by filtering the dark supernatant (0.45 µm) and washing them with 0.2M HCl to pH 1. The solution was allowed to separate into two phases, and the upper one containing fulvic acids was discarded. The retained dark fraction, containing the HA fraction, was rinsed three times with 0.2M HCl and freeze-dried. The reproducibility of these procedures for obtaining different fractions was previously validated at our laboratory using Black Mat samples analyzed by multiple laboratories (Ardelean et al. Reference Ardelean, Israde-Alcántara, González-Hernández, Arroyo-Cabrales, Solis-Rosales, Rodríguez-Ceja, Pears, Watling, Macías-Quintero and Ocampo-Díaz2018).
For 14C AMS analysis, samples containing approximately 1 mg of carbon were processed in an Automated Graphitization Equipment (AGE3, Ion Plus) to transform carbon into CO2 by combustion and then to pure graphite by reaction with hydrogen (Wacker, Němec, and Bourquin Reference Wacker, Němec and Bourquin2009). The total organic carbon TOC(%) and total nitrogen N(%) were directly obtained in an Elemental Analyzer (Vario Micro Cube, Elementar) coupled to AGE3, before the CO2 was transferred to the reactors. The graphite was pressed in an Al cathode for 14C-AMS analysis. Graphite 14C/12C and 13C/12C ratios were measured at LEMA laboratory, Institute of Physics, UNAM with an AMS system based on a 1MV Tandetron (HVEE) (Solís et al. Reference Solís, Chávez-Lomelí, Ortiz, Huerta, Andrade and Barrios2014). The 14C data is expressed as a percentage of modern carbon (pMC=F14C × 100) (Reimer et al. Reference Reimer, Brown, Reimer, Campana, Jones, Cook and Begg2004), and in conventional age 14C (BP), rounded according to community recommendations 14C (Stuiver and Polach Reference Stuiver and Polach1977). Calibration of radiocarbon ages was performed using the OxCal v4.4.4 code (Ramsey and Lee Reference Ramsey and Lee2013; https://c14.arch.ox.ac.uk/oxcal/OxCal.html) with calibration curves IntCal20 (Reimer et al. Reference Reimer, Austin, Bard, Bayliss, Blackwell, Ramsey and Butzin2020) and Bomb21NH2 (Hua et al. Reference Hua, Turnbull, Santos, Rakowski, Ancapichún, De Pol-Holz and Hammer2022). Calibrated ages are reported in calendar years before present (calBP). A Bayesian model was constructed using the calibrated dates; depending on the C/N of the organic fractions, some carbon ages were obtained using the OxCal Mix_Curves function, with IntCal20 and Marine20 calibration curves.
Results
In the cores from Jaina, for 14C analysis, we selected the lower mid-depth section (32–45 cm for JN02 and 38–48 cm for JN03) (Figure 2). The obtained ages for samples of Macros, TOM, Humin and HA taken at different depths, are shown in Figure 2 and Table 1.
Results for organic fractions from JN02 and JN03 samples. Radiocarbon values are reported as percentage of Modern Carbon (pMC). Atmosphere pMC was between 101.6 and 101.2% in 2017–2018, where the cores were extracted (Hua et al. Reference Hua, Turnbull, Santos, Rakowski, Ancapichún, De Pol-Holz and Hammer2022). Calibrated results are based on the present work modeling.
Table 1. Long description
The table presents radiocarbon dating results for organic fractions from JN02 and JN03 samples. It includes columns for lab code, depth in centimeters, carbon fraction type, carbon to nitrogen ratio, percentage of modern carbon with one standard deviation, radiocarbon age in before present, and modeled age with a 95.4% confidence interval. The table has 25 rows and 8 columns. Notable entries include modern ages for several samples and varying radiocarbon ages for different carbon fractions at different depths. The data highlights the variability in radiocarbon ages across different organic fractions and depths.
In the case of the two Jaina cores, variations in pMC are observed for the different fractions studied at different depths. We compared the age offset of average calendar ages modeled of HA and Macros fractions for JN02 core. For example, at 34–35 cm depth an offset of 421 years was found, while at 42–43 cm depth, the offset increased to 796 years (Table 1). Macros and TOM fractions from cores JN02 and JN03 yielded modern ages down to 45 and 48 cm depth respectively. Macros fractions from JN02 averaged 110% pMC, while TOM averaged 104% pMC, indicating a possible incorporation of 14C-depleted carbon, likely from the humic acids which showed much older ages (86.3 to 94.1% pMC). Regarding the modeled ages, both cores, JN02 and JN03, are similar.
An age–depth model for core JN02 was developed using Macros 14C ages (Figure 2, Table 1), supported by C/N ratios indicating a terrestrial origin. In contrast, low C/N values in TOM, Humin, and HA fractions suggest mixed terrestrial, freshwater, and marine sources; these were calibrated using OxCal Mix_Curves function with IntCal20 and Marine20 calibration curves. A local ΔR of −111 ± 25 was applied based on the Marine Reservoir Correction Database for Isla Mujeres, the nearest site to Ría Lagartos.
The C/N ratio has been used as an indicator of the protein content in plants (Müller and Mathesius Reference Müller and Mathesius1999). High C/N values (C/N > 18) have been associated with terrestrial plants, while aquatic plants (algae marine and phytoplankton) with higher content of protein have lower C/N ratios (C/N < 10). After applying the OxCal Mix_Curves function, the best fit of the model was obtained by assuming that less than 15% of the carbon input came from aquatic sources. The ages obtained for TOM and Humin were like those of Macros, indicating that the origin of the carbon is dominated by terrestrial material and that the local reservoir effect (LRE) is not a determining factor. The fact that low C/N ratios do not reflect the contribution of carbon from different sources supports previous criticisms of the use of the C/N ratio to determine the fraction of terrestrially derived organic carbon (or organic matter) in aquatic and sedimentary environments (Perdue and Koprivnjak. Reference Perdue and Koprivnjak2007).
The C/N ratios of the different fractions were also estimated to distinguish the sources of organic carbon in each fraction, as they are frequently used to infer if organic carbon in samples originate from algae in freshwater and marine phytoplankton (C/N of 4 to 10) or terrestrial plants (C/N of 18–40 for C3 plants) (Strunk et al. Reference Strunk, Olsen, Sanei, Rudra and Larsen2020). For the Macros fraction, an average C/N ratio of 41 was obtained, indicating its terrestrial (mainly ligneous) plant carbon sources. The average C/N ratio for the TOM, Humin and HA fractions were 14, 13 and 11, respectively, reflecting a mixture of terrestrial, aquatic and marine organic matter sources. Modeled ages for JN02 and JN03 cores, are similar. The δ13C-AMS values in TOM fraction from JN02 and JN03 ranged between –27.9 to –31.6‰ and –29.1 to –30.9‰ respectively. Despite the isotopic fractionation that may occur during δ13C-AMS measurements, these values generally are consistent with the fact that the TOM in the samples from Jaina is mainly of terrestrial origin.
In the Ría Lagartos site, the top of the RL03 core (0-15 cm) consisted of carbonates. Underneath the carbonate layer, the material consisted of peat accumulated during the mangrove phase, which managed to remain in equilibrium with the rise in sea level. 14C-AMS analyses were performed in the section, between 21 and 45 cm (Figure 3 and Table 2).
Results for organic fractions from RL03 core. Radiocarbon values are reported as percentage of Modern Carbon (pMC). Calibrated results are based on the present work modeling.
Table 2. Long description
The table presents radiocarbon dating results for various organic fractions from the RL03 core. It includes data on depth in centimeters, carbon fraction types (Macros, TOM, Humin, HA), carbon to nitrogen ratio, percentage of modern carbon (pMC) with one standard deviation, radiocarbon age in years before present (BP), and modeled calibrated age ranges. The table has 14 rows and 8 columns, with each row representing a different sample and its corresponding measurements. Notable trends include variations in radiocarbon ages and calibrated age ranges across different depths and carbon fractions.
As with the Jaina cores, samples obtained from each sediment layer were pretreated to obtain Macros, TOM, Humin and HA organic fractions. An age-depth model was constructed based on 14C ages from TOM, using the OxCal v4.4.4 program for the core RL03, and compared with the rest of the fractions. The OxCal P_Sequence function obtained with data from Macros, TOM, Humin and HA fractions is shown in Figure 3. The agreement index was close to 100%. Macrofossils from Ría Lagartos sediment cores showed the youngest ages and HA the oldest ones. TOM and Humin fractions yielded dates between Macros and HA fractions.
We evaluated the offsets in pMC values among different organic fractions from core RL03 to identify variations in the input of ancient organic carbon incorporated into the sediments. At a depth of 21–22 cm, the Macros yielded a pMC of 70.6 ± 0.3, while the TOM fraction was slightly lower at 69.7 ± 0.2, resulting in an offset of 0.9. As depth increased, the pMC values between these fractions converged; notably, at 25–30 cm, both Macros and TOM exhibited an identical pMC of 67.8 ± 0.3. At this same interval, the HA fraction reached a value at 64.8 ± 0.3. Below 40 cm depth, the order was inverted and now the TOM fraction had a higher pMC value than that of the Macros. This can be explained by the presence of fine root particles in the sediment that could not be removed during the chemical treatment. At this depth, the HA fraction had the lowest pMC value 64.7 ± 0.3, with an offset of 1.1 relative to the Macros fraction.
The discrepancies in pMC values across fractions indicate varying proportions of redeposited ancient organic carbon. The provenance of this carbon was further characterized using C/N ratios. While TOM and Humin fractions displayed lower C/N ratios than the Macros, they still maintained a terrestrial signature. In contrast, the significantly lower C/N ratios in the HA fractions suggest that the remobilized ancient carbon may primarily derive from the decomposition of aquatic organisms (Ishiwatari Reference Ishiwatari and Aiken1985).
The sedimentary sequences analyzed provide distinct chronological records: the Jaina sediments cover approximately 1000 years, whereas the Ría Lagartos (RL) core extends back to 3855 cal BP. Differences in C/N ratios between the two sites are driven more by variations in carbon content than in nitrogen. In Jaina, carbon content ranges widely (14% to 58%), whereas in RL, it remains more stable (47% to 56%). Within the RL core, the Macros fraction exhibited the least variability in carbon content (46% to 57%), followed by HA (31% to 50%), Humin (15% to 53%) and TOM, which showed the broadest range (14% to 58%).
Conclusion
The Jaina cores are young cores that, despite being in sites with different characteristics, behave in a similar way. Based on radiocarbon dating, the Macros fraction of the Jaina sediments showed modern ages, whereas the HA fraction showed the oldest ages. The TOM and Humin fractions showed similar ages to each other. The HA fraction in Jaina reached ages of 1060–820 years for JN02 and JN03 the ages reached up to 571 years.
The RL core, even being from mangroves, showed different characteristics from those of Jaina, both in their texture and in the ages of their sedimentary layers. But the different fractions behaved in a similar way, the Macros and TOM being younger than the HA fraction. The value of the C/N ratios were used to estimate the origin of the ancient carbon contributing to the age of each fraction. In the case of Jaina, the 14C depleted values obtained for TOM, Humin and HA fractions appeared to indicate a mixture between terrestrial, aquatic and marine OM sources. This was less evident in the case of the Ría Lagartos core whose old carbon in the organic fractions retained a terrestrial footprint.
However, despite of low C/N values obtained for TOM, Humin and HA fractions from Jaina, applying OxCal Mix Curves yielded similar ages between these materials and Macros, indicating a dominantly terrestrial carbon source and a negligible local reservoir effect. Low C/N ratios failed in this case to reflect source contributions, supporting previous criticisms of their use to quantify terrestrial organic carbon origin in aquatic and sedimentary systems.
The behavior observed in the JN and RL cores reveals that the ages of the organic fractions show a lag with respect to the Macros fraction, which is not constant (Abbott and Stafford Reference Abbott and Stafford1996; Strunk et al. Reference Strunk, Olsen, Sanei, Rudra and Larsen2020). This makes it difficult to select one of the fractions as the best in the absence of macrofossils. At best, it is the TOM or Humin fraction that most closely matches the age of Macros and represents the true age of the deposit. In contrast to the study cores, the HA fraction shows an enrichment of ancient carbon. While the Jaina site reflects a mix of freshwater and marine inputs, Ría Lagartos displays a distinct terrestrial footprint.
Acknowledgments
The authors thank Arcadio Huerta, Sergio Martínez and Carlos Valencia for their technical assistance, Miguel Ángel Martínez for fruitful discussions and Nancy Y. Suárez-Mozo for the sampling work collecting cores at Ría Lagartos. This work was partially supported by grants DGAPA PAPIIT-UNAM IN112023, IA101323, IG101726 and CONAHCYT 2023-LNC-58. Guadalupe Reza Martínez thanks to CONAHCYT and Programa de Apoyo a los Estudios de Posgrado (PAEP) for financial support.
Open access
