Estimates of Nutrient Loads at West Highway Creek in Northeastern Florida Bay

W. Barclay Shoemaker, Mark Zucker and Paul Stumpner
U.S. Geological Survey, Ft Lauderdale, FL, USA

Coastal discharge, total phosphorus (TP) and total nitrogen (TKN) concentrations are continuously monitored by the U.S. Geological Survey (USGS) at West-Highway Creek in northeastern Florida Bay. Data are collected within the transition zone between upstream Everglades wetlands and downstream coastal estuaries, specifically, north of Long Sound near Key Largo. The product of net 3-day discharge and mean 3-day concentration sampling were used to compute TP and TKN loads between the wetlands and bay from the beginning of October 2003 through the end of September 2004 (water year 2004). Net loading was predominately from the wetlands into the bay (positive loading); the mean TP and TKN loads equaled about +0.1 and 8.0 kilograms per day (kg day^-1), respectively. As expected, wet season loads were larger than dry season loads because the loading signal mostly was explained by discharge and positive discharge increased during the wet season. An analysis also was made to determine the potential errors in total loading estimates if water-quality auto-samplers (collecting samples every 18 hours continuously) were replaced with grab samples collected at the end of each month. Although replacing the auto-samplers reduces costs, the errors in loading estimates were considerable. Specifically, monthly grab samples over and under-predicted the total annual TP and TKN loads by about 30% and -7%, respectively.

Regression-defined models were developed that predict TP and TKN loads solely as a function of discharge and time. The loading models were built for "gap-filling" so that annual loads could be estimated, and also for examining how loading will change under hypothetical watermanagement scenarios. The TP model reproduced about 80% of the actual load variability, with a mean absolute error of about 0.06 kg day^-1. The TKN model reproduced about 90% of the actual load variability, with a mean absolute error of about 2.6 kg day^-1. Upon gap-filling, estimated annual TP and TKN loads were about +38 and +3,127 kg, respectively, at West Highway Creek for the 2004 water year.

Loading changes were apparent under alternate water-management scenarios, specifically, retention, detention, and doubling the discharge from the wetlands into the Bay (Fig. 1A, B). The retention scenario captured net discharge greater than +40,000 cubic meters (m³), resembling a case where excess runoff is captured for underground aquifer storage and recovery as part of the Comprehensive Everglades Restoration Plan. The detention scenario also captured net discharge greater than +40,000 m³, but released the captured water at the later time so the total measured discharge was conserved. Doubling the positive coastal discharge from the wetlands into the Bay may resemble flow conditions upon the removal of upstream canals and levees, or construction of bridges that elevate roadways above upstream wetlands. Loading results from this scenario likely are less reliable than the results from the retention and detention scenarios, because doubling the positive discharge creates discharge values that are outside the range of values used for calibration.

Figure 1. Actual and estimated (A) TP and (B) TKN loads from scenario testing. [click on images above for larger version]

Scenario testing results were compiled into duration curves so results could be reported with probabilities that acknowledge uncertainty. In the actual case (flow unaltered) at West Highway Creek, a TP load of +0.15 kg day^-1 was exceeded 40% of the time. In the retention, detention, and doubling scenarios, the probability of exceeding a TP load of +0.15 kg day^-1 was 31%, 45%, and 52%, respectively. For TKN loads in the actual case, a value of +9.0 kg day^-1 was exceeded 51% of the time. In the retention, detention, and doubling scenarios, the probability of exceeding a TKN load of +9.0 kg day^-1 was 40%, 53%, and 55%, respectively. Future plans include application of the methods at North River upstream of the cutoff near Flamingo, in southwestern Florida.

Contact Information: W. Barclay Shoemaker, U.S. Geological Survey, 3110 SW 9th Ave, Ft Lauderdale, FL 33315, Cell: (305) 301-8334, Phone: (954) 377-5956, Fax: (954) 377-5901, Email: bshoemak@usgs.gov

(This abstract is from the 2005 Florida Bay and Adjacent Marine Systems Science Conference)

< Back to Project Homepage (Coastal Gradients of Flow, Salinity and Nutrients)

Related Project: Freshwater Flows to Northeastern Florida Bay