TMDL Report

Fecal Coliform TMDL for Horseshoe Creek

(WBID 1436)

Prepared by:

US EPA Region 4 61 Forsyth Street SW

Atlanta, Georgia 30303

March 2007

In compliance with the provisions of the Federal Clean Water Act, 33 U.S.C §1251 et. seq., as amended by the Water Quality Act of 1987, P.L. 400-4, the U.S. Environmental Protection Agency is hereby establishing the Total Maximum Daily Load (TMDL) for fecal coliforms in Horseshoe Creek, located in the Kissimmee River Basin. Subsequent actions must be consistent with this TMDL. ____________________/s/______________________ _____3/29/07____ James D. Giattina, Director Date Water Management Division

i

Acknowledgments

EPA would like to acknowledge that the contents of this report and the total maximum daily load (TMDL) contained herein were developed by the Florida Department of Environmental Protection (FDEP). Many of the text and figures may not read as though EPA is the primary author for this reason, but EPA is officially establishing the TMDL for fecal coliforms for Horseshoe Creek. EPA is establishing this TMDL in order to meet consent decree requirements pursuant to the Consent Decree entered in the case of Florida Wildlife Federation, et al. v. Carol Browner, et al., Case No. 98-356-CIV-Stafford. The Florida Department of Environmental Protection (Department) appreciates the kind help from Barry Jacobson (SWET) and Joyce Zhang (SFWMD) for providing the WAMView model set up for the Upper Kissimmee River Basin and their willingness to answer questions about the model. The Department would also like to acknowledge URS and CDM for assisting the Department for data collection and model evaluation. This study also benefited from the comments and suggestions on the possible sources of fecal pollution from Eric Pluchino and Robert Renk from the Central District Office, Steve Thompson and Phyllis James from the Southwest District Office, and Dannie Gibson from Polk County. Editorial assistance provided by Daryll Joyner, Jan Mandrup-Poulsen, and Linda Lord. For additional information on the watershed management approach and impaired waters in the Kissimmee River Basin, contact Pat Fricano Florida Department of Environmental Protection Bureau of Watershed Management Watershed Planning and Coordination Section 2600 Blair Stone Road, Mail Station 3565 Tallahassee, FL 32399-2400 Pat.Fricano@dep.state.fl.us Phone: (850) 245-8559; Suncom: 205-8559 Fax: (850) 245-8434 Access to all data used in the development of this report can be obtained by contacting Xueqing Gao Florida Department of Environmental Protection Bureau of Watershed Management Watershed Assessment Section 2600 Blair Stone Road, Mail Station 3555 Tallahassee, FL 32399-2400 Xueqing.Gao@dep.state.fl.us Phone: (850) 245-8464; Suncom: 205-8464 Fax: (850) 245-8536

ii

Contents

Chapter 1: INTRODUCTION___________________________________1

1.1 Purpose of Report ________________________________________________1

1.2 Identification of Waterbody ________________________________________1

1.3 Background _____________________________________________________1

Chapter 2: DESCRIPTION OF WATER QUALITY PROBLEM ________4

2.1 Statutory Requirements and Rulemaking History ______________________4

2.2 Information on Verified Impairment__________________________________4

Chapter 3. DESCRIPTION OF APPLICABLE WATER QUALITY STANDARDS AND TARGETS _______________________6

3.1 Classification of the Waterbody and Criteria Applicable to the TMDL______6

3.2 Applicable Water Quality Standards and Numeric Water Quality Target ____6

Chapter 4: ASSESSMENT OF SOURCES________________________7

4.1 Types of Sources _________________________________________________7

4.2 Potential Sources of Fecal Coliform in the Horseshoe Creek basin ________7 4.2.1 Point Sources ________________________________________________7

Municipal Separate Storm Sewer System Permittees ____________________________ 8 4.2.2 Land Uses and Nonpoint Sources ________________________________9

Land Uses _____________________________________________________________ 9 Source Assessment _____________________________________________________ 11

Chapter 5: DETERMINATION OF ASSIMILATIVE CAPACITY_______19

5.1 Determination of Loading Capacity _________________________________19 5.1.1 Data Used in the Determination of the TMDL _______________________19 5.1.2 TMDL Development Process ___________________________________20

Develop the Flow Duration Curve___________________________________________ 20 Develop the Load Duration Curves for Both the Allowable Load and Existing Loading Capacity _______________________________________________________ 23

5.1.3 Definition of the Critical Condition________________________________24 5.1.4 Establish the needed load reduction by comparing the existing

loading capacity to the allowable load under critical condition_________25

Chapter 6: DETERMINATION OF THE TMDL ____________________27

6.1 Expression and Allocation of the TMDL _____________________________27

iii

6.2 Load Allocation (LA) _____________________________________________28

6.3 Wasteload Allocation (WLA)_______________________________________28 6.3.1 NPDES Wastewater Discharges _________________________________28 6.3.2 NPDES Stormwater Discharges _________________________________28

6.4 Margin of Safety (MOS)___________________________________________28

Chapter 7: NEXT STEPS: IMPLEMENTATION PLAN DEVELOPMENT AND BEYOND _____________________30

7.1 Basin Management Action Plan____________________________________30

References 31

Appendices _______________________________________________32

Appendix A: Background Information on Federal and State Stormwater Programs__________________________________________________________32

iv

List of Tables

Table 2.1. Summary of Fecal Coliform Monitoring Data for Horseshoe Creek (WBID 1436) ................................................................................. 4

Table 4.1. Classification of landuse categories for the Horseshoe Creek basin ........................................................................................................ 9

Table 4.2. Total population in the Horseshoe Creek watershed ............................. 16 Table 4.3. Estimated Septic Tank Numbers and Failure Rates for Polk

County, 1998–2005 ............................................................................... 17 Table 5.1 Fecal coliform concentration at site 21FLPOLKHORSESHOE

CR2. ...................................................................................................... 20 Table 5.2. Allowable loads, existing loading capacities, and need load

reduction for critical flow conditions. ...................................................... 26 Table 6.1. TMDL Components for Fecal Coliform in Horseshoe Creek.................. 28

v

List of Figures

Figure 1.1: Location of Horseshoe Creek and major geopolitical features around these basins ................................................................................ 3

Figure 4.1. Principal Land Uses in basins that drains to Horseshoe Creek ............. 10 Figure 4.2. Spatial distribution of fecal coliform concentration in Horseshoe

Creek in 2004. The 30, 31, and 18 in the graph legend represent stations 21FLCEN 26010030, 21FLCEN 26010031, and 21FLCEN 26010018, respectively. ................................................. 12

Figure 4.3. Locations of water quality stations in the Horseshoe Creek ................. 13 Figure 4.4. Detailed locations of stations 21FLCEN 26010030 and

21FLCEN 26010031; the green triangle marks the location of the spray field for the Shady Oak MHP.Figure 4.5. Detailed location of station 21FLCEN 26010018 ................................................. 14

Figure 4.5. Detailed location of station 21FLCEN 26010018 ................................... 15 Figure 5.1. Locations of rainfall zones and weather stations used for the

Horseshoe Creek flow simulation. ......................................................... 22 Figure 5.2. Flow Duration Curve for Horseshoe Creek ............................................ 23 Figure 5.3. Load Duration Curves for Allowable Load and Existing Loading

Capacities of Fecal Coliform.................................................................. 24 ________________________________________________________________________

vi

Web sites

Florida Department of Environmental Protection, Bureau of Watershed Management

TMDL Program http://www.dep.state.fl.us/water/tmdl/index.htm Identification of Impaired Surface Waters Rule http://www.dep.state.fl.us/water/tmdl/docs/AmendedIWR.pdf STORET Program http://www.dep.state.fl.us/water/storet/index.htm 2002 305(b) Report http://www.dep.state.fl.us/water/docs/2002_305b.pdf Criteria for Surface Water Quality Classifications http://www.dep.state.fl.us/legal/rules/shared/62-302t.pdf Basin Status Report for the Tampa Bay Tributaries Basin http://www.dep.state.fl.us/water/tmdl/stat_rep.htm Allocation Technical Advisory Committee (ATAC) Report http://www.dep.state.fl.us/water/tmdl/docs/Allocation.pdf

U.S. Environmental Protection Agency

Region 4: Total Maximum Daily Loads in Florida http://www.epa.gov/region4/water/tmdl/florida/ National STORET Program http://www.epa.gov/storet/

vii

Chapter 1: INTRODUCTION

1.1 Purpose of Report

This report presents the Total Maximum Daily Load (TMDL) for fecal coliform bacteria in the Kissimmee River Basin. This stream was verified for fecal coliform impairment and therefore was included on the Verified List of impaired waters for the Kissimmee River basin that was adopted by Secretarial Order on May 12, 2006. The TMDL establishes the allowable fecal coliform loading to Horseshoe Creek that would restore the waterbody so that it meets its applicable water quality criteria for fecal coliform.

1.2 Identification of Waterbody

Horseshoe Creek is located in the northeastern part of Polk County (Figure 1.1). The northern part of the creek’s drainage basin extends into the northwestern Osceola County. The total length of the stream is about 11 miles. It drains to the south for the first 6 miles and then turns southeast and discharges to the Reedy Creek Swamp (Figure 1.1). The creek drops about 10 feet, from 35 feet above the see level at its headwater area to about 25 feet above the sea level at the inlet of the Reedy Creek Swamp, with an average slope of 0.07%. The city of Davenport is located to the south of the creek. The creek drains a predominantly agricultural area. Urbanized areas, especially some medium density residential areas, are primarily located in the southern part of the drainage basin. This is also where most of the bacteria samples were taken.

For assessment purposes, the Department has divided the Kissimmee River basin into water assessment polygons with a unique waterbody identification (WBID) number for each watershed or stream reach. This TMDL addresses WBID 1436, Horseshoe Creek – for fecal coliform.

1.3 Background

This report was developed as part of the Florida Department of Environmental Protection’s (Department) watershed management approach for restoring and protecting state waters and addressing TMDL Program requirements. The watershed approach, which is implemented using a cyclical management process that rotates through the state’s 52 river basins over a 5-year cycle, provides a framework for implementing the TMDL Program–related requirements of the 1972 federal Clean Water Act and the 1999 Florida Watershed Restoration Act (FWRA, Chapter 99-223, Laws of Florida).

A TMDL represents the maximum amount of a given pollutant that a waterbody can assimilate and still meet water quality standards, including its applicable water quality criteria and its designated uses. TMDLs are developed for waterbodies that are verified as not meeting their water quality standards. TMDLs provide important water quality restoration goals that will guide restoration activities.

1

This TMDL Report will be followed by the development and implementation of a Basin Management Action Plan, or BMAP, designed to reduce the amount of fecal coliform that caused the verified impairment of Horseshoe Creek. These activities will depend heavily on the active participation of local governments, businesses, and other stakeholders. The Department will work with these organizations and individuals to undertake or continue reductions in the discharge of pollutants and achieve the established TMDLs for impaired waterbodies.

2

Figure 1.1: Location of Horseshoe Creek and major geopolitical features around these basins

3

Chapter 2: DESCRIPTION OF WATER QUALITY PROBLEM

2.1 Statutory Requirements and Rulemaking History

Section 303(d) of the federal Clean Water Act requires states to submit to the U.S. Environmental Protection Agency (EPA) lists of surface waters that do not meet applicable water quality standards (impaired waters) and establish a TMDL for each pollutant causing impairment of listed waters on a schedule. The Department has developed such lists, commonly referred to as 303(d) lists, since 1992. The list of impaired waters in each basin, referred to as the Verified List, is also required by the FWRA (Subsection 403.067[4)] Florida Statutes [F.S.]), and the state’s 303(d) list is amended annually to include basin updates. Florida’s 1998 303(d) list included 25 waterbodies in the Kissimmee River basin. However, the FWRA (Section 403.067, F.S.) stated that all previous Florida 303(d) lists were for planning purposes only and directed the Department to develop, and adopt by rule, a new science-based methodology to identify impaired waters. After a long rulemaking process, the Environmental Regulation Commission adopted the new methodology as Chapter 62-303, Florida Administrative Code (F.A.C.) (Identification of Impaired Surface Waters Rule, or IWR), in April 2001.

2.2 Information on Verified Impairment

The Department used the IWR to assess water quality impairments in the Horseshoe Creek watershed and has verified that the stream is impaired for fecal coliform bacteria. The verification of impairment was based on the observations that 7 out of 31 fecal coliform samples collected during the verified period (January 1, 1998 through June 30, 2005) exceeded applicable fecal coliform water quality criteria (FAC 62-302). Table 2.1 summarizes the fecal coliform monitoring results for the verified period. Table 2.1. Summary of Fecal Coliform Monitoring Data for

Horseshoe Creek (WBID 1436)

Waterbody (WBID) Parameter Fecal Coliform

Total number of samples 31 IWR required number of exceedances for the verified list 6

Number of observed exceedances 7 Number of observed nonexceedances 24 Number of seasons during which samples were collected 4

Horseshoe Creek

Highest observation (MPN/100mL)* 20,000

4

Lowest observation (MPN/100 mL) 2 Median observation (MPN/100 mL) 235 Mean observation (MPN/100 mL) 913 FINAL ASSESSMENT Impaired

* Most probable number per 100 milliliters.

5

Chapter 3. DESCRIPTION OF APPLICABLE WATER QUALITY STANDARDS AND TARGETS

3.1 Classification of the Waterbody and Criteria Applicable to the TMDL

Florida’s surface waters are protected for five designated use classifications, as follows: Class I Potable water supplies Class II Shellfish propagation or harvesting Class III Recreation, propagation, and maintenance of a healthy, well-

balanced population of fish and wildlife Class IV Agricultural water supplies Class V Navigation, utility, and industrial use (there are no state

waters currently in this class)

Horseshoe Creek is a Class III waterbody, with a designated use of recreation, propagation, and maintenance of a healthy, well-balanced population of fish and wildlife. The criteria applicable to this TMDL are the Class III criteria for fecal coliform.

3.2 Applicable Water Quality Standards and Numeric Water Quality Target

Numeric criteria for bacterial quality are expressed in terms of fecal coliform bacteria concentrations. The water quality criteria for protection of Class III waters, as established by Chapter 62-302, F.A.C., state the following:

Fecal Coliform Bacteria: The most probable number (MPN) or membrane filter (MF) counts per 100 ml of fecal coliform bacteria shall not exceed a monthly average of 200, nor exceed 400 in 10 percent of the samples, nor exceed 800 on any one day.

The criteria state that monthly averages shall be expressed as geometric means based on a minimum of 10 samples taken over a 30-day period. During the development of load duration curves for the impaired stream (as described in subsequent chapters), there were insufficient data (fewer than 10 samples in a given month) available to evaluate the geometric mean criterion for fecal coliform bacteria. Therefore, the criteria selected for the TMDLs was not to exceed 400 MPN/100 mL in any sampling event for fecal coliform. The 10 percent exceedance allowed by the water quality criterion for fecal coliform bacteria was not used directly in estimating the target load, but was included in the TMDL margin of safety (as described in subsequent chapters).

6

Chapter 4: ASSESSMENT OF SOURCES

4.1 Types of Sources

An important part of the TMDL analysis is the identification of pollutant source categories, source subcategories, or individual sources of pollutants in the impaired waterbody and the amount of pollutant loadings contributed by each of these sources. Sources are broadly classified as either “point sources” or “nonpoint sources.” Historically, the term point sources has meant discharges to surface waters that typically have a continuous flow via a discernable, confined, and discrete conveyance, such as a pipe. Domestic and industrial wastewater treatment facilities (WWTFs) are examples of traditional point sources. In contrast, the term “nonpoint sources” was used to describe intermittent, rainfall driven, diffuse sources of pollution associated with everyday human activities, including runoff from urban land uses, agriculture, silviculture, and mining; discharges from failing septic systems; and atmospheric deposition.

However, the 1987 amendments to the Clean Water Act redefined certain nonpoint sources of pollution as point sources subject to regulation under the EPA’s National Pollutant Discharge Elimination (NPDES) Program. These nonpoint sources included certain urban stormwater discharges, including those from local government master drainage systems, construction sites over five acres, and a wide variety of industries (see Appendix A for background information on the federal and state stormwater programs).

To be consistent with Clean Water Act definitions, the term “point source” will be used to describe traditional point sources (such as domestic and industrial wastewater discharges) and stormwater systems requiring an NPDES stormwater permit when allocating pollutant load reductions required by a TMDL (see Section 6.1). However, the methodologies used to estimate nonpoint source loads do not distinguish between NPDES stormwater discharges and non-NPDES stormwater discharges, and as such, this source assessment section does not make any distinction between the two types of stormwater.

4.2 Potential Sources of Fecal Coliform in the Horseshoe Creek basin

4.2.1 Point Sources

There are no NPDES permitted facilities located in the basin. However, there are some permitted facilities within the basin that are authorized to discharge to ground water under state permits. These facilities can be divided into three different categories:

(1) Sand mines: These facilities all belong to the Rinker Materials Corporation-Davenport Plant, which include FLG110456, FLG110347, and FLA013244. These facilities are all located to the east of the creek and are not considered significant fecal pollution contributors.

(2) Food processing facility: Only 1 facility belongs to this category, which is Holly Hill Fruit Products Co. Inc (FLA013169). This is a citrus processing plant and wastewater from

7

the facility is treated through screening and land application. The facility is not expected to produce a significant amount of fecal coliform bacteria.

(3) Wastewater treatment facilities: Four facilities belong to this category, which include

Shady Oaks MHP (FLA013023), Bishop Group Inn WWTF (FLA013057), Davenport Elementary School WWTP (FLA012965), and Center Crest RVP WWTF (FLA013110). Three of these facilities, including Bishop Group Inn WWTF, Davenport Elementary School WWTP, and Center Crest RVP WWTF apply the extended aeration to treat the wastewater and all the wastewater is treated in two rapid percolation/evaporation ponds. In addition, these three facilities are located about a mile away from the Horseshoe Creek. Fecal coliform produced by these facilities should not pose serious threat on the water quality of the creek.

Shady Oaks MHP is different from the other three domestic wastewater facilities. It is located on the east bank of the creek and receives wastewater from a mobile home park (Latitude 28010’31”, Longitude 81035’49”). The facility has a 0.010 MGD 3-month Average Daily Flow (3MADF). The wastewater is treated by a Type III, extended aeration domestic wastewater treatment plant, consisting of three aeration basins with a total volume of 11,440 gallons, one clarifier of 1,800 gallons volume and 96 square feet of total surface area, one chlorine contact chamber of 500 gallons volume, and one digester of 825 gallons volume. The plant is operated to provide secondary treatment with basic disinfection. The treated wastewater is disposed at a 16,600 square foot sprayfield located on the west bank of Horseshoe Creek (Latitude 28010’30”, Longitude 81035’50”, location shown in Figure 4.4). The facility has a fecal coliform effluent limit at the discharge point to the spray field (the effluent limit is the same as is defined in the Florida Surface Water Quality Standard, FAC 62-302) and is required by the permit to monitor the fecal coliform concentration at the discharge point to the sprayfield. Based on the facility’s discharge monitoring report in the period from 2000 through 2005, no exceedance of the effluent limit were observed before 2005. Two exceedances were observed in 2005. Although no exceedance was observed at the discharge point to the sprayfield in 2004, as the facility was not required to monitor the fecal coliform concentration in the stormwater runoff from the sprayfield and the sprayfield is very close to Horseshoe Creek, its influence on the fecal coliform concentration of the creek needs to be carefully examined. In fact, the fecal coliform concentrations in Horseshoe Creek were significantly higher at the station downstream of this facility than the concentrations at the station upstream of this facility.

Municipal Separate Storm Sewer System Permittees The stormwater collection systems owned and operated by Polk County and the City of Davenport in the Horseshoe Creek basin are covered by a Phase I NPDES municipal separate storm sewer system (MS4) permit. Polk County is the “unofficial” lead permittee. The Department of Transportation is a co-permitee to this permit.

8

4.2.2 Land Uses and Nonpoint Sources

Land Uses The spatial distribution and acreage of different land use categories were identified using the South West Florida Water Management District (SWFWMD)’s year 1999 land use coverage (scale 1:40,000) contained in the Department’s geographic information system (GIS) library. Land use categories in the watershed were aggregated using the simplified Level 1 codes and tabulated in Table 4.1. Figure 4.1 shows the acreage of the principal land uses in the watershed.

As shown in Table 4.1, the dominant land use category in the basin is agriculture, which accounts for about 40 percent of the total basin area. About 3,562 acres of the basin is occupied by urban and built-up area, which accounts for about 22% of the total basin area. Natural landuse areas, which include water/wetlands and upland forest, occupy about 5,837 acres, accounting for about 35 percent of the total basin area.

Table 4.1. Classification of landuse categories for the Horseshoe Creek basin

Level 1 Code Land Use Acreage Percent Acreage

1000 Urban and Build-Up 1160 7.0% Low-density residential 735 4.4% Medium-density residential 970 5.9% High-density residential 289 1.7%

2000 Agriculture 6633 40.0% 3000 Rangeland 374 2.3% 7000 Barren land 154 0.9% 8000 Transportation, communication, and utilities 408 2.5% 4000 Forest/rural open 872 5.3%

5000/6000 Water/wetland 4965 30.0% TOTAL 16559 100.0%

9

Figure 4.1. Principal Land Uses in basins that drains to Horseshoe Creek

10

Source Assessment Because there was no point sources identified in the watershed that directly discharge to Horseshoe Creek, the primary loadings of fecal coliform into the creek are likely generated by nonpoint sources or MS4-permitted areas in the watershed. Nonpoint sources of coliform bacteria generally, but not always, come from the coliform bacteria that accumulate on land surfaces and wash off as a result of storm events, and the contribution delivered via ground water from sources such as failed septic tanks and the improper land application of domestic wastewater residuals. Typical nonpoint sources of coliform bacteria include the following:

• Wildlife,

• Agricultural animals,

• Pets in residential areas,

• Onsite sewage treatment and disposal systems (septic tanks),

• Land application of domestic wastewater residuals, and

• Urban development (outside of Phase I or II MS4 discharges).

No data were available to specifically identify and quantify the major source(s) of fecal coliform bacteria in the Horseshoe Creek watershed. However, the spatial distribution of the fecal coliform concentration in the creek does shed light on what might be the possible sources. Figure 4.2 shows the spatial distribution of fecal coliform concentrations in Horseshoe Creek. Fecal coliform samples were collected from six water quality stations located in the Horseshoe Creek basin (Figure 4.3). These stations include 21FLCEN 26010030, 21FLCEN 26010031, 21FLCEN 26010018, 21FLGW 7862, 21FLGW 7873, and 21FLPOLKHORSESHOE CR 2. About 66% of the samples were taken from 21FLCEN 26010030, 21FLCEN 26010031, and 21FLCEN 26010018. All the samples taken from these three sites were from sampling events in 2004. Samples were typically taken from all three stations and therefore provided a basis for determining the spatial pattern. Only 1 sample was collected at stations 21FLGW 7862 and 21FLGW 7873, and they were taken in 2000. Therefore, data from these two stations were not included in the spatial pattern analysis. The remaining about 30% of the samples were taken from station 21FLPOLKHORSESHOE CR 2. Some of these samples were taken in 2004. However, none of the samples from this site were taken at the same date as those collected from 21FLCEN 26010030, 21FLCEN 26010031, and 21FLCEN 26010018. Therefore, data from 21FLPOLKHORSESHOE CR 2 were not included in the spatial pattern analysis either. Stations 21FLCEN 26010030, 21FLCEN 26010031, and 21FLCEN 26010018 are located in Horseshoe Creek in an upstream to downstream order (Figure 4.3). Station 21FLCEN 26010030 is the upstream most site and is located immediately upstream of a medium density residential area (Figure 4.4). As is shown in Figure 4.2, fecal coliform concentrations from this site never exceeded the 400 cfu/100 ml water quality criteria. Station 21FLCEN 26010031 is located about 350 meters downstream of station 21FLCEN 26010030, and is also downstream of the medium density residential area. As is shown by Figure 4.2, in the majority of sampling events, fecal coliform concentrations from station 21FLCEN 26010031

11

12

Figure 4.2. Spatial distribution of fecal coliform concentration in Horseshoe Creek in 2004. The 30, 31, and 18 in the graph legend represent stations 21FLCEN 26010030, 21FLCEN 26010031, and 21FLCEN 26010018, respectively.

0

100

200

300

400

500

600

700

800

900

1000

1 3 5 6 8 10 11

Months in 2004

Feca

l col

iform

con

cent

ratio

n (c

fu/1

00 m

l)

30 31 18 Criteria

were higher than those from station 21FLCEN 26010030, which suggested the possible contribution of the fecal coliform loads from the medium density residential area. This medium density residential area is also where the Shady Oaks MHP is located. As described in the previous section, this facility is located on the east bank of the creek. The treated domestic wastewater from this facility is applied to a spray field area. Fecal coliform contributions from the residential area and the facility deserve further investigation. The highest fecal coliform concentrations for all the sampling events in 2004 were always observed at station 21FLCEN 26010018 (Figure 4.5). This is the downstream most site in the creek and is located downstream of low density residential areas and cropland. Based on the information from the Polk County Environmental Affairs, most residences in the Horseshoe Creek basin are on septic tanks, and failure of septic tanks could be an important source of fecal coliform pollution in the creek.

Figure 4.3. Locations of water quality stations in the Horseshoe Creek

13

Figure 4.4. Detailed locations of stations 21FLCEN 26010030 and 21FLCEN 26010031; the green triangle marks the location of the spray field for the Shady Oak MHP.

14

Figure 4.5. Detailed location of station 21FLCEN 26010018

15

A rough estimate of fecal coliform loads from failed septic tanks in the Horseshoe Creek watershed can be made using Equation 1: L = 37.85* N * Q * C * F (1) Where:

L is the fecal coliform daily load (counts/day), N is the total number of septic tanks in the studied area (septic tanks), Q is the discharge rate for each septic tank, C is the fecal concentration for the septic tank discharge, and F is the septic tank failure rate.

The total number of septic tanks (N) existing in the Horseshoe Creek basin was not available to the Department at the time this TMDL was developed. Therefore, the number of septic tanks was calculated using the following method:

(1) Assuming each housing unit has only one septic tank, the total number of septic tanks equals the total number of housing units in the Horseshoe Creek Basin.

(2) As the Department did not have the exact number of housing units in the basin, the number of housing unit was calculated by dividing the total population of the watershed by the average household size.

(3) The total population of the basin was calculated as the sum of the population in each census tract in the basin. Based on the Department GIS library, the Horseshoe Creek basin covers parts of four census tracts, which include tracts 124.02, 408, 125.01, and 125.02. The area of these census tracts, population densities of each tract, total population of each tract, and the total population in the Horseshoe Creek basin are listed in Table 4.2.

Table 4.2. Total population in the Horseshoe Creek watershed

Census Tracts

Population Density (people/square miles)

Area of the tract (square miles)

Total population (people)

124.02 73.2 2.71 199 408 120.1 2.61 313

125.01 334.6 16.93 5664 125.02 121.7 3.60 438 Total 25.84 6613

(4) The average household size for Polk County for year 2000 was about 2.52

people/household based on information from US Census Bureau. (5) The total number of housing units, and also the total number of septic tanks located in

the Horseshoe Creek (N), was then equal to the total population in the watershed (6613 people) divided by the average household size (2.52 people/household) and equals 2,624 houses (septic tanks).

The discharge rate from each septic tank (Q) was calculated by multiplying the average household size by the per-capita wastewater production rate per day. A commonly cited value for per-capita wastewater production rate is 70 gallons/day/person (USEPA, 2001).

16

The commonly cited fecal coliform concentration (C) for septic tank discharge is 1x106 counts/100mL (USEPA, 2001). In this TMDL, the Department assumes that only failed septic tanks contribute fecal coliform pollution to Horseshoe Creek. No measured septic tank failure rate was available for the watershed at the time this report was prepared. Therefore the failure rate was derived from the number of septic tanks and septic tank repair permits for all of Polk County published by the Florida Department of Health (FDOH) (http://www.doh.state.fl.us/environment/ostds/statistics/ostdsstatistics.htm). The number of septic tanks in the county was calculated based on the assumption that none of the installed septic tanks would be removed after installation (Table 4.3). The reported number of septic tank repair permits was also obtained from the FDOH Web site (Table 4.3). Based on this information, a discovery rate of failed septic tanks for each year between 1998 and 2005 was calculated and listed in Table 4.3. The table shows that the annual septic tank failure discovery rate for Polk County from 1998 to 2005 averaged about 1.21 percent. Assuming that failed septic tanks are not discovered for about 5 years, the estimated annual septic tank failure rate is about 5 times the discovery rate, or 6.07 percent. Based on Equation 1, the estimated fecal coliform loading from failed septic tanks in the watershed is about 1.03 x 1012 counts/day. It should be noted that 1.03 x 1012 counts/day represents the potential fecal coliform load that can be generated in the Horseshoe Creek basin. Due to the attenuation in the pollutant transport, the actual fecal coliform loading that eventually reaches Horseshoe Creek could be significantly lower than this number. Table 4.3. Estimated Septic Tank Numbers and Failure Rates for

Polk County, 1998–2005

1998 1999 2000 2001 2002 2003 2004 2005 Average New installation (septic tanks) 1,865 1,784 1,191 1,177 1,079 1,252 1,396 1,333 1,385

Accumulated installation

(septic tanks) 104,969 106,753 107,944 109,121 110,200 111,452 112,848 114,181 109,684

Repair permit (septic tanks) 1300 1438 1322 1326 1278 1439 1299 1238 1,330

Failure discovery rate 1.24% 1.35% 1.22% 1.22% 1.16% 1.29% 1.15% 1.08% 1.21%

Failure rate* 6.19% 6.74% 6.12% 6.08% 5.80% 6.46% 5.76% 5.42% 6.07%

* The failure rate is 5 times the failure discovery rate.

In highly urbanized areas where a large percentage of the basin lands are residential areas, pets such as dogs and cats could become important sources of fecal coliform bacteria (Lim and Oliveri 1982, Trial et al. 1993). However, in Horseshoe Creek basin, the percent basin area occupied by residences is relatively low (Table 4.1). The majority of the basin areas are pervious areas and the opportunity at which fecal coliform bacteria from pet feces being flushed into the receiving water is relatively low. Therefore, fecal coliform contribution from pets was not considered in this study. In addition, based on the information from the Polk County Environmental Affairs, except for some small package plants, the Horseshoe Creek basin is largely unsewered. Therefore, fecal coliform contribution from the sewer line leakage was not considered important in this study.

17

Wildlife is another possible source of fecal coliforms in the Horseshoe Creek basin. As shown in Figure 4.1, there are wetland areas along the creek, and these areas are likely habitats for small wildlife, such as rabbits and raccoons. However, as the fecal coliform concentrations at the most upstream site seldom exceeded water quality criteria, the contribution of wildlife to the fecal coliform pollution probably is not significant.

18

Chapter 5: DETERMINATION OF ASSIMILATIVE CAPACITY

5.1 Determination of Loading Capacity

The methodology used for this TMDL is the load duration method. Also known as the “Kansas Approach” because it was developed by the state of Kansas, this method has been well documented in the literature, with improved modifications used by EPA Region 4. Basically, the method relates the pollutant concentration to the flow of the stream, in order to establish the existing loading capacity and the allowable pollutant load (TMDL) under a spectrum of flow conditions. It then determines the maximum allowable pollutant load and load reduction requirement based on the analysis of the critical flow conditions. This method requires four steps to develop the TMDL and establish the required load reduction:

1. Develop the flow duration curve, 2. Develop the load duration curve for both the allowable load and existing loading, 3. Define the critical conditions, and 4. Establish the needed load reduction by comparing the existing loading with the allowable

load under critical conditions.

5.1.1 Data Used in the Determination of the TMDL

Fecal coliform concentrations and flow measurements were required to estimate both the allowable pollutant load and existing loading to Horseshoe Creek. Figure 4.3 shows the locations of the water quality stations from which fecal coliform data were collected. Fecal coliform data collected during the Verified Period (1998 through 2006) were used in this study. A total of 32 fecal coliform samples were collected from 6 sampling stations in Horseshoe Creek. Data used for this TMDL report were mainly provided by the Department’s Central District office and Polk County. Table 2.1 provides a statistical summary of fecal coliform measurements in Horseshoe Creek.

19

Samples were only taken systematically in 2004 from the creek. Fecal coliform data were collected in other years as well, especially from station 21FLPOLKHORSESHOE CR2, from which data were collected in 2002, 2003, 2004, and 2005. However, most of these samples were collected from discrete sampling events instead of from the whole year. Therefore, it is difficult to establish a robust seasonal pattern based on only one year worth of data. Despite the lack of a multi-year data set, Figure 4.2 does show some rough seasonal trends. For example, although the fecal coliform concentration at 21FLCEN 26010018 exceeded criteria in January and decreased slightly in March, there is a general trend that the concentration increased toward the middle of the year and then decreased toward the end of the year. This trend is shared by data collected from site 21FLCEN 26010031, likely to be dictated by the similarity of the source around these two sites (most likely residential areas) and driven by the rainfall pattern. The data from site 21FLCEN 26010030 does not have a clear seasonal pattern. Data from this site never exceeded the water quality criteria for the period of record, which likely reflect a natural condition.

Table 5.1 shows the data collected from site 21FLPOLKHORSESHOE CR2. As shown in Figure 4.3, this site is very close to site 21FLCEN 26010030, which is upstream of the Shady Oak MHP. The majority of the samples collected from this site did not exceed the 400 counts/100 ml water quality target except for two exceedances observed in February of 2003 and August of 2004. Other than that, there is no clear season pattern for the fecal coliform concentration at this site. Table 5.1 Fecal coliform concentration at site 21FLPOLKHORSESHOE CR2.

Year Month Day Concentration (counts/100 ml)

2002 2 5 2 2002 8 13 360 2003 2 17 670 2003 8 12 120 2004 2 10 260 2004 4 20 50 2004 8 17 20000 2004 10 13 220 2005 3 3 200 2005 6 21 380

It should be noted that one data point, which is the fecal coliform concentration measured at site 21FLPOLKHORSESHOE CR2 on August 17, 2004, was excluded from the load duration analysis. This data point represents a 20,000 counts/100 ml, which is at least an order of magnitude higher than all the rest of the data measured in Horseshoe Creek. Even if the quality of this data point is reliable, it does not represent the typical existing extent of the fecal coliform pollution in Horseshoe Creek, therefore was not included in the calculation of the required percent pollutant load reduction. No flow measurements covering the time period from which water quality samples were collected were available at the time this TMDL was developed. The only gauging station located in the Horseshoe Creek, USGS 02266700, only has data for the period of 1965 to 1966. Therefore, the flow data used in this TMDL were simulated using a calibrated Watershed Assessment Model based on ArcView (WAMView) for the Upper Kissimmee River basin. Application of the Upper Kissimmee Basin WAMView was a joint effort from the JGH Engineering, Soil and Water Engineering Technologies (SWET), Inc., HDR Engineering, Inc., and South Florida Water Management District (SFWMD) to control the phosphorus loading into Lake Okeechobee from the upper basin (JGH Engineering, 2005).

5.1.2 TMDL Development Process

Develop the Flow Duration Curve The first step in the development of load duration curves is to create flow duration curves. A flow duration curve displays the cumulative frequency distribution of daily flow data over the period of record. The duration curve relates flow values measured at a monitoring station to the percent of time the flow values were equalled or exceeded. Flows are ranked from low, which

20

are exceeded nearly 100 percent of the time, to high, which are exceeded less than 1 percent of the time. As no measured flows were available at the time this TMDL was developed. Daily flow of Horseshoe Creek was simulated using the WAMView model calibrated for the Upper Kissimmee River Basin (JGH Engineering, 2005). WAM (Watershed Assessment Model) is a water quality assimilation model that was developed in the mid-1990’s based on similar models written by Dr. Del Bottcher of SWET. The first version of WAM was written for the ARC/INFO Unix workstation platform. This version was written using ESRI’s AML programming language. WAM was one of the first water quality assessment models that used the ARC/INFO component called GRID. GRID provided the ability to perform computations using rasters on a cell-by-cell basis. This resulted in output at a much smaller scale (typically 1 ha) than could be provided using polygons. The concept of WAM is to overlay land use, soils and rainfall zones to produce unique combinations that are submitted to three submodels – Groundwater Loading Effects of Agricultural Management Systems (GLEAMS; Knisel, 1993), Everglades Agricultural Area Model (EAAMod; Botcher et al., 1998; SWET, 1999), and a special-case model written specifically for WAM to handle wetland and urban landscapes. These submodels are bundled into one program called BUCShell, where BUC stands for Basin Unique Cell. BUCShell uses parameter dataset files for land use, soils and rainfall to produce estimates of surface runoff and pollutant loadings per hectare for each unique combination. The results are then spatially distributed onto the watershed by relating the output table to a grid of the unique combinations. The resulting grids represent the pollutant loadings at the source in both surface water runoff and ground water. The source runoff load is then attenuated via overland flow to the nearest downstream feature – stream, wetland or depression. At the stream, the attenuated surface runoff load is combined with the attenuated groundwater load and routed through the stream network where it is further attenuated. Dynamic routing of flows is accomplished within BLASRoute through the use of an algorithm that efficiently solves Manning’s equations (Jacobson et al., 1998). BLASRoute handles the entire attenuation process from the source cells to the basin outlet and provides daily outputs from any stream in the network. For detailed information regarding the WAMView model setup and calibration, please consult the report prepared by JGH Engineering (2005) regarding the development of the Upper Kissimmee Basin WAMView for phosphorus budget analyses. The Upper Kissimmee Basin WAMView simulates the stream flow and pollutant loadings for the period between 1990 and 2000. To simulate the stream flow that covers the time period in which fecal coliform samples were collected (2000 through 2005), rainfall input for the model needed to be updated. Horseshoe Creek is located in an area covered by two rainfall zones defined by two weather stations, including the one located in Lake Marion of Polk County (Latitude 280558, Longitude 812654) and the one located in the City of Kissimmee (Latitude 281726, Longitude 812654). The rainfall data collected at the Lake Marion station ended in 2001. For the period of 2002 to 2005, rainfall data from the Lake Marion station were substituted with the rainfall from the City of Kissimmee station. Figure 5.1 shows the locations of Horseshoe Creek, the rainfall zones covered by the two weather stations, and the two weather stations.

21

Figure 5.1. Locations of rainfall zones and weather stations used for the Horseshoe Creek flow simulation.

22

Daily stream flows for Horseshoe Creek were simulated at the outlet of the creek discharging into the Reedy Creek Swamp. The flow duration curve was created by using the percentile function and estimated flow to generate the flow at a given duration interval. For example, at the 90th duration interval, the percentile function calculates the flow that is equal or exceeded 90 percent of the time. Figure 5.2 shows the flow duration curve for Horseshoe Creek generated from the flow simulated using WAMView. Flows toward the right side of the plot are exceeded in greater frequency and are indicative of low-flow conditions. Flows on the left side of the plot represent high flows and occur less frequently.

Figure 5.2. Flow Duration Curve for Horseshoe Creek

0.0

20.0

40.0

60.0

80.0

100.0

120.0

140.0

160.0

180.0

200.0

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Flow duration interval

Flow

(cfs

)

Develop the Load Duration Curves for Both the Allowable Load and Existing Loading Capacity Flow duration curves are transformed into load duration curves by multiplying the flow values along the flow duration curve by the fecal coliform concentration and the appropriate conversion factors. The final results of the load are typically expressed as MPN per day. The following equations were used to calculate the allowable loads and the existing loading: Allowable load = (observed flow) x (state criteria) (2) Existing loading = (observed flow) x (coliform measurement) (3)

23

On the load duration curve, allowable and existing loads are plotted against the flow duration ranking. The allowable load was calculated based on the water quality criterion and flow values from the flow duration curve, and the line drawn through the data points representing the allowable load is called the target line. The existing load is based on the in-stream fecal coliform concentration measured during ambient monitoring and an estimate of flow in the

24

stream at the time of sampling. As noted previously, because insufficient data were collected to evaluate the fecal coliform geometric mean, 400 MPN/100mL was used as target criteria for fecal coliform. Figures 5.3 shows both the allowable load and the existing loads over the flow duration ranking for Horseshoe Creek. The points of the existing load that were higher than the allowable load at a given flow duration ranking were considered an exceedance of the criteria.

Figure 5.3. Load Duration Curves for Allowable Load and Existing Loading Capacities of Fecal Coliform

1.00E+09

1.01E+11

2.01E+11

3.01E+11

4.01E+11

5.01E+11

6.01E+11

7.01E+11

8.01E+11

0 10 20 30 40 50 60 70 80 90 100

Flow duration interval (%)

Feca

l Col

iform

Loa

d (c

ount

s/da

y)

Existing Load Exceedence Load Target Load

High Low Moist Medium Dry

5.1.3 Definition of the Critical Condition

The critical condition for coliform loadings in a given watershed depends on the existence of point sources and landuse pattern of the watershed. Typically, the critical condition for nonpoint sources is an extended dry period followed by a rainfall runoff event. During the wet weather period, coliform bacteria built up on the land surface under the dry weather condition is washed off by rainfall, resulting in the wet weather exceedence. However, significant nonpoint souce contributions could also appear under the dry weather condition without any major surface runoff event. This usually happens when the nonpoint sources contaminate the surfacial aquifer and the fecal coliform bacteria are brought into the receiving waters through baseflow. The

critical condition for point source loading typically occurs during periods of low stream flow when dilution is minimized. To characterize the critical condition, this study divided the flow duration into several intervals including:

1. High flow (0 to 10%) 2. Moist (10% to 40%) 3. Medium (40% to 60%) 4. Dry (60% to 90%) 5. Low (90% to 100%)

As shown in Figure 5.3, exceedences of the fecal coliform criterion in Horseshoe Creek mostly appeared under medium and dry conditions. In general, exceedences on the right side of the curve occur during low flow events, which implies the contribution from either point source or baseflow. As was discussed in the previous section, this observation appears to be consistent with a discharge from the Shady Oak MHP sprayfield site and the septic tanks used by the low density residential areas immediately upstream of site 21FLCEN 26010018 (Figure 4.5).

5.1.4 Establish the needed load reduction by comparing the existing loading

capacity to the allowable load under critical condition

The fecal coliform load reduction required to achieve the water quality criteria was established by comparing the existing loading capacity to the allowable load under the critical conditions defined in the previous section. The actual needed load reduction was calculated using the following equation:

%100_

___ ×−

=loadingExisting

loadingAllowableLoadingExistingreductionload Where, the Existing_loading of a given flow duration interval was calculated as the median value of all the available existing loading capacities within the flow duration interval that exceeded the allowable load. If only one exceedence appears in a certain flow duration interval, the lone exceeding value will be used as the Existing_Loading for the duration interval. The Allowable_loading of a given flow duration interval was calculated as the medium of all the allowable loads within the flow duration interval in 5% increments. Table 5.2 listed the flow duration intervals, allowable loading with 5% increment, the medium of the allowable loadings for given flow duration intervals, each individual existing loading capacities, the median values of all the existing loading capacities for different flow duration intervals, and needed load reduction under the critical conditions.

25

Table 5.2. Allowable loads, existing loading capacities, and need load reduction for

critical flow conditions.

Flow Duration Interval

Flow Ranking

(%)

Allowable Load

(MPN/day)

Existing loading capacities that exceeded the

allowable loading (MPN/day)

Allowable load for each flow duration

interval** (MPN/day)

Existing loading for each flow duration

interval*** (MPN/day)

Percent Reduction Required

5 1.77E+12 High 10 9.30E+11 ---- 1.77E + 12 ---- ----

15 5.59E+11 20 3.79E+11 25 2.82E+11 30 2.20E+11 35 1.74E+11

Moist

40 1.41E+11

---- 2.51E + 11 ---- ----

45 1.17E+11 50 9.62E+10 55 7.56E+10 Medium

60 5.91E+10

1.69E+11 (48.9%)* 8.59E + 10 1.69E + 11 49%

65 4.50E+10 70 3.65E+10 75 2.99E+10 80 2.31E+10 85 1.68E+10

Dry

90 1.06E+10

5.76E+10 (61.4%)* 6.10E+10 (61.4%)* 7.14E+10 (71.8%)* 1.69E+11 (71.8%)* 3.28E+10 (77.4%)*

2.65E + 10 6.10E + 10 57%

95 4.85E+09 Low

100 0.00E+00 ---- 2.37E + 09 ---- ----

Median 8.59E + 10 53%****

Note: * The numbers in the parentheses are the flow ranking for each existing load that

exceeded the allowable load. ** The allowable load for each flow duration interval was calculated as the midpoint for all

the allowable loads in the flow duration interval with 5% increments. The TMDL was calculated as the average allowable load for midpoints of medium flow and dry condition intervals.

*** The existing loading capacity for each flow duration interval was calculated as the median values of all the existing loading capacities in the flow duration interval that exceed the target loading.

**** The median allowable load (TMDL) and the median needed load reduction were calculated as the median of the allowable load and needed load reduction in all flow duration intervals.

26

Chapter 6: DETERMINATION OF THE TMDL

6.1 Expression and Allocation of the TMDL

The objective of a TMDL is to provide a basis for allocating acceptable loads among all of the known pollutant sources in a watershed so that appropriate control measures can be implemented and water quality standards achieved. A TMDL is expressed as the sum of all point source loads (Waste Load Allocations, or WLAs), nonpoint source loads (Load Allocations, or LAs), and an appropriate margin of safety (MOS), which takes into account any uncertainty concerning the relationship between effluent limitations and water quality:

TMDL = ∑ WLAs + ∑ LAs + MOS

As discussed earlier, the WLA is broken out into separate subcategories for wastewater discharges and stormwater discharges regulated under the NPDES Program:

TMDL ≅ ∑ WLAswastewater + ∑ WLAsNPDES Stormwater + ∑ LAs + MOS

It should be noted that the various components of the revised TMDL equation may not sum up to the value of the TMDL because a) the WLA for NPDES stormwater is typically based on the percent reduction needed for nonpoint sources and is also accounted for within the LA, and b) TMDL components can be expressed in different terms (for example, the WLA for stormwater is typically expressed as a percent reduction, and the WLA for wastewater is typically expressed as mass per day).

WLAs for stormwater discharges are typically expressed as “percent reduction” because it is very difficult to quantify the loads from MS4s (given the numerous discharge points) and to distinguish loads from MS4s from other nonpoint sources (given the nature of stormwater transport). The permitting of stormwater discharges also differs from the permitting of most wastewater point sources. Because stormwater discharges cannot be centrally collected, monitored, and treated, they are not subject to the same types of effluent limitations as wastewater facilities, and instead are required to meet a performance standard of providing treatment to the “maximum extent practical” through the implementation of BMPs.

This approach is consistent with federal regulations (40 CFR § 130.2[I]), which state that TMDLs can be expressed in terms of mass per time (e.g., pounds per day), toxicity, or other appropriate measure. TMDLs for Horseshoe Creek are expressed in terms of MPN/day and percent reduction, and represent the maximum daily fecal coliform load the stream can assimilate and maintain the fecal coliform criterion (Table 6.1).

27

Table 6.1. TMDL Components for Fecal Coliform in Horseshoe Creek

WLA

Parameter TMDL (colonies/day) Wastewater

(colonies/day)

NPDES Stormwater

(percent reduction)

LA (percent

reduction) MOS

Fecal coliform 8.59 x 1010 N/A 53% 53 % Implicit

6.2 Load Allocation (LA)

Based on a loading duration curve approach similar to that developed by Kansas (Stiles, 2002), the load allocation is a 53 percent reduction in fecal coliforms from nonpoint sources. It should be noted that the LA includes loading from stormwater discharges regulated by the Department and the water management districts that are not part of the NPDES Stormwater Program (see Appendix A).

6.3 Wasteload Allocation (WLA)

6.3.1 NPDES Wastewater Discharges

No NPDES-permitted wastewater facilities with fecal coliform limit were identified in the Horseshoe Creek basin.

6.3.2 NPDES Stormwater Discharges

The WLA for stormwater discharges with an MS4 permit is a 53 percent reduction in current fecal coliform. It should be noted that any MS4 permittee will only be responsible for reducing the anthropogenic loads associated with stormwater outfalls that it owns or otherwise has responsible control over, and it is not responsible for reducing other nonpoint source loads in its jurisdiction.

6.4 Margin of Safety (MOS)

28

Consistent with the recommendations of the Allocation Technical Advisory Committee (Florida Department of Environmental Protection, February 2001), an implicit margin of safety (MOS) was used in the development of this TMDL. For fecal coliform, an implicit MOS was inherently incorporated by using 400 MPN/100 mL of fecal coliform as the water quality target for individual samples, instead of setting the criteria as that no more than 10 percent of the samples exceeding 400 MPN/100 mL. Using the load duration curve method to develop TMDLs assumes there is no in-stream decay of fecal coliform bacteria after the watershed loading reaches the receiving waterbody, while in reality fecal coliform loadings could diminish through processes including death, grazing, and deposition. Therefore, the load duration curve method

tends to underestimate allowable fecal coliform loadings that a given waterbody receives and is therefore more conservative in establishing the TMDL. In addition, the existing loading capacity was established based only on samples that exceeded the allowable loadings. This approach has a tendency to make the existing loading estimation more conservative and therefore adds to the MOS.

29

Chapter 7: NEXT STEPS: IMPLEMENTATION PLAN DEVELOPMENT AND BEYOND

7.1 Basin Management Action Plan

Following the adoption of this TMDL by rule, the next step in the TMDL process is to develop an implementation plan for the TMDL, which will be a component of the Basin Management Action Plan (BMAP) for the Middle St. Johns River Basin. This document will be developed over the next year in cooperation with local stakeholders and will attempt to reach consensus on more detailed allocations and on how load reductions will be accomplished. The BMAP will include the following:

• Appropriate allocations among the affected parties,

• A description of the load reduction activities to be undertaken,

• Timetables for project implementation and completion,

• Funding mechanisms that may be utilized,

• Any applicable signed agreement,

• Local ordinances defining actions to be taken or prohibited,

• Local water quality standards, permits, or load limitation agreements, and

• Monitoring and follow-up measures.

30

References

Bottcher, A.B, N.B. Pickering, and A.B. Cooper. 1998. EAAMOD-FIELD: A Flow and Phosphorous Model for High Water Tables. Proceedings of the 7th Annual Drainage Symposium. American Society of Agricultural Engineers, St. Joseph, MI.

Florida Department of Environmental Protection. February 2001. A Report to the Governor and the Legislature on the Allocation of Total Maximum Daily Loads in Florida. Tallahassee, Florida: Bureau of Watershed Management.

Florida Administrative Code. Chapter 62-302, Surface Water Quality Standards.

Florida Administrative Code. Chapter 62-303, Identification of Impaired Surface Waters.

Florida Watershed Restoration Act. Chapter 99-223, Laws of Florida.

Jacobson, Barry M., Adelbert B. Bottcher, Nigel B. Pickering, and Jeffrey G. Hiscock. 1998. Unique routing algorithm for watershed assessment model. ASAE Paper No. 98-2237. Am. Soc. of Agr. Eng., St. Joseph, MI 49085.

JGH Engineering. 2005. Development of a graphical user interface for analyzing phosphorus load and import/export in the Lake Okeechobee protection Plant area. Final report. Contract number RS040634.

Knisel, W. G. 1993. GLEAMS: Groundwater Loading Effects of Agricultural Management Systems. UGA-CPES-BAED Publication no. 5.

Olivieri, V. P., C. W. Kruse, K. Kawata, J. E. Smith. 1977. Microorganisms in urban stormwater. USEPA Report No. EPA-600/2-77-087 (NTIS No. pB-272245). Environmental Protection Agency, Washington, D. C.

SWET (Soil & Water Engineering Technology, Inc.), 1999. EAAMOD Technical and User Manuals. Final Reports to the Everglades Research and Education Center, University of Florida, Belle Glade, FL.

Trial, W. et al. 1993. Bacterial source tracking: studies in an urban seattle watershed. Puget Sound Notes. 30: 1-3

United States Environmental Protection Agency. 2001. Protocol for developing pathogen TMDL, 1st edition. EPA 841-R-00-002.

31

Appendices

Appendix A: Background Information on Federal and State Stormwater Programs

In 1982, Florida became the first state in the country to implement statewide regulations to address the issue of nonpoint source pollution by requiring new development and redevelopment to treat stormwater before it is discharged. The Stormwater Rule, as authorized in Chapter 403, F.S., was established as a technology-based program that relies on the implementation of BMPs that are designed to achieve a specific level of treatment (i.e., performance standards) as set forth in Chapter 62-40, F.A.C. In 1994, the department’s stormwater treatment requirements were integrated with stormwater flood control requirements of the water management districts, along with wetland protection requirements, into the Environmental Resource Permit regulations.

Chapter 62-40 also requires the state’s water management districts (WMDs) to establish stormwater pollutant load reduction goals (PLRGs) and adopt them as part of a SWIM plan, other watershed plan, or rule. Stormwater PLRGs are a major component of the load allocation part of a TMDL. To date, stormwater PLRGs have been established for Tampa Bay, Lake Thonotosassa, the Winter Haven Chain of Lakes, the Everglades, Lake Okeechobee, and Lake Apopka. No PLRG has been developed for Newnans Lake at the time this study was conducted.

In 1987, the U.S. Congress established Section 402(p) as part of the federal Clean Water Act Reauthorization. This section of the law amended the scope of the federal NPDES permitting program to designate certain stormwater discharges as “point sources” of pollution. The EPA promulgated regulations and began implementation of the Phase I NPDES stormwater program in 1990. These stormwater discharges include certain discharges that are associated with industrial activities designated by specific Standard Industrial Classification (SIC) codes, construction sites disturbing five or more acres of land, and master drainage systems of local governments with a population above 100,000, which are better known as municipal separate storm sewer systems (MS4s). However, because the master drainage systems of most local governments in Florida are interconnected, the EPA implemented Phase I of the MS4 permitting program on a countywide basis, which brought in all cities (incorporated areas), Chapter 298 urban water control districts, and the Florida Department of Transportation throughout the fifteen counties meeting the population criteria. The Department received authorization to implement the NPDES stormwater program in 2000.

An important difference between the federal NPDES and the state’s stormwater/environmental resource permitting programs is that the NPDES program covers both new and existing discharges, while the state’s program focus on new discharges only. Additionally, Phase II of the NPDES Program, implemented in 2003, expands the need for these permits to construction sites between one and five acres, and to local governments with as few as 1,000 people. While these urban stormwater discharges are now technically referred to as “point sources” for the purpose of regulation, they are still diffuse sources of pollution that cannot be easily collected and

32

treated by a central treatment facility similar to other point sources of pollution, such as domestic and industrial wastewater discharges. It should be noted that all MS4 permits issued in Florida include a re-opener clause that allows permit revisions to implement TMDLs when the implementation plan is formally adopted.

33