Habitat Assessment of Bocachico (Prochilodus magdalenae) in Ciénaga de Betancí, Colombia, Using a Habitat Suitability Index Model

Abstract

This study evaluates the habitat of the Bocachico fish (Prochilodus magdalenae) in the Ciénaga de Betancí, Colombia, using a habitat suitability index (HSI) model. Wetlands like the Ciénaga de Betancí are under significant pressure from anthropogenic activities, affecting biodiversity and ecosystem health. The Bocachico, a species of immense cultural and economic importance, faces habitat degradation and fragmentation. Using hydrodynamic and water quality data, a numerical model (EFDC+ Explorer 11.5), and field data collected from multiple sampling campaigns, we assessed habitat suitability based on five key parameters: water temperature, dissolved oxygen, ammonia nitrogen, velocity, and depth. The model results indicated that environmental conditions in the wetland remained relatively stable during the dry season, with an average HSI score of 0.67, where 9% of the wetland area displayed acceptable conditions, and the remaining 91% displayed medium conditions. The wet season, on the other hand, had an average HSI score of 0.64, with 7.2% of the area in the acceptable suitability range, and the remaining 92.8% in the medium category. Variations in HSI were primarily driven by ammonia nitrogen levels, water velocity, and depth. Despite limited fluctuations in the HSI, areas of low suitability were identified, particularly in regions impacted by human activities. These findings have practical implications for conservation strategies, providing valuable insights for the sustainable management and conservation of the Ciénaga de Betancí, informing strategies for improving habitat conditions for the Bocachico, and supporting wetland restoration efforts.

1. Introduction

Wetland ecosystems are areas with permanently or seasonally inundated soils, either naturally or artificially [1,2]. Wetlands are among the world’s most productive ecosystems, providing innumerable ecosystem services such as flood control, water depuration and supply, and biodiversity reservoirs. In addition, wetlands provide valuable cultural services [2]. However, more than 35% of wetlands have been degraded worldwide due to anthropogenic actions such as those observed in agriculture, urban expansion, and polluted water spillage [3,4,5,6]. The decrease in wetlands has threatened up to 25% of their species, leading to the loss of the biodiversity associated with these ecosystems [6]. Many wetland species are fish, regarded as bioindicators and a food resource for other species, including humans. Due to their sensitivity to environmental conditions, fish are susceptible to water quality variations, just as bacteria, phytoplankton, and invertebrates. Therefore, a high abundance of fish indicates a healthy environment, whereas their scarcity is a sign of suboptimal conditions [7].

As with many wetlands around the word, the Ciénaga de Betancí is subjected to several stresses caused by anthropogenic activities, including appropriation of inundation areas for agricultural practices, construction of roads, dykes, and jetties, urban expansion, and water pollution, among others. All these changes manifest as habitat degradation for many species, causing progressive faunal and floral decrease [8,9]. Therefore, a precise characterization of the current state of the wetland is necessary to assess environmental conditions for the survival of fish species such as the Bocachico (Prochilodus magdalenae), a culturally and economically valued species in the Sinú River basin of Colombia [10,11]. Bocachico has been cataloged as a threatened species due to water quality decrease and habitat fragmentation [12].

Fish habitat management and conservation demand robust tools for measuring and predicting spatial and temporal variations. Therefore, environmental models have become a crucial technique for this purpose, as they allow us to understand the relationships of species with their environments and spatiotemporal distribution. This way, environmental models can assess ecological fluxes and identify critical conservation areas [13]. Different environmental indicators have been developed to evaluate habitat conditions. Among these is the habitat suitability index (HSI), which represents the ideal environmental conditions for the growth and reproduction of a species so it can thrive in a water body [14]. HSI values range from 0 (least suitable habitat) to 1 (most suitable habitat) [15,16]. Likewise, HSI relates environmental parameter conditions (water depth, velocity, and quality) and species responses to environmental changes, while it allows an understanding of tolerance ranges and the limiting factors of species [17,18,19].

Using numerical models for HSI assessment is critical for managing, conserving, and restoring water bodies [20,21]. It must be noted that modeling allows spatial and temporal analysis based on data measured in the study areas, guaranteeing results close to reality, backed by rigorous calibration and validation using statistical tests to ensure the reliability of the simulations [22,23]. These qualities have led to the application of modeling in a diversity of studies worldwide [16,24,25,26]. This research aims to apply the habitat suitability index (HSI) to the study of hydrodynamics and water quality trends in the Ciénaga de Betancí and how variations in these variables influence habitat conditions for the Bocachico (Prochilodus magdalenae), a species mainly from the Sinú river basin that is highly valued across different Colombian basins, not only for its ecological role, but also for its economic, cultural, and nutritional importance [10,11]. This study provides valuable information for marshy ecosystems’ integral, sustainable management and not only contributes to the conservation of local species such as the Bocachico, but also offers a framework applicable globally for wetland management and restoration. By comparing the results with similar studies in other regions, this work demonstrates how ecological modeling tools can be applied in diverse contexts, providing insights for the sustainability of aquatic ecosystems worldwide.

2. Materials and Methods

2.1. Study Area

The Ciénaga de Betancí is located in the Cordoba Department of Colombia (8°24′6.52″ N, 75°52′10.65″ W, 21 m. a. s. l) and regarded as one of the most important wetlands in the municipality of Montería [27]. The wetland lies 40 km south of Monteria, on the right margin of the Sinú River, in a hollow of the same name (Figure 1). Its basin area is 1194 km², of which approximately 16.27 km² constitute its water mirror, which increases through its connection with Ciénaga Fúnera, reaching up to 27.98 km² during the wet season [28,29,30,31].

The Ciénaga de Betancí currently works as a water reservoir by receiving inputs from precipitation, as well as from the Trementinal, Las Lomitas, and Vueltoso streams together with the El León, Betancí, and El Ñeque creeks. However, it did not always serve that purpose. In 2001, the wetland was dammed with a 14 m long dyke in the Betancí creek, downstream from the confluence of the six water streams entering the marsh. The dam kept the marsh from desiccating during the dry season, thus allowing the use of this water body for controlled fish farming throughout the year [9,32].

The wetlands complex basin has a minimum multiannual precipitation of 1206.44 mm, a maximum of 1579 mm, and an annual average of 1395.5 mm, with a wet period from early April to late October; the average temperature within the basin is 27.5 °C, the average annual evapotranspiration 1300 mm, and surface wind speed 2.5 m/s (http://dhime.ideam.gov.co/atencionciudadano/, accessed on 19 September 2022).

2.2. Field Measurements

Three measurement campaigns were performed in February, July, and August 2021 in 13 points distributed along the swamp (Figure 1). A GPS device (Garmin GPSmap 60CSx Garmin Ltd., Olathe, KS, USA) was used to record the coordinates of each sampling point. Water samples were collected at each end to perform laboratory analyses of parameters such as ammonia nitrogen, nitrates, total nitrogen, phosphates, and suspended solids. Likewise, in situ water quality measurements were taken using a HANNA HI9829 multiparameter portable meter (Hanna Instruments, Woonsocket, RI, USA). This probe allowed us to obtain data on pH, conductivity, turbidity, total dissolved solids, temperature, and dissolved oxygen. A second measurement campaign was performed in September of 2021, focusing on the main tributaries of the marsh. Depth measurements in the Ciénaga de Betancí and the stream were made using a Garmin echo-MAP 73sv sonar chartplotter (Garmin Ltd., Olathe, KS, USA), making zigzag transects along the stream and linear transects along the marsh. The obtained data were directly interpolated with the software EFDC+ Explorer 11.5 using a nearest-neighbor approach to obtain water body depth values.

2.3. Atmospheric Data

We obtained meteorological variables such as air temperature, atmospheric pressure, relative humidity, solar radiation, precipitation, evaporation, cloudiness, and wind speed and direction from the IDEAM meteorological station at Los Garzones Airport [13035501], located at 8°49′33″ N, −75°49′30″ W, through the Hydrology and Meteorology Data Management Information System (DHIME) developed by the Institute of Hydrology, Meteorology and Environmental Studies (IDEAM), in the study area.

2.4. Habitat Suitability Index (HSI)

The U.S. Fish and Wildlife Service (USFWS) describes different components for the HSI study of a species under evaluation. These include measurable physical, chemical, and biological features of the species under study that represent definable and empirically verifiable habitat characteristics, in which variables related to each component can be selected without being limited in their choice. These habitat components are (1) seasonal habitat, (2) vital requirements (food, cover, water quality, and reproductive variables), (3) life stages, and (4) cover types (areas with certain homogeneity) [33]. In this study, components 2 and 4 were chosen based on data availability, computational capacity, and economic resources. Selecting three parameters related to water quality (temperature, dissolved oxygen, and ammonia nitrogen), together with two related to hydrodynamics (velocity and depth), which, according to the literature, significantly influence the survival, growth, and development of the Bocachico (Prochilodus madgalenae). A suitability curve was constructed for each parameter using preferred values based on suitability, stress, and lethal ranges. The curves represent the relationship between the potential values that a parameter might reach in the assessed context (x-axis) and the suitability index (y-axis) ranging between 0 (non-suitable) and 1 (suitable) [34].

2.5. Numerical Model

The EFDC+ Explorer 11.5 modeling software was applied to perform the simulations. This model was developed by DSI, LLC (Edmonds, WA, USA), based on the public domain, open-source version of EFDC (Environmental Fluid Dynamics Code), initially created by the William & Mary’s School of Marine Science (SMS) at the Virginia Institute of Marine Sciences (VIMS) in 1988 [35]. The initiative was subsequently sponsored by the Environmental Protection Agency (EPA) of the United States until 2007 and became a matter of public record. EFDC+ is a coupled model for hydrodynamics, pollutant transport, and water quality kinetics in a surface water system. The integrated code is entirely written in FORTRAN, removing the need for an external coupling between hydrodynamics and transport modules [36,37,38,39]. EFDC+ simulates water quality, hydrodynamics, and habitat suitability index (HSI) in this study.

2.5.1. Hydrodynamics and Water Quality Modules

The hydrodynamic module of EFDC+ solves the Navier–Stokes equations for turbulent flux in three dimensions, in curvilinear and orthogonal coordinates. The model incorporates a second-order turbulence closure, relating turbulence viscosity and diffusivity to length scale and turbulence intensity. In addition, EFDC+ solves transport equations for length scale and turbulence intensity, as well as for salinity, temperature, suspended sediments, and for a passive tracer. The model uses a state equation relating density with pressure, salinity, temperature, and suspended sediment concentration [35].

• Momentum equation in the $x$ direction:

  $$\begin{gathered} \stackrel{1}{\overbrace{{\displaystyle \frac{\partial \left(mHu\right)}{\partial t}}}}+\stackrel{2}{\overbrace{{\displaystyle \frac{\partial \left(m_{y}Huu\right)}{\partial x}}+{\displaystyle \frac{\partial \left(m_{x}Hvu\right)}{\partial y}}+{\displaystyle \frac{\partial \left(mwu\right)}{\partial z}}}}-\stackrel{3}{\overbrace{\left(mf+v{\displaystyle \frac{\partial m_y}{\partial x}}-u{\displaystyle \frac{\partial m_x}{\partial y}}\right)Hv}} \\ =\stackrel{4}{\overbrace{-m_{y}H{\displaystyle \frac{\partial \left(g\zeta +P+P_{atm}\right)}{\partial x}}}} \\ -\stackrel{5}{\overbrace{m_y\left({\displaystyle \frac{\partial h}{\partial x}}-z{\displaystyle \frac{\partial H}{\partial x}}\right){\displaystyle \frac{\partial p}{\partial z}}+{\displaystyle \frac{\partial }{\partial x}}\left({\displaystyle \frac{m_y}{m_x}}{HA}_H{\displaystyle \frac{\partial u}{\partial x}}\right)+{\displaystyle \frac{\partial }{\partial y}}\left({\displaystyle \frac{m_x}{m_y}}{HA}_H{\displaystyle \frac{\partial u}{\partial y}}\right)+{\displaystyle \frac{\partial }{\partial z}}\left({\displaystyle \frac{m}{H}}{A}_V{\displaystyle \frac{\partial u}{\partial z}}\right)}} \\ -m{C}_p{D}_{p }u\sqrt{u^2+v^2}+S_u \end{gathered}$$

  (1)
• Momentum equation in the $y$ direction:

  $$\begin{gathered} \stackrel{1}{\overbrace{{\displaystyle \frac{\partial \left(mHv\right)}{\partial t}}}}+\stackrel{2}{\overbrace{{\displaystyle \frac{\partial \left(m_{y}Huv\right)}{\partial x}}+{\displaystyle \frac{\partial \left(m_{x}Hvv\right)}{\partial y}}+{\displaystyle \frac{\partial \left(mwv\right)}{\partial z}}}}+\stackrel{3}{\overbrace{\left(mf+v{\displaystyle \frac{\partial m_y}{\partial x}}-u{\displaystyle \frac{\partial m_x}{\partial y}}\right)Hu}} \\ =\stackrel{4}{\overbrace{-m_{x}H{\displaystyle \frac{\partial \left(g\zeta +P+P_{atm}\right)}{\partial y}}}} \\ -\stackrel{5}{\overbrace{m_x\left({\displaystyle \frac{\partial h}{\partial y}}-z{\displaystyle \frac{\partial H}{\partial y}}\right){\displaystyle \frac{\partial p}{\partial z}}+{\displaystyle \frac{\partial }{\partial x}}\left({\displaystyle \frac{m_y}{m_x}}{HA}_H{\displaystyle \frac{\partial v}{\partial x}}\right)+{\displaystyle \frac{\partial }{\partial y}}\left({\displaystyle \frac{m_x}{m_y}}{HA}_H{\displaystyle \frac{\partial v}{\partial y}}\right)+{\displaystyle \frac{\partial }{\partial z}}\left({\displaystyle \frac{m}{H}}{A}_V{\displaystyle \frac{\partial v}{\partial z}}\right)}} \\ -m{C}_p{D}_{p }v\sqrt{u^2+v^2}+S_v \end{gathered}$$

  (2)
• Momentum equation in the $z$ direction:

  $${\displaystyle \frac{\partial P}{\partial z}}=-gH{\displaystyle \frac{\rho -{\rho }_0}{{\rho }_0}}=-gHb$$

  (3)

In Equations (1) and (2), term (1) represents the rate of change in velocity concerning time, advective terms (2) represent flow movement due to inertial forces, (3) is the Coriolis parameter, (4) is pressure force, represented through free surface variation, and (5) represents viscous efforts generating turbulence within the flux.

Mass balance equation:

$$\begin{gathered} {\displaystyle \frac{\partial \left(mHS\right)}{\partial t}}+{\displaystyle \frac{\partial \left(m_{y}HuC\right)}{\partial x}}+{\displaystyle \frac{\partial \left(m_{x}HvC\right)}{\partial y}}+{\displaystyle \frac{\partial \left(mwC\right)}{\partial z}}-{\displaystyle \frac{\partial \left(m{w}_{sc}C\right)}{\partial z}} \\ ={\displaystyle \frac{\partial }{\partial x}}\left({\displaystyle \frac{m_y}{m_x}}H{A}_H{\displaystyle \frac{\partial C}{\partial x}}\right)+{\displaystyle \frac{\partial }{\partial y}}\left({\displaystyle \frac{m_x}{m_y}}H{A}_H{\displaystyle \frac{\partial C}{\partial y}}\right)+{\displaystyle \frac{\partial }{\partial z}}\left({\displaystyle \frac{m}{H}}{A}_b{\displaystyle \frac{\partial C}{\partial z}}\right)+m_x{m}_{y}H{S}_c \end{gathered}$$

(4)

where $u$ and $v$ are horizontal and vertical velocity components, in curvilinear coordinates (m/s). The component of vertical velocity (m/s) is represented by $w$; $x$ and $y$ are horizontal orthogonal curvilinear coordinates (m); $z$ is the sigma coordinate (dimensionless); $t$ is time (s); $m_x$ and $m_y$ are the square roots of metric tensor diagonal components (dimensionless); and $m=m_x{m}_y$ is the Jacobian of the metric tensor determinant (dimensionless). $H$ is the water depth (m); $P$ is the physical pressure above reference density hydrostatic pressure (m²/s²); $P_{atm}$ is the barotropic pressure, normalized by the reference water density (m²/s²); ${\rho }_0$ is the water reference density (kg/m³); and $b$ is buoyancy. The Coriolis parameter (1/s) is represented by $f$, and $A_H$ represents horizontal momentum and mass diffusivity (m²/s). $A_V$ is the viscosity of the vertical turbulent eddy (m²/s), and $A_b$ the diffusivity of the vertical turbulent eddy (m²/s). $C_p$ is the resistance coefficient of vegetation (dimensionless), and $D_p$ is the projected vegetation area, normal to the flux per horizontal area unit (dimensionless). $S_u$ and $S_v$ are the source/sink terms for the horizontal moment in the x and y directions, respectively (m²/s²), $S_h$ is source/sink term for the mass conservation equation (m³/s), and $S_c$ is the source/sink term for the constituent that includes subgrid scale horizontal diffusion and thermal sources and sinks. Water density (kg/m³) is represented by $\rho$, water temperature (°C) by $T$, and water salinity (ppt) by $S$. Lastly, $C$ is the concentration of a water quality state variable.

The last term of the mass balance equation represents kinetic processes and external loads for each water quality state variable. State equations for dissolved oxygen, ammonia nitrogen, nitrate nitrogen, and total phosphate in the aqueous phase were applied in this research. As detailed in previous research, state equations were also used for the continuity equation [40,41,42].

2.5.2. Habitat Suitability Module

Normalized habitat preference for five parameters chosen to develop the suitability index was the input for this module. Suitability curves for each parameter were defined with the x-axis as parameter measurement and the y-axis as suitability from 0 to 1 (unsuitable and ideal suitability respectively). Suitability index scores were obtained by comparing existing or predicted conditions in the study area with the relationship represented by the suitability curve, comprising a precise, detailed relationship for each parameter measurement selected [33].

2.6. Model Configuration

Two simulation periods were established when implementing the EFDC+ model for hydrodynamics and water quality to encompass dry and wet seasons. The dry season was established from 8 February to 22 March 2021, and the wet season from 20 September to 10 October 2021.

The domain area was discretized using an orthogonal curvilinear grid generated with the Curvilinear Grid Generator para EFDC (CVLGrid). The grid includes 429 elements in the x-axis direction and 232 in the y-axis direction, with a ΔX of 7.05–45.34 m and ΔY of 10.71–82.64 m, and a 1.45° of mean orthogonal deviation, resulting in 16,125 active cells (Figure 2). A Digital Elevation Model (DEM) (ALOS PALSAR satellite, resolution 12.5 × 12.5 m) was coupled with bathymetric field data taken at the Betancí stream and marsh to process topo-bathymetry data in the domain area grid. The simulation used dynamic time increase so that EFDC+ automatically calculated Δt values in each iteration, ensuring numerical stability by constantly determining the Courant number [39,41].

2.6.1. Initial Conditions

At the initial moment of the simulation, the conditions of physicochemical parameters were defined from field and laboratory data. A Kriging interpolation was applied in those areas without data using ArcGIS 10.5. In addition, initial water elevation in the Ciénaga de Betancí was determined according to the values taken at the Maracayo limnimetric station [13067040] located at 8°24′36″ N, −75°53′24″ W.

2.6.2. Boundary Conditions and Forcings

Boundary conditions are associated with water influx and outflux. An open boundary was established in the Betancí stream, west of the marsh. Five additional boundaries were established at each of the main tributaries (the Las Lomitas, Trementinal, and Vueltoso streams; the El León, Betancí, and El Ñeque creeks), which resulted from fluvial discharge into the marsh (Figure 3). Water level data for the Ciénaga de Betancí were obtained from the Maracayo limnimetric station [13067040] using DHIME, and water level data for the Betancí stream were taken from the Betancí limnimetric station, located at 8°30′0″ N, −75°57′36″ W. The main meteorological forcing data such as air temperature, wind velocity and direction, solar radiation, precipitation, evaporation, and atmospheric pressure were obtained from the IDEAM meteorological station at Los Garzones Airport [13035501].

2.7. Numeric Model Calibration

The numeric model was calibrated by gradually modifying the coefficients and rates of chemical reactions, driving the behavior of physicochemical parameters according to the time series generated in field campaigns during the dry and wet seasons.

Goodness of Fit Tests

Two tests, root mean squared error (RMSE) and mean absolute error (MAE), were performed to assess the goodness of fit for simulated data to in situ measurements. RMSE yields only positive values; a zero value (rarely obtained) indicates a perfect fit to the data, values close to zero represent an ideal fit, and values far from zero indicate a deficient fit to the data [31].

$$\mathrm{R}\mathrm{M}\mathrm{S}\mathrm{E}=\sqrt{{\displaystyle \frac{1}{n}}{\sum }_n^1(\mathrm{s}\mathrm{i}\mathrm{m}\mathrm{u}\mathrm{l}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{d}-\mathrm{o}\mathrm{b}\mathrm{s}\mathrm{e}\mathrm{r}\mathrm{v}\mathrm{e}\mathrm{d}{)}^2}$$

(5)

MAE is the average of the absolute values of all differences between simulated and observed values [43].

$$\mathrm{M}\mathrm{A}\mathrm{E}={\displaystyle \frac{1}{n}}{\sum }_n^1\left|\mathrm{s}\mathrm{i}\mathrm{m}\mathrm{u}\mathrm{l}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{d}-\mathrm{o}\mathrm{b}\mathrm{s}\mathrm{e}\mathrm{r}\mathrm{v}\mathrm{e}\mathrm{d}\right|$$

(6)

2.8. Calculation of Average Suitability Index

Components of water quality (temperature (T), dissolved oxygen (DO), ammonia nitrogen (NH₄), and hydrodynamics (velocity and depth)) were unified to obtain a single habitat suitability index (HSI) result encompassing both climate seasons. This process was performed by calculating the arithmetic mean of the EFDC+ Explorer 11.5 data along the period simulated for each season. Components were grouped using ArcGis 10.5, obtaining an HSI thematic map. To interpret the HSI range (0–1), a classification was developed (

Table 1. HSI score categorization.

HSI Range	Interpretation
0	Unsuitable conditions
0.01–0.25	Inadequate suitability conditions
0.25–0.5	Low suitability conditions
0.5–0.75	Medium suitability conditions
0.75–0.99	Acceptable suitability conditions
1	Ideal conditions

) based on the suitability categories proposed by previous studies [44,45,46].

3. Results and Discussion

3.1. Suitability Curve Values

Preferred values for the Bocachico for the five selected parameters were taken from the literature and previous studies conducted in the region (

Table 2. Preference values for the Bocachico species (Prochilodus magdalenae) for water quality and hydrodynamic components.

Parameter	References	Optimal Range	Stress Range	Lethal Range
Temperature (°C)	[47,48,49]	26–30	<22	>37
Dissolved oxygen (mg/L)	[50]	5–CSat *	1.5–5	<1.5–>CSat *
Ammonia (mg/L)	[50,51,52]	0–0.1	0.1–0.3	>1
Bathymetry or water depth (m)	[50].	1–1.5	0.5–1	2–2.5
Velocity (m s−1)	[53]	0.10–0.25	<0.10 0.25–0.50	0.50–1

Note: * Saturation concentration of dissolved oxygen in the water body.

).

A normalized value was assigned to each parameter (

Table 3. Normalization preference parameters for the Bocachico on a scale from 0 to 1.

Suitability	Parameters
	OD (mg/L)	T (°C)	NH4 (mg/L)	Depth (m)	Velocity (m s−1)
0.0	>CSat *	37.0–36.4	>1.0	0.0	0.49–0.47
	0.0–0.2			4.8–5.0
0.1	0.3–0.7	36.3–35.7	1.0	0.1	0.46–0.45
				4.7–4.4
0.2	0.8–1.3	35.6–35.0	0.9	0.2	0.44–0.42
				4.3–4.0
0.3	1.4–1.8	34.9–34.3	0.8	0.3	0.41–0.40
				3.9–3.7
0.4	1.9–2.3	34.2–33.6	0.7	0.4	0.39–0.38
				3.6–3.3
0.5	2.4–2.8	33.5–32.9	0.6	0.5	0.37–0.35
		21.9–22.6		3.2–3.0	0.01–0.02
0.6	2.9–3.4	32.8–33.2	0.5	0.6	0.34–0.33
		22.7–23.4		2.9–2.6	0.03–0.04
0.7	3.5–3.9	32.1–31.5	0.4	0.7	0.32–0.30
		23.5–24.3		2.5–2.0	0.05–0.06
0.8	4.0–4.4	31.4–30.8	0.3	0.8	0.29–0.28
		24.5–25.1		1.9–1.8	0.07–0.08
0.9	4.5–4.9	30.7–30.1	0.2	0.9	0.27–0.26
		25.2–25.9		1.7–1.6	0.09
1.0	5.0–Csat *	26.0–30.0	0.0–0.1	1.0–1.5	0.10–0.25

Note: * Saturation concentration of dissolved oxygen in the water body.

), fitting the habitat suitability scale between 0 (non-suitable conditions) and 1 (suitable conditions) to generate suitability curves for each selected parameter.

Data normalization followed the equations made in

Table 4. Equations for normalization of Bocachico preference parameters of water quality and hydrodynamics.

Parameter	Equation
HSI OD (mg/L)	$0.19\ast OD+0.05\dots if 0.05\le OD\left(\mathrm{m}\mathrm{g}/\mathrm{L}\right)<5.0$
	$1\dots if 5.0\le OD\left(\mathrm{m}\mathrm{g}/\mathrm{L}\right)\le C_{sat}$
	$0\dots if OD\left(\mathrm{m}\mathrm{g}/\mathrm{L}\right)>C_{sat}$
HSI T (°C)	$0.122\ast T-2.1707\dots if 21.9\le T\left(°\mathrm{C}\right)<26.0$
	$1\dots if 26.0\le T\left(°\mathrm{C}\right)\le 30.0$
	$-0.1429\ast T+5.2857\dots if 30.0<T\left(°\mathrm{C}\right)\le 37.0$
HSI NH4 (mg/L)	$1\dots if 0.0\le N{H}_4\left(\mathrm{m}\mathrm{g}/\mathrm{L}\right)\le 0.1$
	$-N{H}_4+1.1\dots if 0.1<N{H}_4\left(\mathrm{m}\mathrm{g}/\mathrm{L}\right)<1.1$
	$0\dots if 1.1\le N{H}_4\left(\mathrm{m}\mathrm{g}/\mathrm{L}\right)<2.0$
HSI Depth (m)	$1\ast Depth\dots if 0.0\le Depth \left(\mathrm{m}\right)<1.0$
	$1\dots if 1.0\le Depth \left(\mathrm{m}\right)\le 1.5$
	$-0.5\ast Depth+1.75\dots if 1.5<Depth \left(\mathrm{m}\right)\le 2.0$
	$-0.06\ast Depth+0.87\dots if 2.0<Depth\left(\mathrm{m}\right)\le 2.5$
	$-0.28\ast Depth+1.4201\dots if 2.5<Depth\left(\mathrm{m}\right)\le 5.0$
HSI Velocity (m/s)	$50\ast Vel+2\ast {10}^{-16}\dots if 0.001\le Vel\left(\mathrm{m}/\mathrm{s}\right)\le 0.01$
	$5.5556\ast Vel+0.4444\dots if 0.01<Vel\left(\mathrm{m}/\mathrm{s}\right)<0.1$
	$1\dots if 0.1\le Vel\left(\mathrm{m}/\mathrm{s}\right)\le 0.25$
	$-4.1667\ast Vel+2.0417\dots if 0.25<Vel (\mathrm{m}/\mathrm{s})\le 0.49$

, based on concentration values for parameters of interest from Table 3.

Based on the normalized data of Table 3, habitat suitability index (HSI) curves were generated, as shown in Figure 4. These curves were entered into the habitat analysis module of EFDC+ to obtain the spatial distribution of HSI for the Ciénaga de Betancí.

3.2. Calibration of the Hydrodynamic Module

Hydrodynamic calibration of the Ciénaga de Betancí was performed by comparing measured simulated data to water level data, reported by the Maracayo limnimetric station and expressed as a time series (Figure 5). The red points represent water elevation, as observed in the Maracayo station, and the blue line represents simulated water levels.

The model fits the main current traits of the marsh, which are verified with the values obtained in each statistic test. Values of root mean squared error (RMSE) and mean absolute values (MAE) both show an optimal fit (0.001). Overall, the results of hydrodynamic calibration indicate a satisfactory fit between the measured and modeled data [42,54].

3.3. Nutrient Transport Calibration Module

The transport module was calibrated using a time series of water quality data for each parameter obtained during February and March 2021 at sampling point BET1 (Figure 1). The calibration results show a good fit of the model to the measured data for all the analyzed parameters (Figure 6). Statistical errors for each parameter are summarized in

Table 5. Statistical errors for the calibration of the water quality module.

Statistic/Parameter	NH4 (mg/L)	NO3 (mg/L)	PO4 (mg/L)	TN (mg/L)	TSS (mg/L)
RMSE	0.045	0.026	0.011	0.089	6.993
MAE	0.039	0.022	0.007	0.081	5.084

.

3.4. Simulation Results for Ciénaga de Betancí

Hydrodynamics and water quality analyses were performed, and results were obtained at 20 simulation days. In addition, habitat suitability indexes were determined for the Ciénaga de Betancí for both the dry and wet seasons.

3.4.1. Hydrodynamics in the Ciénaga de Betancí

For both climate seasons, the highest velocities, ranging between 0.04 and 1 m s⁻¹, were recorded in the Betancí stream and the zone connecting with the Ciénaga de Fúnera (Figure 7A). For the rest of the marsh, prevailing velocity values were lower than 0.01 m s⁻¹. In the dry season, the direction of water circulation in the marsh was west–east, with an average velocity of 0.08 m s⁻¹. This condition was associated with wind velocities on the surface of the marsh. On the other side, during the wet season, the water flowed into the Betancí stream at an average velocity of 0.09 m s⁻¹; this condition was influenced by water flux from the streams to the marsh.

Low velocities can be associated with barriers (dams or dykes) in the marsh, which reduce water quality for rheophile fish. Low water velocity may cause ecosystem productivity to decrease due to a lesser energy contribution, thus affecting reproduction and survival. In addition, low velocity affects fish population dynamics and restricts habitat availability, impacting their populations and overall ecosystem health [55,56,57]. Regarding water quality, low velocity may lead to increasing nutrient concentration, turbidity, and chlorophyll levels, as well as expanded distribution and mixing processes [58,59,60].

Significant depth changes, reaching up to 1.5 m (Figure 7B), were observed between seasons. Depth variations during the dry season were caused by water evaporation, as the marsh did not exhibit water input or output. In contrast, depth increased in the wet season due to tributary contributions. Maximum depths were reached when the water level was as high as the dam. The overflow discharged into the Betancí stream and connected to the Sinú River [30,61].

Depth might limit ecosystem productivity due to the lower penetration, thus affecting the global primary productivity of the system [62]. Greater depths might decrease water quality, provoking turbidity and pH changes. Likewise, deeper water bodies may become stratified, affecting the distribution of rheophile fish, which tend to move into more suitable zones, driven by their physiological requirements [63,64].

3.4.2. Water Quality Results

During the dry season, the temperature in the Ciénaga de Betancí exhibited slight variation, with prevailing values above 28 °C, except for the mouth of the Betancí stream and the zone near the Las Lomitas stream. In the wet season, temperatures remained below 27.8 °C (Figure 8A). Recorded temperatures in the marsh remained within an optimal range of 26 to 30 °C, thus increasing the chances of successful adaptation and growth for aquatic fauna [65,66]. These results indicate positive effects on water quality and ecosystem productivity, favoring organic matter decay and water oxygen dissolving, which is vital for aquatic organism respiration [67]. Productivity increases due to efficient energy use, affecting species abundance and distribution [68].

According to Figure 8B, dissolved oxygen (DO) did not exhibit variations within the marsh in the dry season. For the wet season, DO concentrations did not feature significant changes in the marsh, excepting minor variations shown at the mouth of the Las Lomitas stream. It is worth emphasizing that despite the slight variations for each climate season, DO concentrations remained within optimal levels, above 6 mg/L. These dissolved oxygen concentrations benefit aquatic organisms’ growth, survival, and resource use efficiency, preventing asphyxiation and mortality due to oxygen scarcity [69,70]. Optimal DO levels support ecosystem health, a key factor for aquatic life and overall water quality [71].

In the dry season, ammonia NH4 nitrogen concentration (Figure 8C) was uniformly distributed, with values close to 0.32 mg/L. In the wet season, ammonia increased by 0.15 mg/L in most of the eastern zone, where the Trementinal stream and El Ñeque creek run into the marsh. Lower concentrations (0.12 mg/L) were recorded in the northern zone during the dry season. Ammonia nitrogen levels may indicate water quality deterioration due to agriculture, cattle ranching, or domestic pollution surrounding water bodies. This pollution can lead to eutrophication and significant changes in wetland structure and function [72]. The presence of NH4 may alter ecosystem productivity and induce algal blooms, biodiversity loss, and ecosystem function damage [73].

3.4.3. Habitat Suitability Index of Hydrodinamic Parameters

For the dry season, HSI analysis was carried out between 8 February and 8 March 2021; for the wet season, it was performed from 20 September to 10 October 2021.

The HSI values of water velocity were between 0.25 and 0.1, for low suitability and unsuitable, respectively (Figure 9A); these conditions were caused by low water velocity above the water body, at approximately between 0 and 0.096 m s⁻¹ in the dry season. Water velocity magnitude and direction were influenced by meteorological variables, bathymetry, and changes in the natural dynamics of the Ciénaga de Betancí, which has been transformed into a water reservoir [30], with the wind being the variable with the most significant effect on HSI values. Water velocity values in the wetland decrease water quality and are considered slow for the Bocachico, a species that depends on currents for movement and foraging [74,75]. In the wet season, water velocity was strongly influenced by tributary flows from the Betancí stream, Trementinal stream, and the El Ñeque creek, with scores ranging between 0.54 and 1, from middle to optimal suitability conditions. Input led to higher velocities up to 0.6 m s⁻¹ within the water mirror; however, to a considerable extent, the central zone of the water body featured scores less or equal to 0.5, even reaching values close to 0 (unsuitable) at specific times. Increasing velocities positively affect the Bocachico habitat by maintaining oxygenation and preventing sedimentation, helping to preserve the overall ecological equilibrium of the marsh [56].

HSI values for depth (Figure 9B) ranged from 0.1 to 0.3 (low suitability to unsuitable) in the central zone of the Ciénaga de Betancí. The average depth of the water body was 3.5 m, outside the range of suitability, with an HSI value of 0.43, within the low suitability category. The pronounced depth of Betancí may affect flux dynamics, reducing flow speed in particular zones, which leads to sediment infill on the wetland [9]. This phenomenon generates low suitability conditions for rheophile fish, reducing food availability and significantly affecting growth rates for the species [76,77]. Acceptable HSI conditions (0.7 to 0.8) were found in the Fúnera marsh, featuring depths between 1.8 and 2.3 m. This depth regime enhanced primary production, increasing the functional surface of habitat available for fish species and improving overall suitability for aquatic organisms. Under these conditions, spawning is favored, and food resource availability is guaranteed [78,79,80]. The habitat suitability index decreased from one climatic season to another due to the increase in water level (approximately 1.46 m). Most of the water mirror presented suitability scores of 0 (unsuitable) in the wet season, Changes in depth were minimal and were related to the flow and sediment input from the tributaries, and excess water was directed towards the Sinú River [30]. These conditions may cause changes in nutrient availability, decreasing primary production and reducing the Bocachico population [81].

3.4.4. Habitat Suitability of Water Quality Parameters

In terms of temperature, the Ciénaga de Betancí remained suitable in both climate seasons (Figure 10A). In the dry season, the temperature HSI remained at a constant value of 1 (suitable conditions) throughout the study period, as temperature fluctuated between 27 °C and 28 °C, and in the wet season, the HSI values were close to 1, with temperatures between 25.6 and 27.6° C. This parameter, which marks the rhythm of physiological processes in fish, did not present sudden changes that could cause physiological or behavioral alterations, including food and energy balances for the Bocachico, given that it is a sensitive species, less able to withstand changes in temperature [82,83,84].

The HSI of DO was 1 in dry season (Figure 10B), with concentrations of 5 mg/L in the Cienaga de Betanci, and the saturation values were between 96% and 99.6%. These values indicate that conditions on the water body’s surface allow fish growth and reproduction at their maximum potential [50,85,86]. In the wet season, the HSI of DO kept a value of 1 (7.2 to 7.8 mg/L) throughout the simulated time interval. These conditions reveal a positive effect for the Bocachico in both climate seasons [74,87]. If saturation decreases, it would cause physiological and behavioral changes, thus affecting fish growth rates and would increase vulnerability to disease [88,89].

Ammonia nitrogen HSI in the dry season is shown in Figure 10C, with values ranging between 0.7 (medium suitability) and 0.87 (acceptable suitability), and with concentrations varying between 0.4 mg/L and 0.23 mg/L, respectively. In the wet season, HSI remained between 0.61 (medium suitability conditions) and 1 (ideal conditions), with concentrations varying between 0.49 mg/L and 0–0.1 mg/L, respectively. Wet conditions lead to increased runoff and tributary input, transporting substances originating in anthropic activities such as agriculture and cattle ranching [90,91,92,93]. The primary agents increasing NH₄ concentration are animal residues, fertilizers, and pesticides [30,94,95]. Unsuitable ammonia conditions may affect rheophile fish’s growth, physiology, and immune response [96,97].

3.4.5. Average HSI of the Ciénaga de Betancí

The average suitability index for the Bocachico (Prochilodus magdalenae) in the Ciénaga de Betancí (Figure 11) shows minimal interseasonal changes. HSI variations were established in well-defined areas, such as the mouth of El Vueltoso stream and part of El León Creek, increasing by 0.02. Conversely, suitability degraded in the remaining tributaries, with HSI decreasing by 0.02 to 0.04. The most significant variation was observed in the Fúnera wetland, where HSI values lowered by 0.08 to 0.2, going from acceptable to medium. Likewise, the central zone of the marsh experienced a slight decay of approximately 0.05. A significant improvement was observed in the Betancí stream, where most of its extent shifted from medium to acceptable suitability conditions.

Zones classified as suitable and unsuitable are presented in Figure 11. The maximum, minimum, and average HSI values in the dry season were 0.84, 0.53, and 0.67, respectively. The total water body area in the dry season was 19.3 km². About 9% of this area (1.73 km²) displayed acceptable conditions, with scores between 0.75 and 0.84, whereas the remaining 91% (17.57 km²) was in the middle HSI category, ranging from 0.5 to 0.75. During the wet season, the maximum, minimum, and average HSI values were 0.94, 0.5, and 0.64, respectively. The total water body area was 23 km² in the wet season, of which 7.2% (1.66 km²) featured HSI values in the acceptable suitability range. In contrast, the remaining 92.8% (21.33 km²) was classified in the category of medium suitability.

Considering that arithmetic mean is influenced by each data value used, the parameters with a more significant effect on average suitability in the Ciénaga de Betanci were ammonia nitrogen, velocity, and depth. Low suitability scores in these parameters indicated a decreasing trend in average suitability. In particular, the use of hydrodynamic models to predict habitat suitability has proven effective in identifying critical areas for the conservation of endangered species in these ecosystems.

4. Conclusions

To determine the habitat suitability index (HSI) for the Bocachico (Prochilodus magdalenae), a hydrodynamics and water quality model was built using the software EFDC+. The model was based on ecological preference data for the species in terms of five parameters affecting its growth and development—deficiencies in one parameter limit the average HSI. Likewise, the combination of hydrodynamic parameters was deemed proper for the initial assessment. Physical and chemical characteristics are vital in defining habitat preferences, particularly in the study area and its environmental problems.

The spatiotemporal simulations in the Ciénaga de Betancí indicate minimal variation in HSI between dry and wet seasons, with only a 1.8% difference in habitat suitability. The greatest variation occurred in the eastern zone, influenced by the runoff and flow discharge area of the Trementinal and El Ñeque streams. The parameters showing the most significant seasonal changes were ammonia nitrogen, water velocity, and depth, primarily due to human activities within the wetland.

Modeling the hydrodynamics and the water quality of the Ciénaga de Betancí (Colombia) provides a solid basis for environmental assessment by environmental authorities. In this way, the model allows an understanding of the state of the marsh, allowing a detailed analysis of hydrodynamic and water quality processes at different points of the water body at various locations and times. The information generated by the model is a critical tool for assessing suitability and marsh management for fishing and other uses.

The results of this study are not only relevant for the local management of the Ciénaga de Betancí but also provide a solid scientific foundation that can be replicated in other wetland ecosystems. The implications for environmental policy-making are clear: the use of models like the HSI can guide decision-making to protect aquatic species and restore degraded ecosystems. Environmental authorities should consider implementing these models in their environmental impact assessments and restoration plans.