Nutrient Loading at Card Sound Canal; a tidally driven canal that experiences highly stratified bi-directional flow conditions

Abstract:

Figure 1A: Card Sound Canal monitoring station and surrounding embayments. [larger image]

Data collection at Card Sound Canal began in October 2003 as part of the Coastal Gradients of Flow, Salinity, and Nutrients Project funded by the U.S. Army Corps. of Engineers, Comprehensive Everglades Restoration Plan (CERP). An evaluation of the field data indicated the potential for data bias due to routinely observed bi-directional flow and salinity/temperature stratification. Due to the unique hydrologic conditions at this site, new methods were employed to collect representative, high quality data.

In order to improve the computation of discharge and eliminate bias due to bi-directional flow, a side looking acoustic Doppler velocity meter (ADVM) was replaced with an up-looking ADVM mounted on the creek bottom. The salinity and temperature stratification problem was handled by implementing a floating probe design at the surface of the water column while at the bottom of the stream a probe was mounted below the observed thermocline. Although automatic samplers are only a point sample, two sampling intakes were deployed in a vertical configuration similar to that of the salinity/temperature probes in order to collect a more representative sample from the stratified estuary. Cross sectional water samples are collected quarterly and the comparison between the automatic sampler and grab sample results appear to correlate well together. The implementation of these field methods was critical for the eventual computation of nutrient loading at this site.

Introduction:

Considerable attention has been paid to computing nutrient loads from rivers and creeks into downstream lakes, oceans, bays, and estuaries. Total phosphorus (TP) and nitrogen (TKN) loads are of particular interest to scientists and water managers, since these nutrients are commonly applied in agricultural and urban areas to facilitate the growth of vegetation. Excessive nutrients can adversely impact fragile ecosystems located downstream causing, for example, algae blooms, fish die-offs, and other harmful ecological effects. Nutrient loading to Florida Bay and adjacent estuaries are monitored as part of the Comprehensive Everglades Restoration Plan (CERP, http://www.evergladesplan.org/). Card Sound Canal monitoring station is part of the Coastal Gradients of Flow, Salinity, and Nutrients Network funded by the U.S. Army Corps of Engineers (Fig. 1A, 1B, 1C, 1D).

Figure 1B: Hydrologic monitoring station at Card Sound Canal. [larger image]

---

Figure 1C: Bottom mounted acoustic Doppler velocity meter. [larger image]

---

Figure 1D: Multiple intake hoses on automatic sampler. [larger image]

---

Methods:

Card Sound Canal was instrumented with water-level, salinity, temperature, and velocity instruments. Velocity data was collected using an acoustic Doppler velocity meter (ADVM) and was calibrated with the use of an acoustic Doppler current profiler (ADCP) for the computation of discharge. The index velocity rating was developed following techniques described by Patino and others (1997), Hittle and others (2001), Morlock and others (2002), and Ruhl and others (2005). Salinity and temperature data were collected at two depths in the water column. A floating salinity probe was implemented for the near surface layer to help determine the presence of freshwater flow, bi-directional flow, and to examine potential effects of stratification on the acoustic signals through the tidal cycle.

Discharge data were computed using area and velocity ratings. The area rating was developed using from a manual cross-section measurement. A computer software program called Areacomp developed by the USGS was used to calculate the area rating (http://il.water.usgs.gov/adcp/). The index-velocity rating was developed through regression analyses by determining relations between instrument velocity (index velocity) and the mean cross-sectional velocity taken from ADCP measurements.

Initially an index velocity rating was developed using a side looking ADVM from 124 discharge measurements. The rating was utilized from October 1, 2003 to November 28, 2004 to which the bias from bi-directional flow was apparent (Fig. 2A, 2B). In the example below (Fig. 2A), positive velocities are observed from the surface to approximately 2.5 ft. of depth, whereas, zero to slightly negative velocities were observed from 2.5 ft. to approximately 8.5 ft. in depth. As a result, a more accurate index rating was developed using an up-looking ADVM and 156 discharge measurements (Fig. 2C).

Figure 2A: A contour map of bi-direction flow at Card Sound Canal. [larger image]

---

Figure 2B: Index velocity rating at Card Sound Canal using a side looking ADVM from October 1, 2003 to November 28, 2004. [larger image]

---

Figure 2C: Index velocity rating at Card Sound Canal using an up-looking ADVM from November 29, 2004 to current. [larger image]

---

The automatic sampler used multiple intakes along a vertical mount to collect water samples for nutrient analysis. In an attempt to collect a representative sample, the intakes were designed to sample water from both the top and the bottom layers of the stratified water column. Every 18 hours a 120 ml water sample was pumped into a pre-preserved (1:1 sulfuric acid) composite bottle that was made up of four pumping sessions over a 3-day period. Field visits to collect sample bottles occurred every 24 days. During site visits, point samples collected from the automatic sampler were used to assess the potential for intake fouling bias as a result of the estuarine environment in which it was deployed. Representative water samples were collected quarterly following a multiple vertical sampling approach. Samples were analyzed at the USGS National Water Quality Laboratory for total and dissolved nutrients. For more background on field methods or products please refer to the USGS South Florida Information Access (SOFIA) web Page (http://sofia.usgs.gov/exchange/patino/methodflow.html).

Results:

Card Sound Canal routinely experiences stratified conditions as well as bi-directional flow. Differences in salinity between the top layer of the water column and the bottom at this site are routinely greater than 15 ppt. This phenomenon has been observed over all four years of data collection at this site. Salinity at Card Sound Canal near the surface was less then 6 ppt approximately 36% of the time from October 1, 2006 to September 15, 2007 (Fig. 3A). This was lower then the average of 50% for all data collected. Average salinity from water year's 2004, 2005, and 2006 were less then 28 ppt 51%, 52%, and 67% of the time, respectively. Salinity near the bottom was less then 26 ppt approximately 33% of the time from October 1, 2006 to September 15, 2007 (Fig. 3A). This was lower then the average of 50% for all data collected. Average salinity from water year's 2004, 2005, and 2006 was less then 28 ppt 56%, 43%, and 66% of the time, respectively. Changes in residence time for lower salinities may be due to reduced flows, the relocation of the sensors during water year 2006, or a combination of these factors.

Figure 3A-B. Duration curves for discharge and salinity at Card Sound Canal. A comparison of the average salinity for all years of data collection vs. water year 2007 is presented. [click on images above for larger versions]

---

Continuous Monitoring Data:

Salinity and temperature stratification was documented by the continuous monitors on May 24, 2005 and coincided with ADCP measurements for index rating development (Fig. 4A-B). ADCP measurements began at approximately 2:00 PM and ended at approximately 7:00 PM. From approximately 10:00 AM to approximately 12:00 PM, the vertical temperature differences were 3.0 °C. Vertical salinity differences were ~ 29.0 ppt for the same period of time. Velocities reversed from positive to negative near 1:00 PM and the water column began to mix causing temperature and salinity values to converge between the top and bottom of the water column. Differences were minor but higher temperature and salinity values were observed as a result of water flowing from Barnes Sound up Card Sound Canal. At 5:00 PM, velocities began to change from negative to positive and as a result, salinity near the surface rapidly changed from approximately 30.0 ppt to 6.0 ppt. Temperature near the surface decreased more gradually than the surface salinity reaching conditions less than 28 °C by midnight.

Figure 4A-B: Continuous temperature, salinity, and velocity profile observed on May 24, 2005. [click on images above for larger versions]

---

Velocity Profile:

The velocity profile data from the bottom mounted ADVM for May 24, 2005 consists of 5 sampling cells at 2.4 ft. intervals and includes the mean velocity shown in black. Cell 1 is closest to the streambed while cell 5 is near the surface (Fig. 4B). The observed bi-directional flow from 9:15 AM to 11:30 AM results in a highly stratified water column (Fig. 4A). Mixing of the water column was observed only when the majority of the velocity profile was negative. The mean velocity was dependent on changes in tidal forcing, density gradients, wind speed and direction. The bottom mounted ADVM and the two sampling intakes provided the necessary data quality required for nutrient loading calculations.

Summary:

Point samples, 3-day composite samples from the automatic sampler, and multiple vertical samples were compared to determine differences between sampling techniques and to identify fouling issues (Fig. 5A-B). The differences between the point and the multiple vertical TKN sample was 0.1 mg/L while TP was more variable. The greatest difference between the point samples, 3-day composite and multiple vertical sample for TP were approximately 0.005 mg/L. The Everglades wetlands is an oligotrophic system with TP concentrations usually < 0.01 mg/L or below the detection limit (0.004 mg/L). No statistical comparison was conducted due to the small sample size. It should be noted the deviations in the 3-day composite samples are to be expected due to changes in the elevation of the boundary layer in relation to tide.

Figure 5A-B: Comparison of TKN and TP point samples, 3-day composites, and multiple vertical samples. [click on images above for larger versions]

---

Figure 5C-D: TKN and TP concentrations throughout water year 2007, Nov. 1, 2006 to current. [click on images above for larger versions]

---

Figure 5E-F: Monthly TKN and TP concentrations in relation to monthly discharge at Card Sound Canal. [click on images above for larger versions]

---

101 TP and TKN samples were collected at Card Sound Canal from November 1, 2006 to September 30, 2007. TP ranged from 0.005 mg/L to 0.055 mg/L with a median concentration of 0.01 mg/L. TKN ranged from 0.4 mg/L to 1.7 mg/L with a median concentration of 0.7 mg/L.

Actual total phosphorus and total nitrogen loads were computed as a product of the 3-day net creek discharges and the 3-day nutrient concentrations. The equation took the form:

L_i = Q_iC_i

where:
L_i was the 3-day total phosphorus or nitrogen load
Q_i was the 3-day net discharge
C_i was the 3-day composite phosphorous or nitrogen concentration

TP and TKN loading at Card Sound Canal from November 1, 2006 to September 30, 2007 equaled 0.19 and 14.4 metric tons, respectively. As restoration of the Florida Everglades continues, results described here will be critical for detecting changes over time.

---

Poster presented November 2007, at the USGS National Water Quality Workshop

Related information:

SOFIA Project: Freshwater Flows to Northeastern Florida Bay

SOFIA Project: Coastal Gradients of Flow, Salinity and Nutrients