2.8. Salinity

This section of the report covers salinity in the lower St. Johns River, and is organized as follows:

  • 8.1 — Overview: Introduction and background of salinity in the river
  • 8.2 — Salinity Models: Models (EFDC-3D and ADCIRC-2D) currently being used on the lower St. Johns River and the (qualitative and quantitative) differences between the models and how they are being used
  • Basis material dealing with salinity in the river — in lieu of reading these sub-sections in full, the below cruxes provide concise statements about each feature of river salinity:
    • 8.3 — Depth Variation of Salinity: Depth variation of salinity by analysis of depth-dependent time-series data including episodic storm events, including assessment of stratification and causal nature
      • Crux: Salinity in the lower St. Johns River is generally well-mixed and not stratified (vertically layered)—astronomic tides with spring tidal range above 2 m constitute the primary factor responsible for the large mixing of salinity in the river.
    • 8.4 — Longitudinal Salinity Variation: Longitudinal salinity variations using depth-integrated time-series data, including assessment of episodic storm events.
      • Crux: Salinity in the lower St. Johns River experiences variations along the longitudinal axis of the river—astronomic tides cause a periodic to-and-fro of salinity along the river’s length while wind-driven storm surges can pulse saltwater up the river and watershed-based rainfall-runoff can pulse freshwater down the river.
    • New material dealing with salinity in the river — in lieu of reading these sub-sections in full, the below cruxes provide concise statements about each feature of river salinity:
      • 8.5 — Long-Term Trend Analysis: Daily-based regression analysis was performed on salinity records (1996–2007) for Dames Point (river km 20) and Acosta Bridge (river km 40)
        • Crux: Salinity in the lower St. Johns River has been on the rise over the past decades, as diagnosed from long-term data alone—of significant importance, some acceleration of salinity rise is evident in the historical record.
      • 8.6 — Climate Change-Impact Assessment: Salinity was simulated for the lower St. Johns River based on freshwater river inflow, offshore water levels and sea-level rise.
        • Crux: Rising salinity in the lower St. Johns River is primarily being driven by sea-level rise, whereby more seawater can encroach further up the river—and more so critical, salinity rise in the river is accelerating.
      • 8.7 — Biological Impacts: The biological impacts of salinity
      • 8.8 — Overall Assessment: Ratings of status and trend of salinity in the river

2.8.1. Overview

Salinity is a measure of the saltiness of a mass of water. As an estuary, the lower St. Johns River experiences variable salinity with more saline waters downstream and more fresh waters upstream. Furthermore, salinity has an impact on water quality and biota in the lower St. Johns River.

The lower St. Johns River can be broken down into three ecological zones based on the salinity regimes experienced (Figure 2.1; Hendrickson and Konwinski 1998; Malecki, et al. 2004):

  • Mesohaline

River km 0-40 (from Mayport Inlet to Downtown Jacksonville/Fuller Warren Bridge)
Narrower and deeper waters, well-mixed with average salinity of 14.5 parts per thousand and fast flow rate

  • Oligohaline

River km 40-75 (from Downtown Jacksonville/Fuller Warren Bridge to Doctors Lake)
Broader and shallower waters, slow-moving and tidally active with average salinity of 2.9 parts per thousand

  • Freshwater Lacustrine

River km 75-200 (from Doctors Lake to Lake George)
Lake-like with weaker tides and average salinity of 0.5 parts per thousand

Salinity in the lower St. Johns River is affected by tides, seasonal rainfall patterns and episodic storm and drought events.  The tides are predictable by the astronomic (ocean) and estuarine (river) tide.  The seasonal pattern of rainfall-derived freshwater input to the lower St. Johns River is predictable, with a majority of the rainfall occurring in the wet season from June to October (Rao, et al. 1989).  Episodic events are less predictable and include hurricanes, tropical storms and (more frequently) nor’easters as well as droughts, like the droughts of the early 1970s, the early 1980s, 1989-1990 and 1999-2001 (DEP 2010d). Storm events can cause surges of coastal waters to propagate up the lower St. Johns River causing a 1-2-day spike in salinity followed by a dramatic reduction in salinity because of the lagged input of freshwater rainfall runoff from the watershed basin.  Salinity increases during period of droughts because of limited freshwater rainfall-runoff input.

Storm events need not necessarily be local in order to drive storm surges and salinity spikes in the lower St. Johns River.  Although non-tidal effects in river flows and salinity can be correlated with wind direction, the principal physical mechanism is not direct surface stress by winds over the river, but rather the response of ocean water level on the adjacent shelf that then forces the flow and salinity in the river.  In short, the lower St. Johns River is primarily affected by remote winds and is secondarily affected by local winds (Bacopoulos et al. 2009). Low frequency, synoptic-scale ocean water level variability is at least as important a factor as storm events in causing distinct pulses of salinity in the river.  Synoptic-scale events have 3- to 12-day time scales and are much more frequent than hurricanes and tropical storms.

Figure 2.65
Figure 2.66 Map of the Ecological (Salinity) Zones of the Lower St. Johns River.

2.8.2. Salinity Models

This sub-section covers models currently being used on the lower St. Johns River and the differences between the models and how they are being used.

As a preface, models are used in science to mimic and study physical processes. For example, Newton’s law governs gravitationally forced motion, like an apple falling from a tree, which can be modeled on a computer to provide simulations for various physical scenarios, like if the object were more or less massive, if the object were dropped from more or less height, etc. The models presented in this sub-section are used to mimic the physical processes of hydrodynamics and salinity. Models are developed, then applied for a specific case (this is called ‘calibration’) and applied again for a different case (this is called ‘validation’). The performance of the model to simulate the physical process(es) is measured in the calibration and validation procedures. There is no binary (yes or no) outcome from the calibration and validation procedures; rather, quantitative measures of performance are used to support the extent that a model has been successfully calibrated and validated. The model cases presented below are from two models that are widely accepted within the coastal and river modeling community and that have undergone extensive development, calibration and validation.

There are two models covered in this sub-section:

  • EFDC — Environmental Fluid Dynamics Code

Used by agencies (United States Army Corps of Engineers; St. Johns River Water Management District)
Used in private industry—examples cited are from Taylor Engineering, Inc

  • ADCIRC — ADvanced CIRCulation

Used in the academic arena — examples cited are from University of North Florida and University of Central Florida

2.8.2.1. EFDC — Environmental Fluid Dynamics Code

The Environmental Fluid Dynamics Code (EFDC) (Hamrick 1992) is a three-dimensional hydrodynamic model capable of simulating changes in water level, velocity, discharge, salinity and water age (a measure of flushing rate) due to changes in inflows from multiple sources and locations. Some of the model features that make EFDC attractive for salinity modeling in the lower St. Johns River include:

  • Advection-Diffusion

Allows for simulation of salinity and water age

  • Surface Wind Stress

Allows for wind forcing to be applied in the model

  • Two-Dimensional Flows

Allows for simulation of horizontal flows and circulation in lakes

  • Three-Dimensional Flows

Allows for simulation of return flows generated by wind setup in lakes

  • Dynamic Coupling of Salinity and Density

Allows for simulation of density stratification and the subsequent baroclinic, estuarine circulation

Further reference information on EFDC can be found online (EPA 2014).

EFDC is used by the United States Army Corps of Engineers (Liu, et al. 2013), the St. Johns River Water Management District (Sucsy and Morris IV 2001; SJRWMD 2012a) and in private industry, e.g., Taylor Engineering, Inc. (Liu, et al. 2013).

The St. Johns River Water Management District continues to employ EFDC, for example, for total maximum daily loadings (Sucsy and Morris IV 2001) and water supply impact study (SJRWMD 2012a) (Figure 2.67). The EFDC hydrodynamic model grid for the lower St. Johns River includes Lake George and Crescent Lake, the main river stem and a portion of the offshore domain.  The salt marshes north and south of the lower St. Johns River near the inlet are included in the model grid as water storage areas.

 

Figure 2.66
Figure 2.67 EFDC Hydrodynamic Model Grid (SJRWMD 2012a).

Resolution of the grid cells is generally 0.5-1 km with finer resolution going down to 50-100 m to define the finer features of the lower St. Johns River, like the model grid spanning four grid cells across the narrow river channel through Downtown Jacksonville (Figure 2.68).

Figure 2.67
Figure 2.68 EFDC Hydrodynamic Model Grid – Zoom of Downtown Jacksonville (SJRWMD 2012a).

Calibration of EFDC for the lower St. Johns River included specifying or adjusting the following data and input parameters (Sucsy and Morris IV 2001):

  • Bottom Bathymetry and Bottom Roughness
  • Tidal Water Level at the Open-Ocean Boundary
  • Salinity at the Open-Ocean Boundary
  • Number of Vertical Layers in the Model
  • Non-Reflective Upstream Open Boundary

The main calibration parameter for EFDC was bottom roughness, which is the usual case for hydrodynamic and salinity modeling.  EFDC calibration was performed for 1997–1999 and EFDC validation was performed for 1996–2005.  Model performance measures were computed for water levels, discharges and salinity simulated by EFDC.  Water levels were simulated by EFDC to within on-average 95% (r2) of observed data and discharges and salinity were simulated by EFDC to within on-average 85% (r2) of observed data. The EFDC model validation proved capable of simulating salinity variations from Mayport Inlet (river km 0) to the entrance of Lake George (river km 190) (Figure 2.69).

One conclusion from the salinity sub-study in the water supply impact study of the lower St. Johns River (SJRWMD 2012a) was stated as (page 5-308) “although the model shows a realistic response to observed salinity at widely spaced locations, there is a paucity of data for confirming the model’s dynamic simulation of salinity at tidal scales,” which becomes further apparent in the later sub-section(s) of this section.

Figure 2.69
Figure 2.69 Salinity Variations Simulated by EFDC Compared to Observed Data for the Lower 190 km of the St. Johns River (SJRWMD 2012a)

2.8.2.2. ADCIRC – ADvanced CIRCulation

The ADvanced CIRCulation (ADCIRC) code (Luettich, et al. 1992) is a two-dimensional hydrodynamic model capable of simulating water levels and depth-integrated velocities for tide, wind and freshwater river-influx forcing scenarios.  Recent developments of ADCIRC (Kubatko, et al. 2006) have given the model the capability to simulate salinity.  Some of the model features that make ADCIRC attractive for salinity modeling in the lower St. Johns River include:

  • Advection-Diffusion – Allows for simulation of salinity
  • Surface Wind Stress – Allows for wind forcing to be applied in the model
  • Unstructured Triangulation – Allows for flexible model-domain definition
  • Two-Dimensional Flows – Allows for simulation of depth-integrated velocities and flows

Further reference information on ADCIRC can be found online (UNC 2014).

ADCIRC is used in the academic area, like at the University of North Florida and the University of Central Florida.

An ADCIRC model has been developed, calibrated and validated for the lower St. Johns River (Bacopoulos, et al. 2012).  The ADCIRC model mesh includes a telescopic view into the lower St. Johns River from the large-scale western North Atlantic Ocean, Caribbean Sea and Gulf of Mexico (Hagen, et al. 2006) (Figure 2.70).  The ADCIRC model mesh represents the lower St. Johns River up to and including Lake George, the salt marshes north and south of the lower St. Johns River near the inlet and a localized offshore zone outside of the inlet (Figure 2.71).  Resolution of the mesh elements ranges from hundreds of meters for the main river stem, or even greater in element size for the offshore zone, to tens of meters for tidal creeks, narrow channels and other fine features of the lower St. Johns River.

Figure 2.70
Figure 2.70 ADCIRC Model Mesh telescoping from the western north Atlantic Ocean, Caribbean Sea, and Gulf of Mexico into the Lower St. Johns River (Hagen, et al. 2006).

ADCIRC calibration was performed for 1995-1997, with bottom roughness as the main calibration parameter used (Bacopoulos, et al. 2012), and ADCIRC validation was performed for May, June and July 2009.  Model performance measures were computed for water levels, discharges and daily discharges simulated by ADCIRC.  Discharges were simulated by ADCIRC to within on-average 20% error (RMS) relative to observed data, water levels were simulated by ADCIRC to within on-average 15% error (RMS) relative to observed and daily discharges were simulated by ADCIRC to within on-average 30% error (RMS).  These errors may seem large upon initial review; however, consider that these errors are with regards to the model’s ability to simulate to complete hydrodynamic signal, including tide-, runoff-, wind- and remote-driven processes, as measured against comparable observations that include the full hydrodynamic signal.

Figure 2.71
Figure 2.71 ADCIRC model mesh localized for the Lower St. Johns River (Bacopoulos, et al. 2012) with four salinity monitoring stations.

Additional simulations were performed for the lower St. Johns River for three one-month periods:

  • June 13 – July 13, 1999
    Referred to as ‘High Extreme’ because the salinity regime in the river was relatively saline for the 30-day period
  • September 21 – October 21, 1999
    Referred to as ‘Most Variable’ because the salinity regime in the river was most longitudinally varied for the 30-day period
  • October 30 – November 29, 1999
    Referred to as ‘Low Extreme’ because the salinity regime in the river was relatively fresh for the 30-day period

The simulations for the three one-month periods in 1999 proved the ADCIRC model’s capability to simulate salinity variations at Dames Point (river km 20) (Figure 2.72) in the mesohaline portion of the river.  This level of performance demonstrates the ADCIRC model’s ability to recreate short-term events, like the reduction in salinity during June 20 – July 4, 1999, the rise and spike in salinity during October 12–19, 1999 and the fluctuations in salinity during November 13-20, 1999.  The model results shown clarify how the ADCIRC model incorporates tide, wind and freshwater river influx as driving forces for the simulation of salinity response in the lower St. Johns River.

Figure 2.72
Figure 2.72 ADCIRC-Modeled salinity versus observations at Dames Point for three one-month periods in 1999.

As a disclaimer, the ADCIRC model simulations for the three one-month periods in 1999 are still in development.  There are model updates, including numerical code updates, calibration and validation still to be done on the ADCIRC simulations for modeling salinity in the lower St. Johns River; however, even as initial results, these ADCIRC model results nonetheless capture the tidal and episodic non-tidal variations in salinity for the lower St. Johns River, as subjected to driving forces of tides, winds and freshwater river influxes.  A fuller presentation of the ADCIRC model simulations of salinity in the lower St. Johns River is presented in Bacopoulos, et al. 2017.

In conclusion, the Environmental Fluid Dynamics Code (EFDC) and the ADvanced CIRCulation (ADCIRC) code were summarized in this sub-section on salinity models.  Salinity fluctuations were shown to be captured by the EFDC model, namely concerning the variability of salinity fluctuations along the 190-km stretch of the lower St. Johns River (Figure 2.69).  Salinity fluctuations were shown to be captured by the ADCIRC model, namely concerning the tidal and short-term (i.e., episodic) salinity fluctuations (Figure 2.72).

2.8.2.3. EFDC (3D) vs. ADCIRC (2D)

It might be useful to clarify which model (EFDC vs. ADCIRC) works better for the different salinity zones of the lower St. Johns River.  This effort will require care, and it will be carried out and shown in the coming years’ reports.  Nonetheless, to begin a comparison of the two models requires the separation into qualitative and quantitative aspects.  In a qualitative sense, the EFDC model can describe depth-variable flow and salinity, but to accomplish this, the horizontal resolution is constrained, whereas the ADCIRC model can flexibly resolve the horizontal features of the geometry and physics, but it solves two-dimensional equations and thusly cannot simulate depth-variable physics.  In this regard, there is a tradeoff between the two models, in general. However, for the St. Johns River, which is a micro- to meso-tidal estuary (with spring tidal ranges greater than 2 m; see Carr-Betts, et al. 2017 for tidal classification), the tide-induced turbulence breaks down any stratification such that the waters are fully mixed, whereby the three-dimensional physics of the flow become diminished and even negligible, especially considering the horizontal (namely, longitudinal) variations of salinity in the river. For a river like the St. Johns, where tidal effects extend up to Lake George (river km 200), the horizontal (longitudinal) variations of salinity are important to consider in the modeling strategy.  To that end, the ADCIRC model, which is far more flexible than the EFDC model in providing horizontal resolution, becomes the preferred model.  On the other, the EFDC model has been shown to perform very well for salinity simulation in the St. Johns River. In a quantitative sense, the EFDC can simulate salinity at Dames Point (river km 20) with accuracy of 2.3-4.1 ppt RMSE and 83-94% R2, while ADCIRC performs at 3.5 ppt RMSE and 90% R2.

2.8.3. Depth Variation of Salinity

This sub-section covers depth variation of salinity by analysis of depth-dependent time-series data including episodic storm events. Continuous salinity data at 1-hour interval are available for six locations in the lower St. Johns River from as  early as 1995, collected as a joint effort between the United States Geological Survey and the St. Johns River Water Management District for the study of total maximum daily loadings (Sucsy and Morris IV 2002). Although these data are ‘old’ (ca. 1990s), they still help us understand important aspects of the river today. The four most downstream stations are focused in this sub-section on longitudinal salinity variations (Figure 2.71):

  • Dames Point
    River km 20
  • Acosta Bridge
    River km 40
  • Buckman Bridge
    River km 60
  • Shands Bridge
    River km 80

Salinity data were collected at three different depths within the local water column: top level; mid-level; and lower level.  Figure 2.8 shows time-series data of salinity at the three different depths, over the four salinity monitoring stations (Figure 2.6) for the ‘Most Variable’ period (September 21–October 21, 1999). The plots show little to no depth variation of salinity at the three upstream stations (Acosta Bridge, Buckman Bridge and Shands Bridge) and show some depth variation of salinity at the downstream station (Dames Point), which is corroborated by the calculated differences of top salinity minus bottom salinity. At Acosta Bridge, Buckman Bridge and Shands Bridge, top-bottom salinity differences generally are negligible, while sometimes the salinity difference can be measureable (e.g., reaching as low as -5 ppt and as high as 2 ppt). At Dames Point, top-bottom salinity differences generally range from -5 to 5 ppt, while sometimes the salinity difference can be more (e.g., reaching as low as -9 ppt and as high as 13 ppt). Considering time averages, top-bottom salinity differences are 1.5, -0.1, -0.3, and -0.1 ppt at Dames Point, Acosta Bridge, Buckman Bridge, and Shands Bridge, respectively.

Figure 2.73
Figure 2.73 Depth-dependent salinity at the four salinity monitoring stations for the ‘Most Variable’ period (September 21-October 21, 1999): a) Dames Point; b) Acosta Bridge; c) Buckman Bridge; d) Shands Bridge; and e) Salinity differences, as computed by top salinity minus bottom salinity, for the four stations.

Figure 2.74 shows the bathymetric depth, the three depth-levels where salinity data were collected and the corresponding top-, mid- and bottom-level salinity data (displayed as average plus/minus standard deviation: AVE ± STD) for the four salinity monitoring stations. The bottom panel of the graphic illustrates the data versus the model, in terms of statistics (average plus and minus one standard deviation), for the three different time periods: HE — high extreme, 30-day duration when salinity was generally high in the river; MV — most variable, 30-day duration when salinity was highly (longitudinally) variable; and LE — low extreme, 30-day duration when salinity was generally low in the river.  At Dames Point, there is some depth variation of salinity such that (for the MV time period) the top-level AVE ± STD salinity is 21.4 ± 4.4 ppt, the mid-level AVE ± STD salinity is 21.6 ± 4.8 ppt and the bottom-level AVE ± STD salinity is 20.1 ± 5.1 ppt.  At Acosta Bridge, Buckman Bridge and Shands Bridge, there is essentially no depth variation of salinity.

Figure 2.74
Figure 2.74 (Top Panel) Depths of Salinity Measurement Device within Vertical Water Column at Four Gaging Stations, along with Depth of Local Channel Bottom; and (Bottom Panel) Salinity Data as Average Plus and Minus One Standard Deviation for Top, Middle and Bottom Measurement Levels (Box Plots) Overlaid on the Statistics (Average Plus and Minus One Standard Deviation) for the Model Results (Area Plots), for Four Gaging Stations and Three Time Periods.

The data analysis suggests that salinity in the lower St. Johns River is vertically well-mixed, especially upstream of Acosta Bridge, located at river km 40; though, the vertical salinity structure can become partially stratified near the river mouth (i.e., downstream of Dames Point, located at river km 20). It is strange that, at Dames Point, the top measurement of salinity would be greater than the bottom measurement of salinity, which is subject for further investigation. At the least, there appears to be some kind of sub-tidal frequency with the top-to-bottom salinity difference. There is question as to the measurement device used, namely with regards to measuring relatively high salinity at Shands Bridge (approximately 3 ppt), which usually has salinity near zero and at most 0.5 ppt (personal communication, Courtney Hackney). Also, there is the question of the amount of salinity in the lower St. Johns River that is ocean-derived versus spring-fed.

Stratification is a common feature of salinity in estuaries, where lighter freshwater inflows ride on top of denser saline tidal flows; however, for meso-tidal estuaries with tidal ranges of 2 m or greater, the tide-induced turbulence can breakdown the stratification and fully mix salinity within the vertical water column. To examine stratification in the lower St. Johns River, the salinity data for Dames Point (river km 20) were statistically analyzed for the depth-integration (average) and depth-variability (standard deviation) of salinity among the three levels (top, middle and bottom) for all time values of the data.  Of the 52,894 time values in the full dataset (1996-2007), 36,742 (69%) of the total time values met the QAQC criterion for extraction of bogus measures, and the longest contiguous record of the QAQC data was 63 days.  The time-average standard deviation (STD) of vertical variance of salinity was computed to be 1.13 ppt, which when compared with the time-average depth integration (AVG) of salinity (23.3 ppt) relates to a STD:AVG ratio of 5%. The time-standard deviation of depth-integrated salinity was computed to be 6.23 ppt, which corresponds to a STD:AVG ratio of 27%. To examine the potential causality of stratification in the lower St. Johns River, water level data were gathered from the nearest offshore tidal gaging station (Fernandina Beach, Florida: NOS ID 8720030) and analyzed for sub-tidal component, which produced a time-series signal of the offshore sub-tidal variability for comparison with the time-series occurrences of salinity stratification in the river.  Tide-filtered discharge data were obtained from the USGS gaging station located at Astor (#02236125).  Figure 2.75 shows time-series plots of the occurrences when vertical variance of salinity was greater than STD = 1.13 ppt. Qualitatively, there is an apparent correlation between salinity stratification and offshore sub-tidal variability, particularly a positive correlation with stratified salinity when offshore sub-tidal water level is high or on the rise, while trend between salinity stratification and upstream daily discharge is indiscernible.  Trend analysis of the data continues with multiple-regression methods, where those results will be added to this subsection in future years.

Figure 2.75
Figure 2.75 Time Series of Depth-Average (Top) and Depth-Variance (Bottom) of Salinity for Occurrences above Threshold

2.8.4. Longitudinal Salinity Variations

This sub-section covers longitudinal salinity variations using time-series data including episodic storm events.  The data are available for the four salinity monitoring stations (Dames Point — river km 20; Acosta Bridge — river km 40; Buckman Bridge — river km 60; and Shands Bridge — river km 80). This sub-section focuses on the depth average (it was cited earlier in the text as ‘depth-integrated,’ which is referring to the same concept) of the salinity data and how salinity varies over the four locations in the lower St. Johns River (from river km 20 to river km 80).  The salinity data were depth-averaged for the three levels (high, middle, and low), over the four salinity monitoring stations (Figure 2.71) for the aforementioned three one-month periods in 1999 (Figure 2.76).

Salinity in the lower St. Johns River fluctuates semi-diurnally (approximately every 12 hours and 25 minutes) which is dictated by the astronomic (ocean) and estuarine (river) tide. The tidally driven salinity fluctuations are greatest (approximately 10 ppt in range at Dames Point) when salinity is generally low, like in the case of the ‘Low Extreme’ period (October 30–November 29, 1999), and least (approximately 6 ppt in range at Dames Point) when salinity is generally high, like in the case of the ‘High Extreme’ period (June 13 – July 13, 1999). The signal (‘signal’ just being a compact word to represent the tidally driven salinity fluctuation) becomes progressively weaker with greater distance upstream, which corresponds with the diminishing tidal hydrodynamics with greater distance up the lower St. Johns River (Bacopoulos, et al. 2012).  Salinity at Shands Bridge varies minimally with only 2-3 ppt of range.

Salinity is generally high for the ‘High Extreme’ period (June 13 – July 13, 1999) with near complete seawater (35 ppt) experienced at Dames Point and salinity reaching near 20, 12 and 5 ppt for Acosta Bridge, Buckman Bridge and Shands Bridge, respectively, for the first two weeks of the record.  On the other hand, consider the generally low salinity for the ‘Low Extreme’ period (October 30 – November 29, 1999), when the lower St. Johns River became nearly entirely fresh (0 ppt).

Salinity spikes on October 15-19, 1999, which is due to Hurricane Irene (Avila 1999), and is especially noticeable at Dames Point and Acosta Bridge while lesser noticeable at Buckman Bridge and Shands Bridge.  As a representative episodic storm event, Hurricane Irene caused salinity levels in the lower St. Johns River to fluctuate by over 10 ppt in just a matter of days.

In closing, the longitudinal variation of salinity in the lower St. Johns River covers the full range of salinity (0–35 ppt), as was evidenced in the salinity data available from the four salinity monitoring stations, which span river km 20–80.  In addition to the longitudinal variation of salinity in the lower St. Johns River, short-term events, like Hurricane Irene that was observed in the salinity data, are able to cause salinity spikes and jumps in salinity by over 10 ppt for two to three days.

Figure 2.76
Figure 2.76 Depth-Averaged Salinity at the Four Salinity Monitoring Stations for the Three One-Month Periods in 1999.

2.8.5. Long-Term Trend Analysis

The full records of salinity data for Dames Point (river km 20) and Acosta Bridge (river km 40) were de-tided and filtered, which resulted in long-term (1996-2007) sub-tidal salinity records for both stations (Figure 2.77). Although these data are ‘old’ (ca. 1990s and 2000s), they still help us understand important aspects of the river today. Each record is essentially a vector of hourly sub-tidal salinity values.  For each record, the data vector was parsed, daily-averaged and organized into a data matrix the size of 365 by 11 (number of rows = number of days in a non-leap year; number of columns = number of years in the trend analysis).  Then each row of the data matrix was pushed through a quadratic regression to generate a (steady coefficient), b (linear coefficient) and c (quadratic coefficient) for each day of a synthetic year.  Yearly-average b values were computed to be 0.99 and 3.31 ppt yr–1 for Dames Point and Acosta Bridge, respectively.  There was a total of 248 days (68% of a synthetic year) when b was positive for Dames Point, and 267 days (73%) for Acosta Bridge.  The sub-tidal salinity signals for Dames Point and Acosta Bridge were synthesized using just the daily-based regression coefficients (an, bn and cn for n = 1,2,…,364,365) and a time vector of 1-11 to correspond with 1996-2007, which shows remarkable fit with the sub-tidal salinity data (Figure 2.77). The long-term trend analysis continues with correlation of the sub-tidal salinity data for Dames Point with sub-tidal offshore water level and the sub-tidal salinity data for Acosta Bridge with sub-tidal upstream river discharge, the results of which will be added to this subsection in future years.

Figure 2.77
Figure 2.77 Time Series of Raw, De-Tided (T_TIDE), Residual and Filtered (SG) Residual Salinity Data for (a) Dames Point and (b) Acosta Bridge.

Hourly salinity data were obtained from Acosta Bridge (USGS 02236125) for late October 2015 – mid April 2016, where the data were de-tided, followed by filtering with Savitzky-Golay scheme (Figure 2.79).  The steady salinity level is 5.4 ppt with tidal range of 10.4 ppt, while the subtidal salinity level ranges between extrema of -7.0 and 16.9 ppt. The SG-filtered residual salinity signal for 2015-2016 was compared with the yearly-average and yearly-range (yearly-average plus/minus yearly-based standard deviation) salinity levels for 1996–2007, where the 2015–2016 salinity never dipped below the lower limit of the 1996-2007 salinity range, while it went above the upper limit for March. The 2015-2016 salinity was 85% greater than 1996-2007 salinity on a yearly-average basis, and 2015–2016 salinity was at the 73rd percentile of 1996-2007 salinity on a yearly-range basis.

Salinity in the river was found to be on rise for the 1996-2007 record at yearly-averaged rates of 1-3 ppt yr–1.  The rise in salinity correlated with elevated offshore (subtidal) water level, low hydrologic (daily) inflow and increasing (daily) flood flow, which explains the rising salinity in the river as the result of larger-than-normal head (potential) from the offshore sea level and lower-than-normal freshwater flux from the upper river stem and watershed basin, thus driving in and entraining more saline water in the river over the long term for 1996-2007. The salinity data for 2015-2016 were found to be in the upper range (73rd percentile on a yearly average) of the salinity data for 1996-2007, while in some cases, the 2015-2016 salinity exceeded the upper limit of the 1996-2007 salinity range. Given the present-day (2015-2016) conditions of offshore sea level and hydrologic inflow, salinity levels in the lower St. Johns River can be expected to rise going into the future.

Figure 2.78
Figure 2.78 Time Series of Synthesized Residual Salinity Signal and Filtered (SG) Residual Salinity Data for (a) Dames Point and (b) Acosta Bridge.
Figure 2.79
Figure 2.79 (a) Time Series of Salinity Data for Acosta Bridge for 2015-2016 with (b) the Residual Signal Compared with the 1996-2007 Synthetic-Year of Residual Signal and (c) the Residual Signal Compared with the Average and Range of the 1996–2007 Synthetic-Year of Residual Signal.

2.8.6. Climate Change-Impact Assessment

Salinity was simulated for the lower St. Johns River based on freshwater river inflow, offshore water levels and sea-level rise. The salinity simulations span over a contiguous, decadal-scale record (1997-2007).  Freshwater river inflow for Astor (see Figure 2.80 for station location) and nine tributaries was applied at the upstream boundary of the model (Figure 2.81).  Offshore water levels for Fernandina Beach (see Figure 2.80 for station location) were applied at the open-ocean boundary of the model (Figure 2.82). Although these model simulations and the inherent data are ‘old’ (ca. 1990s and 2000s), they still help us understand important aspects of the river today.

Figure 2.80
Figure 2.80 Locator Map of Salinity-Gaging Stations Located in the Lower St. Johns River and the Upstream Boundary (Data Derived from Astor) and Open-Ocean Boundary (Data Derived from Fernandina Beach) of the Model.
Figure 2.81
Figure 2.81 Freshwater River Inflow (Data Derived from Astor) of the Upstream Boundary Condition for the Salinity Simulations.
Figure 2.82
Figure 2.82 Water levels (Data Derived from Fernandina Beach) of the Open-Ocean Boundary Condition for the Salinity Simulations.

Five ten-year-long simulations were conducted: based on historical boundary conditions; and based on sea-level rise.  Sea-level rise was applied as an increase in the overall water level of the offshore boundary condition: mild scenario (0.05 m); moderate scenario (0.15 m); and accelerated scenario (0.30 m). The fifth simulation applied a negative (time-backward) sea-level rise of -0.05 m. The five simulations are listed in Table 2.32.

Table 2.32 Five Ten-Year-Long Simulations were Conducted for Simulation of Salinity in the Lower St. Johns River

Simulation name/indexSea-level rise (m)
SIM1-0.05
SIM20.00
SIM30.05
SIM40.15
SIM50.30

As a validation of the model, the results with SIM2 were compared with salinity data for four gaging stations located in the lower St. Johns River (Figure 2.80): Dames Point; Acosta Bridge; Buckman Bridge; and Shands Bridge. Figures 2.83-2.86 display yearly plots of salinity (model vs. data) for the four gaging stations.  Firstly, the spatial feature of salinity in the river is brought out by the longitudinal variation of salinity among the four stations (ranging from river km 20 to river km 80). Salinity at Dames Point is in the higher end of the salinity range (typically in the range of 20–40 ppt) and strongly tidally influenced (the range of fluctuations is on the order of 10 ppt). Salinity at Acosta Bridge is in the middle of the salinity range (typically in the range of 10–30 ppt) and moderately tidally influenced (the range of fluctuations is about 5 ppt). Salinity at Buckman Bridge and Shands Bridge is in the lower end of the salinity range (typically in the range of 0–10 ppt) and mildly tidally influenced (the range of fluctuations is 1–2 ppt). Secondly, salinity at all four stations exhibits sensitivity to non-tidal activity (offshore water level and freshwater inflow), which is most evident in the late summertime events (days 250–300).  Events typically involve a drawdown (sometime rapid, occurring within 1–2 days) of salinity due to high freshwater inflow followed by a more gradual (on the order of a week) return to the antecedent salinity level. Unfortunately, the model simulation only partially captures the more significant events, as though there are physics (not in the model) causing the fast fall and slow rebound of salinity during the more significant events.

Table 2.33 presents quantitative measures of the data-model fits of salinity, including root-mean-square error (RMSE), scatter index (SI = RMSE ÷ mean-value) and index of agreement (IA—analogous to R2).  The model performs good-to-very good at the two downstream stations (Dames Point and Acosta Bridge) with IA values of 0.79 and 0.83, and performs fair-to-good at the two upstream stations (Buckman Bridge and Shands Bridge) with IA values of 0.68 and 0.34.

Table 2.33 Quantitative Measures of Data-Model Fits of Salinity for the Four Gaging Stations Located in the Lower St. Johns River: Root-Mean-Square Error (RMSE); Scatter Index (SI); and Index of Agreement (IA)

StationRMSE (ppt)SI (-)IA (-)
Dames Point (river km 20)5.340.230.79
Acosta Bridge (river km 40)5.010.650.83
Buckman Bridge (river km 60)3.221.030.68
Shands Bridge (river km 80)1.491.550.34

The impact of sea-level rise on salinity was analyzed in a linear sense and a nonlinear sense:

  • Linear: SIMs 1 and 3 compose the linear analysis, i.e., how salinity changes the same over a given step: SALfuture = SALpresent-day + (SAL-CHANGEpresent-day × ΔSLR), where SAL-CHANGEpresent-day is the linear rate-of-change of salinity with respect to sea-level rise at present-day conditions (= [SAL+0.05m – SAL–0.05m] / [0.05 m – {-0.05 m}])
  • Nonlinear: SIMs 2, 3, 4 and 5 compose the nonlinear analysis, i.e., how salinity is a function of sea-level rise: SALfuture = f(SLR) = f(0.00 m, 0.05 m, 0.15 m, 0.30 m)
Figure 2.83
Figure 2.83 Ten-Year Record (1996-2007) of Salinity (Model vs. Data) for Dames Point (River km 20).
Figure 2.84
Figure 2.84 Ten-Year Record (1996-2007) of Salinity (Model vs. Data) for Acosta Bridge (River km 40).
Figure 2.85
Figure 2.85 Ten-Year Record (1996–2007) of Salinity (Model vs. Data) for Buckman Bridge (River km 60).
Figure 2.86
Figure 2.86 Ten-Year Record (1996–2007) of Salinity (Model vs. Data) for Shands Bridge (River km 80).

The polynomial fit of salinity (i.e., spoly = p3δ2 + p4δ + p5, where δ = sea-level rise) was used to generate linear (i.e., summation of terms p5 and p4) and nonlinear (i.e., summation of terms p5, p4 and p3) approximations of salinity change due to sea-level rise (Figure 2.87). The x-axis ranges over the four sea-level rise scenarios (and present-day conditions) investigated in the study.  The left y-axis of each plot is the absolute change in salinity due to sea-level rise, and the right y-axis is the absolute salinity change normalized by the record-average value (p5) of present-day salinity, multiplied by 100 to express the salinity change as a percentage.  There is little to no difference between the nonlinear and linear salinity change for mild sea-level rise (+0.05 m); however, the nonlinear-linear difference is noticeable for moderate sea-level rise (+0.15 m) and appreciable for accelerated sea-level rise (+0.30 m).  The nonlinear-linear difference is greatest (as much as 6-9%) for the two middle stations (Acosta Bridge and Buckman Bridge) and relatively much smaller (as much as 1%) for the downstream and upstream stations (Dames Point and Shands Bridge). Lastly, the nonlinear-linear difference is negative for Shands Bridge.

Figure 2.87
Figure 2.87 Linear and Nonlinear Change in Salinity due to Sea-Level Rise for the Four Salinity-Gauging Stations Located in the Lower St. Johns River.

As some discussion of the results, salinity will increase over the entire 200-km length of the lower St. Johns River due to sea-level rise. In general, risen offshore water levels will enable more saline waters to flux into the river, while subsequently pushing back on the freshwater discharge coming down from upriver, which together will cause salinity in the river to increase.  The salinity increase will be spatially variable over the river’s length, where the greatest change in salinity will occur around river km 40, with a derivative valued at 0.06 ppt cm–1, and lower-valued derivatives of salinity change will occur at the river mouth and upstream of river km 60. Although the derivatives of salinity increase for river km’s 80-200 are near nil on an absolute basis, they are non-negligible when compared with the present-day condition of salinity (0.2-0.5% cm–1). The salinity increase will be nonlinear, particularly for scenarios of moderate to accelerated sea-level rise (0.15-0.30 m). The nonlinearity will exacerbate the salinity increase, especially for river km’s 20-40, where there is a local “hotspot” in the river where salinity will increase nonlinearly (greater than the linear trend) due sea-level rise.  For the case of accelerated sea-level rise (0.30 m), salinity at river km 40 will increase by close to 2 ppt, which corresponds to a greater-than-20% increase in salinity from present-day conditions.

2.8.7. Biological Impacts

This sub-section covers potential biological impacts of salinity on the flora and fauna of LSJRB. Salinity increases as a result of the environment can be looked at in terms of 1) periodic short term events like storms that result in abrupt salinity spikes for less than 14 days. 2) Intermediate term events like droughts that result in elevated salinity for some weeks. 3) Long-term changes as a result of sea levels rising over many years. 4) Salinity can also be altered due to human activities in the basin, such as reduced freshwater inflows to the river caused by dams, surface water withdrawals, or significant pumping of ground water. In addition, activities, such as harbor deepening, tend to increase salt water entering an estuary, thus driving up the salinity (Sucsy 2008).

The LSJRB supports a diverse community of living organisms that are important to the ecosystem, are affected by salinity, and have significant recreational and commercial economic value. Submerged aquatic vegetation and invertebrate bottom dwelling organisms play an important role in shaping habitat so that it is able to support fish and other wildlife. Examples of commercially valuable organisms include blue crabs, bait shrimp, and stone crabs. In 2013, Clay, Duval, Flagler, Putnam, and St. Johns Counties reported a total commercial crab harvest of 1,615,232 lbs (73%); and a fish harvest of some 570,509 lb (FWRI 2017a). In general, striped mullet, whiting, and flounder have been the most caught species, but recreationally, red drum, spotted sea trout, croaker, sheepshead, flounder, largemouth bass, and blue gill are most important to anglers.

For all the species of fish and invertebrates mentioned in this report there are a few themes of importance:

  • Each species plays an essential role in the ecosystem, with many interdependencies (predator, prey relationships).
  • Each species requires essential habitats for an important life stage (coastal and in the river).
  • Each species is of commercial and recreational value that is supported by the rest of the ecosystem, which also has value.

The most recent Supplemental Environmental Impact Statement (SEIS) by the USACE regarding dredging in the St. Johns River indicated that salinity changes, as a result of dredging, would negatively impact the distribution of Submerged Aquatic Vegetation (SAV) in LSJR. The impact would likely be from increased salinity stress on SAVs in the most northern part of their range in LSJR (Duval, Clay, and St. Johns Counties). Moreover, the report states that the 46 feet and 50 feet dredge depth scenarios would increase salinity stress by 32 and 43 acres of potential SAV habitat per day, respectively. This would most likely lead to a reduction in manatee forage habitat, essential fish habitat, benthic macro-invertebrate habitat and freshwater wetlands (USACE 2014a). In addition, the report states that loss of SAVs would represent a small portion of the total available SAVs in the LSJR, also that blue crabs and other marine species may benefit from any increases in salinity. In Appendix E of the report (USACE 2014b), the USACE pledged to monitor salinity and water quality to ensure appropriate mitigation. Furthermore, that the mitigation for SAVs lost is to be accomplished through a Corrective Action Plan that would purchase conservation lands (638 acres of freshwater wetlands, uplands, river shoreline, and saltmarshes).

In the report, the USACE states that the analysis and conclusions were based on modeling efforts that make certain assumptions about the rate of sea level rise (hydrodynamic modeling), and that salinity stress on SAVs was developed from a separate modeling analysis (Taylor 2013a) based on assumptions about levels of salinity stress and SAV acreages (ecological modeling). The hydrodynamic model reports (Taylor 2011; Taylor 2013b; Taylor 2013c) presented error statistics for the EFDC and CE-QUAL-ICM models. However, similar error statistics could not be calculated for the ecological models, and that represents an uncertain risk associated with evaluation of the ecological model results. Moreover, the report stated that, “Future condition hydrodynamic model simulations further rely on assumptions about the rate of sea level rise, quantity of water withdrawal from the middle St. Johns River, patterns of land use, and other factors. Actual conditions will deviate from those used to drive the models. These deviations introduce additional uncertainty in the models’ ability to predict future conditions and impacts. These uncertainties are; however, inherent in the use of numerical models and do not represent an unknown risk” (USACE 2014a; Section 7.2, p. 258).

On February 19, 2016, the DEP issued a Notice of Intent to issue an Environmental Resource Permit and a Variance to allow the Army Corps of Engineers to dredge 13 miles of the St. Johns River from the mouth of the river to Brills Cut from a depth of 40 feet to up to 51 feet. The St. John RIVERKEEPER filed a Petition for Formal Administrative Hearing against DEP on April 1, 2016, based on the contention that the potential environmental impacts were not adequately addressed in the permit and important water quality standards are waived increasing the inherent risks of the proposed deep dredge. The USACE reacted by filing a Notice of Non-Participation asserting sovereign immunity and indicating that it does not plan to participate as a party in the administrative proceeding. This is an unprecedented move, which is likely to create the potential for more risk since the USACE contends that they are immune from abiding to Florida water quality standards. On July 26, 2016 St. Johns RIVERKEEPER filed a notice withdrawing its legal challenge of the state permit due the lack of enforceability with the intent to elevate the challenge to the federal level. On Friday, April 7, 2017, St. Johns RIVERKEEPER filed a Complaint for Declaratory and Injunctive Relief in federal court against the USACE regarding the proposed St. Johns River harbor deepening project. Consequently, any further developments or actions depend on legal issues being resolved.

2.8.7.1. Macroinvertebrates

These are animals without a backbone that live in or on river bottom sediments including small crabs, snails, shrimp, clams, insects, worms, and barnacles among others species (see Section 4.3). These organisms affect oxygen levels in the sediment, as well as sediment size, which in turn affect what is able to live and grow in proximity to them.  Macroinvertebrates are useful indicators of environmental stress and species change as one transitions from higher to low salinity. DEP data from 1974-1999 indicated that the northern river section was dominated by barnacles, polychaetes, and amphipods; and the southern river area was dominated by mollusks, amphipods, polychaetes, oligochaetes, and fly larvae. During the 1980s, the north section was dominated by polychaetes and barnacles, while the southern portion was mostly oligochaetes and fly larvae. In the 1990s, another shift occurred due to salinity, where the northern stations were dominated by amphipods, mollusks, polychaetes, and barnacles and the southern areas by bivalves and snails (Evans, et al. 2004; Montagna, et al. 2011).

Evans, et al. 2004 states that freshwater areas of the river are affected by increasing salinity and that the concern is this will likely change the invertebrate community, the result could be significant negative impacts on the quality and quantity of freshwater fish species harvested from LSJRB. At this time, there is a lack of recent data on macroinvertebrates and how parameters, such as low dissolved oxygen, sediment quality, and toxic substances in the environment, may interact with changes in salinity levels.

2.8.7.2. Blue Crabs

The blue crab is a common benthic predator that represents the largest commercial fishery in LSJRB. Successful crab reproduction relies on a particular set of salinity conditions at specific times in the life cycle. Females carry fertilized eggs and migrate towards the more marine waters near the mouth of the river where they will release their eggs into the water (see section 3.3.2 Fisheries). After some time adrift, wind and currents transport the megalops larvae back to the estuarine parts of the river where they settle in submerged aquatic vegetation (SAV) that serves as a nursery.

One concern that may negatively affect the recruitment of new crabs into the population is that with increasing salinity levels, the salinity transition zone will shift further south increasing the distance that female crabs with eggs will need to travel in order to reach the river mouth. This could ultimately affect recruitment.

Another concern is associated with nursery habitat. Increasing salinity further south in the river will negatively impact submerged aquatic vegetation that is required for young crabs.

Also, since the price of crustaceans in general is dependent on size, yet another concern may be diminishing size of adult crabs. There are several studies mentioned in Tagatz 1968a that report an inverse relationship between salinity and size. The higher the salinity of water in which growth occurs the smaller the adult sizes. This may be due to the crabs absorbing more water in lower salinity conditions when they molt (bigger crab) as opposed to them absorbing less water under higher salinity conditions (smaller crab). As a result, this could translate into lower income per pound for commercial harvesters for a particular level of fishing effort.

Ecologically speaking, blue crabs are very important in both the benthic and planktonic food webs in the St. Johns River. They are important predators that can affect the abundance of many macroinvertebrates, such as bivalves, smaller crabs, and worms. They are also important prey for many species. Smaller crabs provide food for drum, spot, croaker, seatrout and catfish, while sharks and rays eat larger individuals (White, et al. 2009).

2.8.7.3. Shrimp

Three principle shrimp species found in the area include most commonly White Shrimp (Litopenaeus setiferus), Brown Shrimp (Farfantepenaeus aztecus), and Pink Shrimp (Farfantepenaeus duorarum). All are omnivores feeding on worms, amphipods, mollusks, copepods, isopods and organic detritus. White shrimp spawn from April to October; pink shrimp (February to March) and brown shrimp (March to September) (FWRI 2008d). All species spawn offshore in deeper waters with larvae developing in the plankton and eventually settling in salt marsh tidal creeks with appropriate salinities within the estuaries. Changes in salinity will cause a change in the distribution of these early life stages that could potentially affect the number of adults returning offshore. Shrimp are important in both benthic and planktonic food webs in SJR. They affect the abundance of many small macroinvertebrates. They are also important prey for many other species. As small planktonic individuals, the shrimp post‐larvae and juveniles forms provide food for other estuarine species like sheepshead minnows, insect larvae, killifish, and blue crabs. As adult shrimp, they are preyed on by finfish found within the river. The commercial shrimp fishery is one of the largest fisheries in the region, but most shrimp for human consumption are caught offshore.

2.8.7.4. Fish

The SJRWMD (McCloud 2010) compared current FWRI fish data with those collected by Tagatz in 1968 (Tagatz 1968b). The data suggested that at some areas of the river, fish communities were 50% different between 1968 and the 2001-2006 time periods. The differences in fish communities in these areas may have been the result of a transition zone between marine and freshwater moving further upstream (Figures 2.88-2.90). It is important to note that most fish are able to move from an area in response to changes in environmental factors, such as salinity, dissolved oxygen, and temperature. However, sessile species of plants and animals that are closely associated with the bottom substrate cannot move and can be impacted by such variations depending on the frequency and duration of events. Moreover, for the species that can move, there may be important life stages for these that dependent on water quality parameters being relatively stable at essential habitat areas like nursery and spawning grounds. Although fish can move, they may not be able to reproduce effectively because essential habitat has been disrupted that affects a particular life stage.

Figure 2.88
Figure 2.88 Salinity on the bottom of SJR (Station SJR17 near JU) values above the bars indicate the numbers of observations. Solid line (mean), vertical lines (maximum and minimum), and bars (Standard Deviation of the mean) (Data source: Deuerling 2017). SJR17 mean 25.15‰ (SD ± 5.28) for the maxima. Note that only 5 observation were made in 2013; 4 2014, and 3 in 2016.
Figure 2.89
Figure 2.89 Salinity on the bottom of SJR (Mainstem Station SJR40 located mid-channel N. of Piney Pt. 100 m west of green marker 5) values above the bars indicate the numbers of observations. Solid line (mean), vertical lines (maximum and minimum), and bars (Standard Deviation of the mean) (Data source: Deuerling 2017) SJR40 mean 14.05‰ (S.D. ± 5.43 for the maxima). Note that only 5 observations were made in 2013; 2 in 2014, and 3 in 2016.
Figure 2.90
Figure 2.90 Salinity on the bottom of SJR (Station SJR34/34A located ~ 1000 m south of Doctors Lake on the west bank) values above the bars indicate the numbers of observations. Solid line (mean), vertical lines (maximum and minimum), and bars (Standard Deviation of the mean). (Data source: Deuerling 2017). SJR34/34A mean 8.01‰ (SD ± 4.70) for the maxima. Note that only one observation was made in 2014, and 3 in 2016.

With regard to living organisms, changes in water quality parameter averages are not as meaningful as the changes that may occur in the parameter extremes – like salinity maxima and dissolved oxygen minima. If any changes were to persist for an extended time or if they occurred too abruptly then this is likely to be detrimental to survival. Salinity changes may potentially affect the distribution of these fish within estuary creeks and the river by affecting prey distributions for different life stages. As the salinity zone shifts further south, fresh water species are likely to be more impacted than more salt tolerant species.

Red Drum (Sciaenops ocellatus): Red drum is predatory fish that are found in the SJR estuary. The juveniles move into estuary creeks and rivers. Red drum is ecologically in the food web of the St. Johns River where they are bottom feeders that eat crabs, shrimp, worms and small fish. Their predators include larger fish, birds, and turtles. A strong recreational fishery exists; however, drum has not been commercially harvested since 1988.

Spotted Seatrout (Cynoscion nebulosus): The spotted seatrout is another bottom-dwelling predator common to estuaries and shallow coastal habitats. It feeds on small fish species such as anchovies, pinfish and menhaden as well as shrimp. Spotted seatrout larvae feed mostly on copepods, which are part of the plankton. There are a number of predators that feed on seatrout including Atlantic croakers, cormorants, brown pelicans, bottlenose dolphins, and sharks. These fish have significant commercial and recreational value.

Largemouth Bass (Micropterus salmoides): Largemouth bass are predators in brackish to freshwater habitats in SJR, including lakes and ponds. The young feed on zooplankton, insects and crustaceans including crayfish. Adults feed on a variety of larger fish, crayfish, crabs, frogs, and salamanders. Spawning occurs from December to May, with males constructing nests and guarding young in hard-bottom areas near shorelines. Largemouth bass are aggressive predators, significantly affecting the abundance of many organisms in the area. Bass are a popular game fish in the area supporting fishing tournaments.

Channel & White Catfish (Ictalurus punctatus & Ameiurus catus): Channel and white catfish are omnivorous fish found in freshwater rivers, streams, ponds and lakes. During their lifetime, they may feed on insects, crustaceans (including crayfish), mollusks and fish (DeMort 1990). Male will build and guard the nest and fry. Both catfish species are important in benthic food webs that occur in the freshwater sections of the LSJR. Catfish are commercially and recreationally important in SJR.

Striped Mullet (Mugil cephalus): Striped mullet are detritivores that can live in a wide salinity range. They are abundant in most of the SJR, closely associated with bottom mud and feeding on algae, and decaying plant material. Mullet spawn offshore and their larvae drift back into the SJR estuary. They help to transfer energy from detrital matter that they feed on to their predators – birds, seatrout, sharks, and marine mammals. The commercial mullet fishery has been the largest among all fisheries in the St. Johns for many years with over 100,000 lbs harvested annually. Additionally, mullet have significant recreational value as food and bait.

Southern Flounder (Paralichthys lethostigma): These are another common fish in the SJR estuary that are bottom-dwelling predators that eat shrimp, crabs, snails, bivalves and small fish. After spawning offshore in fall and winter, the larvae drift as part of the plankton eventually being transported back to the estuary to settle and grow. They are important in maintaining ecological balance in their roles as both predator and prey. They are food for sharks, marine mammals and birds. Flounders are important both commercially and recreationally in SJR.

Sheepshead (Archosargus probatocephalus): These fish are common to the SJR estuary and coastal waters. They prey on bivalves, crabs and barnacles. The fish spawn off shore in spring and the developing larvae are carried back to the coast by currents. The larvae enter the inlets and settle in shallow grassy areas. These fish are important in maintaining the estuarine and coastal food web as both a predator and prey. Sheepshead are prey for sharks and marine mammals. They are ecologically, recreationally and commercially important.

Atlantic Croaker (Micropogonias undulatus): These are bottom-dwelling predators common around rocks and pilings in the estuary. Spawning takes place in winter and spring in offshore waters, and planktonic offspring are transported back inshore to settle in vegetated shallow marsh areas. Croakers are important in the food web as both predator and particularly as prey. They feed on small invertebrates, and are fed on by fish, such as red drum, seatrout, and sharks. These fish support significant commercial and recreational fisheries in LSJR.

Baitfish (multiple species): There are more than two-dozen small schooling species like anchovies, menhaden, herring, killifish, sheepshead minnows, and sardines. Many baitfish species play a vital role in the ecosystem as planktivores. Others eat small crabs, worms, shrimp and fish. Most spawning occurs at inlets or offshore. Most migrate along or away from the shore. When the larvae hatch they are transported back to the estuary where they grow. Baitfish are important as prey for many larger fish species. They are also important as omnivores that recycle plant and/or animal material making that energy available to higher trophic levels. Commercial uses include bait fish, such as anchovy, menhaden, sardines, and herring which are converted into fertilizers, fishmeal, oil, and pet food (FWC 2000). Smaller fisheries catch killifish, sheepshead minnows, and sardines. For more information see Section 3 Fisheries and Appendix 3.1.

2.8.7.5. Submerged Aquatic Vegetation (SAV)

Submerged aquatic vegetation provides nursery habitat for a variety of aquatic life, helps to reduce erosion, and limits turbidity by trapping sediment. Sunlight is vital for good growth of submerged grasses. Sunlight penetration may be reduced because of increased turbidity, pollution from upland development and/or disturbance of soils. Deteriorating water quality, which may include unusual increases in salinity has been shown to cause a reduction in the amount of viable SAV in an area. This leads to erosion and further deterioration of water quality.

Historical accounts indicate that SAV beds existed in the river since 1773 (Bartram 1928– in 1955 Edition). These SAV beds have shown a gradual decline likely due to a number of cumulative impacts including routine dredging, harbor deepening, filling of wetlands, bulk heading and construction of seawalls, water withdrawals, pumping from wells, along with the contributions from chemical contamination, and sediment and nutrient loading that comes from upland development (DeMort 1990; Dobberfuhl 2007).

Commonly found SAV species within the salinity transition zone in LSJR include: tape grass (Vallisneria americana), wigeon grass (Ruppia maritime), and southern naiad (Najas guadalupensis). The greatest distribution of SAVs in Duval County is in waters south of the Fuller Warren Bridge (Kinnaird 1983a). There are about eight other freshwater species in LSJR (IFAS 2007; Sagan 2007; USDA 2013). These species are all likely to be adversely impacted by increases in salinity.

Under controlled laboratory conditions, tape grass has been shown to grow in 0 to 12 parts per thousand (ppt) of salinity and survive for short periods of time in waters with salinities up to 15‐20 ppt (Twilley and Barko 1990; Boustany, et al. 2003). However, SAV requires more light in a higher salinity environment due to increased metabolic demands (Dobberfuhl 2007). Evidence suggests that greater light availability can lessen the impact of high salinity on SAV (Kraemer, et al. 1999; French and Moore 2003). What is not clearly understood is the ability of SAV to survive higher salinities when combined with environmental variables like temperature, turbidity, and excessive nutrients.

SAV is important ecologically and economically to the LSJRB. SAV persists year round in the LSJRB and forms extensive beds which carry out the ecological role of nursery area for many important invertebrates and fish species, including the endangered Florida manatee (Trichechus manatus latirostris) (White, et al. 2002). Manatees consume from four to 11% of their body weight in SAV daily (Lomolino 1977; Bengtson 1981; Best 1981; Burns Jr, et al. 1997).

Commercial and recreational fisheries, including largemouth bass, catfish, blue crabs, and shrimp, are sustained by healthy SAV habitat (Watkins 1995). Fish and insects forage and avoid predation within the cover of the grass beds (Batzer and Wissinger 1996; Jordan, et al. 1996). For example, Jordan 2000 mentioned that SAV beds in the Lower Basin have three times greater fish abundance and 15 times greater invertebrate abundance than do adjacent sand flats.

The section of the St. Johns River north of Palatka had relatively stable trends with normal seasonal fluctuations. The availability of tape grass decreased significantly in the LSJRB during 2000‐2001, because the drought caused higher than usual salinity values. In 2003, environmental conditions returned to a more normal rainfall pattern. As a result, lower salinity values favored tape grass growth again. In 2004, salinities were initially higher than in 2003 but decreased significantly after August with the arrival of heavy rainfall associated with four hurricanes that skirted Florida (Hurricanes Charley, Francis, Ivan and Jeanne). Grass beds north of the Buckman Bridge regenerated from 2002‐2006 and then declined again in 2007 due to the onset of renewed drought conditions (White and Pinto 2006b). Sagan 2007 notes that at one of her monitoring sites, Sadler Point (the most seaward of all of her monitoring sites), SAV was present in 1998, but after a decline due to drought did not recover as did other SAV beds in the river. She cautions that long-term changes in salinity may be stressing SAV in the estuarine portions of the river. Declining SAV in the river south of Palatka and Crescent Lake is highly influenced by runoff and consequent increases in color of the water.

SAV response to drought and/or periods of reduced flow can provide crucial understanding as to how water withdrawals, harbor deepening and/or the issue of future sea level rise will likely affect the health of the ecosystem by adversely altering salinity profiles. For more information see Section 4.1 SAV and Appendix 4.1.7.1.A-D.

2.8.7.6. Florida Manatee

The Florida manatee (Trichechus manatus latirostris) inhabits the waters of the St. Johns River year-round. Manatees are generally most abundant in the LSJR from late April through August, with few manatees observed during the winter months (December-February). Manatees are protected under State and Federal Laws:  In 1967, under a law that preceded the Endangered Species Act of 1973 the manatee was listed as an endangered species. Manatees are also protected at the Federal level under the Marine Mammal Protection Act of 1972 (Congress 1972b) and at the State level under the Florida Manatee Sanctuary Act of 1978 (FWC 1978). The current federal status of the manatee is “Threatened” (March 30, 2016) having just been down listed by USFWS from “Endangered.”

Jacksonville University has conducted aerial surveys of manatees from 1994 to 2016. Within the SJR manatees were found in greater numbers south of the Fuller Warren Bridge where their food supply is greatest relative to other areas in Duval County. The SJR provides habitat for the manatee along with supporting tremendous recreational and industrial vessel usage. Watercraft deaths of manatees continue to be the most significant threat to survival. Boat traffic in the river is diverse and includes port facilities for large industrial and commercial shippers, commercial fishing, sport fishing and recreational activity. Also, in order to accommodate larger cargo ships more dredging by the port is expected in the future (Appendix 4.1.7.1.F Salinity). Dredging and/or deepening the channel can also affect the salinity conditions in the estuary by causing the salt water wedge to move further upstream (Sucsy 2008), negatively impacting biological communities like the tape grass beds on which manatees rely for food (Twilley and Barko 1990).

The average numbers of manatees observed on aerial surveys in the salinity transition zone area of the SJR decreased during periods of drought (1994-2000 and 2006-2009) and then increased again after the droughts (2000-2005 and 2009-2012) (Section 4.4). The reason for this was that during droughts elevated salinity leads to demise in the grasses that manatees feed on. As a result manatees leave the study area in search for food. Freshwater withdrawals, in addition to harbor deepening, will alter salinity regimes in the LSJRB; however, it is not known yet by how much. If a sufficient change in salinity regimes occurs, it is likely to cause a die-off of the grass bed food resources for the manatee. This result would decrease carrying capacity of the environment’s ability to support manatees

2.8.7.7. Data Sources & Limitations

Various sources of data were identified from DEP’s STORET database, SJRWMD, USGS and COJ. Monthly data obtained from The City of Jacksonville’s Environmental Quality Division “River Run” sampling program was used to determine salinity changes from 1991 to 2015. Other data sources identified include the City’s Station List (122 sites) data from 1995-2009; Tributaries (105 sites) data from 1995-2010; The River Run (10 sites) in the mainstem of SJR from 1980s to 2015; The Timucuan Run (12 Sites) in the Nassau and Ft. George area sampled every other month dating back to 1997; and the recently established Basin Management Action Plan (BMAP) Tributaries sites updated in October 2010. The latter consists of 10 Tributaries (with 2-3 sites each) for a total of 30 sites beginning in 2010.

In addition, there is Water Body ID (WBID) trend data available for Jacksonville from 1994-2015. Older data includes chlorides levels collected at Main Street Bridge from 1954 to 1965 as part of the city’s pollution sampling program around the time of the Buckman sewage plant coming on line (Hendrickson 2014).

Data obtained from The City of Jacksonville’s Environmental Quality Division “River Run” sampling program was used to determine salinity changes from 1991-2016. Data is collected about twice a month at the surface (0.5 m), middle (3-5 m), and bottom (5-10 m) in the water column. However, in recent years the sampling frequency has been significantly curtailed due to budget cuts. Four sites were chosen from the regular ten sampling stations.

1) West bank of SJR 1000 m south of Doctors Lake;
2) East bank of SJR 200 m north of a large apartment complex near Jacksonville University;
3) South bank of SJR just west of Dames Point Bridge, near the western most range marker;
4) Main stem of SJR Mid channel N. of Piney Pt. 100 m west of green marker 5.

Kendall’s Tau correlation analysis revealed that salinity over time had significantly increased at the bottom, middle and surface at SJR near Doctors Lake, Piney Point mid-river, near Jacksonville University and Dames Point Bridge. For a map of the sample sites, analysis results, and graphs showing these trends, see Figures 8-20 in Appendix 4.1.7.1.F Salinity.

Monthly data are limited in that the sampling frequency is relatively low, and short-term events in weather may not be well represented. Continuous water quality data are available on the web through the USGS (USGS 2017). Currently active stations include the Dames Point Bridge, Buckman Bridge (Figure 2.91), and Dancy Point. Other non-active stations for which data is available include Main Street Bridge and Shands Bridge. Yet, another new source for continuous data in LSJR includes NOAA’s PORTS program (NOAA 2017e). This data has some gap years due to budget cuts preventing collection. Data at the Buckman Bridge show an increasing salinity trend in surface waters from 1995-2002 (represents a period of drought), then no data was available from 2004-2007, followed by a downward trend from 2008-early 2015 (represents a period of more normal and stable rainfall). There was another data gap from April to September 2015, and then this was followed by an increasing trend in salinity from October 2015 to March 2017 (representing decreased rainfall and the onset of drought conditions). These data indicate that large salinity fluctuations occurred and persisted for some time.

Figure 2.91
Figure 2.91 Surface salinity for 1995-2002; 2008-2015; and 2015-2017 from USGS continuous data recording station at the Buckman Bridge.

2.8.8. Overall Assessment (Ratings of Status and Trend)

The salinity regime in the LSJRB has changed over the years due to various human activities and natural phenomena, including rising sea level. The river’s ecology has been changed as a result of long-term salinity changes. In addition, there is no regulatory target for salinity in various sections of the river. However, this does not mean that we are not responsible for considering the environmental impacts of activities like surface water withdrawals and dredging, or future changes in rainfall and the amount and quality of surface water runoff given increases in population. All considered, including the historical and present values and trends in salinity, the current STATUS of salinity is rated as unsatisfactory because of its impacts, and the TREND of salinity is rated as worsening because it is increasing.