APPENDIX III - Nonpoint Source Pollution Assessment

Historically, most water quality problems in Florida were associated with point sources, including both domestic and industrial sources. Nonpoint sources have now been determined to account for the majority of the state's water quality problems. This change is due primarily to point source treatment improvements and increases in agriculture and urban developed land (Hand and Paulic 1992). NPS pollution has also been identified as the major factor affecting downstream water quality in about 80 percent of the urban areas targeted in a national survey conducted by the Council on Environmental Quality (1972).

NPS pollution is a major, largely uncontrolled, cause of surface water degradation throughout Florida (Livingston et al. 1989). NPS pollutants in northwest Florida include pesticides, animal wastes, nutrients, and sediments (Wolfe et al. 1988). In north Florida, the progression of natural ecosystems to silvicultural, agricultural, and urban uses has resulted in NPS pollution impacts including increased peak and total discharge, increased concentrations of dissolved solids, nitrates, and ammonia, and increased export rates of pollutants during storms (Livingston et al. 1989).Major contributors of these pollutants include agriculture, stormwater runoff, silviculture, landfills, and septic tanks (US EPA 1989). Land use type and intensity are strongly related to NPS concentrations.

Contaminants associated with NPS can be detrimental to water quality. Nutrients can have direct toxic effects or may stimulate algal growth. Pesticides and other contaminants can be dangerous to the aquatic ecosystem.

Sediments affect water ways by not only reducing their storage capacity but also by increasing the temperature of the water and providing increased opportunities for the growth of water consuming plants (Clark et al. 1985). Additional sediment impacts include damage to the biological health and integrity of the aquatic ecosystem along with a decrease in recreational and aesthetic values.

Stormwater runoff is a significant source of NPS pollution, having solids concentrations equal to or greater than untreated sanitary wastewater, and biological oxygen demand (BOD) values approximately equal to those of secondary effluent. Suspended sediment loads from streams draining urban areas are often an order of magnitude greater than those from nearby forested watersheds (Burton et al. 1977b). In addition, bacterial contamination of stormwater may be two to four orders of magnitude greater than concentrations considered safe for water contact (Field and Turkeltaub 1981).

Urbanization has been shown to fundamentally alter the hydrology of watersheds (Anderson 1970). Increases in impervious surface areas result in substantially increased runoff (Simmons and Reynolds 1982, McElroy 1978). In addition, urbanization, with the associated land clearing and paving of pervious areas, has accelerated the problem of water quality deterioration throughout Florida. Stormwater and associated NPS pollution are responsible for:

1. 80-90 percent of the heavy metals deposited in Florida surface waters;

2. the majority of the sediment deposited in state waters; and 3.450 times the Total Suspended Solids flowing to receiving waters and nine times the BOD load when compared to loads from secondarily-treated wastewater effluent (Livingston et al. 1989).

Water quality changes due to urbanization also affect wildlife habitat. Jones and Clark (1987) indicated biological data were a better discriminator of urbanization stress on an ecological system than chemical parameters used in the same study.

Basin boundaries, land use and land cover data, and selected loading rates were input and processed using the District ARC/INFO GIS. Land use and land cover acreage within the watershed were used in combination with selected loading rates to estimate total loadings by land use.

Satellite imagery and aerial photography were used to quantify existing land use and land cover in the St. Marks and Wakulla Rivers watershed. Potential NPS pollutant loading rates were estimated for existing land use and land cover classes. Due to staffing constraints, Local Comprehensive Plan Future Land Use Maps (FLUM) and future development scenarios were not included. Loading rate relationships were determined for four water quality parameters. Selected loadings were applied to each land use and land cover class and total loads were estimated. Specific areas with exceptionally high loadings were identified.

Land Use Categories and NPS Loading Rates

Existing land use and land cover within the St. Marks and Wakulla Rivers watershed were initially classified into more than 40 categories which included similar land use and land cover types. Due to the impracticality of developing and applying 39 individual loading rates, original categories were aggregated into 15 categories, based on similarities in loading characteristics ( see Table A-3).

Previous NPS pollution studies were reviewed to corroborate loading rate estimates for water quality parameters and land use categories (Rains, et al. 1993). Loading rates for land use/cover categories were based on three studies, the Tampa Bay Watershed Study (Dames and Moore 1990), Florida Department of Environmental Regulation's Watershed Model Users Manual (Camp, Dresser and McKee 1991), and Reikerk (1983).

Loading rates for TN, TP, TSS, and BOD calculations were based on average annual rainfall (inches/acre/year), pollutant runoff coefficient (includes soil type, perviousness, etc.), pollutant runoff concentrations (milligrams/liter), by land use, and St. Marks and Wakulla Rivers mean annual rainfall (58 inches/year). Loading rates were reported in pounds per acre per year (lbs/acre/yr), and total loadings were reported in pounds/year (lbs/yr). Loading rates for TN, TP, BOD, and TSS were estimated for each land use category in developing composite loading rate relationships. Determining whether or not land uses met water quality standards was not within the scope of this study.

Loading rates from three studies (see Table A-4) were selected for use in the NPS assessment. All loading rate calculations were based on local rainfall data . Rainfall data for a five year period from the closest available rainfall stations, (Tallahassee and Wewahitchka) were used to identify an watershed average annual rainfall as 58 inches. (A summary of the 11 studies considered for their potential applicability to northwest Florida is provided in Appendix II).

TN and TP loading rates for four of the land use categories (institutional, transportation/utilities, tree plantations, and natural uplands), were derived from FDEP's Watershed Model Users Manual (Camp Dresser and McKee 1991). The manual provided event-mean concentrations based on percent impervious surface associated with land uses. TN and TP loading rates used to calculate loadings were determined by multiplying a weighted runoff coefficient by average annual rainfall (USGS gauging station) and event-mean concentrations from the Dames and Moore report (1990), FDEP documents, or estimated based on similar land use (see Table A-5).

Table A-3. FLUCCS codes for each existing land use and cover category in the St Marks and Wakulla Rivers watershed. Aggregated categories were based on existing

land use categories and compatibility with future land use categories.

Existing FLUCCS Categories	Existing Aggregated Categories
Residential	Residential
Low density residential	Low density residential
Medium density residential	Medium density residential
High density residential	High density residential
Commercial/services
	Commercial
Industrial	Industrial
Extractive	Extractive
Institutional	Institutional
Transportation	Transportation/utilities
Utilities
Recreation	Recreation/open lands
Open lands (urban)
Sand other than beaches	Spoil
Disturbed land/spoil areas
Cropland/pasture	Cropland/pasture
Open lands (agriculture)
Shrub/brushland
Upland coniferous forest	Upland forests
Upland hardwood forest
Upland mixed coniferous/hardwood forest
Tree plantation	Silviculture
Forest regeneration
Streams/waterways	Streams and lakes
Lakes
Wetland hardwood forest	Wetlands
Gum swamps
Titi swamps
Inland ponds/sloughs
Wetland mixed hardwood forest
Cypress
Wetland mixed coniferous forest
Wetland mixed coniferous/hardwood forest
Freshwater marshes
Saltwater marshes
Non-vegetated wetland
Not applicable	Conservation

Table A - 4. Loading rates (lbs/acre/yr) for land use categories in the St Marks and Wakulla Rivers Watershed .

BORDER=1 CELLSPACING=1> Land use TN TP BOD TSS Urban Low density residential 5.76 0.74 16.12 55.90 Med density residential 10.10 1.63 37.22 100.03 High density residential 19.49 4.36 98.31 677.05 Commercial 21.10 3.14 130.90 894.67 Industrial 17.90 3.10 95.99 935.87 Institutional 5.55 0.71 73.51 475.29 Recreation/open 2.76 0.12 3.20 24.49 Non-urban Cropland/pasture 8.89 1.32 14.57 211.97 Upland forest 2.67 0.42 8.89 118.23 Silviculture 2.67 0.42 8.89 118.23 Lakes and streams 7.88 0.69 10.69 19.54 Wetlands 4.54 0.54 13.33 28.94 Spoil/barren 4.06 0.40 23.45 225.95 Extractive 5.37 0.68 43.70 427.41 Transportation/utilities 8.00 1.01 67.10 459.60

Table A-5. References on which loading rate estimates were based for the St Marks and Wakulla Rivers watershed.

Land cover category	TN	TP	BOD	TSS
Urban
Low density residential	1	1	1	1
Commercial	1	1	1	1
Industrial	1	1	1	1
Institutional	1	1	1	1
Recreation/open	1	1	1	1
Non-urban
Cropland/pasture	1	1	1	1
Upland forest	3	3	4	4
Silviculture	3	3	4	4
Lakes and streams	1	1	1	1
Wetlands	1	1	1	1
Spoil/barren	4	4	4	4
Extractive	1	1	1	1
Transportation/utilities	3	3	4	4

1. Dames and Moore

2. Riekerk

3. Watershed Management Model

4. NWFWMD: derived from 1-3

Silviculture TN and TP loading rates were based on those reported by Riekerk (1988). In a study of impacts of tree regeneration on water quality, Riekerk (1988) reported total Kjeldahl nitrogen (TKN) as the predominant nitrogen form in tree regeneration runoff. Because of the small nitrate differences in tree regeneration runoff due to variations in silvicultural techniques, TKN was used as an estimate of TN for this report.

The BOD and TSS loading rate estimates for the institutional category were calculated as the average of low density residential and commercial category rates. In order to be consistent with FLUCCS land use system, the transportation/utilities category was included in the NPS loadings tables. This category included only major road systems since loading rates for a given land use include roads and other infrastructure associated with the land use. Transportation/utilities category loading rates for BOD and TSS were calculated as the average of recreational/open and commercial category rates.

In the ECFRPC Areawide 208 Study, all waterbodies are assumed to have a runoff coefficient of 1.0 (Dames and Moore 1990). Many waterbodies, however, discharge only under extreme rain events, and some are landlocked with no discharge. Mean runoff value of 0.50 was selected, although varying characteristics made estimation of a representative runoff value for waterbodies difficult.

NPS Total Loadings

Urban. Total acreage for urban land use categories accounted for only 7 percent (13,439 acres) of the area in the St. Marks and Wakulla Rivers watershed (Table A-7, Figure 6 in the main body), while estimated TN, TP, and BOD loadings (Table A-7) for urban categories accounted for 11 percent and 14 percent of the total loadings in the watershed. Of the TN, TP, and BOD loadings associated with urban land uses, 8 percent was associated with low density residential and approximately 1-3 percent were due to commercial and industrial. Urban TSS loadings accounted for only 7 percent of the total TSS loadings in the watershed, and three percent were due to low density residential.

Lowest estimated TN loadings were less than one percent and ranged from 153 pounds per year for institutional (See Table A-7). TP loadings were less than 100 lbs/yr for the institutional, recreation and open urban land categories. Estimated BOD and TSS were also lowest, and similar, for recreation and open urban lands.

Non-urban. Non-urban categories accounted for 85 percent to 93 percent of total estimated loadings for each of the four NPS pollutants (see Table A-7). Non-urban categories accounted for 93 percent of the total acreage in the St. Marks and Wakulla Rivers watershed. Silviculture areas in the St. Marks and Wakulla Rivers watershed

accounted for 35 percent of the watershed and accounted for 39 percent (Table A-7) of the estimated total suspended solids associated with NPS pollutants in the St. Marks and Wakulla Rivers watershed. Estimated TN, TP, and BOD associated with silviculture accounted for 24 percent to 29 percent of the total estimated NPS loadings for these parameters. Natural upland forests made up 25 percent of the watershed acreage. Upland forest estimated TSS loadings were second to those of silviculture.

The agriculture category made up 10 percent of the total acreage in St. Marks and Wakulla Rivers watershed and contributed 13 percent to 24 percent of the total estimated TN, TP, and BOD loadings (Table A-7). Agriculture loadings estimated for TN (178,470 lbs/yr), TP (26,499 lbs/yr), and BOD (292,498 lbs/yr) were second to those for silviculture (Table A 7).

Estimated NPS loadings for wetlands were similar (TN = 163,590 lbs/yr; TP = 19,458 lbs/yr; BOD = 480,319 lbs/yr; TSS = 1,042,794 lbs/yr) when compared with values for silviculture and wetlands. Remaining land use categories contributed less than five percent of the total loadings for any of the four individual parameters (Table A-7) and were consistent with low acreage.

The transportation/utilities category included less than 1 percent of the acreage in the St. Marks and Wakulla Rivers watershed (Table A-7) and corresponded primarily to acreage of major roadways. Estimated loadings due to major roadways made up less than one percent of the total estimated loadings (see Table A-7).

Total existing urban land use acreage accounted for 7 percent of the area within the St. Marks and Wakulla Rivers watershed and estimated NPS urban loadings ranged from nine percent to 11 percent of total TN, TP, and BOD loadings, and 4 percent of the total TSS loadings. Potential per-acre water quality impacts appeared greatest for the commercial and industrial urban land use categories. Overall relationships among categories were consistent with other studies (Wanielista 1975, Hand and Paulic 1992) in which urban land use categories had higher TN, TP, TSS, and BOD values in surface water runoff when compared with values for non-urban runoff.

Although loading rates for non-urban categories were comparatively low on a per- acre basis, silviculture practices have tremendous potential water quality impacts in the St. Marks and Wakulla Rivers watershed due to the amount of acreage and existing land practices.

In contrast with urban land uses, silviculture included 36 percent of the watershed acreage and 39 percent of the total TSS loadings. These findings are consistent with reports in the literature regarding suspended solids runoff from silviculture areas. Variation in nutrient and sediment loadings amounting to several orders of magnitude are common for forestry activities (Novotny et al. 1981), and TSS loadings from logging roads are significantly greater in the absence of appropriate BMPs (Lynch and Corbett 1990). Most sediment reaching waterways from these lands originate from construction of logging roads and from clearcuts which infringe upon natural drainage channels (Novotny 1981).

Water quality maintenance and control are normally achieved by implementing BMPs to reduce extreme pollution problems and promote rapid recovery. The BMPs are intended to control activities in stormflow source areas and promote acceptable road construction and soil conservation practices.

Results of this assessment reflect several areas of concern regarding NPS pollution in the St. Marks and Wakulla Rivers watershed. First, forestry and agriculture are responsible for the largest proportions of NPS pollution in the watershed. Secondly, while the literature indicates water quality impacts associated with these activities can be minimized through recommended BMPs, compliance with these practices is not easily enforced.

The impacts of silviculture and agriculture activities can be substantially reduced if recommended BMPs are comprehensively implemented and rigorously enforced. After evaluating the results of forest BMP surveys the Florida Department of Agriculture reported: "Of the 150 survey sites for 1991, 141 were found to be in compliance with silviculture BMPs which equates to 94 percent compliance",(Florida Department of Agriculture 1991) While many appropriate BMPs were implemented, most surveyed sites had several instances of non-compliance.

In an effort to clarify the use and interpretation of BMPs and make their application more consistent, Florida forestry BMPs have recently undergone substantial revision to increase water resource protection. If these changes are successful and compliance with revised BMPs is achieved, a reduction in NPS loadings from silviculture can be expected. The Division of Forestry is currently reassessing compliance and effectiveness of BMPs in an effort to more accurately reflect BMP compliance.

Three general options exist for abating NPS pollution from urban activity. The options involve prevention, treatment, and control measures, implemented as an integrated abatement approach (Wanielista 1975). In brief, prevention involves practices that are applied before problems arise; treatment involves complete or partial physical, chemical, and/or biological processes for minimizing impacts of stormwater; and control measures would involve reduction or control of pollution sources.

Ideally, limiting discharges from new developments to discharge that would have occurred under natural, undeveloped conditions, in addition to maintaining water quality standards, should result in no increases in NPS pollution.