<%BANNER%>

Nitrogen and Phosphorus Transport in an Urban Watershed

Permanent Link: http://ufdc.ufl.edu/UFE0042263/00001

Material Information

Title: Nitrogen and Phosphorus Transport in an Urban Watershed
Physical Description: 1 online resource (113 p.)
Language: english
Creator: Kamaljit, Kamaljit
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: eutrophication, nitrogen, nutrients, phosphorus, watershed
Soil and Water Science -- Dissertations, Academic -- UF
Genre: Soil and Water Science thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Non-point source pollution is the dominant pathway of nitrogen (N) and phosphorus (P) transport in agricultural, urbanized, and rapidly urbanizing watersheds. We used monthly concentrations data of inorganic and organic forms of N and P in stream waters draining different sub-basins, ranging in size from 19 to 350 km2, of the Alafia River Watershed (total drainage area: 1085 km2), which ultimately drains to Tampa Bay Estuary, to understand N and P transport. The sub-basins were classified based on the percentage of urban land use as three developed (18?24% residential, 1?14% built up) and two undeveloped (3?11% residential, 1?3% built up). Urban land use at two mainstem stations that drained 80?99% of the watershed was 16?17% residential and 3% built up. During 1991?2009, total N concentrations ranged from 0.8 to 2.4 mg L-1 and were greatest in stream waters draining developed (1.7?2.4 mg L-1) than undeveloped (0.8?1.2 mg L-1) sub-basins. Inorganic N (primarily NO3-N) was the dominant form in streams draining developed sub-basins while organic N was greater in streams draining undeveloped sub-basins. Total P concentrations ranged from 0.6 to 3.9 mg L-1 and were not different among developed (0.8?3.9 mg L-1) and undeveloped (0.6?0.9 mg L-1) sub-basins. Of total P, 70?90% was dissolved reactive P while other P forms were 10?30% of total P in both developed and undeveloped sub-basins. The increasing total N and decreasing total P concentrations trends at the mainstem station draining 89% of the watershed over the 19-year period suggests that the development of the watershed resulted in increasing N but not P concentrations in streams. We suggest that the BMP?s to reduce N loss from urban land uses in three developed sub-basins (with total N of 1.7?2.4 mg L-1) may yield greater reductions in N concentrations at watershed outlet (i.e. mainstem) to achieve EPA proposed numeric criteria of total N concentration of 1.798 mg L-1. On the other hand, due to P rich geology and discharge from the wastewaters, most developed and undeveloped sub-basins had greater total P concentrations (0.8?3.9 mg P L-1) than EPA proposed numeric total P value of 0.739 mg L-1 indicating that BMP?s should focus on reducing P loss from phosphate rock mined sub-basins and reduce P inputs from wastewater.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Kamaljit Kamaljit.
Thesis: Thesis (M.S.)--University of Florida, 2010.
Local: Adviser: Toor, Gurpal Singh.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042263:00001

Permanent Link: http://ufdc.ufl.edu/UFE0042263/00001

Material Information

Title: Nitrogen and Phosphorus Transport in an Urban Watershed
Physical Description: 1 online resource (113 p.)
Language: english
Creator: Kamaljit, Kamaljit
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: eutrophication, nitrogen, nutrients, phosphorus, watershed
Soil and Water Science -- Dissertations, Academic -- UF
Genre: Soil and Water Science thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Non-point source pollution is the dominant pathway of nitrogen (N) and phosphorus (P) transport in agricultural, urbanized, and rapidly urbanizing watersheds. We used monthly concentrations data of inorganic and organic forms of N and P in stream waters draining different sub-basins, ranging in size from 19 to 350 km2, of the Alafia River Watershed (total drainage area: 1085 km2), which ultimately drains to Tampa Bay Estuary, to understand N and P transport. The sub-basins were classified based on the percentage of urban land use as three developed (18?24% residential, 1?14% built up) and two undeveloped (3?11% residential, 1?3% built up). Urban land use at two mainstem stations that drained 80?99% of the watershed was 16?17% residential and 3% built up. During 1991?2009, total N concentrations ranged from 0.8 to 2.4 mg L-1 and were greatest in stream waters draining developed (1.7?2.4 mg L-1) than undeveloped (0.8?1.2 mg L-1) sub-basins. Inorganic N (primarily NO3-N) was the dominant form in streams draining developed sub-basins while organic N was greater in streams draining undeveloped sub-basins. Total P concentrations ranged from 0.6 to 3.9 mg L-1 and were not different among developed (0.8?3.9 mg L-1) and undeveloped (0.6?0.9 mg L-1) sub-basins. Of total P, 70?90% was dissolved reactive P while other P forms were 10?30% of total P in both developed and undeveloped sub-basins. The increasing total N and decreasing total P concentrations trends at the mainstem station draining 89% of the watershed over the 19-year period suggests that the development of the watershed resulted in increasing N but not P concentrations in streams. We suggest that the BMP?s to reduce N loss from urban land uses in three developed sub-basins (with total N of 1.7?2.4 mg L-1) may yield greater reductions in N concentrations at watershed outlet (i.e. mainstem) to achieve EPA proposed numeric criteria of total N concentration of 1.798 mg L-1. On the other hand, due to P rich geology and discharge from the wastewaters, most developed and undeveloped sub-basins had greater total P concentrations (0.8?3.9 mg P L-1) than EPA proposed numeric total P value of 0.739 mg L-1 indicating that BMP?s should focus on reducing P loss from phosphate rock mined sub-basins and reduce P inputs from wastewater.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Kamaljit Kamaljit.
Thesis: Thesis (M.S.)--University of Florida, 2010.
Local: Adviser: Toor, Gurpal Singh.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0042263:00001


This item has the following downloads:


Full Text





NITORGEN AND PHOSPHORUS TRANSPORT IN AN URBAN WATERSHED


By

KAMALJIT KAMALJIT


















A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA

2010


































2010 Kamaljit Kamaljit





















This work is dedicated to my mom









ACKNOWLEDGMENTS

I would like to thank my Supervisory Committee for their patience and guidance. Special

appreciation goes to Dr. Gurpal Toor, the chair of my supervisory committee, for his intellectual

guidance, consistent support, and endless efforts to accomplish this project. Dr. Patrick Inglett is

thanked for his critical reviews of chapters. Additionally, unique gratitude goes to Dr. Craig

Stanley for his interests and invaluable suggestions during the period of study.

I would like to thank Mr. Butch Bradley; former laboratory manager at the Soil and Water

Quality Laboratory and my fellow graduate students: Lu Han, Michael Miyittah, and Maninder

Chahal for their help in the laboratory analysis. I express my greatest appreciation to my mother,

brother, and sister who have continually supported all my dreams with patience. Last, and most

importantly, I thank my fiancee Aparna, for her constant encouragement and for standing by me

in all my academic endeavors.

I could not end my acknowledgements without recognizing that ultimately it has been by

the grace of God, and I most gratefully submit my thanks and praise to Him for any good that

comes into my life.









TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ......... ................. .................................................................................... 4

L IST O F T A B L E S ... .. .................. ..... ..................................................................... .............. ....... 8

L IST O F F IG U R E S ................................................................. 9

A B S T R A C T ............................................................................................... ...... 1 1

CHAPTER

1 IN TR O D U C TIO N ...................................................... 13

E utrophication ................ ... ... ...... ....................... ................. ............... 13
Nitrogen and Phosphorus in the Environment......................... ............. 14
N itro g en ................... ................... ...................1.........4
P h o sp h oru s ..................................... ... ............................ ....... .. 14
Sources of Nitrogen and Phosphorus in Watersheds .................................. 15
R research O bjectiv es ............... ............ .......... ............... ... ............ ..... ........... ............... 17
Objective 1. Evaluation of Nitrogen Concentrations in Different Sub-basins of the
Alafia River W watershed. ............... ..... .... ........... ...................... 18
Objective 2. Evaluation of Phosphorus Concentrations in Different Sub-basins of
the Alafia River Watershed .......... ............................. 18

2 STUDY SITE DESCRIPTION ........... ............................. 20

Location .. ..... ........................................................ 20
C lim ate ............................................................................... 20
Sub-basins of the Alafia River W atershed ............................................................. .... 20
D ev elo p ed Su b -b a sin s ................................................................... ............................. 2 1
T u rk ey C re ek ................................................................................................... 2 1
E n g lish C reek .................................................................. 2 1
N o rth P ro n g .................................................................................................... 2 2
U developed Sub-basins .................................................... ............... 23
S o u th P ro n g .................................................................................................... 2 3
F ishhaw k C reek ...................................................... 23

3 NITROGEN TRANSPORT IN AN URBAN WATERSHED ..............................................32

A b stra c t ............. ...... ........... ................. ........................ 3 2
Introduction ........................................ ..................... 33
Materials and Methods ......................................... 37
Study Site D description ..................................................... 37
Data Collection .............................................................. 37
Stream -w ater Collection and Analysis ..................................................................... 37









S statistical A n aly sis............................................................... .......................................... 3 8
R e su lts .......................................................................................................................................... 3 9
Chemical Characteristics of Stream Waters.................................... .............. 39
Concentrations of Nitrogen Forms in Streams Draining Different Sub-basins ..............40
Seasonal Variations in Chemical Characteristics in Stream Waters Draining
Different Sub-basins ........................................................... ... ......................... 42
Seasonal Variations in Concentrations of Nitrogen Forms in Streams Draining
D different Sub -basins .................................. ............... .... ............. ... ....... ..... 43
Long Term Trends in Flow Un-weighted Nitrogen Forms in Streams Draining
D different Su b -b a sin s ................................................. ................................................. 4 4
Total Nitrogen ..... ........................................ ................ 44
O rg an ic N itro g en ............................................................... ..................................... 4 5
N itrate N itrogen............................................. ........ .......... ...... 4 5
A m m onium N nitrogen ........................................... .............. ...... ... ....... ..... 46
Long Term Trends in Flow Weighted Nitrogen Forms in Streams Draining at
M ainstem Station .................................... ...... ....... ..... .. ............... 46
Long Term Trends in Nitrogen Loads at Mainstem Station .................. .............. ....46
Relationship between Land Use and Nitrogen Forms........................................... 47
D iscussion......................................... .. .............. ........... ............... 47
Influence of Land Uses on Total Nitrogen Concentrations in Stream Waters ................47
Land Uses and Forms of Nitrogen Concentrations in Stream Waters............................ 49
Seasonal Impacts on Nitrogen Forms in Stream Waters............. ........ ........... .... 50
Long Term Trends in Nitrogen Concentrations.................... ................... 51
Sum m ary ... ......... ............. ............. ........ ................................... ......... ...... 57

4 PHOSPHORUS TRANSPORT IN AN URBAN WATERSHED...........................................71

A b stra c t ........................................................................................................................................ 7 1
Introduction ........................................ ..................... 72
M materials and M ethod s................................................................................... ........................ 7 5
Stu dy Site D description .............................................................................. ........... .......... 7 5
Data Collection ..................................................................... 76
Stream -w ater C collection and A analysis ........................................ ......................... .. 76
Statistical Analysis ....................................................... 76
R e su lts .......................................................................................................................................... 7 8
Chemical Characteristics of Stream Waters....................... .......................... 78
Concentrations of Phosphorus Forms in Streams Draining Different Sub-basins ..........78
Seasonal Variation in Phosphorus Concentrations in Streams Draining Different
Sub -basins............................ ......... ...... ....... ......... ............ 80
Long Term Trends in Concentrations of Flow Un-weighted Phosphorus Forms in
Stream s D raining D different Sub-basins..................................................... ................ 81
Long Term Trends in Flow Weighted Concentrations at Mainstem Station ................... 83
Long Term Trends in Phosphorus Loads at Mainstem Station.............. ............... 83
D discussion ................................................................ ........ .... ................. 83
Land Use Impacts on Phosphorus Concentrations in Streams............... .............. 83
Seasonal Impact on Phosphorus Concentrations in Streams ..................................... 85
Long Term Trends in Concentration of Phosphorus Forms in Stream Waters................ 86


6









Su m m ary ............... .......... ............................................... ........... ...... 8 8

5 SUMMARY, CONCLUSIONS, AND RECOMMENDATION....................................... 100

L IST O F R E F E R E N C E S ................................................................................................................. 104

B IO G R A PH IC A L SK E T C H ...................................................... ............................................... 113
















































7









LIST OF TABLES


Table page

2-1 Station characteristics and associated land uses in the various sub-basins on the
A lafia R iv er W atersh ed .............................................................................. ...................... 2 4

2-2 Grouping of FLUCCS codes into major land uses...........................................................25

3-1 Station characteristics of the Alafia River Watershed............................................. 58

3-2 Long-term trends in flow weighted and loads ofN forms at Bell Shoals ........................ 58

4-1 Station characteristics of the Alafia River Watershed ......................................................90

4-2 Long-term trends in flow weighted and loads of P forms at Bell Shoals........................... 90









LIST OF FIGURES


Figure pe

1-1 Estimated sources of nitrogen in the Tampa Bay...................... .................... .......... 19

2-1 Location map and land uses in the various sub-basins of the Alafia River Watershed.....26

2-2 Land use in the Turkey Creek sub-basin in 1990, 1999, and 2007................................. 27

2-3 Land use in the English Creek sub-basin in 1990, 1999, and 2007...............................28

2-4 Land use in the North Prong sub-basin in 1990, 1999, and 2007. ...............................29

2-5 Land use in the South Prong sub-basin in 1990, 1999, and 2007. .................................. 30

2-6 Land use in the Fishhawk Creek sub-basin in 1990, 1999, and 2007.............................31

3-1 Location map of the Alafia River W atershed........ ....... ..... ............. ... ............ 59

3-2 Chemical characteristics of the stream waters during two time periods from 1991 to
2009....... ......................................... ................ 60

3-3 Summary of mean monthly concentrations of total, organic, nitrate, and ammonium
nitrogen during 1991-2009 in mainstem (Alafia and Bell Shoals), developed
(English Creek, Turkey Creek, and North Prong), and undeveloped (South Prong and
Fishhawk Creek) of the Alafia River W atershed............................................................ 61

3-4 Seasonal variation in chemical characteristics of the stream waters during 1991-
2009 ......................... ................ 62

3-5 Seasonal variation in mean monthly concentration of nitrogen forms during 1991-
2009....... ......................................... ................ 63

3-6 Seasonal variation in proportion of organic, nitrate, and ammonium nitrogen during
1991-2009 ...... .......................... ........... ................. .. 64

3-7 Long-term (1991-2009) trends in monthly flow un-weighted total N concentrations
in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and
North Prong), and undeveloped (South Prong and Fishhawk Creek) sub-basins of the
A lafia R iv er W atersh ed ................................................... ............................................... 6 5

3-8 Long-term (1991-2009) trends in monthly flow un-weighted organic N
concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek,
Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk
Creek) sub-basins of the Alafia River Watershed........................... ............... 66

3-9 Long-term (1991-2009) trends in monthly nitrate N concentrations in mainstem
(Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North









Prong), and undeveloped (South Prong and Fishhawk Creek) sub-basins of the Alafia
R iver W watershed. .............................................................................. 67

3-10 Long-term (1991-2009) trends in monthly flow un-weighted ammonium N
concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek,
Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk
Creek) sub-basins of the Alafia River Watershed............ .............................68

3-11 Relationship between percent urban and agricultural land use and nitrogen forms in
different sub -basin s. .......................................................................... ..................... ...... 69

3-12 Relationship between pasture and forest land use and nitrogen forms in streams
draining different su b -basin s ......................................... ................................................. 70

4-1 Location map of the Alafia River Watershed............... ................................ 91

4-2 Summary of mean monthly concentrations of total, dissolved reactive, and other
phosphorus forms during 1991-2009 in mainstem (Alafia and Bell Shoals),
developed (English Creek, Turkey Creek, and North Prong), and undeveloped (South
Prong and Fishhawk Creek) sub-basins of the Alafia River Watershed .......................... 92

4-3 Seasonal variation in mean monthly concentration of phosphorus forms during
1991-2009 ...... ................................... .......... .......... 93

4-4 Seasonal variation in contribution of organic, nitrate, and ammonium nitrogen to
total nitrogen during 1991-2009....................... ................................... 94

4-5 Long-term (1991-2009) trends in mean monthly total P concentrations in mainstem
(Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North
Prong), and undeveloped (South Prong and Fishhawk Creek) sub-basins of the Alafia
R iver W watershed ......................................... .. .................. ............... 95

4-6 Long-term (1991-2009) trends in mean monthly dissolved reactive P concentrations
in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and
North Prong), and undeveloped (South Prong and Fishhawk Creek) sub-basins of the
A lafia R iv er W atersh ed .................................................. ............................................... 9 6

4-7 Long-term (1991-2009) trends in mean monthly other phosphorus forms
concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek,
Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk
Creek) sub-basins of the Alafia River Watershed.................................................. 97

4-8 Relationship between percent urban and agricultural land use and phosphorus forms
in different su b -b a sin s. .................................................... ................................................. 9 8

4-9 Relationship between pasture and forest land use and phosphorus forms in streams
drain in g different su b -b asin s. ................................................................................................ 9 9









Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Thesis of Master of Science

NITORGEN AND PHOSPHORUS TRANSPORT IN AN URBAN WATERSHED

By

Kamaljit Kamaljit

August 2010

Chair: Gurpal Toor
Major: Soil and Water Science

Non-point source pollution is the dominant pathway of nitrogen (N) and phosphorus (P)

transport in agricultural, urbanized, and rapidly urbanizing watersheds. We used monthly

concentrations data of inorganic and organic forms of N and P in stream waters draining

different sub-basins, ranging in size from 19 to 350 km2, of the Alafia River Watershed (total

drainage area: 1085 km2), which ultimately drains to Tampa Bay Estuary, to understand N and P

transport. The sub-basins were classified based on the percentage of urban land use as three

developed (18-24% residential, 1-14% built up) and two undeveloped (3-11% residential, 1-3%

built up). Urban land use at two mainstem stations that drained 80-99% of the watershed was

16-17% residential and 3% built up. During 1991-2009, total N concentrations ranged from 0.8

to 2.4 mg L-1 and were greatest in stream waters draining developed (1.7-2.4 mg L-1) than

undeveloped (0.8-1.2 mg L1) sub-basins. Inorganic N (primarily NO3-N) was the dominant form

in streams draining developed sub-basins while organic N was greater in streams draining

undeveloped sub-basins. Total P concentrations ranged from 0.6 to 3.9 mg L-1 and were not

different among developed (0.8-3.9 mg L1) and undeveloped (0.6-0.9 mg L-1) sub-basins. Of

total P, 70-90% was dissolved reactive P while other P forms were 10-30% of total P in both

developed and undeveloped sub-basins. The increasing total N and decreasing total P









concentrations trends at the mainstem station draining 89% of the watershed over the 19-year

period suggests that the development of the watershed resulted in increasing N but not P

concentrations in streams. We suggest that the BMP's to reduce N loss from urban land uses in

three developed sub-basins (with total N of 1.7-2.4 mg L-1) may yield greater reductions in N

concentrations at watershed outlet (i.e. mainstem) to achieve EPA proposed numeric criteria of

total N concentration of 1.798 mg L-1. On the other hand, due to P rich geology and discharge

from the wastewaters, most developed and undeveloped sub-basins had greater total P

concentrations (0.8-3.9 mg P L-1) than EPA proposed numeric total P value of 0.739 mg L-1

indicating that BMP's should focus on reducing P loss from phosphate rock mined sub-basins

and reduce P inputs from wastewater.









CHAPTER 1
INTRODUCTION

Eutrophication

Eutrophication is a broad term used to describe enhanced phytoplankton growth in water

bodies such as lakes, rivers, reservoirs, and estuaries that receive excess nitrogen (N) and

phosphorus (P) from the landscape (Jansson and Dahlberg, 1999; Paerl, 2009). The

consequences of eutrophication include hypoxia, acidification of natural waters, degradation of

coastal waters including increased episodes of noxious algal blooms, and reductions in aquatic

macrophyte communities often leading to substantial shifts in ecosystem structure and function

(Carpenter et al., 1998; Dodds et al., 2009). In the US, eutrophication is one of the greatest

threats to the health of the estuaries. For example, Bricker et al. (1999) in their assessment of 138

estuaries reported that nearly 60% of estuaries exhibited moderate to severe eutrophic conditions.

In Florida, threshold concentrations of 1.20-1.79 mg N L-1 and 0.107-0.739 mg P L-1 have been

proposed for stream waters in three of four regions of Florida (EPA, 2010).

The phytoplankton growth in waterbodies is dependent upon the N: P ratio. For example,

total N: P ratio of 16:1 is suggested for optimum phytoplankton growth, termed as Redfield Ratio

(Redfield, 1934). An N: P ratio of<16:1 is indicative ofN limitation while >16:1 indicates P

limitation. In a review from 40 studies, Koerselman and Meuleman, (1996) reported that at an N:

P ratio of>16, P would be a limiting nutrient and at N:P <14, N would be limiting, and at

intermediate values (14-16) either N and/or P would be limiting nutrients for phytoplankton

growth. In the Tampa Bay estuary, monthly water quality concentrations data from 1981-2004

showed that the N: P ratio in the stream waters was about 5:1 suggesting that this water body is

N limited (Dixon et al., 2009). Further, it has been suggested that the loss of seagrass beds in the

Tampa Bay estuary is a direct consequence of N loading to the Bay from several point and non









point sources (TBEP, 2009). Therefore, source control on N needs greater attention than P, for

controlling eutrophication in the Tampa Bay. Although, recent research has suggested that

controls on both N and P transport might be needed to control eutrophication in freshwater-

marine continuum (Conley et al., 2009; Paerl, 2009).

Nitrogen and Phosphorus in the Environment

Nitrogen

The largest global pool of N exists as dinitrogen gas (N2) comprising up to 78% in the

lithosphere. However, only specialized microbes and cyanobacteria with the enzyme nitrogenase

can directly use N2 via N fixation, while for >99% of the organisms, N2 is made available by

inorganic N fertilizers using Haber-Bosch process where N2 is converted to ammonia (NH3).

Living organisms utilize inorganic N in the metabolic processes and convert it into organic N

(ON) forms such as amino acids, proteins, and nucleic acids. After the organisms die, micro-

organisms break down ON to ammonium (NH4+) which can be oxidized to NO3- via nitrification.

Finally, the denitrification process, in which micro-organisms oxidize organic matter using NO3s

as electron acceptor under reduced conditions close the N cycle by converting NO3- back into N2

(Galloway et al., 1996). Therefore, in different steps ofN cycle, NO3s, NH4 and ON forms are

either produced or consumed while the excess amount of these forms at each step has the

potential to be transported to waterbodies resulting in water quality deterioration.

Phosphorus

Like N, P in waterbodies exists in several combinations of organic and inorganic forms.

Haygarth and Sharpley (2000) suggested a physicochemical classification (i.e. filtration and

chemical methodology) to differentiate inorganic and organic P forms in water. According to this

classification, P can be divided into two main forms: dissolved (<0.45 [im) and particulate

(>0.45 pm). Dissolved P can be further divided into dissolved reactive P (DRP: orthophosphate)









and dissolved unreactive P (DUP: organic P forms such as sugar phosphates, mononucleotides,

DNA, RNA, and phospholipids). Similarly, particulate P can be divided into particulate reactive

P (PRP: P sorbed on sediments, Fe, Al, or Ca oxides) and particulate unreactive P (PUP: P

sorbed on mineral-humic acid complexes) (Toor et al., 2004). Dissolved and particulate P forms

in water bodies change from one form to another in response to a variety of environmental and

biological responses. For example, microbial decomposition or chemical desorption can convert

P from particulate to dissolved forms. Similarly organisms can take up dissolved P and transform

them into particulate P forms. As a result, P in the waterbodies is present in organic and

inorganic forms and is continually recycled.

Sources of Nitrogen and Phosphorus in Watersheds

Anthropogenic activities such as application of fertilizers, manures, industrial effluents,

and wastewater discharge are the major known sources of N and P in watersheds (Anisfeld et al.,

2007; Russell et al., 2008). Therefore, the inputs ofN and P are greater in human dominated land

uses (agricultural and urban) as compared to relatively undeveloped land areas such as natural

forests (Boyer et al., 2002; Kaushal et al., 2008; Russell et al., 2008).

Nutrient input sources can be divided into "point sources" such as wastewater and

industrial effluents and "non-point sources" such as runoff and leaching from urban and

agricultural areas. With the implementation of the Clean Water Act in the late 20th century, N

and P concentrations from point sources have been substantially reduced in the US (Howarth et

al., 2002). However, non-point source pollution is dominant in most of the watersheds in the US

and elsewhere. Non-point source pollution is difficult to control because the pollution sources

cannot be attributed to one particular discharge location but rather to a diffused landscape

(Rhodes et al., 2001). For example, since 1987, the United States Department of Agriculture

Conservation Reserve Program (CRP) has distributed $29.7 billion to agricultural land owners to









implement conservation practices to reduce soil loss, restore wetlands, and conserve forested

areas (USDA, 2006). However, these conservation measures showed a little evidence of

improvements in stream water quality at broad spatial scales as a greater emphasis of this

program was to reduce soil erosion to control nutrient losses, which was not successful in

controlling dissolved N losses (Boesch et al., 2001; Meals, 1996). Secondly, it was assumed that

all areas in the landscape contribute uniformly to nutrient loads, which resulted in less favorable

outcomes of reducing nutrient losses to water bodies.

Recent research has provided insights about contribution of various non-point sources to

nutrient loading in watersheds. For example, Poe et al. (2006) estimated that storm water runoff

contributes 63% of annual N loads in the Tampa Bay (Fig. 1-1). They further estimated that in

the storm water runoff, the residential areas were the major N contributors (20%) followed by

pasture/rangelands (15%), intensive agriculture (12%), and mining lands (6%). The second most

important source of N in Tampa Bay is atmospheric deposition (21%), while the contribution of

point sources such as domestic wastewater (9%) and industrial wastewaters (3%) is

comparatively lower than the non-point sources. The reduction of non-point source pollution is

urgently needed to protect and conserve water resources (USEPA, 2002). In case of P, such a

detailed analysis of various sources is lacking, however, it can be construed that the contribution

of different land uses to storm water runoff may be different, with higher contribution of P from

mined lands and wastewater discharges. In addition to different sources in the non-point

category, we also know that in each watershed, there are "hot spot" areas, termed as variable

source areas, which contribute a majority of nutrient losses (Poe et al., 2006; Diebel, 2009).

Therefore, a first step in controlling non-point source pollution is to develop a quantitative

understanding of their sources (i. e. hot spot areas) in the landscape followed by using best









management practices (BMPs) to control nutrient losses from these areas (Diebel et al., 2009;

Maxted et al., 2009). For example, Diebel et al. (2009) reported that targeting 10% watersheds in

Wisconsin, US decreased total P loads by 20% for the entire state. Therefore, the conservation

programs targeting the hot spot areas present an effective way to control nutrient losses and

improve water quality in a watershed while using less resources rather than attempting to use

BMP's for an entire watershed. Secondly, understanding how land uses impact nutrient losses

can help to unravel mechanisms of nutrient transport, which can lead to fine-tune BMP's to

reduce nutrient losses from land to water and protect water resources.

Research Objectives

In the Southern US, population is anticipated to increase from approximately 8 million in

1992 to 22 million in 2020 and 33 million in 2040 (Wear, 2002). Florida is one of the rapidly

developing states in the US and has serious water quality problems such as eutrophication of

coastal waters (Dame et al., 2002). Therefore, it is important to assess the impact of

anthropogenic activities on water quality of coastal waters. Very little is known about N and P

fate and transport in urban watersheds in Florida, which have a high proportion of sandy soils,

high ground water table, and altered hydrology due to storm water retention ponds. The Alafia

River Watershed (1085 km2) which drains into the Tampa Bay estuary was our study site to

understand the N and P transport as several years of historic water quality data was available.

Secondly, this watershed represents a typical urbanizing watershed in the region with diverse

mix of urban, agricultural, and mined land uses. The main objectives of this research are

presented below along with specific aims for each objective.









Objective 1. Evaluation of Nitrogen Concentrations in Different Sub-basins of the Alafia
River Watershed.

Aim la. Determine how different sub-basins influence concentrations of inorganic and

organic N forms.

Aim lb. Evaluate the influence of low (dry season) and high (wet season) flow conditions

on stream N concentrations in different sub-basins.

Aim Ic. Determine the long term trends of N concentrations in different sub-basins.

Objective 2. Evaluation of Phosphorus Concentrations in Different Sub-basins of the Alafia
River Watershed.

Aim 2a. Determine how different sub-basins influence concentrations of P forms.

Aim 2b. Evaluate the influence of low (dry season) flow and high (wet season) flow

conditions on stream P concentrations in different sub-basins.

Aim 2c. Determine the long term trends of P concentrations in different sub-basins.












3%
Groundwater
& Springs

9%
Municipal
Wastewater

3%
Industrial
Wastewater

1%
Accidental
Fertilizer Losses




















6%
Mining


21%
Atmospheric
Deposition


20%
Residential



1%
/Undeveloped
Land

9%
Commercial/
Industrial


15%
Pasture/Range
Lands


Figure 1-1. Estimated sources of nitrogen in the Tampa Bay (Adapted from Poe et al., 2006).









CHAPTER 2
STUDY SITE DESCRIPTION

Location

Alafia River Watershed is located in the central Florida and drains 1085 km2 of land area

(Fig. 2-1). The headwaters of the Alafia River originate from the swamp and prairie lands of

Polk County and extend 38.6 km long flowing west into lower Hillsborough Bay, ultimately

discharging into Tampa Bay Estuary (SWFWMD, 2007). The soils in the watershed are sandy,

with moderate to slow infiltration and are dominated by Myakka, Winder, Zolfo, Lake, and

Chandler soil groups (USDA, 2010).

Climate

The climate in the area is humid subtropical, with an annual mean temperature of 22.3 C.

Long term (1891-2009) annual average precipitation was 120 cm; -60% of precipitation

occurred during a four-month period from June to September while 40% of the rainfall occurred

during eight months period from October to May (Florida Climate Center, 2009). Therefore, a

water year is divided into wet season i.e. high flow conditions from June-September and dry

season i.e. low flow conditions from October-May.

Sub-basins of the Alafia River Watershed

Two mainstem stations namely Bell Shoals and Alafia drain 80 to 99% of the watershed

(Fig 2-1). Bell Shoals drains 89% of the watershed and includes discharge from five sub-basins

i.e. North Prong, South Prong, and English Creek, Turkey Creek, and Fishhawk Creek. While the

Alafia station drains 99% of the watershed and include discharges seven sub-basins including

five from Bell Shoals station and Bell Creek and Buckhorn Creek. We grouped the FLUCCS

codes at level IV into residential, built up, agricultural, pasture, mined, and forest land uses

(Table 2-2). The commercial, industrial, institutional, and transportation (such as roads) were









included under the built up land use. On the other hand, residential land use included the low,

medium, and high density residential. In this study, the residential and built up land uses were

considered as urban land use.

Overall, at two mainstem stations (Alafia and Bell Shoals), watershed land use was

dominated by mined (32-34%), followed by forest (19%), residential (16-17%), built up (3%),

pasture (11%), and agricultural (8-9%) (SWFWMD, 2007; Table 2-1). We grouped different

sub-basins of the Alafia River Watershed using percent residential land use into two categories:

1) three developed (18-24% residential land use) and 2) two undeveloped (3-11% residential

land use). A detailed description of sub-basins is given below.

Developed Sub-basins

Turkey Creek

In 2007, Turkey Creek had 20% residential, 3% built up, 24% agricultural, and 16%

pasture land use (Fig. 2-2; Table 2-1). Other land uses in the sub-basin include 12% forest, 17%

reclaimed, and 2% mined lands. Turkey Creek was under active mining operations during 1990,

however, all of the mining land use in the sub-basin was reclaimed by 2007 (Fig. 2-2). Other

significant land use changes in the sub-basin include a 9% increase (from 11% to 20%) in

residential and 9% decrease (from 25% to 16%) in pasture land use during 1990-2007. In

contrast to changes in mined and residential land use, the percent agricultural land use remained

similar at 22-24% in the sub-basin during 1990-2007. One domestic wastewater treatment plant

discharges 0.13 m3 sec-1 of wastewater with total N and total P concentration of 2.25 mg L-1 and

0.36 mg L-1, respectively in this sub-basin (NPDES, 2009).

English Creek

In 1990, the land use in the English Creek was 38% pasture, 19% agricultural, 10%

residential, 1% built up (Fig. 2-3). The sub-basin has undergone significant land use changes









during 1990-2007. For example, the residential land use in the sub-basin increased by 11%

(from 10% to 21%) followed by 13% increase (from 1% to 14%) in built up during 1990-2007.

The increase in residential and built up land use has occurred at the expense of pasture land use,

which has decreased by 24% from 38% in 1990 to 16% in 2007 (Fig. 2-3). During 1990-2007,

the forest and agricultural land use remained similar at 17-19% and 25-28%, respectively.

Florida Department of Health has identified 943 housing units on septic tanks in this sub-

basin (Florida Department of Health, 2009). In addition, several small animal feed additive

producing plants discharge wastewater in the sub-basin (NPDES, 2009).

North Prong

North Prong is the largest sub-basin draining 350 km2 of the Alafia River Watershed.

Significant changes in the residential, pasture, and mined land use have occurred in this sub-

basin during 1990-2007. In 1990, North Prong had 13% residential, 13% pasture, and 44%

mined land use (Fig. 2-4; Table 2-1). While in 2007, there was 39% mined, 18% residential, 6%

built up, 5% pasture, and 4% agricultural land uses (Fig. 2-4; Table 2-1). In this sub-basin, one

domestic wastewater treatment plant discharges -0.35 m3 sec-1 of wastewater with total N and P

concentrations of 1.19 mg L-1 and 3.1 mg L-1, respectively

(http://cfpub2.epa.gov/npdes/index.cfm).

Mined land use can be divided into four categories: active mine lands, reclaimed mined

lands, lands owned by mine interests that are yet to be mined, and lands owned by mine interests

that cannot be mined (SWFWMD, 2007). However, the Florida land use classification system

does not discriminate among these four mined land uses categories.









Undeveloped Sub-basins

South Prong

South Prong is the second largest sub-basin that drains 277 km2 of the Alafia River

Watershed. In 1990, 93% area in the South Prong sub-basin was under mining operations (Fig. 2-

5). However, reclamation of the mined land occurred during 1990-2007. For example, in 2007,

only 66% of the sub-basin was under mined land use (Fig. 2-5; Table 2-1) while the remainder

was forest (15%), pasture (9%), and agriculture (4%). In contrast to North Prong, South Prong is

less developed with residential land use of 3%. Several small industrial wastewater plants

(phosphate mines) discharge wastewater into South Prong during high rainfall events.

Fishhawk Creek

In 1990, Fishhawk Creek was undeveloped with 1% residential, 43% forest, 29% pasture,

and 16% mined land use (Fig. 2-6). All of the mined land in the sub-basin has been reclaimed

and comprised 10% of the sub-basin in 2007 (SWFWMD, 2007). During 1990-2007, residential

land use increased from 1% to 11% (Fig. 2-6; Table 2-1) while other land uses were 11%

agricultural and 23% pasture land.









1 Table 2-1. Station characteristics and associated land uses in the various sub-basins on the Alafia River Watershed
Sub-basin Station Sampling location Drainage area Land Use in 2007
Lat Long km2 % Residential Built up Agricultural Pasture Forest Mined


Alafia
Bell Shoals

English Creek
Turkey Creek
North Prong


Mainstem Stations
2301718 27.87 -82.32 1072 99 17
2301638 27.86 -82.26 974 89 16


-t
-t
2301000


South Prong 2301300
Fishhawk Creek t
2 -t USGS station not present


27.93
27.91
27.86

27.86
27.85


Developed
-82.06 99 9 2
-82.18 128 13 2
-82.13 350 32 1
Undeveloped
-82.13 277 26
-82.24 70.6 7 1









Table 2-2. Grouping of FLUCCS codes into major land uses (Source: SWFWMD, 2007).
Land Use Description FLUCCS Code
Residential Low density 1100
Medium density 1200
High density 1300
Built Up Commercial and services 1400
Industrial 1500
Institutional 1700
Transportation 8100
Communications 8200
Utilities 8300
Agriculture Row crops 2140
Tree crops 2200
Feeding operations 2300
Nurseries and vineyards 2400
Specialty farms 2500
Fish farms 2550
Other open lands 2600
Pasture Pastureland 2100
Forest Herbaceous 3100
Shrubs and brush land 3200
Mixed rangeland 3300
Upland coniferous forest 4100
Pine woodlands 4110
Upland hardwood forests 4200
Hardwood conifer mixed 4340
Tree plantations 4400
Wetland hardwood forests 6100
Cypress 6210
Wetland forested mixed 6300
Freshwater marshes 6410
Salt water marshes 6420
Wet prairies 6430
Emergent aquatic vegetation 6440
Extractive 1600
Mined Reclaimed lands 1650
Reclaimed Recreational 1800
Recreational Golf courses 1820
Open lands 1900
Streams and waterways 5100
Other Lakes and Reservoirs 5200, 5300
Bays and estuaries 5400
FLUCCS: Florida Land Use and Cover Classification System











LAND USE ... .


O 2 4 8 12 16 South Prong
km Tr I I / k


Alafia River Watershed


Figure 2-1. Location map and land uses in the various sub-basins of the Alafia River Watershed.






















































LAND USE
RESIDENTIAL EXTRACTIVE
BUILT UP m RECLAIMED


0 1 2 4 6 8
km


RECREATIONAL ACTIVE AGRICULTURE
PASTURELAND FOREST


Figure 2-2. Land use in the Turkey Creek sub-basin in 1990, 1999, and 2007.


1999


2007


OTHER









1990


location


0 1 2 4 6 8
km


RESIDENTIAL M EXTRACTIVE
BUILTUP RECREATIONAL


PASTURELAND
M ACTIVE AGRICULTURE


Figure 2-3. Land use in the English Creek sub-basin in 1990, 1999, and 2007.


N

+


LAND USE


M FOREST
OTHER












1990


0 2.5 5 10 15 20
km


LAND USE


RESIDENTIAL
BUILT UP


EXTRACTIVE
RECLAIMED


RECREATIONAL = ACTIVE AGRICULTURE
PASTURELAND FOREST


Figure 2-4. Land use in the North Prong sub-basin in 1990, 1999, and 2007.


1999


OTHER









































Sampling location









N


3 8 12 16


RESIDENTIAL EXTRACTIVE
BUILT UP RECLAIMED


RECREATIONAL
PASTURELAND


i ACTIVE AGRICULTURE
i FOREST


Figure 2-5. Land use in the South Prong sub-basin in 1990, 1999, and 2007.


1990















1999















2007


LAND USE


0 2


Si km


OTHER











1990 1999



















2007






Sampling locations












0 1 2 4 6 8
Land Use n
RESIDENTIAL RECLAIMED PASTURELAND FOREST OTHER
BUILT UP RECREATIONAL ACTIVE AGRICULTURE EXTRACTIVE


Figure 2-6. Land use in the Fishhawk Creek sub-basin in 1990, 1999, and 2007.









CHAPTER 3
NITROGEN TRANSPORT IN AN URBAN WATERSHED

Abstract

Eutrophication affects freshwater, estuarine, and marine ecosystems worldwide. Many

nitrogen (N) limiting water bodies such as Tampa Bay estuary show accelerated eutrophication

due to increase in nitrogen (N) concentrations. Understanding the processes and mechanisms

controlling the N transport in a watershed are essential to devise strategies to prevent coastal

eutrophication. In this study, we used 5 to 19 years of monthly stream-water concentrations data

of inorganic and organic forms of N for seven sub-basins (19-350 km2) of the Alafia River

Watershed (total drainage area: 1085 km2). Results showed that total N in stream waters ranged

from 0.84 to 2.43 mg L-1, was greater in the developed (23-35% urban land use) than

undeveloped (3-14% urban land use) sub-basins. Total N was significantly (P<0.05) correlated

with urban (r=0.83) but not with agricultural (r=0.49) land uses thereby indicating that the losses

of N are greater from urban areas. Among N forms, organic N was dominant in undeveloped

(66-71% of total N) than developed (30-44% total N) sub-basins while NO3-N was dominant in

developed (53-68% of total N) than undeveloped (25-30% of total N) sub-basins.

Concentrations of NO3-N were lower in wet season (June-September) compared with dry season

(October- May) due to higher rainfall-runoff in wet season that may have caused dilution of

NO3-N. In contrast, greater concentration of organic N in wet season is possibly due to the

greater transport of organic materials such as leaves and plant residues in high rainfall-runoff

events that occurred during wet season. Trend analysis indicated that total N concentrations

increased by 1.01% per year at the most downstream station during 1991-2009 and this increase

was primarily due to the increased concentrations of organic N. Total N at two mainstem stations

was 1.77-1.91 mg L-1, which is close to the EPA proposed numeric total N of 1.79 mg L-1.









Greater total N concentrations in three developed sub-basins (1.7-2.4 mg L1) suggest that

targeting best management practices in these three urban sub-basins may be a resource efficient

way to reduce N concentrations in this urban watershed.

Introduction

Nitrogen is a limiting nutrient for eutrophication in many coastal waterbodies in the US

and elsewhere, including the Tampa Bay Estuary in Florida (USEPA, 2002; Dixon, 2009; TBEP,

2009). Major sources of N in water bodies are from point such as wastewater and industrial

effluents and non-point such as runoff and leaching from urban and agricultural lands. Point

source contributions have substantially reduced in the US following passage of Clean Water Act

in early 1970 (Howarth et al., 2002). While non-point source N pollution is the major contributor

ofN loads in water bodies because of its diffuse nature (Diebel et al., 2009; Maxted et al., 2009).

Major non-point sources in urban watersheds can be fertilized urban lawns, septic tanks, and pet

waste (Brett et al., 2005; Rhodes et al., 2001). A logical approach to control non-point N

pollution may be to develop an understanding of major N sources and hot spot areas in a

watershed. This can be achieved by determining the N forms in streams that drain sub-basins

with different land uses.

Different land uses in a watershed have been shown to receive different N inputs (Boyer et

al., 2002; Kaushal et al., 2008). For example, in Northeastern watersheds of the US, Boyer et al.

(2002) reported that the N inputs were positively correlated with the percentage of agricultural

and urban land uses (r=+0.96) and negatively correlated with the forest land use (r =-0.77).

Similarly, the net anthropogenic N inputs were six-times greater in agricultural (77 kg ha-1) and

two-times greater in suburban (25 kg ha-1) than the forest (11.2 kg ha-1) watersheds in Baltimore,

US (Groffman et al., 2004). Losses of N in watersheds depend upon N inputs and therefore the

anthropogenically influenced agricultural or urban watersheds result in greater N losses than









forested watersheds (Boyer et al., 2002; Han et al., 2009). Groffman et al. (2004) reported that

total N was greater in streams draining agricultural (4-5 mg L-1) and sub-urban (1-4 mg L-1) as

compared to the forested watersheds (< 0.25 mg L-1). Similarly, concentrations of total N were

greater in streams draining urban (1.5 mg L1) than forest (1.1 mg L1) watersheds in Seattle, US

(Brett et al., 2005). Vidon et al. (2008) observed total N concentrations of 6.2-9.4 mg L-1 in

intensively agricultural (corn-soybean) dominated watersheds in Indiana (Vidon et al., 2008).

These and other studies have shown that total N concentrations are higher in streams that drain

agricultural and urban watersheds as compared with forested watershed, primarily due to the

greater inputs of N in the former than the later watersheds.

Land uses in a watershed also influence the forms of N lost (Scott et al., 2007). To

understand the terrestrial N cycling and transport across spatial and temporal scales in

watersheds with different land uses, understanding of different N forms in streams that drain

different land uses is essential (Pellerin et al., 2006). Organic N (ON) forms have been shown to

be the dominant forms in less developed watersheds such as forested or mixed cover than more

developed watersheds such as intensive agriculture. Stanley and Maxted (2008) reported that ON

was 68% (<0.3 mg L1) of total dissolved N (NO3-N + NH4-N + dissolved ON) in 84 Wisconsin

streams. While in more anthropogenically influenced watersheds (urban and agricultural),

inorganic N forms such as NO3-N were dominant (Pellerin et al., 2006; Stanley and Maxted,

2008). For example, increased agriculture and urban land uses in the Mississippi River basin

have been shown to double N concentration from 1.1 mg L1 in 1902 to 2.2 mg L1 in 1999 and

most of this increase was in the NO3-N, which increased from <55% (total N: 1.1 mg L-1) in

1902 to >75% (total N: 2.2 mg L1) in 1999 (Goolsby and Battaglin, 2001; Turner and Rabalais,

2003). In a study on N budgets in 348 watersheds with varying land uses in North America,









Pellerin et al. (2006) reported that of the total dissolved N, ON was 35% in urban and 55% in

forest streams. These studies provide evidence of land use specific controls on the fate and

transport of different N forms.

Although the percent ON is lower in streams draining agricultural and urban land uses, the

concentrations of ON may increase with anthropogenic activities. For example, Pellerin et al.

(2006) observed that dissolved ON concentrations were 2-4 times greater in urban and

agricultural than forest watersheds. In two watersheds dominated by pastures in Japan,

Hayakawa et al. (2006) reported that urban land use of 0.2-4.3% was significantly correlated

(r=+0.52) with dissolved ON. High concentrations of ON in anthropogenically influenced

watersheds could be due to the inputs of organic N rich sources such as manures, composts, and

other organic amendments in crops and urban lawns, contribution of organic N from onsite septic

systems, and direct discharge of sewage effluents in streams (Kroeger et al., 2006; Scott et al.,

2007). It is also plausible that NO3-N in urban and agricultural watersheds is first taken up by

microbes and vegetation, which on decomposition releases N in the ON forms, which can be

subsequently transported to waters. Long-term N fertilization studies at the Harvard forest in

Massachusetts have shown that dissolved ON concentrations in the forest floor increased as a

result of elevated inorganic N deposition thereby representing the conversion of inorganic N to

ON (McDowell et al., 2004). Wetlands and artificial retention ponds in many urban landscapes

are common features, designed to reduce flooding events, which may stimulate additional

reinforcing processes such as plant or microbial uptake and can denitrify NO3-N resulting in

reduced transport ofNO3-N (Groffman et al., 2004; Pellerin et al., 2004).

Another consequence of land use change primarily in urbanizing watersheds is the

alteration of physical and hydrologic features that increase impervious area in the form of









rooftops, roadways, parking lots, sidewalks, and driveways (Carpenter et al., 1998; Tufford et al.,

1998). The increased impervious area in urban watersheds leads to increased runoff due to

altered and short flow paths compared to natural systems (Arnold and Gibbons, 1996; Lee and

Heaney, 2003; Paul and Meyer, 2001). In addition, the loss of natural vegetation minimizes

recycling and uptake ofN due to reduced microbial and vegetative processes that immobilize N

in the forest canopy, litter, soils, and organic matter (Abelho, 2001; Wahl et al., 1997). Together,

increased flow and reduced vegetation decrease the residence time of water and N in the urban

watersheds, which decreases N processing such as denitrification and thus make more N

available for transport (Alexander et al., 2000; Green et al., 2004; Peterson et al., 2001).

Therefore, the watersheds with greater development and modified hydrology are likely to export

a higher amount of N to water bodies than less developed watersheds.

The Alafia River Watershed is an example of such a watershed, where land use has been

continuously changing from natural areas to urban lands for the last 30+ years. As a result,

several sub-basins of the watershed have modified hydrology-due to stormwater retention ponds

and water convergence structures-to drain excess water during high flow events to avoid

flooding. Two sub-basins of the watershed are dominant in mined land where phosphate mining

is a commercial enterprise. We hypothesize that the 1) concentration of N forms may be different

in sub-basins with diverse land uses and 2) the urbanization may increase the N concentrations in

streams of the Alafia River Watershed. The objectives of this study were to (1) determine how

different sub-basins (with different land uses) influence concentrations of different forms of N in

stream waters; (2) investigate the influence of two distinct seasons (dry and wet) on

concentrations of N forms in stream waters, and (3) evaluate the long term trends of N

concentrations in different sub-basins.









Materials and Methods

Study Site Description

Refer to Chapter 2 for detailed description of sub-basins of the Alafia River Watershed.

Data Collection

Monthly concentrations data from 1991 to 2009 of total Kjeldahl N (TKN), NO3-N, and

NH4-N were obtained from the Environmental Protection Commission of Hillsborough County

(EPCHC) water quality database (http://www.epchc.org). Among the two mainstem stations,

water quality data were available only for the Bell Shoals station for 19 years (1991-2009),

while for the Alafia station the data were available for 9 years (1999-2009) (Table 3-1). Among

other sub-basins in the watershed, 19 years of water quality data were available only for North

Prong, South Prong, and Turkey Creek while the data availability for English Creek was from

1999 to 2009 and for Fishhawk Creek from 2005 to 2009.

Stream-water Collection and Analysis

Each month, a grab sample from surface water was collected from the channel thalweg

(center of the stream flow) in plastic water bottles by EPCHC staff as per the surface water

quality collection standard operating procedures. In brief, the water bottles were rinsed three

times with the stream water before collection of the samples. The collected samples were chilled

with ice and transported to the laboratory where samples were stored at 4 C prior to analysis.

Environmental Protection Commission Hillsborough County staff analyzed the surface water

samples for TKN (EPA 350.2 method), NO2-N+NO3-N (SM 4500 NO3), and NH4-N (EPA

350.1) using a discrete analyzer. Other N fractions were calculated as follows: total N = NO2-

N+NO3-N + TKN; organic N (ON) = TKN-NH4.









Statistical Analysis

Mean, median, and standard deviations ofN forms were calculated using MS Excel 2007.

The GLIMMIX time series procedure in SAS Version 9.1 (SAS Institute, Cary, NC) was used to

compare the N forms across different sub-basins as well as to determine the long term trends in

N forms. The GLIMMIX procedure fits models to data with correlations or non constant

variability and assumes normal random effects (SAS Institute, 2008). In this time series analysis,

two kinds of effects were used: (1) fixed effects which included stream station and season of the

year and (2) random effects which was the date of sample collection (monthly intervals). For first

fixed effect (stream station), each station was treated as a group and the variance was pooled

over the sites. Similarly for the second fixed effect (season), each wet (June-September) and dry

(October-May) season at each station was treated as one group of observations. The date of

sample collection from 1991-2009 for the each station was considered as random effect and

therefore a time series was constructed. With this, we compared (1) the difference in N forms at

each of the station averaged over all sampling dates and (2) differences in pooled mean

concentrations ofN forms during wet and dry seasons at each station using P<0.05 as

significance level. In addition, same time series with fixed and random effects were used to

determine the trends in N forms over the period of study. The concentrations of different forms

ofN were logarithmically transformed to equalize variances and normalize skewed data and

were back transformed to present means ofN forms in more relevant manner.

Various statistical techniques have been used to determine the long term water quality

trends (Goodrich et al., 2009; Johnson et al., 2009; Richards et al., 2008). Most common

methods include non parametric Mann-Kendall and Seasonal Kendall for determination of long

term trends in water quality (Daroub et al., 2009; Johnson et al., 2009). Only a few studies have

used GLIMMIX procedure (e.g., Goodrich et al., 2009) and no studies have attempted to use a









mix of more than one statistical procedure. Therefore, in addition to GLIMMIX procedure, we

determined the long term trends using Seasonal Kendall procedure. In Seasonal Kendall, trend is

calculated by comparing all potential data pairs. If the later value in the pair (in time) is higher

than the first, a plus sign is scored. If the later value in the pair is lower than the first, a minus

sign is scored. If the results find an equal number of pluses and minuses, then there is no

discernible trend. In other words, there isjust as much of a likelihood that a pair of data values

will be higher (or lower) than the next one. If the results show more pluses than minuses, this

would indicate that a positive trend is likely (Hirsch et al., 1982).

The trend in both GLIMMIX and Seasonal Kendall procedures is defined by the rate of

change in the concentration of a constituent over time, which is referred to as the trend slope.

The trend slope can be expressed either as change in original units per year [S (0)], or as a

percent of the mean concentration of water quality variable [S (%)]. The former is the median

slope of all pair wise comparisons (each pair wise difference is divided by the number of years

separating the pair of observations) while the latter is produced by dividing the slope (in original

units per year) by the mean and multiplying by 100.

The N forms concentrations data were divided into two time periods: 1991-2000 and

2001-2009. To determine the effect of land use change on N forms in corresponding sub-basins,

the land use data of 1999 and 2007 was correlated with N concentrations data of 1991-2000 and

2001-2009, respectively using SAS PROC CORR procedure.

Results

Chemical Characteristics of Stream Waters

During 1991-2009, mean monthly temperature in stream waters was 22-23 C (Fig. 3-2).

The concentration of DO in stream waters was lowest at Alafia (3.1 mg L-1) than Bell Shoals (6.4

mg L-1) and other sub-basins (5.3-7.3 mg L-1) of the Alafia River Watershed. Among the sub-









basins, the DO concentration was significantly (P< 0.05) lower at one developed (English Creek;

5.3 mg L-1) than other sub-basins. Mean pH of the stream waters was similar (7.2-7.5) at all the

stations of the Alafia River Watershed (Fig. 3-2). In contrast to pH, the mean EC of the stream

waters was atleast one order of magnitude greater at one mainstem station near to the Tampa Bay

(Alafia; 7548 iS cm-1) than Bell Shoals (443 iS cm-1) and other sub-basins (214-556 iS cm-1).

Among the sub-basins, EC was significantly lower at one undeveloped sub-basin (Fishhawk

Creek: 214 iS cm-1) than other developed and undeveloped sub-basins.

Concentrations of Nitrogen Forms in Streams Draining Different Sub-basins

During 1991-1999, mean monthly total N concentration at the mainstem station (Bell

Shoals) that drains 89% of the watershed was 1.65 mg L-1 (Fig. 3-3A). Total N was significantly

(P< 0.05) greater at two developed (1.45-1.74 mg L-1) than one undeveloped sub-basin (1.06 mg

L-). Among the two developed sub-basins, Turkey Creek had significantly (P< 0.05) greater

total N than North Prong.

Total N concentrations were greater during 2000-2009 than 1991-1999 at one mainstem

(Bell Shoals), two developed (North Prong and Turkey Creek), and one undeveloped (South

Prong) sub-basins of the Alafia River Watershed (Fig. 3-3A). Mean monthly total N at two

mainstem stations (Alafia and Bell Shoals; urban land use 19-20%) was 1.77-1.91 mg L1 (Fig.

3-3A). Total N concentrations in stream waters were significantly (P<0.05) greater in all the

developed sub-basins (English Creek, Turkey Creek, and North Prong; 1.67-2.43 mg L-1) than

undeveloped sub-basins (South Prong and Fishhawk Creek; 0.84-1.21 mg L-1). Differences in

total N concentrations in streams among developed and undeveloped sub-basins were observed.

For example, among the developed sub-basins, English Creek had significantly greater total N

(2.43 mg L-1) than Turkey Creek (1.96 mg L1) and North Prong (1.67 mg L-1). Similarly in the









undeveloped sub-basins, total N was significantly higher in South Prong (1.21 mg L-1) as

compared to Fishhawk Creek (0.84 mg L-1).

During 1991-1999, mean monthly ON concentration at the mainstem station (Bell Shoals)

was 0.49 mg L-1 (30% of total N) (Fig. 3-3B). The ON concentration in the stream waters was

similar in all the sub-basins (0.57-0.63 mg L-1). In contrast to the concentrations, the percent of

ON was greater in undeveloped (53% of total N) than two developed (36-41% of total N) sub-

basins.

Concentrations of ON were greater during 2000-2009 than 1991-1999 in streams draining

mainstem, two developed, and one undeveloped sub-basins. Among the mainstem stations, mean

monthly concentration of ON was significantly (P< 0.05) greater at the most downstream Alafia

(1.02 mg L-1; 58% of total N) as compared to next upstream Bell Shoals (0.75 mg L-1; 39% of

total N) (Fig.3-3B). All the developed and undeveloped sub-basins had a narrow range of stream

ON (0.73-0.77 mg L-) except for a significant lower concentration at Fishhawk Creek (0.60 mg

L-1). However, the percent of ON was much greater in streams draining undeveloped (66-71% of

total N) than developed (30-44% of total N) sub-basins.

At the mainstem station (Bell Shoals), the NO3-N was the dominant N form (1.13 mg L-1;

69% of total N) during 1991-1999 (Fig. 3-3C). In contrast to ON, NO3-N concentrations were

two folds greater in two developed (0.82-1.07 mg L-1; 57-62% of total N) than one undeveloped

sub-basin (0.47 mg L-1; 45% of total N). Among the two developed sub-basins, Turkey Creek

had significantly (P< 0.05) greater NO3-N than North Prong.

The concentrations of NO3-N were similar during 1991-1999 and 2000-2009 in all the

streams draining different sub-basins (Fig. 3-3C). Concentrations of NO3-N at Bell shoals were

significantly (P< 0.05) greater (1.14 mg L-1; 60% of total N) than the most downstream Alafia









(0.67 mg L-1; 38% of total N). Among the sub-basins, NO3-N followed the total N pattern and

was significantly greater in developed (0.89-1.66 mg L-1; 53-68% of total N) than undeveloped

(0.21-0.37 mg L-1; 25-30% of total N) sub-basins.

At all study stations, mean monthly concentrations of NH4-N were much lower (<0.1 mg

L-1; 2-5% of total N) than both ON and NO3-N during 1991-1999 and 2000-2009 (Fig. 3-3D).

Seasonal Variations in Chemical Characteristics in Stream Waters Draining Different Sub-
basins

At two mainstem stations (Alafia and Bell Shoals), the mean temperature of the stream

waters was significantly (P< 0.05) greater in wet (26-27 C) than dry season (20-21 C) (Fig. 3-

4). A similar increase in temperature of stream waters in wet (26-27 C) than dry (18-20 C)

seasons was found in developed and undeveloped sub-basins. The seasonal variation showed

opposite trends in DO concentrations than temperature of the stream waters possibly due to

greater biological activity in stream waters with greater temperature. For example, at two

mainstem stations, the DO concentration of the stream waters was greater in dry (3.4-6.8 mg L-1)

than wet season (2.7-5.9 mg L-1). Among the sub-basins, the concentration of DO in stream

waters was 5.6-8.1 mg L-1 in dry season and decreased to 4.6-6.1 mg L-1 in wet season (Fig. 3-

4). This decrease in DO concentration of the stream waters was significant (P< 0.05) in all the

sub-basins except for English Creek.

At two mainstem stations, the pH of stream waters was greater but not significantly (P<

0.05) in dry than wet seasons (Fig. 3-4). Among sub-basins, the pH of the stream waters was

greater in dry season (7.4-7.6) and decreased in wet season (6.9-7.3); however the decrease was

significant (P<0.05) only at Fishhawk Creek. Similarly, the EC of the stream waters was greater

but not significantly (P<0.05) different in dry than wet season in all study basins.









Seasonal Variations in Concentrations of Nitrogen Forms in Streams Draining Different
Sub-basins

At two mainstem stations (Alafia and Bell Shoals), the mean total N was slightly higher,

but not significant (P< 0.05), in dry (1.83-1.84 mg L-1) than wet (1.61-1.69 mg L-1) season (Fig.

3-5A). All the developed sub-basins (Turkey Creek, English Creek, and North Prong) had greater

total N in dry than wet season while the undeveloped sub-basins had greater total N in wet than

dry season (Fig. 3-5A). However, the seasonal variation was significant (P< 0.05) only in

Fishhawk Creek and English Creek.

Mean ON concentrations were lower in dry season and increased in wet season at all sites

(Fig. 3-5B; Fig. 3-6). The magnitude of this increase was greater at Bell Shoals (from 0.52 to

0.82 mg L-1) as compared to Alafia (from 0.99 to 1.03 mg L-1). Among the developed sub-basins,

the increase in ON concentration during wet season was greatest at English Creek (from 0.57 to

1.03 mg L-1) than Turkey Creek and North Prong (from 0.60-0.64 to 0.80-0.83 mg L-1). In

undeveloped sub-basins, ON was 0.48-0.58 mg L-1 in dry season which increased to 0.83-0.88

mg L-1 in wet season.

In contrast, NO3-N was lower in wet than dry season at all sites (Fig. 3-5C; Fig. 3-6). For

example, at two mainstem stations, NO3-N in dry season was 0.75-1.30 mg L-1, which decreased

to 0.51-0.84 mg L-1 in the wet season. Similarly, among developed sub-basins, the decrease in

NO3-N in wet season compared to dry season was significant at English Creek (from 2.07 to 0.91

mg L-1) but not at Turkey Creek (from 1.24 to 0.85 mg L-1) and North Prong (from 0.98 to 0.65

mg L-1). In undeveloped sub-basins, the decrease in NO3-N was significant only at Fishhawk

Creek (from 0.23 to 0.17 mg L-1).









Concentrations ofNH4-N were not significantly different among dry and wet seasons at all

study stations (Fig. 3-5D; Fig. 3-6). But, NH4-N was greater at the most downstream station

(Alafia) than other sites in both wet and dry seasons.

Long Term Trends in Flow Un-weighted Nitrogen Forms in Streams Draining Different
Sub-basins

Total Nitrogen

The longest data record (19 years) was only available for one mainstem station (Bell

Shoals) that drains 89% of the watershed, two developed (Turkey Creek and North Prong), and

one undeveloped (South Prong) sub-basin (Table 3-7). According to Seasonal Kendall trend

analysis, none of the total N concentrations trends were significant (P< 0.05) for these four sub-

basins during the study period. However, the concentrations of total N were increasing at Bell

Shoals (+0.17% per year; 2.79 [g L-1 per year) (Fig. 3-7). Among the sub-basins, total N showed

an increasing trend +1.7% or 18.9 pg L-1 per year at South Prong and +1.5% or 29.1 g L-1 per

year at Turkey Creek and a decreasing trend at North Prong (-0.42% or 6.7 pg L-1 per year).

Among the stations with 10 years of data record, the mainstem station Alafia and developed

English Creek showed decreasing but not significant (P< 0.05) total N trends.

The total N concentration trends estimated with the GLIMMIX procedure were similar to

Seasonal Kendall but the magnitude of trend slopes was variable. The GLIMMIX model showed

a significant (P<0.09) overall increasing total N trend of+1.01% per year at the mainstem Bell

Shoals station which equates to approximately 20% increase (0.33 mg L-1) in the total N over 19-

years (Fig. 3-7). Total N significantly increased at Turkey Creek (+1.02% per year; P< 0.005)

and South Prong (+1.07% per year; P< 0.03) but not at North Prong (-1.11% per year; P< 0.64)

during 1991-2009. The study stations with 10 years of data record (Alafia and English Creek)

showed decreasing but not significant (P< 0.05) total N concentration trends.









Organic Nitrogen

Seasonal Kendall trend analysis showed that the ON concentrations were increasing at all

the four study stations with 19-years of data record (Fig. 3-8). At Bell Shoals, total N showed an

increasing but not significant (P< 0.45) trend (+1.8% per year) which equates to 35% increase in

mean ON concentration during 1991-2009. The increasing ON trends were significant (P<

0.001) at Turkey Creek (+2.5% per year) but not at North Prong (+0.72% per year; P< 0.46) and

South Prong (+1.95% per year; P< 0.25). None of the trends were significant in two sub-basins

with 10 years of data record.

The GLIMMIX procedure showed increasing ON concentration trends at four sub-basins

with 19 years of data record (Fig. 3-8). Overall, at Bell Shoals the ON concentration trend was

(+0.94% per year; P= 0.001) which equates to approximately 19% increase in ON concentration

over 19-years. Among the sub-basins, the ON concentrations significantly (P< 0.001) increased

at Turkey Creek (+0.94% per year), South Prong (+0.95% per year) but not at North Prong

(+1.00% per year). During 1999-2009, Alafia station showed decreasing (-2.45% per year) as

compared to increasing trends at English Creek (+0.65% per year).

Nitrate Nitrogen

As per Seasonal Kendall trend analysis, none of the NO3-N concentrations trends were

significant (P< 0.05), however, the NO3-N decreased at Bell Shoals (-0.27% per year) during

1991-2009 (Fig. 3-9). Among the sub-basins, the NO3-N concentration trend was decreasing at

North Prong (-0.1% per year) and South Prong (-2.7% per year) but not at Turkey Creek

(+0.18% per year) during 1991-2009. During 1999-2009, Alafia station showed an increasing

NO3-N concentration trend (+5.4% per year).

Using GLIMMIX procedure, none of the NO3-N concentrations trend was significant (P<

0.05); however the overall trend at Bell Shoals was decreasing (-1.04% per year) during 1991-









2009 (Fig. 3-9). Turkey Creek showed an increasing NO3-N concentration trend (+1.13% per

year) in contrast to decreasing trends at North Prong (-0.31% per year) and South Prong (-

0.98% per year). During 1999-2009, the NO3-N concentration trends were not significant at

Alafia (+0.96% per year) and English Creek (-1.74% per year).

Ammonium Nitrogen

Both the Seasonal Kendall trend analysis and GLIMMIX procedure showed non-

significant trends NH4-N at all the sub-basins of the Alafia River Watershed (Fig. 3-10). The

concentrations of NH4-N remained lower (<0.4 mg L1) during the study period at all the sub-

basins.

Long Term Trends in Flow Weighted Nitrogen Forms in Streams Draining at Mainstem
Station

At the mainstem station that drains 89% of the Alafia River Watershed showed increasing

but not significant (P< 0.241) total N concentration trends during 1991-2009 (Table 3-2). Mean

concentrations of total N were greater during 2000-2009 (1.82 mg L1) than 1991-1999 (1.53 mg

L-1) reflecting the increasing GLIMMIX trends in terms of actual increase in concentrations. The

ON concentrations showed significant (P< 0.01) increasing trends of+1.05% per year (+7.0 [g

L-1 per year; 0. 14 mg L-1 in 19 years). Similar to total N, mean ON was greater during 2000-

2009 (0.81 mg L-U) than 1991-1999 (0.55 mg L-1). In contrast to ON, flow weighted NO3-N

concentration trends were not significantly (P< 0.487) different for 1991-1999 (0.94 mg L1) and

2000-2009 (0.97 mg L1).

Long Term Trends in Nitrogen Loads at Mainstem Station

At mainstem station (Bell Shoals), total N loads were not significant in wet and dry

seasons during 1991-2009 (Table 3-2). The non significant trends in the N loads during wet and









dry seasons could be due to greater variations in N loads and reduced number of data points in

wet and dry seasons.

Relationship between Land Use and Nitrogen Forms

Total N concentrations were significantly (P<0.05) correlated with the percent urban

(residential + built up) land use (r=+0.83) in the watershed (Fig. 3-11). This positive correlation

was probably due to the influence of urban land use on NO3-N (r=+0.78; P<0.05) rather than ON

(r=+0.28) concentrations during the study period. The percent agricultural land use was

positively but not significantly (P<0.05) correlated with total N (r= +0.49) and NO3-N (r=

+0.53). The pasture and forest land uses were not significantly correlated with any of the N

forms (r- <0.30).

Discussion

Influence of Land Uses on Total Nitrogen Concentrations in Stream Waters

Mean monthly concentrations of total N at two mainstem stations (1.7-1.9 mg L-1)

approached the EPA's proposed numeric total N value of 1.79 mg L-1 for the region (EPA,

2010). These total N concentrations were greater than reported for Hillsborough County lakes

value of 1.12 mg L-1 (Florida Lakewatch Program, 2005). However, total N was much lower

than corn-soybean dominated watersheds (6.2-9.4 mg L-1) in Indiana, US (Vidon et al., 2008).

Lower total N concentrations in the Alafia River Watershed may be to the low agricultural land

use (8%) while watersheds in Indiana had more than 60% of row crops land use. In our study,

total N concentrations were similar to values reported for the urban and forest watersheds (1.1-

1.5 mg L-) in Seattle, US (Brett et al., 2005).

The three developed sub-basins (23-35% urban land use) had greater total N (1.67-2.43

mg L1) than undeveloped sub-basins (3-14% urban land use) (0.84-1.21 mg L-1). Total N

concentrations were significantly (P< 0.05) positively correlated with urban land use (r= +0.83)









(Fig. 3-11). While total N was positively but not significantly (P< 0.05) correlated with

agricultural (r=+0.49) and no relationship was observed between total N and forest and pasture

land uses (r= <0.30). This indicates that the N concentrations significantly increase with

urbanization of the watershed. The strong correlation between percent urban land use and NO3-N

concentrations (r=+0.78) may be due to the runoff of fertilizers containing N from

urban/residential lawns. Overall, greater concentrations of total N in urban than forest and

pasture land uses may be due to the higher N inputs in the former than later land uses (Boyer et

al., 2002; Groffman et al., 2004). Furthermore, in urban lands impervious areas result in greater

transport of N (Alexander et al., 2000; Green et al., 2004; Peterson et al., 2001) resulting in

higher total N concentrations in developed than undeveloped sub-basin streams.

Among the three developed sub-basins, the total N concentrations were significantly

(P<0.05) different during 2000-2009. Greater total N concentrations at English Creek than

Turkey Creek and North Prong could be due to the greater urban land use (35%) in English

Creek than North Prong and Turkey Creek (23-24%). A wastewater treatment plant discharges

0.11 m3 sec-1 wastewater with total N concentration of 2.76 mg L-1 into the English Creek

(SWFWMD, 2007). The total N concentrations in developed sub-basin vary inversely with size

of the sub-basin. The length of the streams increases with the size of the sub-basin, which

provides opportunities for the N removal processes such as denitrification in riparian buffer

zones as well as within the stream resulting in lower concentrations of N (Alexander et al., 2000;

Seitzinger et al., 2002; Alexander et al., 2009). These factors might have resulted in significant

differences in N concentrations in three developed sub-basins. On the other hand, lower total N

concentration in Fishhawk Creek that had 14% urban land (0.84 mg L-1) than South Prong that

had 3% urban land (1.21 mg L-1) could be because of greater land area in forest in the former









(32%) than later (15%) sub-basin (Table 3-1). This highlights the importance of other land uses

such as forest in addition to urban and agricultural land use in modifying the net total N losses in

sub-basins.

Land Uses and Forms of Nitrogen Concentrations in Stream Waters

In our study, NO3-N was 25-30% of total N in undeveloped and 53-68% of total N in

developed sub-basins. The percent urban land use was strongly correlated with NO3-N

concentrations (r=+0.78) but not with ON concentrations (r=+0.28). This suggested that the

increase in total N concentrations in urban land uses is primarily driven by NO3-N probably due

to the runoff losses of the N fertilizers applied to the urban lawns. Similarly, Stanley and Maxted

(2008) in 84 streams in Wisconsin reported that NO3-N was 75% of total dissolved N in urban

compared to 30% in the forested streams. In a review on 348 watersheds across North America,

Pellerin et al. (2006), reported that NO3-N was 65% of total N in urban (>50% urban land use)

and 35% of total N in forest (>90% forest land use) watersheds.

In contrast to decreasing percentages of ON in developed sub-basins, the concentrations of

ON remained similar in developed (0.73-0.77 mg L-1) and undeveloped sub-basins (0.60-0.79

mg L-1). This indicates the existence of enough ON sources in urban areas such as contributions

from septic tanks, grass cuttings of urban lawns, and deciduous leaves fallen on the ground

(Kroeger et al., 2006; Scott et al., 2007). In addition, more impervious areas in urban areas may

increase the runoff resulting in greater transport of these organic materials that contain ON

(Carpenter et al., 1998; Tufford et al., 1998). Recent studies suggest that microbial

transformations of NO3-N to ON may act as an important source of ON in urban watersheds

(McDowell et al., 2004; Pellerin et al., 2006). In our study, urban land use was slightly positively

correlated (r=+0.28) with ON concentrations. Florida has a warm climate where the existence of

wetlands and artificial stormwater retention ponds might play a stimulating role for biotic









conversion of NO3-N into ON resulting in greater concentrations of ON in streams draining

urban land uses (Groffman et al., 2004; Pellerin et al., 2004). Similarly, greater proportion of ON

was found at the most downstream Alafia stations (58% of total N) than Bell Shoals (39% of

total N) (Fig. 3-6). Alafia station is close to the Tampa Bay estuary where the velocity of the

water decreases and thus this may provide more favorable conditions for denitrification and/or

conversion of NO3-N to ON via vegetation/microbial uptake (Groffman et al., 2004).

Seasonal Impacts on Nitrogen Forms in Stream Waters

The NO3-N concentrations in stream waters were greater in dry season and decreased in

wet season in all the sub-basins (Fig. 3-5). This can be explained by the significantly (P<0.05)

greater temperature in wet season that might increase the biological uptake and denitrification of

NO3-N resulting in lower losses ofNO3-N in wet than dry season (Alexander et al., 2009; Chen

and Driscoll, 2009). Due to greater biological activities in the stream waters, the concentration of

DO is lower in wet than dry season in all the streams (Fig. 3-4). This provides an indication of

greater denitrification losses of NO3-N resulting in lower concentrations of NO3-N in wet than

dry season. Another probable reason for decreased NO3-N concentrations in wet season could be

the shift in flow-paths from groundwater to stormwater in wet season. In the Alafia River

Watershed, EC of stream water is lower in wet than dry season thereby representing the dilution

of the stream waters with stormwater runoff during the wet season. Similar to other salts (EC),

NO3-N concentrations decrease with greater stormwater runoff in wet season. This dilution of

NO3-N concentration in the stream waters during the wet season contradicts previous studies,

who reported higher concentrations (flushing) ofNO3-N with greater runoff in wet season (Chen

and Driscoll, 2009; Green et al., 2004; Qian et al., 2007; Royer et al., 2006). In general, the

flushing of the NO3-N with stormwater in wet season usually occurs when excess of NO3-N is

present in the watershed soils and other sources. For example, in the Indian River Lagoon,









Florida, fertilizers are applied to citrus crop during the wet season and the excess ofNO3-N is

flushed with rain resulting in greater concentrations ofNO3-N in wet than dry season (Qian et al.,

2007). Similarly, the watersheds in Midwest, US, are row crops dominated which receive greater

N inputs and therefore the NO3-N concentrations increase with greater drainage from the

watersheds (Vidon et al., 2008). The Alafia River Watershed is not as intensively cultivated as

the watersheds of the US Midwest and the Indian River Lagoon, therefore greater water is

available due to modified urban drainage for dilution of the dissolved forms ofN (NO3-N)

resulting in decreased concentration of NO3-N during wet than dry season.

In contrast to NO3-N, transport of particulate forms of N which represents ON is greater

with greater storm water runoff in wet than dry season (data not shown). The particulate N forms

could be decomposing products of leaf litter, grass cuttings of lawns, and soil sediments (Culbert

and France, 1995). Therefore, concentrations of ON were greater in wet season than dry season

in the Alafia River Watershed.

Long Term Trends in Nitrogen Concentrations

In our study, the Seasonal Kendall and GLIMMIX procedures showed similar trends

(increasing or decreasing), though the magnitude of the trend slopes differed by 0 to 14% per

year among two procedures (Fig. 3-7, 3-8, 3-9, 3-10). These differences in trend slopes appear to

be associated with the conceptual difference in the procedures of two methods. For example,

Seasonal Kendall trend analysis compares the change in concentrations of nutrients with time

(slope) for each point and the median slope is calculated as a summary statistic describing the

magnitude of the trend (Johnson et al., 2009; Qian et al., 2007). On the other hand, GLIMMIX

procedure fits the linear lines through the data and slope is the rate of change in concentration

with time at any of two points (SAS, 2008). Therefore, the differences in the magnitude of slopes

appear in the trend analysis in two methods. In general, Seasonal Kendall trend analysis









determines the trends at a single location. However, GLIMMIX procedure can be used to

compare the mean concentrations at multiple locations as well as has the ability to include many

parameters such as stream discharge, seasonal variation, distance of the mainstem, and location

of the station etc (Goodrich et al., 2009). In this study, GLIMMIX procedure has shown the

significant variations in mean and seasonal variations in N concentrations at different stations.

The causes for the concentration trends can be either 1) natural such as the changes in

precipitation and evapotranspiration resulting in variations in flow, or 2) anthropogenic such as

changes in land use, nutrient management practices in the watersheds. The flow weighted and

flow un-weighted concentration trends can help discriminate the effects of natural and

anthropogenic changes on nutrient concentrations in streams. In the analysis of trends in flow

weighted concentrations, the effects of natural changes in stream flow on concentration are

removed and therefore flow adjustment of nutrient concentrations allows trends due to

anthropogenic changes to be assessed. In the analysis of trends in observed flow un-weighted

concentrations, no adjustments are made for any natural or anthropogenic influence. Therefore,

the net effects of all simultaneous influences on concentration are evaluated, allowing for the

assessment of nutrient concentrations in streams relative to water quality standards of

Environmental Protection Agency and the condition of aquatic communities (Sprague and

Lorenz, 2009).

At Bell Shoals that drains 89% of the watershed, increasing trends were found in flow un-

weighted (+1.01% per year; P<0.08) and flow weighted (+1.08% per year; P<0.24) total N

concentrations during 1991-2009 (Fig. 3-7; Table 3-2). Flow remained similar and did not show

any significant trends (P<0.75) during 1991-2009. Therefore, the possibility of concentration or

dilution of total N with variations in the natural causes (flow) is ruled out. Non-significant trends









in flow weighted concentrations could be due to greater variations in flow resulting in greater

variance in data and therefore non-significant results (Johnson et al., 2009). Similarly the loads

data at Bell Shoals is insignificant due to greater variations in the data. Greater variation in the

loads could be due to the large variation in smaller streams of the Alafia River Watershed.

However, increasing trends in both flow weighted and flow un-weighted total N concentrations

indicated the possibility of anthropogenic influence in increasing the N concentrations in the

Alafia River. In this regard, the Alafia River Watershed is similar to most of the watersheds in

US which have shown increasing total N concentration trends during the last few decades

(Goolsby and Battaglin, 2001; Turner and Rabalais, 2003; Weston et al., 2009). For example, in

Mississippi River Basin, Turner and Rabalais (2003) reported the increasing total N

concentration trends primarily due to clearing of forests for the agricultural land use in the last

200 years. Similarly, in Minnesota River, Mann-Kendall trend analysis showed increasing NO3-

N concentration trends (from +2.90 to +6.29% per year) during 1976-2003 (Johnson et al.,

2009). They attributed the increasing N losses to the increased drainage of the corn-soybean

dominated watersheds during 1976-2003. Contrary to these studies, in 58 watersheds of Eastern

US, Sprague and Lorenz, (2009) using Mann-Kendall trend analysis have not found any

significant total N concentrations trends during 1993-2003.

In the Alafia River Watershed which is located in the Hillsborough County, population has

grown from 0.83 million in 1990 to 0.99 million in 2000 and 1.20 million in 2009 (US Census

Bureau, 2010). This represents 19.8% increase during 1990-2000 and 19.7% during 2000-2009.

The population growth has resulted in the land use changes in the watershed. For example, the

urban land use (residential and built up) has increased by 8% (from 12 to 20%) of the watershed

during 1990-2007 (SWFWMD, 2007). This is coupled with a decrease of 5% in forests (from 23









to 18%) and 8% decrease in pastures (from 19 to 11%) while agricultural land use remained

similar (8%) during the study period. In our study, mean total N concentrations were positively

correlated with urban (r=0.83) (Fig. 3-11; Fig. 3-12). Therefore, increasing trends of total N

concentration in the Alafia River Watershed may be attributed to the increase in urban land use

of the watershed. Similarly, studies in the Altamaha River, GA by Weston et al., (2009) reported

the significant increasing total N trends (+7.42 tg L-1 per year) during 1970-2000. In their study,

the average population density was increased from 35 to 70 person km-2 while the agricultural

land use had decreased from 44 to 29%; thereby reflecting the impact of urbanization on total N

losses. In general, the increase in urban land use results in increase in the N inputs (fertilizer

application to lawns, septic tanks, leaf litter, and grass cutting of lawns) and changes the

hydrology of the watershed which can increase the N losses to the streams (Carpenter et al.,

1998; Han et al., 2009; Tufford et al., 1998). In our study, two developed sub-basins (English

Creek and North Prong) have shown the decreasing total N concentration trends where the urban

land use has increased by 8-23% during 1990-2007. It is important to note that at sub-basin

scale several anthropogenic changes takes place simultaneously which affect the net N losses

from a watershed. For example, N concentrations in streams may increase with the urbanization

in the watershed while the improvements in the wastewater treatment facilities might decrease N

losses resulting in decreasing or non significant results (Sprague and Lorenz, 2009). Two sub-

basins (North Prong and English Creek) receive discharges from two domestic wastewater

discharges in additions to several small industrial wastewaters (NPDES, 2009). We believe that

in English Creek and North Prong the increasing total N concentration trends due to urbanization

are offset by the improvements in the domestic as well industrial wastewater discharges.









In our study, NO3-N (25-68% of total N) has not any significant increasing or decreasing

trends at all the study stations (Fig. 3-9). Among the sub-basins, only Turkey Creek has shown

the increasing flow un-weighted NO3-N concentration trends during 1991-2009. In the Alafia

River Watershed, several programs such as fertilizer application based on crop/soil tests, N

pollution reduction from residential areas, improvements in wastewater treatment facilities were

initiated to control N (especially NO3-N) pollution during last two decades (TBEP, 2009). It

seems the success of the N abatement programs is partially offset by the increasing urban land

use and population growth resulting in insignificant NO3-N concentration trends during the study

period. In contrast toNO3-N, the ON which contributed 30-71% of total N, has shown

significantly increasing trends most of the study stations (Fig. 3-8). In general, increasing ON

concentration trends due to increases in urban land use in the Alafia River Watershed contradicts

the common belief that urban land use increase the NO3-N rather than ON losses from

watersheds (Pellerin et al., 2006; Stanley and Maxted, 2008). In our study, the proportion of ON

decreased with increasing urban land use while the concentrations of ON were similar in both

developed and undeveloped sub-basins. This suggests that there are sources of ON even in the

urban land uses. In urban land uses, the impervious areas increase the runoff resulting in greater

transport of particulate ON which may ultimately lead to increasing ON concentration trends

with urbanization (Carpenter et al., 1998; Han et al., 2009; Tufford et al., 1998). In a review of

348 watersheds, Pallerin et al. (2006) reported that the ON concentrations in the surface waters

were 2-3 folds greater in urban/suburban (0.47-0.49 mg L-1) than forest watersheds (0.18 mg L-

1). They attributed the greater concentrations in urban land uses to the influence of septic tanks,

wastewater treatment plants, and the biotic conversions of NO3-N to ON. Using isotopic

signatures of carbon and N, Peebles et al. (2009) reported that inorganic N from the fertilizers









was the dominant source of N to the sediments in the Safety Harbor part of the Tampa Bay.

Similarly, long-term N fertilization studies at the Harvard Forest in Massachusetts have shown

that dissolved ON concentrations in the forest floor have increased as a result of elevated

inorganic N deposition (McDowell et al., 2004). In our study, wetlands and artificial retention

ponds might have played a stimulating role for biotic conversion ofNO3-N into ON which

resulted in increased ON concentration trends with urbanization during the study period

(Groffman et al., 2004; Pellerin et al., 2004). In this way, increasing ON concentration trends in

this urbanizing watershed has raised two important questions 1) Is the increase in ON

concentration due to ON sources (such as septic tanks, urban manures/composts, runoff of grass

cuttings, and deciduous leaves fallen on the urban surfaces, 2) Is the ON increase a product of

microbial transformations of NO3-N into ON. Further studies on the source characterization of

the ON can help in devising the BMP's to control N pollution in the Alafia River Watershed.

Results of our trend analysis suggested that the increasing total N concentrations could be

due to the increasing urban land use in the watershed during 1991-2009. None of the sub-basin

showed significant decreasing or increasing NO3-N trends. This has suggested that the

conversion from forest and pastures to urban land use has offset the success of the N abatement

programs in controlling NO3-N pollution in the Alafia River Watershed. On the other hand, lack

of BMP's to target ON (30-71% of total N) has caused increase in ON concentrations ultimately

resulting in increasing total N trends. For controlling N pollution, Turkey Creek with greater

total N concentrations and increasing total N concentration trends should be targeted followed by

English Creek to control N pollution. In these two developed sub-basins, the BMP's should also

focus ON rather than NO3-N alone to protect water quality in the Alafia River Watershed.









Summary

At two mainstem stations (Alafia and Bell Shoals) that drain 89-99% of the Alafia River

Watershed, the mean total N concentrations approached the EPA proposed numeric total N

criteria values for the region. During 2000-2009, total N concentrations were greater in three

developed (1.67-2.43 mg L-1) than two undeveloped sub-basins (0.84-1.21 mg L1) thereby

suggesting that the urbanization of the watershed has increased N losses from watersheds.

Greater proportion of NO3-N in developed (53-68% of total N) than undeveloped sub-basins

(25-30%) indicate that in urban land uses the losses of NO3-N are greater than ON. In our study,

the concentration of NO3-N decreased in wet season due to greater biotic uptake and

denitrification of NO3-N with greater temperature as well as dilution of NO3-N with greater

runoff in wet than dry season. The greater concentrations of ON in wet than dry season were

probably due to greater transport of organic material with runoff water. The trend analysis

showed increasing total N concentration at the mainstem station during 1991-2009. This was

primarily due to the greater increases in ON than NO3-N thereby suggesting the need to control

N losses from ON sources. Three developed sub-basins (Turkey Creek, English Creek, and North

Prong) that had greater total N concentrations should be first targeted to reduce N pollution in the

Alafia River Watershed.









Table 3-1. Station characteristics of the Alafia River Watershed


Sub-basin


Alafia
Bell Shoals

English Creek
Turkey Creek
North Prong


South Prong
Fishhawk Creek
-t USGS st


Station Sampling location
Lat Long


Drainage area
km2 % Residential


Land Use in 2007
Built up Agricultural Pasture Forest Mined


Mainstem Stations
2301718 27.87 -82.32 1072 99 17
2301638 27.86 -82.26 974 89 16
Developed
-t 27.93 -82.06 99 9 21
-t 27.91 -82.18 128 13 20
2301000 27.86 -82.13 350 32 18
Undeveloped
2301300 27.86 -82.13 277 26 3
t 27.85 -82.24 70.6 7 11
ation not present


Table 3-2. Long-term trends in flow weighted and loads ofN forms at Bell Shoals
Parameter data Flow weighted concentrations Loads
Trend Slope (%) p value Trend Slope p value
Total Nitrogen 1991-2009 Increasing +1.01 0.242 No trend 0.746


Organic Nitrogen
Nitrate Nitrogen


1991-2009 1
1991-2009


increasing +1.08 0.016
No trend 0.491








































I
0 2 4 8 12 16 South Prong
Skm I I /


Alafia River Watershed


Figure 3-1. Location map of the Alafia River Watershed.












A. Temperature (Celcius)


A A A A A A A








Alafia BellShoals English Creek Turkey Creek North Prong South Prong FishhawkCreek

B. Dissolved Oxygen (mg L')


30



20


10



12


8


4


0


9


8


7


6


5


1200


800



400


Alafia Bell Shoals English Creek Turkey Creek North Prong South Prong Fishhawk Creek

D. Electrical Conductivity (uS cm1)

BC B BC C B A





S *,m


Alafia BellShoals English Creek Turkey Creek North Prong South Prong FishhawkCreek


Figure 3-2. Chemical characteristics of the stream waters during two time periods from 1991 to
2009. Values indicated by different letters are significantly different at P<0.05 for
each graph.


percentile
tlan
percentile


-


Alafia Bell Shoals English Creek TurkeyCreek North Prong South Prong Fishhawk Creek

C. pH


A A A A A A A



M,:1 No 0


A B AB



- -i -


C C B


--ON











A. Total Nitrogen
1991-1999 2000-2009
B C B A C C D C C B A






N= 120 225 120 207 225 225 74


Bell Shoals Turkey Creek North Prong South Prong


2000-2009
BC C D D C B A






*1i*s .


75 percentile
Median
25 percentile


Alafia Bell Shoals English Turkey North South Fishhawk
Creek Creek Prong Prong Creek


B. Organic Nitrogen
1991-1999 2000-2009

A A A A C AB B B B B A

***ii = -


Bell Shoals


Bell Shoals Turkey Creek North Prong South Prong Alafia Bell Shoals English Turkey North South Fishhawk
Creek Creek Prong Prong Creek
D. Ammonium-N
1991-1999 2000-2009

A A A A B A A A A A A






_-_ I......


Bell Shoals Turkey Creek North Prong South Prong


Alafia Bell Shoals English Turkey North South Fishhawk
Creek Creek Prong Prong Creek


Figure 3-3. Summary of mean monthly concentrations of total, organic, nitrate, and ammonium
nitrogen during 1991-2009 in mainstem (Alafia and Bell Shoals), developed (English
Creek, Turkey Creek, and North Prong), and undeveloped (South Prong and
Fishhawk Creek). Values indicated by different letters are significantly different
according to GLIMMIX procedure at P< 0.05. N in each sub-basin indicates number
of months/ob servations.


4

3

2

. 1


Turkey Creek North Prong South Prong Alafia Bell Shoals English Turkey North South Fishhawk
Creek Creek Prong Prong Creek
C. Nitrate-N


1991-1999
C


C B A


*


* j












A. Temperature (Celcius)
Dry Season Wet Season



*i* *i *n i*



lim~je!!|


Alafia Bell Shoals English Turkey
Creek Creek


Alafia Bell Shoals English Turkey
Creek Creek


75 percentile
Median
25 percentile


North South Fishhawk Alafia Bell Shoals English Turkey North South Fishhawk
Prong Prong Creek Creek Creek Prong Prong Creek
B. Dissolved Oxygen (mg L1)


North South Fishhawk Alafia Bell Shoals English Turkey North South Fishhawk
Prong Prong Creek Creek Creek Prong Prong Creek
C. pH


Dry Season Wet Season










Alafia Bell Shoals English Turkey North South Fishhawk Alafia Bell Shoals English Turkey North South Fishhawk
Creek Creek Prong Prong Creek Creek Creek Prong Prong Creek
D. Electrical Conductivity (uS cm1)
Dry Season Wet Season










Bell Shoals English Turkey North South Fishhawk Bell Shoals English Turkey North South Fishhawk
Creek Creek Prong Prong Creek Creek Creek Prong Prong Creek


Figure 3-4. Seasonal variation in chemical characteristics of the stream waters during 1991-
2009. Values indicated by different letters are significantly different according to
GLIMMIX procedure at P< 0.05. Dotted line represents the mean value.











A. Total Nitrogen
Dry Season Wet Season








N=120 225 120 207 225 225 74


Alafia Bell Shoals English Turkey
Creek Creek


North South Fishhawk Alafia Bell Shoals English Turkey North South Fishhawk
Prong Prong Creek Creek Creek Prong Prong Creek
B. Organic Nitrogen


Dry Season Wet Season









Alafia Bell Shoals English Turkey North South Fishhawk Alafia Bell Shoals English Turkey North South Fishhawk
Creek Creek Prong Prong Creek Creek Creek Prong Prong Creek
C. Nitrate-N
Dry Season Wet Season









Alafia Bell Shoals English Turkey North South Fishhawk Alafia Bell Shoals English Turkey North South Fishhawk
Creek Creek Prong Prong Creek Creek Creek Prong Prong Creek
D. Ammonium-N


Alafia Bell Shoals English Turkey North South Fishhawk Alafia Bell Shoals English Turkey North South Fishhawk
Creek Creek Prong Prong Creek Creek Creek Prong Prong Creek


Figure 3-5. Seasonal variation in mean monthly concentration of nitrogen forms during 1991
2009. Values indicated by different letters are significantly different according to
GLIMMIX procedure at P< 0.05. Dotted line represents the mean value.


75 percentile
Median
?5 percentile


0
.2

4
0
. 4
O
L)


Wet Season








;i *o


Dry Season








1Li16







U Ammonium-N Nitrate-N Organic-N


1011 lllll li,,'
100

60
40
20
0 -
Dry Wet Dry Wet Dry Wet Dry Wet Dry Wet Dry Wet Dry Wet
Alafia Bell Shoals English Turkey North Prong South Prong Fishhawk
Creek Creek Creek
Figure 3-6. Seasonal variation in proportion of organic, nitrate, and ammonium nitrogen during
1991-2009.















5 Mann-Kendall Slope = -2.75% peryear, P= 0.18
SGLIMMIX Slope= -1.02% peryear, P= 0.17


Alafia


N= 120


Numeric nutrient criteria (1.79 mg L )

/


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09

5 Mann-Kendall Slope = +0.17% per year, P= 0.82 Bell Shoals N= 225

4

3

2

1-

0
91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
5 Mann-Kendall Slope = -4.84% peryear, P= 0.19 English Creek N=120
GLIMMIX Slope =-1.14% per year, P = 0.44
4

3
2

1-

0
91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09


5 Mann-Kendall Slope = +1.51% per year, P= 0.21
GLIMMIX Slope =+1.02% per year, P = 0.005
4


Turkey Creek N= 217


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09


5 -Mann-Kendall Slope = -0.42% per year, P= 0.64
GLIMMIX Slope =-1.11% per year, P = 0.76
4

3
2

1


North Prong


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
Mann-Kendall Slope = +1.70% per year, P= 0.13 South Prong N= 225
SGLIMMIX Slope =+1.07% per year, P = 0.03








91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09


Figure 3-7. Long-term (1991-2009) trends in monthly flow un-weighted total N concentrations

in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and

North Prong), and undeveloped (South Prong and Fishhawk Creek) sub-basins of the

Alafia River Watershed. The dotted line indicates the proposed numeric nutrient (1.79

mg L'1) criteria for the region (EPA, 2010).






















91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
ManKnalSoe=+.0 e yaP .5 Bl hasN 2


2-






= Mann-Kendall Slope = +2.51% per year, P= 0.005 Turkey Creek N= 211
Q 3 GLIMMIX Slope =+0.94% per year, P = 0.001
0 2
0
0 2-

1-

0
91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
Mann-Kendall Slope = +0.72% per year, P=- 0.46 North Prong N= 225
3 -GLIMMIX Slope =+1.00% per year, P = 0.99

20

1 -

0
91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09


Figure 3-8. Long-term (1991-2009) trends in monthly flow un-weighted organic N
concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek,
Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk
Creek) sub-basins of the Alafia River Watershed.


Mann-Kendall Slope = +1.80% per year, P= 0.45 Bell Shoals
GLIMMIX Slope =+0.94% per year, P= 0.001


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09


Mann-Kendall Slope = -3.72% per year, P= 0.41 English Creek
GLIMMIX Slope =+0.65% per year, P= 0.38


Mann-Kendall Slope = +1.95% per year, P= 0.25 South Prong
GLIMMIX Slope =+0.95% per year, P = 0.001


N= 225


N=120


N= 225


4
























91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09

4 Mann-Kendall Slope = -3.23% per year, P= 0.42 English Creek N=12(
GLIMMIX Slope =-1.74% per year, P = 0.32
3


E

o
0
S 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09

4 Mann-Kendall Slope = +0.18% per year, P= 0.90 Turkey Creek N= 217
C GLIMMIX Slope =+1.13% per year P= 0.21
0
3






91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09

SJMann-Kendall Slope = -0.11% per year, P= 0.98 North Prong N= 22!
GLIMMIX Slope =-0.31% per year, P = 0.64


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09


Figure 3-9. Long-term (1991-2009) trends in monthly nitrate N concentrations in mainstem

(Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North Prong),

and undeveloped (South Prong and Fishhawk Creek) sub-basins of the Alafia River

Watershed.


3 Mann-Kendall Slope = -0.27% per year, P= 0.90
I GLIMMIX Slope = -1.04% per year, P= 0.64


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
Mann-Kendall Slope = -2.7% per year, P= 0.98 South Prong N= 225
GLIMMIX Slope = -0.98% per year, P= 0.54















0.5 Mann-Kendall Slope =-3.7% per year, P= 0.81 Alafia
0.4 GLIMMIX Slope =--1.00% per year, P= 0.76 Alaia N120

0.3 -

0.2 -

0.1

0.0
91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
0.5
Mann-Kendall Slope =-1.6% per year, P= 0.38 Bell Shoals N= 225
0.4 GLIMMIX Slope = -1.00% per year, P= 0.66

0.3

0.2

0.1

0.0
91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
0.5
Mann-Kendall Slope =-15.7% peryear, P=0.12 English Creek N=120
0.4 GLIMMIX Slope =-0.92% per year, P= 0.30

0.3 -

0.2

E 0.1

0 0 .. .. ...
91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
S0.5
) Mann-Kendall Slope = +2.3% per year, P= 0.34 Turkey Creek N= 217
0.4 GLIMMIX Slope = +0.99% per year, P = 0.02
0.
0 0.3

0.2

0.1

0.0
91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
0.5
Mann-Kendall Slope = -3.6% per year, P= 0.22 North Prong N= 225
0.4 GLIMMIX Slope =-0.87% per year, P= 0.34

0.3-

0.2

0.1

0.0 -L
91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
0.5
Mann-Kendall Slope = +1.6% per year, P=- 0.44 South Prong N= 225
0.4 GLIMMIX Slope= +1.00% peryear, P= 0.14

0.3

0.2-

0.1

0.0
91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09





Figure 3-10. Long-term (1991-2009) trends in monthly flow un-weighted ammonium N

concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek,

Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk

Creek) sub-basins of the Alafia River Watershed.


















Total Nitrogen

y= 0.039x+0.857, r= 0.83 y
P<0.05
2.0


t 1.5-


1.0


I 0.5


0.0-
0 10 20 30 40 0


Total Nitrogen

y= 0.033x+1.296, r= 0.49
P>0.05





r11


II 0.0 -
0 10 20 30 0


Nitrate Nitrogen


I-I 0.0 +
10 20 30 40 0

Urban Land Use (%)


Nitrate Nitrogen

y- 0.025x+0.5367, r= 0.53
2.0 P>0.05


1.5


1.0


0.5-


10 20

Agricultural Land Use (%)


Organic Nitrogen

y= 0.0006x+0.676, r= 0.28
P>0.05
2.0-


1.5


1.0 -


0.5 -I


10 20 30 40


Organic Nitrogen
y= 0.003x+0.7515, r= 0.03
.0 P>0.05


.5


.0- I





.0


30 0


10 20 30


Figure 3-11. Relationship between percent urban and agricultural land use and nitrogen forms in

different sub-basins (* significantly correlated at P< 0.05). Mean monthly data of

1991-1999 and 2000-2009 with land use of 1999 and 2007 respectively.


.















Total Nitrogen

--0.032x+1.639, r=- 0.06
P>0.05





I


--I '0.0 +
10 20 30 0


Total Nitrogen

y- -0.015x+1.893, r= -0.17
P>0.05



I
w r


SI 0.0 -
10 20 30 40 10


Nitrate Nitrogen


10 20

Pasture Land Use (%)


Nitrate Nitrogen


20 30

Forest Land Use (%)


Organic Nitrogen

y= -0.009x+0.876, r= -0.34
2.0- P>0.05
2.0-


1.5


1.0-


0.5-


-I 0.0 +
30 0


-I 0.0 -
40 10


10 20 30




Organic Nitrogen


20 30 40


Figure 3-12. Relationship between pasture and forest land use and nitrogen forms in streams

draining different sub-basins (* significantly correlated at P< 0.05). Mean monthly

data of 1991-1999 and 2000-2009 with land use of 1999 and 2007 respectively.


y 0.006x+0.713, r= 0.05
2.0- >0.05
2.0


1.5-


1.0-o


0.5-


T-

0






o
E

U 4
C
0


y= -0.007x+0.949, r= -0.09
2.0- P>0.05
2.0

I
1.5-


1.0o-


0.5- [
Fm


y= -0.007x+0.884, r= -0.28
2.0- P>0.05
2.0


1.5-


1.0- w


0.5









CHAPTER 4
PHOSPHORUS TRANSPORT IN AN URBAN WATERSHED

Abstract

Non point source phosphorus (P) pollution is a significant concern in several waterbodies.

In this study, we determined the concentrations of total P, dissolved reactive P (DRP), and other

P forms in stream waters draining developed and undeveloped sub-basins, ranging in size from

19 to 350 km2, of the Alafia River Watershed (total drainage area: 1085 km2). During 1991-

2009, mean monthly total P concentrations ranged from 0.56 to 3.95 mg P L-1. Of total P,

dissolved reactive P (DRP) was dominant (70-90% of total P) than other P (10-30% of total P)

in developed and undeveloped sub-basins. None of the P forms were significantly (P< 0.05)

correlated with urban, agricultural, forest, and pasture land use of the sub-basin (r<0.50)

indicating that the P concentrations are not controlled by land use of the Alafia River Watershed.

Greater concentration of total P were greater in two developed (North Prong and English Creek:

2.18-2.53 mg P L1) may be due to P rich geology, active mined lands, and discharges of P rich

wastewater in these sub-basins. In all the developed and undeveloped sub-basins, the

concentrations of P forms were greater in wet than dry season. This represents that the flushing

of the P with greater rainfall-runoff in wet season that might have accumulated due to dissolution

and desorption of P from soil minerals. Long term trend analysis showed decreasing total P and

DRP trends in both flow weighted and flow un-weighted concentrations. The decreasing trends

in P concentrations indicated that the P abatement programs such as increased regulations on P

discharges from mined lands as well as wastewater discharges were successful in controlling P

pollution in the Alafia River Watershed. During 2000-2009, all the sub-basin except Fishhawk

Creek had greater total P concentrations (0.80-2.53 mg P L1) than EPA proposed numeric total

P value of 0.739 mg L-1 for the region. Results suggests that P source controls from mined and









wastewater discharges in North Prong, English Creek, Turkey Creek, and South Prong are

needed to control P pollution in the Alafia River Watershed.

Introduction

Phosphorus (P) pollution is the primary source of water quality degradation in the US

(Bricker et al., 1998; USEPA, 2001) as total P concentration as low as 0.050 mg P L-1 in lakes

and 0.10 mg P L-1 in stream waters can impair the water quality (USEPA, 1986). In the US, 45%

of lakes and 35% of rivers are degraded and approximately 90% of the rivers show signs of

eutrophication due to P enrichment (USEPA, 1996, 2000; Dodds et al., 2009; Paerl, 2009). The

consequences of eutrophication include hypoxia, acidification of natural waters, degradation of

coastal waters including increased episodes of noxious algal blooms, and reductions in aquatic

macrophyte communities often leading to substantial shifts in ecosystem structure and function

(Carpenter et al., 1998; Dodds et al., 2009). The cost of eutrophication has been estimated at

$2.2 billion per annum due to losses of recreational water usage, spending on recovery of

threatened and endangered species, and drinking waters (Dodds et al., 2009). Therefore, controls

on the sources of P can help to protect the water resources and reduce water quality deterioration

in a region.

Phosphorus in water bodies can come from either point sources such as wastewater and

industrial effluents or non-point sources which include the storm water runoff losses from urban

areas, agricultural fields, animal feedlots, roadways, and mined areas (Edwards and Withers,

2008). Following the passage of Clean Water Act in 1970, P contributions from the point sources

have decreased and consequently non-point source has become the dominant form of P pollution

in many watersheds in the US (Bricker et al., 1998; USEPA, 2001; Diebel et al., 2009; Maxted et

al., 2009). A logical approach to control non-point P pollution may be to determine the hot spot

areas that contribute greater P losses and to develop best management practices (BMPs) to









reduce P pollution from these hot spots (Haygarth, 2005). In Wisconsin, US, Diebel et al. (2009)

reported that targeting 10% of watersheds decreased total P losses by 20% for the entire state.

Therefore, federal, state and local government, nonprofits (e.g., The Nature Conservancy), and

stakeholder groups (e.g., watershed associations, soil and water conservation districts) find

watershed scale pollution assessment valuable for identifying and targeting the land areas which

contribute greater pollution (hot spots) in a watershed.

In order to identify the hot spots areas at a watershed scale, it is vital to fully understand

how different land uses impact non-point P pollution in a watershed. In general, land use affects

the net anthropogenic inputs of P in watersheds. For example, Russell et al. (2008) reported that

net anthropogenic P inputs were greater in agricultural and urban land uses (15.8-19.6 kg P ha-1)

as compared to forest watersheds (1.6 kg P ha-1). Another consequence of the land use change is

the alteration of the flow paths. For example, in urban watersheds, impervious surfaces lead to

increased runoff due to altered hydrology compared to forest watersheds (Arnold and Gibbons,

1996; Lee and Heaney, 2003; Paul and Meyer, 2001). In this way, urban land uses result in

greater flows and facilitates the transport of suspended solids, an important P transport source

(Mulliss et al., 1996; Stone and Droppo, 1994). Further, the loss of natural vegetation in urban

land uses reduces recycling and uptake of P by vegetation that can immobilize P (Abelho, 2001;

Wahl et al., 1997). This can result in a greater amount of P available and thus export to water

bodies in urban land use dominating watersheds.

The influence of land use on P exports from watersheds can be seen in Lake Washington,

US, where Ellison and Brett, (2006) reported greater total P concentrations in streams draining

agricultural (0.13 mg P L1) and urban (0.07 mg P L-1) than forest sub-basins (0.03 mg P L-1).

Similarly, in 17 watersheds dominated with urban (22-87%) or forest (6-73%) land use, Brett et









al. (2005) reported that with 10% conversion of forest to urban land use, total P concentrations

increased by 0.07 mg P L-1. In contrast to these studies, other researchers have not observed any

significant difference in P losses from different land use dominated watersheds (Dodds and

Oakes, 2006; Johnson et al., 1997). Therefore, losses of P may or may not be impacted by land

uses in a watershed despite the fact that land use alters inputs as well as mode of delivery of P

from watersheds to streams.

Land use in a watershed may influence the proportion of the different P forms such as

dissolved reactive P (DRP) and particulate P (PP) in the streams (Stone and Droppo, 1994). In

general, PP losses are often associated with the erosion of soil particles that are enhanced by the

anthropogenic land use and soil disturbance typical in agricultural dominated watersheds

(Wallbrink et al., 2003). On the other hand, DRP in the stream waters represents the losses from

the anthropogenic sources such as fertilizer application to the agricultural fields as well as to the

weathering of the P minerals if present in a watershed (Harrison et al., 2005). Most of the studies

have demonstrated that about 50% of the P losses occur as PP (Omernik, 1977; Sharpley and

Menzel, 1987; Vaithiyanathan and Correll, 1992). However, in Lake Washington, Ellision and

Brett (2006) reported that of total P, the total dissolved P was 72% in urban (0.05 mg P L-1) and

60-64% in forest and mixed (0.02-0.04 mg P L-1) compared to 50% (0.07 mg P L-1) in

agricultural streams. Information on the losses of different P forms from sub-basins under

contrasting land use activities may act as a useful tool in studying the cycling of P, which may be

valuable to develop source control of P in sub-basins draining different land uses.

In addition to the land uses, stream flow conditions affect the concentration and

proportions of P losses from the watersheds (Ellison and Brett, 2006; Qian et al., 2007). For

example, Royer et al. (2006) reported that in corn-soybean dominated (80-90%) watersheds in









Illinois, US, >80% of the total P losses occur only during the high flow conditions (>90%

percentile flow). In Indian River Lagoon, Florida, Qian et al. (2007) reported that the

concentrations of P forms were two times greater during wet (DRP: 0.23 mg P L-1; total P: 0.31

mg P L1) than dry season (DRP: 0.11 mg P L 1; total P: 0.15 mg P L-1). In general, P is sorbed

onto the soil particles and greater losses of P occur when sufficient water is available to transport

soil particles from land to streams. In this way, ecological impacts due to P pollution may depend

upon the flow conditions (Edwards et al., 2000; Jarvie et al., 2006; Svendsen et al., 1995).

The Alafia River Watershed is an example of such a watershed, where land use has been

continuously changing from natural areas to urban lands. As a result, several sub-basins of the

watershed have modified hydrology due to stormwater retention ponds and water convergence

structures that are meant to drain water during high flow events to avoid flooding. In addition,

two sub-basins of the watershed are dominant in mined lands where phosphate mining is a

commercial enterprise. We hypothesize that the 1) concentration of P forms may be different in

sub-basins that drain different land uses, 2) concentration of P forms may differ in dry and wet

seasons due to difference in rainfall, and 3) urbanization of the watershed may increase the P

concentrations. The objectives of this study were to (1) determine how different sub-basins (with

different land uses) influence concentrations of different forms of P in stream waters; (2)

investigate the influence of two distinct seasons (dry and wet) on concentrations of P forms in

stream waters, and (3) evaluate the long term trends of P losses in different sub-basins of the

watershed.

Materials and Methods

Study Site Description

Refer to Chapter 2 for detailed description of sub-basins of the Alafia River Watershed.









Data Collection

Monthly concentrations of total P, and dissolved reactive P (DRP) from 1991 to 2009 were

obtained from the Environmental Protection Commission ofHillsborough County

(http://www.epchc.org). Among the two mainstem stations, water quality data was available only

for Bell Shoals station for 19-years study period, while for Alafia station, the data was available

for 1999 to 2009 (Table 4-1). Among other sub-basins in the watershed, 19 years of water quality

data were available only for North Prong, South Prong, and Turkey Creek while the data

availability for English Creek was from 1999 to 2009 and for Fishhawk Creek from 2005 to

2009.

Stream-water Collection and Analysis

Each month, a grab sample from surface water was collected from the surface of channel

thalweg (centre of the stream flow) in plastic water bottles by the EPCHC staff. Before collection

of the samples, the water bottles were rinsed three times with the stream water. The collected

samples were chilled with ice and transported to the laboratory where samples were stored at 4

C prior to analysis. Environmental Protection Commission Hillsborough County staff analyzed

the surface water samples for total P (EPA 365.4 method) and DRP (SM 4500-P method) using a

discrete analyzer (Seal Analytical, Model AQ2 Mequon, WI). Other P (OP) was calculated as

follows: OP = total P-DRP.

Statistical Analysis

Mean, median, and standard deviations were calculated using MS Excel 2007. We used

GLIMMIX time series procedure SAS Version 9.1 (SAS Institute, Cary, NC) to compare the P

forms across different sub-basins as well as to determine the long term trends in P forms. The

GLIMMIX procedure fits models to data with correlations or non constant variability and

assumes normal random effects (SAS Institute, 2008). In this time series analysis two kinds of









effects were used 1) fixed effects which included stream station and season of the year, and 2)

random effects, which were the date of sample collection. For the first fixed effect (stream

station), each station was treated as a group and the variance was pooled over the sites; similarly

for the second fixed effect (season), each wet (June-September) and dry (October-May) season

at each station was treated as one group of observations. The date of sample collection from

1991-2009 for the each station was considered as random effect and therefore a time series was

constructed. With this, we compared 1) the difference in P forms at each of the station averaged

over all sampling dates and 2) differences in pooled mean concentrations of P forms during wet

and dry seasons at each station using P<0.05 as significance level. In addition, same time series

with fixed and random effects were used to determine the trends in P forms over the period of

study. The concentrations of different forms of P were logarithmically transformed to equalize

variances and normalize skewed data and were back transformed to present means of P forms in

more relevant manner.

For the long term water quality data, various techniques have been used (Richards, 2006;

Johnson et al., 2009; Goodrich et al., 2009). However, non parametric Mann-Kendall and

Seasonal Kendall are among the most common methods for determination of long term trends of

water quality (Daroub et al., 2009; Johnson et al., 2009). Therefore, in addition to GLIMMIX

procedure, we determined the long term trends using Seasonal Kendall procedure. In Seasonal

Kendall trend is calculated by comparing all potential data pairs. If the later value in the pair (in

time) is higher than the first, a plus sign is scored. If the later value in the pair is lower than the

first, a minus sign is scored. If the results find an equal number of pluses and minuses, then there

is no discernible trend. In other words, there is just as much of a likelihood that a pair of data









values will be higher (or lower) than the next one. If the results show more pluses than minuses,

this would indicate that a positive trend is likely (Gilbert, 1987; Helsel and Hirsch, 1991).

The trend in both the procedures is defined by the rate of change over time, which is

referred to as the trend slope (Sen 1968; Schertz et al. 1991). The trend slope can be expressed

either as change in original units per year [S (0)], or as a percent of the mean concentration of

water quality variable [S (%)]. The former is the median slope of all pair wise comparisons (each

pair wise difference is divided by the number of years separating the pair of observations) while

the latter is produced by dividing the slope (in original units per year) by the mean and

multiplying by 100.

The P forms concentrations data were divided into two time periods: 1991-2000 and

2001-2009. To determine the effect of land use change on P forms in corresponding sub-basins,

the land use data of 1999 and 2007 was correlated with P concentrations data of 1991-2000 and

2001-2009, respectively, using SAS PROC CORR procedure.

Results

Chemical Characteristics of Stream Waters

Refer to Chapter 3 for detailed description of chemical characteristics of the stream waters.

Concentrations of Phosphorus Forms in Streams Draining Different Sub-basins

Mean monthly total P concentration at the mainstem station (Bell Shoals) that drains 89%

of the watershed was 1.76 mg P L-1 during 1991-1999 (Fig. 4-2A). Among the sub-basins, total

P concentrations were 4-5 folds greater in one developed (North Prong: 3.95 mg P L1) than one

developed (Turkey Creek: 0.78 mg P L-1) and one undeveloped (South Prong: 0.89 mg L-1) sub-

basin. The concentrations of total P decreased with time in the Alafia River Watershed (Fig. 4-

2A). For example, at mainstem stations (Bell Shoals), mean total P concentrations were lower

during 2000- 2009 (1.24 mg P L-1) than 1991-1999 (1.76 mg P L-1). Similar to the mainstem









station, at North Prong, South Prong, and Turkey Creek, the total P concentrations decreased

from 0.9-3.9 mg P L-1 in 1991-1999 to 0.6-2.5 mg P L1 in 2000-2009. Among the sub-basins,

two developed sub-basins had significantly (P< 0.05) greater total P concentrations (North Prong

and English Creek: 2.18-2.53 mg P L1) than one developed (Turkey Creek; 0.90 mg P L1) and

two undeveloped sub-basins (0.56-0.80 mg P L-1).

At the mainstem station (Bell Shoals), mean DRP concentration was greater (1.46 mg L-1)

during 1991-1999 than 2000-2009 (1.00 mg L1) (Fig. 4-2B). Similarly, at two developed (North

Prong and Turkey Creek) and one undeveloped (South Prong), the concentration of DRP

decreased from 0.8-3.3 mg L1 in 1991-1999 to 0.7-2.2 mg L1 in 2000-2009. During 2000-

2009, all the three developed sub-basins had significantly (P< 0.05) greater DRP concentrations

(North Prong, South Prong, and Turkey Creek: 0.63-2.18 mg P L-1: 70-90% of total P) than one

undeveloped sub-basin (Fishhawk Creek: 0.40 mg P L-1: 72% of total P). However, North Prong

and English Creek had 3-4 folds greater DRP concentrations (1.95-2.18 mg P L1) than Turkey

Creek (0.63 mg P L-1). Among the undeveloped sub-basins, South Prong had significantly (P<

0.05) greater total P concentration (0.68 mg P L-1) than Fishhawk Creek (0.40 mg P L-1).

At the mainstem station (Bell Shoals), mean monthly OP concentrations were similar

during 1991- 1999 (0.30 mg P L-1; 17% of total P) and 2000-2009 (0.29 mg P L-1; 19% of total

P) (Fig. 4-2C). At two developed (North Prong and Turkey Creek) and one undeveloped (South

Prong) sub-basin, the concentrations of OP were slightly greater (0.09-0.6 mg P L1) during

1991-1999 than 2000-2009 (0.1-0.3 mg P L-1). Among the sub-basins, mean OP concentrations

were slightly greater in developed (0.21-0.34 mg P L-1) than undeveloped sub-basins (0.12-0.15

mg P L1). The greatest concentrations of other P forms were at North Prong (0.34 mg P L1) and

lowest at South Prong (0.12 mg P L-1) during 2000-2009.









Seasonal Variation in Phosphorus Concentrations in Streams Draining Different Sub-
basins

The water quality data from 1991-2009 showed that total P concentrations in stream water

were greater, but not statistically significant (P< 0.05) in wet than dry season at all study stations

(Fig. 4-3A). At two mainstem stations, total P concentration was 40-45% greater in wet (1.26-

1.88 mg P L-1) than dry season (0.90-1.30 mg P L-1). In developed sub-basins, the concentration

of total P was 63% greater at English Creek (2.93 mg P L-1) followed by 38% greater at Turkey

Creek (1.04 mg P L-1) and 9% greater at North Prong (3.42 mg P L-1) in wet than dry season. In

undeveloped sub-basins, total P concentrations were 9-20% greater in wet (0.60-0.95 mg P L-1)

than dry season (0.55-0.79 mg P L-1).

Similar to total P, mean concentration of DRP was greater, but statistically (P<0.05)

similar in wet and dry seasons at all the study stations (Fig. 4-3B). At two mainstem stations, the

DRP concentration was 21-46% greater in wet (1.03-1.39 mg P L-1) than dry season (0.71-1.14

mg P L-1). Among the developed sub-basins, DRP increased by 55-60% (from 0.53-1.60 mg P

L-1 to 0.83-2.56 mg P L1) at Turkey Creek and English Creek followed a comparatively lower

increase at North Prong (from 2.68 to 2.85 mg P L-1; 6%) in wet than dry season. In undeveloped

sub-basins, the increase in total P concentration was -10% (from 0.39-0.71 mg P L-1 to 0.42-

0.79 mg P L-1) in wet than dry season.

Proportion of DRP was 71-90% of total P in wet and dry seasons at all the study stations

(Fig. 4-4). Among the developed sub-basins, overall proportion of DRP was greater at English

Creek and North Prong (83-89% of total P) than Turkey Creek (71-79% of total P). Among the

undeveloped sub-basins, the greater proportion of DRP was at South Prong (83-89% of total P)

than Fishhawk Creek (71%).









Mean concentration of OP forms during dry and wet seasons were statistically (P<0.05)

similar; however, the magnitude of variation was greater than total P and DRP at all the sub-

basins (Fig. 4-3C). For example, other P was -220% greater at Bell Shoals (0.49 mg P L-) and

10% greater at Alafia (0.23 mg P L-1) in wet compared to dry season. Among the developed sub-

basins, other P concentrations increased by 80% at English Creek (from 0.20 to 0.36 mg P L-1),

26% at North Prong (from 0.45 to 0.57 mg P L-1), 2% at Turkey Creek (from 0.22-0.24 mg P L-

1) in the wet than dry season. Among the undeveloped sub-basins, the increase in OP

concentration in wet compared to dry season was 100% at South Prong (from 0.08 to 0.16 mg P

L-1) and 8% at Fishhawk Creek (from 0.15 to 0.17 mg P L-1).

Proportion of OP to total P was similar: 11-43% in dry season and 14-44% in wet season

at all sub-basins (Fig. 4-4). However, at two mainstem stations, proportion of OP was greater in

wet (22-40%) than dry (15-29%).

Long Term Trends in Concentrations of Flow Un-weighted Phosphorus Forms in Streams
Draining Different Sub-basins

The longest data record (19 years) was available only for one mainstem station (Bell

Shoals) that drains 89% of the watershed area and three other sub-basins (Fig. 4-5). The results

of the Seasonal Kendall trend analysis showed a significant (P<0.002) decreasing flow un-

weighted total P concentration trend (-3.8% per year; 57.6 tg P L-1 per year) at Bell Shoals. This

decreasing trend equates to -72% decrease in mean total P concentrations during 1991-2009.

Total P concentrations trends were decreasing at North Prong (-4.8% per year; 104.2 tg P L-1

per year; P<0.001) and South Prong (-2.2% per year; 14.4 tg P L-1 per year; P< 0.03). In

contrast to these sub-basins, Turkey Creek showed non-significant (P<0.29) total P

concentrations trend during 1991-2009. Among the stations with 10-years of data, English Creek

showed a significant decreasing total P trends (-36.5% per year; P< 0.008).









Similar to the Seasonal Kendall trend analysis, GLIMMIX model showed a significant (P

<0.001) overall decreasing total P trend at Bell Shoals (-0.9% per year) during 1991-2009 (Fig.

4-5). At Turkey Creek, total P showed an increasing (P< 0.001) trend (+1.1% per year). In

contrast to Turkey Creek, other two sub-basins with 19-years data record showed significant (P

<0.001) but with lower magnitude decreasing total P concentration trends (-0.3 and -1.1% per

year). The total P trends at two stations with 10-years of data record were decreasing (from -1.09

to -1.11% per year).

Seasonal Kendall trend analysis showed a decreasing DRP concentration trend (-4.4% per

year; P<0.003) during 1991-2009 (Fig. 4-6). Among the sub-basins with 19-years of data record,

DRP concentration trends showed a non significant increase at Turkey Creek (+0.5% per year;

P<0.55) compared to a significant decreasing DRP trends at North Prong (-4.2% per year;

P<0.002) and South Prong (-2.3% per year; P<0.09). The stations with 10-years of data record

showed decreasing DRP concentration trends at Alafia (-5.6% per year; P<0.37) and English

Creek (-34.3% per year; P<0.009) during 1999-2009 (Fig. 4-6).

GLIMMIX model showed a significant (P<0.001) decreasing DRP concentration trend (-

1.14% per year; P< 0.001) at Bell Shoals (Table 4-6). Turkey Creek had increasing DRP

concentrations trend (+0.99% per year; P<0.0001) as compared to decreasing trends at North

Prong (-1.03% per year; P<0.001) and South Prong (-0.98% per year; P< 0.03). During 1999-

2009, decreasing DRP trends were found at English Creek (-1.23% per year; P<0.001) and

Alafia (-0.89% per year; P<0.018).

Both the Seasonal Kendall and GLIMMIX procedures showed non-significant trends for

OP concentration at all sub-basins (Fig. 4-7). However, other P concentration trends were

decreasing at Bell Shoals (from -1.06 to -1.60% per year) during 1991-2009.









Long Term Trends in Flow Weighted Concentrations at Mainstem Station

At the mainstem station that drains 89% of the Alafia River Watershed showed a

significant (P< 0.041) decreasing total P concentration trend (-1.14% per year) during 1991-

2009 (Table 4-2). Mean total P concentration at the mainstem station was greater during 1991-

1999 (1.76 mg P L-1) and 2000-2009 (1.24 mg P L-1) thereby reflecting the decreasing total P

concentration trends in terms of actual concentrations. Similarly, DRP showed significant (P<

0.016) decreasing trend (-1.16% per year) during 1991-2009 (Table 4-2). In contrast to total P

and DRP, the flow weighted concentration trends were insignificant in OP forms (P<0.561) at

Bell Shoals.

Long Term Trends in Phosphorus Loads at Mainstem Station

At mainstem station (Bell Shoals), total P loads were not significant during 1991-2009

(Table 4-2). The non significant trends in the P loads during wet and dry seasons could be due to

greater variations in P loads.

Discussion

Land Use Impacts on Phosphorus Concentrations in Streams

During 2000-2009, at two mainstem stations (Alafia and Bell Shoals), total P

concentrations were greater (1.02-1.76 mg P L-) than the EPA's proposed numeric total P value

of 0.739 mg P L-1 for rivers in the Tampa Bay (EPA, 2010). In our study, total P concentrations

were an order of magnitude greater than total P in Miller Creek Watersheds in Kansas, US

(0.008-0.22 mg P L-1, mean= 0.031 mg P L-1) reported by Dodds and Oakes (2006).

Concentrations of total P in our study were also greater than corn-soybean dominated (>60%)

watersheds of Indiana, USA (0.13-0.23 mg P L1) (Vidon et al., 2008). It is important to note that

the Alafia River flows over a geologic P rich formation (fluorapatite) that is commercially mined









for P (Lane, 1994). As a result, P concentration is comparatively higher in Alafia River than

other studies conducted across US.

Mean monthly concentrations of total P were greater in two developed sub-basins (North

Prong and English Creek) than other sub-basins (Fig. 4-2). This could be attributed to active

phosphate mining in North Prong (39% mined land use) and small animal feeding operations in

English Creek may have elevated the P concentrations in stream waters (NPDES, 2009). In

addition, a wastewater treatment plant discharges 0.35 m3 sec-1 of wastewater with mean total P

concentration of 3.1 mg P L-1 in the North Prong. In our study, total P concentrations were not

significantly (P<0.05) correlated with urban, agricultural, forest, and pasture land uses thereby

indicating that P losses are not significantly affected by land uses in the Alafia River Watershed

(Fig. 4-8, 4-9). Negative (r= -0.47: P>0.05) correlations between total P and pasture land uses

might be due to the fact that pastures may act as a potential buffer to reduce the transport of P

from watersheds to streams (Abelho, 2001; Wahl et al., 1997; Zaimes et al., 2008). On the other

hand, slightly positive but not significant correlation between total P and urban land use (r=

+0.33) may be due to greater impervious areas which increase the runoff generation to facilitate

the transport of P from the watersheds (Arnold and Gibbons, 1996; Lee and Heaney, 2003; Paul

and Meyer, 2001). It appears that the P concentration in the streams of the Alafia River

Watershed is controlled by the geology and presence of active mined land use along with point

source input from domestic and industrial (phosphate mining) wastewater. In our study, lower

concentrations of total P at South Prong (66% mined land use) may be due to the fact that most

of the mined lands discharge wastewaters to the sub-basin only during extremely high flow

events and therefore is not impacted with mining activities.









Seasonal Impact on Phosphorus Concentrations in Streams

In our study, the concentrations ofP forms were greater but not significant (P< 0.05) in

wet than dry season at all the sub-basins (Fig. 4-3). The increase in concentrations of P forms in

wet season was not significant due to greater variations in P concentrations. However, higher

concentrations of P forms in the wet season represent the influence of non-point sources on P

pollution in the Alafia River Watershed. Previous studies have reported greater losses ofP

during high flow conditions. For instance, in high fertilizer inputs corn-soybean dominated

watersheds of the Midwest, US, Royer et al. (2006) reported that >80% of the P losses occurred

during the extreme discharges (>90% percentile flow). Similarly, in Indian River Lagoon,

Florida, due to application of fertilizers to agricultural crops in wet season the concentrations of

both DRP and total P were two times greater in wet than dry season (Qian et al., 2007). Research

has indicated that most solute losses occur when large quantities of solutes are present in the

landscape coupled with increased flows (Boyer et al., 1997; Verseveld et al., 2009). Although

mechanisms controlling the temporal pattern in P concentrations were not directly investigated, it

seems probable that during the dry season the dissolution and desorption of P occurs in

geologically P rich Alafia River Watershed which is subsequently flushed with increasing

discharge in wet season (Chen and Driscoll, 2009; Wetzel, 2001). Another probable reason for

greater concentration of P forms could be the release of P due to suspension of the stream

sediments with the high flow events during the wet season (Svendsen et al., 1995).

In our study, the concentrations of other P forms were greater in wet than dry season;

however, the proportions of DRP and other P forms remained similar during both wet and dry

seasons (Fig. 4-4). This contradicts previous studies which reported the lower proportion of DRP

in wet season (Ellison and Brett, 2006; Pacini and Gachter, 1999). In general, streams are

characterized by higher proportions of DRP in dry season when the sediment transport capacity









is low as fine bed sediments are the primary source of PP and the proportions of PP increase with

the transport of greater particulates into the streams during high flow conditions in wet season

(Ellison and Brett, 2006; Pacini and Gachter, 1999). Similar proportions of DRP and other P

forms during dry and wet seasons in our study suggests that desorption of P from the soil

particles and slow dissolution of P minerals during dry season is the dominant source of P which

flushed during wet season.

Long Term Trends in Concentration of Phosphorus Forms in Stream Waters

The Seasonal Kendall and GLIMMIX procedures showed similar trends (positive or

negative), though with different magnitude of change (slope %) (Fig. 4-5; 4-6; 4-7). These

differences in trend slopes appear to be associated with the conceptual differences in the way of

calculating the slopes in both the methods. For example, Seasonal Kendall compares the change

in concentrations of nutrients with time (slope) for each point and the median slope is calculated

as a summary statistic describing the magnitude of the trend (Johnson et al., 2009; Qian et al.,

2007). On the other hand, GLIMMIX procedure fits the linear lines through the data and slope is

the rate of change in concentration with time at any of two points (SAS, 2008). Therefore, the

differences in the magnitude of slopes appear in the trend analysis in two methods. In general,

Kendall trend analysis determines the trends at a single location. However, GLIMMIX procedure

can be used to compare the mean concentrations at multiple locations as well as has the ability to

include many parameters such as stream discharge, seasonal variation, distance of the mainstem,

and location of the station etc (Goodrich et al., 2009). In this study, GLIMMIX procedure has

shown the significant variations in mean and seasonal variations in P concentrations at different

stations.

In general, flow weighted concentration trends represent the changes due to anthropogenic

activities since the natural changes due to variations in flow are minimized. On the other hand,









flow un-weighted concentrations trends reflect the net effects of all natural and anthropogenic

influences on concentration, allowing for the assessment of nutrient concentrations in streams

relative to water quality standards of Environmental Protection Agency and the condition of

aquatic communities (Sprague and Lorenz, 2009; Johnson et al., 2009). In our study, the flow

weighted and flow un-weighted concentrations showed significantly decreasing trends at Bell

Shoals during 1991-2009. This has suggested that the anthropogenic activities were responsible

for decreasing P trends in the Alafia River Watershed. Similar to the Alafia River Watershed, the

controls on P losses from watersheds have been documented in 50% rivers in the US in last 2-3

decades (Alexander and Smith, 2006). For instance, in Minnesota River, US, Johnson et al.,

(2009) using Seasonal Kendall trend analysis reported the decreasing trends of total P (-2.0% per

year) and DRP (-1.83% per year) during 1976-2003. They attributed the decrease in P

concentrations trends to the conservation programs such as plantation of native grasses, buffer

strips, wetland restoration, and reduction in agricultural activities near the streams. Similarly, in

the Everglades Agricultural Area, FL, Daroub et al. (2009) reported decreasing total P

concentration trends due to implementation of agricultural BMP's during 1992-2002. In contrast

to these watersheds, the Alafia River Watershed has P rich geology and therefore is

commercially mined for P. The discharges of P rich wastewater from the mining operations have

been reduced as a result of greater regulations in the watershed (NPDES, 2009). Further, the

mined land uses have been reclaimed in most of the sub-basins during 1991-2009 resulting in

decreased losses of P from the mined lands.

In the Alafia River Watershed, population has grown from 0.83 million in 1990 to 0.99

million in 2000 and 1.20 million in 2009 (US Census Bureau, 2010). This represents 19.8%

increase during 1990-2000 and 19.7% during 2000-2009. The population growth has resulted in









the land use changes in the watershed. For example, the urban land use (residential and built up)

has increased by 8% (from 12 to 20%) of the watershed during 1990-2007 (SWFWMD, 2007).

This is coupled with a decrease of 5% in forests (from 23 to 19%) and 8% decrease in pastures

(from 19 to 11%) while agricultural land use remained similar (8%) during the study period. The

decreasing total P and DRP concentrations trends suggest that the land use change has not

resulted in increasing P concentrations in the Alafia River Watershed. This contradicts the

previous studies which have suggested the greater losses of P with urbanization (Brett et al.,

2005; Ellison and Brett, 2006). In our study watershed, the constructions of mandatory urban

storm water retention ponds is an important feature which might have played an important role in

P removal through burial of P in retention ponds sediments (Alexander et al., 2000; Bosch, 2008;

Evans et al., 2004). Further, increased regulations on the discharges of P from the wastewaters

and mining activities might have masked the effect of urbanization (Alexander and Smith, 2006).

Summary

The mainstem stations of the Alafia River Water (Alafia and Bell Shoals) had greater mean

total P concentrations (1.01-1.76 mg P L1) than EPA's proposed numeric total P value of 0.739

mg P L-1for the region. Greater total P concentrations at two developed sub-basins (North Prong

and English Creek: 2.2-2.5 mg P L-1) may be due to P rich geology as well the discharge of P

rich wastewater in these sub-basins. Of total P, DRP was dominant (70-90% of total P) than

other P forms (10-30% of total P) probably because of dissolution of P rich minerals in the

watershed. Greater concentrations of P forms in wet than dry season may be due to flushing of P

accumulated due to dissolution ofP minerals. Long-term trend analysis showed decreasing flow

un-weighted and flow weighted concentrations of total P and DRP thereby suggesting that the P

abatement programs such as increased regulations on the wastewaters facilities and reclamation

of mined lands might have resulted in reducing concentrations of P in the Alafia River









Watershed. Despite the decreasing P trends, the concentration of total P was still greater than the

EPA's proposed numeric nutrient criteria (0.739 mg P L-1) for the region. Therefore, the

reduction in P loss from mined lands and wastewater discharges in four sub-basins (North Prong,

English Creek, Turkey Creek, and South Prong) may result in reducing total P concentrations in

the Alafia River Watershed.









Table 4-1. Station characteristics of the Alafia River Watershed


Sub-basin


Alafia
Bell Shoals

English Creek
Turkey Creek
North Prong


South Prong
Fishhawk Creek
-t USGS st


Station Sampling location
Lat Long


Drainage area
km2 % Residential


Land Use in 2007
Built up Agricultural Pasture Forest Mined


Mainstem Stations
2301718 27.87 -82.32 1072 99 17
2301638 27.86 -82.26 974 89 16
Developed
-t 27.93 -82.06 99 9 21
-t 27.91 -82.18 128 13 20
2301000 27.86 -82.13 350 32 18
Undeveloped
2301300 27.86 -82.13 277 26 3
t 27.85 -82.24 70.6 7 11
ation not present


Table 4-2. Long term trends in flow weighted and loads of P forms at Bell Shoals
Parameter Data Flow weighted concentrations Loads
Trend Slope (%) p value Trend Slope p value
Total Phosphorus 1991-2009 Decreasing -1.14 0.041 No trend 0.535
Dissolved Reactive P 1991-2009 Decreasing -1.16 0.016
Other P 1991-2009 No trend 0.561











LAND USE ...


I
O 2 4 8 12 16 South Prong
km r 1 IIk


Alafia River Watershed


Figure 4-1. Location map of the Alafia River Watershed.











A. Total Phosphorus
1991-1999 2000-2009
B A C A B B BC A C A A



_i< r75 percentile
< Median
25 percentile



Bell Shoals Turkey Creek North Prong South Prong Alafia Bell Shoals English Turkey North South Fishhawk
Creek Creek Prong Prong Creek
B. Dissolved Reactive Phosphorus
1991-1999 2000-2009
B A C A B B C B C B A









Bell Shoals Turkey Creek North Prong South Prong Alafia Bell Shoals English Turkey North South Fishhawk
Creek Creek Prong Prong Creek
C. Other Phosphorus
1991-1999 2000-2009
A B B A A A A A A A A



'j~l^^*i


Bell Shoals


Turkey Creek North Prong South Prong Alafia Bell Shoals English Turkey North South Fishhawk
Creek Creek Prong Prong Creek


Figure 4-2. Summary of mean monthly concentrations of total, dissolved reactive, and other
phosphorus forms during 1991-2009 in mainstem (Alafia and Bell Shoals),
developed (English Creek, Turkey Creek, and North Prong), and undeveloped (South
Prong and Fishhawk Creek). Values indicated by different letters are significantly
different according to GLIMMIX procedure at P< 0.05. N in each sub-basin indicates
number of months/observations.













A. Total Nitrogen
Dry Season Wet Season








S120 225 120 207 225 225 74
N= 120 225 120 207 225 225 74


Alafia Bell Shoals English Turkey North
Creek Creek Prong


South Fishhawk Alafia Bell Shoals English Turkey North South Fishhawk
Prong Creek Creek Creek Prong Prong Creek
B. Organic Nitrogen


Dry Season Wet Season











Alafia BellShoals English Turkey North South Fishhawk Alafia BellShoals English Turkey North South Fishhawk
Creek Creek Prong Prong Creek Creek Creek Prong Prong Creek
C. Nitrate-N

Dry Season Wet Season











Alafia Bell Shoals English Turkey North South Fishhawk Alafia Bell Shoals English Turkey North South Fishhawk
Creek Creek Prong Prong Creek Creek Creek Prong Prong Creek
D. Ammonium-N

Dry Season Wet Season











Alafia Bell Shoals English Turkey North South Fishhawk Alafia Bell Shoals English Turkey North South Fishhawk
Creek Creek Prong Prong Creek Creek Creek Prong Prong Creek


Figure 4-3. Seasonal variation in mean monthly concentration of phosphorus forms during
1991-2009. Values indicated by different letters are significantly different according
to GLIMMIX procedure at P< 0.05. Dotted line represents the mean value.


'5 percentile
Median
.5 percentile











Dissolved Reactive P U Other P


|I'l| I I II


.4 r r


Dry Wet


Dry Wet


Alafia I Bell Shoals


Dry Wet

English
Creek


Dry Wet

Turkey
Creek


Dry Wet


Dry Wet


North Prong|South Prong


Figure 4-4. Seasonal variation in contribution of organic, nitrate, and ammonium nitrogen to
total nitrogen during 1991-2009.


100 -

80-
o-

o 60
o
Ln 40 -
-^

20
0


Dry Wet

Fishhawk
Creek




























Mann-Kendall Slope = -36.5% per year, P- 0.008
GLIMMIX Slope =-1 11% per yea 1


Alafia


English Creek N=120


6

IX.*, 0 ---------------------------------------
S2

0 o
S 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
Mann-Kendall Slope = +0.9% per year, P= 0.29 Turkey Creek N= 21
S 4 GLIMMIX Slope =-1.08% per year, P= 0.0001. N
O t


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09


Figure 4-5. Long-term (1991-2009) trends in mean monthly total P concentrations in mainstem
(Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North Prong),
and undeveloped (South Prong and Fishhawk Creek) sub-basins of the Alafia River
Watershed. The dotted line indicates the proposed numeric nutrient (0.79 mg L-1)
criteria for the region (EPA, 2010).


Mann-Kendall Slope = -8.3% per year, P= 0.45
GLIMMIX Slope = -1.09% per year, P =0.003


Numeric nutrient criteria (0.79 mg L1)
A


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
SMann-Kendall Slope = -3.8% per year, P= 0.002 Bell Shoals N= 22,
SGLIMMIX Slope = -1.07% per year, P= 0.001


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
SMann-Kendall Slope = -2.2% per year, P= 0.03 South Prong N= 225
GLIMMIX Slope =-0.93% per year, P= 0.024 t


N= 120


AA^rW












Mann-Kendall Slope = -5.6% per year, P= 0.37 Alafia N= 120
SGLIMMIX Slope = -0.89% per year, P =0.018
3-]




0
91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
8
Mann-Kendall Slope = -4.2% peryear, P= 0.002 North Prong N= 225
GLIMMIX Slope = -1.03% per year, P= 0.0001
6

4

2

0
91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
Mann-Kendall Slope = -2.3% per year, P= 0.09 South Prong N= 225
GLIMMIX Slope =-0.98% per year, P = 0.024



i ... ... L ... -


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09


Figure 4-6. Long-term (1991-2009) trends in mean monthly dissolved reactive P concentrations
in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and
North Prong), and undeveloped (South Prong and Fishhawk Creek) sub-basins of the
Alafia River Watershed.


N= 225


92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09


Mann-Kendall Slope = -34.3% per year, P= 0.009 English Creek
GLIMMIX Slope =-1.23% per year, P= 0.001


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
4
Mann-Kendall Slope = +0.5% per year, P= 0.55 Turkey Creek N= 217
GLIMMIX Slope =+0.99% per year, P= 0.0001
3-

2-

1- V A ^^^^


1


N=120

























91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09
2.0 Mann-Kendall Slope = -15.7% per year, P= 0.12 English Creek N=120
GLIMMIX Slope =-0.99% per year, P= 0.004
1.5 -






C
1.0-

0.5 -

0.0
O 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09

2.0 Mann-Kendall Slope = +2.33% per year, P= 0.34 Turkey Creek N= 217
C GLIMMIX Slope =+0.67% per year, P= 0.0003
O 1.5 -



0.5

0.0
91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09


Mann-Kendall Slope = -1.66% per year, P= 0.44
GLIMMIX Slope =-0.87% per year, P = 0.321


91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09


Figure 4-7. Long-term (1991-2009) trends in mean monthly other phosphorus forms

concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek,

Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk

Creek) sub-basins of the Alafia River Watershed.


South Prong


Mann-Kendall Slope = -1.6% per year, P= 0.38 Bell Shoals
GLIMMIX Slope = -1.06% per year, P= 0.170


Mann-Kendall Slope = -3.64% per year, P= 0.22
GLIMMIX Slope = -0.95% per year, P = 0.017


N= 225















Total Phosphorus

y- 0.056x+0.709, r= 0.33
P>0.05





r
I^-- 2


ii-ii


S01-
10 20 30 40 0


4 -




2-





( 0-
0)
E

o

0)

C
0 4-




2-




0


Dissolved Reactive Phosphorus
4
y= 0.046x+0.580, r= 0.36
P>0.05






2 -
ifw


S'I 0.0 -
10 20 30 40 0
Urban Land Use (%)


Dissolved Reactive Phosphorus
5
y= -0.401x+1.658, r= -0.36
P>0.05
4 -


3 -


2 -


-- 0 -
30 0


10 20
Agricultural Land Use (%)


Other Phosphorus
1.0
y= 0.012x+0.103, r= 0.22
P>0.05
0.8


0.6


0.4


0.2 _
a F


10 20 30 40



Other Phosphorus


10 20


-I 0.0 -
30 0


Figure 4-8. Relationship between percent urban and agricultural land use and phosphorus forms

in different sub-basins (* significantly correlated at P< 0.05). The mean monthly data

of 2001-2009 was correlated with the land use during 2007.











































98


Total Phosphorus

y-0.045x+1.984, r= -0.35
P>0.05







I0

10 20

1 10 20


1.0
y= -0.0028x+0.315, r=-0.29
P>0.05
0.8


0.6


0.4


0.2 j














Total Phosphorus

y- -0.100x+2.843, r= -47
P>0.05







^-^ 2


4




2


T-J
_10

0
E

0


U
0
4




2




0


Dissolved Reactive Phosphorus
y=-0.0905x+2.439, r= -0.50
P>0.05


i


0 10 20
Pasture Land Use (%)


Dissolved Reactive Phosphorus
y= -0.022x+1.667, r= -0.09
P>0.05
4-




2


T


10 20 30 40 10 20 30
Forest Land Use (%)


30 0




-1.0


40 10


Other Phosphorus
y=-0.008x+0.394, r= -0.37
P>0.05









"^T-L,-


10 20 30



Other Phosphorus


20 30 40


Figure. 4-9. Relationship between pasture and forest land use and phosphorus forms in streams

draining different sub-basins (* significantly correlated at P< 0.05).










































99


0 10 20 31



Total Phosphorus

y -0.023x+1.973, r= -0.10
P>0.05




-i



~-{w j-


y= -0.001x+0.2631, r= -0.12
P>0.05
0.8

0.6

0.4

0.2 IF


-









CHAPTER 5
SUMMARY, CONCLUSIONS, AND RECOMMENDATION

The anthropogenic activities such as urbanization and agriculture have increased

concentrations of nitrogen (N) and phosphorus (P) in lakes, rivers, reservoirs, and estuaries

resulting in eutrophication of water bodies. In the US, nearly 60% of 138 estuaries exhibit

moderate to severe eutrophic conditions and >90% rivers have either N or P concentrations

greater than the respective reference levels. In general, anthropogenic influences (urban and

agricultural) in watersheds result in greater nutrient inputs such as fertilizer application to crops,

urban lawns, and septic tanks which may lead to greater losses of nutrients to streams. Another

consequence of changes in land use, primarily in urbanizing watersheds, is the increase in

impervious areas such as rooftops, roadways, parking lots, sidewalks, and driveways. These

impervious areas increase runoff and minimize the biotic uptake processes that immobilize

nutrients in the forest canopy, litter, soils, and organic matter and thereby result in greater losses

of nutrients. Florida is one of the rapidly developing states in the US and has serious water

quality problems with nutrients especially eutrophication of coastal waters. Therefore, it is

important to assess the impact of anthropogenic activities on water quality in waterbodies of

Florida. Very little is known about N and P transport in urban watersheds in Florida, which is

dominated by sandy soils, high ground water table, P rich geology, and altered hydrology due to

construction of stormwater retention ponds to avoid flooding.

In this study, we used monthly collected data (5-19 years for various sites) of stream

water N and P forms for seven sub-basins (19-350 km2) of the Alafia River Watershed (total

drainage area: 1085 km2). In the Hillsborough County where this watershed is located,

population has increased from 0.83 million in 1990 to 0.99 million in 2000 and 1.20 million in

2009. This increase in population has resulted in significant changes in land uses in the


100









watershed. We used, Florida Land Use and Cover Classification Codes (FLUCCCS) at level IV

and grouped land uses into six main categories: residential, built up, agricultural, pasture, forest,

and mined using land use data of three time periods (1990, 1999, and 2007). During 1990-2007,

urban land use (residential and built up) in the watershed increased by 8% (from 12 to 20%)

while forest decreased by 6% (from 25 to 19%) and pasture decreased by 8% (from 19 to 11%).

The agricultural land use remained similar at 8% during 1990-2007. Based on the urban land use

of 2007, we classified the sub-basins into: (1) three developed (18-24% residential; 1-14% built

up; 4-24% agricultural; 0-39% mined) and (2) two undeveloped (3-11% residential; 1-3% built

up; 4-14% agricultural; 0-66% mined). In addition, two mainstem stations draining 89-99% of

the watershed had 16-17% residential, 3% built up, 8% agricultural, and 32-34% mined land

uses in 2007.

Long term monthly collected data showed that at mainstem stations (Alafia and Bell

Shoals), total N concentrations of 1.77-1.91 mg L-1 were similar to EPAs proposed numeric total

N value of 1.79 mg L-1 for the region (EPA, 2010). Total N concentrations were significantly

(P<0.05) correlated with percent urban land use (r=0.83) but not with agricultural, pasture,

mined, and forest land uses (r<0.50) suggesting that urbanization has increased N concentrations

in stream waters. This is further reflected in significant (P<0.05) total N in three developed

(1.67-2.43 mg L-) than two undeveloped (0.84-1.21 mg L-) sub-basins during 2000-2009.

Greater proportion ofNO3-N was observed in developed (53-68% of total N) than undeveloped

sub-basins (25-30% of total N). Concentration of NO3-N was lower in wet than dry season due

to greater biotic uptake and greater denitrification ofNO3-N due to higher temperature in wet

season. In contrast, ON concentrations were greater in wet than dry season probably due to the

greater transport of organic materials (leaves, grass) with more runoff During 1991-2009,


101









concentrations showed increasing trends at the mainstem station thereby indicating that

urbanization has increased total N in the streams. Interestingly, the increased total N

concentration trends were primarily due to increases in ON rather than NO3-N. This suggest

processing ofNO3-N in our watershed and likely conversion to ON in stormwater retention

ponds. In addition, greater runoff generation in urban land uses may enhance the transport of

organic materials such as composts, grass cuttings from urban lawns, deciduous leaves fallen on

the ground. As these are rich sources of ON, decomposition or leaching of N may have resulted

in increased ON concentrations in the watershed. The increased ON concentration trends in this

urbanizing watershed has raised two important questions (1) Is the increase in ON concentration

due to greater ON sources in urban land uses? and (2) Is the ON increase a product of microbial

transformations of NO3-N into ON?. Further studies on the source characterization of the ON

can help in devising the accurate BMP's to control N pollution in the Alafia River Watershed.

Total P concentrations ranged from 1.01 to 1.23 mg P L-1 and were much greater than

EPA's proposed numeric total P value of 0.739 mg P L1 for the region. Of total P, dissolved

reactive P (DRP) was dominant (70-90% of total P) than other P (10-30% of total P) in both

developed and undeveloped sub-basins. None of the P forms were significantly (P< 0.05)

correlated with urban, agricultural, forest, and pasture land use (r<0.50) indicating that the P

concentrations are not controlled by these land uses in the Alafia River Watershed. Two

developed sub-basins had significantly greater total P concentrations (North Prong and English

Creek: 2.18-2.53 mg P L-1) probably due to P rich geology, active mined lands, and discharges

of P rich wastewaters. In all the developed and undeveloped sub-basins, the concentrations of P

forms were greater in wet than dry season. This indicates perhaps the flushing of P from

dissolution and desorption of P from soil minerals and suspension of the stream sediments with


102









greater runoff in wet season might have elevated P concentrations in stream waters. The

decreasing flow weighted and flow un-weighted total P concentration trends indicated that the

anthropogenic activities such as increased regulations on P discharges from mined lands and,

wastewater discharges together with reclamation of mined lands were probably successful in

controlling P pollution in the Alafia River Watershed.

If the EPA proposed numeric total N and P criteria for Florida streams is established, it

will be increasingly difficult to maintain concentrations of total N below 1.798 mg L-1 and total P

below 0.739 mg L-1 in this urban watershed, unless the mechanisms controlling N and P

transport from the landscape are clearly understood and BMPs to control nutrient losses from

watershed are developed and implemented. In the short-term, our results can aid in planning

efforts to reduce N and P concentrations. We suggest that BMPs should be targeted to control N

loss in three developed sub-basins (English Creek, Turkey Creek, and North Prong) that had total

N concentrations of 1.7-2.4 mg L^as these may yield greater reductions in N concentrations at

watershed scale. On the other hand, due to P rich geology and discharges from wastewaters, all

developed and one undeveloped sub-basins had greater total P concentrations (0.8-2.5 mg P L-1)

than EPA proposed numeric value of 0.739 mg P L-1. Therefore, the reduction in P loss from

mined lands and wastewater discharges in four sub-basins (North Prong, English Creek, Turkey

Creek, and South Prong) may result in reducing total P concentrations in stream waters.


103









LIST OF REFERENCES


Abelho, M. 2001. From litterfall to breakdown in streams: a review. Scientific World Journal
1:656-80.

Alexander, R.B., and R.A. Smith. 2006. Trends in the nutrient enrichment of US rivers during
the late 20th century and their relation to changes in probable stream trophic conditions.
Limnol. Oceanogr. 51:639-654.

Alexander, R.B., R.A. Smith, and G.E. Schwarz. 2000. Effect of stream channel size on the
delivery of nitrogen to the Gulf of Mexico. Nature 403:758-761.

Alexander, R. B., J. K. Bohlke, E. W., Boyer, M. B. David, J. W. Harvey, P. J. Mulholland, S. P.
Seitzinger, C. R. Tobias, C. Tonitto, and W. M. Wollheim. 2009. Dynamic modeling of
nitrogen losses in river networks unravels the coupled effects of hydrological and
biogeochemical processes. Biogeochemistry. doi:10.1007/sl0533-008-9274-8.

Anisfeld, S.C., R.T. Barnes, M.A. Altabet, and T.X. Wu. 2007. Isotopic apportionment of
atmospheric and sewage nitrogen sources in two Connecticut rivers. Environ. Sci.
Technol. 41:6363-6369.

Arnold, C.L., and C.J. Gibbons. 1996. Impervious surface coverage The emergence of a key
environmental indicator. J. American Plan. Assoc. 62:243-258.

Boesch, D.F., R.B. Brinsfield, and R.E. Magnien. 2001. Chesapeake Bay eutrophication:
Scientific understanding, ecosystem restoration, and challenges for agriculture. J.
Environ. Qual. 30:303-320.

Bosch, N.S. 2008. The influence of impoundments on riverine nutrient transport: An evaluation
using the Soil and Water Assessment Tool. J. Hydrol.355:131-147.

Boyer, E.W., G.M. Hornberger, K.E. Bencala, and D.M. McKnight. 1997. Response
characteristics of DOC flushing in an alpine catchment. Hydrol. Processes
11:1635-1647.

Boyer, E.W., C.L. Goodale, N.A. Jaworsk, and R.W. Howarth. 2002. Anthropogenic nitrogen
sources and relationships to riverine nitrogen export in the northeastern USA.
Biogeochemistry 57:137-169.

Brett, M.T., G.B. Arhonditsis, S.E. Mueller, D.M. Hartley, J.D. Frodge, and D.E. Funke. 2005.
Non-point-source impacts on stream nutrient concentrations along a forest to urban
gradient. Environ. Manage. 35:330-342.

Bricker, S.B., C.G. Clement, D.E. Pirhalla, S.P. Orlando, and D.R.G. Farrow. 1999. National
Estuarine Eutrophication Assessment. Effect of nutrient enrichment in the nation's
estuaries, NOAA-NOS Special Projects Office pp 71.


104









Carpenter, S.R., N.F. Caraco, D.L. Correll, R.W. Howarth, A.N. Sharpley, and V.H. Smith.
1998. Nonpoint pollution of surface waters with phosphorus and nitrogen. Ecol. Appl.
8:559-568.

Chen, X.X., and C.T. Driscoll. 2009. Watershed Land Use Controls on Chemical Inputs to Lake
Ontario Embayments. J. Environ. Qual. 38:2084-2095.

Conley, D.J., H.W. Paerl, R.W. Howarth, D.F. Boesch, S.P. Seitzinger, K.E. Havens, C.
Lancelot, and G.E. Likens. 2009. Ecology Controlling Eutrophication: Nitrogen and
Phosphorus. Science 323:1014-1015.

Culbert, H., and R. France. 1995. Laboratory prediction of phosphorus release from deciduous
leaves to urban runoff Water Qual. Res. J. Can. 30:243-246.

Dame, R., D. Bushek, D. Allen, A. Lewitus, D. Edwards, E. Koepfler, and L., Gregory. 2002.
Ecosystem response to bivalve density reduction: management implications. Aquatic
Ecology 36: 51-65.

Daroub, S.H., T.A. Lang, O.A. Diaz, and S. Grunwald. 2009. Long-term water quality trends
after implementing best management practices in south Florida. J. Environ. Qual.
38:1683-1693.

Diebel, M.W., J.T. Maxted, D.M. Robertson, S.B. Han, and M.J.v. Zanden. 2009. Landscape
planning for agricultural nonpoint source pollution reduction. III: Assessing phosphorus
and sediment reduction potential. Environ. Manage. 43:69-83.

Dixon, L.K., G.A. Vargo, J.O.R. Johansson, R. Montgomery, and M.B. Neely. 2009. Trends and
explanatory variables for the major phytoplankton groups of two southwestern Florida
estuaries, USA. J. Sea Res. 61:95-102.

Dodds, W.K., and R.M. Oakes. 2006. Controls on nutrients across a prairie stream watershed:
Land use and riparian cover effects. Environ. Manage. 37:634-646.

Dodds, W.K., W.W. Bouska, J.L. Eitzmann, T.J. Pilger, K.L. Pitts, A.J. Riley, J.T. Schloesser,
and D.J. Thornbrugh. 2009. Eutrophication of U.S. freshwaters: analysis of potential
economic damages. Environ. Sci. Technol. 43:12-19.

Edwards, A.C., and P.J.A. Withers. 2008. Transport and delivery of suspended solids, nitrogen
and phosphorus from various sources to freshwaters in the UK. J. Hydrol. 350:144-153.

Edwards, A.C., H. Twist, and G.A. Codd. 2000. Assessing the impact of terrestrially derived
phosphorus on flowing water systems. J. Environ. Qual. 29:117-124.

EPA, 2010. Water Quality Standards for the State of Florida's Lakes and Flowing Waters.
Federal Register. 75:16, January 26, 2010.

Ellison, M.E., and M.T. Brett. 2006. Particulate phosphorus bioavailability as a function of
stream flow and land cover. Water Resour. Res.40:1258-1268.


105









Evans, D.J., P.J. Johnes, and D.S. Lawrence. 2004. Physico-chemical controls on phosphorus
cycling in two lowland streams. Part 2 The sediment phase. Sci. Total Environ.
329:165-182.

Florida Climate Center, 2009. http://www.coaps.fsu.edu/climatecenter/index.shtml. Accessed
on 06, December, 2009.

Florida Department of Health, 2009.
http://www.doh.state.fl.us/environment/programs/EhGis/EhGisDownload.htm. Accessed
on 07, December, 2009.

Florida Lakewatch Program. 2005. Florida Lakewatch Data Book (Hillsborough County).
http://lakewatch.ifas.ufl.edu/2005DataReport/Hillsborough Marion.pdf. Accessed on 1
November, 2010).

Galloway, J.N., R.W. Howarth, A.F. Michaels, S.W. Nixon, J.M. Prospero, and F.J. Dentener.
1996. Nitrogen and phosphorus budgets of the North Atlantic Ocean and its watershed.
Biogeochemistry 35:3-25.

Gilbert, R.O. 1987. Statistical methods for environmental monitoring. John Wiley & Sons, New
York.

Goodrich, B.A., R.D. Koski, and W.R. Jacobi. 2009. Monitoring surface water chemistry near
magnesium chloride dust suppressant treated roads in Colorado. J. Environ. Qual.
38:2373-2381.

Goolsby, D.A., and W.A. Battaglin. 2001. Long-term changes in concentrations and flux of
nitrogen in the Mississippi River Basin, USA. Hydrol. Processes 15:1209-1226.

Granlund, K., A. Raike, P. Ekholm, K. Rankinen, and S. Rekolainen. 2005. Assessment of water
protection targets for agricultural nutrient loading in Finland. J. Hydrol. 304:251-260.

Green, P.A., C.J. Vorosmarty, M. Meybeck, J.N. Galloway, B.J. Peterson, and E.W. Boyer.
2004. Pre-industrial and contemporary fluxes of nitrogen through rivers: a global
assessment based on typology. Biogeochemistry 68:71-105.

Groffman, P.M., N.L. Law, K.T. Belt, L.E. Band, and G.T. Fisher. 2004. Nitrogen fluxes and
retention in urban watershed ecosystems. Ecosystems 7:393-403.

Han, H.J., J.D. Allan, and D. Scavia. 2009. Influence of Climate and Human Activities on the
Relationship between Watershed Nitrogen Input and River Export. Environ. Sci. Technol.
43:1916-1922.

Harrison, J.A., N. Caraco, and S.P. Seitzinger. 2005. Global patterns and sources of dissolved
organic matter export to the coastal zone: Results from a spatially explicit, global model.
Global Biogeochem. Cycles 19.


106









Hayakawa, A., M. Shimizu, K.P. Woli, K. Kuramochi, and R. Hatano. 2006. Evaluating stream
water quality through land use analysis in two grassland catchments: Impact of wetlands
on stream nitrogen concentration. J. Environ. Qual. 35:617-627.

Haygarth, P.M. 2005. Linking landscape sources of phosphorus and sediment to ecological
impacts in surface waters. Sci. Total Environ. 344:1-3.

Haygarth,P.M. and A. N, Sharpley, 2000. Terminology for phosphorus transfer. J. Environ. Qual.
29: 10-15.

Helsel, D.R., and R.M. Hirsch. 1992. Statistical methods in water resources. Elsevier Publ.,
Amsterdam, the Netherlands.

Hirsch, R.M., J.R. Slack, and R.A. Smith. 1982. Techniques of trend analysis for monthly water-
quality data. Water Resour. Res. 18:107-121.

Howarth, R.W., and R. Marino. 2006. Nitrogen as the limiting nutrient for eutrophication in
coastal marine ecosystems: Evolving views over three decades. Limnol. Oceanogr.
51:364-376.

Howarth, R.W., A. Sharpley, and D. Walker. 2002. Sources of nutrient pollution to coastal
waters in the United States: Implications for achieving coastal water quality goals.
Estuaries 25:656-676.

Howarth, R.W., G. Billen, D. Swaney, A. Townsend, N. Jaworski, K. Lajtha, J.A. Downing, R.
Elmgren, N. Caraco, T. Jordan, F. Berendse, J. Freney, V. Kudeyarov, P. Murdoch, and
Z.L. Zhu. 1996. Regional nitrogen budgets and riverine N&P fluxes for the drainages to
the North Atlantic Ocean: Natural and human influences. Biogeochemistry 35:75-139.

Jansson, B.O., and K. Dahlberg. 1999. The environmental status of the Baltic Sea in the 1940s,
today, and in the future. Ambio 28:312-319.

Jarvie, H.P., C. Neal, and P.J.A. Withers. 2006. Sewage-effluent phosphorus: A greater risk to
river eutrophication than agricultural phosphorus? Sci. Total Environ. 360:246-253.

Johnson, H.O., S.C. Gupta, A.V. Vecchia, and F. Zvomuya. 2009. Assessment of water quality
trends in the Minnesota River using non-parametric and parametric methods. J. Environ.
Qual. 38:1018-1030.

Johnson, L.B., C. Richards, G.E. Host, and J.W. Arthur. 1997. Landscape influences on water
chemistry in Midwestern stream ecosystems. Freshwater Biology 37:193-198.

Kaushal, S.S., and W.M. Lewis. 2003. Patterns in the chemical fractionation of organic nitrogen
in Rocky Mountain streams. Ecosystems 6:483-492.

Kaushal, S.S., P.M. Groffman, L.E. Band, C.A. Shields, R.P. Morgan, M.A. Palmer, K.T. Belt,
C.M. Swan, S.E.G. Findlay, and G.T. Fisher. 2008. Interaction between urbanization and


107









climate variability amplifies watershed nitrate export in Maryland. Environ. Sci. Technol.
42:5872-5878.

Koerselman, W., and A.F.M. Meuleman. 1996. The vegetation N:P ratio: A new tool to detect
the nature of nutrient limitation. J. Applied Ecology 33:1441-1450.

Kroeger, K.D., M.L. Cole, J.K. York, and I. Valiela. 2006. Nitrogen loads to estuaries from
waste water plumes: Modeling and isotopic approaches. Ground Water 44:188-200.

Lane, E. (Ed.), 1994. Florida's Geological History and Geological Resources. Special Publication
No. 35. Florida Geological Survey, Tallahassee, FL.

Mainstone, C.P., R.M. Dils, and P.J.A. Withers. 2008. Controlling sediment and phosphorus
transfer to receiving waters A strategic management perspective for England and Wales.
J. Hydrol.350:131-143.

Maxted, J., M. Diebel, and M. Vander Zanden. 2009. Landscape Planning for Agricultural Non-
Point Source Pollution Reduction. II. Balancing Watershed Size, Number of Watersheds,
and Implementation Effort. Environ. Manage. 43:60-68.

McDowell, W.H., A.H. Magill, J.A. Aitkenhead-Peterson, J.D. Aber, J.L. Merriam, and S.S.
Kaushal. 2004. Effects of chronic nitrogen amendment on dissolved organic matter and
inorganic nitrogen in soil solution. Forest Ecology and Management 196:29-41.

Meals, D.W. 1996. Watershed-scale response to agricultural diffuse pollution control programs
in Vermont, USA. Water Sci. Technol. 33:197-204.

Mulliss, R.M., D.M. Revitt, and R.B. Shutes. 1996. The impacts of urban discharges on the
hydrology and water quality of an urban watercourse. Sci. Total Environ. 190:385-390.

National Pollution Discharge Elimination, Systems, 2009. http://cfpub.epa.gov/npdes/. Accessed
on 09, December, 2009.

Omernik, J.M. 1977. Nonpoint source-stream nutrient level relationships: a nationwide study.
Nonpoint source-stream nutrient level relationships: a nationwide study. 12: 151.

Pacini, N., and R. Gachter. 1999. Speciation of riverine particulate phosphorus during rain
events. Biogeochemistry 47:87-109.

Paerl, H.W. 2009. Controlling eutrophication along the freshwater-marine continuum: dual
nutrient (N and P) reductions are essential. Estuaries and Coasts 32:593-601.

Paul, M.J., and J.L. Meyer. 2001. Streams in the urban landscape. Annual Review of Ecology
and Systematics 32:333-365.

Peebles, E., D. Hollander, S. Locker, P. Swarzenski and G. Brooks. 2009. Areal Extent, Source
and Ecological Status of Organic Sediment Accumulation in Safety Harbor, Tampa Bay.
Technical Report #07-09.


108









Pellerin, B.A., S.S. Kaushal, and W.H. McDowell. 2006. Does anthropogenic nitrogen
enrichment increase organic nitrogen concentrations in runoff from forested and human-
dominated watersheds? Ecosystems 9:852-864.

Pellerin, B.A., W.M. Wollheim, C.S. Hopkinson, W.H. McDowell, M.R. Williams, C.J.
Vorosmarty, and M.L. Daley. 2004. Role of wetlands and developed land use on
dissolved organic nitrogen concentrations and DON/TDN in northeastern US rivers and
streams. Limnol. Oceanogr. 49:910-918.

Peterson, B.J., W.M. Wollheim, P.J. Mulholland, J.R. Webster, J.L. Meyer, J.L. Tank, E. Marti,
W.B. Bowden, H.M. Valett, A.E. Hershey, W.H. McDowell, W.K. Dodds, S.K.
Hamilton, S. Gregory, and D.D. Morrall. 2001. Control of nitrogen export from
watersheds by headwater streams. Science 292:86-90.

Poe, A, K. Hackett, S. Janicki, R. Pribble, and S. Janicki. 2005. Estimates of total nitrogen, total
phosphorous, and total suspended solids to Tampa Bay, Florida. Technical Report #02-
05.

Qian, Y., K.W. Migliaccio, Y. Wan, Y.C. Li, and D. Chin. 2007. Seasonality of selected surface
water constituents in the Indian river lagoon, Florida. J. Environ. Qual. 36:416-425.

Redfield A.C., 1934. On the proportions of organic derivations in sea water and their relation to
the composition of plankton. James Johnstone Memorial Volume. (ed. R.J. Daniel).
University Press of Liverpool, 177-192.

Rhodes, A.L., R.M. Newton, and A. Pufall. 2001. Influences of land use on water quality of a
diverse new England watershed. Environ. Sci. Technol.35:3640-3645.

Richards, R.P., and J. Holloway. 1987. Monte-carlo studies of sampling strategies for estimating
tributary loads. Water Resour. Res. 23:1939-1948.

Richards, R. P., D. B. Baker, J. P. Crumrine, J. W. Kramer, D. E. Ewing, and B. J. Merryfield.
2008. Thirty-year trends in suspended sediment in seven Lake Erie tributaries. J. Environ.
Qual. 37:1894-1908.

Richards, R.P., D.B. Baker, J.P. Crumrine, J.W. Kramer, D.E. Ewing, and B.J. Merryfield. 2008.
Thirty-year trends in suspended sediment in seven Lake Erie tributaries. J. Environ. Qual.
37:1894-1908.

Royer, T.V., M.B. David, and L.E. Gentry. 2006. Timing of riverine export of nitrate and
phosphorus from agricultural watersheds in Illinois: Implications for reducing nutrient
loading to the Mississippi River. Environ. Sci. Technol.40:4126-4131.

Russell, M.J., D.E. Weller, T.E. Jordan, K.J. Sigwart, and K.J. Sullivan. 2008. Net anthropogenic
phosphorus inputs: spatial and temporal variability in the Chesapeake Bay region.
Biogeochemistry 88:285-304.


109









SAS Institute. 2008. SAS/STAT 9.1. The Glimmix Procedure. User's guide. SAS Inst., Cary,
NC.

Schertz, T.L., R.B. Alexander, and D.J. Ohe. 1991. The computer program Estimate Trend
(ESTREND), a system for the detection of trends in water quality data. U.S. Geological
Survey Water- Resources Investigations Rep. 91-4040. U.S. Geological Survey, Reston,
VA.

Scott, D., J. Harvey, R. Alexander, and G. Schwarz. 2007. Dominance of organic nitrogen from
headwater streams to large rivers across the conterminous United States. Global
Biogeochem. Cycles 21:1003.

Seitzinger, S.P., R.V. Styles, E.W. Boyer, R.B. Alexander, G. Billen, R. Howarth, B. Mayer, and
N. van Breemen. 2002. Nitrogen retention in rivers: Model development and application
to watersheds in the northeastern USA. Biogeochemistry 57:199-237.

Sen, P.K. 1968. Estimates of the regression coefficient based on Kendall's tau. J. Am. Stat.
Assoc. 63:1379-1389.

Setaro, F.V., and J.M. Melack. 1984. Responses of phytoplankton to experimental nutrient
enrichment in an amazon floodplain lake. Limnol. Oceanogr. 29:972-984.

Sharpley, A.N., and R.G. Menzel. 1987. The impact of soil and fertilizer phosphorus on the
environment. Adv. Agron. 41:297-324.

Southwest Florida Water Management District (SWFWMD). 2007. The Determination of
Minimum Flows for the Lower Alafia River Estuary. Southwest Florida Water
Management District. Brooksville, Florida.

Sprague, L.A., and D.L. Lorenz. 2009. Regional Nutrient Trends in Streams and Rivers of the
United States, 1993-2003. Environ. Sci. Technol.43:3430-3435.

Stanley, E.H., and J.T. Maxted. 2008. Changes in the dissolved nitrogen pool across land cover
gradients in Wisconsin streams. Ecol. Appl.18:1579-1590.

Stone, M., and I.G. Droppo. 1994. In-channel surficial fine-grained sediment laminae .2.
chemical characteristics and implications for contaminant transport in fluvial systems.
Hydrol. Process. 8:113-124.

Svendsen, L.M., B. Kronvang, P. Kristensen, and P. Graesbol. 1995. Dynamics of phosphorus-
compounds in a lowland river system importance of retention and nonpoint sources.
Hydrol. Process. 9:119-142.

Tampa Bay Estuary Program, 2009. http://www.tbep.org/. Accessed on 10 December, 2009.

Toor, G. S., S. Hunger, D. J. Peak, T. J. Sims, and D. L. Sparks, 2004. Advances in the
characterization of phosphorus in organic wastes: environmental and agronomic
applications. Adv. Agron. 89: 1-65.


110









Tufford, D.L., H.N. McKellar, and J.R. Hussey. 1998. In-stream nonpoint source nutrient
prediction with land-use proximity and seasonality. J. Environ. Qual. 27:100-111.

Turner, R.E., and N.N. Rabalais. 2003. Linking landscape and water quality in the Mississippi
river basin for 200 years. Bioscience 53:563-572.

United States Census Bureau, 2009. http://quickfacts.census.gov/qfd/states/12/12057.html.
Accessed on 09, December, 2009.

United States Department of Agriculture, 2006. Conservation reserve program: summary and
enrollment statistics-2005. Farm Service Agency, Washington, DC.

United States Department of Agriculture, 2010. National resource conservation service.
http://soils.usda.gov/survey/printedsurveys/state.asp? state=Florida&abbr=FL. Accessed
on 06, June, 2010.

United States Environmental Protection Agency (2002). A review of statewide watershed
management approaches. Office of Water, Washington, DC.

United States Environmental Protection Agency (1986). Quality criteria for water 1986 and
update no. 1 and update no. 2 (May 1, 1987): Washington, D.C., EPA-440/5-86-001.

Vaithiyanathan, P., and D.L. Correll. 1992. The Rhode river watershed phosphorus distribution
and export in forest and agricultural soils. J. Environ. Qual. 21:280-288.

van Kessel, C., T. Clough, and J.W. van Groenigen. 2009. Dissolved Organic Nitrogen: An
Overlooked Pathway of Nitrogen Loss from Agricultural Systems? J. Environ. Qual.
38:393-401.

Verseveld, W.J.v., J.J. McDonnell, and K. Lajtha. 2009. The role of hillslope hydrology in
controlling nutrient loss. J. Hydrol. (Amsterdam) 367:177-187.

Vidon, P., L.P. Tedesco, D.L. Pascual, M.A. Campbell, L.R. Casey, J. Wilson, and M. Gray.
2008. Seasonal changes in stream water quality along an agricultural/urban land-use
gradient. Proceedings of the Indiana Academy of Science 117:107-123.

Wahl, M.H., H.N. McKellar, and T.M. Williams. 1997. Patterns of nutrient loading in forested
and urbanized coastal streams. Journal of Experimental Marine Biol. Ecol. 213:111-131.

Wallbrink, P.J., C.E. Martin, and C.J. Wilson. 2003. Quantifying the contributions of sediment,
sediment-P and fertiliser-P from forested, cultivated and pasture areas at the landuse and
catchment scale using fallout radionuclides and geochemistry. Soil Tillage Res.69:53-68.

Wear, D. and J. Greis, 2002. Southern forest resource assessment: summary of findings. J.
Forestry 100:6-14.


111









Weston, N.B., J.T. Hollibaugh, and S.B. Joye. 2009. Population growth away from the coastal
zone: Thirty years of land use change and nutrient export in the Altamaha River, GA. Sci.
Total Environ. 407:3347-3356.

Wetzel, R.G. 2001. Fundamental processes within natural and constructed wetland ecosystems:
short-term versus long-term objectives. Water Sci. Technol. 44:1-8.

Zaimes, G.N., R.C. Schultz, and T.M. Isenhart. 2008. Streambank soil and phosphorus losses
under different riparian land-uses in Iowa. J. Am. Water Resour. Assoc. 44:935-947.


112









BIOGRAPHICAL SKETCH

Kamaljit was born in Northwest India (Punjab). The eldest of the three children, he spent

earlier period of his life in Punjab. After his bachelor's at Punjab Agricultural University, India

in 2005, he completed his master's in soil science and agricultural chemistry at University of

Agricultural Sciences, Bangalore, India. In August 2008, he started another master's program in

soil and water sciences under the supervision of Dr. Gurpal Toor at the Gulf Coast Research and

Education Center-Wimauma, University of Florida. Kamaljit received his master's degree from

the University of Florida in the summer 2010.


113





PAGE 1

1 NITORGEN AND PHOSPHORUS TRANSPORT IN AN URBAN WATERSHED By KAMALJIT KAMALJIT A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2010

PAGE 2

2 2010 Kamaljit Kamaljit

PAGE 3

3 This work is dedicated to my mom

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank my Supervisory Committee for their patience and guidance. Special appreciation goes to Dr. Gurpal Toor, the chair of my supervisory committee, for his intellectual guidance, consistent support, and endless efforts to accomplish this project. Dr. Patrick Inglett is thanke d for his critical reviews of chapters Additionally, unique gratitude goes to Dr. Craig Stanley for his interests and invaluable suggestions during the period of study. I would like to thank Mr. Butch Bradley; former laboratory manager at the Soil and Wa ter Q uality L aboratory and my fellow gradua te students : Lu Han, Michael Miyitta h, and Maninder Chahal for their help in the laboratory analysis I express my greatest appreciation to my mother, brother, and sister who have continually supported all my drea ms with patience. Last, and most importantly, I thank my fiance Aparna, for her constant encouragement and for standing by me in all my academic endeavors. I could not end my acknowledgements without recognizing that ultimately it has been by the grace o f God, and I most gratefully submit my thanks and praise to Him for any good that comes into my life.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................................... 4 LIST OF TABLES ................................................................................................................................ 8 LIST OF FIGURES .............................................................................................................................. 9 ABSTRACT ........................................................................................................................................ 11 CHAPTER 1 INTRODUCTION ....................................................................................................................... 13 Eutrophica tion ............................................................................................................................. 13 Nitrogen and Phosphorus in the Environment ........................................................................... 14 Nitrogen ................................................................................................................................ 14 Phosphorus ........................................................................................................................... 14 Sources of Nitrogen and Phosphorus in Watersheds ................................................................ 15 Research Objectives .................................................................................................................... 17 Objective 1. Evaluation of Nitrogen Concentrations in Different Subbasins of the Alafia River Watershed. .................................................................................................. 18 Objective 2. Evaluation of Phosphorus Concentrations in Different Subbasins of the Alafia River Watershed. ............................................................................................ 18 2 STUDY SITE DESCRIPTION .................................................................................................. 20 Location ....................................................................................................................................... 20 Climate ......................................................................................................................................... 20 Sub -basins of the Alafia River Watershed ................................................................................. 20 Developed Subbasins ......................................................................................................... 21 Turkey Creek ................................................................................................................ 21 English Creek ............................................................................................................... 21 North Prong .................................................................................................................. 22 Undeveloped Sub-basins ..................................................................................................... 23 South Prong .................................................................................................................. 23 Fishhawk Creek ............................................................................................................ 23 3 NITROGEN TRANSPORT IN AN URBAN WATERSHED ................................................. 32 Abstract ........................................................................................................................................ 32 Introduction ................................................................................................................................. 33 Materials and Methods ................................................................................................................ 37 Study Site Description ......................................................................................................... 37 Data Colle ction .................................................................................................................... 37 Stream -water Collection and Analysis ............................................................................... 37

PAGE 6

6 Statistical Analysis ............................................................................................................... 38 Results .......................................................................................................................................... 39 Chemical Characteristics of Stream Waters ....................................................................... 3 9 Concentrations of Nitrogen Forms in Streams Draining Different Sub basins ............... 40 Seasonal Variations in Chemical Characteristics in Stream Waters Draining Different Sub -basins ........................................................................................................ 42 Seasonal Variations in Concentrations of Nitrogen Forms in Streams Draining Different Sub -basins ........................................................................................................ 43 Lo ng Term Trends in Flow Un-weighted Nitrogen Forms in Streams Draining Different Sub -basins ........................................................................................................ 44 Total Nitrogen .............................................................................................................. 44 Organic Nitrogen .......................................................................................................... 45 Nitrate Nitrogen ............................................................................................................ 45 Ammonium Nitroge n ................................................................................................... 46 Long Term Trends in Flow Weighted Nitrogen Forms in Streams Draining at Mainstem Station ............................................................................................................. 46 Long Term Trends in Nitrogen Loads at Mainstem Station ............................................. 46 Relationship between Land Use and Nitrogen Forms ....................................................... 47 Discussion .................................................................................................................................... 47 Influence of Land Uses on Total Nitrogen Concentrations in Stream Waters ................ 47 Land Uses and Forms of Nitrogen Concentrations in Stream Waters .............................. 49 Seasonal Impacts on Nitrogen Forms in Stream Waters ................................................... 50 Long Term Trends in Nitrogen Concentrations ................................................................. 51 Summary ...................................................................................................................................... 57 4 PHOSPHORUS TRANSPORT IN AN URBAN WATERSHED ........................................... 71 Abstract ........................................................................................................................................ 71 Introduction ................................................................................................................................. 72 Materials and Methods ................................................................................................................ 75 Study Site Description ......................................................................................................... 75 Data Colle ction .................................................................................................................... 76 Stream -water Collection and Analysis ............................................................................... 76 Statistical Analysis ............................................................................................................... 76 Results .......................................................................................................................................... 78 Chemical Characteristics of Stream Waters ....................................................................... 78 Concentrations of Phosphorus Forms in Streams Draining Different Subbasins .......... 78 Seasonal Variation in Phosphorus Concentrations in Streams Draining Different Sub -basins ......................................................................................................................... 80 Long Term Trends in Concentrations of Flow Un-weighted Phosphorus Forms in Streams Draining Different Sub -basins .......................................................................... 81 Long Term Trends in Flow Weighted Concentrations at Mainstem Station ................... 83 Long Term Trends in Phosphorus Loads at Mainst em Station ......................................... 83 Discussion .................................................................................................................................... 83 Land Use Impacts on Phosphorus Concentrations in Streams .......................................... 83 Seasonal Impact on Phosphorus Concentrations in Streams ............................................ 85 Long Term Trends in Concentration of Phosphorus Forms in Stream Waters ................ 86

PAGE 7

7 Summary ...................................................................................................................................... 88 5 SUMMARY, CONCLUSIONS, AND RECOMMENDATION ........................................... 100 LIST OF REFERENCES ................................................................................................................. 104 BIOGRAPHICAL SKETCH ........................................................................................................... 113

PAGE 8

8 LIST OF TABLES Table page 2 1 Station characteristics and associated land uses in the various sub -basins on the Alafia River Watershed .......................................................................................................... 24 2 2 Grouping of FLUCCS codes into major land uses. .............................................................. 25 3 1 Station characteristics of the Alafia River Watershed ......................................................... 58 3 2 Long -term trends in flow weighted and loads of N forms at Bell Shoals .......................... 58 4 1 Station characteristics of the Alafia River Watershed ......................................................... 90 4 2 Long -term trends in flow weighted and loads of P forms at Bell Shoals ........................... 90

PAGE 9

9 LIST OF FIGURES Figure page 1 1 Estimated sources of nitrogen in the Tampa Bay. ................................................................ 19 2 1 Location map and land uses in the various sub -basins of the Alafia River Watershed. .... 26 2 2 Land use in the Turkey Creek subbasin in 1990, 1999, and 2007. .................................... 27 2 3 Land use in the English Creek sub-basin in 1990, 1999, and 2007. ................................... 28 2 4 Land use in the North Prong sub-basin in 1990, 1999, and 2007. ...................................... 29 2 5 Land use in the South Prong sub-basin in 1990, 1999, and 2007. ...................................... 30 2 6 Land use in the Fishhawk Creek sub-basin in 1990, 1999, and 2007. ................................ 31 3 1 Location map of the Alafia River Watershed. ...................................................................... 59 3 2 Chemical characteristics of the stream waters during two time periods from 1991 to 2009.. ....................................................................................................................................... 60 3 3 Summary of mean monthly concentrations of total, organic, nitrate, and ammonium nitrogen during 1991 2009 in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk Creek) of the Alafia River Watershed .. ............................................................... 61 3 4 Seasonal variation in chemical chara cteristics of the stream waters during 1991 2009.. ....................................................................................................................................... 62 3 5 Seasonal variation in mean monthly concentration of nitrogen forms during 1991 2009.. ....................................................................................................................................... 63 3 6 Seasonal variation in proportion of organic, nitrate, and ammonium nitrogen during 1991 2009. ............................................................................................................................. 64 3 7 Long -term (1991 2009) trends in monthly flow un-weighted total N concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk Creek) sub-basins of the Alafia River Watershed. ........................................................................................................ 65 3 8 Long -term (1991 2009) trends in monthly flow un-weighted organic N concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk Creek) sub -basins of the Alafia Rive r Watershed ............................................................... 66 3 9 Long -term (1991 2009) trends in monthly nitrate N concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North

PAGE 10

10 Prong), and undeveloped (South Prong and Fishhawk Creek) sub-basins of the Alafia River Watershed. .................................................................................................................... 67 3 10 Long -term (1991 2009) trends in monthly flow un-weighted ammonium N concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk Creek) sub -basins of the Alafia River Watershed. ............................................................... 68 3 11 Relationship between percent urban and agricultural land use and nitrogen forms in different sub -basins. ............................................................................................................... 69 3 12 Relationship between pasture and forest land use and nitrogen forms in streams draining different sub -basins ................................................................................................. 70 4 1 Location map of the Alafia River Watershed. ...................................................................... 91 4 2 Summary of mean monthly concentrations of total, dissolved reactive, and other phosphorus forms during 1991 2009 in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk Creek) subbasins of the Alafia River Watershed ............................. 92 4 3 Seasonal variation in mean monthly concentration of phosphorus forms during 1991 2009. ............................................................................................................................. 93 4 4 Seasonal variation in contribution of organic, nitrate, and ammonium nitrogen to total nitrogen during 1991 2009. .......................................................................................... 94 4 5 Long -term (1991 2009) trends in mean monthly total P concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk Creek) sub-basins of the Alafia River Watershed.. ................................................................................................................... 95 4 6 Long -term (1991 2009) trends in mean monthly dissolved reactive P concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk Creek) sub-basins of the Alafia River Watershed. ........................................................................................................ 96 4 7 Long -term (1991 2009) trends in mean monthly other phosphorus forms concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk Creek) sub -basins of the Alafia River Watershed. ............................................................... 97 4 8 Relationship between percent urban and agricultural land use and phosphorus forms in different sub basins. ........................................................................................................... 98 4 9 Relationship between pasture and forest land use and phosphorus forms in streams draining different sub -basins. ................................................................................................ 99

PAGE 11

11 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Thesis of Master o f Science NITORGEN AND PHOSPHORUS TRANSPORT IN AN URBAN WATERSHED By Kamaljit Kamaljit August 2010 Chair: Gurpal Toor Major: Soil and Water Science Non -point source pollution is the dominant pathway of nitrogen (N) and phosphorus (P) transport in agricultural urbanized and rapidly urbanizing watersheds. We used monthly c oncentrations data of inorgan ic and organic forms of N and P in stream waters draining different sub -basins, ranging in size from 19 to 350 km2, of the Alafia River Watershed (total drainage area: 1085 km2), which ultimately drains to Tamp a Bay Estuary to understand N and P transport The sub basins were classified based on the percentage of urban land use as three developed (18 24% residential 1 14% built up ) and two undeveloped (3 11% residential 1 3% built up). Urban l and use at t wo mainstem stations that drained 80 99% of the watershed was 16 17% residential and 3% built up. D uring 1991 2009, t otal N concentrations ranged from 0.8 to 2.4 mg L1 and were greatest in stream waters draini n g developed (1. 7 2.4 mg L1) than undeveloped ( 0.8 1. 2 mg L1) sub basins. Inorganic N (primarily NO3N) was the dominant form in streams draining developed sub -basins while organic N was greater in streams draining undeveloped sub-basins. Total P concentrations ranged from 0. 6 to 3.9 mg L1 and were n ot different among developed (0.8 3.9 mg L1) and undeveloped (0. 6 0. 9 mg L1) sub -basins. Of total P, 70 90% was dissolved reactive P while other P forms were 1 0 30% of total P in both developed and undeveloped subbasins. The increasing total N and decre asing total P

PAGE 12

12 concentrations trends at the mainstem station draining 89% of the watershed over the 19-year period suggests that the development of the watershed resulted in increasing N but not P concentrations in streams. We suggest that the BMPs to reduce N loss from urban land uses in three developed sub basins (with total N of 1.7 2.4 mg L1) may yield greater reductions in N concentrations at watershed outlet (i.e. mainstem) to achieve EPA proposed numeric criteria of total N concentration of 1.798 mg L1. On the other hand, due to P rich geology and discharge from the wastewaters, most devel oped and undeveloped sub basins had greater total P concentrations (0.8 3.9 mg P L1) than EPA proposed numeric total P value of 0.739 mg L1 indicating that B MP s should focus on reducing P loss from phos phate rock mined sub-basins and reduce P input s from wastewater

PAGE 13

13 CHAPTER 1 INTRODUCTION Eutrophication Eutrophication is a broad term used to describe enhanced phytoplankton growth in water bodies such as lakes, rivers, reservoirs, and estuaries that receive excess nitrogen (N) and phosphorus (P) from the landscape (Jansson and Dahlberg, 1999; Paerl, 2009) The consequences of eutrophication include hypoxia, acidification of natural waters, de gradation of coastal waters including increased episodes of noxious algal blooms, and reductions in aquatic macrophyte communities often leading to substantial shifts in ecosystem structure and function (Carpenter et al., 1998; Dodds et al., 2009) In the US, eutrophication is one of the greatest threats to the health of the estuaries. For example, Bricker et al. (1999) in their assessment of 138 estuaries reported that nearly 60% of estuaries exhibited moderate to severe eutrophic conditions. In Florida, threshold conce ntrations of 1.20 1.79 mg N L1 and 0.107 0.739 mg P L1 have been proposed for stream waters in three of four regions of Florida (EPA, 2010). The phytoplankton growth in waterbodies is dependent upon the N: P ratio. For example, total N: P ratio of 16:1 is suggested for optimum phytoplankt on growth, termed as Redfield Ratio (Redfield, 1934). An N: P ratio of <16:1 is indicative of N limitation while >16:1 indicates P limitation. In a review from 40 studies, Koerselman and Meuleman, (1996) reported that at an N: P ratio of >16, P would be a limiting nutrient and at N:P <14, N would be limiting, and at intermediate values (14 16) either N and/o r P would be limiting nutrients for phytoplankton growth. In the Tampa Bay estuary, monthly water quality concentrations data from 1981 2004 showed that the N: P ratio in the stream water s was about 5:1 suggesting that this water body is N limited (Dixon et al., 2009) Further it has been suggested that the loss of seagrass beds in the Tampa Bay estuary is a direct consequence of N loading to the Bay from several point and non

PAGE 14

14 point sources (TBEP, 2009). Therefore, source control on N needs greater attention than P, for contro lling eutrophication in the Tampa Bay. Although, recent research has suggested that controls on both N and P transport might be needed to control eutrophication in freshwater marine continuum (Conley et al., 2009; Paerl, 2009) Nitrogen and Phosphorus in the Environment Nitrogen The largest global pool of N exists as dinitrogen gas (N2) comprising up to 78% in the lithosphere. However, only specialized microbes and cyanob acteria with the enzyme nitrogenase can directly use N2 via N fixation, while for >99% of the organisms, N2 is made available by inorganic N fertilizers using Haber Bosch process where N2 is converted to ammonia (NH3). Living organisms utilize inorganic N in the metabolic processes and convert it into organic N (ON) forms such as amino acids, proteins, and nucleic acids. After the organisms die, microorganisms break down ON to ammonium (NH4 +) which can be oxidized to NO3 via nitrification. Finally, the denitrification process, in which micro organisms oxidize organic matter using NO3 as electron acceptor under reduced conditions close the N cycle by converting NO3 back into N2 (Galloway et al., 1996) Therefore, in different steps of N cycle, NO3 -, NH4 +, and ON forms are either produced or consumed while the excess amount of these forms at each step has the potentia l to be transported to waterbodies resulting in water quality deterioration. Phosphorus Like N, P in waterbodies exists in several combinations of organic and inorganic forms. Haygarth and Sharpley (2000) suggested a physicochemical classification (i.e. f iltration and chemical methodology) to differentiate inorganic and organic P forms in water. According to this classification, P can be divided into two main forms: dissolved (<0.45 m) and particulate (>0.45 m). Dissolved P can be further divided into di ssolved reactive P (DRP: orthophosphate)

PAGE 15

15 and dissolved unreactive P (DUP: organic P forms such as sugar phosphates, mononucleotides, DNA, RNA, and phospholipids). Similarly, particulate P can be divided into particulate reactive P (PRP: P sorbed on sediments, Fe, Al, or Ca oxides) and particulate unreactive P (PUP: P sorbed on mineral humic acid complexes) (Toor et al., 2004). D issolved and p articulate P forms in water bodies change from one form to another in response to a variety of environmental and biol ogical responses. For example, microbial decomposition or chemical desorption can convert P from particulate to dissolved forms. Similarly organisms can take up dissolved P and transform them into particulate P forms. As a result, P in the waterbodies is p resent in organic and inorganic forms and is continually recycled. Sources of Nitrogen and Phosphorus in Watersheds Anthropogenic activities such as application of fertilizers, manures, industrial effluents, and wastewater discharge are the major known so urces of N and P in watersheds (Anisfeld et al., 2007; Russell et al., 2008) Therefore, the inputs of N and P are greater in human dominated land uses (agricultural and urban) as compared to relatively undeve loped land areas such as natural forests (Boyer et al., 2002; Kaushal et al., 2008; Russell et al., 2008) Nutrient input sources can be divided into point sources such as wastewater and industrial ef fluents and non -point sources such as runoff and leaching from urban and agricultural areas. With the implementation of the Clean Water Act in the late 20th century, N and P concentrations from point sources have been substantially reduced in the US (Howarth et al., 2002) However, non-point source pollution is dominant in most of t he watersheds in the US and elsewhere. Non -point source pollution is difficult to control because the pollution sources cannot be attributed to one particular discharge location but rather to a diffused landscape (Rhodes et al., 2001) For example, since 1987, the United States Department of Agriculture Conservation Reserve Program (CRP) has distributed $29.7 billion to agricultural land owners to

PAGE 16

16 implement conservation practices to reduce soil loss, restore wetlands, and conserve fore sted areas (USDA, 2006). However, these conservation measures showed a little evidence of improvements in stream water quality at broad spatial scales as a greater emphasis of this program was to reduce soil erosion to control nutrient losses, which was no t successful in controlling dissolved N losses (Boesch et al., 2001; Meals, 1996) Secondly, it was assumed that all areas in the landscape contribute uniformly to nutrient loads, which resulted in less favorable outcomes of reducing nutrient losses to water bodies. Recent research has provided insights about cont ribution of various non -point sources to nutrient loading in watersheds. For example, Poe et al. (2006) estimated that storm water runoff contributes 63% of annual N loads in the Tampa Bay (Fig. 1 1). They further estimated that in the storm water runoff, the residential areas were the major N contributors (20%) followed by pasture/rangelands (15%), intensive agriculture (12%), and mining lands (6%). The second most important source of N in Tampa Bay is atmospheric deposition (21%), while the contribution of point sources such as domestic wastewater (9%) and industrial wastewaters (3%) is comparatively lower than the non -point sources. The reduction of non -point source pollution is urgently needed to protect and conserve water resources (USEPA, 2002). In cas e of P, such a detailed analysis of various sources is lacking however it can be construed that the contribution of different land uses to storm water runoff may be different, with higher contribution of P from mined lands and wastewater discharges. In a ddition to different sources in the non -point category, we also know that in each watershed, there are hot spot areas, termed as variable source areas, which contribute a majority of nutrient losses (Poe et al., 2006; Diebel, 2009). Therefore, a first st ep in controlling non -point source pollution is to develop a quantitative understanding of their sources (i. e. hot spot areas) in the landscape followed by using best

PAGE 17

17 management practices (BMPs) to control nutrient losses from these areas (Diebel et al., 2009; Maxted et al., 2009) For example, Diebel et al. (2009) reported that targeting 10% watersheds in Wisconsin, US decreased total P loads by 20% for the entire state. Therefore, the conservation programs targeting the hot spot areas present an effec tive way to control nutrient losses and improve water quality in a watershed while using less resources rather than attempting to use BMPs for an entire watershed. Secondly, understanding how land uses impact nutrient losses can help to unravel mechanisms of nutrient transport, which can lead to fine tune BMPs to reduce nutrient losses from land to water and protect water resources. Research Objectives In the Southern US, population is anticipated to increase from approximately 8 million in 1992 to 22 mil lion in 2020 and 33 million in 2040 (Wear, 2002). Florida is one of the rapidly developing states in the US and has serious water quality problems such as eutrophication of coastal waters (Dame et al., 2002). Therefore, it is important to assess the impact of anthropogenic activities on water quality of coastal waters. Very little is known about N and P fate and transport in urban watersheds in Florida, which have a high pr oportion of sandy soils, high ground water table, and altered hydrology due to storm water retention ponds. The Alafia River Watershed (1085 km2) which drains into the Tampa Bay estuary was our study site to understand the N and P transport as several years of historic water quality data was available. Secondly, this watershed represents a typical urbanizing watershed in the region with diverse mix of urban, agricultural, and mined land uses. The main objectives of this research are presented below along with specific aims for each objective.

PAGE 18

18 Objective 1. Evaluation of Nitrogen Concentrati ons in Different Subbasins of the Alafia River Watershed. Aim 1a. Determine how different sub-basins influence concentrations of inorganic and organic N forms Aim 1b. Evaluate the influence of low (dry season) and high (wet season) flow conditions on str eam N concentrations in different sub basins. Aim 1c. Determine the long term trends of N concentrations in different sub-basins. Objective 2. Evaluation of Phosphorus Concentrations in Different Subbasins of the Alafia River Watershed. Aim 2a. Determine how different sub -basins influence concentrations of P forms Aim 2b. Evaluate the influence of low (dry season) flow and high (wet season) flow conditions on stream P concentrations in different sub-basins. Aim 2c. Determine the long term trend s of P concentrations in different sub -basins.

PAGE 19

19 Figure 1 1. Estimated sources of nitrogen in the Tamp a Bay (Adapted from Poe et al., 2006).

PAGE 20

20 CHAPTER 2 STUDY SITE DESCRIPTI ON Location Alafia River Watershed is located in the central Florida and drains 1085 km2 of land area (Fig. 2 1). The headwaters of the Alafia River originate from the swamp and prairie lands of Polk County and extend 38.6 km long flowing west into lower Hillsborough Bay, ultimately discharging into Tampa Bay Estuary (SWFWMD, 2 007). The soils in the watershed are sandy, with moderate to slow infiltration and are dominated by Myakka, Winder, Zolfo, Lake, and Chandler soil groups (USDA, 2010). Climate The climate in the area is humid subtropical, with an annual mean temperature o f 22.3 Long term (1891 2009) annual average precipitation was 120 cm; ~60% of precipitation occurred during a four -month period from June to September while 40% of the rainfall occurred during eight months period from October to May (Florida Climate Cen ter, 2009). Therefore, a water year is divided into wet season i.e. high flow conditions from June September and dry season i.e. low flow conditions from October May. Sub-basins of the Alafia River Watershed Two mainstem stations namely Bell Shoals and Ala fia drain 80 to 99% of the watershed (Fig 2 1). Bell Shoals drains 89% of the watershed and includes discharge from five su b -basins i.e. North Prong, South Prong, and English Creek Turkey Creek and Fishhawk Creek. While the Alafia station drains 99% of t he watershed and include discharges seven sub-basins including five from Bell Shoals station and Bell Creek and Buckhorn Creek. We grouped the FLUCCS codes at level IV into residential, built up, agricultural, pasture, mined, and forest land uses (Table 2 2). The commercial, industrial, institutional, and transportation (such as roads) were

PAGE 21

21 included under the built up land use. On the other hand, residential land use included the low, medium, and high density residential. In this study, the residential and built up land uses were considered as urban land use. Overall, at two mainstem stations ( Alafia and Bell Shoals ) watershed land use was dominated by mined (32 34%), followed by forest (19%), residential (16 17%), built up (3%), pasture (11%), and agricult ural (8 9%) (SWFWMD, 2007; Table 2 1). We grouped different sub -basins of the Alafia River Watershed using percent residential land use into t wo categories: 1) three developed (18 24% residential land use) and 2 ) two undeveloped (3 11% residential land use ). A detailed description of sub -basins is given below. Developed Subbasins Turkey Creek In 2007, Turkey Creek had 20% residential, 3% built up, 24% agricultural, and 16% pasture land use (Fig. 2 2 ; Table 2 1). Other land uses in the sub-basin include 12% forest, 17% reclaimed, and 2% mined lands. Turkey Creek was under active mining operations during 1990, however all of the mining land use in the sub-basin was reclaimed by 2007 (Fig. 2 2 ). Other significant land use changes in the sub-basin include a 9% increase (from 11% to 20%) in residential and 9% decrease (from 25% to 16%) in pasture land use during 1990 2007. In contrast to changes in mined and residential land use, the percent agricultural land use remained similar at 22 24% in the sub-basin durin g 1990 2007. One domestic wastewater treatment plant discharges 0.13 m3 sec1 of wastewater with total N and total P concentration of 2.25 mg L1 and 0.36 mg L1, respectively in this sub -basin ( NPDES, 2009). English Creek In 1990, the land use in the English Creek was 38% pasture, 19% agricultural, 10% residential, 1% built up (Fig. 2 3 ). The sub-basin has undergone significant land use changes

PAGE 22

22 during 1990 2007. For example, the residential land use in the sub-basin increased by 11% (from 10% to 21%) followed by 13% increase (from 1% to 14%) in built up during 1990 2007. The increase in residential and built up land use has occurred at the expense of pasture land use, which has decreased by 24% from 38% in 1990 to 16% in 2007 (Fig. 2 3 ). During 1990 2007, t he forest and agricultural land use remained similar at 17 19% and 25 28% respectively. Florida Department of Health has indentified 943 housing units on septic tanks in this subbasin (Florida Depar tment of Health, 2009). In addition, several small animal feed additive producing plants discharge wastewater in the sub -basin (NPDES, 2009). North Prong North Prong is the largest sub -basin draining 350 km2 of the Alafia River Watershed. Significant chan ges in the residential, pasture, and mined land use have occurred in this sub basin during 1990 2007. In 1990, North Prong had 13% residential, 13% pasture, and 44% mined land use (Fig. 2 4 ; Table 2 1). While in 2007, there was 39% mined, 18% residential, 6% built up 5% pasture, and 4% agricultural land use s (Fig. 2 4 ; Table 2 1). In this sub -basin, one domestic wastewater treatment plant discharges ~0.35 m3 sec1 of wastewater with total N and P concentrations of 1.19 mg L1 and 3.1 mg L1, respectively (http://cfpub2.epa.gov/npdes/index.cfm ). Mined land use can be divided into four categories: active mine lands, reclaimed mined lands, lands owned by mine interests that are yet to be mined, and lands ow ned by mine interests that cannot be mined (SWFWMD, 2007). However, the Florida land use classification system does not discriminate among these four mined land uses categories.

PAGE 23

23 Undeveloped Subbasins South Prong South Prong is the second largest sub-basi n that drains 277 km2 of the Alafia River Watershed. In 1990, 93% area in the South Prong sub-basin was under mining operations (Fig. 2 5 ). However, reclamation of the mined land occurred during 1990 2007. For example, in 2007, only 66% of the sub-basin wa s under mined land use (Fig. 2 5 ; Table 2 1) while the remainder was forest (15%), pasture (9%), and agriculture (4%). In contrast to North Prong, South Prong is less developed with residential land use of 3%. Several small industrial wastewater plants (ph osphate mines) discharge wastewater into South Prong during high rainfall events. Fishhawk Creek In 1990, Fishhawk Creek was undeveloped with 1% residential, 43% forest, 29% pasture, and 16% mined land use (Fig. 2 6 ). All of the mined land in the sub-basin has been reclaimed and comprise d 10% of the sub-basin in 2007 (SWFWMD, 2007). During 1990 2007, residential land use increased from 1% to 11% (Fig. 2 6 ; Table 2 1) while other land uses were 11% agricultural and 23% pasture land.

PAGE 24

24 Table 2 1. Station characteristics and associated land uses in the various sub basins on the Alafia River Watershed 1 Sub basin Station Sampling location Drainage area Land Use in 2007 Lat Long km 2 % Residential Built up Agricultural Pasture Forest Mined Mainstem Stations Alafia 2301718 27.87 82.32 1072 99 17 3 8 11 18 32 Bell Shoals 2301638 27.86 82.26 974 89 16 3 8 12 18 33 Developed English Creek 27.93 82.06 99 9 21 14 19 23 25 3 Turkey Creek 27.91 82.18 128 13 20 3 24 16 12 0 North Prong 2301000 27.86 82.13 350 32 18 6 4 5 16 39 Undeveloped South Prong 2301300 27.86 82.13 277 26 3 1 4 9 15 66 Fishhawk Creek 27.85 82.24 70.6 7 11 3 14 23 32 0 USGS station not present 2

PAGE 25

25 Table 2 2 Grouping of FLUCCS codes into major land u se s (Source: SWFWMD, 2007). FLUCCS : Florida Land Use and Cover Classification System Land Use Description FLUCCS Code Residential Low density 1100 Medium density 1200 High density 1300 Built Up Commercial and services 1400 Industrial 1500 Institutional 1700 Transportation 8100 Communications 8200 Utilities 8300 Agriculture Row crops 2140 Tree crops 2200 Feeding operations 2300 Nurseries and vineyards 2400 Specialty farms 2500 Fish farms 2550 Other open lands 2600 Pasture Pastureland 2100 Forest Herbaceous 3100 Shrubs and brush land 3200 Mixed rangeland 3300 Upland coniferous forest 4100 Pine woodlands 4110 Upland hardwood forests 4200 Hardwood conifer mixed 4340 Tree plantations 4400 Wetland hardwood forests 6100 Cypress 6210 Wetland forested mixed 6300 Freshwater marshes 6410 Salt water marshes 6420 Wet prairies 6430 Emergent aquatic vegetation 6440 Extractive 1600 Mined Reclaimed lands 1650 Reclaimed Recreational 1800 Recreational Golf courses 1820 Open lands 1900 Streams and waterways 5100 Other Lakes and Reservoirs 5200, 5300 Bays and estuaries 5400

PAGE 26

26 Figure 2 1. Location map and land uses in the various sub -basins of the Alafia River Watershed.

PAGE 27

27 Figure 2 2 Land use in the Turkey Creek sub -basin in 1990, 1999, and 2007.

PAGE 28

28 Figure 2 3 Land use in the English Creek sub -basin in 1990, 1999, and 2007.

PAGE 29

29 Figure 2 4 Land use in the North Prong sub -basin in 1990, 1999, and 2007.

PAGE 30

30 Figure 2 5 Land use in the South Prong sub -basin in 1990, 1999, and 2007.

PAGE 31

31 Figure 2 6 Land use in the Fishhawk Creek sub -basin in 1990, 1999, and 2007.

PAGE 32

32 CHAPTER 3 NITROGEN TRANSPORT IN AN URBAN WATERSHED Abstract Eutrophication affect s freshwater, estuarine, and marine eco systems worldwide. Many nitrogen (N) limiting water bodies such as Tampa Bay estuary show accelerated eutrophication due to increase in nitrogen (N) concentrations. U nderstanding the processes and mechanisms controlling the N transport in a watershed are essential to devise strategies to prevent coastal eutrophication. In this study, we used 5 to 19 years of monthly stream -water concentrations data of inorganic and organic forms of N for seven sub -basins (19 350 km2) of the Alafia River Watershed (total drainage area: 1085 km2). Results showed that t otal N in stream waters ranged from 0.84 to 2.4 3 mg L1, was greater in the developed (23 35% urban land use) than undeveloped (3 14% urban land use) sub basins. Total N was significantly ( P <0.05) correlated with urban (r=0.83) but not with agricultural (r=0.49) land uses thereby indicating that the losses of N are greater from urban areas. Among N forms, o rganic N was dominant in undeveloped (6 6 7 1 % of total N) than developed (30 4 4 % total N) sub basins while NO3N was dominant in developed (5 3 68% of total N) than undeveloped (25 3 0 % of total N) sub -basins. Concentrations of NO3-N were lower in wet season (June September) compared with dry season (October May) due to higher rainfall runoff in wet season that may have caused dilution of NO3N. In contrast, greater concentration of organic N in wet season is possibly due to the greater transport of organic materials such as leaves and plant residues in high rainfall runoff events that occurred during wet season. Trend analysis indicated that total N concentrations increas ed by 1.01% per year at the most downstream station during 1991 2009 and th is increase was primarily due to the increas ed concentrations of organic N T otal N at two mainstem stations was 1.77 1.91 mg L1, which is close to the EPA proposed numeric total N of 1.79 mg L1.

PAGE 33

33 Greater total N concentrations in three developed sub basins (1. 7 2. 4 mg L1) suggest that targeting best management practices in these th ree urban sub basins may be a resource efficient way to reduce N concentrations in this urban watershed. Introduction Nitrogen is a limiting nutrient for eutrophication in many coastal waterbodies in the US and elsewhere including the Tampa Bay E stuary in Florida ( US EPA, 2002 ; Dixon, 2009; TBEP, 2009). Major s ources of N in water bodie s are from point such as wastewater and industrial effluents and non-point such as runoff and leaching from urban and agricultural lands Point source contributions have su bstantially reduced in the US following passage of Clean Water Act in early 1970 (H owarth et al., 2002) While non -point source N pollution is the major contributor of N loads in water bodies because of its diffuse nature (Diebel et al., 2009; Maxted et al., 2009) Major non-point sources in urban watersheds can be fertilized urban lawns septic tanks, and pet waste (Brett et al., 2005; Rhodes et al., 2001) A logical approach to control non-point N pollution may be to develop an understanding of major N sources and hot spot areas in a watershed This can be achieved by determining the N forms in streams that drain sub -basins with different land uses Different land use s in a watershed have been shown to receive different N inputs (Boyer et al., 2002; Kaushal et al., 2008) For example, i n Northeastern watersheds of the US, Boyer et al. (2002) reported that the N inputs were positively correlated with the percentage of agricultural and urban land uses (r=+0.96) and negatively correlated with the forest land use (r = 0.77). Similarly, the n et anthropogenic N inputs were sixtimes greater in agricultural (77 kg ha1) and two times greater in suburban (25 kg ha1) than the forest (11.2 kg ha1) watersheds in Baltimore, US (Groffman et al., 2004) L osses of N in watersheds depend upon N inputs and therefore the anthropogenically influenced agricultural or urban watersheds result in greater N losses than

PAGE 34

34 forest ed watersheds (Boyer et al., 2002; Han et al., 2009) Groffman et al. (2004) reported that total N w as greater in streams draining agricultural (4 5 mg L1) and suburban (1 4 mg L1) as compared to the forest ed watersheds (< 0.25 mg L1). Similarly, concentrations of total N were greater in streams draining urban (1.5 mg L1) than forest (1.1 mg L1) watersheds in Se attle, US (Brett et al., 2005). Vidon et al. (2008) observed total N concentrations of 6.2 9.4 mg L1 in intensively agricultural ( corn soybean ) dominated watersheds in Indiana (Vidon et al., 2008) These and other studies have shown that total N concentrations are higher in streams that drain agricultural and urban watersheds as compared with forest ed watershed, primarily due to the greater inputs of N in the former than the later watersheds. Land use s in a watershed also influence the forms of N lost (Scott et al., 2007). To understand the terrestrial N cycling and transport across spatial and temporal scales in watersheds with dif ferent land uses, understanding of different N forms in streams that drain different land uses is essential (Pellerin et al., 2006) O rganic N (ON) forms ha ve been shown to be the dominant form s in less developed watersheds such as fores ted or mixed cover than more developed watersheds such as intensive agriculture Stanley and Maxted (2008) reported that ON was 68% (<0.3 mg L1) of total dissolved N (NO3-N + NH4N + dissolved ON) in 84 Wisconsin streams. W hile in more anthropogenic ally influence d watersheds (urban and agricultural) inorganic N forms such as NO3-N were dominant (Pellerin et al., 2006; Stanley and Maxted, 2008) For example, increased agriculture and urban land uses in the Mississippi River basin have been shown to double N concentration from 1.1 mg L1 in 1902 to 2.2 mg L1 in 1999 and most of this increase was in the NO3N which increas ed from <55% (total N: 1.1 mg L1) in 1902 to > 75% (total N: 2.2 mg L1) in 1999 (Goolsby and Battaglin, 2001; Turner and Rabalais, 2003) In a study on N budgets in 348 watersheds with varying land uses in North America

PAGE 35

35 Pellerin et al. (2006) reported that of the total dissolved N, ON was 35% in urban and 55% in forest streams. These studies provide evidence of land use specific controls on the fate and transport of different N forms. Although the percent ON is lower in streams draining agricultur al and urban land uses, the concentrations of ON may increase with anthropogenic activities. For example, P e llerin et al. (2006) observed that dissolved ON concentrations were 2 4 times greater in urban and agricultural than forest watersheds. In two watersheds dominated b y pastures in Japan Hayakawa et al. (2006) reported that urban land use of 0.2 4.3% was significantly correlated (r=+0.52) with dissolved ON. High concentrations of ON in anthropogenically influenced watersheds could be due to the inputs of organic N rich sources such as manures composts, and other organic amendments in crops and urban lawns, contribution of organic N from onsite septic systems, and direct discharge of sewage effluents in streams (Kroeger et al., 2006; Scott et al., 2007) It is also plausible that NO3N in urban and agricultural watersheds is first taken up by microbes and vegetation which on decomposition releases N in the ON forms which can be subsequently transported to waters Long term N fertilization studies at the Harvard forest in Massachusetts have shown that dissolved ON concentrations in the forest floor increased as a result of elevated inorganic N deposition thereby rep resenting the conversion of inorganic N to ON (McDowell et al., 2004) Wetlands and artificial retention ponds in many urban l andscapes are common features, designed to reduce flooding events, which may stimulate additional reinforcing processes such as plant or microbial uptake and can denitrify NO3N resulting in reduced transport of NO3N (Groffman et al., 2004; Pellerin et al., 2004) Another consequence of land use change primarily in urbanizing wat ersheds is the alteration of physical and hydrologic features that increase impervious area in the form of

PAGE 36

36 rooftops, roadways, parking lots, sidewalks, and driveways (Carpenter et al., 1998; Tufford et al., 1998) The increased impervious area in urban w atersheds leads to increased runoff due to altered and short flow paths compared to natural systems (Arnold and Gibbons, 1996; Lee and Heaney 2003; Paul and Meyer, 2001) In addi tion, t he loss of natural vegetation minimizes recycling and uptake of N due to reduced microbial and vegetative processes that immobilize N in the forest canopy, litter, soils, and organic ma tter (Abelho, 2001; Wahl et al., 1997) Together, i ncreased flow and reduced vegetation decrease the residence time of water and N in the urban watersheds, which decreases N processing such as denitrification and thus mak e mo re N available for transport (Alexander et al., 2000; Green et al., 2004; Peterson et al., 2001) Therefore, the watersheds with greater develop ment and modified hydrology are likely to export a higher amount of N to water bodies than less developed watersheds. The Alafia River Watershed is an example of such a watershed, where land use has been continuously changing from natural areas to urban lands for the last 30+ years As a result, several sub -basins of the watershed have modified hydrology due to stormwater retention ponds and water convergence structures to drain excess water during high flow events to avoid flooding T wo sub -ba sins of the watershed are dominant in mined land where phosphate mining is a commercial enterprise. We hypothesize that the 1) concentration of N forms may be different in sub basins with diverse land uses and 2) the urbanization may increase the N concent rations in streams of the Alafia River Watershed. The objectives of this study were to (1) determine how different sub -basins (with different land uses) influence concentrations of different forms of N in stream waters ; (2) investigate the influence of two distinct seasons (dry and wet) on concentrations of N forms in stream waters and (3) evaluate the long term trends of N concentrations in different sub-basins.

PAGE 37

37 Materials and Methods Study Site D escription Refer to C hapter 2 for detailed description of sub -basins of the Alafia River Watershed. Data C ollection Monthly concentrations data from 1991 to 2009 of total Kjeldahl N (TKN), NO3N, and NH4N were obtained from the Environmental Protection Commission of Hillsborough County (EPCHC) water quality database ( http://www.epchc.org). Among the two mainstem stations, water quality data w ere available only for the Bell Shoals station for 19 years (1991 2009) while for the Alafia station the data w ere available for 9 yea rs ( 1999 2009) (Table 3 1). Among other sub -basins in the watershed, 19 years of water quality data w ere available only for North Prong, South Prong, and Turkey Creek while the data availability for English Creek was f r om 1999 to 2009 and for Fishhawk Creek from 2005 to 2009. Stream water Collection and A nalysis Each month, a grab sample from surface water was collected from the channel thalweg (cent er of the stream flow) in plastic water bottles by EPCHC staff as per the surface water quality collecti on standard operating procedures In brief, the water bottles were rinsed three times with the stream water b efore collection of the samples. The collected samples were chilled with ice and transported to the laboratory where samples were stored at 4 oC pr ior to analysis. Environmental Protection Commission Hillsborough County staff analyzed the surface water samples for TKN ( EPA 350.2 method), NO2-N+ NO3N ( SM 4500 NO3), and NH4N (EPA 350.1) using a discrete analyzer. Other N fractions were calculated as f ollows: total N = NO2N+ NO3N + TKN; organic N (ON) = TKN NH4.

PAGE 38

38 Statistical A nalysis Mean, median, and standard deviations of N forms were calculated using MS Excel 2007. The GLIMMIX time series procedure in SAS Version 9.1 (SAS Institute, Cary, NC) was use d to compare the N forms across different sub basins as well as to determine the long term trends in N forms. The GLIMMIX procedure fits models to data with correlations or non constant variability and assumes normal random effects (SAS Institute, 2008). I n this time series analysis two kinds of effects were used : (1) fixed effects which included stream station and season of the year and (2) random effects which w as the date of sample collection (monthly intervals) For first fixed effect (stream station), each station was treated as a group and the variance was pooled over the sites S imilarly for the second fixed effect (season), each wet (June September) and dry (October May) season at each station was treated as one group of observations. The date of sa mple collection from 1991 2009 for the each station was considered as random effect and therefore a time series was constructed. With this, we compared (1) the difference in N forms at each of the station averaged over all sampling dates and (2) difference s in pooled mean concentrations of N forms during wet and dry seasons at each station using P significance level. In addition, same time series with fixed and random effects were used to determine the trends in N forms over the period of study. The concentrations of different forms of N were logarithmically transformed to equalize variances and normalize skewed data and were back transformed to present means of N forms in more relevant manner. V arious statistical techniques have been used to determi ne t he long term water quality trends (Goodrich et al., 2009; Johnson et al., 2009; Rich ards et al., 2008) Most common methods include non parametric Mann -Ke ndall and Seasonal Kendall for determination of lon g term trends in water quality (Daroub et al., 2009; Johnson et al., 2009). Only a few studies have used GLIMMIX procedure (e.g., Goodrich et al., 2009) and no studies have attempted to use a

PAGE 39

39 mix of more than one statistical p rocedure. Therefore, in addition to GLIMMIX procedure, we determined the long term trends using S easonal Kendall procedure. In Seasonal Kendall trend is calculated by comparing all potential data pairs. If the later value in the pair (in time) is higher t han the first, a plus sign is scored. If the later value in the pair is lower than the first, a minus sign is scored. If the results find an equal number of pluses and minuses, then there is no discernible trend. In other words, there is just as much of a likelihood that a pair of data values will be higher (or lower) than the next one. If the results show more pluses than minuses, this would indicate that a positive trend is likely (Hirsch et al., 1982) The trend in both GLIMMIX and Seasonal Kendall procedures is defined by the rate of change in the concentration of a constituent over time, which is referred to as the trend slope. The trend slope can be expressed either as change in original units per year [S (0)], or as a percent of the mean concentration of water quality variable [S (%)]. The former is the median slope of all pair wise comparisons (each pair wise difference is divided by the number of years separating the pair of observations) while the latter is produced by dividing the slope (in original units per year) by the mean and multiplying by 100. The N forms concentrations data w ere divided into two tim e periods : 1991 2000 and 2001 2009. To determine the effect of land use change on N forms in corresponding sub -basins, the land use data of 1999 and 2007 was correlated with N concentrations data of 1991 2000 and 2001 2009, respectively using SAS PROC CORR procedure Results Chemical Characteristics of Stream Waters During 1991 2009, mean monthly temperature in stream waters was 22 23 oC (Fig. 3 2). The concentration of DO in stream waters was lowest at Alafia (3.1 mg L1) than Bell Shoals (6.4 mg L1) and other sub-basins (5.3 7.3 mg L1) of the Alafia River Watershed. Among the sub-

PAGE 40

40 basins, the DO concentration was significantly ( P < 0.05) lower at one developed (English Creek; 5.3 mg L1) than other sub -basins. Mean pH of the stream waters was similar ( 7.2 7.5) at all the stations of the Alafia River Watershed (Fig. 3 2) In contrast to pH, the mean EC of the stream waters was atleast one order of magnitude greater at one mainstem station near to the Tampa Bay (Alafia; 7548 S cm1) than Bell Shoals (443 S cm1) and other sub -basins (214 556 S cm1). Among the sub basins, EC was significantly lower at one undeveloped sub-basin (Fishhawk Creek: 214 S cm1) than other developed and undeveloped sub -basins. Concentrations of Nitrogen Forms in Streams Dr aining Different Subbasins During 1991 1999, mean monthly total N concentration at the mainstem station (Bell Shoals) that drains 89% of the w atershed was 1. 6 5 mg L1 (Fig. 3 3A). T otal N was significantly (P < 0.05) greater at two developed (1.45 1.74 mg L1) than one undeveloped sub-basin (1.06 mg L1). Among the two developed sub-basins, Turkey Creek had significantly ( P < 0.05) greater total N than North Prong. T otal N concentrations were greater during 2000 2009 than 1991 1999 at one mainstem (Bell Sh oals), two developed (North Prong and Turkey Creek), and one undeveloped (South Prong) sub -basins of the Alafia River Watershed (Fig. 3 3A). M ean mo nthly total N at two mainstem stations (Alafia and Bell Shoals; urban land use 19 20%) w as 1.7 7 1. 91 mg L1 (Fig. 3 3 A). Total N concentrations in stream waters were significantly ( P <0.05) greater in all the developed sub basins (English Creek, Turkey Creek, and North Prong; 1. 6 7 2.4 3 mg L1) than undeveloped sub-basins (South Prong and Fishhawk Creek; 0.84 1. 21 mg L1). Differences in total N concentrations in streams among developed and undeveloped subbasins were observed. For example, among the developed sub-basins, English Creek had significantly greater total N (2.4 3 mg L1) than Turkey Creek (1. 96 mg L1) and North Prong (1. 6 7 mg L1). Similarly in the

PAGE 41

41 undeveloped sub-basins, total N was significantly higher in South Prong (1. 21 mg L1) as compared to Fishhawk Creek (0.84 mg L1). During 1991 1999, mean monthly ON concentration at the mainstem station (Be ll Shoals) was 0.49 mg L1 (30% of total N) (Fig. 3 3B). The ON concentration in the stream waters was similar in all the sub -basins (0.57 0.63 mg L1). In contrast to the concentrations, the percent of ON was greater in undeveloped (53% of total N) than t wo developed (36 41% of total N) sub basins Concentrations of ON were greater during 2000 2009 than 1991 1999 in streams draining mainstem, two developed, and one undeveloped sub-basins. Among the mainstem stations, m ean monthly concentration of ON was significantly ( P < 0.05) greater at the most downstream Alafia (1.0 2 mg L1; 5 8 % of total N) as compared to next upstream Bell Shoals (0. 75 mg L1; 3 9 % of total N) (Fig.3 3 B). All the developed and undeveloped sub-basins had a narrow range of stream ON (0. 7 3 0.7 7 mg L1) except for a significant lower concentration at Fishhawk Creek (0.60 mg L1). However, the percent of ON was much greater in streams draining undeveloped (66 7 1 % of total N) than developed (30 4 4 % of total N) sub -basins. At the mainstem station (Bell Shoals), the NO3N was the dominant N form (1.13 mg L1; 69% of total N) during 1991 1999 (Fig. 3 3C). In contrast to ON, NO3N concentrations were two folds greater in two developed (0.82 1.07 mg L1; 57 62% of total N) than one undeveloped sub -basin (0.47 mg L1; 45% of total N). Among the two developed sub-basins, Turkey Creek had significantly ( P < 0.05) greater NO3N than North Prong. The concentrations of NO3N were similar during 1991 1999 and 2000 2009 in all the stream s draining different sub -basins (Fig. 3 3C). Concentrations of NO3N at Bell shoals were significantly ( P < 0.05) greater (1.14 mg L1; 6 0 % of total N) than the most downstream Alafia

PAGE 42

42 (0.6 7 mg L1; 38% of total N). Among the subbasins, NO3N followed the t otal N pattern and was signif icantly greater in developed (0.89 1. 66 mg L1; 5 3 68% of total N) than undeveloped (0.21 0. 37 mg L1; 25 3 0 % of total N) sub -basins. At all study stations, m ean monthly concentrations of NH4N were much lower (<0.1 mg L1; 2 5% of total N) than both ON and NO3N during 1991 1999 and 2000 2009 (Fig. 3 3 D) Seasonal Variations in Chemical Characteristics in Stream Waters Draining Different Subbasins At two mainstem stations (Alafia and Bell Shoals), the mean t emperature of t he stream waters was significantly ( P < 0.05) greater in wet ( 26 27 oC) than dry season ( 20 21 oC) (Fig. 3 4). A similar increase in temper ature of stream waters in wet ( 26 27 oC) than dry ( 18 20 oC) seasons was found in developed and undeveloped sub -basins. The seasonal variation showed opposite trends in DO concentrations than temperature of the stream waters possibly due to greater biological activity in stream waters with greater temperature For example, at two mainstem stations, the DO concentration of the stream waters was greater in dry (3.4 6.8 mg L1) than wet season (2.7 5.9 mg L1). Among the sub-basins, the concentration of DO in stream waters was 5.6 8.1 mg L1 in dry season and decreased to 4.6 6.1 mg L1 in wet season (Fig. 3 4). Thi s decrease in DO concentration of the stream waters was significant ( P < 0.05) in all the sub -basins except for English Creek. At two mainstem stations, the pH of stream waters was greater but not significantly ( P < 0.05) in dry than wet seasons (Fig. 3 4). Among sub-basins, the pH of the stream waters was greater in dry season (7.4 7.6) and decreased in wet season (6.9 7.3); however the decrease was significant ( P <0.05) only at Fishhawk Creek. Similarly, the EC of the stream waters was greater but not signi ficantly ( P <0.05) different in dry than wet season in all study basins

PAGE 43

43 Seasonal Variations in C oncentrations of Nitrogen Forms in Streams Draining Different Sub-basins At two mainstem stations (Alafia and Bell Shoals), the mean total N was slightly high er but not significant ( P < 0.05) in dry (1.83 1.84 mg L1) than wet (1.61 1.69 mg L1) season (Fig. 3 5 A). All t he developed sub-basins (Turkey Creek, English Creek, and North Prong) had greater total N in dry than wet season while the undeveloped sub basins had greater total N in wet than dry season (Fig. 3 5 A). However, the seasonal variation was significant ( P < 0 .05) only in Fishhawk Creek and English Creek. Mean ON concentrations were lower in dry season and increased in wet season at all sites (Fig. 3 5 B; Fig. 3 6 ). Th e magnitude of this increase was greater at Bell Shoals (from 0.52 to 0.82 mg L1) as compared to Alafia (from 0.99 to 1.03 mg L1). Among the developed sub-basins, the increase in ON concentration during wet season was greate st at English Creek (from 0.57 to 1.03 mg L1) than Turkey Creek and North Prong (from 0.60 0.64 to 0.80 0.83 mg L1). In und eveloped sub-basins, ON was 0.48 0.58 mg L1 in dry season which increased to 0.83 0.88 mg L1 in wet season In contrast, NO3N w as lower in wet than dry season at all sites (Fig. 3 5 C; Fig. 3 6 ). For example, a t two mainstem stations, NO3N in dry seaso n was 0.75 1.30 mg L1, which decreased to 0.51 0.84 mg L1 in the wet season. Similarly, a mong developed subbasins, the decrease in NO3N in wet season compared to dry season was significant at English Creek (from 2.07 to 0.91 mg L1) but not at Turkey C reek (from 1.24 to 0.85 mg L1) and North Prong (from 0.98 to 0.65 mg L1). In undeveloped subbasins, the decrease in NO3N was significant only at Fishhawk Creek (from 0.23 to 0.17 mg L1).

PAGE 44

44 Concentration s of NH4-N were not significantly different among dry and wet seasons at all study stations (Fig. 3 5 D ; Fig. 3 6 ). But NH4-N w as greater at the most downstream station (Alafia) than other sites in both wet and dry seasons. Long T erm Trends in Flow Un weighted Nitrogen Forms in Streams Draining Different Sub-basins Total Nitrogen The longest data record (19 years) was only available for one mainstem station (Bell Shoals ) that drains 89% of the watershed, two developed (Turkey Creek and North Prong), and one undeveloped (South Prong) sub basin (T able 3 7 ). According to Seasonal Kendall trend analysis, none of the total N concentrations trends were significant ( P < 0.05) for these four sub basins during the study period. However, the concentrations of total N w ere increasing at Bell Shoals (+0.17% per year; 2.79 g L1 per year) (Fig. 3 7 ). Among the sub-basins total N showed an increasing trend +1.7% or 18. 9 g L1 per year at South Prong and +1.5% or 29. 1 g L1 per year at Turkey Creek and a decreasing trend at North Prong ( 0.42% or 6. 7 g L1 per year). Among the stations with 10 years of data record, the mainstem station Alafia and developed English Creek showed decreasing but not significant ( P < 0.05) total N trend s The total N concentration trends estimated with the GLIMMIX procedure were similar to Seasonal Kendall but the magnitude of trend slopes was variable The GLIMMIX model showed a significant ( P <0.09) overall increasing total N trend of +1.01% per year at the mainstem Bell Shoals station which equates to approximately 20% increase (0.33 mg L1) in the total N over 19 years (Fig. 3 7 ). Total N significantly increased at Turkey Creek (+1.02% per year; P < 0.005) and South Prong (+1.07% per year; P < 0.03) but not at North Prong ( 1.11% per year; P < 0.64) during 1991 2009. The study stat ions with 10 years of data record (Alafia and English Creek) showed decreasing but not significant ( P < 0.05) total N concentration trends.

PAGE 45

45 Organic Nitrogen Seasonal Kendall trend analysis showed that the ON concentrations were increasing at all the four study stations with 19 years of data record (Fig. 3 8 ). At Bell Shoals, total N showed an increasing but not significant ( P < 0.45) trend (+1.8% per year) which equates to 35% increase in mean ON concentration during 1991 2009. The increasing ON trends were significant ( P < 0.001) at Turkey Creek (+2.5% per year) but not at North Prong (+0.72% per year; P < 0.46) and South Prong (+1.95% per year; P < 0.25). None of the trends w ere significant in two sub -basins with 10 years of data record. The GLIMMIX procedure showed increasing ON concentration trends at four sub-basins with 19 years of data record (Fig. 3 8 ). Overall, at Bell Shoals the ON concentration trend was (+0.94% per year; P = 0.001) which equates to approximately 19% increase in ON concentration over 1 9 -years. Among the sub basins, the ON concentrations significant ly ( P < 0.001) increased at Turkey Creek (+0.94% per year), South Prong (+0.95% per year) but not at North Prong (+1.00% per year). During 19992009, Alafia station showed decreasing ( 2.45% pe r year) as compared to increasing trends at English Creek (+0.65% per year). Nitrate Nitrogen As per Seasonal Kendall trend analysis, none of the NO3N concentrations trends w ere significant ( P < 0.05) however the NO3-N decreas ed at Bell Shoals ( 0.27% per year) during 1991 2009 (Fig. 39 ). Among the sub-basins, the NO3N concentration trend was decreasing at North Prong ( 0.1% per year) and South Prong ( 2.7% per year) but not at Turkey Creek (+0.18% per year) during 1991 2009. During 1999 2009, Alafia station showed an increasing NO3N concentration trend (+5.4% per year). Using GLIMMIX procedure, none of the NO3N concentrations trend was significant ( P < 0.05); however the overall trend at Bell Shoals was decreasing ( 1.04% per year) during 1991

PAGE 46

46 2009 (Fig. 3 9 ). Turkey Creek showed an increasing NO3-N concentration trend (+1.13% per year) in contrast to decreasing trends at North Prong ( 0.31% per year) and South Prong ( 0.98% per year). During 1999 2009, the NO3N concentration trends were not signi ficant at Alafia (+0.96% per year) and English Creek ( 1.74% per year). Ammonium Nitrogen Both the Seasonal Kendall trend analysis and GLIMMIX procedure showed non significant trends NH4N at all the sub -basins of the Alafia River Watershed (Fig. 3 10). Th e concentrations of NH4N remained lower (<0.4 mg L1) during the study period at all the subbasins. Long T erm Trends in Flow Weighted Nitrogen Forms in Streams Draining at Mainstem Station At the mainstem station that drains 89% of the Alafia River Wate rshed showed increasing but not significant ( P < 0. 241) total N concentration trends during 1991 2009 (Table 3 2). Mean concentrations of total N were greater during 2000 2009 (1.82 mg L1) than 1991 1999 (1.53 mg L1) reflecting the increasing GLIMMIX trends in terms of actual increase in concentrations The ON concentrations showed significant ( P < 0. 01) increasing trends of +1.05% per year (+7.0 g L1 per year ; 0. 14 mg L1 in 19 years ) Similar to total N, mean ON w as greater during 2000 2009 (0.81 mg L1) than 1991 1999 (0.55 mg L1). In contrast to ON, flow weighted NO3N concentration trends were not significant ly ( P < 0. 487) different for 1991 1999 (0.94 mg L1) and 2000 2009 (0.97 mg L1). Long Term Trends in Ni trogen Loads at Mainstem Station At mainstem station (Bell Shoals), total N loads were not significant in wet and dry seasons during 19912009 (Table 32). The non significant trends in the N loads during wet and

PAGE 47

47 dry seasons could be due to greater variations in N loads and reduced number of data points in wet and dry seasons. Relationship between Land Use and Nitrogen Forms T otal N concentration s w ere significantly ( P <0. 05) correlated with the percent urban (resi dential + built up) land use (r=+0. 83) in the watershed (Fig. 3 11). This positive correlation was probably due to the influence of urban land use on NO3N (r=+0. 78; P <0. 05) rather than ON (r= + 0. 28) concentrations during the study period. The p ercent agric ultural land use was positively but not significantly ( P <0. 05) correlated with total N (r= +0. 49) and NO3N (r= +0. 53) The pasture and forest land uses were not significantly correlated with any of the N forms (r= <0.30). Discussion Influence of Land Use s on Total Nitrogen Concentrations in Stream Waters Mean monthly concentrations of total N at two mainstem stations (1.7 1.9 mg L1) approached the EPAs proposed numeric total N value of 1.79 mg L1 for the region (EPA, 2010). These t otal N concentrations were greater than reported for Hillsborough County lakes value of 1.12 mg L1 (Florida Lakewatch Program, 2005). However, total N w as much lower than corn -soybean dominated watersheds (6.2 9.4 mg L1) in Indiana, US (Vidon et al., 2008). Lower total N con centrations in the Alafia River Watershed may be to the low agricultural land use (8%) while watersheds in Indiana had more than 60% of row crops land use. In our study, t otal N concentrations were similar to values reported for the urban and forest waters heds ( 1.1 1.5 mg L1) in Seattle, US (Brett et al., 2005) The three developed sub basins (23 35% urban land use) had greater total N (1.67 2.43 mg L1) than undeveloped sub-basins (3 14% urban land use) (0.84 1.21 mg L1). T otal N concentrations were significantly ( P < 0.05) positively correlated with urban land use (r= +0. 83)

PAGE 48

48 (Fig. 3 11). While total N was positively but not significantly ( P < 0.05) correlated with agricultural (r=+0. 49) and no relationship was observed between total N and forest and pasture land uses (r= < 0. 3 0). This indicates that the N concentrations signific antly increase with urbanization of the watershed. The strong correlation between percent urban land use and NO3N concentrations (r=+0. 78) may be due to the runoff of fertilizers containing N from urban/residential lawns Overall, greater concentrations of total N in urban than forest and pasture land uses may be due to the higher N inputs in the former than later land uses (Boyer et al., 2002; Groffman et al., 2004) Furthermore, in urban lands impervious areas result in greater transport of N (Alexander et al., 2000; Green et al., 2004; Peterson et al., 2001) resulting in higher total N concentrations in developed tha n undeveloped sub -basin stream s. Among the three developed sub-basins the total N concentrations were significantly (P <0.05) different during 2000 2009. G reater total N concentrations at English Creek than Turkey Creek and North Prong could be due to the greater urban land use (35%) in English Creek than North Prong and Turkey Creek (23 24%). A wastewater treatment plant discharges 0.11 m3 sec1 wastewater with total N concentration of 2.76 mg L1 into the English Creek (SWFWMD, 2007). The total N concentrations in developed sub-basin vary inversely with size of the sub-basin. T he length of the streams increases with the size of the sub -basin, which provides opportunities for the N removal processe s such as denitrification in riparian buffer zones as well as within the stream resulting in lower concentrations of N (Alexander et al., 2000; Seitzinger et al., 2002; Alexander et al., 2009) These factors might have resulted in significant differences i n N concentrations in three developed sub -basins. On the other hand, lower total N concentration in Fishhawk Creek that had 14% urban land (0.8 4 mg L1) than South Prong that had 3% urban land (1. 21 mg L1) could be because of greater land area in forest i n the former

PAGE 49

49 (32%) than later (15%) sub -basin (Table 3 1). Th is highlight s the importance of other land uses such as forest in addition to urban and agricultural land use in modifying the net total N losses in sub -basins. Land Uses and Forms of Nitrogen Concentrations in Stream Waters In our study, NO3N was 25 3 0 % of total N in undeveloped and 5 3 68% of total N in developed sub basins. The percent urban land use was strongly correlated with NO3N concentrations (r= +0.78) but not with ON concentrations (r= +0.2 8 ). This suggested that the increase in total N concentrations in urban land uses is primarily driven by NO3N probably due to the runoff losses of the N fertilizers applied to the urban lawns Similarly, Stanley and Maxted (2008) in 84 streams in Wisc onsin reported that NO3-N was 75% of total dissolved N in urban compared to 30% in the forested streams. In a review on 348 watersheds across North America, Pellerin et al. (2006), reported that NO3N was 65% of total N in urban (>50% urban land use) and 3 5% of total N in forest (>90% forest land use) watersheds In contrast to decreasing p ercentages of ON in developed sub-basins, the concentrations of ON remained similar in developed (0. 73 0.7 7 mg L1) and undeveloped sub -basins (0.60 0. 79 mg L1). This indicates the existence of enough ON sources in urban areas such as contributions from septic tanks, grass cuttings of urban lawns, and deciduous leaves fallen on the ground (Kroeger et al., 2006; Scott et al., 2007) In addition, more impervious areas in urban areas may increase the runoff resulting in greater transport of these organic materials that contain ON (Carpenter et al., 1998; Tufford et al., 1998) Recent studies suggest that microbial transformations of NO3N to ON may act as an important source of ON in urban watersheds (McDowell et al., 2004; Pellerin et al., 2006) In our study, urban land use was slightly positively correlated (r= +0.2 8 ) with ON concentrations. Florida has a warm climate where the existence of wetlands and artificial stormwater retention ponds might play a stimulating role for biotic

PAGE 50

50 conversion of NO3N into ON resulting in greater concentrations of ON in streams draining urban land uses (Groffman et al., 2004; Pellerin et al., 2004) Similarly, gr eater proportion of ON was found at the most downstream Alafia stations (5 8 % of total N) than Bell Shoals (3 9 % of total N) (Fig. 3 6 ). Alafia station is close to the Tampa Bay estuary where the velocity of the water decreases and thus this may provide more favorable conditions for denitrification and/or conversion of NO3N to ON via vegetation/microbial uptake (Groffman et al., 2004) Seasonal Impact s on Nitrogen F orms in Stream Waters The NO3-N concentrations in stream waters were greater in dry season and decreased in wet season in all the sub -basins (Fig. 3 5 ). This can be explained by the significantly ( P <0.05) greater temperature in wet season that might increase the biological uptake and denitrification of NO3N resulting in lower losses of NO3-N in wet than dry season (Alexander et al., 2009; Chen and Driscoll, 2009). Due to greater biological activities in th e stream waters the concentration of DO is lower in wet than dry season in all the streams (Fig. 3 4). This provides an indication of greater denitrification losses of NO3N resulting in lower concentrations of NO3N in wet than dry season. Another probab le reason for decreased NO3N concentrations in wet season could be the shift in flow -paths from groundwater to stormwater in wet season. In the Alafia River Watershed, EC of stream water is lower in wet than dry season thereby representing the dilution of the stream waters with stormwater runoff during the wet season Similar to other salts (EC), NO3N concentrations decrease with greater stormwater runoff i n wet season. This dilution of NO3N concentration in the stream waters during the wet season contra dicts previous studies, who reported higher concentrations (flushing) of NO3-N with greater runoff in wet season (Chen and Driscoll, 2009; Green et al., 2004; Qian et al., 2007; Royer et al., 2006) In general, the flushing of the NO3N with stormwater in wet season usually occurs when excess of NO3N is present in the watershed soils and other sources For example, in the Indian River Lagoon,

PAGE 51

51 Florida, fertilizers are applied to citrus crop during the wet season and the excess of NO3N is flushed with rain resulting in greater concentrations of NO3N in wet than dry season (Qian et al., 2007). Similarly, the watersheds in Midwest, US, are row crops dominated which receive greater N inputs and therefore the NO3N concentrations i ncrease with greater drainage from the watersheds (Vidon et al., 2008). The Alafia River Watershed is not as intensively cultivated as the watersheds of the US Midwest and the Indian River Lagoon, therefore greater water is available due to modified urban drainage for dilution of the dissolved forms of N (NO3N) resulting in decreased concentration of NO3N during wet than dry season. In contrast to NO3-N, transport of particulate forms of N which represents ON is greater with greater storm water runoff in wet than dry season (data not shown) The particulate N forms could be decomposing products of leaf litter, grass cuttings of lawns, and soil sediments (Culbert and France, 1995) Therefore, concentrations of ON were greater in wet season than dry season in the Alafia River Watershed. Long Term Trends in Nitrogen Concentrations In our study, the Seasonal Kendall and GLIMMIX procedures showed similar trends (increasing or decreasing), though the magnitude of the trend slopes differed by 0 to 14% per year among two procedures (Fig. 3 7, 3 8, 3 9, 3 10). These differences in trend slopes appear to be associated with the conceptual difference in the procedures of two methods. For example, Seasonal Kendall trend analysis compares the change in concentrations of nutrients with time (slope) for each point and the median slope is calcula ted as a summary statistic describing the magnitude of the trend (Johnson et al., 2009; Qian et al., 2007) On the other hand, GLIMMIX procedure fits the linear lines through the data and slope is the rate of change in concentration with time at any of two points ( SAS 2008). Therefore, the differences in the magnitude of slopes appear in the trend analysis in two methods. In general, Seasonal Kendall trend analysis

PAGE 52

52 determines the trends at a single location. However, GLIMMIX procedure can be used to compare the mean co ncentrations at multiple locations as well as has the ability to include many parameters such as stream discharge, seasonal variation, distance of the mainstem, and location of the station etc (Goodrich et al., 2009) In this study, GLIMMI X procedure has shown the significant variations in mean and seasonal variations in N concentrations at different stations. The causes for the concentration trends can be either 1) natural such as the changes in precipitation and evapotranspiration result ing in variations in flow, or 2) anthropogenic such as changes in land use, nutrient management practices in the watersheds. The flow weighted and flow un -weighted concentration trends can help discriminate the effects of natural and anthropogenic changes on nutrient concentrations in streams. In the analysis of trends in flow weighted concentrations, the effects of natural changes in stream flow on concentration are removed and therefore flow adjustment of nutrient concentrations allows trends due to anthr opogenic changes to be assessed. In the analysis of trends in observed flow un-weighted concentrations, no adjustments are made for any natural or anthropogenic influence. Therefore, the net effects of all simultaneous influences on concentration are evalu ated, allowing for the assessment of nutrient concentrations in streams relative to water quality standards of Environmental Protection Agency and the condition of aquatic communities (Sprague and Lorenz, 2009) At Bell Shoals that drains 89% of the watershed, increasing trends were found in flow un weighted (+1.01% per year; P<0.08) and flow weighted (+1.08% per year; P<0.24) total N concentrations during 1991 2009 (Fig. 37; Table 3 2). Flow remained similar and did not show any significant trends ( P<0.75) during 1991 2009. Therefore, the possibility of concentration or dilution of total N with variations in the natural causes (flow) is ruled out. Non-significant trends

PAGE 53

53 in flow weighted concentrations could be due to greater variations in flow resulting in greater variance in data and therefore non -significant results ( Johnson et al., 2009). Similarly the loads data at Bell Shoals is insignificant due to greater variations in the data. Greater variation in the loads could be due to the large variation in smaller streams of the Alafia River Watershed. However, increasing trends in both flow weighted and flow un-weighted total N concentrations indicated the possibility of anthropogenic influence in increasing the N concentrations in the Alafia River. In this regard, the Alafia River Watershed is similar to most of the watersheds in US which have shown increasing total N concentration trends during the last few decades (Goolsby and Battaglin, 2001; Turner and Rabalais, 2003; Weston et al., 2009) For example, in Mississippi River Basin, Turner and Rabalais (2003) reported the increasing total N concentration trends primarily due to clearing of forests for the agricultural land use in the last 200 ye ars. Similarly, in Minnesota River, Mann -Kendall trend analysis showed increasing NO3N concentration trends (from +2.90 to +6.29% per year) during 1976 2003 (Johnson et al., 2009) They attributed the increasing N losses to the increased drainage of the corn -soybean dominated watersheds during 1976 2003. Contrary to these studies, in 58 watersheds of Eastern US, Sprague and Lorenz, (2009) using Mann-Kendall trend analysis have not found any significant total N concentrations trends during 1993 2003. In the Alafia River Watershed which is located in the Hillsborough County, population has grown from 0.83 million in 1990 to 0.99 million in 2000 and 1.20 million in 2009 (US Census Bureau, 2010). This represents 19.8% increase during 1990 2000 and 19.7% during 2000 2009. The population growth has resulted in the land use changes in the watershed. For exa mple, the urban land use (residential a nd built up) has increased by 8 % (from 12 to 20%) of the watershed during 1990 2007 (SWFWMD, 2007). This is coupled with a de crease of 5% in forests (from 23

PAGE 54

54 to 18 %) and 8% decrease in pastures (from 19 to 11%) while agricultural land use remained similar (8%) during the study period. In our study, mean total N concentrations were positively correlated with urban (r=0. 83) (Fig. 3 11; Fig. 3 12). Therefore, increasing trends of total N concentration in the Alafia River Watershed may be attributed to the increase in urban land use of the watershed. Similarly, studies in the Altamaha River, GA by Weston et al., (2009) reported the significant increasing total N trends (+7.42 g L1 per year) during 1970 2000. In their study, the average population density was increased from 35 to 70 person km2 while the agricultural land use had decreased from 44 to 29%; thereby reflecting the impact of urbanization on total N losses. In general, the increase in urban land use results in i ncrease in the N inputs (fertilizer application to lawns, septic tanks, leaf litter, and grass cutting of lawns) and changes the hydrology of the watershed which can increase the N losses to the streams (Carpenter et al., 1998; Han et al., 2009; Tufford et al., 1998) In our study, two developed subbasins (English Creek and North Prong) have shown the decreasing total N concentration tr ends where the urban land use has increased by 8 23% during 1990 2007. It is important to note that at subbasin scale several anthropogenic changes takes place simultaneously which affect the net N losses from a watershed. For example, N concentrations in streams may increas e with the urbanization in the watershed while the improvements in the wastewater treatment facilities might decrease N losses resulting in decreasing or non significant results (Sprague and Lorenz, 2009) Two sub basins (North Prong and English Creek) rec eive discharges from two domestic wastewater discharges in additions to several small industrial wastewaters (NPDES, 2009). We believe that in English Creek and North Prong the increasing total N concentration trends due to urbanization are offset by the i mprovements in the domestic as well industrial wastewater discharges.

PAGE 55

55 In our study, NO3N (25 68% of total N) has not any significant increasing or decreasing trends at all the study stations (Fig. 3 9 ). Among the sub basins, only Turkey Creek has shown th e increasing flow un-weighted NO3N concentration trends during 1991 2009. In the Alafia River Watershed, several programs such as fertilizer application based on crop/soil tests, N pollution reduction from residential areas, improvements in wastewater tre atment facilities were initiated to control N (especially NO3N) pollution during last two decades (TBEP, 2009). It seems the success of the N abatement programs is partially offset by the increasing urban land use and population growth resulting in insign ificant NO3N concentration trends during the study period. In contrast to NO3N, the ON which contributed 30 71% of total N, has shown significantly increasing trends most of the study stations (Fig. 3 8 ). In general, increasing ON concentration trends du e to increases in urban land use in the Alafia River Watershed contradicts the common belief that urban land use increase the NO3N rather than ON losses from watersheds (Pellerin et al., 2006; Stanley and Maxted, 2008). In our study, the proportion of ON decreased with increasing urban land use while the concentrations of ON were similar in both developed and undeveloped subbasins. This suggests that there are sources of ON even in the urban land uses. In urban land uses, the impervious areas increase the runoff resulting in greater transport of particulate ON which may ultimately lead to increasing ON concentration trends with urbanization (Carpenter et al., 1998; Han et al., 2009; Tufford et al., 1998) In a review of 348 watersheds, Pallerin et al. (2006) reported that the ON concentrations in the surface waters were 2 3 folds greater in urban/suburban (0.47 0.49 mg L1) than forest watersheds (0.18 mg L1). They attributed the greater concentrations in urban land uses to the influence of septic tanks, wastewater treatment plants, and the biotic conversions of NO3N to ON. Using isotopic signatures of carbon and N, Peebles et al. (200 9) reported that inorganic N from the fertilizers

PAGE 56

56 was the dominant source of N to the sediments in the Safety Harbor part of the Tampa Bay. Similarly, long term N fertilization studies at the Harvard Forest in Massachusetts have shown that dissolved ON con centrations in the forest floor have increased as a result of elevated inorganic N deposition (McDowell et al., 2004) In our study, wetlands and artificial retention ponds might have played a stimulating role for biotic conversion of NO3-N into ON which resulted in increased ON concentration trends with urbanization during the study period (Groffman et al., 2004; Pellerin et al., 2004) In this way, increasing ON concentration trends in this urbanizing wa tershed has raised two important questions 1) Is the increase in ON concentration due to ON sources (such as septic tanks, urban manures /composts runoff of grass cuttings, and deciduous leaves fallen on the urban surfaces, 2) Is the ON increase a product of microbial transformations of NO3N into ON. Further studies on the source characterization of the ON can help in devising the BMPs to control N pollution in the Alafia River Watershed. Results of our trend analysis suggested that the increasing total N concentrations could be due to the increasing urban land use in the watershed during 1991 2009. None of the sub-basin showed significant decreasing or increasing NO3N trends. This has suggested that the conversion from forest and pastures to urban land use has offset the success of the N abatement programs in controlling NO3N pollution in the Alafia River Watershed. On the other hand, lack of BMPs to target ON (30 7 1 % of total N) has caused increase in ON concentrations ultimately resulting in increasi ng total N trends. For controlling N pollution, Turkey Creek with greater total N concentrations and increasing total N concentration trends should be targeted followed by English Creek to control N pollution. In these two developed sub-basins, the BMPs s hould also focus ON rather than NO3N alone to protect water quality in the Alafia River Watershed.

PAGE 57

57 Summary At two mainstem stations (Alafia and Bell Shoals) that drain 89 99% of the Alafia River Watershed, the mean total N concentrations approached the EPA proposed numeric total N criteria values for the region. During 2000 2009, t otal N concentrations were greater in three developed (1. 6 7 2.4 3 mg L1) than two undeveloped sub -basins (0.8 4 1. 21 mg L1) thereby suggesting that the urbanization of the watershed has increased N losses from watersheds Greater proportion of NO3N in developed (53 68% of total N) than undeveloped subbasins (25 3 0 %) indicate that in urban land uses the losses of NO3N are g reater than ON. In our study, the concentration of NO3N decreased in wet season due to greater biotic uptake and denitrification of NO3-N with greater temperature as well as dilution of NO3N with greater runoff in wet than dry season. T he greater concent rations of ON in wet than dry season were probably due to greater transport of organic material with runoff water. The trend analysis showed increasing total N concentration at the mainstem station during 1991 2009. Th is was primarily due to the greater in creases in ON than NO3N thereby suggesting the need to control N losses from ON sources T hree developed sub-basins (Turkey Creek English Creek and North Prong ) that had greater total N con centrations should be first targeted to reduce N pollution in the Alafia River Watershed.

PAGE 58

58 Table 3 1. Station characteristics of the Alafia River Watershed Sub basin Station Sampling location Drainage area Land Use in 2007 Lat Long km 2 % Residential Built up Agricultural Pasture Forest Mined Mainstem Stations Alafia 2301718 27.87 82.32 1072 99 17 3 8 11 18 32 Bell Shoals 2301638 27.86 82.26 974 89 16 3 8 12 18 33 Developed English Creek 27.93 82.06 99 9 21 14 19 23 25 3 Turkey Creek 27.91 82.18 128 13 20 3 24 16 12 0 North Prong 2301000 27.86 82.13 350 32 18 6 4 5 16 39 Undeveloped South Prong 2301300 27.86 82.13 277 26 3 1 4 9 15 66 Fishhawk Creek 27.85 82.24 70.6 7 11 3 14 23 32 0 USGS station not present Table 3 2. Longterm trends in flow weighted and loads of N forms at Bell Shoals Parameter data Flow weighted concentrations Loads Trend Slope (%) p value Trend Slope p value Total Nitrogen 1991 2009 Increasing +1.01 0.242 No trend 0.746 Organic Nitrogen 1991 2009 Increasing +1.08 0.016 Nitrate Nitrogen 1991 2009 No trend 0.491

PAGE 59

59 Figure 3 1. Location map of the Alafia River Watershed

PAGE 60

60 10 20 30 40 0 4 8 12 5 6 7 8 9 A. Temperature (Celcius) B. Dissolved Oxygen (mg L-1) C. pH 75 percentile 25 percentile Median 0 400 800 1200 D. Electrical Conductivity (uS cm-1)Alafia Bell Shoals English Creek Turkey Creek North Prong South Prong Fishhawk CreekA A A A A A A A B AB C C C B A A A A A A A BC B BC C B AAlafia Bell Shoals English Creek Turkey Creek North Prong South Prong Fishhawk Creek Alafia Bell Shoals English Creek Turkey Creek North Prong South Prong Fishhawk Creek Alafia Bell Shoals English Creek Turkey Creek North Prong South Prong Fishhawk Creek Figure 3 2. Chemical characteristics of the stream waters during two time periods from 1991 to 2009. Values indicated by different letters are significantly different at P <0.05 for each graph

PAGE 61

61 0 1 2 3 4 5 Concentration (mg L -1 ) 0 1 2 3 4 0 1 2 3 4 A. Total Nitrogen B. Organic Nitrogen C. Nitrate-NN 120 225 120 207 225 225 74 = 75 percentile 25 percentile Median 0.0 0.1 0.2 0.3 D. Ammonium-N Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek1991-1999 2000-2009 1991-1999 2000-2009 1991-1999 2000-2009 1991-1999 2000-2009Bell Shoals Turkey Creek North Prong South Prong Bell Shoals Turkey Creek North Prong South Prong Bell Shoals Turkey Creek North Prong South Prong Bell Shoals Turkey Creek North Prong South ProngC C D C C B A B C B A C AB B B B B A A A A A BC C D D C B A C C B A B A A A A A A A A A A Figure 3 3. Summary of mean monthly concentrations of total, organic, nitrate, and ammonium nitrogen during 1991 2009 in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk Creek). Values indicated by different letters are significantly different according to GLIMMIX procedure at P< 0.05. N in each sub-basin indicates number of months/observations

PAGE 62

62 10 20 30 40 0 2 4 6 8 10 12 5 6 7 8 9 Dry Season Wet SeasonA. Temperature (Celcius) B. Dissolved Oxygen (mg L-1) C. pH 75 percentile 25 percentile Median 0 200 400 600 800 1000 D. Electrical Conductivity (uS cm-1) Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Dry Season Wet Season Dry Season Wet Season Dry Season Wet Season Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek* * * * * * Figure 3 4. Seasonal variation in chemical characteristics of the stream waters during 1991 2009. Values indicated by dif ferent letters are significantly different according to GLIMMIX procedure at P< 0.05. Dotted line represents the mean value

PAGE 63

63 0 1 2 3 4 Concentration (mg L -1 ) 0 1 2 3 4 0 1 2 3 4 Dry Season Wet SeasonA. Total Nitrogen B. Organic Nitrogen C. Nitrate-NN 120 225 120 207 225 225 74 = 75 percentile 25 percentile MedianDry Season Dry Season Wet Season Wet Season 0.0 0.1 0.2 0.3 0.4 0.5 D. Ammonium-NDry Season Wet Season* * Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek* * Figure 3 5. Seasonal variation in mean monthly concentration of nitrogen forms during 1991 2009. Values indicated by different letters are significantly different according to GLIMMIX procedure at P< 0.05. Dotted line represents the mean value

PAGE 64

6 4 Figure 3 6. Se asonal variation in proportion of organic, nitrate, and ammonium nitrogen duri ng 1991 2009. 0 20 40 60 80 100 Dry Wet Dry Wet Dry Wet Dry Wet Dry Wet Dry Wet Dry Wet Alafia Bell Shoals English Creek Turkey Creek North Prong South Prong Fishhawk Creek Ammonium N Nitrate N Organic N

PAGE 65

65 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 4 5 Bell ShoalsConcentration (mg L -1 ) Numeric nutrient criteria (0.79 mg L-1) 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 4 5 Numeric nutrient criteria (1.79 mg L-1)Alafia 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 4 5 English Creek 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 4 5 Turkey Creek 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 4 5 North Prong 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 4 5 South ProngN= 120 N= 225 N=120 N= 217 N= 225 N= 225 Mann-Kendall Slope = -2.75% per year, P = 0.18 GLIMMIX Slope = -1.02% per year, P = 0.17 Mann-Kendall Slope = +0.17% per year, P = 0.82 GLIMMIX Slope =+1.01% per year, P = 0.09 Mann-Kendall Slope = -4.84% per year, P = 0.19 GLIMMIX Slope =-1.14% per year, P = 0.44 Mann-Kendall Slope = +1.51% per year, P = 0.21 GLIMMIX Slope =+1.02% per year, P = 0.005 Mann-Kendall Slope = -0.42% per year, P = 0.64 GLIMMIX Slope =-1.11% per year, P = 0.76 Mann-Kendall Slope = +1.70% per year, P = 0.13 GLIMMIX Slope =+1.07% per year, P = 0.03 Figure 3 7. Longterm (1991 2009) trends in monthly flow un-weighted total N concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk Creek) sub-basins of the Alafia River Watershed. The dotted line indicates the proposed numeric nutrient (1.79 mg L1) criteria for the region (EPA, 2010).

PAGE 66

66 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 Bell ShoalsConcentration (mg L -1 ) Numeric nutrient criteria (0.79 mg L-1) 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 Alafia 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 English Creek 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 Turkey Creek 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 North Prong 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 South ProngN= 120 N= 225 N=120 N= 217 N= 225 N= 225 Mann-Kendall Slope = -6.24% per year, P = 0.45 GLIMMIX Slope =-2.45% per year, P = 0.04 Mann-Kendall Slope = +1.80% per year, P = 0.45 GLIMMIX Slope =+0.94% per year, P = 0.001 Mann-Kendall Slope = -3.72% per year, P = 0.41 GLIMMIX Slope =+0.65% per year, P = 0.38 Mann-Kendall Slope = +2.51% per year, P = 0.005 GLIMMIX Slope =+0.94% per year, P = 0.001 Mann-Kendall Slope = +0.72% per year, P = 0.46 GLIMMIX Slope =+1.00% per year P = 0.99 Mann-Kendall Slope = +1.95% per year, P = 0.25 GLIMMIX Slope =+0.95% per year, P = 0.001 Figure 3 8. Longterm (1991 2009) trends in monthly flow un-weighted organic N concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk Creek) sub -basins of the Alafia River Watershe d

PAGE 67

67 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 Bell ShoalsConcentration (mg L -1 ) Numeric nutrient criteria (0.79 mg L-1) 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 Alafia 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 4 English Creek 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 4 Turkey Creek 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 North Prong 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 South ProngN= 120 N= 225 N=120 N= 217 N= 225 N= 225 Mann-Kendall Slope =+5.44% per year, P = 0.16 GLIMMIX Slope =+0.96% per year, P = 0.50 Mann-Kendall Slope = -0.27% per year, P = 0.90 GLIMMIX Slope = -1.04% per year, P = 0.64 Mann-Kendall Slope = -3.23% per year, P = 0.42 GLIMMIX Slope =-1.74% per year, P = 0.32 Mann-Kendall Slope = +0.18% per year, P = 0.90 GLIMMIX Slope = +1.13% per year P = 0.21 Mann-Kendall Slope = -0.11% per year, P = 0.98 GLIMMIX Slope =-0.31% per year, P = 0.64 Mann-Kendall Slope = -2.7% per year, P = 0.98 GLIMMIX Slope = -0.98% per year, P = 0.54 Figure 3 9. Longterm (1991 2009) trends in monthly nitrate N concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk Creek) sub -basins of the Alafia River Watershed

PAGE 68

68 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0.0 0.1 0.2 0.3 0.4 0.5 Bell ShoalsConcentration (mg L -1 ) Numeric nutrient criteria (0.79 mg L-1) 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0.0 0.1 0.2 0.3 0.4 0.5 Alafia 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0.0 0.1 0.2 0.3 0.4 0.5 English Creek 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0.0 0.1 0.2 0.3 0.4 0.5 Turkey Creek 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0.0 0.1 0.2 0.3 0.4 0.5 North Prong 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0.0 0.1 0.2 0.3 0.4 0.5 South ProngN= 120 N= 225 N=120 N= 217 N= 225 N= 225 Mann-Kendall Slope =-3.7% per year, P = 0.81 GLIMMIX Slope =--1.00% per year, P = 0.76 Mann-Kendall Slope =-1.6% per year, P = 0.38 GLIMMIX Slope = -1.00% per year, P = 0.66 Mann-Kendall Slope = -15.7% per year, P = 0.12 GLIMMIX Slope =-0.92% per year, P = 0.30 Mann-Kendall Slope = +2.3% per year, P = 0.34 GLIMMIX Slope = +0.99% per year, P = 0.02 Mann-Kendall Slope = -3.6% per year, P = 0.22 GLIMMIX Slope =-0.87% per year, P = 0.34 Mann-Kendall Slope = +1.6% per year, P = 0.44 GLIMMIX Slope = +1.00% per year, P = 0.14 Figure 3 10. Longterm (1991 2009) trends in monthly flow un-weighted ammonium N concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk Creek) sub -basins of the Alafia River Watershed .

PAGE 69

69 0 10 20 30 40 0 1 2 3 4 0 10 20 30 40 0.0 0.5 1.0 1.5 2.0 0 10 20 30 40 0.0 0.5 1.0 1.5 2.0 y= 0.039x+0.857, r= 0.83 P<0.05Total Nitrogen Nitrate Nitrogen Organic Nitrogeny= 0.076x-0.014, r= 0.78 P<0.05 y= 0.0006x+0.676, r= 0.28 P>0.05 0 10 20 30 0 1 2 3 4 0 10 20 30 0.0 0.5 1.0 1.5 2.0 0 10 20 30 0.0 0.5 1.0 1.5 2.0 y= 0.033x+1.296, r= 0.49 P>0.05Total Nitrogen Nitrate Nitrogen Organic Nitrogeny= 0.025x+0.5367, r= 0.53 P>0.05 y= 0.003x+0.7515, r= 0.03 P>0.05Urban Land Use (%) Agricultural Land Use (%) Figure 3 11. Relationship between percent urban and agricultural land use and nitrogen forms in different sub -basins (* significant ly correlated at P< 0.05). M ean monthly data of 1991 1999 and 2000 2009 with land use of 1999 and 2007 respectively

PAGE 70

70 0 10 20 30 0 1 2 3 4 0 10 20 30 0.0 0.5 1.0 1.5 2.0 0 10 20 30 0.0 0.5 1.0 1.5 2.0 y=-0.032x+1.639, r=0.06 P>0.05Total Nitrogen Nitrate Nitrogen Organic Nitrogeny= 0.006x+0.713, r= 0.05 P>0.05 y= -0.009x+0.876, r= -0.34 P>0.05 10 20 30 40 0 1 2 3 4 10 20 30 40 0.0 0.5 1.0 1.5 2.0 10 20 30 40 0.0 0.5 1.0 1.5 2.0 y= -0.015x+1.893, r= -0.17 P>0.05Total Nitrogen Nitrate Nitrogen Organic Nitrogeny= -0.007x+0.949, r= -0.09 P>0.05 y= -0.007x+0.884, r= -0.28 P>0.05Pasture Land Use (%) Forest Land Use (%)Concentration (mg L -1 ) Figure 3 12. Relationship between pasture and forest land use and nitrogen forms in streams draining different sub -basins (* significantly correlated at P < 0.05) M ean monthly data of 1991 1999 and 2000 2009 with land use of 1999 and 2007 respectively

PAGE 71

71 CHAPTER 4 PHOSPHORUS TRANSPORT IN AN URBAN WATERSHED Abstract Non point source phosphorus (P) pollution is a significant concern in several waterbodies. In this study, we determined the concentrations of total P, dissolved reactive P (DRP), and other P forms in stream waters draining developed and undeveloped sub -basins, ranging in size from 19 to 350 km2, of the Alafia River Watershed (total drainage area: 1085 km2). During 1991 2009, mean monthly total P concentrations ranged from 0.56 to 3.95 mg P L1. Of total P, dissolved reactive P (DRP) was dominant (70 90% of total P) than other P (1 0 30% of total P) in developed and undeveloped sub-basins. None of the P forms were significantly ( P < 0.05) correlated with urban, agricultural, forest, and pasture la nd use of the sub-basin (r<0.50) indicating that the P concentrations are not controlled by land use of the Alafia River Watershed. Greater concentration of total P were greater in two developed (North Prong and English Creek: 2.18 2.53 mg P L1) may be du e to P rich geology, active mined lands and discharges of P rich wastewater in these sub basins. In all the developed and undeveloped sub-basins, the concentrations of P forms were greater in wet than dry season. This r epresents that the flushing of the P with greater rainfall runoff in wet season that might have accumulated due to dissolution and desorption of P from soil minerals. Long term trend analysis showed decreasing total P and DRP trends in both flow weighted and flow un-weighted concentrations. The decreasing trends in P concentrations indicated that the P abatement programs such as increased regulations on P discharges from mined lands as well as wastewater discharges were successful in controlling P pollution in the Alafia River Watershed. Duri ng 2000 2009, all the subbasin except Fishhawk Creek had greater total P concentrations (0.80 2.53 mg P L1) than EPA proposed numeric total P value of 0.739 mg L1 for the region. Results suggest s that P source controls from mined and

PAGE 72

72 wastewater discharg es in North Prong, English Creek, Turkey Creek, and South Prong are needed to control P pollution in the Alafia River Watershed. Introduction Phosphorus (P) pollution is the primary source of water quality degradation in the US (Bricker et al., 1998; USEP A, 2001) as total P concentration as low as 0.050 mg P L1 in lakes and 0.10 mg P L1 in stream waters can impair the water quality (US EPA, 1986) In the US, 45% of lakes and 35% of rivers are degraded and approximately 90% of the rivers show signs of eutrophication due to P enrichment (USEPA, 1996, 2000; Dodds et al., 2009; Paerl, 2009) The consequences of eutrophication include hypoxia, acidification of na tural waters, degradation of coastal waters including increased episodes of noxious algal blooms, and reductions in aquatic macrophyte communities often leading to substantial shifts in ecosystem structure and function (Carpenter et al., 1998; Dodds et al., 2009) The cost of eutrophication has been es timated at $2.2 billion per annum due to losses of recreational water usage, spending on recovery of threatened and endangered species, and drinking waters (Dodds et al., 2009) Therefore, controls on the sources of P can help to protect the water resources and reduce water quality deterioration in a region. Phosphorus in water bodies can come from either point sources such as wastewater and industrial effluents or non -point sources which include the storm water runoff losses from urban areas, agr icultural fields, animal feedlots, roadways, and mined areas (Edwards and Withers, 2008) Following the passage of Clean Water Act in 1970, P contributions from the point sources have decreased and consequently nonpoint source has become the dominant form of P pollution in many watersheds in the US (Bricker et al., 1998; U SEPA, 2001; Diebel et al., 2009; Maxted et al., 2009) A logical approach to control non -point P pollution may be to determine the hot spot areas that contribute greater P losses and to develop b est m anagement p ractices (BMPs) to

PAGE 73

73 reduce P pollution from these hot spots (Haygarth, 2005) In Wisconsin, US, Di ebel et al. (2009) reported that targeting 10% of watersheds decreased total P losses by 20% for the entire state. Therefore, federal, state and local government, nonprofits (e.g., The Nature Conservancy), and stakeholder groups (e.g., watershed associations, soil and water conservation districts) find watershed scale pollution assessment valuable for identifying and targeting the land areas which contribute greater pollution (hot spots) in a watershed. In order to identify the hot spots areas at a watershe d scale, it is vital to fully understand how different land uses impact non -point P pollution in a watershed. In general, l and use affects the net anthropogenic inputs of P in watersheds. For example, Russell et al. (2008) reported that net anthropogenic P inputs were greater in agricultural and urban land uses (15.8 19.6 kg P ha1) as compared to forest watersheds (1.6 kg P ha1). Another consequence of the land use change is the alteration of the flow paths. For example, in urban watersheds, impervious surfaces lead to increased runoff due to altered hydrology compared to forest watersheds (Arnold and Gibbons, 1996; Lee and Heaney, 2003; Paul and Meyer, 2001) In this way, urban land uses result in greater flows and facilitates the transport of suspended solids, an important P transport source (Mulliss et al., 1996; Stone and Droppo, 1994) Further, the loss of natural vegetation in urban land uses reduces recycling and uptake of P by vegetation that can immobilize P (Abelho, 2001; Wahl et al., 1997) This can result in a greater amount of P available and thus export to water bodies in urban land use dominating watersheds. The influence of land use on P exports from watersheds can be seen in Lake Washington, US, where Ellison and Brett, (2006) reported greater total P concentrations in streams draining agricultural (0.13 mg P L1) and urban (0.07 mg P L1) than forest sub basins (0.03 mg P L1). Similarly, in 17 watersheds domina ted with urban (22 87%) or forest (6 73%) land use, Brett et

PAGE 74

74 al. (2005) reported that with 10% conversion of forest to urban land use, total P concentrations increased by 0.07 mg P L1. In contrast to these studies, other researchers have not observed any significant difference in P losses from different land use dominated watersheds (Dodds and Oakes, 2006; Johnson et al., 1997) Therefore, losses of P may or may not be impacted by land uses in a watershed despite the fact that land use alters inputs as well as mode of delivery of P from watersheds to streams. Land use in a watershed may influence the proportion of the different P forms such as dissolved reactive P (DRP) and particulate P (PP) in the streams (Stone and Droppo, 1994) In general, PP losses are often associated with the erosion of soil particles that are enhanced by the anthropogenic land use and soil disturbance typical in agricultural dominated watersheds (Wallbrink et al., 2003) On the other hand, DRP in the stream waters represents the losses from the anthropogenic sources such as fertilizer application to the agricultural fields as well as to the weathering of the P minerals if present in a watershed (Harrison et al., 2005) Most of the studies have demonstrated that about 50% of the P losses occur as PP (Om ernik, 1977; Sharpley and Menzel, 1987; Vaithiyanathan and Correll, 1992) However, in Lake Washington, Ellision and Brett (2006) reported that of total P, the total dissolved P wa s 72% in urban (0.05 mg P L1) and 60 64% in forest and mixed (0.02 0.04 mg P L1) compared to 50% (0.07 mg P L1) in agricultural streams. Information on the losses of different P forms from sub-basins under contrasting land use activities may act as a useful tool in studying the cycling of P, which may be valuable to develop source control of P in sub -basins draini ng different land uses. In addition to the land uses, stream flow conditions affect the concentration and proportions of P losses from the watersheds (Ellison and Brett, 2006; Qian et al., 2007). For example, Royer et al. (2006) reported that in corn-soyb ean dominated (80 90%) watersheds in

PAGE 75

75 Illinois, US, >80% of the total P losses occur only during the high flow conditions (>90% percentile flow). In Indian River Lagoon, Florida, Qian et al. (2007) reported that the concentrations of P forms were two times greater during wet (DRP: 0.23 mg P L1; t otal P: 0.31 mg P L1) than dry season (DRP: 0.11 mg P L1; t otal P: 0.15 mg P L1). In general, P is sorbed onto the soil particles and greater losses of P occur when sufficient water is available to transport soil particles from land to streams. In this way, ecological impacts due to P pollution may depend upon the flow conditions (Edwards et al., 2000; Jarvie et al., 2006; Svendsen et al., 1995) The Alafia River Watershed is an example of such a watershed, where land use has bee n continuously changing from natural areas to urban lands. As a result, several subbasins of the watershed have modified hydrology due to stormwater retention ponds and water convergence structures that are meant to drain water during high flow events to avoid flooding. In addition, two sub basins of the watershed are dominant in mined lands where phosphate mining is a commercial enterprise. We hypothesize that the 1) concentration of P forms may be different in sub -basins that drain different land uses, 2 ) concentration of P forms may differ in dry and wet seasons due to difference in rainfall, and 3) urbanization of the watershed may increase the P concentrations. The objectives of this study were to (1) determine how different sub-basins (with different land uses) influence concentrations of different forms of P in stream waters; (2) investigate the influence of two distinct seasons (dry and wet) on concentrations of P forms in stream waters, and (3) evaluate the long term trends of P losses in different sub -basins of the watershed. Materials and Methods Study Site Description Refer to C hapter 2 for detailed description of sub-basins of the Alafia River Watershed.

PAGE 76

76 Data Collection Monthly concentrations of total P, and dissolved reactive P (DRP) from 1991 to 2009 were obtained from the Environmental Protection Commission of Hillsborough County (http://www.epchc. org). Among the two mainstem stations, water quality data was available only for Bell Shoals station for 19 -years study period, while for Alafia station, the data was available for 1999 to 2009 (Table 41). Among other sub -basins in the watershed, 19 years of water quality data w ere available only for North Prong, South Prong, and Turkey Creek while the data availability for English Creek was f r om 1999 to 2009 and for Fishhawk Creek from 2005 to 2009. Stream water Collection and Analysis Each month, a grab sa mple from surface water was collected from the surface of channel thalweg (centre of the stream flow) in plastic water bottles by the EPCHC staff. Before collection of the samples, the water bottles were rinsed three times with the stream water. The collec ted samples were chilled with ice and transported to the laboratory where samples were stored at 4 oC prior to analysis. Environmental Protection Commission Hillsborough County staff analyzed the surface water samples for total P (EPA 365.4 method) and DRP ( SM 4500P method) using a discrete analyzer (Seal Analytical, Model AQ2 Mequon, WI). Other P (OP) was calculated as follows: OP = total P DRP. Statistical Analysis Mean, median, and standard deviations were calculated using MS Excel 2007. We used GLIMMIX time series procedure SAS Version 9.1 (SAS Institute, Cary, NC) to compare the P forms across different sub -basins as well as to determine the long term trends in P forms. The GLIMMIX procedure fits models to data with correlations or non constant variabi lity and assumes normal random effects (SAS Institute, 2008). In this time series analysis two kinds of

PAGE 77

77 effects were used 1) fixed effects which included stream station and season of the year, and 2) random effects, which were the date of sample collection For the first fixed effect (stream station), each station was treated as a group and the variance was pooled over the sites; similarly for the second fixed effect (season), each wet (June September) and dry (October May) season at each station was treate d as one group of observations. The date of sample collection from 1991 2009 for the each station was considered as random effect and therefore a time series was constructed. With this, we compared 1) the difference in P forms at each of the station averag ed over all sampling dates and 2) differences in pooled mean concentrations of P forms during wet and dry seasons at each station using P < 0.05 as significance level. In addition, same time series with fixed and random effects were used to determine the tre nds in P forms over the period of study. The concentrations of different forms of P were logarithmically transformed to equalize variances and normalize skewed data and were back transformed to present means of P forms in more relevant manner. For the long term water quality data, various techniques have been used (Richards, 2006; Johnson et al., 2009; Goodrich et al., 2009). However, non parametric Mann-Kendall and Seasonal Kendall are among the most common methods for determination of long term trends of water quality (Daroub et al., 2009; Johnson et al., 2009). Therefore, in addition to GLIMMIX procedure, we determined the long term trends using Seasonal Kendall procedure. In Seasonal Kendall trend is calculated by comparing all potential data pairs. If t he later value in the pair (in time) is higher than the first, a plus sign is scored. If the later value in the pair is lower than the first, a minus sign is scored. If the results find an equal number of pluses and minuses, then there is no discernible tr end. In other words, there is just as much of a likelihood that a pair of data

PAGE 78

78 values will be higher (or lower) than the next one. If the results show more pluses than minuses, this would indicate that a positive trend is likely (Gilbert, 1987; Helsel and Hirsch, 1991). The trend in both the procedures is defined by the rate of change over time, which is referred to as the trend slope (Sen 1968; Schertz et al. 1991). The trend slope can be expressed either as change in original units per year [S (0)], or a s a percent of the mean concentration of water quality variable [S (%)]. The former is the median slope of all pair wise comparisons (each pair wise difference is divided by the number of years separating the pair of observations) while the latter is prod uced by dividing the slope (in original units per year) by the mean and multiplying by 100. The P forms concentrations data w ere divided into two time periods : 1991 2000 and 2001 2009. To determine the effect of land use change on P forms in corresponding sub -basins, the land use data of 1999 and 2007 was correlated with P concentrations data of 1991 2000 and 2001 2009, respectively using SAS PROC CORR procedure Results Chemical Characteristics of Stream Waters Refer to Chapter 3 for detailed description of chemical characteristics of the stream waters Concentrations of Phosphorus Forms in Streams Draining Different Sub-basins Mean monthly total P concentration at the mainstem station (Bell Shoals) that drains 89% of the watershed was 1.76 mg P L1 during 1991 1999 (Fig. 4 2A). Among the sub-basins, total P concentrations were 4 5 folds greater in one developed (North Prong: 3.95 mg P L1) than one developed (Turkey Creek: 0.78 mg P L1) and one undeveloped (South Prong: 0.89 mg L1) sub basin. The concentrations of total P decreased with time in the Alafia River Watershed (Fig. 4 2A) For example, at mainstem st ations (Bell Shoals), mean total P con centrations were lower during 2000 2009 (1.24 mg P L1) than 1991 1999 ( 1.76 mg P L1) Similar to the mainstem

PAGE 79

79 station, at North Prong, South Prong, and Turkey Creek, the total P concentrations decreased from 0.9 3.9 mg P L1 in 1991 1999 to 0.6 2.5 mg P L1 in 2000 2009. Among the subbasins, t wo developed sub-basins had significantly ( P < 0.05) gre ater total P concentrations (North Prong and English Creek: 2.18 2.53 mg P L1) than one developed (Turkey Creek; 0.90 mg P L1) and two undeveloped sub-basins (0.56 0.80 mg P L1). At the mainstem station (Bell Shoals), m ean DRP concentration was greater ( 1.46 mg L1) during 1991 1999 than 2000 2009 (1.00 mg L1) (Fig. 4 2 B). Similarly, at two developed (North Prong and Turkey Creek) and one undeveloped (South Prong), the concentration of DRP decreased from 0.8 3.3 mg L1 in 1991 1999 to 0.7 2.2 m g L1 in 2000 2009. During 2000 2009, a ll the three developed sub basins had significantly ( P < 0.05) greater DRP concentrations (North Prong, South Prong, and Turkey Creek: 0.63 2.18 mg P L1: 70 90% of total P) than one undeveloped sub-basin (Fishhawk Cre ek: 0.40 mg P L1: 72% of total P). However, North Prong and English Creek had 3 4 folds greater DRP concentrations (1.95 2.18 mg P L1) than Turkey Creek (0.63 mg P L1). Among the undeveloped sub-basins, South Prong had significantly ( P < 0.05) greater to tal P concentration (0.68 mg P L1) than Fishhawk Creek (0.40 mg P L1). At the mainstem station (Bell Shoals), m ean monthly OP concentration s w ere similar during 1991 1999 (0.30 mg P L1; 17% of total P) and 2000 200 9 (0. 29 mg P L1; 1 9 % of total P) (Fig. 4 2C). At two developed (North Prong and Turkey Creek) and one undeveloped (South Prong) sub-basin, the concentrations of OP were slightly greater (0.09 0.6 mg P L1) during 1991 1999 than 2000 2009 (0.1 0.3 mg P L1). Among the sub -basins, mean OP concentrations were slightly greater in developed (0.21 0.34 mg P L1) than undeveloped sub -basins (0.12 0.15 mg P L1). The greatest concentrations of other P forms were at North Prong (0.34 mg P L1) and lowest at South Prong (0.12 mg P L1) during 2000 2 009

PAGE 80

80 Seasonal Variation in Phosphorus Concentrations in Streams Draining Different Sub basins The water quality data from 1991 2009 showed that total P concentrations in stream water were greater, but not statistically significant ( P < 0.05) in wet than d ry season at all study stations (Fig. 4 3A). At two mainstem stations, total P concentration was 40 45% greater in wet (1.26 1.88 mg P L1) than dry season (0.90 1.30 mg P L1). In developed sub-basins, the concentration of total P was 63% greater at Engli sh Creek (2.93 mg P L1) followed by 38% greater at Turkey Creek (1.04 mg P L1) and 9% greater at North Prong (3.42 mg P L1) in wet than dry season. In undeveloped sub-basins, total P concentrations were 9 20% greater in wet (0.60 0.95 mg P L1) than dry season (0.55 0.79 mg P L1). Similar to total P, mean concentration of DRP was greater, but statistically ( P <0.05) similar in wet and dry season s at all the study stations (Fig. 4 3B). At two mainstem stations, the DRP concentration was 21 46% greater in wet (1.03 1.39 mg P L1) than dry season (0.71 1.14 mg P L1). Among the developed sub-basins, DRP increased by 55 60% (from 0.53 1.60 mg P L1 to 0.83 2.56 mg P L1) at Turkey Creek and English Creek followed a comparatively lower increase at North Prong (from 2.68 to 2.85 mg P L1; 6%) in wet than dry season. In undeveloped sub -basins, the increase in total P concentration was ~10% (from 0.39 0.71 mg P L1 to 0.42 0 .79 mg P L1) in wet than dry season. Proportion of DRP was 71 90% of total P in wet and dry seasons at all the study stations (Fig. 4 4). Among the developed sub-basins, overall proportion of DRP was greater at English Creek and North Prong (83 89% of to tal P) than Turkey Creek (71 79% of total P). Among the undeveloped sub-basins, the greater proportion of DRP was at South Prong (83 89% of total P) than Fishhawk Creek (71%).

PAGE 81

81 Mean concentration of O P forms during dry and wet season s were statistically ( P < 0.05) similar; however the magnitude of variation was greater than total P and DRP at all the sub basins (Fig. 4 3C). For example, other P was ~220% greater at Bell Shoals (0.49 mg P L1) and 10% greater at Alafia (0.23 mg P L1) in wet compared to dry se ason. Among the developed subbasins, other P concentrations increased by 80% at English Creek (from 0.20 to 0.36 mg P L1), 26% at North Prong (from 0.45 to 0.57 mg P L1), 2% at Turkey Creek (from 0.22 0.24 mg P L1) in the wet than dry season. Among the undeveloped subbasins, the increase in O P concentration in wet compared to dry season was 100% at South Prong (from 0.08 to 0.16 mg P L1) and 8% at Fishhawk Creek (from 0.15 to 0.17 mg P L1). Proportion of O P to total P was similar: 11 43% in dry seas on and 14 44% in wet season at all sub -basins (Fig. 4 4). However, a t two mainstem stations, proportion of O P was greater in wet (22 40% ) than dry (15 29%). Long T erm Trends in Concentrations of Flow Unweighted Phosphorus Forms in Streams Draining Different Sub -basins The longest data record (19 years) was available only for one mainstem station (Bell Shoals ) that drain s 89% of the watershed area and three other subbasins (Fig. 4 5). The results of the Seasonal Kendall trend analy sis showed a significant ( P <0.002) decreasing flow unweighted total P concentration trend ( 3.8% per year; 57.6 g P L1 per year) at Bell Shoals. This decreasing trend equates to ~72% decrease in mean total P concentrations during 1991 2009. Total P conc entrations trends were decreasing at North Prong ( 4.8% per year; 104.2 g P L1 per year; P <0.001) and South Prong ( 2.2% per year; 14.4 g P L1 per year; P < 0.03). In contrast to these sub-basins, Turkey Creek showed non-significant ( P <0.29) total P con centrations trend during 1991 2009. Among the stations with 10-years of data, English Creek showed a significant decreasing total P trends ( 36.5% per year; P < 0.008).

PAGE 82

82 Similar to the Seasonal Kendall trend analysis, GLIMMIX model showed a significant ( P <0 .001) overall decreasing total P trend at Bell Shoals ( 0.9% per year) during 1991 2009 (Fig. 4 5). At Turkey Creek, total P showed an increasing ( P < 0.001) trend (+1.1% per year) In contrast to Turkey Creek, other two sub-basins with 19 -years data record showed significant ( P <0.001) but with lower magnitude decreasing total P concentration trends ( 0.3 and 1.1% per year) The total P trends at two stations with 10 -years of data record were decreasing (from 1.09 to 1.11% per year). Seasonal Kendall tre nd analysis showed a decreasing DRP concentration trend ( 4.4% per year; P <0.003) during 1991 2009 (Fig. 4 6). Among the sub-basins with 19 -years of data record, DRP concentration trends showed a non significant increase at Turkey Creek (+0.5% per year; P < 0.55) compared to a significant decreasing DRP trends at North Prong ( 4.2% per year; P <0.002) and South Prong ( 2.3% per year; P<0.09) The stations with 10-years of data record showed decreasing DRP concentration trends at Alafia ( 5.6% per year; P<0.37) and English Creek ( 34.3% per year; P<0.009) during 1999 2009 (Fig. 46). GLIMMIX model showed a significant ( P <0.001) decreasing DRP concentration trend ( 1.14% per year; P< 0.001) at Bell Shoals (Table 4 6). Turkey Creek had increasing DRP concentrati ons trend ( +0.99% per year; P<0.0001) as compared to decreasing trends at North Prong ( 1.03% per year; P<0.001) and South Prong ( 0.98% per year; P< 0.03) During 1999 2009, decreasing DRP trends were found at English Creek ( 1.23% per year; P<0.001) and Alafia ( 0.89% per year; P<0.018). Both the Seasonal Kendall and GLIMMIX procedures showed non -significant trends for O P concentration at all sub-basins (Fig. 4 7). However, other P concentration trends were decreasing at Bell Shoals (from 1.06 to 1.60% per year) during 1991 2009.

PAGE 83

83 Long Term Trends in Flow Weighted Concentrations at Mainstem Station At the mainstem stati on that drains 89% of the Alafia River Watershed showed a significant ( P < 0.041) decreasing total P concentration trend ( 1.14% per year) during 1991 2009 (Table 4 2). Mean total P concentration at the mainstem station was greater during 1991 1999 (1.76 mg P L1) and 2000 2009 (1.24 mg P L1) thereby reflecting the decreasing total P concentration trends in terms of actual concentrations Similarly, DRP showed significant ( P < 0.016) decreasing trend ( 1.16% per year) during 1991 2009 (Table 4 2). In contras t to total P and DRP, the flow weighted concentration trends were insignificant in O P forms ( P <0.561) at Bell Shoals. Long Term Trends in Phosphorus Loads at Mainstem Station At mainstem station (B ell Shoals), total P loads were not significant during 1 991 2009 (Table 4 2). The non significant trends in the P loads during wet and dry seasons could be due to greater variations in P loads Discussion Land Use Impacts on Phosphorus Concentrations in Streams During 2000 2009, at two mainstem stations (Alafi a and Bell Shoals), total P concentrations were greater (1.02 1.76 mg P L1) than the EPAs proposed numeric total P value of 0.739 mg P L1 for rivers in the Tampa Bay (EPA, 2010). In our study, total P concentrations were an order of magnitude greater th an total P in Miller Creek Watersheds in Kansas, US (0.008 0.22 mg P L1, mean= 0.031 mg P L1) reported by Dodds and Oakes (2006). Concentrations of total P in our study were also greater than corn -soybean dominated (>60%) watersheds of Indiana, USA ( 0.13 0.23 mg P L1) ( Vidon et al., 2008). It is important to note that the Alafia River flows over a geologic P rich formation (fluorapatite) that is commercially mined

PAGE 84

84 for P (Lane, 1994). As a result, P concentration is comparatively higher in Alafia River th an other studies conducted across US Mean monthly concentrations of total P were greater in two developed sub-basins (North Prong and English Creek) than other sub-basins (Fig. 4 2). This could be attributed to active phosphate mining in North Prong (39% mined land use) and small animal feeding operations in English Creek may have elevated the P concentrations in stream waters (NPDES, 2009). In addition, a wastewater treatment plant discharge s 0.35 m3 sec1 of wastewater with mean total P concentration of 3.1 mg P L1 in the North Prong. In our study, total P concentrations were not significantly ( P <0.05) correlated with urban, agricultural, forest, and pasture land uses thereby indicating that P losses are not significantly affected by land uses in the Alafia River Watershed (Fig. 4 8, 4 9). Negative (r= 0.47: P >0.05) correlations between total P and pasture land uses might be due to the fact that pastures may act as a potential buffer to reduce the transport of P from wa tersheds to streams (Abelho, 2001; Wahl et al., 1997; Zaimes et al., 2008) On the other hand, slightly positive but not significant correlation between total P and urban land use (r= +0.33) may be due to greater impervious areas which increase the runoff generation to facilitate the transport of P from the watersheds (Arnold and Gibbons, 1996; Lee and Heaney, 2003; Paul and Meyer, 2001) It appears that the P concentration in the streams of the Alafia River Watershed is controlled by the geology and presence of active mined land use along with point source input from domestic and industrial (phosphate mining) wastewater In our study, lower concentrations of total P at South Prong (66% mined land use) may be due to the fact that most of the mined lands discharge wastewaters to the sub-basin only during extremely high flow events and therefore is not impacted with mining activities.

PAGE 85

85 Seasonal Impact on Phosphorus Concentrations in Streams In our study, the concentrations of P forms were greater but not significant ( P < 0.05) in wet than dry season at all the sub -basins (Fig. 4 3). The increase in concentrations of P forms in wet season was not significant due to greater variations in P concentrations. How ever, higher concentrations of P forms in the wet season represent the influence of nonpoint sources on P pollution in the Alafia River Watershed. Previous studies have reported greater losses of P during high flow conditions. For instance, in high fertil izer inputs corn -soybean dominated watersheds of the Midwest, US, Royer et al. (2006) reported that >80% of the P losses occurred during the extreme discharges (>90% percentile flow). Similarly, in Indian River Lagoon, Florida, due to application of fertilizers to agricultural crops in wet season the concentrations of both DRP and total P were two times greater in wet than dry season (Qian et al., 2007) Research has indicated that most solute losses occur when large quantities of solutes are present in the landscape coupled with increased flows (Boyer et al., 1997; Verseveld et al., 2009) Although mechanisms controlling the temporal pattern in P concentrations were not directly investigate d, it seems probable that during the dry season the dissolution and desorption of P occurs in geologically P rich Alafia River Watershed which is subsequently flushed with increasing discharge in wet season (Chen and Driscoll, 2009; Wetzel, 2001) Another probable reason for greater concentration of P forms could be the release of P due to suspension of the stream sediments with the high flow events during the wet season (Svendsen et al., 1995) In our study, the concentratio ns of other P forms were greater in wet than dry season; however the proportions of DRP and other P forms remained similar during both wet and dry seasons (Fig. 4 4). This contradicts previous studies which reported the lower proportion of DRP in wet seas on (Ellison and Brett, 2006; Pacini and Gachter, 1999) In general, streams are characterized by higher proportions of DRP in dry season when the sediment transport capacity

PAGE 86

86 is low as fine bed sediments are the primary source of PP and the proportions of PP increase with the transport of greater particulates into the streams during high flow conditions in wet season (Ellison and Brett, 2006; Pacini and Gachter, 1999) Similar proportions of DRP and other P forms during dry and wet seasons in our study suggests that desorption of P from the soil particles and slow dissoluti on of P minerals during dry season is the dominant source of P which flushed during wet season. Long T erm Trends in Concentration of Phosphorus Forms in Stream Waters The Seasonal Kendall and GLIMMIX procedures showed similar trends (positive or negativ e), though with different magnitude of change (slope %) (Fig. 4 5; 4 -6; 4 7). These differences in trend slopes appear to be associated with the conceptual differences in the way of calculating the slopes in both the methods. For example, Seasonal Kendall compares the change in concentrations of nutrients with time (slope) for each point and the median slope is calculated as a summary statistic describing the magnitude of the trend (Johnson et al., 2009; Qian et al., 2007) On the other hand, GLIMMIX procedure fits the linear lines through the data and slope is the rate of change in concentrat ion with time at any of two points ( SAS, 2008). Therefore, the differences in the magnitude of slopes appear in the trend analysis in two methods. In general, Kendall trend analysis determines the trends at a single location. However, GLIMMIX procedure can be used to compare the mean concentrations at multiple locations as well as has the ability to include many parameters such as stream discharge, seasonal variation, distance of the mainstem, and l ocation of the station etc (Goodrich et al., 2009) In this study, GLIMMIX procedure has shown the significant variations in mean and seasonal variations in P concentrations at different stations. In general, flow weighted concentration t rends represent the changes due to anthropogenic activities since the natural changes due to variations in flow are minimized. On the other hand,

PAGE 87

87 flow un -weighted concentrations trends reflect the net effects of all natural and anthropogenic influences on concentration, allowing for the assessment of nutrient concentrations in streams relative to water quality standards of Environmental Protection Agency and the condition of aquatic communities (Sprague and Lorenz, 2009; Johnson et al., 2009). In our study, the flow weighted and flow un -weighted concentrations showed significantly decreasing trends at Bell Shoals during 1991 2009. This has suggested that the anthropogenic activities were responsible for decreasing P trends in the Alafia River Watershed. Simi lar to the Alafia River Watershed, the controls on P losses from watersheds have been documented in 50% rivers in the US in last 2 3 decades (Alexander and Smith, 2006). For instance, in Minnesota River, US, Johnson et al., (2009) using Seasonal Kendall tr end analysis reported the decreasing trends of total P ( 2.0% per year) and DRP ( 1.83% per year) during 1976 2003. They attributed the decrease in P concentrations trends to the conservation programs such as plantation of native grasses, buffer strips, we tland restoration, and reduction in agricultural activities near the streams. Similarly, in the Everglades Agricultural Area FL, Daroub et al. (2009) reported decreasing total P concentration trends due to implementation of agricultural BMPs during 1992 2002. In contrast to these watersheds, the Alafia River Watershed has P rich geology and therefore is commercially mined for P. The discharges of P rich wastewater from the mining operations have been reduced as a result of greater regulations in the water shed (NPDES, 2009). Further, the mined land uses have been reclaimed in most of the sub-basins during 1991 2009 resulting in decreased losses of P from the mined lands. In the Alafia River Watershed, population has grown from 0.83 million in 1990 to 0.99 million in 2000 and 1.20 million in 2009 (US Census Bureau, 2010). This represents 19.8% increase during 1990 2000 and 19.7% during 2000 2009. The population growth has resulted in

PAGE 88

88 the land use changes in the watershed. For example, the urban land use (res idential and built up) has increased by 8% (from 12 to 20 %) of the watershed during 1990 2007 (SWFWMD, 2007). This is coupled with a decrease of 5 % in forests (from 2 3 to 19%) and 8% decrease in pastures (from 19 to 11%) while agricultural land use remained similar (8%) during the study period. The decreasing total P and DRP concentrations trends suggest that the land use change has not resulted in increasing P concentr ations in the Alafia River Watershed. This contradicts the previous studies which have suggested the greater losses of P with urbanization (Brett et al., 2005; Ellison and Brett, 2006). In our study watershed, the constructions of mandatory urban storm wat er retention ponds is an important feature which might have played an important role in P removal through burial of P in retention ponds sediments (Alexander et al., 2000; Bosch, 2008; Evans et al., 2004) Further, increased regulations on the discharges of P from the wastewaters and mining activities might have masked the effect of urbanization (Alexander and Smith, 2006) Summary T he mainstem stations of the Alafia River Water (Alafia and Bell Shoals) had greater mean total P concentrations (1.01 1. 76 mg P L1) than EPAs proposed numeric total P value of 0.739 mg P L1for the region. Greater total P concentrations at two developed sub -basins (North Prong and English Creek: 2.2 2.5 mg P L1) may be due to P rich geology as well the discharge of P rich wastewater in these sub -basins. Of total P, DRP was dominant (70 90% of total P) than other P form s (10 30% of total P) probably because of disso lution of P rich minerals in the w atershed. Greater concentrations of P forms in wet than dry season may be due to flushing of P accumulated due to dissolution of P minerals Longterm trend analysis showed d ecreasing flow un -weighted and flow weighted concentrations of total P and DRP thereby suggesting that the P abatement programs such as increased re gulations on the wastewaters facilities and reclamation of mined lands might have resulted in reducing concentrations of P in the Alafia River

PAGE 89

89 Watershed. Despite the decreasing P trends, the concentration of total P was still greater than the EPAs propose d numeric nutrient criteria (0.739 mg P L1) for the region. Therefore, the reduction in P loss from mined lan ds and wastewater discharges in four sub -basins (North Prong, English Creek, Turkey Creek, and South Prong) may result in reducing total P concent rations in the Alafia River Watershed

PAGE 90

90 Table 4 1. Station characteristics of the Alafia River Watershed Sub basin Station Sampling location Drainage area Land Use in 2007 Lat Long km 2 % Residential Built up Agricultural Pasture Forest Mined Mainstem Stations Alafia 2301718 27.87 82.32 1072 99 17 3 8 11 18 32 Bell Shoals 2301638 27.86 82.26 974 89 16 3 8 12 18 33 Developed English Creek 27.93 82.06 99 9 21 14 19 23 25 3 Turkey Creek 27.91 82.18 128 13 20 3 24 16 12 0 North Prong 2301000 27.86 82.13 350 32 18 6 4 5 16 39 Undeveloped South Prong 2301300 27.86 82.13 277 26 3 1 4 9 15 66 Fishhawk Creek 27.85 82.24 70.6 7 11 3 14 23 32 0 USGS station not present Table 4 2. Long term trends in flow weighted and loads of P forms at Bell Shoals Parameter D ata Flow weighted concentrations Loads Trend Slope (%) p value Trend Slope p value Total Phosphorus 1991 2009 Decreasing 1.14 0.041 No trend 0.535 Dissolved Reactive P 1991 2009 Decreasing 1.16 0.016 Other P 1991 2009 No trend 0.561

PAGE 91

91 Figure 4 1. Location map of the Alafia River Watershed

PAGE 92

92 0 2 4 6 Concentration (mg P L -1 ) 0 1 2 3 4 5 6 0 1 2 A. Total Phosphorus B. Dissolved Reactive Phosphorus C. Other Phosphorus 75 percentile 25 percentile Median Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek1991-1999 2000-2009 1991-1999 2000-2009 1991-1999 2000-2009Bell Shoals Turkey Creek North Prong South Prong Bell Shoals Turkey Creek North Prong South Prong Bell Shoals Turkey Creek North Prong South ProngB B BC A C A A B A C A B B C B C B A B A C A A A A A A A A A B B A Figure 4 2. Summary of mean monthly concentrations of total, dissolved reactive, and other phosphorus forms during 1991 2009 in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk Creek). Values indicated by different letters are significantly different according to GLIMMIX procedure at P< 0.05. N in each sub-basin indicates number of months/observations

PAGE 93

93 0 1 2 3 4 Concentration (mg P L -1 ) 0 1 2 3 4 0 1 2 3 4 Dry Season Wet SeasonA. Total Nitrogen B. Organic Nitrogen C. Nitrate-NN 120 225 120 207 225 225 74 = 75 percentile 25 percentile MedianDry Season Dry Season Wet Season Wet Season 0.0 0.1 0.2 0.3 0.4 0.5 D. Ammonium-NDry Season Wet Season* * Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek Alafia Bell Shoals English Turkey North South Fishhawk Creek Creek Prong Prong Creek* * Figure 4 3. Seasonal variation in mean monthly concentration of phosphorus forms during 1991 2009. Values indicated by diff erent letters are significantly different according to GLIMMIX procedure at P< 0.05. Dotted line represents the mean value

PAGE 94

94 Figure 4 4. Seasonal variation in contribution of organic, nitrate, and ammonium nitrogen to total nitrogen during 1991 2009. 0 20 40 60 80 100 Dry Wet Dry Wet Dry Wet Dry Wet Dry Wet Dry Wet Dry Wet Alafia Bell Shoals English Creek Turkey Creek North Prong South Prong Fishhawk Creek Percentage of total P (%) Dissolved Reactive P Other P

PAGE 95

95 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 2 4 Bell ShoalsConcentration (mg P L -1 ) Numeric nutrient criteria (0.79 mg L-1) 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 2 4 Numeric nutrient criteria (0.79 mg L -1 ) Mann-Kendall Slope = -8.3% per year, P = 0.45 GLIMMIX Slope = -1.09% per year P =0.003 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 2 4 6 8 English Creek 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 2 4 Turkey Creek 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 2 4 6 8 North Prong 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 2 4 South ProngN= 120N= 225 N=120 N= 217 N= 225 N= 225 Alafia Mann-Kendall Slope = -3.8% per year, P = 0.002 GLIMMIX Slope = -1.07% per year P = 0.001 Mann-Kendall Slope = -36.5% per year, P = 0.008 GLIMMIX Slope =-1.11% per year P = 0.001 Mann-Kendall Slope = +0.9% per year, P = 0.29 GLIMMIX Slope =-1.08% per year P = 0.0001 Mann-Kendall Slope = -4.8% per year, P = 0.002 GLIMMIX Slope = -1.03% per year, P = 0.0001 Mann-Kendall Slope = -2.2% per year, P = 0.03 GLIMMIX Slope =-0.93% per year P = 0.024 Figure 4 5. Longterm (1991 2009) trends in mean monthly total P concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhaw k Creek) sub-basins of the Alafia River Watershed. The dotted line indicates the proposed numeric nutrient (0.79 mg L1) criteria for the region (EPA, 2010)

PAGE 96

96 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 4 Bell ShoalsConcentration (mg P L -1 ) Numeric nutrient criteria (0.79 mg L-1) 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 4 Alafia 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 2 4 6 8 English Creek 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 4 Turkey Creek 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 2 4 6 8 North Prong 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0 1 2 3 4 South Prong N= 120 N= 225 N=120 N= 217 N= 225 N= 225 Mann-Kendall Slope = -5.6% per year, P = 0.37 GLIMMIX Slope = -0.89% per year P =0.018 Mann-Kendall Slope = -4.4% per year, P = 0.003 GLIMMIX Slope = -1.14% per year P = 0.001 Mann-Kendall Slope = -34.3% per year, P = 0.009 GLIMMIX Slope =-1.23% per year P = 0.001 Mann-Kendall Slope = +0.5% per year, P = 0.55 GLIMMIX Slope =+0.99% per year P = 0.0001 Mann-Kendall Slope = -4.2% per year, P = 0.002 GLIMMIX Slope = -1.03% per year, P = 0.0001 Mann-Kendall Slope = -2.3% per year, P = 0.09 GLIMMIX Slope =-0.98% per year P = 0.024 Figure 4 6. Longterm (1991 2009) trends in mean monthly dissolved reactive P concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk Creek) sub-basins of the Alafia River Watershed

PAGE 97

97 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0.0 0.5 1.0 1.5 2.0 Bell ShoalsConcentration (mg P L -1 ) Numeric nutrient criteria (0.79 mg L-1) 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0.0 0.5 1.0 Alafia 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0.0 0.5 1.0 1.5 2.0 English Creek 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0.0 0.5 1.0 1.5 2.0 Turkey Creek 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0.0 1.0 2.0 3.0 North Prong 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 0.0 0.5 1.0 South Prong Mann-Kendall Slope = -3.7% per year, P = 0.81 GLIMMIX Slope = -1.03% per year P =0.071 Mann-Kendall Slope = -1.6% per year, P = 0.38 GLIMMIX Slope = -1.06% per year P = 0.170 Mann-Kendall Slope = -15.7% per year, P = 0.12 GLIMMIX Slope =-0.99% per year P = 0.004 Mann-Kendall Slope = +2.33% per year, P = 0.34 GLIMMIX Slope =+0.67% per year P = 0.0003 Mann-Kendall Slope = -3.64% per year, P = 0.22 GLIMMIX Slope = -0.95% per year, P = 0.017 Mann-Kendall Slope = -1.66% per year, P = 0.44 GLIMMIX Slope =-0.87% per year P = 0.321N= 120 N= 225 N=120 N= 217 N= 225 N= 225 Figure 4 7. Longterm (1991 2009) trends in mean monthly other phosphorus forms concentrations in mainstem (Alafia and Bell Shoals), developed (English Creek, Turkey Creek, and North Prong), and undeveloped (South Prong and Fishhawk Creek) sub -basi ns of the Alafia River Watershed

PAGE 98

98 0 10 20 30 40 0 2 4 0 10 20 30 40 0 2 4 0 10 20 30 40 0.0 0.2 0.4 0.6 0.8 1.0 y= 0.056x+0.709, r= 0.33 P>0.05Total Phosphorus Dissolved Reactive Phosphorus Other Phosphorusy= 0.046x+0.580, r= 0.36 P>0.05 y= 0.012x+0.103, r= 0.22 P>0.05 0 10 20 30 0 2 4 0 10 20 30 0 1 2 3 4 5 0 10 20 30 0.0 0.2 0.4 0.6 0.8 1.0 y=-0.045x+1.984, r= -0.35 P>0.05 y= -0.401x+1.658, r= -0.36 P>0.05 y= -0.0028x+0.315, r=-0.29 P>0.05Urban Land Use (%) Agricultural Land Use (%)Concentration (mg P L -1 )Total Phosphorus Dissolved Reactive Phosphorus Other Phosphorus Figure 4 8. Relationship between percent urban and agricultural land use and phosphorus forms in different sub basins (* significantly correlated at P< 0.05). The mean monthly data o f 2001 2009 was correlated with the land use during 2007.

PAGE 99

99 0 10 20 30 0 2 4 0 10 20 30 0 2 4 0 10 20 30 0.0 0.2 0.4 0.6 0.8 1.0 y= -0.100x+2.843, r= -47 P>0.05Total Phosphorus Dissolved Reactive Phosphorus Other Phosphorusy=-0.0905x+2.439, r= -0.50 P>0.05 y=-0.008x+0.394, r= -0.37 P>0.05 10 20 30 40 0 2 4 10 20 30 40 0 2 4 10 20 30 40 0.0 0.2 0.4 0.6 0.8 1.0 y= -0.023x+1.973, r= -0.10 P>0.05y= -0.022x+1.667, r= 0.09 P>0.05y= -0.001x+0.2631, r= -0.12 P>0.05Pasture Land Use (%) Forest Land Use (%)Concentration (mg P L -1 )Total Phosphorus Dissolved Reactive Phosphorus Other Phosphorus Figure. 4 9. Relationship between pasture and forest land use and phosphorus forms in streams draining different sub -basins (* significantly correlated at P< 0.05)

PAGE 100

100 CHAPTER 5 SUMMARY, CONCLUSIONS AND RECOMMENDATION The anthropogenic activities such as urbanization and agriculture have increased concentrations of nitrogen (N) and phosphorus (P) in lakes, rivers, reservoirs, and estuaries resulting in eutrophication of water bodies. In the US, nearly 60% of 138 estuaries exhibit moderate to severe eutrophic conditions and >90% rivers have either N or P concentrations greater than the respective reference levels. In genera l anthropogenic influence s (urban and agricultural) in watersheds result in greater nutrient inputs such as fertilizer application to crops, urban lawns, and septic tanks which may lead to greater losses of nutrients to streams. Another consequence of change s in land use primarily i n urbanizing watersheds, is the increase in impervious areas such as rooftops, roadways, parking lots, sidewalks, and driveways. These impervious areas increase runoff and minimize the biotic uptake processes that immobilize nutrients in the forest canopy, litter, soils, and organic matter and thereby result in greater losses of nutrients. Florida is one of the rapidly developing states in the US and has serious water quality problems with nutrients especially eutrophication of coastal waters. Therefore, it is important to assess the impact of anthropogenic activities on water quality in waterbodies of Florida Very little is known about N and P transport in urban watersheds in Florida, whi ch is dominated by sandy soils, high ground water table, P rich geology, and altered hydrology due to construction of stormwater retention ponds to avoid flooding. In this study, we used monthly collected data (5 19 years for various sites) of stream water N and P forms for seven sub-basins (19 350 km2) of the Alafia Rive r Watershed (total drainage area: 1085 km2). In the Hillsborough County where th is w atershed is located population has increased from 0.83 million in 1990 to 0.99 million in 2000 and 1.20 million in 2009. This increase in population has resulted in signif icant changes in land use s in the

PAGE 101

101 w atershed. We used, Florida Land Use and Cover Classification Codes (FLUCCCS) at level IV and grouped land uses into six main categories: residential, built up, agricultural, pasture, forest, and mined using land use data of three time periods ( 1990, 1999, and 2007). During 1990 2007, urban land use (residential and built up) in the watershed increased by 8 % (from 12 to 20 %) while forest decrease d by 6% (from 25 to 19%) and pasture decreased by 8% (from 19 to 11%) The agricultural land use remained similar at 8% during 1990 2007. Based on the urban land use of 2007, we classified the sub-basins into: (1) three developed (18 24% residential; 1 14% built up; 4 24% agricultural; 0 39% mined) and (2) two undeveloped (3 11% residential; 1 3% built up; 4 14% agricultural; 0 66% mined). In addition, two mainstem stations draining 89 99% of the watershed had 16 17% residential, 3% built up, 8% agricultural, and 32 34% mined land uses in 2007. Long term monthly collected data sh owed that at mainstem stations (Alafia and Bell Shoals), total N concentrations of 1.77 1.91 mg L1 were similar to EPAs proposed numeric total N value of 1.79 mg L1 for the region (EPA, 2010). T otal N concentrations were significantly (P <0.05) correlated with percent urban land use (r=0.83) but not with agricultural, pasture, mined, and forest land uses (r<0.50) suggesting that urbanization has increased N concentrations in stream waters This is further reflected in significant ( P <0.05) total N in three developed (1.67 2.43 mg L1) than two undeveloped (0.84 1.21 mg L1) sub -basins during 2000 2009. Greater proportion of NO3N was observed in developed (53 68% of total N) than undeveloped sub -basins (25 30% of total N ). C oncentration of NO3N w as lower in wet than dry season due to greater biotic uptake and greater denitrification of NO3N due to higher temperature in wet season In contrast, ON concentrations were greater in wet than dry season probably due to the greater transport of organic materials (leaves, grass) with more runoff. During 1991 2009,

PAGE 102

102 concentrations showed increasing trends at the mainstem station thereby indicating that urbanization has increased total N in the streams Interestingly, the increas ed total N concentration trends were pr imarily due to increases in ON rather than NO3-N. This suggest processing of NO3N in our watershed and likely conversion to ON in stormwater retention ponds In addition, greater runoff generation in urban land uses may enhance the transport of organic ma terials such as composts grass cuttings from urban lawns, deciduous leaves fallen on the ground. As these are rich sources of ON, decomposition or leaching of N may have resulted in increased ON concentrations in the watershed. The increased ON concentrat ion trends in this urbanizing watershed has raised two important questions (1) Is the increase in ON c oncentration due to greater ON sources in urban land uses? and (2) Is the ON increase a product of microbial transformations of NO3N into ON ? Further s tudies on the source characterization of the ON can help in devising the accurate BMPs to control N pollution in the Alafia River Watershed. T otal P concentrations ranged from 1.01 to 1.23 mg P L1 and were much greater than EPAs proposed numeric total P value of 0.739 mg P L1 for the region. Of total P, dissolved reactive P (DRP) was dominant (70 90% of total P) than other P (10 30% of total P) in both developed and undeveloped subbasins. None of the P forms were significantly ( P < 0.05) correlat ed with urban, agricultural, forest, and pasture land use (r<0.50) indicating that the P concentrations a re not controlled by these land uses in the Alafia River Watershed. Two developed sub basins had significantly greater total P concentrations (North Pr ong and English Creek: 2.18 2.53 mg P L1) probably due to P rich geology, active mined lands, and discharges of P rich wastewaters. In all the developed and undeveloped sub-basins, the concentrations of P forms were greater in wet than dry season. This in dicate s perhaps the flus hing of P from dissolution and desorption of P from soil minerals and suspension of the stream sediments with

PAGE 103

103 greater runoff in wet season might have elevated P concentrations in stream waters The decreasing flow weighted and flow un-weighted total P concentration trends indicated that the anthropogenic activities such as increased regulations on P discharges from mined lands and, wastewater discharges together with reclamation of mined lands were probably successful in controlling P pollution in the Alafia River Watershed. If the EPA proposed numeric total N and P criteria for Florida streams is established it will be increasingly difficult to maintain concentrations of total N below 1.798 mg L1 and total P below 0.739 mg L1 in this urban watershed, unless the mechanisms controlling N and P transport from the landscape are clearly understood and BMPs to control nutrient losses from watershed are developed and implemented In the short term, our results can aid in planning efforts to reduce N and P concentrations. We suggest that BMPs should be target ed to control N loss in three developed sub-basins (English Creek, Turkey Creek, and North Prong ) that had total N concentrations of 1.7 2.4 mg L1as these may yield greater reductions in N concentrations at watershed scale. On the other hand, due to P rich geology and discharges from wastewaters, all developed and one undeveloped sub-basins had greater total P concentrations (0.8 2.5 mg P L1) than EPA proposed numeric value of 0.739 mg P L1. Therefore, the reduction in P loss from mined lan ds and wastewater discharges in four sub -basins (North Prong, English Creek, Turkey Creek, and South Prong) may result in reducing total P concentrations in stream waters

PAGE 104

104 LIST OF REFERENCES Abelho, M. 2001. From litterfall to breakdown in streams: a review. Scientific World Jo urnal 1:656 80. Alexander, R.B., and R.A. Smith. 2006. Trends in the nutrient enrichment of US rivers during the late 20th century and their re lation to changes in probable stream trophic conditions. Limnol. Oceanogr. 51:639 654. Alexander, R.B., R.A. Smith, and G.E. Schwarz. 2000. Effect of stream channel size on the delivery of nitrogen to the Gulf of Mexico. Nature 403:758 761. Alexander, R. B ., J. K. Bhlke, E. W., Boyer, M. B. David, J. W. Harvey, P. J. Mulholland, S. P. Seitzinger, C. R. Tobias, C. Tonitto, and W. M. Wollheim. 2009. Dynamic modeling of nitrogen losses in river networks unravels the coupled effects of hydrological and biogeoc hemical processes. Biogeochemistry. doi: 10.1007/s105330089274 8 Anisfeld, S.C., R.T. Barnes, M.A. Altabet, and T.X. Wu. 2007. Isotopic apportionment of atmospheric and sewage nitrogen sources in two Connecticut rivers. Environ. Sci. Technol. 41:6363 6369. Arnold, C.L., and C.J. Gibbons. 1996. Impervious surface coverage The emergence of a key environmental indicator. J. American Plan. Assoc. 62:243 258. Boesch, D.F., R.B. Brinsfield, and R.E. M agnien. 2001. Chesapeake Bay eutrophication: Scientific understanding, ecosystem restoration, and challenges for agriculture. J. Environ. Qual. 30:303 320. Bosch, N.S. 2008. The influence of impoundments on riverine nutrient transport: An evaluation using the Soil and Water Assessment Tool. J. Hydrol. 355:131 147. Boyer, E.W., G.M. Hornberger, K.E. Bencala, and D.M. McKnight. 1997. Response characteristics of DOC flushing in an alpine catchment. Hydrol. Processes 11:1635 1647. Boyer, E.W., C.L. Goodale, N.A. Jaworsk, and R.W. Howarth. 2002. Anthropogenic nitrogen sources and relationships to riverine nitrogen export in the northeastern USA. Biogeochemistry 57:137 169. Brett, M.T., G.B. Arhonditsis, S.E. Mueller, D.M. Hartley, J.D. Frodge, and D.E. Funke. 2005 Non -point -source impacts on stream nutrient concentrations along a forest to urban gradient. Environ. Manage. 35:330 342. Bricker, S.B., C.G. Clement, D.E. Pirhalla, S.P. Orlando, and D.R.G. Farrow. 1999. National Estuarine Eutrophication Assessment. Effect of nutrient enrichment in the nation's estuaries, NOAA NOS Special Projects Office pp 71.

PAGE 105

105 Carpenter, S.R., N.F. Caraco, D.L. Correll, R.W. Howarth, A.N. Sharpley, and V.H. Smith. 1998. Nonpoint pollution of surface waters with phosphorus and nitroge n. Ecol. Appl. 8:559 568. Chen, X.X., and C.T. Driscoll. 2009. Watershed Land Use Controls on Chemical Inputs to Lake Ontario Embayments. J. Environ. Qual. 38:2084 2095. Conley, D.J., H.W. Paerl, R.W. Howarth, D.F. Boesch, S.P. Seitzinger, K.E. Havens, C. Lancelot, and G.E. Likens. 2009. Ecology Controlling Eutrophication: Nitrogen and Phosphorus. Science 323:1014 1015. Culbert, H., and R. France. 1995. Laboratory prediction of phosphorus release from deciduous leaves to urban runoff. Water Qual. Res. J. Can. 30:243 246. Dame R., D. Bushek, D. Allen, A. Lewitus, D. Edwards, E. Koepfler, and L., Gregory. 2002. Ecosystem response to bivalve density reduction: management implications. Aquatic Ecology 36: 51 65. Daroub, S.H., T.A. Lang, O.A. Diaz, and S. Grunwald. 2009. Longterm water quality trends after implementing best management practices in south Florida. J. Environ. Qual. 38:1683 1693. Diebel, M.W., J.T. Maxted, D .M. Robertson, S.B. Han, and M.J.v. Zanden. 2009. Landscape planning for agricultural nonpoint source pollution reduction. III: Assessing phosphorus and sediment reduction potential. Environ. Manage. 43:69 83. Dixon, L.K., G.A. Vargo, J.O.R. Johansson, R. Montgomery, and M.B. Neely. 2009. Trends and explanatory variables for the major phytoplankton groups of two southwestern Florida estuaries, USA. J. Sea Res. 61:95 102. Dodds, W.K., and R.M. Oakes. 2006. Controls on nutrients across a prairie stream waters hed: Land use and riparian cover effects. Environ. Manage. 37:634 646. Dodds, W.K., W.W. Bouska, J.L. Eitzmann, T.J. Pilger, K.L. Pitts, A.J. Riley, J.T. Schloesser, and D.J. Thornbrugh. 2009. Eutrophication of U.S. freshwaters: analysis of potential economic damages. Environ. Sci. Technol. 43:12 19. Edwards, A.C., and P.J.A. Withers. 2008. Transport and delivery of suspended solids, nitrogen and phosphorus from various sources to freshwaters in the UK. J. Hydrol. 350:144 153. Edwards, A.C., H. Twist, and G .A. Codd. 2000. Assessing the impact of terrestrially derived phosphorus on flowing water systems. J. Environ. Qual. 29:117 124. EPA, 2010. Water Quality Standards for the State of Floridas Lakes and Flowing Waters. Federal Register. 75:16, January 26, 2010. Ellison, M.E., and M.T. Brett. 2006. Particulate phosphorus bioavailability as a function of stream flow and land cover. Water Resour. Res. 40:1258 1268.

PAGE 106

106 Evans, D.J., P.J. Johnes, and D.S. Lawrence. 2004. Physico-chemical controls on phosphorus cycling in two lowland streams. Part 2 The sediment phase. Sci. Total Environ. 329:165 182. Florida Climate Center, 2009. http://www.coaps.fsu.edu/climate_center/index.shtml Accessed on 06, Dece mber, 2009. Florida Department of Health, 2009. http://www.doh.state.fl.us/environment/programs/EhGis/EhGisDownload.htm Accessed on 07, December, 2009. Florida Lakewatch Program. 2005. Florida Lakewatch Data Book (Hillsborough County). http://lakewatch.ifas.ufl.edu/2005DataReport/Hillsborough_Marion.pdf Accessed on 1 November, 2010). Galloway, J.N., R.W. Howarth, A.F. Michaels, S.W. Nixon, J.M. Prospero, and F.J. Dentener. 1996. Nitrogen and phosphorus budgets of the North Atlantic Ocean and its watershed. Biogeochemistry 35:3 25. Gilbert, R.O. 1987. Statistical methods for environmental m onitoring. John Wiley & Sons, New York. Goodrich, B.A., R.D. Koski, and W.R. Jacobi. 2009. Monitoring surface water chemistry near magnesium chloride dust suppressant treated roads in Colorado. J. Environ. Qual. 38:2373 2381. Goolsby, D.A., and W.A. Battag lin. 2001. Longterm changes in concentrations and flux of nitrogen in the Mississippi River Basin, USA. Hydrol. Processes 15:1209 1226. Granlund, K., A. Raike, P. Ekholm, K. Rankinen, and S. Rekolainen. 2005. Assessment of water protection targets for agr icultural nutrient loading in Finland. J. Hydrol. 304:251 260. Green, P.A., C.J. Vorosmarty, M. Meybeck, J.N. Galloway, B.J. Peterson, and E.W. Boyer. 2004. Pre industrial and contemporary fluxes of nitrogen through rivers: a global assessment based on typology. Biogeochemistry 68:71 105. Groffman, P.M., N.L. Law, K.T. Belt, L.E. Band, and G.T. Fisher. 2004. Nitrogen fluxes and retention in urban watershed ecosystems. Ecosystems 7:393 403. Han, H.J., J.D. Allan, and D. Scavia. 2009. Influence of Climate and Human Activities on the Relationship between Watershed Nitrogen Input and River Export. Environ. Sci. Technol. 43:1916 1922. Harrison, J.A., N. Caraco, and S.P. Seitzinger. 2005. Global patterns and sources of dissolved organic matter export to the coa stal zone: Results from a spatially explicit, global model. Global Biogeochem. Cycles 19.

PAGE 107

107 Hayakawa, A., M. Shimizu, K.P. Woli, K. Kuramochi, and R. Hatano. 2006. Evaluating stream water quality through land use analysis in two grassland catchments: Impact of wetlands on stream nitrogen concentration. J. Environ. Qual. 35:617 627. Haygarth, P.M. 2005. Linking landscape sources of phosphorus and sediment to ecological impacts in surface waters. Sci. Total Environ. 344:1 3. Haygarth,P.M. and A. N, Sharpley, 20 00. Terminology for phosphorus transfer. J. Environ. Qual. 29: 10 15. Helsel, D.R., and R.M. Hirsch. 1992. Statistical methods in water resources. Elsevier Publ., Amsterdam, the Netherlands. Hirsch, R.M., J.R. Slack, and R.A. Smith. 1982. Techniques of tr end analysis for monthly water quality data. Water Resour. Res. 18:107 121. Howarth, R.W., and R. Marino. 2006. Nitrogen as the limiting nutrient for eutrophication in coastal marine ecosystems: Evolving views over three decades. Limnol. Oceanogr. 51:364 376. Howarth, R.W., A. Sharpley, and D. Walker. 2002. Sources of nutrient pollution to coastal waters in the United States: Implications for achieving coastal water quality goals. Estuaries 25:656 676. Howarth, R.W., G. Billen, D. Swaney, A. Townsen d, N. Jaworski, K. Lajtha, J.A. Downing, R. Elmgren, N. Caraco, T. Jordan, F. Berendse, J. Freney, V. Kudeyarov, P. Murdoch, and Z.L. Zhu. 1996. Regional nitrogen budgets and riverine N&P fluxes for the drainages to the North Atlantic Ocean: Natural and hu man influences. Biogeochemistry 35:75 139. Jansson, B.O., and K. Dahlberg. 1999. The environmental status of the Baltic Sea in the 1940s, today, and in the future. Ambio 28:312 319. Jarvie, H.P., C. Neal, and P.J.A. Withers. 2006. Sewage -effluent phosphoru s: A greater risk to river eutrophication than agricultural phosphorus? Sci. Total Environ. 360:246 253. Johnson, H.O., S.C. Gupta, A.V. Vecchia, and F. Zvomuya. 2009. Assessment of water quality trends in the Minnesota River using non -parametric and param etric methods. J. Environ. Qual. 38:1018 1030. Johnson, L.B., C. Richards, G.E. Host, and J.W. Arthur. 1997. Landscape influences on water chemistry in Midwestern stream ecosys tems. Freshwater Biology 37:193 198 Kaushal, S.S., and W.M. Lewis. 2003. Pattern s in the chemical fractionation of organic nitrogen in Rocky Mountain streams. Ecosystems 6:483 492. Kaushal, S.S., P.M. Groffman, L.E. Band, C.A. Shields, R.P. Morgan, M.A. Palmer, K.T. Belt, C.M. Swan, S.E.G. Findlay, and G.T. Fisher. 2008. Interaction b etween urbanization and

PAGE 108

108 climate variability amplifies watershed nitrate export in Maryland. Environ. Sci. Technol. 42:5872 5878. Koerselman, W., and A.F.M. Meuleman. 1996. The vegetation N:P ratio: A new tool to detect the nature of nutrient limitation. J. Applied Ecology 33:1441 1450. Kroeger, K.D., M.L. Cole, J.K. York, and I. Valiela. 2006. Nitrogen loads to estuaries from waste water plumes: Modeling and isotopic approaches. Ground Water 44:188 200. Lane, E. (Ed.), 1994. Florida's Geological History and Geological Resources. Special Publication No. 35. Florida Geological Survey, Tallahassee, FL. Mainstone, C.P., R.M. Dils, and P.J.A. Withers. 2008. Controlling sediment and phosphorus transfer to receiving waters A strategic management perspective for England and Wales. J. Hydrol. 350:131 143. Maxted, J., M. Diebel, and M. Vander Zanden. 2009. Landscape Planning for Agricultural NonPoint Source Pollution Reduction. II. Balancing Watershed Size, Number of Watersheds, and Implementation Effort. Environ. Manage. 43:60 68. McDowell, W.H., A.H. Magill, J.A. Aitkenhead -Peterson, J.D. Aber, J.L. Merriam, and S.S. Kaushal. 2004. Effects of chronic nitrogen amendment on dissolved organic matter and inorganic nitrogen in soil solution. Forest Ecology and Management 196:29 41. Meals, D.W. 1996. Watershed-scale response to agricultural diffuse pollution contro l programs in Vermont, USA. Water Sci. Technol. 33:197 204. Mulliss, R.M., D.M. Revitt, and R.B. Shutes. 1996. The impacts of urban discharges on the hydrology and water quality of an urban watercourse. Sci. Total Environ. 190:385 390. National Pollution D ischarge Elimination, Systems, 2009. http://cfpub.epa.gov/npdes/ Accessed on 09, December, 2009. Omernik, J.M. 1977. Nonpoint source -stream nutrient level relationships: a nationwide study. Nonpoint source -stream nutrient level relationships: a nationwide study. 12: 151. Pacini, N., and R. Gachter. 1999. Speciation of riverine particulate phosphorus during rain events. Biogeochemistry 47:87 109. Paerl, H.W. 2009. Controlling eutrophication along the freshwater -mari ne continuum: dual nutrient (N and P) reductions are essential. Estuaries and Coasts 32:593 601. Paul, M.J., and J.L. Meyer. 2001. Streams in the urban landscape. Annual Review of Ecology and Systematics 32:333 365. Peebles, E., D. Hollander, S. Locker, P. Swarzenski and G. Brooks. 2009. Areal Extent, Source and Ecological Status of Organic Sediment Accumulation in Safety Harbor, Tampa Bay. Technical Report #0709.

PAGE 109

109 Pellerin, B.A., S.S. Kaushal, and W.H. McDowell. 2006. Does anthropogenic nitrogen enrichment increase organic nitrogen concentrations in runoff from forested and human dominated watersheds? Ecosystems 9:852 864. Pellerin, B.A., W.M. Wollheim, C.S. Hopkinson, W.H. McDowell, M.R. Williams, C.J. Vorosmarty, and M.L. Daley. 2004. Role of wetlands and developed land use on dissolved organic nitrogen concentrations and DON/TDN in northeastern US rivers and streams. Limnol. Oceanogr. 49:910 918. Peterson, B.J., W.M. Wollheim, P.J. Mulholland, J.R. Webster, J.L. Meyer, J.L. Tank, E. Marti, W.B. Bowden, H .M. Valett, A.E. Hershey, W.H. McDowell, W.K. Dodds, S.K. Hamilton, S. Gregory, and D.D. Morrall. 2001. Control of nitrogen export from watersheds by headwater streams. Science 292:86 90. Poe, A, K. Hackett, S. Janicki, R. Pribble, and S. Janicki. 2005. Es timates of total nitrogen, total phosphorous, and total suspended solids to Tampa Bay, Florida. Technical Report #0205. Qian, Y., K.W. Migliaccio, Y. Wan, Y.C. Li, and D. Chin. 2007. Seasonality of selected surface water constituents in the Indian river lagoon, Florida. J. Environ. Qual. 36:416 425. Redfield A.C., 1934. On the proportions of organic derivations in sea water and their relation to the composition of plankton. James Johnstone Memorial Volume (ed. R.J. Daniel). University Press of Liverpool 177 192. Rhodes, A.L., R.M. Newton, and A. Pufall. 2001. Influences of land use on water quality of a diverse new England watershed. Environ. Sci. Technol. 35:3640 3645. Richards, R.P., and J. Holloway. 1987. Monte -carlo studies of sampling strategies fo r estimating tributary loads. Water Resour. Res. 23:1939 1948. Richards, R. P., D. B. Baker, J. P. Crumrine, J. W. Kramer, D. E. Ewing, and B. J. Merryfield. 2008. Thirtyyear trends in suspended sediment in seven Lake Erie tributaries. J. Environ. Qual. 37:1894 1908. Richards, R.P., D.B. Baker, J.P. Crumrine, J.W. Kramer, D.E. Ewing, and B.J. Merryfield. 2008. Thirty -year trends in suspended sediment in seven Lake Erie tributaries. J. Environ. Qual. 37:1894 1908. Royer, T.V., M.B. David, and L.E. Gentry. 2006. Timing of riverine export of nitrate and phosphorus from agricultural watersheds in Illinois: Implications for reducing nutrient loading to the Mississippi River. Environ. Sci. Technol. 40:4126 4131. Russell, M.J., D.E. Weller, T.E. Jordan, K.J. Sigwa rt, and K.J. Sullivan. 2008. Net anthropogenic phosphorus inputs: spatial and temporal variability in the Chesapeake Bay region. Biogeochemistry 88:285 304.

PAGE 110

110 SAS Institute. 2008. SAS/STAT 9.1. The Glimmix Procedure. User's guide. SAS Inst., Cary, NC. Schertz, T.L., R.B. Alexander, and D.J. Ohe. 1991. The computer program Estimate Trend (ESTREND), a system for the detection of trends in water quality data. U.S. Geological Survey Water Resources Investigations Rep. 91 4040. U.S. Geological Survey, Resto n, VA. Scott, D., J. Harvey, R. Alexander, and G. Schwarz. 2007. Dominance of organic nitrogen from headwater streams to large rivers across the conterminous United States. Global Biogeochem. Cycles 21:1003. Seitzinger, S.P., R.V. Styles, E.W. Boyer, R.B Alexander, G. Billen, R. Howarth, B. Mayer, and N. van Breemen. 2002. Nitrogen retention in rivers: Model development and application to watersheds in the northeastern USA. Biogeochemistry 57:199 237. Sen, P.K. 1968. Estimates of the regression coefficie nt based on Kendalls tau. J. Am. Stat. Assoc. 63:1379 1389. Setaro, F.V., and J.M. Melack. 1984. Responses of phytoplankton to experimental nutrient enrichment in an amazon floodplain lake. Limnol. Oceanogr. 29:972 984. Sharpley, A.N., and R.G. Menzel. 19 87. The impact of soil and fertilizer phosphorus on the environment. Adv. Agron. 41:297 324. Southwest Florida Water Management District (SWFWMD). 2007. The Determination of Minimum Flows for the Lower Alafia River Estuary. Southwest Florida Water Manageme nt District. Brooksville, Florida. Sprague, L.A., and D.L. Lorenz. 2009. Regional Nutrient Trends in Streams and Rivers of the United States, 1993 2003. Environ. Sci. Technol. 43:3430 3435. Stanley, E.H., and J.T. Maxted. 2008. Changes in the dissolved nitr ogen pool across land cover gradients in Wisconsin streams. Ecol. Appl.18:1579 1590. Stone, M., and I.G. Droppo. 1994. In-channel surficial fine -grained sediment laminae .2. chemical characteristics and implications for contaminant transport in fluvial sys tems. Hydrol. Process. 8:113124. Svendsen, L.M., B. Kronvang, P. Kristensen, and P. Graesbol. 1995. Dynamics of phosphorus compounds in a lowland river system importance of retention and nonpoint sources. Hydrol. Process. 9:119 142. Tampa Bay Estuary P rogram, 2009. http://www.tbep.org/ Accessed on 10 December, 2009. Toor, G. S., S. Hunger, D. J. Peak, T. J. Sims, and D. L. Sparks, 2004. Advances in the characterization of phosphorus in organic wastes: environmental and agronomic applications. Adv. Agron. 89: 1 65.

PAGE 111

111 Tufford, D.L., H.N. McKellar, and J.R. Hussey. 1998. In -stream nonpoint source nutrient prediction with land use proximity and seasonality. J. Environ. Qual. 27:100 111. Turner, R.E., and N.N. Rabalais. 2003 Linking landscape and water quality in the Mississippi river basin for 200 years. Bioscience 53:563 572. United States Census Bureau, 2009. http://quickfacts.ce nsus.gov/qfd/states/12/12057.html. Accessed on 09 December, 2009. United States Department of Agriculture, 2006. Conservation reserve program: summary and enrollment statistics 2005. Farm Service Agency, Washington, DC. United States Department of Agriculture, 2010. National resource conservation service. http://soils.usda.gov/survey/printed_surveys/state.asp?state=Florida&abbr=FL Accessed on 06, June, 2010. United States Environmental Protection Agency (2002). A review of statewide watershed management approaches. Office of Water, Washington, DC. United States Environmental Protection Agency (1986). Quality criteria for water 1986 and update no. 1 and update no. 2 (May 1, 1987): Washington, D.C., EPA 440/586 001. Vaithiyanathan, P., and D.L. Correll. 1992. The Rhode river watershed phosphorus distribution and export in forest and agricultural soils. J. Environ. Qual. 21:280 288 van Kessel, C., T. Clough, and J.W. van Groenigen. 2009. Dissolved Organic Nitrogen: An Overlooked Pathway of Nitrogen Loss from Agricultural Systems? J. Environ. Qual. 38:393 401. Verseveld, W.J.v., J.J. McDonnell, and K. Lajtha. 2009. The role of hills lope hydrology in controlling nutrient loss. J. Hydrol. (Amsterdam) 367:177 187. Vidon, P., L.P. Tedesco, D.L. Pascual, M.A. Campbell, L.R. Casey, J. Wilson, and M. Gray. 2008. Seasonal changes in stream water quality along an agricultural/urban landuse g radient. Proceedings of the Indiana Academy of Science 117:107 123. Wahl, M.H., H.N. McKellar, and T.M. Williams. 1997. Patterns of nutrient loading in forested and urbanized coastal streams. Journal of Experimental Marine Biol. Ecol. 213:111 131. Wallbrin k, P.J., C.E. Martin, and C.J. Wilson. 2003. Quantifying the contributions of sediment, sediment -P and fertiliser P from forested, cultivated and pasture areas at the landuse and catchment scale using fallout radionuclides and geochemistry. Soil Tillage Re s.69:53 68. Wear D. and J. Greis, 2002. Southern forest resource assessment: summary of findings. J. Forestry 100: 6 14.

PAGE 112

112 Weston, N.B., J.T. Hollibaugh, and S.B. Joye. 2009. Populatio n growth away from the coastal zone: Thirty years of land use change and nutrient export in the Altamaha River, GA. Sci. Total Environ. 407:3347 3356. Wetzel, R.G. 2001. Fundamental processes within natural and constructed wetland ecosystems: short term versus longterm objectives. Water Sci. Technol. 44:1 8. Zaimes, G.N., R.C. Schultz, and T.M. Isenhart. 2008. Streambank soil and phosphorus losses under different riparian land uses in Iowa. J. Am. Water Resour. Assoc. 44:935 947.

PAGE 113

113 BIOGRAPHICAL SKETCH Kam aljit was born in Northwest India (Punjab). The eldest of the three children, he spent earlier period of his life in Punjab. After his bachelors at Punjab Agricultural University, I ndia in 2005, he completed his m aster s in soil science and agricultural c hemistry at University of Agricultural Sciences, Bangalore, India. In A ugust 2008, he started another m asters program in soil and water sciences under the supervision of Dr. Gurpal Toor at the Gulf Coast Research and Education Center -Wimauma, University o f Florida. Kamaljit received his m asters degree from the University of Florida in the summer 2010.