Ecohydrogeochemical functioning of coastal freshwater herbaceous wetlands in the Protected Natural Area, Ciénaga del Fuerte (American tropics): Spatiotemporal behaviour.

Coastal zones are characterized by the interactions between continents and oceans and, therefore, between fresh and salt surface and groundwater. The wetlands of coastal zones represent transitional ecosystems that are affected by these conditions, although little is known about the hydrogeochemistry of wetlands, especially coastal wetlands. In the present study, the hydrogeochemical characterization of coastal freshwater herbaceous wetlands in the Ciénaga del Fuerte Protected Natural Area in Veracruz, Mexico, in the American tropics was carried out per plant community. Four herbaceous wetlands (alligator flag, saw grass, cattail, and floodplain pasture) were monitored to understand the origin of the water feeding these ecosystems, the hydrogeochemical composition of groundwater, and the relationship between the groundwater and ecology of these ecosystems during dry and rainy seasons. The results indicate that Ciénaga del Fuerte is located in a regional discharge area and receives local recharge, so it is fed by both regional and local flows. The chemical composition varied temporally and spatially, creating unique conditions that determined the habitat occupied by the hydrophytic vegetation. The spatiotemporal behaviour of groundwater is one factor that, along with the hydroperiod, determines wetland dynamics and affects wetland biota (ecohydrogeochemistry). Generalist plant communities established in zones of local recharge, whereas other more specialized and/or plastic communities inhabited zones receiving regional flows with greater ion concentrations. This information forms the basis for establishing an appropriate scale (municipal, state, or larger regions) for the sustainable management of goods and services provided by the wetlands.

INTRODUCTION

Coastal zones are highly dynamic and productive areas where the continents, oceans, and atmosphere interact. Diverse ecosystems are present in these zones, including beaches, dunes, and wetlands such as reefs, mangroves, floodplain forests, coastal lagoons, estuaries, and herbaceous wetlands (e.g., tulares, carrizales, and popales). Evidence of the dynamic nature of coastal zones includes erosion processes, which affect beaches and dunes, and meteorological phenomena, such as hurricanes and torrential rainfall, which affect extensive coastal areas. Additionally, coastal zones serve as discharge areas for groundwater and surface water originating from the continent. However, coastal regions are also vulnerable to climate change and rising sea levels and phreatic levels as well as the marine intrusion and/or soil salinization (Spalding et al., 2014). These factors can lead to the loss and/or migration of wetlands, habitat fragmentation, and the reduction or ecosystem services loss (Intergovernmental Panel on Climate Change, 2014; Nicholls & Cazenave, 2010).

Wetlands, particularly coastal wetlands, are important ecosystems because of the large quantity and variety of ecosystem services that they provide to society (Costanza et al., 1997). As transitional ecosystems between terrestrial and aquatic environments (Mitsch & Gosselink, 2015), wetlands are strongly influenced by hydrology (Gusyev & Haitjema, 2011). Additionally, wetlands are often located in discharge areas and therefore have an important contribution of groundwater (Winter, 1999), which is crucial for nutrient transport and for wetland salinity (Jolly et al., 2008). Both of them (nutrient and salinity) and water have ecological effects, influencing the presence or type of vegetation that establishes, for example (Morris, 1995). Finally, in coastal wetlands, all these factors should consider interactions with both groundwater and marine water too (Qu et al., 2017).

Traditionally, ecological studies in wetlands have been carried out to characterize vegetation composition, biodiversity, water quality, and its effects on wetlands dynamics. Water studies in wetlands tend to be limited to surface water (Kors et al., 2012) and, occasionally, to interstitial water (water present at the root level, Weterbach et al., 2016), without understanding the role of groundwater in wetlands, its chemical composition, or the hydrogeochemical processes influencing these ecosystems (Hunt, Krabbenhoft, & Anderson, 1997; Liu & Mou, 2016). With respect to hydrogeochemical studies in coastal areas, this have been carried out to understand the water quality of aquifers for human use (Chidambaram et al., 2018; Lee & Song, 2007) and, more recently, to identify the causes of salinization (salt intrusion, geological processes, or anthropic contamination, Böhlke & Denver, 1995; Lee & Song, 2007; Bouzurra, Bouhlila, Elango, Slama, & Ouslati, 2017). Some hydrogeochemical studies have been focused to understand dynamics between groundwater and surface water to reduce anthropic effects but without an ecological approximation (Ladouche & Weng, 2005). Liu and Mou (2016) described some interactions between groundwater–surface water and wetlands and the necessity of a new approach to study this ecosystem. Few hydrogeochemical studies have been carried out in tropical coastal wetlands, so the hydrogeology of these ecosystems is largely unknown, including the origin and evolution of the water that feeds them (Hunt et al., 1997; Carol, Mas‐Pla, & Kruse, 2013). In a temperate climate (Scotland), Malcolm and Soulsby (2001) performed the hydrogeochemical characterization of a coastal aquifer associated with an interdunal wetland complex to understand its biodiversity and its capacity to maintain the quality of fresh water despite its location and land‐use change. House and Sorensen (2015) characterized a riparian wetland in the United Kingdom using a model that incorporated temperature and botanical indicators (hydrophytes species present in sampling sites) in order to determine the dynamics between groundwater and surface water. In Latinamerica, Carol et al. (2013) characterized the wetlands in the Bay of Samboronbón, Argentina, with the objective of establishing criteria for the conservation of their water resources, and Yetter (2004) characterized the hydrogeochemistry, the origin, and the quantity of water that is feeding mangroves and herbaceous wetlands in the Ramsar Site La Mancha in Mexico.

Generally, these authors conclude that information is lacking on groundwater (quality and quantity), one of the main flows to wetlands, and that such information is crucial for the adequate management and conservation of these ecosystems and their goods and services, especially due to the population growth in coastal areas. It is estimated that 50% of the world population is living within 100 km of the coast (Small & Nicholls, 2003) and that 10% of the population in coastal areas is located at an elevation of lower than 10 masl (Spalding et al., 2014). In the state of Veracruz, for example, 27% of the population lives less than 20 km from the coast (Mendoza‐González, Martínez, Lithgow, Pérez‐Maqueo, & Simonin, 2012), which is a concern, considering that this Mexican state is one of the most vulnerable to climate change (Monterroso et al., 2014) and to rising sea levels, with reported increases of 1.9 mm year−1 (Zavala‐Hidalgo, de Buen Kalman, Romero‐Centeno, & Hernández Maguey, 2010).

The present study presents a hydrogeochemical characterization of coastal freshwater herbaceous wetlands in the Protected Natural Area (PNA) Ciénaga del Fuerte (American tropics). The objective was to understand the hydrogeochemistry of the wetland in the Ciénaga del Fuerte PNA and the origin of the water feeding this wetland in addition to how these latter factors could influence the plant communities. Ultimately, this information would provide a basis for understanding the local and regional functioning of these wetlands. Additionally, this key information is the first step for quantifying ecosystem services and for generating policies to conserve wetland ecosystems through sustainable management.

METHODOLOGY

Study area

Ciénaga del Fuerte PNA

The Ciénaga del Fuerte PNA is a state reserve of 4,269 ha located in the municipality of Tecolutla, Veracruz, in the touristic region of Costa Esmeralda, on the Gulf of Mexico (Figure 1). The climate is subhumid warm, with rainfall in summer. The annual average temperature is 27.5°C, and the total annual rainfall is 1,490.8 mm (Coordinación Estatal de Medio Ambiente, 2002, Figure 2). It is located in the hydrographic watershed of the Tecolutla River, which belongs to the hydrological region of Tuxpan‐Nautla (RH‐27). It has an area of 7,446 km2 and is fed by four rivers originating in the Northern Sierra of Puebla: the Necaxa, Lajajalpan, Tecuantepec, and Apulco Rivers (from north to south; Osuna‐Osuna et al., 2015).

Figure 1

Location of the Ciénaga del Fuerte Protected Natural Area in the municipality of Tecolutla, Veracruz, Mexico, and monitoring sites location

Figure 2

Climatogram of the study area (1980–2010). Data from meteorological station no. 030153 in San Rafael, Veracruz, CONAGUA

Geology of the study area

The study area is located near the regional discharge area of the Western Sierra Madre (WSM). This latter mountain range is composed of numerous geological units of marine origin that are strongly folded due to orogenic processes, forming cavities with differing degrees of competence (Moran‐Ramírez et al., 2018) as well as regional fractures and faults, which increase the hydraulic conductivity of these materials (Morán‐Ramírez, Ramos‐Leal, López‐Álvarez, Carranco‐Lozada, & Santacruz‐De León, 2013).

The underlying Pimienta Formation of the Jurassic period it is composed of lime mudstone and wackestone and black or dark grey clay–limestone intercalated with thin layers of calcareous or carbonate shale, with a maximum width of approximately 600 m (Cantú‐Chapa, 1971; Heim, 1926). Over this latter unit, the undifferentiated Tamaulipas Formation was deposited during the Middle Cretaceous. It is composed of lime mudstone and wackestone with some intercalations of shale and loam, with a thickness of 400 m (Stephenson, 1922). The Agua Nueva Formation of the Upper Cretaceous covers this latter unit. It is formed by lime mudstone and wackestone with intercalations of shale, with a thickness of 127 m (Stephenson, 1922). Over this latter formation, the Méndez Formation from the Upper Cretaceous is found. It is composed of shale and loam alternated with bentonitic shale, with a varying thickness of 100 to 1,000 m (Jeffreys, 1910). The Cretaceous formations are covered by the Chicontepec Formation of the Tertiary period. It is constituted by an alternation of clay sandstone with siltstone, sandy loam, and grey shale, with a varying thickness of 1,500 to 3,300 m (Dumble, 1918). Additional geological units composed of clay–sand materials cover the Chicontepec Formation (Figure 3). The sediments in the coastal plain in the Gulf of Mexico are composed of clays of which Cruz‐Orozco, Machado Navarro, Alba Cornejo, and Téllez Ortíz (1987) identified: montmorillonite 32% to 50%, illite from 20% to 34%, kaolinite from 19% to 34%, and chlorite with the lowest values from 10% to 29%. According to vegetation, Campos et al. (2011) detailed major soil properties by depth in different vegetation communities from Ciénaga del Fuerte (Table 1). The volcanic rocks that have been reported in the region correspond to rhyolites with quartz, potassium feldspar, oligoclase, biotite, amphibole, and pyroxene, as well as dacites with plagioclase (oligoclase, albite) with a higher content of potassium and amphiboles (SGM, 2004).

Figure 3

Geology of the study area and location of the regional geological section of the Western Sierra Madre (shown in Figure 4) in Tecolutla, Veracruz

Table 1

Soil properties in a popal‐carrizal (Thalia geniculata and Cyperus giganteus) reported by Campos et al. (2011)

Depth (cm)	Layer type	Texture class
0–10	Organic	Clay
10–20	Organic	Clay
20–42	Organic	Clay
42–76	Organic	Clay
76–105	Organic	Clay

Figure 4

Regional transversal section of geological characteristics and groundwater flow from the Western Sierra Madre to the coastal plain of the Gulf of Mexico

With respect to the hydrogeological functioning (Figure 4), the recharge area of the WSM experiences high rainfall and humidity, favouring the infiltration of water. Once water infiltrates, it flows towards regions of lower hydraulic head according to Darcy's law or, in this case, towards the Gulf of Mexico. Along this route, groundwater interacts with different geological materials and becomes enriched with ions, resulting in the modification of its hydrogeochemical signature until its capture and/or arrival at the discharge area. Because of the stratification of the region, multiple sedimentary units of low hydraulic conductivity are present. Water contained in the sedimentary units of the Cretaceous period is confined to aquifers and can only ascend to the surface or mix with more local flows through faults or fractures (Figure 4). Ciénaga del Fuerte is located near the coast, in the discharge zone, within a regional hydrogeological system. In this context, the recharge occurs mainly in the limestone of the mountainous area in the Sierra Madre Oriental. These sedimentary rocks are covered by other less permeable clay formations, giving it the character of a confined aquifers. On the other hand, on the Cretaceous sedimentary rocks, there are volcanic rocks with a certain permeability that collect the local recharge along the Gulf Slope.

Vegetation

The PNA contains a wetland complex with different types of tropical wetlands. There are different herbaceous freshwater wetlands, and these are locally known according to the dominant species: popales (broadleaf species, i.e., Thalia geniculata), tulares (vegetation with long and thin leaves, e.g., Typha domingensis and Cladium jamaicense), carrizales (vegetation with cylindrical stems and thin and/or modified leaves, e.g., Cyperus giganteus and Phragmites australis) as well as native floodplain pastures (Leersia hexandra and Leersia oryzoides with creepers, e.g., Ipomea tilacea and Ipomea Indica) and floodplain pastures (introduced grass species mixed with native species, such as Lippia nodiflora, Coordinación Estatal de Medio Ambiente, 2002). In addition, there are several patches of floodplain forest containing different aquatic and forest species, such as Pachira aquatica, Zygia latifolia, Diospyrus digyna, Attalea butiracea, Pithecellobium recorddii, and Inga vera. The PNA is surrounded by Valencian orange orchards and cattle ranching (Infante, Moreno‐Casasola, Madero‐Vega, Castillo‐Campos, & Warner, 2011; Sánchez‐Luna, 2018).

Monitoring sites

The sites with the greatest presence of hydrophytic species in Ciénaga del Fuerte were selected for monitoring. In total, four sites representative of each herbaceous wetland community were selected (Figure 1): one popal, two tular‐carrizal communities with different species composition located in distinct sites, and one floodplain pasture (this last site was previously a wetland, Figure 5).

Figure 5

Monitoring sites of the Ciénaga del Fuerte Protected Natural Area labelled according to their dominant vegetation: (a) Thalia, (b) Cladium‐Typha‐Cyperus, (c) Typha‐Cladium, and (d) floodplain pasture

In each site, three linear 100‐m transects were established parallel to one another and separated by a distance of 50 m, except in the floodplain pasture, where transects were separated by larger distances because of the water flow and vegetation (Figure 5d). The transects were placed in the direction of the water flow. A nest of standpipe piezometers was placed at the extreme ends of each transect and was constructed according to Peralta, Infante, and Moreno‐Casasola (2009). Piezometers reached a depth of up to 6 m except in the site Typha‐Cladium with 4 m depth. To characterize dominant vegetation, in each 1 m × 1 m quadrant, the per cent cover and height of each species were monitored.

Hydrogeochemical analysis

In the four monitoring sites, a total of 96 groundwater samples were taken directly from the piezometers in the rainy season (October 2016) and the dry season (April 2017), corresponding with four Samples × Transect × Site × Season (4 × 3 × 4 × 2 = 96 samples). The laboratory analyses were performed according to the methods established in Welch et al. (1996) and Apha (2005). In each site, physical parameters (temperature, electric conductivity, Total Dissolved Solids (TDS), pH, salinity, dissolved oxygen, and oxidation‐reduction potential (ORP)) were measured in situ using a YSI 556© multiparametric probe. All groundwater samples were collected in high‐density polyethylene bottles for major ion analysis; prior to use, all bottles were triple washed with abundant deionized water.

The water samples for cation and metal analyses were immediately acidified after being taken with ultrapure nitric acid until reaching a pH < 2. All samples were conserved at 4°C. The principal ions were analysed in the Ecological Engineering and Wetland Biogeochemistry Laboratory (Laboratorio de Ingeniería Ecológica y Biogeoquímica de Humedales) of the Institute of Ecology using ion chromatography (Dionex Chromeleon ICS‐1100). The anions were analysed using an eluent of 4.5‐mM Na3CO3/0.8‐mM NaHCO3 in an Anion Self‐Regenerating Suppressor (ASRS 300 2 mm) in Auto Suppression Recycle Mode at a flow rate of 0.25 ml min−1, a temperature of 30°C, an applied current of 7 mA, and an injection volume of 5 μl. The cations were analysed using an eluent of 20‐mM methanesulfonic acid via suppressed conductivity detection in a Cation Self‐Regenerating Suppressor (CSRS® Ultra II, 2 mm) in Auto Suppression Recycle Mode at a flow rate of 0.25 ml min−1, an ambient temperature, an applied current of 15 mA, and an injection volume of 2.5 ml. The alkalinity was quantified in situ by the titration method (Association of Official Analytical Chemists, 1980). All chemical analyses had an ionic balance lower than 10%, which was considered acceptable.

To characterize the dominant vegetation, in each quadrant, per cent cover and height of each species were monitored. Per cent cover was calculated using the Westhoff, and van deer Maarel (1978) cover‐abundance scale.

The resulting hydrogeochemical data were used to generate Piper diagrams to determine the water families and descriptive statistics and are presented in Table 2; Mifflin diagrams were used to define the flow systems; Gibbs diagrams helped to identify the influence of evaporation processes, water–rock interactions, and rainfall; and biplots were used to identify the occurrence of other processes such as ionic exchange, mixing, and the overall evolution of groundwater.

Table 2

Descriptive statistics of ion data per group and season (dry and rainy)

Group	Season		Cl	SO4	HCO3	Na	K	Mg	Ca
		(mEq L−1)
1	Rainy	Min	0.63	0.037	1.229	0.161	0.076	0.358	1.13
		Max	0.254	0.115	2.049	0.46	0.084	0.46	1.463
		M	0.126	0.072	1.639	0.269	0.08	0.42	1.31
		SD	0.087	0.035	0.335	0.131	0.003	0.045	0.138
	Dry	Min	0.299	0.023	1.639	0.37	0.081	0.477	0.378
		Max	2.141	0.258	3.688	5.923	0.113	0.614	1.381
		M	1.012	0.103	2.54	2.145	0.095	0.507	0.854
		SD	0.865	0.098	0.788	2.465	0.013	0.059	0.455
2	Rainy	Min	0	0	3.278	1.411	0.079	0.873	0
		Max	3.71	3.017	12.701	12.55	0.144	5.472	2.535
		M	2.104	0.647	7.221	5.877	0.106	2.225	1.488
		SD	1.363	1.066	3.277	3.511	0.027	1.734	0.925
	Dry	Min	0.242	0.096	1.639	0.419	0.077	0	0
		Max	5.587	5.044	3.278	13.03	0.249	1.967	1.467
		M	2.412	1.262	2.123	4.109	0.114	0.958	0.636
		SD	1.707	1.599	0.478	3.726	0.057	0.573	0.563

RESULTS AND DISCUSSION

The first site is located near the centre of the PNA. It is a popal wetland dominated by T. geniculata (commonly known as alligator flag) and L. hexandra (southern cut grass) containing one to five species; it was labelled as Thalia (Figure 5a). The second site is located near the outer limit of the PNA and is surrounded by agricultural fields. It is a tular community dominated by C. jamaicense (saw grass) followed by T. domingensis (cattail) and C. giganteus (giant flatsedge or Mexican papyrus) containing one to six species; it was labelled as Cladium‐Typha‐Cyperus (Figure 5b). Notably, this site presented the highest flood levels. The third site is located near the centre of the PNA in an area surrounded by hills dedicated to livestock ranching. It is also a tular community dominated by T. domingensis followed by C. jamaicense containing two to four species; it was labelled as Typha‐Cladium (Figure 5c). The fourth and final site is located in an area of extensive livestock ranching adjacent to the PNA. It is dominated by Lippia nodiflora (tangle fogfruit) but contains one to 10 species of diverse introduced and native grasses; it was labelled as floodplain pasture (Figure 5d).

Hydrogeochemistry

Hydrochemical facies

In the Piper diagram (Figure 6), five main water types were identified: 40% of the samples corresponded with CaHCO3 type (Type II), 20% with NaCl type (Type I), 27% with NaCaHCO3 type (Type III), 12% with NaHCO3 type (Type VI), and only one sample with CaMgCl type (Type IV).

Figure 6

Piper diagram of the hydrochemical facies in the Ciénaga del Fuerte Protected Natural Area

The samples in the rainy season (October) were mainly distributed in Types II, III, and VI (CaHCO3, NaCaHCO3, and NaHCO3, respectively). The samples in the dry season (April) were mainly concentrated in Types I and II (NaCl and CaHCO3, respectively) and some in Type III (NaCaHCO3).

Spatially, two groups can be observed. Group 1 includes the sites of the floodplain pasture and Cladium‐Typha‐Cyperus, and Group 2 includes the sites of Thalia and Typha‐Cladium. Overall, these groups present temporal behaviour. During the rainy season, all Group 1 samples were Type II (CaHCO3), corresponding with recently infiltrated rainwater. The Group 2 samples were mainly Type III (NaCaHCO3) and Type VI (NaHCO3). During the dry season, the Group 1 samples did not vary considerably with respect to the rainy season, but some samples did correspond with Type III (NaCaHCO3), coinciding with the increased temperatures in this season. Meanwhile, Group 2 differed the most between seasons. During the dry season, the large majority of its samples were Type 1 (NaCl). A few samples were Type II (CaHCO3), and a couple of samples were Type III (NaCaHCO3) and Type VI (NaHCO3). Therefore, Group 2 is characterized by the presence of more evolved water samples.

Origin of groundwater and water–rock interactions

The evolution of groundwater is related to its physicochemical content, due to the interaction of the medium through which it circulates. It is also a function of residence time and distance travelled; in such a way, that recent infiltration waters have low concentrations of their physicochemical parameters, whereas more evolved waters have a higher concentration of these components (Tóth, 1999). Therefore, groundwater will have less dissolved solids in the recharge zone. These solids will increase as water circulates. This type of relationship was addressed by Mifflin (1968), which related the content of Na + K vs Cl + SO4 with flow systems (small local, local, and regional) and was corroborated with tritium.

The SO4 + Cl versus Na + K relationship in the Mifflin diagram (Figure 7a) enabled the water samples to be characterized into three flow types: local, intermediate, and regional. Local and regional flows were dominant. Only a few samples were indicative of intermediate flow and, possibly, mixing between local and regional flows.

Figure 7

(a) Mifflin diagram showing the types of groundwater flow and (b) Gibbs diagram showing the dominant processes affecting water chemistry in the Ciénaga del Fuerte Protected Natural Area

Spatially, the samples are separated into two groups. Group 1 is composed of samples from the floodplain pasture and Cladium‐Typha‐Cyperus sites. These samples are concentrated near the origin, corresponding with the lowest SO4 + Cl− and Na + K values. Accordingly, these samples are characteristic of local recharge or rainwater. Group 2 is associated with samples from the Thalia and Typha‐Cladium sites. These samples tend to correspond with more evolved waters and regional flow.

Meanwhile, the Cl/(HCO3 + Cl) versus TDS relationship is presented in the Gibbs diagram (Figure 7b). The graph is indicative of the major processes controlling groundwater chemistry: evaporation and crystallization, rock dominance, and atmospheric precipitation. As observed in most of the samples, evaporation processes are most influential followed by water–rock interactions.

Overall, during the rainy season, both groups present a greater concentration of TDS compared with the dry season. This may be due to the transportation of sediments and dissolved contaminants in water originating from anthropogenic activities, such as land‐use change. Both groups presented a lower concentration of chlorides in the rainy season versus the dry season. The increase in chlorides during the dry season is more notable in Group 2 and is associated with evaporation as a result of high temperatures.

The diagram of the HCO3 + SO4 versus Ca + Mg relationship is presented in Figure 8a to discern whether groundwater circulates in carbonate rocks or rocks rich in silicates. Most samples (both dates) fall in the field of silicate alteration, whereas only a few correspond with carbonate dissolution, which is likely related with the previously described geological context. The Group 1 samples present low dispersion independently of season. Meanwhile, the Group 2 samples present higher dispersion: In the month of October, alteration by silicates appears to be more influential, whereas during the dry season, interaction with carbonates is evidenced.

Figure 8

Water–rock interaction diagrams based on the relationships of (a) HCO3 + SO4 versus Ca + Mg, (b) Ca/Na versus HCO3/Na, and (c) Na/Ca versus Mg/Ca

The Ca/Na versus HCO3/Na relationship is presented in Figure 8b. This graph has been used to understand the interactions between water and the geological medium through which it circulates (water–rock interactions). In the figure, evolution curves initially representing local recharge or rainwater were identified (I and II). The first curve (I) corresponds with water that has had a greater interaction with volcanic rock, leading to a higher alteration by silicates. The second curve (II) also corresponds with rainwater that has interacted with volcanic rock and has undergone silicate alteration, although it additionally shows evidence of evaporative processes. The evaporative processes observed in curves I and II are consistent with the characteristics of Ciénaga del Fuerte: Evapotranspiration occurs in shallow water bodies, and the aquifer has a depth of only 4 m.

In the month of April, the samples are once again divided into two groups: One group contains the samples from the floodplain pasture and Cladium‐Typha sites, and a second group contains the samples from the Thalia and Typha‐Cladium sites. During the rainy season, the Group 1 samples are located in the field of meteoric water, whereas the Group 2 samples show greater evolution and interaction with rocks rich in silicates. However, during the month of April, the Group 1 samples show little interaction with silicate rocks (as rhyolite and dacite), whereas the Group 2 samples present greater evolution and interaction with silicates and indicate evaporation, which is consistent with the high temperatures during this month.

The Na/Ca versus Mg/Ca relationship is presented in Figure 8c. This graph lets us to identify the interaction between groundwater and rocks rich in silicates (as rhyolite and dacite), limestone, and dolostone as well as mixtures with sea water/saltwater intrusion and irrigation return. Most samples are located in Quadrants 1, 3, and 4 (indicating the influence marine and weathering silicates). The groups are similarly separated. Overall, in both October and April, the Group 1 samples are mainly distributed in the limestone and silicate zone and indicate little evaporation. The Group 2 samples show interaction with silicates and the influence of evaporative processes, irrigation return, and/or evolution. This finding is congruent given the location of the study area in a discharge area near citrus orchards and the temperatures of the region.

The relationship between the cation chloride ratio index and Cl/ (alkanility + Cl) is shown in Figure 9a. Two groundwater evolution curves can be identified: The first begins with Ca + Mg>> Na + K but shifts to Ca + Mg = Na + K as water evolves. Some samples begin at Ca + Mg << Na + K and tend to show similar values as water evolves, initiating with ionic exchange and subsequent alteration of silicates and evapotranspiration as water evolves. Temporally, in the month of October, the Group 1 samples tend to group on the left and show characteristics of rainwater that has undergone little evolution. These samples show greater vertical distribution and, correspondingly, little variation in HCO3 and Cl. The Group 2 samples are distributed towards the right and are associated with more evolved waters, evapotranspiration processes, and ionic exchange with the clay material present in the soil. In April, the Group 1 samples are slightly dispersed towards the right, showing some influence from evaporation but maintaining their vertical distribution. Meanwhile, the Ca, Mg, and Na values of the Group 2 samples are similar and show a horizontal dispersion associated with evapotranspiration and silicate alteration.

Figure 9

Diagrams of the relationships between (a) Cl and residual alkalinity, (b) the CCR index and Cl/ (alkalinity + Cl), and (c) Cl and Na. CCR, cation chloride ratio

The Cl versus residual alkalinity relationship is shown in Figure 9b. The samples are divided into two regions. The upper region indicates greater presence of bicarbonate ions and ionic exchange. The lower region indicates a lower concentration of HCO3 than Ca–Mg as well as greater evapotranspiration. Two evolution curves can be observed, with values of residual alkalinity starting near zero. A grouping can also be observed near the right side of the graph, indicating an increase in chlorides and, accordingly, greater evolution and evapotranspiration. Temporally, the residual alkalinity is positive in October and negative in April (dry season), which is congruent with the temperatures during this latter month (25°C to 27°C). On both dates, Groups 1 and 2 are separated. Group 1 tends to have a residual alkalinity near zero, whereas Group 2 has a positive or negative residual alkalinity depending on the season (rainy or dry, respectively).

The Cl versus Na relationship is shown in Figure 9c. This relationship can be used to identify processes of ionic exchange and alteration by albite or processes of reversible ionic exchange. The Group 1 samples tend to be concentrated near the origin and have low Na and Cl values, whereas the Group 2 samples show higher concentrations of both ions. Notably, all samples are located in the region of ion exchange and albite alteration, which is consistent with the presence of clay materials and rocks rich in silicates.

Ecological implications

Wetlands distribution

The local recharge sites (Group 1) are mostly stable with respect to ion concentrations, which are low in both seasons. In contrast, regional recharge sites (Group 2) show slightly diluted ion concentrations in the rainy season, so some water mixing processes likely occur during this season. These differences can be visualized on species distribution (Table 3).

Table 3

Wetland types and dominant vegetation by per groundwater flow system

Groundwater flow system	Group	Monitoring sites	Dominant plant species per site
Local flow	1	Floodplain pasture	Lippia nodiflora‐native and introduced grass
		Cladium‐Typha‐Cyperus	Cladium jamaicense‐Typha domingensis‐Cyperus giganteus
Regional flow	2	Thalia	Thalia geniculata‐Leersia hexandra
		Typha‐Cladium	Typha domingensis‐Cladium jamaicense

Wetlands in local recharge areas

The groundwater samples from the floodplain pasture and Cladium‐Typha‐Cyperus (Group 1) showed characteristics of rainwater and varied little between seasons (rainy and dry). So the sites of this group are likely present in the local recharge area despite being located near the coastline and the outer border of the PNA (Figure 5b,d).

The floodplain pasture is the site with the highest diversity of plant species because it has the lowest level of flooding, which favours the presence of both hydrophytic and terrestrial species. Also, it has lower ion concentrations, which create more favourable conditions for most species. This site experienced disturbance 20 years ago and is therefore regenerating and is subjected to the constant perturbation by grazing cattle. On the other hand, the Cladium‐Typha‐Cyperus site had the highest level of flooding of all sites, even in the dry season. Its dominant species was C. jamaicense, even though T. domingensis is reportedly one of the most tolerant species to flooding in terms of flood depth (Chen & Vaughan, 2014) and time exposed to flooding (Cronk & Fennessy, 2001) because it possesses several mechanisms that enable it to tolerate prolonged periods of flooding (Armstrong, Justin, Beckett, & Lythe, 1991; Colmer, 2003; Cronk & Fennessy, 2001; Voesenek & Bailey‐Serres, 2015). Interestingly, these adaptions can signify that low levels of flooding (−10 cm) are stressful for this species (unpublished results). Meanwhile, C. jamaicense is reportedly more sensitive to flooding (Newman, Grace, & Koebel, 1996) and to water with high phosphorus concentrations. For these reasons, it was displaced by T. domingensis in the Everglades as this environment transformed from oligotrophic to eutrophic (Davis, 1991). In the present study, the dominance of C. jamaicense in the most flooded site may be due to its higher sensitivity to other factors that are not present as the major ion concentration. Accordingly, C. jamaicense may take advantage of this opportunity niche despite the flood level favouring T. domingensis. Species present in Group 1 can be considered more generalist and are fed with water characteristics of rainwater and inhabit environments that are favourable for the establishment of a greater number of species and/or environments that may be considered oligotrophic.

Wetlands located in regional discharge zones

The sites of Group 2 (Thalia and Typha‐Cladium) presented both flooded and saturated conditions. Based on the chemical composition of groundwater, these sites are fed by regional flows. These sites have communities of hydrophytic plants that can possibly tolerate higher ion concentrations and/or the double stress of reduced water levels and increased salt concentrations as a result of evapotranspiration during the dry season.

T. geniculata can reportedly respond to changes in the hydroperiod. Under drought conditions, it responds by increasing its number of leaves (unpublished results). One assumption may be that T. geniculata is more sensitive to ions than to variation in the flood level. Meanwhile, T. domingensis was present under saturated conditions, which complies with its basic requirements (nonstress conditions) although is more specialized to major floods (Inoue & Tsuchiya, 2006; unpublished results), and it likely dominates the Typha‐Cladium sites because of its greater tolerance to the higher ion concentrations in regional discharge compared with C. jamaicense. Overall, greater plasticity is observed in the Group 2 species, which are mainly represented by T. geniculata and T. domingensis. These latter species occupy more specialized habitats compared with the Group 1 species (considered more generalist).

Implications for management

The Ciénaga del Fuerte PNA is a discharge area fed by groundwater from local and regional flows, indicating that some of the wetlands of the PNA are fed by water from regional flows whereas others are fed by local recharge or rainwater. This indicate that PNA Ciénaga del Fuerte requires management at the local level (municipality) as well as at the regional level, including federal entities and their municipalities with jurisdiction over regional recharge area in WSM. This information is an indication of the necessity to conserve and recover forested mountain zones in recharge of watersheds to secure the provision of freshwater to communities in lowlands and to maintain the environmental flow to wetlands and coastal wetlands.

The water chemistry of Ciénaga del Fuerte PNA is influenced by its location near the coastline and extensive citrus orchards. In fact, the area outlying the PNA is part of the most important region for the production of Valencian orange in the country (Sistema de Información Agroalimentaria y Pesquera, 2018). Accordingly, a large portion of the water samples were taken near the zone of saltwater intrusion and/or irrigation return, especially those in Group 2, so if salinization is due to irrigation return, strategies targeting the sustainable development and management of agriculture should be implemented to maintain the quantity and quality of water inflow (whether local or regional in origin) in Ciénaga del Fuerte. If salinization is due to salt intrusion, then it is even more important to maintain the quality and quantity of local and regional flows to minimize or stop it. Further studies are necessary to distinguish which of these latter two processes is most influential for the water chemistry of the PNA.

Currently, there are important efforts to restore the forest high up in the mountains. This work is giving accurate information about a necessity to take care of these recharge areas to conserve wetlands coast.

The flow characteristics of Ciénaga del Fuerte demonstrated that future studies and management should consider the surface characteristics (vegetation and surface hydrology) and the underground characteristics (especially water chemistry) of wetlands in order to understand their functioning and plant communities as well as the origin of water.

CONCLUSIONS

The Ciénaga del Fuerte PNA is located in a regional discharge area and, based on the Mifflin hydrogeochemical diagram, it is mainly fed by local and regional flows.

Two groups of groundwater were identified in the Gibbs diagram. Group 1 was associated with the floodplain pasture and Cladium‐Typha‐Cyperus sites, and Group 2 was associated with the Thalia and Typha‐Cladium sites. The Group 1 samples had characteristics similar to those of local recharge or rainwater, whereas the Group 2 samples appeared to be more evolved or originate from regional flows.

The groundwater of Ciénaga del Fuerte was more influenced by silicate alteration than by carbonate dissolution.

The chemical signatures of the groundwater of Ciénaga del Fuerte were influenced by interaction with the rocks through which groundwater circulated. Spatiotemporally, the characteristics of the Group 1 samples are more similar to those of local recharge, whereas those of the Group 2 samples evidence the occurrence of water–rock interaction and evaporation.

The spatiotemporal behaviour of the water samples shows that groundwater underwent evapotranspiration, possibly as a result of saltwater intrusion and/or irrigation return from agricultural activities.

Groundwater in Ciénaga del Fuerte first undergoes processes of ion exchange and, as it evolves, evapotranspiration occurs. This is congruent with the region's climate and the clay content in wetland soil.

The residual alkalinity showed that, as groundwater begins to evolve, spatiotemporal variation is near zero. However, in the rainy season, the Group 2 samples had positive residual alkalinity, which then turned negative in the dry season. This indicates that the hydrophytic communities associated with Group 2 likely exhibit plasticity to variation in the chemical components of water. The hydrophytic communities associated with Group 1 are more restricted in their habitat to sites with lower salt concentration.

The chemical composition of groundwater in Ciénaga del Fuerte creates unique conditions that determine the areas or habitats occupied by hydrophytic species.

The spatiotemporal behaviour of groundwater chemistry is one factor that, along with the hydroperiod, determines the dynamics and functioning of wetlands (eco‐hydrogeoperiod).

The characterization and spatiotemporal behaviour of the hydrogeochemistry of wetlands are critical for understanding the origin of water feeding wetlands and the interaction between different water flows and wetland ecosystems and their dynamics. Ultimately, this information will provide a better understanding of the influence of water chemistry on the ecology of wetlands, over their floristic composition and to propose management plans including discharges areas (wetland zones) and recharges areas (upstream).