USF Libraries
USF Digital Collections

Spatial and temporal trends in water quality in the Alafia River watershed

MISSING IMAGE

Material Information

Title:
Spatial and temporal trends in water quality in the Alafia River watershed
Physical Description:
Book
Language:
English
Creator:
Aragon, Jennifer M
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Seasonal kendall
Nonparametric
Land use
Tampa
Florida
Geographic information systems
Dissertations, Academic -- Geography -- Masters -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Water quality data and land use information were analyzed within the Alafia River watershed in Florida to determine spatial and temporal trends in these variables over a 16 year time period from 1991-2006. Monthly water quality data (for dissolved oxygen, turbidity, fecal coliform, total phosphorus, and total nitrogen) were statistically analyzed using the modified seasonal Kendall nonparametric test for trends that accounts for serial correlation. The statistical trend analysis was conducted for the entire study period, but monthly, seasonal, and land use trends were also examined. Land use information was examined using Geographic Information Systems to determine the percent change in land use proportion from 1990 to 1999, 1999 to 2006, and 1990 to 2006. The proportions of each land use and their percent change were then related to the trends in water quality.The results of this analysis showed that water quality for the parameters turbidity and total phosphorus have been shown to be improving with statistically significant decreasing trends for turbidity at stations 74, 111, 116, and 139 and for total phosphorus at stations 74, 114, and 115. A statistically significant decreasing trend in dissolved oxygen was determined for stations 116 and an increasing trend in total nitrogen for stations 114, 115, and 151 implying water quality for these parameters is degrading. Other noted trends were high fecal coliform and total nitrogen at station 111, which has higher proportions of agricultural land use and an increasing proportion of urban and built-up land use. Also, low dissolved oxygen was noted at station 74. The proportions of land use for the entire study area have changed from predominantly wetlands to now urban and built-up land use.While agricultural, rangeland, and wetlands land use have shown a reduction in the proportion of coverage in the contributing zone of almost every station, urban and built-up land use has increased in proportion at every station.
Thesis:
Thesis (M.S.)--University of South Florida, 2009.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Jennifer M. Aragon.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 143 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002069273
oclc - 608084231
usfldc doi - E14-SFE0003163
usfldc handle - e14.3163
System ID:
SFS0027479:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Spatial and Temporal Trends in Water Quality in the Alafia River Watershed by Jennifer M. Aragon A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science Department of Geography College of Arts and Sciences University of South Florida Major Professor: Philip Reeder, Ph.D. Steven Reader, Ph.D. Kamal Alsharif, Ph.D. Date of Approval: November 16, 2009 Keywords: seasonal kendall, nonparametric, land use Tampa, Florida, geographic information systems Copyright 2009 Jennifer M. Aragon

PAGE 2

Dedication I would like to dedicate this thesis to my loving f amily. To my parents, Richard Kabat and Joan Kabat who have always stressed the i mportance of education and who continually encouraged me throughout my college car eer. To my brother, Steven, who set a good example for me while growing up and my l oving husband, Nelson, who supported me and provided encouragement throughout the thesis writing process. His patience and understanding every time I retreated t o the computer room for hours on end made everything easier on me.

PAGE 3

Acknowledgements I would like to begin by thanking my major professo r, Dr. Philip Reeder, who was always easily accessible when I needed information, support, and assistance throughout my thesis. Without his encouragement and confidenc e in me, I would not have been able to complete my master’s thesis. I also appreciate Dr. Steven Reader and Dr. Kamal Alsharif for all of their comments and suggestions that have helped guide me through my research. A special thanks goes to Dr. Robert Hirsch, Researc h Hydrologist, U.S. Geological Survey, and Dr. Steven Millard, Probabil ity, Statistics, and Information, for responding to all of my statistical inquiries. Dr. Hirsch helped me grasp the seasonal Kendall test, a modification of the Mann-Kendall te st that he played a main role in developing. Dr. Millard provided assistance with l earning and applying the statistical software, EnvironmentalStats for S-Plus, which he c reated. I am very grateful for their help and without their valuable support; I would no t have been able to complete my master’s thesis. I would also like to thank the Environmental Protec tion Commission of Hillsborough County, Florida for supplying the hist orical water quality data used in this study and for providing knowledge of laboratory and water sampling related information.

PAGE 4

i Table of Contents List of Tables..................................... ................................................... .............................iv List of Figures.................................... ................................................... ............................vii Abstract........................................... ................................................... ................................xi Introduction....................................... ................................................... ...............................1 Literature Review.................................. ................................................... ..........................6 Water Quality Studies.............................. ................................................... ....................6 Water Quality Studies in the Tampa Bay Region...... ................................................... ..9 Alafia River Watershed Technical Reports........... ................................................... .....10 Statistical Analysis of Water Quality Trends....... ................................................... ......13 Water Quality Management........................... ................................................... ............16 Watershed Management Policy........................ ................................................... ..........18 Research Design.................................... ................................................... ........................21 The Problem........................................ ................................................... .......................21 Research Questions................................. ................................................... ...................22 Project Significance and Rationale................. ................................................... ...........22 Study Area......................................... ................................................... ............................24 Tampa Bay Watershed................................ ................................................... ...............24 Alafia River Watershed............................. ................................................... .................25 Alafia River Watershed Sub-Drainage Basins......... ................................................... ..26

PAGE 5

ii Methodology........................................ ................................................... ..........................28 Contents of Dataset................................ ................................................... ....................28 Sample Analysis Methods............................ ................................................... ..............30 Sample Design...................................... ................................................... ......................31 Water Quality Data Analysis........................ ................................................... ..............32 Data Storage and Manipulation ................................................... .............................32 Statistical Analysis ................................................... ..................................................3 4 Land Use Data Analysis............................. ................................................... ................36 Data Storage and Manipulation ................................................... .............................36 Geographic Information System Data Analysis ................................................... .....37 Results and Discussion............................. ................................................... .....................39 Dissolved Oxygen................................... ................................................... ...................41 Turbidity.......................................... ................................................... ...........................57 Fecal Coliform..................................... ................................................... .......................71 Total Phosphorus................................... ................................................... .....................86 Total Nitrogen..................................... ................................................... .....................102 Land Use Trends.................................... ................................................... ..................116 Summary and Conclusions............................ ................................................... ..............124 Water Quality Seasonal Trends...................... ................................................... ..........124 Overall Trends by Sample Station................... ................................................... ........126 Research Goals..................................... ................................................... ....................130 Implications....................................... ................................................... .......................132 Additional Research................................ ................................................... .................135

PAGE 6

iii List of References................................. ................................................... .......................137 Appendices......................................... ................................................... ..........................141 Appendix A: Land Use Tables....................... ................................................... .............142

PAGE 7

iv List of Tables Table 1. Sample Methods of Analysis............... ................................................... ............31 Table 2-A. Descriptive statistics of dissolved oxy gen (mg/L) for station 74...................43 Table 2-B. Descriptive statistics of dissolved oxy gen (mg/L) for station 111.................44 Table 2-C. Descriptive statistics of dissolved oxy gen (mg/L) for station 114.................45 Table 2-D. Descriptive statistics of dissolved oxy gen (mg/L) for station 115.................46 Table 2-E. Descriptive statistics of dissolved oxy gen (mg/L) for station 116..................47 Table 2-F. Descriptive statistics of dissolved oxy gen (mg/L) for station 139..................48 Table 2-G. Descriptive statistics of dissolved oxy gen (mg/L) for station 151.................49 Table 2-H. Descriptive statistics of dissolved oxy gen (mg/L) for station 154.................50 Table 3. Modified Seasonal Kendall trend test resu lts (significance level = 0.05) for dissolved oxygen by sample station....... .............................................57 Table 4-A. Descriptive statistics of turbidity (NT U) for station 74.................................. 58 Table 4-B. Descriptive statistics of turbidity (NT U) for station 111................................5 9 Table 4-C. Descriptive statistics of turbidity (NT U) for station 114................................6 0 Table 4-D. Descriptive statistics of turbidity (NT U) for station 115................................6 1 Table 4-E. Descriptive statistics of turbidity (NT U) for station 116................................6 2 Table 4-F. Descriptive statistics of turbidity (NT U) for station 139................................6 3 Table 4-G. Descriptive statistics of turbidity (NT U) for station 151................................6 4 Table 4-H. Descriptive statistics of turbidity (NT U) for station 154................................6 5

PAGE 8

v Table 5. Modified Seasonal Kendall trend test resu lts (significance level = 0.05) for turbidity by sample station.............. ................................................... .71 Table 6-A. Descriptive statistics of fecal colifor m (cfu/100mL) for station 74...............73 Table 6-B. Descriptive statistics of fecal colifor m (cfu/100mL) for station 111..............74 Table 6-C. Descriptive statistics of fecal colifor m (cfu/100mL) for station 114..............75 Table 6-D. Descriptive statistics of fecal colifor m (cfu/100mL) for station 115.............76 Table 6-E. Descriptive statistics of fecal colifor m (cfu/100mL) for station 116..............77 Table 6-F. Descriptive statistics of fecal colifor m (cfu/100mL) for station 139..............78 Table 6-G. Descriptive statistics of fecal colifor m (cfu/100mL) for station 151.............79 Table 6-H. Descriptive statistics of fecal colifor m (cfu/100mL) for station 154.............80 Table 7. Modified Seasonal Kendall trend test resu lts (significance level = 0.05) for fecal coliform by sample station......... ................................................86 Table 8-A. Descriptive statistics of total phospho rus (mg/L) for station 74....................88 Table 8-B. Descriptive statistics of total phospho rus (mg/L) for station 111...................89 Table 8-C. Descriptive statistics of total phospho rus (mg/L) for station 114...................90 Table 8-D. Descriptive statistics of total phospho rus (mg/L) for station 115..................91 Table 8-E. Descriptive statistics of total phospho rus (mg/L) for station 116...................92 Table 8-F. Descriptive statistics of total phospho rus (mg/L) for station 139...................93 Table 8-G. Descriptive statistics of total phospho rus (mg/L) for station 151..................94 Table 8-H. Descriptive statistics of total phospho rus (mg/L) for station 154..................95 Table 9. Modified Seasonal Kendall trend test resu lts (significance level = 0.05) for total phosphorus by sample station....... ............................................102 Table 10-A. Descriptive statistics of total nitrog en (mg/L) for station 74......................103 Table 10-B. Descriptive statistics of total nitrog en (mg/L) for station 111....................104

PAGE 9

vi Table 10-C. Descriptive statistics of total nitrog en (mg/L) for station 114....................105 Table 10-D. Descriptive statistics of total nitrog en (mg/L) for station 115....................106 Table 10-E. Descriptive statistics of total nitrog en (mg/L) for station 116....................107 Table 10-F. Descriptive statistics of total nitrog en (mg/L) for station 139....................108 Table 10-G. Descriptive statistics of total nitrog en (mg/L) for station 151....................109 Table 10-H. Descriptive statistics of total nitrog en (mg/L) for station 154....................110 Table 11. Modified Seasonal Kendall trend test res ults (significance level = 0.05) for total nitrogen by sample station......... .............................................116 Table 12. Water quality seasonal trend summaries.. ................................................... ...125 Table A-1. The proportion of each land use area to the total area within the contributing zone of each sample station........... ..........................................142 Table A-2. The percent change in the proportion of each land use area from 1990 – 1999, 1999 – 2006, and 1990 – 2006 for each sampl e station..................143

PAGE 10

vii List of Figures Figure 1. Alafia River Watershed’s Impaired WBIDs. ................................................... ....2 Figure 2. The Tampa Bay Watershed................. ................................................... ...........25 Figure 3. Alafia River Major Sub-basins........... ................................................... ............27 Figure 4. Alafia River Monthly Sampling Stations.. ................................................... .....30 Figure 5. The eight sample stations in the Alafia River watershed used in this study.............................................. ................................................... ..........41 Figure 6. Each sample station’s monthly dissolved oxygen median................................51 Figure 7-A. Dry versus wet annual seasonal median trends of dissolved oxygen (mg/L) for station 74.............................. ................................................... ...52 Figure 7-B. Dry versus wet annual seasonal median trends of dissolved oxygen (mg/L) for station 111............................. ................................................... ..53 Figure 7-C. Dry versus wet annual seasonal median trends of dissolved oxygen (mg/L) for station 114............................. ................................................... ..53 Figure 7-D. Dry versus wet annual seasonal median trends of dissolved oxygen (mg/L) for station 115............................. ................................................... ..54 Figure 7-E. Dry versus wet annual seasonal median trends of dissolved oxygen (mg/L) for station 116............................. ................................................... ..54 Figure 7-F. Dry versus wet annual seasonal median trends of dissolved oxygen (mg/L) for station 139............................. ................................................... ..55 Figure 7-G. Dry versus wet annual seasonal median trends of dissolved oxygen (mg/L) for station 151............................. ................................................... ..55 Figure 7-H. Dry versus wet annual seasonal median trends of dissolved oxygen (mg/L) for station 154............................. ................................................... ..56

PAGE 11

viii Figure 8. Each sample station’s monthly turbidity median............................................. .66 Figure 9-A. Dry versus wet annual seasonal median trends of turbidity (NTU) for station 74......................................... ................................................... ..........67 Figure 9-B. Dry versus wet annual seasonal median trends of turbidity (NTU) for station 111........................................ ................................................... .........67 Figure 9-C. Dry versus wet annual seasonal median trends of turbidity (NTU) for station 114........................................ ................................................... .........68 Figure 9-D. Dry versus wet annual seasonal median trends of turbidity (NTU) for station 115........................................ ................................................... .........68 Figure 9-E. Dry versus wet annual seasonal median trends of turbidity (NTU) for station 116........................................ ................................................... .........69 Figure 9-F. Dry versus wet annual seasonal median trends of turbidity (NTU) for station 139........................................ ................................................... .........69 Figure 9-G. Dry versus wet annual seasonal median trends of turbidity (NTU) for station 151........................................ ................................................... .........70 Figure 9-H. Dry versus wet annual seasonal median trends of turbidity (NTU) for station 154........................................ ................................................... .........70 Figure 10. Each sample station’s monthly fecal col iform median....................................81 Figure 11-A. Dry versus wet annual seasonal median trends of fecal coliform (cfu/100mL) for station 74......................... ................................................82 Figure 11-B. Dry versus wet annual seasonal median trends of fecal coliform (cfu/100mL) for station 111........................ ...............................................82 Figure 11-C. Dry versus wet annual seasonal median trends of fecal coliform (cfu/100mL) for station 114........................ ...............................................83 Figure 11-D. Dry versus wet annual seasonal median trends of fecal coliform (cfu/100mL) for station 115........................ ...............................................83 Figure 11-E. Dry versus wet annual seasonal median trends of fecal coliform (cfu/100mL) for station 116........................ ...............................................84 Figure 11-F. Dry versus wet annual seasonal median trends of fecal coliform (cfu/100mL) for station 139........................ ...............................................84

PAGE 12

ix Figure 11-G. Dry versus wet annual seasonal median trends of fecal coliform (cfu/100mL) for station 151........................ ...............................................85 Figure 11-H. Dry versus wet annual seasonal median trends of fecal coliform (cfu/100mL) for station 154........................ ...............................................85 Figure 12. Each sample station’s monthly total pho sphorus median................................96 Figure 13-A. Dry versus wet annual seasonal median trends of total phosphorus (mg/L) for station 74.............................. ................................................... .97 Figure 13-B. Dry versus wet annual seasonal median trends of total phosphorus (mg/L) for station 111............................. ................................................... 98 Figure 13-C. Dry versus wet annual seasonal median trends of total phosphorus (mg/L) for station 114............................. ................................................... 98 Figure 13-D. Dry versus wet annual seasonal median trends of total phosphorus (mg/L) for station 115............................. ................................................... 99 Figure 13-E. Dry versus wet annual seasonal median trends of total phosphorus (mg/L) for station 116............................. ................................................... 99 Figure 13-F. Dry versus wet annual seasonal median trends of total phosphorus (mg/L) for station 139............................. .................................................10 0 Figure 13-G. Dry versus wet annual seasonal median trends of total phosphorus (mg/L) for station 151............................. .................................................10 0 Figure 13-H. Dry versus wet annual seasonal median trends of total phosphorus (mg/L) for station 154............................. .................................................10 1 Figure 14. Each sample station’s monthly total nit rogen median...................................111 Figure 15-A. Dry versus wet annual seasonal median trends of total nitrogen (mg/L) for station 74.............................. ..................................................1 12 Figure 15-B. Dry versus wet annual seasonal median trends of total nitrogen (mg/L) for station 111............................. .................................................11 2 Figure 15-C. Dry versus wet annual seasonal median trends of total nitrogen (mg/L) for station 114............................. .................................................11 3 Figure 15-D. Dry versus wet annual seasonal median trends of total nitrogen (mg/L) for station 115............................. .................................................11 3

PAGE 13

x Figure 15-E. Dry versus wet annual seasonal median trends of total nitrogen (mg/L) for station 116............................. .................................................11 4 Figure 15-F. Dry versus wet annual seasonal median trends of total nitrogen (mg/L) for station 139............................. .................................................11 4 Figure 15-G. Dry versus wet annual seasonal median trends of total nitrogen (mg/L) for station 151............................. .................................................11 5 Figure 15-H. Dry versus wet annual seasonal median trends of total nitrogen (mg/L) for station 154............................. .................................................11 5 Figure 16-A. Changes in the proportions of land us e at station 74.................................120 Figure 16-B. Changes in the proportions of land us e at station 111...............................120 Figure 16-C. Changes in the proportions of land us e at station 114...............................121 Figure 16-D. Changes in the proportions of land us e at station 115...............................121 Figure 16-E. Changes in the proportions of land us e at station 116...............................122 Figure 16-F. Changes in the proportions of land us e at station 139...............................122 Figure 16-G. Changes in the proportions of land us e at station 151...............................123 Figure 16-H. Changes in the proportions of land us e at station 154...............................123

PAGE 14

xi Spatial and Temporal Trends in Water Quality in the Alafia River Watershed Jennifer M. Aragon Abstract Water quality data and land use information were a nalyzed within the Alafia River watershed in Florida to determine spatial and temporal trends in these variables over a 16 year time period from 1991-2006. Monthly water quality data (for dissolved oxygen, turbidity, fecal coliform, total phosphorus and total nitrogen) were statistically analyzed using the modified seasonal Kendall nonpar ametric test for trends that accounts for serial correlation. The statistical trend anal ysis was conducted for the entire study period, but monthly, seasonal, and land use trends were also examined. Land use information was examined using Geographic Informati on Systems to determine the percent change in land use proportion from 1990 to 1999, 1999 to 2006, and 1990 to 2006. The proportions of each land use and their p ercent change were then related to the trends in water quality. The results of this analysis showed that water qual ity for the parameters turbidity and total phosphorus have been shown to be improvin g with statistically significant decreasing trends for turbidity at stations 74, 111 116, and 139 and for total phosphorus at stations 74, 114, and 115. A statistically sign ificant decreasing trend in dissolved

PAGE 15

xii oxygen was determined for stations 116 and an incre asing trend in total nitrogen for stations 114, 115, and 151 implying water quality f or these parameters is degrading. Other noted trends were high fecal coliform and tot al nitrogen at station 111, which has higher proportions of agricultural land use and an increasing proportion of urban and built-up land use. Also, low dissolved oxygen was noted at station 74. The proportions of land use for the entire study area have changed from predominantly wetlands to now urban and built-up land use. While agricultural, r angeland, and wetlands land use have shown a reduction in the proportion of coverage in the contributing zone of almost every station, urban and built-up land use has increased in proportion at every station.

PAGE 16

1 Introduction This study will explore temporal and spatial change s in water quality and land use in the Alafia River watershed, located within Hills borough and Polk counties, Florida. Ten Waterbody Identification (WBID) codes are liste d as impaired (those that do not meet water quality standards) in the Alafia River w atershed, including WBIDs 1621G, 1660, 1635, 1592C, 1621E, 1675, 1583, 1653, 1639, a nd 1578B (Figure 1) (U.S. EPA, 2007). As noted from Figure 1, most of the impaire d waters lie along the main tributary to the south and towards the headwaters of the main tributary in the northern portion of the watershed. The main cause of impairment of the se waters is from nutrients (U.S. EPA, 2007). Nutrients including nitrogen and phosp horous are essential for aquatic plant growth, but at high levels they can cause harmful a lgal blooms, poor water clarity, and harmful effects on wildlife. It is important to st udy temporal trends in water quality, the spatial distribution of these trends in the water b ody, as well as how land use changes correlate to these trends in order to better focus time and resources on improving water quality in the watershed. Part of the Florida Department of Environmental Pro tection’s (FDEP) duties are to identify impaired waters and determine the total amount of the impairing pollutant that the water bodies can handle and still meet water qu ality standards [a Total Maximum Daily Load (TMDL)]. The establishment of a TMDL fo r WBID 1639 (Thirty Mile

PAGE 17

2 Figure 1. Alafia River Watershed’s Impaired WBIDs. Each delineated WBID boundary and identification number are shown within the Alafia River Watershed with all of the impaired WBIDs, as listed by the U. S. EPA, shown in red. Creek) in the Alafia River watershed was determined to be of high priority due to its degree of impairment while the remaining impaired W BIDs were listed as low priority. On December 20, 2005, the target to meet the nutrie nts and dissolved oxygen water quality standards for Thirty Mile Creek was determi ned to be a TMDL of 3.0 mg/L for total nitrogen. The suggested TMDLs for the remain ing impaired WBIDs are slated to be submitted to the U.S. Environmental Protection Agen cy (U.S. EPA) by December 31, 2008 (U.S. EPA, 2007). In Florida, water quality degradation is a major en vironmental concern since many of its waters are used most of the year for re creational activities and some are used

PAGE 18

3 as sources of drinking water. Water quality is def ined as the physical, chemical, and biological condition of a given water body (Wang 20 01). There are two main sources of water quality pollution including point source poll ution and non-point source pollution. Point source pollution can be linked to a specific origin such as the discharge from a domestic wastewater treatment plant. These sources are easier to locate and control through environmental regulations. Non-point sourc e pollution; however, cannot be linked to one specific source or pipe. Two main so urces of non-point source pollution are from atmospheric deposition and stormwater runoff. Humans can have a great impact on the amount of harmful pollutants introduced into su rface waters in stormwater runoff depending on the types of land use surrounding the water body. Non-point source pollution (primarily stormwater ru noff) is the main cause of water quality degradation in the United States (U.S EPA, 2008b). This type of pollution is very difficult to regulate since the origin of t he pollution cannot be linked to one specific source such as a discharge pipe from an in dustry. The effects of stormwater runoff on water quality are dependent upon the type s of land use in the area. For example, residential and agricultural lands may hav e more fertilizers and pesticides entering receiving waters while pasture lands may h ave more fecal coliform bacteria runoff. To minimize the amount of pollution enteri ng surface waters from stormwater runoff, it is important to study the types of land use in an area to determine the possible contributing sources of stormwater pollution. To address non-point source pollution, states recei ve grant money from the U.S. EPA under Section 319 of the Clean Water Act to use in their non-point source management programs (FDEP, 2008). In Florida, the FDEP’s program reviews proposals

PAGE 19

4 for non-point source pollution reduction projects a nd provides funding of up to 60 percent to implement the plan. This program is bas ed on a watershed approach and projects may only be implemented in watersheds that have been highly impacted by nonpoint source pollution including those listed in th e Surface Water Improvement and Management (SWIM) program (FDEP, 2008). According to the U.S. EPA (2007), there are 827 wat er bodies listed impaired in the state of Florida as of June 11, 2003. Major so urces of water quality pollution in Florida are due to human activities such as increas ing amounts of impervious surfaces, domestic and industrial discharges, and agricultura l runoff with most of the sources of pollution stemming from non-point sources (DeBusk, 2002). The top two causes of water quality impairment in Florida are due to oxyg en depletion and excessive nutrients (U.S. EPA, 2007). According to the U.S. EPA (2006), each state in the U.S. is required by Section 303(d) of the Clean Water Act to identify impaired surface water bodies and establish TMDLs for those pollutants impairing the water. A TMDL is the total amount of a pollutant that a water body can handle and still me et its water body designation. TMDLs are established on a watershed based approach. A w atershed is an area of land in which all the water drains to a common body of water. Th e FDEP designates a WBID code for each watershed sub-basin in the state of Florida. Water quality is assessed in each WBID to determine whether the established water quality standards are being achieved. If the WBID is determined to be impaired, a priority ranki ng is given to the water body and TMDLs are created. This value should also take int o account seasonal water quality variations and a margin of safety so the water body maintains its designated usage such

PAGE 20

5 as a drinking water source, recreational usage, or fishery. Establishing TMDLs for impaired WBIDs helps states and local governments s et goals for restoration and protection of the water bodies (U.S. EPA, 2006). The Alafia River is currently listed as a Class III surface water with the designated uses identified for recreation and the p ropagation of healthy fish and wildlife. Most water bodies in the state are designated as Cl ass III waters (FDEP, 2009). However, officials from Tampa Bay Water are recomme nding to change the classification of the Alafia River to a Class I wat er body since they draw water from the river for treatment and distribution for potable wa ter usage (Tampa Bay Water, 2008). As a Class I water, the Alafia River would be desig nated as a potable water supply, in which stricter water quality standards would be req uired. Tampa Bay Water has forwarded its request for a reclassification to the FDEP for review and if approved, the request will be forwarded to the Florida Environmen tal Regulation Commission for final determination (Tampa Bay Water, 2008). With the rising population in Hillsborough County [ 1990 population 834,054 and 2006 population 1,157,738, according to the U.S. Ce nsus Bureau (2008)], it is assumed that the land use will change throughout the county over time as well. As a result of land use change, different pollutants or concentration o f pollutants may be entering the waters of the county during storm events than had been sev eral years ago. It is important to examine water quality trends to determine if these changes have affected local waterways and to better focus environmental restoration effor ts to impacted waters.

PAGE 21

6 Literature Review Water Quality Studies Several studies have analyzed the effects of non-p oint source pollution on water quality on a watershed scale. Many of these studie s obtained their land use information from remote sensing sources and the data was statis tically analyzed to determine the effect of land use on water quality, including stud ies by Maillard and Pinheiro Santos (2008), Kebede et al. (2003), and Tong and Chen (20 02). Maillard and Pinheiro Santos (2008) studied water q uality, specifically turbidity, fecal coliform, nitrate, nitrite, and phosphorus fr om 16 sample stations in the Velhas River watershed in Brazil. Water quality data was separated based on the wet season (January) and dry season (July) and least squares r egression analysis was used to determine the effects of non-point source pollution on water quality in the river. The analysis generated linear models which were calcula ted based on the percent, not the total measurable area, of each type of land use classific ation within five different buffer widths (riparian zones) of the river. An analysis of vari ance was conducted on the coefficients to test the null hypothesis. The results of this stud y showed that barren land contributed the most to increased turbidity and fecal coliform, urb an land use was the greatest contributor of nitrite but it was not heavily weighted since la rge urban centers were only located near two of the sample points, and almost all of the lan d use classes were found to contribute

PAGE 22

7 to increases in nitrate and phosphorus. A good inh ibitor of all these pollutants was determined to be riparian forest land use. Each wa ter quality variable also varied by season with turbidity and fecal coliform higher dur ing the wet season, nitrate and nitrite higher during the dry season, and phosphorus had li ttle variability between the seasons. Finally, comparison of the five riparian zones show ed that there should not be one set buffer for all land use types; it should vary depen ding on the land use in the area (Maillard and Pinheiro Santos 2008). Kebede et al. (2003) studied the link between non-p oint source pollution and land use distribution from five sub-basins in the J.B.Co nverse watershed in Alabama. They examined water quality data from each of these subbasins for the parameters of dissolved phosphorous, dissolved nitrogen, and tota l nitrogen. To determine the types of land use that contributed to non-point source pollu tion, the ordinary least square regression model was also used in this study. In t his model, the concentration of each pollutant was the dependent variable and stream flo w and the percent land cover for each type of land use were the independent variables. B ased on the results of the Prediction Criterion (PC) they determined that the log linear function was more appropriate for this study. The results of this study showed that the n itrogen levels (both dissolved and total nitrogen) decreased with increases in water dischar ge. They also determined that with an increase in forest area, there was a decrease in th e amount of nitrogen runoff. However, with an increase in pasture and urban land use, the nitrogen concentration increased. The model results for dissolved phosphorous were not st atistically significant in this watershed to show a strong connection between land use and dissolved phosphorous concentration, so the results were not included in this study (Kebede et al. 2003).

PAGE 23

8 Tong and Chen (2002) analyzed the relationship betw een land use, stream flow, and water quality on a regional scale for the state of Ohio and on a watershed scale for the East Fork Little Miami River basin. For the an alysis at the state level, Spearman’s Rank Correlation (non-parametric statistical analys is) was used to determine the relationship between percent land use cover and mea n water quality from each Hydrological Unit (HUC) in the state for the parame ters total nitrogen, total phosphorus, and fecal coliform and the results were spatially a nalyzed using ArcView GIS. The results of the analysis for the state of Ohio showe d that total nitrogen, total phosphorus, and fecal coliform were positively correlated with agriculture, residential, and commercial land uses and they were all negatively c orrelated with forest land use. When total nitrogen and total phosphorus were compared t o each land use type in an analysis of variance, they determined that each of parameter’s mean ranked higher in areas of agricultural land use than in urban ones. These re sults brought up the point that most efforts in restoring water bodies are focused on re ducing nitrogen loading; however, more attention should be focused on reducing both nitrog en and phosphorous runoff since each are shown to be strongly correlated to agricultural and urban land use (Tong and Chen 2002). For the East Fork Little Miami River basin, Tong an d Chen (2002) used the Better Assessment Science Integrating Point and Non point Sources (BASINS) hydrologic model designed by the U.S. EPA to determine the amo unt of runoff and quality of runoff based on land use in the basin. The watershed was delineated into eleven sub-basins and the Nonpoint Source Model (NPSM) was used from BASI NS to build a hydrologic model for the river basin. The NPSM was also used to build models for each water

PAGE 24

9 quality parameter studied; including nitrogen, phos phorous, and fecal coliform. The results of this analysis showed that impervious lan d surface had greater than 55 percent more runoff than pervious surfaces and that out of all the land use types, agricultural lands in the sub-basins produced the greatest conce ntration of all the water quality parameters (Tong and Chen 2002). Water Quality Studies in the Tampa Bay Region In studying the effects of urban land use change a nd population distribution on water quality in the Tampa Bay watershed, Xian et a l. (2007) also used remote sensing and regression analysis. A multivariate regression liner model was used to estimate the amount of large impervious surface area in the wate rshed. To keep the heterogeneity in the urban land use and land cover in the watershed, estimating the percent impervious surface area was also conducted on a sub-pixel scal e using high resolution imagery. Population density was then calculated from U.S. Ce nsus data and defined as the amount of people per square kilometer in a particular area of the watershed. Areas of high population density were often associated with areas of high impervious surface area (Xian et al., 2007). Water quality was examined using data provided in GIS format from the Engineering Division of Hillsborough County. Estim ates of pollutant loads to the Hillsborough River, Alafia River, and Little Manate e River watersheds were assessed based on the Hillsborough County Pollutant Loading and Removal Model (PLRM). The Alafia River watershed had the highest annual avera ge non-point source loading for most of the water quality parameters out of each of thes e watersheds (Xian et al., 2007).

PAGE 25

10 The spatial distribution of the pollutant loadings was also analyzed. A high concentration of five day biological oxygen demand (BOD5) and total dissolved phosphorous (TDP) was found in the southern portion of Hillsborough County where there is mainly agricultural land. High total susp ended solids (TSS) concentration was found in the northwest portion of the Tampa Bay wat ershed where population density was high. High total Kjeldahl nitrogen (TKN) and t otal nitrogen (TN) values were observed in the northwest portion of the Tampa Bay watershed where agricultural land use is minimal at between two to nine percent of th e total area land use. Each of these pollutants showed clear regional distributions with in the watershed (Xian et al., 2007). Polynomial regression analysis was then used to co rrelate pollutant loadings with population density and impervious surface area (“ur banization”). The results of this analysis indicated that as urbanization increased, so did the pollutant loadings for each of the parameters sampled. Through further regression analysis, percent impervious surface area and population density were each tested for co rrelations with pollutant loading rates. It was determined that population density was a bet ter estimator for pollutant loading rates. These regression analyses are good for thes e more broad scale water quality studies involving a large watershed such as Tampa B ay (Xian et al., 2007). Alafia River Watershed Technical Reports The Southwest Florida Water Management District (D istrict) (2001) is focusing on developing a Comprehensive Watershed Management (CWM) plan for 11 different watersheds within their 16 county districts located in west-central Florida. In 2001, a CWM plan was developed for the Alafia River watersh ed to identify areas of the

PAGE 26

11 watershed where additional resources and time shoul d be focused in order to effectively manage the watershed. The main areas of interest i n regards to resource management that the District focused on were water supply, flo od protection, natural systems, and water quality (SWFWMD, 2001). The results of the w ater quality analysis will be the focal point of this discussion. The District determined water quality trends in th e Alafia River based on the analysis of the Florida Water Quality Assessment 19 94 and 1996 305(b) report results and water quality data collected by the Hillsboroug h County Environmental Protection Commission (EPC). They examined EPC water quality data collected from stations 74, 114, 111, 151, 115, 116, and 139 along the river du ring the time period between 1974 and 1995 (Figure 4). They sectioned their analysis int o three groups; North Prong, South Prong, and Lower Alafia River (SWFWMD, 2001). The North Prong is the portion of the Alafia River which originates in the northeastern portion of the watershed and eventuall y converges with the South Prong at the Alderman Ford State Park at the center of the w atershed. The main water quality impact in this portion of the watershed was determi ned to be from phosphate mining (SWFWMD, 2001). Based on the results of the Florid a Water Quality Assessment, most of this portion of the river and its tributaries we re given a “fair” water quality rating while Thirty Mile Creek was deemed most polluted and give n a “poor” rating on a scale of good, threatened, fair, poor, and unknown resulted. Based on water quality analysis from EPC data (station 115), it was determined that inor ganic nitrogen and phosphorous had decreased over time, while turbidity had increased (SWFWMD, 2001).

PAGE 27

12 The South Prong is the portion of the Alafia River which originates in the southeastern portion of the watershed and eventuall y converges with the North Prong at the center of the watershed. Phosphate mining was determined to be a major source of pollution in the South Prong (SWFWMD, 2001). Most of this portion of the river was given a “good” rating. While the headwaters histor ically received a “fair” rating, it was given a “good” rating showing an improvement in wat er quality in this area. Water quality analysis from EPC data (station 116) also s howed an increase in turbidity. A slight decrease in phosphorous and a major decrease in bacteria (since 1980) were observed in data from this station. At station 139 the only major change was determined to be a large decrease in phosphorous (SWFWMD, 2001 ). The Lower Alafia River consists of the portion of the river that originates at the merging of the North and South Prongs and flows to Hillsborough Bay and all of the tributaries that feed this part of the river. Sect ions of this part of the river were given anywhere from a “poor” to “good” rating (SWFWMD, 20 01). There was not enough data from the station 151 to detect any water quali ty trends. A slight increase in turbidity and a decrease in phosphorous and bacteria were obs erved from the data at station 114. The total and fecal coliform results from station 1 11 did not meet water quality standards more than 80 percent of the time. It was determine d that more resources should be devoted to identifying the source of these high col iform counts (SWFWMD, 2001). The SWFWMD (2001) determined potential sources of high nutrients in the Alafia River were attributed to atmospheric deposit ion, point source pollution, non-point source pollution, and groundwater in the form of na tural springs. The District anticipates higher nutrient loads into the river as development increases and expands into existing

PAGE 28

13 agricultural and upland portions of the watershed. Some of the goals of this CWM plan include preventing an average increase in nutrient loads to the river, strengthening interagency coordination concerning water quality i ssues in the watershed, increase the amount of preservation lands along the river shorel ine, improve industry best management practices, and conduct further studies o n water quality impairment in the river (SWFWMD, 2001). Statistical Analysis of Water Quality Trends Either parametric or nonparametric statistical tes ts are used in the analysis of trends. Parametric tests are based on the assumpti on that the data is normally distributed, the observations are independent, and the variance is constant, while nonparametric tests are used when the assumption of normality is violat ed (Berryman et al., 1988). Parametric statistical analysis is not typically re commended for water quality analysis since the data is often not normally distributed an d transformations to normality are not usually uniform across multiple datasets and may bi as the results of the analysis (Hirsch et al., 1991). Nonparametric tests have been shown to be more powerful than parametric tests when the data is not normally distributed and there is evidence of heterogeneity in the dataset (Berryman et al., 1988). Previous studies involving a monotonic trend hypot hesis (the assumption that there is a steady or abrupt trend over time with no previously determined change in the population, N) have often used the nonparametric Se asonal Kendall test to detect trends in water quality data due to the seasonal variabili ty that is typically exhibited in this type of dataset (Chang, 2008; Fraser, 1986; Qian et al., 2007a; Qian et al., 2007b; Smith et al.,

PAGE 29

14 1987; Yu et al., 1993). The Seasonal Kendall test is a test of the overall trend over the entire study period and is best suited for systemat ically sampled data (Hirsch et al., 1982). It is a modification of the Kendall tau tes t that accounts for seasonality and it is effective on datasets with missing data, outliers, tied values, and values reported as less than the detection limit (Smith et al., 1987; Yu et al., 1993). In the Seasonal Kendall test, seasonality is removed by making comparisons betwee n each of the values for each season (i.e. January values are compared to other J anuary values in the study period) and a Kendall tau value (rank) is assigned to that seas on between -1.0 and +1.0, with a -1.0 indicating a strong negative trend, +1.0 indicating a strong positive trend, and 0 indicating no trend (Cavanaugh and Mitsch, 1989). A Kendall tau statistic is then computed for each season and they are summed to obt ain the seasonal Kendall test statistic and determine the trend over the entire s tudy period (Millard and Neerchal, 2001). The Seasonal Kendall test is used to examine the r andomness of the dataset based on the null hypothesis (Ho) that the monthly dataset is a sample of independe nt random variables and there is no trend, and the alternativ e hypothesis (Ha) that one or more seasons in the dataset are not identically distribu ted. The formulas for computing this test statistic are described by Hirsch et al. (1982). T o estimate the magnitude of the trend, the Seasonal Kendall slope estimator is used. The Seas onal Kendall slope estimator is based on Sen’s estimator of slope, however, it accounts f or seasonality in the dataset. Like the Seasonal Kendall test for trend, the Seasonal Kenda ll slope estimator is a nonparametric test that is not affected by outliers or missing da ta. The Seasonal Kendall slope estimator

PAGE 30

15 is obtained by computing the median of all the indi vidual slope estimates for each season, the formula for which is described by Hirsch et al. (1982). In order to determine which specific statistical t est to use to analyze temporal trends in water quality data, it is important to in itially examine the characteristics of the dataset. Studies by Cavanaugh et al. (1989), Qian et al. (2007b), and Yu et al. (1993), have all used some or all of the following descript ive statistics for preliminary analysis of their dataset: Range (minimum and maximum values), mean, median, and standard deviation. Qian et al. (2007b) in particular, used the descriptive statistics information to explain the differences in variable concentration b etween the wet and dry seasons as well as make comparisons between the different sample st ations. Normality assumptions were tested using the Kolmogorov-Smirnov goodness of fit test in studies conducted by Sliva and Williams (2001) and Yu et al. (1993); while Leh rter (2006) used normal probability plots of the residuals versus the predicted values. To evaluate whether a dataset exhibits seasonal fluctuations, the Kruskal-Wallis test is o ften used (Hirsch et al., 1982; Sliva and Williams, 2001; Yu et al., 1993), while visually ex amining box-plots and running the Wilcoxan signed rank test have also been shown effe ctive (Qian et al., 2007b). Other aspects to check are whether there is homoge neity of the trends in the different seasons and whether the dataset values ar e serially correlated. Testing for the homogeneity of the trend will indicate whether ther e are differences in the trends between seasons with one going up and another going down. The Seasonal Kendall test does not account for heterogeneity in the trend and may prov ide inaccurate results indicating there is no significant trend in the dataset when actuall y the trend shifts between seasons. The Chi-Square test is often recommended to detect thes e fluctuations in the trend and if the

PAGE 31

16 trend is heterogeneous, the Mann-Kendall test is us ed to determine the trend for each of the season (Berryman et al., 1988; Gilbert, 1987; v an Belle and Hughes, 1984; Yu et al., 1993). It also is important to test for serial correlatio n to prevent type I errors of indicating there is a trend when actually no trend is present (rejecting the null hypothesis when the null hypothesis is true). There is a modi fied version of the Seasonal Kendall test for trend that is powerful even with serially correlated data and it has been shown to be effective with a dataset of greater than or equa l to 10 years of monthly data (12 months) and with a lag-one autocorrelation of less than or equal to 0.6 (Hirsch and Slack, 1984; Millard and Neerchal, 2001). The modified Se asonal Kendall test assumes the data are serially dependent so the covariance (the stren gth of the correlation) between the Seasonal Kendall test statistics is calculated whil e the original Seasonal Kendall test assumes the data are serially independent and the c ovariance terms are all zero. The modified test has been determined to be more precis e than the original Seasonal Kendall test concerning the significance of the trend when the data are serially correlated (Hirsch and Slack, 1984). Water Quality Management In water quality studies from Kebede et al. (2003) Tong and Chen (2002), and Xian et al. (2007), strong correlations have been m ade between land use and water quality. Each of these studies has made a strong n egative correlation between areas of urban land use and water quality degradation. If t he areas of urban lands continue to

PAGE 32

17 expand, there will have to be more coordination bet ween land use planning and water quality management to reduce the amount of pollutio n entering local water bodies. Land use planning and water quality management are typically managed separately with differing purposes. Land use plann ing often involves getting the most use from the land by humans in the future without negat ively impacting humans’ wellbeing. It is also usually confined by political boundaries such as by city, county, or state and non-local views are not equally incorporated into t he decision making process, if at all. On the other hand, water quality management is base d on monitoring and enhancing water quality (Wang, 2001). Watersheds often cross political boundaries, for example, the Apalachicola River watershed covers portions of Florida, Georgia, and Alabama and the Alafia River watershed cover portions of Hillsb orough and Polk Counties. Since water quality is evidently affected by urban develo pment, land use planning should be done in collaboration with water quality management to be successful at maintaining and improving water quality. In a staff report completed by Hillsborough County and the Southwest Florida Water Management District (SWFWMD), a collaborative effort was established between these governmental agencies in the form of an Alafi a River Basin Agency Team to better manage the Alafia River watershed. Both Hillsborou gh County and the SWFWMD’s jurisdictional boundaries include sections of as we ll as the entire Alafia River watershed, respectively. Hillsborough County’s responsibility in this watershed is growth management and one of SWFWMD’s main responsibilitie s is to protect water quality in the watershed. The main goals of this report were to improve the interagency communication, incorporate land use planning into w atershed management, and to give

PAGE 33

18 stakeholders the opportunity to participate in the land use and watershed planning (The Alafia River Basin Agency Team 2008). In the process of developing this staff report, th e Alafia River Basin Agency Team first gathered public and stakeholder input on their objectives and priorities for land and water management in the watershed. Once t his information was collected, the agencies and experts in watershed management determ ined the plans necessary for implementation and the actions necessary to accompl ish these goals. A three year plan was established which focused on water management, community enjoyment, and a sustainable ecosystem, slated to begin in 2000 and end in 2003. This plan identified the actions to be accomplished each year and the primar y agency responsible for completing each action (The Alafia River Basin Agency Team, 20 08). Watershed Management Policy Under the Federal Water Pollution Control Act of 1 972, now known as the Clean Water Act (CWA), guidelines were set for regulating point source pollution from municipal and industrial wastewater discharges to w aters of the states. This piece of legislation granted the U.S. EPA the authority to s et these effluent standards. The CWA established the National Pollutant Discharge Elimin ation System (NPDES) program in which a permit has to be obtained by point source p olluters that discharge directly to surface waters of the United States. These regulat ions for point source pollution have improved water quality since the enactment of the C WA; however, challenges still remain. It has become increasingly apparent that n on-point source pollution is a major contributor to water quality degradation. Many of the waters of the United States remain

PAGE 34

19 impaired due to non-point source pollution and an e ver-increasing population; therefore, it is becoming more difficult to meet surface water quality standards. In 1987, amendments were made to the CWA to focus more atten tion on non-point source pollution (Brady, 1996). The U.S. EPA is endorsing the Watershed Protection Approach (WPA) which is a framework for determining various sources of enviro nmental pollutants and their additive effects on surface and groundwater quality on a wat ershed scale. These stressors can then be prioritized and such programs as the TMDL p rogram can be implemented in the WPA. The U.S. EPA follows four main principles in applying the WPA. The first principle is for the application of a deep-rooted s cientific approach to assess all of the stressors within a watershed. This allows for all resources to be efficiently focused within that geographic area. The second principle involves stakeholder (anyone with a special interest in the watershed) participation in the identification and resolution of watershed problems. It is assumed that watershed a ctions will more likely be accepted by the public if stakeholder involvement is incorporat ed into the process. The third principle involves the integration of multiple stakeholders’ knowledge and expertise in implementing the actions determined in the analysis including watershed protection, monitoring, and education. The fourth and final pr inciple is the evaluation of the results to measure the success of the plan (Brady, 1996). In the mid-1990s, Florida began applying ecosystem management at the watershed scale in pilot projects located in the Ap alachicola River and Bay, Suwannee River, Lower St. Johns River, Florida Bay, Hillsbor ough River, and Wekiva River. Stakeholders from state and local governments, envi ronmental groups, and universities

PAGE 35

20 were involved in watershed monitoring, standards se tting, and developing plans to implement their findings. Studying these ecosystem s at the watershed scale gives environmental agencies insight into the effectivene ss of current environmental regulations. The goals of these projects are to de termine what changes need to be made in environmental regulations, financial allocations and administrative actions in order to effectively employ large scale ecosystem management (Brown and Marshall, 1996).

PAGE 36

21 Research Design Many Tampa Bay area residents use the Alafia River for recreational purposes and since 2003, the river has become a source of dr inking water for these people as well. As a result, the water quality in the river is an e nvironmental concern for many residents. Current environmental regulations are mainly based on point-source pollution. However, improper land use maintenance can contribute greatl y to polluted stormwater runoff entering area surface waters. It is becoming incre asingly apparent that non-point source pollution is a major contributing factor to surface water quality degradation. By understanding the trends in water quality in the Al afia River watershed and the possible sources of water quality degradation from improper land use practices, regulatory efforts can be better focused on implementing management pl ans to target the impaired areas of the watershed. The Problem This project examined trends in water quality, spec ifically dissolved oxygen, turbidity, fecal coliform, total nitrogen, and tota l phosphorus, in the Alafia River watershed during the time period between 1991 and 2 006. From this examination, a secondary analysis was conducted by visually compar ing these trends to changes in land use in the watershed. Gaining a better understandi ng of water quality trends and how they relate to land use changes in the Alafia River is beneficial in developing and

PAGE 37

22 implementing TMDLs for specific water quality param eters of concern and for implementing environmental regulations requiring th e use of best management practices for specific land use types. Research Questions The following research questions concerning the sp atial and temporal trends in water quality in the Alafia River watershed and how land use relates to these trends were addressed from this research: 1. How has water quality in the Alafia River watershed varied over time [i.e. by season (Wet/Dry), over the eight year time perio ds of 1991-1998 and 1999-2006, and over the entire study period (1991-2 006)]? 2. How has proportion of land use changed within th e “contributing zone” in between each sample station? 3. How does the change in the proportions of land u se relate to water quality relative to each sample station in the Alafia River ? Project Significance and Rationale Analysis of water quality and land use within the Alafia River watershed was conducted with the purposes of finding out how wate r quality has changed over the time period of study, examining how land use has changed over the study period, and

PAGE 38

23 determining how the water quality trends and land u se changes relate. Previous water quality research in the Tampa Bay area has focused on the larger watershed of the Tampa Bay which consists of ten major drainage basins, in cluding the Alafia River watershed. For example, the study by Xian et al. (2007) examin ed the effects of impervious surface area and population density on water quality in the Tampa Bay watershed. Few studies have been completed which examine water quality tre nds in the Alafia River watershed. This study provides information on trends in water quality throughout the Alafia River watershed and relates that information to changes i n land use within the watershed. This study can also serve as a model for future watershe d studies in the country, region, or the state.

PAGE 39

24 Study Area Tampa Bay Watershed Tampa Bay is located along the coast of west-centr al Florida and is Florida’s largest open-water estuary. The Tampa Bay watershe d spans roughly 6600 square kilometers and extends into Pinellas, Pasco, Polk, Manatee, Sarasota, and all of Hillsborough Counties (Figure 2) (Xian et al., 2007 ). The climate in this region is humid subtropical. The annual average temperature in thi s part of Florida is 72.2 degrees Fahrenheit with an average maximum temperature reac hing 91 degrees Fahrenheit during July and August and an average minimum temperature reaching 49 degrees Fahrenheit in January (SWFWMD, 2001). In the Tampa Bay area, urban development has grown substantially in the past several decades. Between 1960 and 2001, Hillsborou gh County has had the highest population growth out of all the counties in the Ta mpa Bay area at 158 percent rise in population (Xian et al., 2007). The Tampa Bay area serves many different economical uses including tourism, recreation, fisheries, as a major shipping channel, and a withdrawal source for future drinking water needs i n the area. Four major rivers drain into the Tampa Bay including the Hillsborough, Alaf ia, Little Manatee, and Manatee rivers. The Alafia River watershed will be the stu dy site of this project.

PAGE 40

25 Figure 2. The Tampa Bay Watershed. The Tampa Bay, Hillsborough River, Alafia River, Little Manatee River, and Manatee River wate rsheds all comprise the Tampa Bay watershed by each draining into the Tampa Bay. Alafia River Watershed The Alafia River watershed is located in central H illsborough County and portions of western Polk County and spans a total a rea of 1084.21 square kilometers (Xian et al., 2007). The Alafia River watershed is bordered by the Hillsborough River

PAGE 41

26 watershed to the north and the Little Manatee River watershed to the south (Figure 2). The Alafia River primarily originates at the conver gence of the North Prong creek and South Prong creek and flows about 38.6 kilometers ( World Wide Metric, 2006) to the Lower Hillsborough Bay of Tampa Bay and eventually to the Gulf of Mexico. The elevations in this watershed mainly range between a bout 7.6 to 22.9 meters (World Wide Metric, 2006), with some areas as high as about 39. 6 meters (World Wide Metric, 2006). On average, the Alafia River watershed receives abo ut 52.30 inches of rainfall each year, predominantly from the months of June through Septe mber (SWFWMD, 2001). Major land uses in the area consist of mining, agr iculture, and urban development. Phosphate mining is a major industry in this region and is mainly found in the eastern portion of the watershed. The majority of the agri cultural land use is found in the north eastern portion of the watershed in the form of dai ries, row crops such as strawberries, poultry, and citrus farms. Urban development is ma inly concentrated in the western portion but continues to expand east throughout the watershed (SWFWMD, 2001). Alafia River Watershed Sub-Drainage Basins The Alafia River watershed is divided into eight ma jor sub-drainage basins delineated by Hillsborough County government (Figur e 3). The main soil types in this watershed include Myakka, Basinger, and Holopaw (po orly drained) which is mainly found in the upland areas in the Northern and South ern portions of the watershed; Candler and Lake (excessively drained, rapidly perm eable) which is found in the central portion of the River major sub-drainage basin above the Alafia River; Winder, Chobee, and St. Johns (deep, poorly drained, moderate to sl owly permeable soil) is found along

PAGE 42

27 the North and South Prongs and where they converge; and Arents, Haplaquents, Quartzipsamments (soils of manmade areas) which is found in most of the Polk County section of the watershed and in the Turkey major su b-drainage basin (SWFWMD, 2001; USDA, 2007). Figure 3. Alafia River Major Sub-basins. The 8 ma jor sub-drainage basins in the Alafia River watershed, including Buckhorn, River, Bell, Fishhawk, Turkey, English, North Prong, and South Prong sub-basins.

PAGE 43

28 Methodology Contents of Dataset In order to determine the temporal trends in water quality in the Alafia River watershed, ambient water quality data from 1991 thr ough 2006 were obtained from the Hillsborough County Environmental Protection Commis sion (EPC). The EPC collects water quality samples from the Alafia River and its tributaries on a monthly basis at stations 74, 153, 114, 155, 166, 151, 111, 115, 116 154, and 139 (Figure 4). The eight monitoring stations that will be studied for this p roject will be stations 74, 114, 151, 111, 115, 116, 139, and 154 since the sample data has be en collected since at least 1991 (except station 154), allowing for a sufficient dat a set to be analyzed and the distribution of these monitoring stations provides sufficient sp atial coverage over the entire watershed. Station 154 was chosen based on its loc ation in the watershed. It is the only active monitoring station located in the northeaste rn main tributary of the Alafia River. The data from this station dates back to the year 1 999. All of the water quality values in the dataset are based on the actual value reported by the sample device/machine, so there are no values reported as less than the limit of de tection. Land use data was obtained from the Florida Geogra phic Data Library for the years 1990, 1999, and 2006. The shape files are ar ranged by county within the water management districts. The features were obtained f rom 1:24,000 and 1:12,000 USGS color infrared digital orthophoto quarter quadrangl es (DOQQ) for the years 1990 and

PAGE 44

29 1999, respectively, and from 1:12,000 color infrare d digital aerial photographs for the year 2006. The land use data in the shape file is organized by Florida Land Use and Cover Classification System (FLUCCS) codes. There are nine main land use classifications in the study area for the years 199 0 and 2006 including agriculture, barren land, range land, special classifications (vegetati on), transportation, communication, and utilities, upland forests, urban and built up, wate r, and wetlands. The shape file for the year 1999 includes all of the same land use classif ications except for special classifications (vegetation) which is included in t he wetlands classification. The National Hydrography Dataset (NHD) for the Alaf ia River watershed was obtained from the United States Geological Survey ( USGS) in a geographic information system (GIS) geodatabase format. The NHD data are at a resolution of 1:24,000-scale (high resolution) and contain the network of water flow for the watershed’s rivers, streams, and artificial paths.

PAGE 45

30 Figure 4. Alafia River Monthly Sampling Stations. The EPC ambient water quality sampling stations in the Alafia River watershed. T he sample stations are symbolized with red stars are the stations used in this study. Sample Analysis Methods All water quality sample data utilized in this stu dy were obtained from the EPC. The EPC laboratory is certified by the Florida Depa rtment of Health to test water samples for several different analytes, including those use d in this study. The analytes and their corresponding analysis methods are listed in Table 1 (Clesceri et al., 1998; U.S. EPA, 2008a). Another sample parameter that will be used in this study is dissolved oxygen which is measured in the field with a Hydrolab mult iprobe sonde. Florida Department of

PAGE 46

31 Environmental Protection Standard Operating Procedu res for Field Activities (DEP-SOP001/01), last updated February 1, 2004, are followe d when collecting water quality samples. Table 1. Sample Methods of Analysis. Analytes and their corresponding method(s) of analysis used by the EPC to test water quality s amples (Clesceri et al., 1998; U.S. EPA, 2008a). Analyte Method Nutrients Total Nitrogen SM 4500 N 20th Edition Total Phosphorous EPA 365.4 Microbiology Fecal Coliform SM 9222D 20th Edition General Chemistry Turbidity EPA 180.1 Sample Design The Alafia River watershed was chosen as the study area due to its ease of accessibility and the availability of monthly ambie nt water quality sampling data from the EPC. The water quality sampling by the EPC has bee n conducted off and on at various sampling locations along the Alafia River since 197 2, providing a significant historical database of water quality information. Since the o riginal date of sampling began at different times for each of the sampling stations a nd some stations are no longer actively sampled, the time period chosen for this analysis i s from 1991 through 2006. This time period provides a sufficient amount of sampling sta tions to make more accurate determinations based on the data. An additional sa mple station (station 154) which has data that dates back to 1999 was also chosen since no other active sample stations are located near the headwaters of the North Prong. Th is study area was also chosen due to

PAGE 47

32 the limited detailed evaluations of land use change s and water quality trends in this specific watershed. Water Quality Data Analysis Data Storage and ManipulationMonthly ambient water quality data provided by the EPC is stored in a Microsoft Office Access ( Microsoft Corporation, 2003a) database. A query of the water quality parameters that will be used in this study including dissolved oxygen (DO), turbidity, fecal c oliform, total nitrogen (TN), and total phosphorous (TP) along with the sample station and sample date was conducted. The information queried from the Microsoft Office Acces s (Microsoft Corporation, 2003a) database was converted into a Microsoft Office Exce l (Microsoft Corporation, 2003b) spreadsheet for manipulation based on sample statio n and date. Although the statistical tests used in this study are powerful, even with outliers, the standardized and studentized residuals of the d atasets were examined for obvious incorrect or highly extreme values using SPSS (SPSS Inc., 2007); however, no datum was excluded based on the test for outliers alone ( Gilbert, 1987). Five values were removed from the dataset including: the August 200 4 turbidity outlier (173 NTU) for station 116 since it was suspected of being an inco rrectly transcribed value missing a decimal point and the December 1997 total nitrogen and total phosphorus outliers for station 115 (46.26 mg/L and 220.21 mg/L, respective ly) and downstream at station 114 (37.43 mg/L and 234.83 mg/L, respectively) due to a large acidic wastewater spill from a gypsum stack dam breach at the Mulberry Phosphates fertilizer plant that discharged to the north prong of the Alafia River three days prio r to sample collection.

PAGE 48

33 Since some of the statistical tests used in the ana lysis are not effective with missing values (e.g. Rank von Neumann test) or with more than one missing value per season per year (e.g. Seasonal Kendall test adjuste d for serial correlation), SPSS (SPSS Inc., 2007) was used to replace missing values with the mean of the surrounding values (Yu et al., 1993). Since sampling did not begin un til more than half-way through the year for stations 151 and 154, the sample data from 1991 (August – December) were deleted from station 151 and the sample data from 1999 (Sep tember – December) were deleted from station 154, rather than replacing the missing values. Although this caused station 151 to have one year less of data than the study pe riod analyzed, Hirsch et al. (1991) indicate that if there are only one to two years of missing values in the dataset, or no more than 20 percent missing per period when the da taset is divided into three equal time periods, then the dataset should still be included in the analysis. Although the dataset for station 154 does not match these criteria, it was s till included in the analysis since it was the only station located in the northeastern portio n of the watershed. Sample dates (in the format mm/dd/yyyy) were transf ormed using the SPSS’s (SPSS Inc., 2007) Date and Time Wizard into the two new variables: sample year and sample month. The sample months were then recoded into seasons (dry season, dummy variable = 1 and wet season, dummy variable = 2) wi th the dry season covering the months of October through May and the wet season co vering the months of June through September (SWFWMD, 2001). The sample years were al so recoded into two equally split groups; one for the first eight years (1991 – 1998, dummy variable = 1) and one for the second eight years (1999 – 2006, dummy variable = 2). These dataset groupings (year, season, and eight-year time periods) were us ed to analyze the descriptive statistics

PAGE 49

34 [mean, median, range (minimum and maximum), and sta ndard deviation] for each variable at each sample station. The dataset was t hen analyzed using statistical procedures. Statistical Analysis. All statistical analyses were carried out using eit her SPSS (SPSS Inc., 2007) or EnvironmentalStats (Millard, 2 002) for S-PLUS (Insightful Corp., 2007) with a significance level of 0.05 ( p -value 0.05). SPSS (SPSS Inc., 2007) was utilized to conduct the Shapiro-Wilk test, which te sts the null hypothesis that the dataset is normally distributed (Millard and Neerchal, 2001 ). The normality assumption was tested for every sample station and turbidity, feca l coliform, total nitrogen (except station 154), and total phosphorus were all not assumed to come from a normal distribution. For dissolved oxygen, three out of the eight station’s datasets show evidence the values are not normally distributed while the others (stations 74, 115, 116, 151, and 154) were normally distributed. Although some groups showed signs of normality, nonparametric tests will be utilized in this study since most of the data has shown departures from normality, comparisons will be made between all sam ple stations, and the nonparametric Seasonal Kendall test for trend has been shown to o nly be slightly less powerful than its parametric counterpart when the data are normally d istributed (Hirsch et al., 1982). The complete dataset was transferred over to the st atistical program S-PLUS (Insightful Corp., 2007) for further statistical an alysis. The nonparametric KruskalWallis one-way analysis of variance test was used t o determine whether the datasets showed signs of seasonality (Hirsch et al., 1982; S liva and Williams, 2001; Yu et al., 1993). All of the water quality parameters, except fecal coliform, exhibited seasonality at

PAGE 50

35 the majority of sample stations. Fecal Coliform sh owed significant seasonality at only two out of the eight sample stations. Examination of the annual seasonal trends revealed that although the fecal coliform annual seasonal me dians were typically similar for the wet and dry season, the wet season showed more extr eme variability from year to year than in the dry season (Figures 10-A through 10-H). As a result, it was determined that a nonparametric test that accounts for seasonality wo uld to be used to test for trends in the datasets. The nonparametric rank von Neumann ratio test for lag-1 serial correlation was used to test the null hypothesis that the time seri es distribution is completely random. Lag-1 serial correlation or autocorrelation tests t he correlation of observations one time unit apart. This test does not account for seasona lity in the dataset, so the seasonal variability was removed by subtracting the seasonal median from each observation in the season for every sample station (Qian et al., 2007b ). The results of the test indicated that the majority of the datasets were serially dependen t and only dissolved oxygen from sample station 74 and dissolved oxygen, turbidity, and fecal coliform from station 154 were determined to be serially independent. The Chi-Square test for heterogeneity was used to test the null hypothesis that the seasonal trends are heterogeneous with some having an upward trend and some a downward trend (Berryman et al., 1988; Gilbert, 198 7; van Belle and Hughes, 1984; Yu et al., 1993). Based on this test, the trends betw een seasons for all datasets showed no significant heterogeneity. As a result of conducting these preliminary tests, it was determined that the modified Seasonal Kendall test that accounts for se rial correlation would be used to test

PAGE 51

36 for trends in water quality and the Seasonal Kendal l slope estimator would be used to estimate the slope of the trend (Hirsch and Slack, 1984; Millard and Neerchal, 2001). The modified Seasonal Kendall test was used to redu ce the probability of making a Type I error (rejecting the null hypothesis when the nul l hypothesis is true) since the data showed signs of serial dependence. The modified Se asonal Kendall test is used to test the null hypothesis that there is no trend in the conce ntrations of a water quality parameter. If the modified Seasonal Kendall test for trend is significant, indicating an overall trend in the data, then the Seasonal Kendall slope estima tor provides an estimate of the slope and the direction (upwards or downwards) of the tre nd. Land Use Data Analysis Data Storage and Manipulation Before determining the temporal changes in land use in the study area, the geographic data fir st had to be manipulated using the ArcMap (ESRI, 2008) GIS program. Land use informat ion for the years 1990, 1999, and 2006 is stored in separate GIS shape files for Hill sborough and Polk Counties. The two shape files had to be merged for each study year to combine the land use attributes for analysis. Since the 1999 land use file only listed eight different types of land use and the special classification (vegetation) descriptors wer e encompassed in the wetlands classification, the 1990 and 2006 land use informat ion was adjusted to resemble the 1999 data. To do so, a new table was created in Microso ft Office Excel (Microsoft Corporation, 2003b) that contained the revised land use classifications and it was joined to the attribute table in ArcMap (ESRI, 2008) using common field identification numbers.

PAGE 52

37 To examine the contributing land use area for each sample station, the upstream flow path had to first be established by evaluating the National Hydrography Dataset (NHD) geodatabase from the United States Geological Survey (USGS) using ArcMap (ESRI, 2008) and by visually examining aerial image s. The upstream flow path for each sample station was traced using the Utility Network Analysis toolbar and the NHDFlowline (the main surface water drainage system ) and HYDRO_NET_Junctions (nodes connecting the NHDFlowlines used to determin e the flow path) from the NHD geodatabase for the Alafia River watershed. Any ga ps in the upstream path were visually analyzed using aerial images and, as deemed necessa ry, were connected by creating a new feature using the Editor toolbar. Once the ups tream flow path was obtained for a sample station, the river section was selected and converted into a new shapefile. A 100 meter buffer was created around each river s ection to represent the “contributing zone” of land use pollution to each s ample station (Brown and Vivas, 2005; Johnson et al., 1997; Sliva and Williams, 2001). A ll overlapping sections of the buffer were removed using the Editor toolbar, so there was no redundancy in information. The land use shapefile for each year was then clipped t o the 100 meter buffer around the river for each sample station for future analysis. Geographic Information System Data AnalysisTo determine the temporal changes in land use within 100 meters of the Alafia River, sample stations were analyzed individually and comparisons were made between each land use classification area of coverage and the total contributing area. The attr ibute tables for each sample station’s land use layer contain information about the area o f coverage for each polygon

PAGE 53

38 representing a particular land use classification. The area of the polygons did not automatically update once the land use layer was cl ipped to the contributing zone, so the area had to be recalculated for each sample station for each year. For 1990 and 2006 land use shapefiles, the land use area is given in acres while the 1999 shapfiles are in square meters. The recalculation was done in the attribut e table by calculating the geometry of the area with the output units all set to square ki lometers for consistency. Once all of the area values were converted to squar e kilometers, a report was generated and exported to Microsoft Office Word (Mi crosoft Corporation, 2003c). After converting the Microsoft Office Word (Microso ft Corporation, 2003c) document into a Microsoft Office Excel (Microsoft Corporati on, 2003b) spreadsheet for formatting, the tables of land use classifications and associated square kilometer values for each polygon were imported into a Microsoft Off ice Access (Microsoft Corporation, 2003a) database. In Microsoft Office Access (Microsoft Corporation, 2003a), a query was conducted to calculate the sum of the square kilometers from each of the land use classifications for the individual sam ple stations. Microsoft Office Excel (Microsoft Corporation, 20 03b) was used to calculate the proportion of land use coverage for each sample station as well as calculate the percent change in the proportion of each land use a rea between the years 1990, 1999, and 2006. The proportion of land use coverage for each sample station was determined by dividing the area of each land use classification b y the total area within the sample stations contributing zone. The percent change in the proportion of each land use was calculated using the formula: [(new value – old va lue)/ |old value|]*100.

PAGE 54

39 Results and Discussion The descriptive statistics, specifically the mean, median, range (minimum to maximum), and standard deviation, were examined for each water quality variable at each sample station. The descriptive statistics we re grouped by year, season (wet vs. dry), by land use years (1991 – 1998 and 1999 – 200 6 except station 154 since its study period began in 2000), and by the entire study peri od to get a preliminary idea of the basic features of the datasets [Tables 2, 4, 6, 8, and 10 (all A through H)]. Monthly medians (the median of each of the 12 individual mo nths over the entire study period) were plotted for each of the water quality paramete rs to examine monthly trends for each sample station and to make comparisons between samp le stations based on the month the sample was collected (Figures 5, 7, 9, 11, and 13). Annual seasonal median water quality concentration s were also examined to determine the seasonal trends over the entire study period and to make comparisons between seasons and sample stations [Figures 6, 8, 10, 12, and 14 (all A through H)]. The seasons were defined as the wet season (June th rough September) and the dry season (January through May and October through December) (SWFWMD, 2001). Along with analyzing the water quality data for monthly and se asonal trends, the data was statistically analyzed to determine whether signifi cant trends (significance level of =

PAGE 55

40 0.05) existed over the entire study period of 1991 – 2006 (except station 151, analyzed from 1992 – 2006 and station 154, analyzed from 200 0 – 2006). Land use data for the years 1990, 1999, and 2006 w ere analyzed and compared to determine how proportions of land use have changed over the period of study and to see if correlations could be made with water quality tr ends for each sample station. The land use proportions and percent change tables discussed in this section were important in the overall scheme of this study, but were included in the appendix due to the amount of data included in each of the tables. Figure 5 depicts the locations of the eight sample stations within the Alafia River watershed that were used in this study. The map wa s included in this section as a quick reference to the locations of each of the sample st ations when reviewing the results of this analysis.

PAGE 56

41 Figure 5. The eight sample stations in the Alafia River watershed used in this study. Each sample station is symbolized with a red star. Dissolved Oxygen According to state standards for Class III fresh w aters [Rule 62-302, Florida Administrative Code (F.A.C.)], dissolved oxygen sho uld be maintained at levels at or above 5.0 mg/L. Review of the descriptive statisti cs for sample station 74 revealed the median dissolved oxygen concentration over the enti re study period was below the state standard with a value of 4.6 mg/L and the range ind icated levels had dipped to as low as 0.20 mg/L (Table 2-A). This implies that dissolved oxygen is degraded at station 74, located at the mouth of the river (Figure 5). Prev ious studies done by the SWFWMD

PAGE 57

42 (2001) and Parsons Engineering Science, Inc. (2002) have also noted low dissolved oxygen levels in this portion of the river. Some o f the low levels of dissolved oxygen at this station could be attributed to samples being c ollected from this station earlier in the morning than the other stations, since dissolved ox ygen is highest in the afternoon due to increased photosynthesis. Also, this station is lo cated at the mouth of the river and higher salinity levels can lower dissolved oxygen. The descriptive statistics also showed that dissolv ed oxygen concentrations were lower during the wet season for every station, with the greatest difference in the median concentration between the seasons exhibited at stat ion 74 (2.60 mg/L during the wet season and 5.60 mg/L during the dry season) (Table 2-A). This indicates that water quality for dissolved oxygen declines during the we t season and improves during the dry season. When compared to sample station 74, all other stati ons exhibited less variability in dissolved oxygen based on the overall range and in the median concentration when grouped by year and by season, with station 114 sho wing the least amount of variability (Tables 2-A through 2-H). This indicates that diss olved oxygen at station 114 has remained fairly consistent throughout the 16 year s tudy period, and if conditions remain the same, it would be expected that levels at this station would continue on this trend. When grouped by land use years, all of the sample s tations except station 74 showed a lower median concentration during the seco nd half of the study period (19992006) (Tables 2-A through 2-H). This indicates the median concentration at those stations had degraded during the second half of the study period when compared to the first, while the median dissolved oxygen concentrat ion at station 74 slightly improved.

PAGE 58

43 Table 2-A. Descriptive statistics of dissolved oxy gen (mg/L) for station 74. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 5.00 4.90 2.50 7.10 1.39 1992 4.56 4.35 0.20 7.80 2.51 1993 4.45 3.80 2.60 7.20 1.62 1994 4.39 4.75 1.50 6.20 1.48 1995 4.68 4.63 0.60 8.47 2.35 1996 4.48 3.95 1.00 8.00 2.39 1997 4.92 4.75 1.70 8.50 1.91 1998 4.51 4.75 1.40 6.70 1.52 1999 4.29 4.55 0.80 7.70 2.23 2000 4.95 5.15 1.40 10.30 2.28 2001 4.37 5.00 0.70 6.30 1.64 2002 4.18 3.90 0.40 8.80 2.46 2003 4.64 4.60 1.30 7.10 1.79 2004 5.08 5.65 2.06 6.79 1.46 2005 4.30 4.25 0.66 7.34 2.31 2006 4.39 4.12 1.18 8.05 2.41 Grouped by Season (wet vs. dry) Season Mean Median Range Standard Deviation Wet 2.68 2.60 0.20 5.70 1.35 Dry 5.52 5.60 2.80 10.30 1.47 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 4.62 4.50 0.20 8.50 1.88 1999-2006 4.53 4.72 0.40 10.30 2.05 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 4.57 4.60 0.20 10.30 1.96

PAGE 59

44 Table 2-B. Descriptive statistics of dissolved oxy gen (mg/L) for station 111. Descriptive statistics of dissolved oxygen (mg/L) f or sample station 111. Separated by year (n=12), season [wet (n=64) vs. dry (n=128)] grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Group by Year Year Mean Median Range Standard Deviation 1991 6.27 6.30 4.10 9.30 1.19 1992 7.32 7.10 5.70 8.70 1.07 1993 7.31 7.65 3.70 9.40 1.42 1994 6.98 6.30 5.30 10.50 1.46 1995 7.52 7.35 4.60 10.20 1.66 1996 7.66 7.30 4.70 9.60 1.42 1997 6.91 6.65 3.70 9.00 1.55 1998 7.28 7.20 5.00 8.60 1.03 1999 7.35 7.26 3.00 9.70 1.65 2000 6.02 6.00 2.90 9.30 2.06 2001 7.14 6.91 5.90 8.60 0.82 2002 7.12 6.84 5.30 10.30 1.27 2003 7.36 7.00 6.30 9.50 1.06 2004 7.61 7.00 5.48 11.30 2.09 2005 7.57 7.39 5.28 10.79 1.78 2006 7.35 7.14 5.55 9.82 1.34 Group by Season Season Mean Median Range Standard Deviation Wet 6.11 6.26 3.00 7.40 0.86 Dry 7.70 7.60 2.90 11.30 1.44 Group by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 7.15 7.10 3.70 10.50 1.38 1999-2006 7.19 7.00 2.90 11.30 1.58 Group by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 7.17 7.00 2.9 11.30 1.48

PAGE 60

45 Table 2-C. Descriptive statistics of dissolved oxy gen (mg/L) for station 114. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 6.33 6.25 5.80 7.40 0.50 1992 6.42 6.25 5.10 7.50 0.69 1993 6.82 6.85 5.70 8.20 0.79 1994 6.54 6.35 5.80 8.10 0.71 1995 6.88 6.55 5.70 9.00 0.99 1996 6.78 6.60 5.80 8.20 0.79 1997 6.57 6.30 5.20 8.60 1.12 1998 6.63 6.50 5.50 8.20 0.77 1999 6.45 6.35 6.00 7.20 0.38 2000 6.34 6.25 5.80 7.50 0.53 2001 6.01 6.00 5.50 6.70 0.37 2002 6.63 6.30 5.50 8.30 0.84 2003 6.64 6.36 5.72 8.80 0.95 2004 6.28 6.30 4.26 8.03 1.07 2005 6.98 6.99 5.24 8.82 1.13 2006 6.77 6.14 5.55 9.67 1.40 Grouped by Season Season Mean Median Range Standard Deviation Wet 5.93 5.97 4.26 7.74 0.44 Dry 6.88 6.76 5.42 9.67 0.85 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 6.62 6.40 5.10 9.00 0.81 1999-2006 6.51 6.30 4.26 9.67 0.92 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 6.57 6.30 4.26 9.67 0.86

PAGE 61

46 Table 2-D. Descriptive statistics of dissolved oxy gen (mg/L) for station 115. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 7.74 7.80 5.90 10.40 1.37 1992 8.00 8.30 6.90 9.50 0.92 1993 7.70 7.75 5.10 9.30 1.07 1994 7.33 7.50 5.60 9.20 1.12 1995 7.64 7.55 5.90 10.00 1.14 1996 7.62 7.70 6.20 9.20 0.90 1997 7.79 7.90 6.20 10.20 1.08 1998 7.29 7.15 6.00 8.30 0.73 1999 7.59 7.60 6.10 9.00 0.93 2000 7.88 8.05 6.20 9.40 1.06 2001 7.61 7.80 5.50 9.30 1.28 2002 7.50 7.10 6.10 10.30 1.28 2003 7.14 6.64 5.73 9.28 1.29 2004 6.78 6.67 4.19 9.87 1.88 2005 7.02 7.22 4.85 8.79 1.36 2006 7.82 7.94 5.74 10.19 1.52 Grouped by Season Season Mean Median Range Standard Deviation Wet 6.35 6.40 4.19 7.70 0.70 Dry 8.12 8.13 4.53 10.40 0.95 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 7.64 7.65 5.10 10.40 1.04 1999-2006 7.42 7.43 4.19 10.30 1.35 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 7.53 7.50 4.19 10.40 1.21

PAGE 62

47 Table 2-E. Descriptive statistics of dissolved oxy gen (mg/L) for station 116. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 8.17 8.10 5.80 11.00 1.54 1992 8.08 7.80 6.50 10.70 1.37 1993 7.61 7.70 5.20 9.40 1.05 1994 7.13 7.40 5.30 9.00 1.21 1995 7.36 7.85 5.40 9.50 1.36 1996 7.85 8.00 5.90 9.10 0.97 1997 7.88 7.95 5.80 10.40 1.33 1998 7.05 7.00 5.60 8.10 0.77 1999 7.63 8.00 5.80 9.20 1.11 2000 7.65 8.05 5.00 9.70 1.46 2001 7.13 7.10 4.60 9.50 1.57 2002 7.23 6.75 5.90 9.80 1.32 2003 6.81 6.33 5.22 9.20 1.42 2004 6.98 6.96 3.95 10.07 1.94 2005 6.99 7.05 4.54 9.03 1.40 2006 7.49 7.35 5.47 10.05 1.70 Grouped by Season Season Mean Median Range Standard Deviation Wet 6.15 5.99 3.95 8.00 0.83 Dry 8.08 8.10 4.90 11.00 1.12 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 7.64 7.75 5.20 11.00 1.24 1999-2006 7.24 7.17 3.95 10.07 1.48 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 7.44 7.50 3.95 11.00 1.38

PAGE 63

48 Table 2-F. Descriptive statistics of dissolved oxy gen (mg/L) for station 139. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 7.38 7.50 4.80 9.90 1.41 1992 7.33 7.40 4.80 9.10 1.53 1993 7.08 6.95 5.30 9.20 1.05 1994 6.82 7.05 5.00 8.60 1.30 1995 6.68 6.50 4.90 9.10 1.30 1996 7.12 7.40 5.50 8.40 1.08 1997 7.38 7.70 4.80 9.70 1.27 1998 6.65 6.50 4.90 7.80 0.83 1999 6.93 7.60 5.20 8.90 1.40 2000 7.91 7.70 6.50 9.10 0.89 2001 6.88 7.05 3.00 8.80 1.74 2002 6.99 7.25 5.20 8.60 1.09 2003 6.25 6.15 4.51 8.30 1.28 2004 6.13 6.47 3.83 8.94 1.85 2005 6.60 6.53 4.64 9.33 1.40 2006 6.88 7.08 4.29 9.54 1.88 Grouped by Season Season Mean Median Range Standard Deviation Wet 5.60 5.50 3.00 7.80 0.98 Dry 7.61 7.70 4.82 9.90 1.01 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 7.05 7.20 4.80 9.90 1.23 1999-2006 6.82 6.94 3.00 9.54 1.51 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 6.94 7.10 3.00 9.90 1.38

PAGE 64

49 Table 2-G. Descriptive statistics of dissolved oxy gen (mg/L) for station 151. Separated by year (n=12), season [wet (n=60) vs. dr y (n=120)], grouped by land use information [1992-1998 (n=84) and 1999-2006 (n=96)] and by the entire study period [1992-2006 (n=180)]. Grouped by Year Year Mean Median Range Standard Deviation 1992 6.90 6.65 4.90 9.00 1.52 1993 6.97 7.60 4.30 9.40 1.59 1994 6.35 5.95 4.00 10.10 1.73 1995 7.54 7.45 4.70 10.50 1.65 1996 8.26 8.70 6.50 9.80 1.03 1997 6.78 6.45 4.00 10.30 1.94 1998 7.73 7.80 6.70 8.90 0.80 1999 7.86 7.70 6.80 9.30 0.86 2000 6.90 6.40 5.10 9.40 1.35 2001 7.09 7.20 5.20 9.60 1.24 2002 7.18 6.75 5.80 9.40 1.19 2003 7.36 6.85 5.73 9.80 1.49 2004 7.58 7.42 4.95 11.11 1.98 2005 7.73 7.47 4.90 11.05 1.85 2006 7.62 7.49 5.51 10.19 1.53 Grouped by Season Season Mean Median Range Standard Deviation Wet 6.00 6.00 4.00 7.50 0.85 Dry 7.98 7.95 4.00 11.11 1.33 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1992-1998 7.22 7.35 4.00 10.50 1.58 1999-2006 7.42 7.20 4.90 11.11 1.46 Grouped by the Entire Study Period Years Mean Median Range Standard Deviation 1992-2006 7.32 7.20 4.00 11.11 1.51

PAGE 65

50 Table 2-H. Descriptive statistics of dissolved oxy gen (mg/L) for station 154. Separated by year (n=12), season [wet (n=28) vs. dr y (n=56)], and by the entire study period [2000-2006 (n=84)]. Grouped by Year Year Mean Median Range Standard Deviation 2000 7.07 7.25 5.20 8.60 1.02 2001 6.48 6.45 3.30 8.10 1.30 2002 6.53 6.56 4.60 8.20 1.14 2003 6.35 5.81 5.33 8.80 1.17 2004 6.22 5.91 4.11 9.03 1.47 2005 6.40 6.22 4.66 8.94 1.37 2006 6.77 6.57 5.20 9.00 1.31 Grouped by Season Season Mean Median Range Standard Deviation Wet 5.55 5.67 3.30 6.80 0.71 Dry 7.05 7.18 4.60 9.03 1.15 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 2000-2006 6.55 6.41 3.30 9.03 1.25 In Figure 6, the dissolved oxygen monthly trends w ere compared for each of the sample stations. There is evidence of seasonal var iability for each of the sample stations as depicted by the decrease in the monthly median c oncentration of dissolved oxygen during the wet season (June through September), wit h it being most evident at station 74. When analyzed by month, the trends in dissolved oxy gen over the entire study period support what was observed in the descriptive statis tics with levels declining during the wet season. All of the stations exhibited similar median monthly concentrations except station 74 which had the lowest monthly median conc entration of 1.7 mg/L for the month of July; 3.4 mg/L lower than the next lowest concen tration which was at station 139. The

PAGE 66

51 highest median concentration of dissolved oxygen fo r every sample station was evident during the dry season for the months of December an d January. 0 1 2 3 4 5 6 7 8 9 10 1234567891011 MonthDissolved Oxygen (mg/L) Station 74 Station 111 Station 114 Station 115 Station 116 Station 139 Station 151 Station 154 Figure 6. Each sample station’s monthly dissolved oxygen median. The monthly median for all 12 months over the entire study peri od [1991 – 2006, except Station 151 (1992 – 2006) and Station 154 (2000 – 2006)]. The seasonal trends were examined for dissolved ox ygen for each sample station (Figures 7-A through 7-H). For every sample statio n, the annual median concentration was lower during the wet season. This decline in d issolved oxygen levels during the wet season could be the result of increased stormwater runoff which carries sediment into the river and warmer temperatures, both reducing the ab ility of oxygen to dissolve in the water. The opposite was determined in the study by Sliva and Williams (2001), conducted in Ontario, Canada, which determined that dissolved oxygen was highest during Ontario’s wet season (spring) and lowest in the dry season (summer). However,

PAGE 67

52 they were unable to show any correlation between di ssolved oxygen and land use which they attributed to temporal variability (Sliva and Williams, 2001). Station 74 tended to have the largest difference i n annual median concentrations between the wet and the dry seasons, with an averag e difference of 2.7 mg/L. Station 74 also exhibited the lowest overall annual median con centrations for each season with all values less than 6.1 mg/L. As indicated before, th is could be due to higher salinity in this portion of the river or due to the time of day the samples were collected when compared to the other stations. In any case, the degraded d issolved oxygen at this station should be further investigated. 0 1 2 3 4 5 6 7 8 9 10 19911993199519971999200120032005 YearDissolved Oxygen (mg/L) Dry Season Wet Season Figure 7-A. Dry versus wet annual seasonal median trends of dissolved oxygen (mg/L) for station 74.

PAGE 68

53 0 1 2 3 4 5 6 7 8 9 10 19911993199519971999200120032005 Year Dissolved Oxygen (mg/L) Dry Season Wet Season Figure 7-B. Dry versus wet annual seasonal median trends of dissolved oxygen (mg/L) for station 111. 0 1 2 3 4 5 6 7 8 9 10 19911993199519971999200120032005 Year Dissolved Oxygen (mg/L) Dry Season Wet Season Figure 7-C. Dry versus wet annual seasonal median trends of dissolved oxygen (mg/L) for station 114.

PAGE 69

54 0 1 2 3 4 5 6 7 8 9 10 19911993199519971999200120032005 Year Dissolved Oxygen (mg/L) Dry Season Wet Season Figure 7-D. Dry versus wet annual seasonal median trends of dissolved oxygen (mg/L) for station 115. 0 1 2 3 4 5 6 7 8 9 10 19911993199519971999200120032005 Year Dissolved Oxygen (mg/L) Dry Season Wet Season Figure 7-E. Dry versus wet annual seasonal median trends of dissolved oxygen (mg/L) for station 116.

PAGE 70

55 0 1 2 3 4 5 6 7 8 9 10 19911993199519971999200120032005 Year Dissolved Oxygen (mg/L) Dry Season Wet Season Figure 7-F. Dry versus wet annual seasonal median trends of dissolved oxygen (mg/L) for station 139. 0 1 2 3 4 5 6 7 8 9 10 19921994199619982000200220042006 YearDissolved Oxygen (mg/L) Dry Season Wet Season Figure 7-G. Dry versus wet annual seasonal median trends of dissolved oxygen (mg/L) for station 151.

PAGE 71

56 0 1 2 3 4 5 6 7 8 9 10 2000200120022003200420052006 Year Dissolved Oxygen (mg/L) Dry Season Wet Season Figure 7-H. Dry versus wet annual seasonal median trends of dissolved oxygen (mg/L) for station 154. Analysis of the datasets for dissolved oxygen reve aled that sample station 116 had a statistically significant decreasing trend (p-val ue = 0.0278) over the study period with an estimated annual trend of -0.0527 mg/year (Table 3). This provides evidence that the water quality is slowly degrading over time; howeve r, from the review of the descriptive statistics from this site, the median dissolved oxy gen sample results indicate this water quality variable is still being maintained over the minimum state standard of 5.0 mg/L. None of the other sample stations were determined t o have significant trends in the overall dissolved oxygen data.

PAGE 72

57 Table 3. Modified Seasonal Kendall trend test resu lts (significance level = 0.05) for dissolved oxygen by sample station. Station N Z-statistic modified Seasonal Kendall p-value Seasonal Kendall slope estimator Trend 74 192 -0.6215 0.5343 -0.0087 No Trend 111 192 0.5150 0.6065 0.0107 No Trend 114 192 -0.3677 0.7131 0.0000 No Trend 115 192 -1.8113 0.0701 -0.0323 No Trend 116 192 -2.2004 0.0278 -0.0527 Decreasing 139 192 -1.5731 0.1157 -0.0333 No Trend 151 180 0.8251 0.4093 0.0284 No Trend 154 84 -1.4312 0.1524 -0.0967 No Trend Turbidity According to state standards for Class III surface waters, turbidity should be maintained less than or equal to 29 NTU above natur al background conditions (Rule 62302, F.A.C.). The descriptive statistics for turbi dity showed the greatest deviation from the mean at sample station 74 with a mean concentra tion of 6.06 NTU 6.13 (Table 4A). This indicates that at this station, turbidity concentrations showed the highest variability. Some of this high variability may hav e led to depressed dissolved oxygen levels that were exhibited at this station since di ssolved oxygen is less easily dissolved in waters with high suspended solids. Stations 111 an d 115 had the highest median concentration of turbidity out of all the sample st ations over the entire study period with a concentration of 6.00 NTU (Tables 4-B and 4-D). Al l of the sample stations had a higher median turbidity concentration during the first hal f of the study period when grouped by land use years which provides evidence of improveme nt in water quality for this variable during the second half of the study period (Tables 4-A through 4-H).

PAGE 73

58 Table 4-A. Descriptive statistics of turbidity (NT U) for station 74. Separated by year (n=12), season [wet (n=64) vs. dry (n=128)], g rouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)], and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 13.17 6.50 2.00 76.00 20.32 1992 6.17 5.50 1.00 12.00 3.64 1993 5.08 4.00 3.00 9.00 2.11 1994 6.75 6.00 4.00 16.00 3.44 1995 4.83 4.50 3.00 8.00 1.40 1996 5.25 5.00 3.00 10.00 2.30 1997 7.50 7.00 3.00 12.00 2.58 1998 7.58 6.00 3.00 26.00 6.19 1999 5.17 4.00 2.00 12.00 3.07 2000 5.42 4.50 2.00 13.00 3.26 2001 5.50 5.50 2.00 8.00 2.02 2002 5.08 4.00 2.00 12.00 3.00 2003 6.14 4.44 1.75 20.00 4.80 2004 6.06 4.15 1.70 17.70 4.37 2005 3.73 3.70 2.70 5.20 0.67 2006 3.48 3.15 2.20 6.20 1.13 Grouped by Season Season Mean Median Range Standard Deviation Wet 7.06 5.10 1.00 76.00 9.44 Dry 5.56 5.00 1.70 26.00 3.38 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 7.04 5.50 1.00 76.00 7.98 1999-2006 5.07 4.00 1.70 20.00 3.13 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 6.06 5.00 1.00 76.00 6.13

PAGE 74

59 Table 4-B. Descriptive statistics of turbidity (NT U) for station 111. Separated by year (n=12), season [wet (n=64) vs. dry (n=128)], g rouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)], and by the entire study period [1991-2006 (n=192)]. Group by Year Year Mean Median Range Standard Deviation 1991 10.42 8.50 3.00 28.00 7.17 1992 8.50 6.50 3.00 23.00 5.65 1993 6.08 5.00 3.00 13.00 3.00 1994 9.58 7.00 4.00 21.00 5.99 1995 7.08 5.00 3.00 22.00 5.38 1996 5.08 4.50 2.00 9.00 2.31 1997 8.08 5.75 2.00 29.00 7.03 1998 6.08 5.00 4.00 12.00 2.64 1999 5.58 3.50 2.00 17.00 4.60 2000 6.00 6.00 2.00 14.00 3.57 2001 6.58 7.00 4.00 8.00 1.38 2002 14.25 9.50 3.00 56.00 14.40 2003 5.50 5.35 2.56 9.00 2.08 2004 5.20 4.25 0.90 11.10 3.24 2005 5.38 3.95 1.10 21.60 5.62 2006 4.64 2.55 0.80 20.20 5.68 Group by Season Season Mean Median Range Standard Deviation Wet 8.36 7.00 2.00 29.00 5.57 Dry 6.51 5.00 0.80 56.00 6.28 Group by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 7.61 6.00 2.00 29.00 5.33 1999-2006 6.64 5.45 0.80 56.00 6.78 Group by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 7.13 6.00 0.80 56.00 6.10

PAGE 75

60 Table 4-C. Descriptive statistics of turbidity (NT U) for station 114. Separated by year (n=12), season [wet (n=64) vs. dry (n=128)], g rouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)], and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 7.75 3.50 1.00 23.00 8.05 1992 7.58 4.00 1.00 41.00 11.56 1993 5.83 4.50 1.00 16.00 5.13 1994 5.33 5.00 1.00 12.00 4.03 1995 5.75 4.50 2.00 17.00 4.79 1996 3.67 3.00 1.00 11.00 2.67 1997 3.67 4.00 1.00 9.00 2.46 1998 5.50 3.50 2.00 15.00 4.40 1999 4.00 2.50 1.00 10.00 3.41 2000 2.50 1.50 1.00 9.00 2.39 2001 3.83 3.00 1.00 14.00 3.71 2002 5.67 4.50 2.00 14.00 3.17 2003 6.86 5.00 1.55 18.00 4.72 2004 7.40 3.80 1.10 27.20 7.82 2005 5.41 4.10 2.10 16.00 4.17 2006 5.73 2.40 0.90 26.00 8.12 Grouped by Season Season Mean Median Range Standard Deviation Wet 7.60 5.35 1.00 26.00 5.67 Dry 4.31 3.00 0.90 41.00 5.24 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 5.64 4.00 1.00 41.00 6.03 1999-2006 5.17 3.80 0.90 27.20 5.14 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 5.40 4.00 0.90 41.00 5.59

PAGE 76

61 Table 4-D. Descriptive statistics of turbidity (NT U) for station 115. Separated by year (n=12), season [wet (n=64) vs. dry (n=128)], g rouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)], and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 12.33 9.00 1.00 26.00 8.92 1992 11.00 6.00 3.00 39.00 10.74 1993 8.00 7.50 2.00 16.00 4.16 1994 7.58 8.50 1.00 15.00 4.36 1995 7.17 6.50 4.00 11.00 2.72 1996 6.92 5.50 2.00 17.00 4.27 1997 6.67 7.00 1 .0024.00 6.04 1998 9.00 9.00 3.00 16.00 4.20 1999 6.17 4.50 2.00 15.00 4.28 2000 3.17 2.00 1.00 8.00 2.25 2001 5.25 3.50 1.00 18.00 5.59 2002 8.25 6.50 2.00 23.00 6.20 2003 8.75 7.12 3.99 15.00 4.11 2004 8.17 6.20 2.60 19.80 5.67 2005 5.63 5.80 3.50 7.70 1.16 2006 6.01 3.60 1.60 24.50 6.73 Grouped by Season Season Mean Median Range Standard Deviation Wet 10.45 9.00 2.00 26.00 5.80 Dry 6.03 5.00 1.00 39.00 5.19 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 8.58 7.00 1.00 39.00 6.31 1999-2006 6.42 5.05 1.0024.50 5.00 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 7.50 6.00 1.00 39.00 5.78

PAGE 77

62 Table 4-E. Descriptive statistics of turbidity (NT U) for station 116. Separated by year (n=12), season [wet (n=64) vs. dry (n=128)], g rouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)], and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 7.67 7.50 3.00 17.00 4.12 1992 7.67 6.00 3.00 21.00 5.84 1993 5.50 4.50 2.00 10.00 2.58 1994 4.67 4.00 2.00 10.00 2.57 1995 4.58 3.50 2.00 9.00 2.57 1996 3.75 3.00 1.00 8.00 1.86 1997 2.92 3.00 1.00 5.00 1.08 1998 3.92 3.50 1.00 13.00 3.12 1999 3.75 2.50 1.00 21.00 5.50 2000 2.42 2.00 1.00 8.00 1.93 2001 2.25 2.00 1.00 5.00 1.29 2002 3.83 3.00 1.00 9.00 2.25 2003 3.25 3.43 1.74 5.00 1.20 2004 2.86 2.55 1.30 5.04 1.36 2005 2.29 2.30 1.70 2.90 0.40 2006 3.69 2.00 1.00 14.30 4.09 Grouped by Season Season Mean Median Range Standard Deviation Wet 5.14 4.00 1.00 21.00 3.62 Dry 3.52 3.00 1.00 21.00 3.04 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 5.08 4.00 1.00 21.00 3.55 1999-2006 3.04 2.25 1.00 21.00 2.73 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 4.06 3.00 1.00 21.00 3.32

PAGE 78

63 Table 4-F. Descriptive statistics of turbidity (NT U) for station 139. Separated by year (n=12), season [wet (n=64) vs. dry (n=128)], g rouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)], and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 6.42 5.50 4.00 12.00 2.39 1992 6.58 6.50 3.00 12.00 2.75 1993 4.58 4.00 2.00 13.00 2.81 1994 4.50 4.00 2.00 9.00 2.15 1995 4.67 5.00 2.00 10.00 2.06 1996 4.83 5.00 2.00 10.00 1.99 1997 3.83 3.50 2.00 7.00 1.53 1998 4.25 4.00 2.00 9.00 2.01 1999 4.33 3.00 1.00 13.00 4.05 2000 1.50 1.00 1.00 4.00 0.90 2001 2.56 2.38 1.00 5.00 1.37 2002 2.67 3.00 1.00 4.00 0.89 2003 3.88 3.63 2.00 6.00 1.34 2004 3.93 3.70 2.40 5.70 1.14 2005 3.29 3.30 1.70 4.50 0.82 2006 2.86 2.70 1.80 4.00 0.73 Grouped by Season Season Mean Median Range Standard Deviation Wet 4.70 4.00 1.00 13.00 2.86 Dry 3.72 3.45 1.00 13.00 1.92 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 4.96 5.00 2.00 13.00 2.36 1999-2006 3.13 3.00 1.00 13.00 1.89 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 4.04 3.85 1.00 13.00 2.32

PAGE 79

64 Table 4-G. Descriptive statistics of turbidity (NT U) for station 151. Separated by year (n=12), season [wet (n=60) vs. dry (n=120)], g rouped by land use information [1992-1998 (n=84) and 1999-2006 (n=96)], and by the entire study period [1992-2006 (n=180)]. Grouped by Year Year Mean Median Range Standard Deviation 1992 7.08 6.50 1.00 24.00 5.66 1993 3.92 3.50 2.00 10.00 2.11 1994 6.58 5.50 2.00 15.00 4.19 1995 4.58 4.00 2.00 11.00 2.78 1996 3.00 3.00 1.00 4.00 0.95 1997 3.25 2.00 1.00 7.00 2.42 1998 4.75 4.00 3.00 9.00 1.96 1999 3.25 3.00 1.00 5.00 1.29 2000 4.00 3.00 1.00 11.00 2.98 2001 5.38 5.00 2.00 11.00 2.44 2002 5.08 4.50 2.00 12.00 2.78 2003 5.25 4.64 2.75 12.00 2.67 2004 4.82 3.80 2.10 9.86 2.71 2005 4.04 3.10 2.30 6.80 1.73 2006 5.25 2.40 1.10 31.40 8.43 Grouped by Season Season Mean Median Range Standard Deviation Wet 5.86 4.24 1.00 31.40 4.48 Dry 4.09 3.10 1.00 24.00 2.83 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1992-1998 4.74 4.00 1.00 24.00 3.43 1999-2006 4.63 4.00 1.00 31.40 3.69 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1992-2006 4.68 4.00 1.00 31.40 3.56

PAGE 80

65 Table 4-H. Descriptive statistics of turbidity (NT U) for station 154. Separated by year (n=12), season [wet (n=28) vs. dry (n=56)], an d by the entire study period [2000-2006 (n=84)]. Grouped by Year Year Mean Median Range Standard Deviation 2000 4.17 3.50 1.00 10.00 2.89 2001 4.42 3.00 1.00 12.00 3.87 2002 4.58 4.50 2.00 8.00 1.62 2003 4.40 3.43 1.30 14.00 3.33 2004 5.37 3.95 1.30 18.00 4.74 2005 5.31 3.90 1.50 22.10 5.49 2006 5.83 3.45 1.30 26.40 7.33 Grouped by Season Season Mean Median Range Standard Deviation Wet 7.43 5.45 2.00 26.40 5.30 Dry 3.59 3.00 1.00 22.10 3.23 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 2000-2006 4.87 3.83 1.00 26.40 4.40 When comparing the monthly median turbidity concen tration for each station (Figure 8), there is evidence of seasonal fluctuati ons in median concentration with several of the station’s trends peaking during the wet seas on, around July and August. This indicates that water quality with respect to turbid ity is degraded during the wet season. The increased levels during the wet season are like ly due to increased stormwater runoff and stream flow which causes more particles to be s uspended in the water column. The most dramatic increase in median concentration duri ng the wet season was noted for stations 114 and 115. The highest overall monthly median concentration was for station 115 with a turbidity concentration of 12.7 NTU duri ng the month of July (Figure 8).

PAGE 81

66 0 5 10 15 20 25 1234567891011 MonthTurbidity (NTU) Station 74 Station 111 Station 114 Station 115 Station 116 Station 139 Station 151 Station 154 Figure 8. Each sample station’s monthly turbidity median. For all 12 months over the entire study period [1991 – 2006, except Statio n 151 (1992-2006) and Station 154 (2000-2006)] for the water quality parameter turbid ity. Figures 9-A through 9-H show that the annual media n turbidity trend for the wet season was typically higher than the dry season and also exhibited greater variability in median concentrations. It is also interesting to n ote that the trend in median concentration of turbidity was often in the opposit e direction between the wet and the dry seasons. For instance, at station 115 (Figure 9-D) the trend in the annual median concentration in 1994 decreased in the dry season w hile it increased in the wet season; the opposite happened at this station between 1996 and 1998. The cause of these opposite seasonal trends is unknown.

PAGE 82

67 0 5 10 15 20 25 19911993199519971999200120032005 YearTurbidity (NTU) Dry Season Wet Season Figure 9-A. Dry versus wet annual seasonal median trends of turbidity (NTU) for station 74. 0 5 10 15 20 25 19911993199519971999200120032005 YearTurbidity (NTU) Dry Season Wet Season Figure 9-B. Dry versus wet annual seasonal median trends of turbidity (NTU) for station 111.

PAGE 83

68 0 5 10 15 20 25 19911993199519971999200120032005 YearTurbidity (NTU) Dry Season Wet Season Figure 9-C. Dry versus wet annual seasonal median trends of turbidity (NTU) for station 114. 0 5 10 15 20 25 19911993199519971999200120032005 YearTurbidity (NTU) Dry Season Wet Season Figure 9-D. Dry versus wet annual seasonal median trends of turbidity (NTU) for station 115.

PAGE 84

69 0 5 10 15 20 25 19911993199519971999200120032005 YearTurbidity (NTU) Dry Season Wet Season Figure 9-E. Dry versus wet annual seasonal median trends of turbidity (NTU) for station 116. 0 5 10 15 20 25 19911993199519971999200120032005 YearTurbidity (NTU) Dry Season Wet Season Figure 9-F. Dry versus wet annual seasonal median trends of turbidity (NTU) for station 139.

PAGE 85

70 0 5 10 15 20 25 19921994199619982000200220042006 YearTurbidity (NTU) Dry Season Wet Season Figure 9-G. Dry versus wet annual seasonal median trends of turbidity (NTU) for station 151. 0 5 10 15 20 25 2000200120022003200420052006 YearTurbidity (NTU) Dry Season Wet Season Figure 9-H. Dry versus wet annual seasonal median trends of turbidity (NTU) for station 154. The results of the statistical analysis of turbidit y from each sample station revealed half of the sample stations exhibited a st atistically significant trend over the entire study period. Sample stations 74, 111, 116, and 139 had overall statistically significant decreasing trends with the greatest mag nitude of the estimated slope occurring at station 116 (estimated slope of -0.2000 NTU/year ) (Table 5). This is the opposite

PAGE 86

71 result of what had been determined in a previous re port conducted between 1974 and 1995 which showed turbidity increasing at station 1 16 (SWFWMD, 2001). This is a good sign that turbidity values are beginning to tu rn around and that water quality for this variable is improving in the South Prong of the Ala fia River, in the upper portion of Turkey Creek, and at the mouth of the river. The r emaining stations (stations 114, 115, 151 and 154) showed no significant trend over the s tudy period. Table 5. Modified Seasonal Kendall trend test resu lts (significance level = 0.05) for turbidity by sample station. Station N Z-statistic modified Seasonal Kendall p-value Seasonal Kendall slope estimator Trend 74 192 -2.7886 0.0053 -0.1191 Decreasing 111 192 -2.2405 0.0251 -0.1667 Decreasing 114 192 -0.0753 0.9400 0.0000 No Trend 115 192 -1.5029 0.1329 -0.1429 No Trend 116 192 -3.2703 0.0011 -0.2000 Decreasing 139 192 -2.6882 0.0072 -0.1667 Decreasing 151 180 -0.8577 0.3911 0.0000 No Trend 154 84 0.8709 0.3838 0.0675 No Trend Fecal Coliform There did not appear to be any noticeable trends i n the annual and land use grouped medians for fecal coliform for any sample s tation based on the descriptive statistics (Tables 6-A through 6-H). The highest m edian concentration of fecal coliform for the entire study period was at sample station 1 11 with a concentration of 1050.00 cfu/100mL (Table 6-B). The overall mean concentrat ion for this station (2084.38 cfu/100mL) was much higher than the monthly mean st ate standard of 200 cfu/100mL and even the limit of 800 cfu/100mL on any one day (Rule 62-302, F.A.C.) (Table 6-B).

PAGE 87

72 Very high fecal coliform sample results for this st ation have been noted previously in the studies conducted by Parsons Engineering Science, I nc. (2002) and the Southwest Florida Water Management District (2001). The greatest var iability from the mean concentration was also at station 111 with a mean concentration o f 2084.38 cfu/100mL 4219.39 (Table 6-B). This indicates that fecal coliform le vels at this station, on average, are not only higher at this station, but the concentrations also vary more from the mean than any other station, with concentrations reaching as high as 38,000 cfu/100mL (Table 6-B). Station 154 also had a high overall median fecal c oliform concentration (240.00 cfu/100mL) and an overall mean concentration above the monthly mean state standard of 200 cfu/100mL with a concentration of 641.55 cfu/10 0mL (Table 6-H). High fecal coliform concentrations were also noted for this si te in the studies by Parsons Engineering Science, Inc. (2002) and Southwest Flor ida Water Management District (2001). Although these concentrations were not as high as those observed at station 111, they are still above state standards and sources of the high bacteria should be further investigated.

PAGE 88

73 Table 6-A. Descriptive statistics of fecal colifor m (cfu/100mL) for station 74. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 186.67 145.00 10.00 590.00 155.52 1992 270.00 35.00 10.00 2000.00 571.81 1993 131.67 100.00 10.00 400.00 117.92 1994 302.75 200.00 10.00 1333.00 370.60 1995 100.00 80.00 10.00 480.00 127.14 1996 100.83 80.00 10.00 240.00 82.62 1997 29.17 10.00 10.00 200.00 54.52 1998 131.67 45.00 10.00 1130.00 315.42 1999 25.83 15.00 10.00 60.00 19.29 2000 44.17 15.00 10.00 290.00 79.48 2001 40.00 20.00 10.00 230.00 61.50 2002 195.00 20.00 5.00 2000.00 569.47 2003 182.92 120.00 10.00 1020.00 272.85 2004 238.33 60.00 10.00 2020.00 564.93 2005 60.83 50.00 10.00 240.00 63.60 2006 74.17 30.00 10.00 360.00 100.95 Grouped by Season Season Mean Median Range Standard Deviation Wet 142.31 60.00 5.00 1333.00 232.60 Dry 127.03 40.00 5.00 2020.00 320.76 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 156.594 70 10.00 2000.00 283.092 1999-2006 107.656 35 5.00 2020.00 303.598 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 132.13 42.50 5.00 2020.00 293.78

PAGE 89

74 Table 6-B. Descriptive statistics of fecal colifor m (cfu/100mL) for station 111. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Group by Year Year Mean Median Range Standard Deviation 1991 6891.67 2350.00 1000.00 30000.00 9174.81 1992 4608.33 1000.00 100.00 38000.00 10601.07 1993 608.33 300.00 100.00 1600.00 519.54 1994 1158.33 900.00 100.00 3000.00 874.34 1995 1450.00 1150.00 200.00 4100.00 1076.61 1996 1033.33 900.00 200.00 3500.00 962.32 1997 2200.00 1050.00 400.00 14500.00 3891.72 1998 1508.33 1200.00 400.00 5800.00 1471.21 1999 775.00 700.00 400.00 1200.00 280.02 2000 1445.83 1387.50 100 .003700.00 1006.62 2001 3181.25 1750.00 100.00 12500.00 3547.17 2002 1581.25 1400.00 500.00 3900.00 958.83 2003 3075.00 1350.00 500.00 20000.00 5403.72 2004 1366.67 950.00 600.00 3400.00 841.36 2005 758.33 750.00 200.00 1600.00 442.02 2006 1708.33 1650.00 100.00 5500.00 1529.98 Group by Season Season Mean Median Range Standard Deviation Wet 2418.36 1100.00 100.00 30000.00 4645.69 Dry 1917.38 1000.00 100.00 38000.00 3998.05 Group by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 2432.29 1000.00 100.00 38000.00 5417.02 1999-2006 1736.46 1100.00 100.00 20000.00 2491.02 Group by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 2084.38 1050.00 100.00 38000.00 4219.39

PAGE 90

75 Table 6-C. Descriptive statistics of fecal colifor m (cfu/100mL) for station 114. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 1250.00 200.00 100.00 11500.00 3255.35 1992 1341.67 150.00 100.00 12300.00 3484.63 1993 115.00 100.00 60.00 240.00 49.08 1994 115.00 100.00 20.00 300.00 92.29 1995 86.67 60.00 20.00 220.00 70.50 1996 58.33 50.00 20.00 140.00 42.18 1997 53.33 60.00 20.00 100.00 28.71 1998 183.33 80.00 20.00 1140.00 310.17 1999 45.00 40.00 20.00 80.00 22.76 2000 230.00 50.00 20.00 1840.00 513.84 2001 75.00 80.00 20.00 180.00 45.23 2002 427.50 125.00 20.00 2340.00 689.60 2003 557.50 100.00 40.00 4000.00 1149.16 2004 1890.00 120.00 20.00 9300.00 3396.61 2005 201.67 120.00 40.00 860.00 234.40 2006 713.33 110.00 60.00 4000.00 1327.84 Grouped by Season Season Mean Median Range Standard Deviation Wet 725.94 140.00 20.00 11500.00 1934.03 Dry 325.47 80.00 20.00 12300.00 1361.47 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 400.42 100.00 20.00 12300.00 1708.33 1999-2006 517.50 80.00 20.00 9300.00 1450.93 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 458.96 100.00 20.00 12300.00 1581.80

PAGE 91

76 Table 6-D. Descriptive statistics of fecal colifor m (cfu/100mL) for station 115. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 958.33 550.00 200.00 3200.00 869.13 1992 500.00 150.00 100.00 2800.00 781.61 1993 100.00 100.00 20.00 200.00 55.27 1994 115.00 40.00 20.00 420.00 146.50 1995 116.67 100.00 20.00 340.00 89.78 1996 73.33 20.00 20.00 300.00 84.14 1997 135.00 60.00 20.00 600.00 175.01 1998 198.33 140.00 80.00 540.00 150.99 1999 78.33 50.00 20.00 200.00 62.93 2000 65.00 40.00 20.00 200.00 56.65 2001 55.00 40.00 20.00 180.00 51.96 2002 292.50 155.00 40.00 1280.00 396.88 2003 469.17 105.00 40.00 4000.00 1118.81 2004 438.33 80.00 20.00 1600.00 663.19 2005 96.67 90.00 20.00 220.00 61.99 2006 635.00 70.00 20.00 4000.00 1284.00 Grouped by Season Season Mean Median Range Standard Deviation Wet 421.25 100.00 20.00 4000.00 876.07 Dry 195.00 100.00 20.00 2800.00 359.41 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 274.58 100.00 20.00 3200.00 502.97 1999-2006 266.25 60.00 20.004000.00 672.04 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 270.42 100.00 20.00 4000.00 592.02

PAGE 92

77 Table 6-E. Descriptive statistics of fecal colifor m (cfu/100mL) for station 116. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 1008.33 400.00 100.00 8200.00 2275.35 1992 416.67 250.00 100.00 1800.00 489.59 1993 86.67 100.00 20.00 140.00 35.51 1994 71.67 30.00 20.00 320.00 88.81 1995 181.67 100.00 20.00 620.00 185.95 1996 64.17 60.00 20.00 180.00 44.61 1997 101.67 70.00 20.00 300.00 98.15 1998 130.00 70.00 20.00 520.00 161.25 1999 81.67 80.00 20.00 140.00 34.60 2000 391.67 80.00 20.00 4000.00 1137.00 2001 51.67 50.00 20.00 120.00 31.29 2002 222.50 95.00 10.00 1140.00 312.62 2003 152.50 100.00 40.00 500.00 125.85 2004 555.00 70.00 20.00 3600.00 1108.58 2005 53.33 40.00 20.00 120.00 39.39 2006 738.33 70.00 20.00 4000.00 1440.57 Grouped by Season Season Mean Median Range Standard Deviation Wet 478.59 100.00 20.00 8200.00 1288.48 Dry 164.53 80.00 10.00 3600.00 375.72 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 257.60 100.00 20.00 8200.00 854.21 1999-2006 280.83 80.00 10.00 4000.00 777.20 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 269.22 80.00 10.00 8200.00 814.55

PAGE 93

78 Table 6-F. Descriptive statistics of fecal colifor m (cfu/100mL) for station 139. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 750.00 400.00 100.00 2600.00 830.66 1992 225.00 150.00 100.00 700.00 205.05 1993 123.33 100.00 80.00 300.00 63.15 1994 268.33 80.00 20.00 2340.00 655.30 1995 115.00 60.00 20.00 640.00 172.28 1996 111.67 90.00 20.00 300.00 90.03 1997 223.33 170.00 20.00 960.00 251.08 1998 108.33 90.00 20.00 280.00 78.84 1999 181.67 130.00 60.00 660.00 169.16 2000 326.67 200.00 40.00 1060.00 345.60 2001 248.33 190.00 20.00 700.00 203.51 2002 318.33 130.00 60.00 1170.00 344.85 2003 96.67 80.00 40.00 280.00 66.51 2004 171.67 100.00 20.00 1000.00 267.03 2005 136.67 140.00 20.00 280.00 88.15 2006 603.33 200.00 40.00 2400.00 789.15 Grouped by Season Season Mean Median Range Standard Deviation Wet 304.38 120.00 20.00 2600.00 559.84 Dry 223.59 120.00 20.00 2200.00 297.33 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 240.63 100.00 20.00 2600.00 433.96 1999-2006 260.42 130.00 20.00 2400.00 374.75 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 250.52 120.00 20.00 2600.00 404.50

PAGE 94

79 Table 6-G. Descriptive statistics of fecal colifor m (cfu/100mL) for station 151. Separated by year (n=12), season [wet (n=60) vs. dr y (n=120)], grouped by land use information [1992-1998 (n=84) and 1999-2006 (n=96)] and by the entire study period [1992-2006 (n=180)]. Grouped by Year Year Mean Median Range Standard Deviation 1992 2225.00 450.00 100.00 20000.00 5650.60 1993 165.00 100.00 20.00 500.00 135.95 1994 188.33 110.00 20.00 680.00 221.48 1995 93.33 40.00 20.00 580.00 158.94 1996 73.33 30.00 20.00 420.00 113.24 1997 130.00 20.00 20.00 940.00 261.95 1998 305.00 60.00 20.00 2700.00 757.47 1999 55.00 40.00 20.00 140.00 41.89 2000 557.50 70.00 20.00 4000.00 1146.87 2001 300.00 190.00 20.00 1440.00 399.43 2002 435.00 120.00 20.00 3080.00 883.58 2003 673.33 195.00 40.00 4000.00 1137.21 2004 838.33 190.00 40.00 5800.00 1613.14 2005 250.00 240.00 20.00 680.00 219.01 2006 498.33 80.00 20.00 4400.00 1238.08 Grouped by Season Season Mean Median Range Standard Deviation Wet 391.25 140.00 20.00 4400.00 780.94 Dry 483.13 70.00 20.00 20000.00 1961.21 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1992-1998 454.29 90.00 20.00 20000.00 2205.59 1999-2006 450.94 120.00 20.00 5800.00 974.78 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1992-2006 452.50 100.00 20.00 20000.00 1661.31

PAGE 95

80 Table 6-H. Descriptive statistics of fecal colifor m (cfu/100mL) for station 154. Separated by year (n=12), season [wet (n=28) vs. dr y (n=56)], and by the entire study period [2000-2006 (n=84)]. Grouped by Year Year Mean Median Range Standard Deviation 2000 580.00 300.00 100.00 4000.00 1082.42 2001 356.67 170.00 20.00 920.00 341.53 2002 373.33 355.00 20.00 740.00 235.77 2003 596.67 210.00 60.00 4000.00 1095.53 2004 1389.17 210.00 60.00 10200.00 2890.13 2005 505.00 160.00 40.00 4000.00 1114.76 2006 690.00 300.00 100.00 4000.00 1163.02 Grouped by Season Season Mean Median Range Standard Deviation Wet 1164.64 230.00 40.00 10200.00 2184.72 Dry 380.00 240.00 20.00 4000.00 554.60 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 2000-2006 641.55 240.00 20.00 10200.00 1376.57 Review of the monthly median fecal coliform concen trations over the entire study period revealed that there does not appear to be se asonal variability in concentrations over the study period (Figure 10). However, there was an obvious difference in monthly median fecal coliform concentration between sample station 111 and the rest of the stations. The lowest monthly median concentration for station 111 (800 cfu/100mL) was still 440 cfu/100mL higher than the next largest va lue at station 154 and was much higher than the monthly mean state standard of 200 cfu/100 mL (Rule 62-302, F.A.C.) (Figure 10). The highest overall monthly median concentrat ion was at station 111 during the month of December; however, this trend was not pres ent at the other stations.

PAGE 96

81 0 500 1000 1500 2000 2500 1234567891011 MonthFecal Coliform (cfu/100mL) Station 74 Station 111 Station 114 Station 115 Station 116 Station 139 Station 151 Station 154 Figure 10. Each sample station’s monthly fecal col iform median. For all 12 months over the entire study period [1991 – 2006, except S tation 151 (1992 – 2006) and Station 154 (2000 – 2006)] for the water quality pa rameter fecal coliform. For the annual seasonal median trends in fecal col iform (Figures 11-A through 11-H), there was not a distinguishable difference i n the overall trends between the wet and the dry seasons. However, every sample station did show substantial peaks in annual seasonal median concentration during the wet season with most of the peaks exhibited in the very beginning and at the end of the study peri od. This indicates that the river is likely being affected by high surges of fecal colif orm runoff from the land during the wet season for those years. Sources of fecal coliform are animal wastes on agricultural lands and pet and human wastes (from septic systems) from residential lands which can be carried to the river during storm events. While most stations had seasonal median concentrat ions around 500 cfu/100mL or below, except for the peaks in the wet season, s tation 111 had much greater variability and overall substantially higher concentrations. T he highest annual seasonal median concentration for station 111 was in the 1991 wet s eason with a value off the chart at

PAGE 97

82 14,000 cfu/100mL (Figure 11-B). As indicated befor e, fecal coliform has historically been a problem at this station and specific sources of bacteria in the contributing zone of this station should be further investigated, especi ally during the wet season. 0 500 1000 1500 2000 2500 19911993199519971999200120032005 YearFecal Coliform (cfu/100mL) Dry Season Wet Season Figure 11-A. Dry versus wet annual seasonal median trends of fecal coliform (cfu/100mL) for station 74. 0 500 1000 1500 2000 2500 19911993199519971999200120032005 YearFecal Coliform (cfu/100mL) Dry Season Wet Season Figure 11-B. Dry versus wet annual seasonal median trends of fecal coliform (cfu/100mL) for station 111. The wet season median value outside the graph range for 1991 was 14,000 cfu/100mL.

PAGE 98

83 0 500 1000 1500 2000 2500 19911993199519971999200120032005 YearFecal Coliform (cfu/100mL) Dry Season Wet Season Figure 11-C. Dry versus wet annual seasonal median trends of fecal coliform (cfu/100mL) for station 114. 0 500 1000 1500 2000 2500 19911993199519971999200120032005 YearFecal Coliform (cfu/100mL) Dry Season Wet Season Figure 11-D. Dry versus wet annual seasonal median trends of fecal coliform (cfu/100mL) for station 115.

PAGE 99

84 0 500 1000 1500 2000 2500 19911993199519971999200120032005 YearFecal Coliform (cfu/100mL) Dry Season Wet Season Figure 11-E. Dry versus wet annual seasonal median trends of fecal coliform (cfu/100mL) for station 116. 0 500 1000 1500 2000 2500 19911993199519971999200120032005 YearFecal Coliform (cfu/100mL) Dry Season Wet Season Figure 11-F. Dry versus wet annual seasonal median trends of fecal coliform (cfu/100mL) for station 139.

PAGE 100

85 0 500 1000 1500 2000 2500 19921994199619982000200220042006 YearFecal Coliform (cfu/100mL) Dry Season Wet Season Figure 11-G. Dry versus wet annual seasonal median trends of fecal coliform (cfu/100mL) for station 151. 0 500 1000 1500 2000 2500 2000200120022003200420052006 YearFecal Coliform (cfu/100mL) Dry Season Wet Season Figure 11-H. Dry versus wet annual seasonal median trends of fecal coliform (cfu/100mL) for station 154. Based on the statistical analysis of fecal colifor m from each sample station, it was determined that no significant trends were present over the study period (Table 7). Although there were no statistically significant tr ends detected in the datasets over the entire study period, fecal coliform has exhibited t rends as noted earlier with large spikes in concentration during the wet season, particularl y during the beginning and end of the

PAGE 101

86 study period, and a much higher monthly median tren d at station 111 than at the other stations (Figures 10 and 11-A through 11-H). It is likely that the spikes in levels during the wet season are due to greater amounts of fecal coliform polluted stormwater runoff during certain years when compared to others. Also the historically high fecal coliform levels at station 111 should be further investigate d to determine the specific source of bacteria contributing to this station. Table 7. Modified Seasonal Kendall trend test resu lts (significance level = 0.05) for fecal coliform by sample station. Station N Z-statistic modified Seasonal Kendall p-value Seasonal Kendall slope estimator Trend 74 192 -1.5325 0.1254 -1.2917 No Trend 111 192 0.2085 0.8349 0.0000 No Trend 114 192 0.0799 0.9364 0.0000 No Trend 115 192 -1.2798 0.2006 -3.3333 No Trend 116 192 -1.2207 0.2222 -2.5000 No Trend 139 192 -0.2277 0.8199 0.0000 No Trend 151 180 0.7574 0.4488 0.0000 No Trend 154 84 -1.1597 0.2462 -15.8333 No Trend Total Phosphorus State standards for nutrients, including total pho sphorus, indicate that the discharge of total phosphorus should be limited to prevent violations of other water quality standards and should not alter the natural population of aquatic flora and fauna (Rule 62-302, F.A.C.). Review of the descriptive s tatistics showed that the total phosphorus median concentrations were higher during the first half of the study period for every sample site except 151 (which exhibited l ittle variability between the land use groupings), when comparing the values grouped by la nd use years (Tables 8-A through 8-

PAGE 102

87 H). This preliminary analysis indicated that total phosphorus improved at all stations over the second half of the study period when compa red to the first half, while station 151 only slightly degraded. Sample station 154 sho wed the greatest deviation from the mean with a mean concentration of 2.67 mg/L 2.35 (Table 8-H) which means there is a lot of variability in concentrations at this statio n, while station 115 had the highest concentration of a single sample at 24.86 mg/L when comparing ranges (Table 8-D).

PAGE 103

88 Table 8-A. Descriptive statistics of total phospho rus (mg/L) for station 74. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 1.09 0.95 0.44 2.22 0.52 1992 0.75 0.79 0.43 1.04 0.23 1993 1.02 0.89 0.42 2.24 0.47 1994 1.15 1.15 0.57 1.94 0.49 1995 1.10 1.04 0.37 1.84 0.45 1996 0.69 0.71 0.24 1.06 0.24 1997 0.54 0.49 0.31 1.33 0.28 1998 0.80 0.79 0.40 1.30 0.26 1999 0.50 0.45 0.26 1.14 0.24 2000 0.48 0.44 0.19 0.96 0.24 2001 0.57 0.32 0.10 2.03 0.57 2002 0.64 0.56 0.17 1.39 0.37 2003 0.91 0.85 0.38 1.45 0.37 2004 1.02 1.00 0.30 2.07 0.55 2005 0.62 0.54 0.30 1.24 0.27 2006 0.51 0.49 0.24 1.02 0.21 Grouped by Season Season Mean Median Range Standard Deviation Wet 0.96 0.75 0.21 2.24 0.54 Dry 0.68 0.61 0.10 1.76 0.33 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 0.89 0.80 0.24 2.24 0.43 1999-2006 0.66 0.51 0.10 2.07 0.41 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 0.77 0.68 0.10 2.24 0.43

PAGE 104

89 Table 8-B. Descriptive statistics of total phospho rus (mg/L) for station 111. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Group by Year Year Mean Median Range Standard Deviation 1991 1.08 1.11 0.43 1.89 0.42 1992 0.71 0.57 0.39 1.92 0.43 1993 0.65 0.63 0.40 1.25 0.22 1994 0.97 0.73 0.56 1.77 0.42 1995 0.89 0.78 0.61 1.89 0.36 1996 0.68 0.63 0.39 1.20 0.23 1997 0.87 0.77 0.61 1.90 0.36 1998 1.14 1.15 0.54 2.63 0.59 1999 0.58 0.57 0.37 1.09 0.20 2000 0.77 0.56 0.32 1.53 0.41 2001 0.75 0.73 0.52 1.16 0.18 2002 1.22 1.01 0.49 3.34 0.82 2003 0.71 0.61 0.44 1.09 0.25 2004 0.65 0.53 0.33 1.13 0.29 2005 0.90 1.00 0.30 1.30 0.37 2006 0.70 0.65 0.48 1.38 0.23 Group by Season Season Mean Median Range Standard Deviation Wet 1.02 0.99 0.50 2.63 0.38 Dry 0.73 0.62 0.30 3.34 0.41 Group by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 0.87 0.75 0.39 2.63 0.42 1999-2006 0.78 0.67 0.30 3.34 0.42 Group by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 0.83 0.74 0.30 3.34 0.42

PAGE 105

90 Table 8-C. Descriptive statistics of total phospho rus (mg/L) for station 114. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 1.95 1.87 1.24 2.81 0.56 1992 1.51 1.49 0.65 2.44 0.54 1993 1.66 1.64 1.07 2.50 0.43 1994 1.66 1.55 0.88 2.68 0.53 1995 2.53 1.51 1.10 10.49 2.68 1996 1.58 1.56 0.70 2.75 0.61 1997 1.59 1.65 1.02 2.00 0.28 1998 1.99 1.94 1.31 2.82 0.52 1999 1.43 1.38 0.85 1.98 0.34 2000 1.06 1.08 0.50 1.73 0.36 2001 1.33 1.02 0.62 2.49 0.66 2002 1.65 1.47 1.06 2.95 0.59 2003 1.52 1.60 0.96 1.79 0.25 2004 1.41 1.54 0.62 2.11 0.45 2005 1.41 1.37 0.88 1.95 0.33 2006 1.28 1.08 0.44 3.06 0.71 Grouped by Season Season Mean Median Range Standard Deviation Wet 2.03 1.80 0.50 10.49 1.26 Dry 1.38 1.36 0.44 2.82 0.41 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 1.81 1.60 0.65 10.49 1.07 1999-2006 1.39 1.35 0.44 3.06 0.50 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 1.60 1.49 0.44 10.49 0.86

PAGE 106

91 Table 8-D. Descriptive statistics of total phospho rus (mg/L) for station 115. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 4.89 4.75 3.44 6.81 1.02 1992 3.71 3.51 2.41 4.99 0.81 1993 3.71 3.68 2.27 4.76 0.80 1994 3.63 3.13 2.62 6.20 1.20 1995 5.78 3.82 2.41 24.86 6.12 1996 3.85 3.66 2.60 5.45 0.91 1997 3.56 3.29 2.24 5.45 0.81 1998 4.90 4.86 3.47 6.14 0.90 1999 3.34 3.42 2.19 4.67 0.86 2000 2.90 2.64 2.05 6.11 1.11 2001 3.22 3.28 0.67 5.38 1.27 2002 3.41 3.05 1.78 5.48 1.13 2003 3.08 2.97 2.35 4.38 0.63 2004 2.84 2.79 1.84 4.07 0.72 2005 2.67 2.85 0.66 3.67 0.85 2006 1.68 1.55 0.96 2.41 0.52 Grouped by Season Season Mean Median Range Standard Deviation Wet 4.03 3.52 0.96 24.86 2.94 Dry 3.35 3.32 0.66 6.81 1.13 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 4.25 3.79 2.24 24.86 2.38 1999-2006 2.89 2.83 0.66 6.11 1.02 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 3.57 3.35 0.66 24.86 1.95

PAGE 107

92 Table 8-E. Descriptive statistics of total phospho rus (mg/L) for station 116. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 0.95 0.95 0.51 1.34 0.26 1992 0.86 0.74 0.52 1.45 0.32 1993 0.79 0.86 0.49 .99 0.16 1994 1.08 0.97 0.59 2.57 0.52 1995 0.93 0.99 0.41 1.34 0.31 1996 0.95 0.88 0.59 1.63 0.31 1997 0.77 0.72 0.60 1.03 0.14 1998 1.07 0.87 0.62 3.52 0.79 1999 0.66 0.66 0.48 .89 0.11 2000 0.70 0.64 0.52 1.22 0.20 2001 0.78 0.71 0.29 1.30 0.31 2002 0.86 0.86 0.40 1.52 0.31 2003 0.84 0.87 0.58 1.08 0.17 2004 1.15 0.90 0.54 4.09 0.95 2005 0.82 0.80 0.65 1.06 0.14 2006 0.66 0.60 0.34 1.02 0.19 Grouped by Season Season Mean Median Range Standard Deviation Wet 0.96 0.90 0.34 4.09 0.51 Dry 0.82 0.73 0.29 3.52 0.34 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 0.92 0.87 0.41 3.52 0.40 1999-2006 0.81 0.72 0.29 4.09 0.40 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 0.87 0.80 0.29 4.09 0.41

PAGE 108

93 Table 8-F. Descriptive statistics of total phospho rus (mg/L) for station 139. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 1.15 1.07 0.61 1.94 0.39 1992 1.13 0.88 0.43 2.49 0.64 1993 1.18 1.11 0.54 2.03 0.52 1994 1.50 1.60 1.00 2.15 0.35 1995 1.21 1.18 0.67 1.91 0.37 1996 1.31 1.22 0.86 2.41 0.45 1997 1.29 1.19 0.71 2.13 0.42 1998 1.01 0.99 0.76 1.30 0.14 1999 1.13 1.10 0.94 1.35 0.12 2000 0.98 0.93 0.69 1.23 0.18 2001 1.21 1.13 0.66 2.07 0.40 2002 1.11 1.21 0.66 1.39 0.25 2003 1.00 0.96 0.69 1.41 0.21 2004 1.23 1.25 0.75 1.47 0.20 2005 1.35 1.17 1.03 2.27 0.41 2006 0.84 0.86 0.34 1.28 0.31 Grouped by Season Season Mean Median Range Standard Deviation Wet 1.20 1.18 0.34 2.49 0.45 Dry 1.14 1.08 0.43 2.41 0.34 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 1.22 1.12 0.43 2.49 0.44 1999-2006 1.11 1.09 0.34 2.27 0.31 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 1.16 1.09 0.34 2.49 0.46

PAGE 109

94 Table 8-G. Descriptive statistics of total phospho rus (mg/L) for station 151. Separated by year (n=12), season [wet (n=60) vs. dr y (n=120)], grouped by land use information [1992-1998 (n=84) and 1999-2006 (n=96)] and by the entire study period [1992-2006 (n=180)]. Grouped by Year Year Mean Median Range Standard Deviation 1992 0.99 0.75 0.43 3.58 0.86 1993 0.71 0.63 0.45 1.45 0.28 1994 0.96 0.77 0.57 1.69 0.41 1995 0.77 0.78 0.45 1.54 0.29 1996 0.63 0.66 0.35 0.83 0.14 1997 0.72 0.79 0.46 1.03 0.19 1998 0.71 0.74 0.54 0.86 0.13 1999 0.84 0.60 0.43 3.87 0.96 2000 0.77 0.62 0.32 1.56 0.42 2001 0.91 0.93 0.43 1.46 0.32 2002 0.95 0.91 0.57 1.84 0.35 2003 0.78 0.70 0.35 1.22 0.27 2004 0.81 0.76 0.39 1.49 0.35 2005 0.85 0.93 0.35 1.38 0.34 2006 0.74 0.67 0.46 1.57 0.29 Grouped by Season Season Mean Median Range Standard Deviation Wet 1.01 0.97 0.60 1.69 0.27 Dry 0.71 0.62 0.32 3.87 0.46 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1992-1998 0.79 0.73 0.35 3.58 0.41 1999-2006 0.83 0.74 0.32 3.87 0.45 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1992-2006 0.81 0.73 0.32 3.87 0.43

PAGE 110

95 Table 8-H. Descriptive statistics of total phospho rus (mg/L) for station 154. Separated by year (n=12), season [wet (n=28) vs. dr y (n=56)], and by the entire study period [2000-2006 (n=84)]. Grouped by Year Year Mean Median Range Standard Deviation 2000 2.15 1.40 0.73 5.15 1.47 2001 2.16 1.34 1.01 5.87 1.68 2002 4.45 3.61 1.60 11.01 3.12 2003 4.32 3.49 0.63 14.04 3.54 2004 2.61 2.61 0.80 5.99 1.43 2005 1.78 1.42 0.70 5.33 1.27 2006 1.19 0.88 0.40 2.88 0.75 Grouped by Season Season Mean Median Range Standard Deviation Wet 3.67 3.25 0.93 11.01 2.20 Dry 2.16 1.53 0.40 14.04 2.28 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 2000-2006 2.67 1.84 0.40 14.04 2.35 Examination of the monthly median concentrations o f total phosphorus showed evidence of seasonal variability with the monthly t rends being the highest during the wet season (Figure 12). This seasonal variability in t otal phosphorus was also noted in the study by Fraser (1986) and Qian et al. (2007a; 2007 b) also observed higher total phosphorus concentrations in the wet season in thei r study of water quality in the Southern Indian River Lagoon, Florida. Higher tota l phosphorus loading during the wet season may be attributed to polluted stormwater run off. The greatest fluctuations in monthly median total p hosphorus concentration occurred at sample stations 114, 115, and 154 (Figu re 12), located in the main portion and the North Prong of the Alafia River (Figure 5). The highest monthly median

PAGE 111

96 concentration of total phosphorus was at station 15 4 with a concentration of 5.15 mg/L for the month of September. Two of the stations lo cated downstream of this station, stations 115 and 114, exhibited similar trends by h aving the next highest peaks in monthly median concentration. However, station 115 had overall higher monthly median total phosphorus concentrations than any other stat ion. These results indicate that the North Prong of the river and downstream into the ma in section of the river are more degraded for total phosphorus than the other portio ns of the river, especially during the wet season. 0 1 2 3 4 5 6 1234567891011 MonthTotal Phosphorus (mg/L) Station 74 Station 111 Station 114 Station 115 Station 116 Station 139 Station 151 Station 154 Figure 12. Each sample station’s monthly total pho sphorus median. For all 12 months over the entire study period [1991 – 2006, e xcept Station 151 (1992 – 2006) and Station 154 (2000 – 2006)] for the water qualit y parameter total phosphorus. As noted from the examination of the monthly media ns, total phosphorus annual seasonal median trends were typically higher during the wet season for every sample station (Figures 13-A through 13-H). As indicated above, previous studies by Frasier (1986) and Qian et al. (2007a; 2007b) have also not ed seasonal variability in total

PAGE 112

97 phosphorus levels. Sources of total phosphorus can be from excessive use of fertilizers, animal and human wastes [from septic systems and wa stewater residuals application (SWFWMD, 2001)], and phosphate mining. The total p hosphorus annual seasonal median concentrations were noticeably much higher a nd more variable at station 115 compared to the other stations; however, it appears the trends in concentration at this station for both seasons have decreased over the st udy period (Figure 13-D). A similar trend was evident for station 154 where the concent rations peaked around 2002 and 2003 for both seasons; and they began to decrease toward s the end of the study period in 2006 (Figure 13-H). This indicates that although total phosphorus was high at both of these stations, water quality for this parameter has been improving. 0 1 2 3 4 5 6 19911993199519971999200120032005 YearTotal Phosphorus (mg/L) Dry Season Wet Season Figure 13-A. Dry versus wet annual seasonal median trends of total phosphorus (mg/L) for station 74.

PAGE 113

98 0 1 2 3 4 5 6 19911993199519971999200120032005 YearTotal Phosphorus (mg/L) Dry Season Wet Season Figure 13-B. Dry versus wet annual seasonal median trends of total phosphorus (mg/L) for station 111. 0 1 2 3 4 5 6 19911993199519971999200120032005 YearTotal Phosphorus (mg/L) Dry Season Wet Season Figure 13-C. Dry versus wet annual seasonal median trends of total phosphorus (mg/L) for station 114.

PAGE 114

99 0 1 2 3 4 5 6 19911993199519971999200120032005 YearTotal Phosphorus (mg/L) Dry Season Wet Season Figure 13-D. Dry versus wet annual seasonal median trends of total phosphorus (mg/L) for station 115. 0 1 2 3 4 5 6 19911993199519971999200120032005 YearTotal Phosphorus (mg/L) Dry Season Wet Season Figure 13-E. Dry versus wet annual seasonal median trends of total phosphorus (mg/L) for station 116.

PAGE 115

100 0 1 2 3 4 5 6 19911993199519971999200120032005 YearTotal Phosphorus (mg/L) Dry Season Wet Season Figure 13-F. Dry versus wet annual seasonal median trends of total phosphorus (mg/L) for station 139. 0 1 2 3 4 5 6 19921994199619982000200220042006 YearTotal Phosphorus (mg/L) Dry Season Wet Season Figure 13-G. Dry versus wet annual seasonal median trends of total phosphorus (mg/L) for station 151.

PAGE 116

101 0 1 2 3 4 5 6 2000200120022003200420052006 YearTotal Phosphorus (mg/L) Dry Season Wet Season Figure 13-H. Dry versus wet annual seasonal median trends of total phosphorus (mg/L) for station 154. Total phosphorus was shown to have a statistically significant trend at three sample stations while the remaining five did not ha ve a significant trend over the study period. Sample stations 74, 114, and 115 had a dec reasing trend with station 115 having the greatest magnitude with an estimated slope of 0.1233 mg/year (Table 9). This provides evidence that water quality is improving a t these sample stations, especially station 115. This is a good sign since examination of the descriptive statistics and monthly median trends had indicated that this stati on had the highest single sample concentration and overall monthly median trend conc entrations, respectively, when compared to the other stations. The SWFWMD (2001) observed similar decreasing trends at stations 74, 114, and 115, but also obser ved decreasing trends at stations 116 and 139 based on water quality data from these stat ions between 1974 and 1995.

PAGE 117

102 Table 9. Modified Seasonal Kendall trend test resu lts (significance level = 0.05) for total phosphorus by sample station. Station N Z-statistic modified Seasonal Kendall p-value Seasonal Kendall slope estimator Trend 74 192 -1.9930 0.0463 -0.0238 Decreasing 111 192 -0.4252 0.6707 -0.0031 No Trend 114 192 -2.3605 0.0182 -0.0293 Decreasing 115 192 -3.2376 0.0012 -0.1233 Decreasing 116 192 -1.5497 0.1212 -0.0086 No Trend 139 192 -0.5602 0.5753 -0.0039 No Trend 151 180 0.2716 0.7860 0.0010 No Trend 154 84 -0.7979 0.4249 -0.1079 No Trend Total Nitrogen As with total phosphorus, the state standard for t otal nitrogen indicates that the discharge of total nitrogen should be limited to pr event violations of other water quality standards and should not alter the natural populati on of aquatic flora and fauna (Rule 62302, F.A.C.). For every sample station except stat ions 74 and 111, the median concentration of total nitrogen was higher during t he second half of the study period than the first half when grouped by land use years (Tabl es 10-A through 10-H). This implies that total nitrogen at most of the stations decline d during the second half of the study period when compared to the first half. Possible s ources of total nitrogen are from animal and human wastes (from septic systems), fertilizer in stormwater runoff, and atmospheric deposition. The highest median concentration of to tal nitrogen over the entire study period was at station 111 with a value of 3.24 mg/L (Table 10-B). The highest total nitrogen concentration from a single sample was at station 151 with a value of 7.50 mg/L (Table 10-G) with station 111 just below that with a concentration of 7.35 mg/L (Table 10-B).

PAGE 118

103 Table 10-A. Descriptive statistics of total nitrog en (mg/L) for station 74. Separated by year (n=12), season [wet (n=64) vs. dry (n=128)] grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Group by Year Year Mean Median Range Standard Deviation 1991 1.10 1.12 0.46 1.58 0.33 1992 1.11 1.03 0.83 1.65 0.24 1993 1.16 1.16 0.68 1.78 0.35 1994 1.61 1.64 0.35 2.29 0.48 1995 1.23 1.36 0.58 1.70 0.38 1996 1.44 1.16 0.74 3.23 0.83 1997 1.01 0.96 0.75 1.53 0.24 1998 1.10 1.08 0.77 1.43 0.26 1999 0.96 0.89 0.70 1.32 0.19 2000 1.20 1.20 0.35 2.19 0.46 2001 1.13 1.09 0.71 2.21 0.41 2002 1.36 1.20 0.76 3.02 0.66 2003 1.28 1.29 0.96 1.69 0.17 2004 1.48 1.31 0.51 4.03 0.95 2005 0.94 1.04 0.56 1.38 0.28 2006 0.82 0.86 0.11 1.20 0.30 Group by Season Season Mean Median Range Standard Deviation Wet 1.29 1.22 0.69 3.01 0.42 Dry 1.13 1.06 0.11 4.03 0.52 Group by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 1.22 1.16 0.35 3.23 0.46 1999-2006 1.15 1.14 0.11 4.03 0.52 Group by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 1.18 1.15 0.11 4.03 0.49

PAGE 119

104 Table 10-B. Descriptive statistics of total nitrog en (mg/L) for station 111. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Group by Year Year Mean Median Range Standard Deviation 1991 2.75 2.57 1.08 5.15 1.12 1992 2.86 2.66 2.02 4.05 0.64 1993 3.05 3.02 1.16 4.64 0.87 1994 3.13 3.18 2.33 4.00 0.45 1995 3.38 3.39 2.23 4.74 0.76 1996 3.85 3.95 0.74 6.25 1.32 1997 3.42 3.35 1.78 4.52 0.68 1998 3.83 4.03 2.49 5.31 1.05 1999 3.50 3.41 1.36 4.97 0.93 2000 3.72 3.53 1.20 7.35 1.61 2001 3.35 3.24 2.42 4.94 0.77 2002 3.41 3.61 2.20 4.20 0.61 2003 2.71 2.60 1.03 3.58 0.68 2004 2.90 2.75 2.07 4.18 0.67 2005 2.98 2.78 2.24 4.15 0.66 2006 3.70 3.49 2.15 5.65 1.23 Group by Season Season Mean Median Range Standard Deviation Wet 3.04 2.79 1.03 5.14 0.89 Dry 3.41 3.34 0.74 7.35 0.98 Group by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 3.28 3.33 0.74 6.25 0.95 1999-2006 3.28 3.16 1.03 7.35 0.98 Group by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 3.28 3.24 0.74 7.35 0.96

PAGE 120

105 Table 10-C. Descriptive statistics of total nitrog en (mg/L) for station 114. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 1.8942 1.89 1.49 2.63 0.31047 1992 1.795 1.72 1.38 2.35 0.30243 1993 1.3717 1.305 1.13 1.85 0.22695 1994 1.4008 1.355 1.15 2.09 0.2521 1995 1.4383 1.455 1.18 2.04 0.24071 1996 1.5925 1.545 1.24 2.26 0.30224 1997 1.8481 1.795 1.30 2.46 0.34531 1998 1.8642 1.88 1.28 2.37 0.33926 1999 1.6683 1.65 1.28 2.00 0.24094 2000 2.1117 2.125 1.45 2.56 0.3013 2001 1.9875 2.035 1.24 2.55 0.32017 2002 1.6683 1.472 1.35 2.41 0.38059 2003 1.8197 1.9156 1.03 2.54 0.46097 2004 2.165 1.7 1.30 5.01 1.05784 2005 1.9154 1.9135 1.43 2.50 0.35092 2006 2.0314 2.0795 1.00 2.80 0.46995 Grouped by Season Season Mean Median Range Standard Deviation Wet 1.6775 1.5665 1.00 2.63 0.38782 Dry 1.8399 1.84 1.06 5.01 0.49002 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 1.6506 1.5787 1.13 2.63 0.35233 1999-2006 1.9209 1.91 1.00 5.01 0.52108 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 1.7858 1.771 1.00 5.01 0.46385

PAGE 121

106 Table 10-D. Descriptive statistics of total nitrog en (mg/L) for station 115. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 1.48 1.49 0.87 2.24 0.38 1992 1.25 1.14 0.76 2.11 0.41 1993 1.23 1.19 0.98 1.76 0.23 1994 1.24 1.26 0.54 1.64 0.27 1995 1.47 1.38 1.10 3.07 0.52 1996 1.26 1.32 0.88 1.52 0.21 1997 1.60 1.59 1.21 2.10 0.31 1998 1.73 1.81 1.22 2.10 0.30 1999 1.87 1.46 1.15 5.31 1.19 2000 1.76 1.77 1.08 2.33 0.37 2001 1.57 1.41 0.90 3.00 0.58 2002 1.43 1.35 0.89 2.17 0.37 2003 1.53 1.51 1.03 1.72 0.20 2004 2.03 1.60 0.86 4.86 1.10 2005 2.07 1.97 1.37 3.23 0.58 2006 1.80 1.78 1.02 2.37 0.42 Grouped by Season Season Mean Median Range Standard Deviation Wet 1.46 1.39 0.89 2.75 0.35 Dry 1.64 1.47 0.54 5.31 0.67 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 1.41 1.34 0.54 3.07 0.37 1999-2006 1.76 1.61 0.86 5.31 0.70 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 1.58 1.47 0.54 5.31 0.59

PAGE 122

107 Table 10-E. Descriptive statistics of total nitrog en (mg/L) for station 116. Separated by year (n=12), season [wet (n=64) vs. dr y (n=128)], grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 1.25 1.23 0.90 1.70 0.21 1992 1.27 1.28 0.80 1.66 0.29 1993 0.97 0.99 0.60 1.34 0.24 1994 0.86 0.92 0.44 1.22 0.27 1995 0.85 0.85 0.54 1.28 0.22 1996 1.16 0.95 0.69 2.79 0.61 1997 1.16 1.18 0.85 1.48 0.21 1998 1.00 0.97 0.72 1.70 0.27 1999 1.07 1.00 0.87 1.39 0.17 2000 0.97 0.99 0.45 1.83 0.36 2001 1.07 1.06 0.51 2.20 0.45 2002 1.18 1.11 0.93 1.83 0.24 2003 1.25 1.09 0.86 1.71 0.35 2004 1.40 1.21 0.92 3.71 0.75 2005 1.09 1.06 0.98 1.25 0.09 2006 1.33 1.31 0.79 1.64 0.26 Grouped by Season Season Mean Median Range Standard Deviation Wet 1.22 1.16 0.67 2.20 0.30 Dry 1.07 1.03 0.44 3.71 0.39 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 1.06 1.00 0.44 2.79 0.34 1999-2006 1.17 1.10 0.45 3.71 0.39 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 1.12 1.08 0.44 3.71 0.37

PAGE 123

108 Table 10-F. Descriptive statistics of total nitrog en (mg/L) for station 139. Separated by year (n=12), season [wet (n=64) vs. dry (n=128)] grouped by land use information [1991-1998 (n=96) and 1999-2006 (n=96)] and by the entire study period [1991-2006 (n=192)]. Grouped by Year Year Mean Median Range Standard Deviation 1991 1.69 1.67 1.12 2.36 0.35 1992 1.82 1.74 1.38 2.61 0.34 1993 1.11 1.02 0.76 1.70 0.28 1994 1.17 1.29 0.59 1.83 0.41 1995 0.92 0.92 0.65 1.49 0.23 1996 0.91 0.89 0.44 1.30 0.25 1997 1.21 1.12 0.76 1.77 0.37 1998 1.09 0.89 0.62 3.29 0.72 1999 1.30 1.25 0.81 2.24 0.45 2000 1.49 1.49 1.06 2.05 0.29 2001 1.53 1.37 0.97 3.17 0.61 2002 1.50 1.36 0.90 2.37 0.46 2003 1.05 1.06 0.84 1.32 0.10 2004 1.33 1.35 0.90 1.67 0.21 2005 1.54 1.54 0.90 2.31 0.33 2006 1.50 1.53 0.87 2.27 0.42 Grouped by Season Season Mean Median Range Standard Deviation Wet 1.32 1.23 0.78 3.29 0.49 Dry 1.32 1.35 0.44 2.37 0.45 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1991-1998 1.24 1.08 0.44 3.29 0.50 1999-2006 1.41 1.37 0.81 3.17 0.41 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1991-2006 1.32 1.31 0.44 3.29 0.46

PAGE 124

109 Table 10-G. Descriptive statistics of total nitrog en (mg/L) for station 151. Separated by year (n=12), season [wet (n=60) vs. dr y (n=120)], grouped by land use information [1992-1998 (n=84) and 1999-2006 (n=96)] and by the entire study period [1992-2006 (n=180)]. Grouped by Year Year Mean Median Range Standard Deviation 1992 2.27 2.26 0.28 7.50 1.85 1993 1.68 1.82 0.54 2.86 0.70 1994 1.73 1.84 0.25 2.41 0.58 1995 1.77 1.82 0.75 2.63 0.50 1996 1.57 1.61 0.72 2.84 0.63 1997 1.23 0.67 0.18 2.76 1.02 1998 1.79 1.71 0.75 2.86 0.68 1999 1.93 1.93 0.32 3.87 1.04 2000 1.78 1.61 0.32 4.08 1.03 2001 1.84 1.80 0.30 3.18 0.81 2002 1.99 2.20 0.63 3.22 0.76 2003 1.69 1.74 1.03 2.04 0.25 2004 1.89 1.96 1.49 2.15 0.20 2005 2.11 1.95 1.50 3.20 0.57 2006 2.32 2.21 1.02 4.16 0.88 Grouped by Season Season Mean Median Range Standard Deviation Wet 1.68 1.18 0.40 3.14 0.63 Dry 1.92 1.81 0.18 7.50 0.95 Grouped by Land Use Years Land Use Years Mean Median Range Standard Deviation 1992-1998 1.72 1.76 0.18 7.50 0.97 1999-2006 1.95 1.86 0.30 4.16 0.75 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 1992-2006 1.84 1.81 0.18 7.50 0.86

PAGE 125

110 Table 10-H. Descriptive statistics of total nitrog en (mg/L) for station 154. Separated by year (n=12), season [wet (n=28) vs. dr y (n=56)], and by the entire study period [2000-2006 (n=84)]. Grouped by Year Year Mean Median Range Standard Deviation 2000 2.81 2.84 1.21 3.93 0.71 2001 2.37 2.35 0.79 4.22 0.91 2002 2.07 2.06 1.15 3.28 0.55 2003 2.47 2.55 1.03 3.52 0.58 2004 2.48 2.57 1.46 4.06 0.84 2005 2.89 2.86 1.60 3.79 0.64 2006 2.79 2.89 0.98 4.13 0.88 Grouped by Season Season Mean Median Range Standard Deviation Wet 2.16 2.01 0.98 3.79 0.67 Dry 2.75 2.76 0.79 4.22 0.74 Grouped by Entire Study Period Years Mean Median Range Standard Deviation 2000-2006 2.55 2.56 0.79 4.22 0.77 While total phosphorus exhibited seasonal fluctuat ions with the monthly trend peaking during the wet season, total nitrogen tende d to dip during the wet season (Figure 14). As shown in Figure 14, sample stations 111, 1 14,115, 139, and 154 all tended to dip to their lowest monthly median concentration around July or August. Upon comparison of the monthly trends between each of the stations, station 111 had the average highest monthly median total nitrogen concentrations each m onth with station 154 being the next highest. Both of these stations have shown high to tal nitrogen and high fecal coliform values during the study period. This provides evid ence that there are strong water pollution sources from animal and/or human waste in the contributing zones of these stations.

PAGE 126

111 0 1 2 3 4 5 1234567891011 MonthTotal Nitrogen (mg/L) Station 74 Station 111 Station 114 Station 115 Station 116 Station 139 Station 151 Station 154 Figure 14. Each sample station’s monthly total nit rogen median. For all 12 months over the entire study period [1991 – 2006, except S tation 151 (1992 – 2006) and Station 154 (2000 – 2006)] for the water quality pa rameter total nitrogen. The seasonal median trends for total nitrogen (Fig ures 15-A through 15-H) did not vary much between the wet and dry seasons for a ll of the sample stations except station 154, which had a definite separation in con centrations. While two stations, (stations 74 and 116) (Figures 15-A and 15-E), had higher annual seasonal median total nitrogen trends in the wet season, most of the stat ions (stations 111, 114, 115, 151, and 154) had higher trends during the dry season (Figur es 15-B through 15-D, 15-G, and 15H). Although it would be expected that total nitro gen levels would be higher during the wet season due to increased stormwater pollution, s imilar to total phosphorus; total nitrogen and nitrate levels have been shown to be n egatively correlated to water discharge in the study by Kebede et al. (2003) and Lehrter (2006), respectively, and both correlations were attributed to the effects of dilu tion. Sample station 139 had a higher trend in the wet season for the first half of the s tudy period and had a higher trend in the

PAGE 127

112 dry season for the second half of the study period, but overall did not have a distinguishable difference between the seasons (Fig ure 15-F). Station 111 overall had the highest annual seasonal median concentrations compa red to all other stations (Figure 15B). 0 1 2 3 4 5 19911993199519971999200120032005 YearTotal Nitrogen (mg/L) Dry Season Wet Season Figure 15-A. Dry versus wet annual seasonal median trends of total nitrogen (mg/L) for station 74. 0 1 2 3 4 5 19911993199519971999200120032005 YearTotal Nitrogen (mg/L) Dry Season Wet Season Figure 15-B. Dry versus wet annual seasonal median trends of total nitrogen (mg/L) for station 111.

PAGE 128

113 0 1 2 3 4 5 19911993199519971999200120032005 YearTotal Nitrogen (mg/L) Dry Season Wet Season Figure 15-C. Dry versus wet annual seasonal median trends of total nitrogen (mg/L) for station 114. 0 1 2 3 4 5 19911993199519971999200120032005 YearTotal Nitrogen (mg/L) Dry Season Wet Season Figure 15-D. Dry versus wet annual seasonal median trends of total nitrogen (mg/L) for station 115.

PAGE 129

114 0 1 2 3 4 5 19911993199519971999200120032005 YearTotal Nitrogen (mg/L) Dry Season Wet Season Figure 15-E. Dry versus wet annual seasonal median trends of total nitrogen (mg/L) for station 116. 0 1 2 3 4 5 19911993199519971999200120032005 YearTotal Nitrogen (mg/L) Dry Season Wet Season Figure 15-F. Dry versus wet annual seasonal median trends of total nitrogen (mg/L) for station 139.

PAGE 130

115 0 1 2 3 4 5 19921994199619982000200220042006 YearTotal Nitrogen (mg/L) Dry Season Wet Season Figure 15-G. Dry versus wet annual seasonal median trends of total nitrogen (mg/L) for station 151. 0 1 2 3 4 5 2000200120022003200420052006 YearTotal Nitrogen (mg/L) Dry Season Wet Season Figure 15-H. Dry versus wet annual seasonal median trends of total nitrogen (mg/L) for station 154. The statistical analysis of total nitrogen at each sample station revealed a statistically significant increasing trend at sampl e stations 114, 115, and 151 with the greatest magnitude of the slope estimated at statio n 115 (slope estimate of 0.0353 mg/year) (Table 11). This provides evidence that w ater quality is slowly degrading at these sample stations. Previous technical reports indicated that sample analysis from

PAGE 131

116 Lithia Springs, located upstream of station 114, ha ve shown excessive nitrate levels (Berg et al., 2003; FDEP, 1998) with evidence of a gradua l increase over time (Berge et al., 2003). Since total nitrogen concentration is deter mined by adding nitrate/nitrite levels to total kjeldahl nitrogen levels, then the excessive and rising nitrate levels discharging from Lithia Springs could be a strong source of the incr easing total nitrogen at this sample station. The increased total nitrogen concentratio ns at station 151 is likely due to the high levels coming from upstream at station 111. Since total nitrogen is increasing with the highest magnitude at station 115, possible sources of nitrogen contributing to this station should be further investigated. The remaining five stations did not have a significant trend over the study period. Table 11. Modified Seasonal Kendall trend test res ults (significance level = 0.05) for total nitrogen by sample station. Station N Z-statistics modified Seasonal Kendall p-value Seasonal Kendall slope estimator Trend 74 192 -1.4229 0.1548 -0.0115 No Trend 111 192 0.3879 0.6981 0.0072 No Trend 114 192 2.0803 0.0375 0.0256 Increasing 115 192 2.8990 0.0037 0.0353 Increasing 116 192 1.2194 0.2227 0.0115 No Trend 139 192 0.5313 0.5952 0.0079 No Trend 151 180 2.0028 0.0452 0.0252 Increasing 154 84 0.9144 0.3605 0.0510 No Trend Land Use Trends In Table A-1 and Figures 16-A through 16-H, it is evident that in 1990, the highest proportions of land use coverage in the con tributing zone were of wetlands, particularly at stations 74, 114, 115, 116, and 139 These sample stations are located

PAGE 132

117 along the main section of the Alafia River, at the split of the North and South Prongs, and further upstream in the South Prong. Wetlands prov ide benefits to water quality by slowing down the speed of stormwater runoff as well as uptake nutrients in the runoff. In 1990, the highest proportions of agricultural l and use were at station 111 with a proportion of 0.5167 [located in the upper portion of Turkey Creek (Figure 5)] and at station 154 with a proportion of 0.3991 [located in the northeastern portion of the watershed (Figure 5)] with the majority of it being crop and pastureland for both stations (Table A-1). Agricultural land use has previously been shown to be a large contributor to nutrient loads in the study area, especially phosph orus (Parsons Engineering Science, Inc., 2002). Both of these stations exhibited high er concentrations of nutrients with station 111 high in total nitrogen and station 154 high in both total nitrogen and total phosphorus (Figures 12 and 14, respectively). For station 151 in 1990, urban and builtup land usage was the highest proportion at 0.8798; however, most of this usage was reclaimed land from past mining operations and very little was from low density residential land use (Table A-1). In 1999, the predominant land use types began to s hift from wetlands to urban and built-up land use (Table A-1), particularly at stat ion 139 (Figure 16-F). When comparing the percent change in the proportion of land use fo r each station between 1990 and 1999, the proportions of wetlands have decreased for each station except station 115, which increased by 2.77% (Table A-2). The proportion of urban and built-up land use increased for every station between 1990 and 1999 with the hi ghest percent change occurring at station 154 at 83.18% (Table A-2). Residential lan d use has been shown to be positively correlated with fecal coliform concentrations (Sliv a and Williams, 2001) and a strong

PAGE 133

118 contributor of nitrate (Basnyat et al., 1999). Fec al coliform and total nitrogen were both shown to be high at station 154 (Figures 10 and 14, respectively) and may be the result of the increase in urban and built-up land use in the contributing zone of this station. While rangeland did not comprise a large proportion of land use coverage for any station in 1990 and 1999, there was a decrease in t his land use coverage for every station between 1990 and 1999, except station 151 which did not contain this land use type in either year (Tables A-1 and A-2 and Figures 16-A th rough 16-H). Rangeland comprises mainly shrub and brushland and dry prairies. This type of land use also benefits water quality by removing pollutants from stormwater runo ff and when decreased, it could negatively affect water quality. In 2006, it becomes even more evident that the lan d use coverage shifted more towards urban and built-up land use with stations 7 4, 139, 151, and 154 all having this type of land use as the highest proportion (Table A -1 and Figures 16-A, and 16-F through 16-H). From 1999 to 2006, the percent change in ur ban and built-up land use had again increased for every station with the greatest chang e at station 111 at 34.35% (Table A-2 and Figures 16-A through 16-H). This increase in t he proportion of urban and built-up land use at this station may be contributing to the increase in fecal coliform from the first half of the study period to the second half that wa s noted from the descriptive statistics for this station (Table 6-B). For every station, the amount of agricultural land use decreased, while wetlands appeared to recover a little between 1999 and 2006 (Table A-2 and Figures 16-A through 16-H). The highest percent change in the proportio n of land use between 1999 and 2006 was for barren land use at station 114 which had a 3098445.08% increase (Table A-2).

PAGE 134

119 The proportion of barren land use changed for this station from 0.000007 to 0.0097 between 1999 and 2006 (Table A-1 and Figure 16-C). Although the percent change in the proportion of this land use type at station 114 increased substantially, it still only comprised a small portion of the contributing zone of this station and likely would not have a great effect on water quality. Overall, between 1990 and 2006, there was a decrea se in the proportion of agricultural land use for every station, with the h ighest reduction noted at station 151 (in the lower portion of Turkey Creek) at 47.78% (Table A-2). During this time period, there was also a reduction in rangeland at every station except station 151 which did not contain any rangeland (Table A-2 and Figures 16-A t hrough 16-H). There was also a decrease in wetland land use for every station exce pt 115 which was shown to have an overall 0.19% increase in proportion (Table A-2 and Figures 16-A through 16-H). This increase in the proportion of wetlands may be contr ibuting to the decreasing trend in total phosphorus at this station (Table 9). Urban and bu ilt-up land use over the entire study period has been shown to have increased at every st ation in the Alafia River, with the highest change in proportion at station 154 with an increase of 140.63% (Table A-2 and Figures 16-A through 16-H).

PAGE 135

120 0.0000 0.2000 0.4000 0.6000 0.8000 1.0000 199019992006 YearProportion Agricultural Barren Rangeland Transport, Comm, & Util Upland Forest Urban & Built-Up Water Wetlands Figure 16-A. Changes in the proportions of land us e at station 74. The proportion is the part of the entire contributing zone that is represented by each land use type. The years are each of the land use years used in th is study. 0.0000 0.2000 0.4000 0.6000 0.8000 1.0000 199019992006 YearProportion Agricultural Barren Rangeland Transport, Comm, & Util Upland Forest Urban & Built-Up Water Wetlands Figure 16-B. Changes in the proportions of land us e at station 111. The proportion is the part of the entire contributing zone that is represented by each land use type. The years are each of the land use years used in th is study.

PAGE 136

121 0.0000 0.2000 0.4000 0.6000 0.8000 1.0000 199019992006 YearProportion Agricultural Barren Rangeland Transport, Comm, & Util Upland Forest Urban & Built-Up Water Wetlands Figure 16-C. Changes in the proportions of land us e at station 114. The proportion is the part of the entire contributing zone that is represented by each land use type. The years are each of the land use years used in th is study. 0.0000 0.2000 0.4000 0.6000 0.8000 1.0000 199019992006 YearProportion Agricultural Barren Rangeland Transport, Comm, & Util Upland Forest Urban & Built-Up Water Wetlands Figure 16-D. Changes in the proportions of land us e at station 115. The proportion is the part of the entire contributing zone that is represented by each land use type. The years are each of the land use years used in th is study.

PAGE 137

122 0.0000 0.2000 0.4000 0.6000 0.8000 1.0000 199019992006 YearProportion Agricultural Barren Rangeland Transport, Comm, & Util Upland Forest Urban & Built-Up Water Wetlands Figure 16-E. Changes in the proportions of land us e at station 116. The proportion is the part of the entire contributing zone that is represented by each land use type. The years are each of the land use years used in th is study. 0.0000 0.2000 0.4000 0.6000 0.8000 1.0000 199019992006 YearProportion Agricultural Barren Rangeland Transport, Comm, & Util Upland Forest Urban & Built-Up Water Wetlands Figure 16-F. Changes in the proportions of land us e at station 139. The proportion is the part of the entire contributing zone that is represented by each land use type. The years are each of the land use years used in th is study.

PAGE 138

123 0.0000 0.2000 0.4000 0.6000 0.8000 1.0000 199019992006 YearProportion Agricultural Barren Rangeland Transport, Comm, & Util Upland Forest Urban & Built-Up Water Wetlands Figure 16-G. Changes in the proportions of land us e at station 151. The proportion is the part of the entire contributing zone that is represented by each land use type. The years are each of the land use years used in th is study. 0.0000 0.2000 0.4000 0.6000 0.8000 1.0000 199019992006 YearProportion Agricultural Barren Rangeland Transport, Comm, & Util Upland Forest Urban & Built-Up Water Wetlands Figure 16-H. Changes in the proportions of land us e at station 154. The proportion is the part of the entire contributing zone that is represented by each land use type. The years are each of the land use years used in th is study.

PAGE 139

124 Summary and Conclusions Water quality data for the parameters of dissolved oxygen, turbidity, fecal coliform, total phosphorus, and total nitrogen were examined both graphically and statistically to determine temporal trends in data from the Alafia River. The water quality data was preliminarily studied based on the descrip tive statistics [mean, median, range (minimum to maximum), and standard deviation] for e ach site with the intention of examining the distinguishing features of the datase ts. The water quality data was graphically represented to depict monthly and seaso nal trends with the seasons defined as the wet season (June through September) and the dry season (January through May and October through December) (SWFWMD, 2001). The wate r quality data was also statistically analyzed using the modified Seasonal Kendall test for trend to determine whether there were statistically significant trends in the datasets and in which direction the trends were heading. Land use data from the co ntributing zone of the Alafia River was also examined for 1990, 1999, and 2006 using ge ographic information systems. Water Quality Seasonal Trends Overall, water quality was degraded more during th e wet season than the dry season for every water quality parameter except fec al coliform, which showed no overall season trend, and total nitrogen, which was higher during the dry season (Table 12). Analysis of the datasets for turbidity revealed tha t there was a higher and more variable

PAGE 140

125 trend exhibited by each of the sample stations duri ng the wet season with monthly trends generally peaking around the month of July. Howeve r, all of the overall median concentrations for each sample station were general ly low ( 6.0 NTU). It was also noted that although there was no distinguishable ov erall trend between seasons for any of the sample stations for fecal coliform, very high p eaks in median annual seasonal concentration were exhibited during the wet season. Total phosphorus was shown to typically have higher median concentrations during the wet season than the dry season. It was noted that total nitrogen for the majority of t he sample stations, had higher concentrations during the dry season as opposed to the wet season (Table 12). This trend in total nitrogen was noted previously in the study by Kebede et al. (2003) which attributed the lower wet season concentrations to d ilution from high stream flow. Table 12. Water quality seasonal trend summaries. Parameter Seasonality Seasonal Trend Comments Dissolved Oxygen Yes Lower Wet Turbidity Yes Higher & More Variable in Wet Monthly trends peaking around July. Overall, turbidity generally low ( 6 NTU) Fecal Coliform No N/A High peaks in wet season Total Phosphorus Yes Higher Wet Total Nitrogen Yes Higher Dry Previous study (Kebede et al., 2003) where total nitrogen decreased with increased water discharge

PAGE 141

126 Overall Trends by Sample Station For sample station 74 [located at the mouth of the Alafia River (Figure 5)], there was an obvious separation in monthly median dissolv ed oxygen concentrations when compared to the rest of the stations. This station ’s median concentration of dissolved oxygen over the entire study period (4.6 mg/L) was lower than the state standard of 5.0 mg/L for Class III fresh water bodies (Rule 62-302, F.A.C.) and single sample results for this station showed the levels have dipped much low er (to 0.20 mg/L). Some of the low levels of dissolved oxygen could be attributed to s amples being collected from this station earlier in the morning than the other stati ons, when photosynthesis is low. Also, this station is located at the mouth of the river a nd higher salinity levels can lower dissolved oxygen. Although there has been an incre ase in the proportion of urban and built-up land use in the contributing zone of this station, there have been improvements in water quality for turbidity and total phosphorus, a s indicated by statistically significant decreasing trends resulting from the modified Seaso nal Kendall test. This could possibly be attributed to the decrease in agricultural land use and increase in upland forest in the contributing zone of this station. For sample station 111 [located in the upper porti on of Turkey Creek (Figure 5)], although water quality was shown to be improving wi th a statistically significant decrease in turbidity; fecal coliform and total nitrogen rem ained high at this station. Fecal coliform concentrations at this station (overall me dian concentration of 1050.00 cfu/100mL) were much higher than any other station. The overall mean fecal coliform concentration at this station (2084.38 cfu/100mL) w as much higher than the state monthly average standard of 200 cfu/100mL and also higher than the single sample limit

PAGE 142

127 of 800 cfu/100mL (Rule 62-302, F.A.C). Very high f ecal coliform sample results for this station have been noted previously in the study con ducted by the SWFWMD (2001). This station also had the highest total nitrogen me dian concentration over the entire study period (3.24 mg/L) than any other station. The hig hest proportion of land use for this station throughout the entire study period was agri cultural land use. Agricultural land has been shown to be a strong contributor of nitrogen w ithin contributing zones in studies conducted by Basnyat et al. (1999) and Johnson et a l. (1997). It was also determined that urban and built-up land use has increased in the co ntributing zone of this station over the study period which could be the cause of higher tot al nitrogen and fecal coliform concentrations from animal and human wastes (septic systems). Station 114 [located in the main portion of the Al afia River (Figure 5)] was shown to have improving water quality for total phosphoru s with a statistically significant decreasing trend. A decreasing trend was also note d between 1974 and 1995 in the Comprehensive Watershed Management Plan by the Sout hwest Florida Water Management District (2001). Although water quality was shown to be improving for total phosphorus, total nitrogen declined at this s tation with a statistically significant increasing trend. The main proportion of land use in the contributing zone of this station is wetlands, which have only slightly decreased ove r the study period. It is expected the wetlands would work to filter out the nutrients fro m stormwater runoff, which would explain the reduction in total phosphorus at this s tation, but it does not explain the increasing trend in total nitrogen. Previous techn ical reports indicated that sample analysis from Lithia Springs, located upstream of s tation 114, have shown excessive nitrate levels (Berg et al., 2003; FDEP, 1998) with evidence of a gradual increase over

PAGE 143

128 time (Berge et al., 2003). Since total nitrogen co ncentration is determined by adding nitrate/nitrite levels to total kjeldahl nitrogen l evels, then the excessive and rising nitrate levels discharging from Lithia Springs could be a s trong source of the increasing total nitrogen at this sample station. At station 115 [located in the North Prong, just u pstream of the split from the main section of the Alafia River and upstream of st ation 114 (Figure 5)], the overall total phosphorus monthly median concentrations were highe r at this station than any other station. Although the total phosphorus levels are high, this station has a statistically significant decreasing trend with water quality gen erally improving for this variable. A similar decreasing trend in phosphorus levels at th is station was noted between 1974 and 1995 in the Comprehensive Watershed Management Plan by the SWFWMD (2001). The predominant proportion of land use in the contribut ing zone of this station was wetlands. Over the period of study, the proportion of wetland s for this station increased [percent change of 0.19% (Table A-2)]. The ability of the w etland plants to uptake nutrients is likely to be helping reduce total phosphorus at thi s station. Station 116 [located in the South Prong, just upst ream of the split from the main section of the Alafia River and upstream of station 114 (Figure 5)] showed a statistically significant decreasing trend in dissolved oxygen (s lope estimate of -0.0527mg/year) (Table 3). This information provides evidence that water quality for this variable is degrading at this station; however, the magnitude o f the trend is low and examination of the descriptive statistics showed that annual media n concentrations have stayed above the state standard of 5.0 mg/L (Rule 62-302, F.A.C.). Turbidity was shown to be improving at this station with a statistically significant de crease in concentration. These results are

PAGE 144

129 the opposite of what had been determined in a previ ous report conducted between 1974 and 1995 which showed turbidity increasing that thi s station (SWFWMD, 2001). This is a good sign that turbidity values are beginning to turn around and that water quality for this variable is improving. The highest proportion of land use for the contributing zone of this station is wetlands. At sample station 139 [located in the South Prong of the Alafia River, upstream of station 116 (Figure 5)], there was also a statistic ally significant decrease in turbidity. This affirms that water quality, with respect to tu rbidity, is improving overall in the South Prong of the Alafia River. The highest proportion of land use in the contributing zone for this station is urban and built-up and the proporti on of this type of land use has risen over the study period (by 29.85%) (Table A-2). However, the urban and built-up areas are mainly low-density residential land use with less t han two units per acre. Station 151 [located in the lower portion of Turke y Creek, downstream of station 111 (Figure 5)], was shown to have a statistically significant increase in total nitrogen over the study period. This indicates that water q uality for this parameter is degrading at this station. Although the main land use type in t he contributing zone of this station is urban and built-up, that land is reclaimed land fro m previous mining operations. The reclaimed land appears to be helping remove the hig h fecal coliform levels exhibited upstream at station 111 and reduce the amount seen at this station. Since station 111 is located upstream of this station and the total nitr ogen levels were shown to be high there, the reclaimed land might be helping reduce the amou nt reaching this station, but the vegetation may be reaching its threshold for uptake causing levels to rise at this station.

PAGE 145

130 At station 154 [located in the North Prong, upstre am of station 115 (Figure 5)], although there were no statistically significant tr ends exhibited at this station, higher concentrations of fecal coliform, total phosphorus, and total nitrogen were exhibited at this station. The median concentration of fecal co liform for this station over the entire study period (240.00 cfu/100mL) (Table 6-H) was the second highest median concentration out of all of the stations (the highe st at station 111). Monthly median total phosphorus values tended to have a strong peak duri ng the wet season. Monthly median total nitrogen values were also the second highest out of all the stations (the highest at station 111). The predominant proportion of land u se in the contributing zone of this station over the study period has changed from agri cultural to now urban and built-up. The urban and built-up area for this station includ es, but is not limited to, low residential, industrial, golf courses, extractive, and commercia l land uses. The previously higher agricultural areas or the now high proportion of ur ban land use could be contributing to the excessive nutrients and bacteria. Research Goals The goal of this research was to determine the spa tial and temporal trends in water quality in the Alafia River watershed and to determ ine how those trends relate to land use within the contributing zone of the river. This an alysis was conducted for eight separate sample stations throughout the river to locally ana lyze sections of the river as well as make comparisons between sample stations. The majo r conclusions from this study are: 1. Water quality in the Alafia River has varied over t ime, specifically:

PAGE 146

131 a. By season: Dissolved oxygen was overall lower in t he wet season than in the dry season, turbidity and total phosphorus w ere higher in the wet season than the dry season, fecal coliform had high peaks in the wet season and not in the dry season, and total nitroge n was generally higher in the dry season than in the wet season for most of the river. b. By land use years (1991-1998 and 1999-2006): Disso lved oxygen did not change much between the first eight years and t he second eight years of the study period. Turbidity and total pho sphorus median concentrations were higher in the first half of the study period compared to the second half. Total nitrogen was hi gher the second half of the study period compared to the first half for the majority of the sample stations, and fecal coliform varied for each sample station with the stations on Turkey Creek (stations 111 and 151) and the South Prong headwaters (station 139) having a higher medi an concentration during the second half of the study period and the others, the opposite. c. Over the entire study period (1991-2006): Dissolve d oxygen was determined to be degrading at station 116 with an o verall decreasing trend. Turbidity had a statistically significant d ecreasing trend at stations 74, 111, 116, and 139; thereby, improving water quality for this variable at these stations. Fecal coliform wa s determined to have no statistically significant trends. While total p hosphorus was shown

PAGE 147

132 to be improving with statistically significant decr easing trends at stations 74, 114, and 115; water quality for total nitrogen was generally degrading with overall increasing trends at stations 114, 115, and 151. 2. The proportion of land use within the “contributing zone” of the Alafia River watershed was shown to be predominantly wetla nd and agricultural land use in the beginning of the study period, but changed to urban and built-up land use by the end of the study period. 3. The increasing trends in total nitrogen at stations 114, 115, and 151 could be due to the increased proportion of urban land us e in the area; however, some of the increase in total nitrogen at station 1 14 was likely due to excessive nitrates discharging from Lithia Springs. The high fecal coliform and total nitrogen at station 111 was like ly attributed to the high proportion of agricultural and urban and built-up l and use around this station. Other observed trends in turbidity and to tal phosphorus were reductions in the concentrations throughout the wat ershed which is a positive sign considering urban and built-up land u se has increased. Implications The quality of the Alafia River is important to un derstand since this river is used for recreational purposes and it is also a source o f drinking water for local residents. By

PAGE 148

133 determining the trends in water quality in the rive r, changes can be made to implement new environmental regulatory requirements, make lif estyle changes, and otherwise be more aware of the quality of this natural resource. Previous water quality studies conducted in the Al afia River watershed have mainly been technical reports that focused on estab lishing an overall rating of either “poor”, “fair”, or “good” for each section of the r iver based on calculation of the Water Quality Index. Other studies have examined the Ala fia River watershed only as a part of the Tampa Bay watershed and focused mainly on overa ll non-point source loadings from the Alafia River to the Tampa Bay. In each of thes e studies, the analysis was based on data that had been converted over to annual average s and not on the raw monthly or quarterly sample results. This study has not only provided a more recent, lo ng-term trend analysis of monthly sampled water quality variables within the Alafia River watershed, but has also incorporated seasonal trends which are lacking in p revious studies of this watershed. This evaluation provided new information about this watershed indicating total nitrogen was more degraded (higher) during the dry season an d dissolved oxygen was more degraded (lower) during the wet season. This study also included a detailed statistical tr end analysis of water quality over the entire study period in the Alafia River watersh ed which is also lacking in the previous studies of this watershed. This information provid es a base knowledge of how water quality has changed over time in the watershed and bases that knowledge on statistical analysis. The steps used in this study to determin e the appropriate statistical test to use in

PAGE 149

134 the analysis of trends in water quality can also be used as a guideline for future studies on water quality trend analysis. This study also incorporated land use as a variabl e and examined how changes in land use may be contributing to the trends exhibite d in water quality. Previous studies have only analyzed the land use in the area at one point in time. Examination of changes in land use provides a basis for correlating land u se and water quality to determine how a change in land use can affect water quality. This research has demonstrated the need for more a ttention and resources to be focused on improving water quality in the Alafia Ri ver. Efforts should be focused on identifying potential contributors of fecal colifor m and total nitrogen to the Alafia River, especially upstream of station 111 in Turkey Creek and in the North Prong of the river. Investigations should also be made into the source of the low dissolved oxygen concentrations consistently experienced at the mout h of the river. This research has provided information on seasonal trends in the watershed which indicate turbidity, fecal coliform, and total phosp horus all have concentrations higher during the wet season than in the dry season. More focus can be put on reducing the impact of these pollutants during storm events. Th is research has also demonstrated that turbidity and total phosphorus have shown decreasin g trends within the watershed while total nitrogen has an increasing trend. This infor mation can be used by environmental regulators to help prioritize specific water qualit y pollutants to devote more resources and time to reducing. The land use trend information examined in this re search study has indicated that urban and built-up land use continues to rise in th e watershed while wetlands are

PAGE 150

135 decreasing. This information provides a general ba ckground of changes occurring in the watershed and reaffirms the need for continued educ ation on best management practices and low impact development to try to curb the effec ts of growth on water quality. Additional Research Recommendations for further research into the spat ial and temporal trends in water quality in the Alafia River watershed and how land use relates to these trends: 1. Determine the trends in water quality on flow-adjus ted data. Water quality is often correlated to river flow. Since f lows vary throughout the watershed and some water quality concentrations may be diluted with higher discharge while others may be more concentra ted, adjusting for the effects of flow on the data may provide more accura te results for trend by eliminating this variability. 2. Further examine trends in water quality grouped by season (wet vs. dry). Since concentrations of some water quality variable s vary greatly between seasons, an overall trend in the dataset may not be detected. By grouping the data based on the median seasonal value for eac h year and running a statistical test for trend, the data may no longer be serially correlated and the amount of statistically significant trends dete cted may change.

PAGE 151

136 3. Examine the trends for the entire watershed as a wh ole. By examining the watershed as a whole, total pollutant loadings from the watershed can be determined. 4. Determine the correlation between each water qualit y variable and each land use type in the watershed. By examining corre lations between water quality and land use, better associations can be ma de about the effects of land use and land use change on water quality. 5. Incorporate the effects of precipitation on water q uality in the watershed. Since non-point source pollution is the main cause of water quality pollution, examining the effects of precipitation o n water quality would provide a better understanding of the impacts of st ormwater runoff on the river, especially if samples are collected during s torm events. 6. Conduct a more detailed GIS analysis which includes layers such as stormwater inlets and pipes, septic systems, and/or land use subgroups within the watershed.

PAGE 152

137 List of References Basnyat, P., L.D. Teeter, K.M. Flynn, and B.G. Lock aby. (1999). Relationships Between Landscape Characteristics and Nonpoint Source Pollu tion Inputs to Coastal Estuaries. Environmental Management 23(4), 539-549. Berg, B., L. McCoy, M. Forhand, J. Miller, L.A. Elg in. (2003). Lithia Springs: A Biological Assessment : University of South Florida. Berryman, D., B. Bobee, D. Cluis, and J. Haemmerli. (1988). Nonparametric Tests for Trend Detection in Water Quality Time Series. Water Resources Bulletin 24(3), 545-556. Brady, D. J. (1996). The Watershed Protection Appro ach. Water Science and Technology 33(4-5), 17-21. Brown, M. T., and M.B. Vivas. (2005). Landscape Dev elopment Intensity Index. Environmental Monitoring and Assessment 101, 289-309. Brown, R. S., and K. Marshall. (1996). Ecosystem Ma nagement in State Governments. Ecological Applications 6(3), 721-723. Cavanaugh, T. M., and W.J. Mitsch. (1989). Water Qu ality Trends of the Upper Ohio River from 1977 to 1987. The Ohio Journal of Science 89(5), 153-163. Chang, H. (2008). Spatial Analysis of Water Quality Trends in the Han River Basin, South Korea. Water Research 42, 3285-3304. Clesceri, L. S., A.E. Greenberg, A.D. Eaton (Ed.). (1998). Standard Methods for the Examination of Water and Wastewater (20th ed.). Washington, DC: American Public Health Association, American Water Works Ass ociation, and Water Environment Federation. DeBusk, W. (2002). Surface Water Quality Assessment if Florida: The 305(b) Report and 303(d) List. Retrieved September 20, 2008, fr om http://edis.ifas.ufl.edu/SS405 ESRI Inc. (2008). ArcMap 9.3. Redlands, CA.

PAGE 153

138 Florida Department of Environmental Protection (FDE P). (1998). EcoSummary: Lithia Springs, Hillsborough County. Alafia River Watershed Florida Department of Environmental Protection (FDE P). (2008). Florida Section 319 Grant Work Plans and Project Summaries. Nonpoint Source Management Retrieved September 20, 2008, from http://www.dep.state.fl.us/water/nonpoint/319h.htm Florida Department of Environmental Protection (FDE P). (2009). Florida’s Surface Water Quality Standards. Refining Designated Uses & Classifications Frequently Asked Questions. Retrieved November 17, 2009, from http://www.dep.state.fl.us/northeast/CurrentTopics/Surface_WQ_StandardsFAQ.pdf Fraser, T. H. (1986). Long-Term Water-Quality Characteristics of Charlott e Harbor, Florida : U.S. Department of the Interior, U.S. Geological Survey. Gilbert, R. O. (1987). Statistical Methods for Environmental Pollution Mon itoring New York, NY: John Wiley & Sons, Inc. Hirsch, R. M., J.R. Slack, and R.A. Smith. (1982). Techniques of Trend Analysis for Monthly Water Quality Data. Water Resources Research 18(1), 107-121. Hirsch, R. M., and J.R. Slack. (1984). A Nonparamet ric Trend Test for Seasonal Data with Serial Dependence. Water Resources Research 20(6), 727-732. Hirsch, R. M., R.B. Alexander, and R.A. Smith. (199 1). Selection of Methods for the Detection and Estimation of Trends in Water Quality Water Resources Research 27(5), 803-813. Insightful Corp. (2007). S-PLUS (Version 8.0 for Wi ndows). Johnson, L. B., C. Richards, G.E. Host, and J.W. Ar thur. (1997). Landscape Influences on Water Chemistry in Midwestern Stream Ecosystems. Freshwater Biology 37, 193-208. Kebede, E., J. Gan, and Z. Gebeyehu. (2003). Non-Po int Source Pollution and Land Use Pattern Linkage: A Watershed Approach. Journal of the Alabama Academy of Science 74(3/4), 197-209. Lehrter, J. C. (2006). Effects of Land Use and Land Cover, Stream Discharge, and Interannual Climate on the Magnitude and Timing of Nitrogen, Phosphorus, and Organic Carbon Concentrations in Three Coastal Plai n Watersheds. Water Environment Research 78(12), 2356-2368.

PAGE 154

139 Maillard, P., and N. Pinheiro Santos. (2008). A Spa tial-Statistical Approach for Modeling the Effect of Non-point Source Pollution on Differe nt Water Quality Parameters in the Velhas River Watershed Bazil. Journal of Environmental Management 86, 158-170. Microsoft Corporation. (2003a). Microsoft Office Ac cess 2003. Microsoft Corporation. (2003b). Microsoft Office Ex cel 2003. Microsoft Corporation. (2003c). Microsoft Office Wo rd 2003. Millard, S. P., and N.K. Neerchal. (2001). Environmental Statistics with S-PLUS Boca Raton, FL: CRC Press LLC. Millard, S. P. (2002). EnvironmentalStats for S-PLU S (Version 2.0). Probability, Statistics & Information, Seattle, WA. Parsons Engineering Science, Inc. (2002). Alafia Ri ver Watershed Management Plan. Volume II, Chapters 7 through 12. Retrieved Octob er 19, 2009, from http://www.hillsborough.wateratlas.usf.edu/upload/d ocuments/Volume02.pdf Qian, Y., K.W. Migliaccio, Y. Wan, and Y. Li. (2007 a). Trend Analysis of Nutrient Concentrations and Loads in Selected Canals of the Southern Indian River Lagoon, Florida. Water, Air, and Soil Pollution 186, 195-208. Qian, Y., K.W. Migliaccio, Y. Wan, Y.C. Li, and D. Chin. (2007b). Seasonality of Selected Surface Water Constituents in the Indian R iver Lagoon, Florida. Journal of Environmental Quality 36, 416-425. Sliva, L., and D.D. Williams. (2001). Buffer Zone V ersus Whole Catchment Approaches to Studying Land Use Impact on River Water Quality. Water Research 35(14), 3462-3472. Smith, R. A., R.B. Alexander, M.G. Wolman. (1987). Water-Quality Trends in the Nation's Rivers. Science 235(4796), 1607-1615. Southwest Florida Water Management District (SWFWMD ). (2001). Alafia River Comprehensive Watershed Management Plan. Retrieve d September 20, 2008, from http://www.swfwmd.state.fl.us/documents/plans/cwm/c wm-alafiariver.pdf SPSS Inc. (2007). SPSS 16.0 Command Syntax Referenc e. Chicago, IL. Tampa Bay Water. (2008). Alafia River and Tampa Byp ass Canal Reclassification Petitions. Reclassification Petitions Retrieved October 7, 2008, from http://www.tampabaywater.org/watersupply/reclassifi cation.aspx

PAGE 155

140 Tong, S., and W. Chen. (2002). Modeling the Relatio nship Between Land Use and Surface Water Quality. Journal of Environmental Management 66, 377-393. U.S. Census Bureau. (2007). Table 7: Population Es timates for the 100 Largest U.S. Counties Based on July 1, 2006 Population Estimates : April 1, 2000 to July 1, 2006. Population Estimates Retrieved November 20, 2007, from http://www.census.gov/popest/counties/CO-EST2006-07 .html U.S. Census Bureau. (2008). Geographic Area: Hills borough County, Florida. General Population and Housing Characteristics: 1990 Retrieved November 20, 2007, from http://factfinder.census.gov/servlet/QTTable?_bm=n& _lang=en&qr_name=DEC_ 1990_STF1_DP1&ds_name=DEC_1990_STF1_&geo_id=05000US 12057 U.S. Environmental Protection Agency (U.S. EPA). (2 006). Clean Water Act Section 303. Retrieved February 15, 2008, from http://www.epa.gov/waterscience/standards/303.htm U.S. Environmental Protection Agency (U.S. EPA). (2 007). 2002 Section 303(d) List Fact Sheet for Florida. Retrieved November 20, 20 07, from http://oaspub.epa.gov/waters/state_rept.control?p_s tate=FL U.S. Environmental Protection Agency (U.S. EPA). (2 008a). Approved General-Purpose Methods. Clean Water Act Analytical Methods Retrieved October 20, 2009, from http://www.epa.gov/waterscience/methods/method/ U.S. Environmental Protection Agency (U.S. EPA). (2 008b). What is Nonpoint Source (NPS) Pollution? Questions and Answers. Polluted Runoff (Nonpoint Source Pollution) Retrieved November 17, 2009, from http://www.epa.gov/owow/nps/qa.html van Belle, G., and J.P. Hughes. (1984). Nonparametr ic Tests for Trend in Water Quality. Water Resources Research 20(1), 127-136. Wang, X. (2001). Integrating Water-Quality Manageme nt and Land-Use Planning in a Watershed Context. Journal of Environmental Management 61, 25-36. World Wide Metric. (2006). Conversion Calculator. Retrieved February 10, 2008, from http://www.worldwidemetric.com/metcal.htm Xian, G., M. Crane, and J. Su. (2007). An Analysis of Urban Development and its Environmental Impact on the Tampa Bay Watershed. Journal of Environmental Management 85, 965-976. Yu, Y. S., S. Zou, and D. Whittemore. (1993). Non-P arametric Trend Analysis of Water Quality Data of Rivers in Kansas. Journal of Hydrology 150, 61-80.

PAGE 156

141 Appendices

PAGE 157

142 Appendix A: Land Use Tables Table A-1. The proportion of each land use area to the total area within the contributing zone of each sample station.* The proportions are separated based on the year of the land use data (1990, 1999, and 2006). Each station ’s highest land use proportion is in bold. Land Use Classification Year Station Agricultural Barren Rangeland Transport., Comm., & Util. Upland Forest Urban & Built-Up Water Wetlands 1990 74 0.1779 0.0019 0.0079 0.0093 0.1494 0.2454 0.1121 0.2960 111 0.5167 0.0000 0.0001 0.0055 0.0242 0.2103 0.0057 0.2375 114 0.2279 0.0010 0.0103 0.0063 0.1331 0.0864 0.0389 0.4960 115 0.1294 0.0000 0.0182 0.0270 0.0911 0.2155 0.0204 0.4984 116 0.1410 0.0000 0.0323 0.0024 0.0837 0.0764 0.0083 0.6558 139 0.0697 0.0000 0.0071 0.0031 0.1473 0.3825 0.0023 0.3879 151 0.0468 0.0000 0.0000 0.0272 0.0000 0.8798 0.0073 0.0390 154 0.3991 0.0021 0.0448 0.0234 0.1473 0.1153 0.0064 0.2615 1999 74 0.1710 0.0000 0.0062 0.0084 0.1573 0.2594 0.1121 0.2856 111 0.5249 0.0000 0.0001 0.0059 0.0287 0.2425 0.0101 0.1878 114 0.2427 0.0000 0.0072 0.0064 0.1366 0.1098 0.0443 0.4530 115 0.1039 0.0000 0.0113 0.0312 0.0842 0.2396 0.0174 0.5122 116 0.1468 0.0000 0.0239 0.0014 0.0807 0.1047 0.0078 0.6347 139 0.0615 0.0000 0.0000 0.0024 0.1220 0.4779 0.0059 0.3303 151 0.0416 0.0000 0.0000 0.0000 0.0153 0.9139 0.0000 0.0292 154 0.2734 0.0076 0.0413 0.0185 0.1847 0.2112 0.0128 0.2505 2006 74 0.1315 0.0000 0.0042 0.0093 0.1516 0.3055 0.1267 0.2713 111 0.4299 0.0000 0.0000 0.0073 0.0281 0.3258 0.0135 0.1954 114 0.2074 0.0097 0.0057 0.0064 0.1319 0.1324 0.0505 0.4560 115 0.0831 0.0027 0.0104 0.0312 0.0719 0.2854 0.0159 0.4993 116 0.1301 0.0000 0.0212 0.0014 0.0852 0.1253 0.0072 0.6296 139 0.0432 0.0000 0.0000 0.0024 0.1220 0.4967 0.0028 0.3330 151 0.0244 0.0000 0.0000 0.0000 0.0153 0.9310 0.0000 0.0292 154 0.2204 0.0003 0.0345 0.0234 0.1716 0.2774 0.0176 0.2547 All proportions were rounded from six decimal pla ces to four for table formatting.

PAGE 158

143 Appendix A (continued) Table A-2. The percent change* in the proportion o f each land use area from 1990 – 1999, 1999 – 2006, and 1990 – 2006 for each sample station. Land Use Classification Years Station Agricultural Barren Rangeland Transport., Comm., & Util. Upland Forest Urban & Built-Up Water Wetlands 74 -3.88% -100.00% -21.50% -10.03% 5.27% 5.69% 0.02 % -3.52% 1990 1999 111 1.57% N/A -27.94% 5.88% 18.85% 15.31% 78.92% -2 0.90% 114 6.49% -99.97% -30.50% 1.45% 2.61% 27.08% 13.91 % -8.68% 115 -19.69% N/A -37.48% 15.62% -7.53% 11.18% -14.6 7% 2.77% 116 4.11% N/A -26.00% -40.76% -3.63% 37.01% -6.35% -3.22% 139 -11.77% N/A -100.00% -21.44% -17.16% 24.93% 15 2.83% -14.86% 151 -11.12% N/A N/A -100.00% N/A 3.87% -100.00% -2 4.96% 154 -31.51% 255.57% -7.79% -20.82% 25.43% 83.18% 99.25% -4.23% 74 -23.10% N/A -32.26% 11.03% -3.62% 17.76% 12.95% -5.02% 1999 2006 111 -18.09% N/A -100.00% 25.30% -2.30% 34.35% 32.84 % 4.03% 114 -14.55% 3098445.08% -20.41% -0.47% -3.45% 20.5 7% 13.96% 0.67% 115 -20.05% N/A -8.06% 0.03% -14.69% 19.13% -8.65% -2.52% 116 -11.38% N/A -11.41% 0.00% 5.60% 19.64% -7.79% -0.79% 139 -29.86% N/A N/A 0.01% -0.03% 3.93% -52.64% 0.8 2% 151 -41.25% N/A N/A N/A 0.00% 1.88% N/A 0.00% 154 -19.36% -95.75% -16.42% 25.97% -7.09% 31.36% 37.64% 1.69% 74 -26.08% -100.00% -46.83% -0.10% 1.46% 24.46% 12. 98% -8.36% 1990 2006 111 -16.80% N/A -100.00% 32.67% 16.12% 54.92% 137.6 9% -17.72% 114 -9.01% 908.99% -44.69% 0.98% -0.93% 53.22% 29. 82% -8.08% 115 -35.79% N/A -42.52% 15.66% -21.11% 32.45% -22. 05% 0.19% 116 -7.74% N/A -34.44% -40.76% 1.76% 63.91% -13.64 % -3.99% 139 -38.12% N/A -100.00% -21.43% -17.19% 29.85% 19 .75% -14.16% 151 -47.78% N/A N/A -100.00% N/A 5.83% -100.00% -2 4.96% 154 -44.77% -84.89% -22.93% -0.26% 16.53% 140.63% 174.25% -2.61% The percent changes were all rounded to two decim al places for table formatting.


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200409Ka 4500
controlfield tag 001 002069273
005 20100420151250.0
007 cr bnu|||uuuuu
008 100420s2009 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0003163
035
(OCoLC)608084231
040
FHM
c FHM
043
n-us-fl
049
FHMM
090
G116 (Online)
1 100
Aragon, Jennifer M.
0 245
Spatial and temporal trends in water quality in the Alafia River watershed
h [electronic resource] /
by Jennifer M. Aragon.
260
[Tampa, Fla] :
b University of South Florida,
2009.
500
Title from PDF of title page.
Document formatted into pages; contains 143 pages.
502
Thesis (M.S.)--University of South Florida, 2009.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
520
ABSTRACT: Water quality data and land use information were analyzed within the Alafia River watershed in Florida to determine spatial and temporal trends in these variables over a 16 year time period from 1991-2006. Monthly water quality data (for dissolved oxygen, turbidity, fecal coliform, total phosphorus, and total nitrogen) were statistically analyzed using the modified seasonal Kendall nonparametric test for trends that accounts for serial correlation. The statistical trend analysis was conducted for the entire study period, but monthly, seasonal, and land use trends were also examined. Land use information was examined using Geographic Information Systems to determine the percent change in land use proportion from 1990 to 1999, 1999 to 2006, and 1990 to 2006. The proportions of each land use and their percent change were then related to the trends in water quality.The results of this analysis showed that water quality for the parameters turbidity and total phosphorus have been shown to be improving with statistically significant decreasing trends for turbidity at stations 74, 111, 116, and 139 and for total phosphorus at stations 74, 114, and 115. A statistically significant decreasing trend in dissolved oxygen was determined for stations 116 and an increasing trend in total nitrogen for stations 114, 115, and 151 implying water quality for these parameters is degrading. Other noted trends were high fecal coliform and total nitrogen at station 111, which has higher proportions of agricultural land use and an increasing proportion of urban and built-up land use. Also, low dissolved oxygen was noted at station 74. The proportions of land use for the entire study area have changed from predominantly wetlands to now urban and built-up land use.While agricultural, rangeland, and wetlands land use have shown a reduction in the proportion of coverage in the contributing zone of almost every station, urban and built-up land use has increased in proportion at every station.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
590
Advisor: Philip Reeder, Ph.D.
653
Seasonal kendall
Nonparametric
Land use
Tampa
Florida
Geographic information systems
690
Dissertations, Academic
z USF
x Geography
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.3163