<%BANNER%>

Development of Bankfull Discharge and Channel Geometry Regressions for Peninsular Florida Streams

Permanent Link: http://ufdc.ufl.edu/UFE0024089/00001

Material Information

Title: Development of Bankfull Discharge and Channel Geometry Regressions for Peninsular Florida Streams
Physical Description: 1 online resource (326 p.)
Language: english
Creator: Blanton, Kristen
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: bankfull, channel, curves, discharge, geometry, hydraulic, regional, restoration, stream
Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre: Environmental Engineering Sciences thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Regional curves, which relate bankfull discharge and channel geometry (cross-sectional area, width, and mean depth) to drainage area in regions of similar climate, geology, and vegetation, have greatly aided in creating target natural channel designs for stream restoration efforts. Regional curves were developed for peninsular Florida based on cross-sectional and longitudinal survey data collected at 17 gaged and 28 ungaged as near-to-natural streams, ranging in drainage area from 0.2 to 311 square miles and valley slope from 0.02 to 2.27%. Based on an analysis of prevalence among sites, slopes, and hydrologic data, the elevation of the flat floodplain was determined to be the most reliable bankfull indicator at sites with a wetland floodplain, while the elevation of the inflection on the bank was the most reliable indicator at sites with an upland floodplain. Analysis of bankfull indicator slopes further revealed that a water slope threshold of approximately 0.5% exists, above which bankfull indicators appear to more reliable, suggesting that slope-area techniques for calculating the bankfull discharge may be unreliable in peninsular Florida streams with a water slope less than 0.5%. The dataset was further divided based on physiography (flatwoods versus highlands), geography (northern versus southern peninsula), and floodplain types (wetland versus upland and cypress-dominated versus non-cypress-dominated) to determine if significant differences exist in the bankfull regressions and/or various dimensionless ratios (sinuosity, width-to-depth, maximum depth-to-mean depth, valley slope, and maximum discharge-to-mean annual discharge) among various peninsular Florida stream subsets. Streams with wetland floodplains were found to have a significantly greater bankfull area and bankfull width than streams with an upland floodplain. Also, streams with cypress-dominated floodplains had a greater width-to-depth ratio than streams with non-cypress-dominated floodplains. Further, streams draining flatwoods physiographies were found to be flashier. These differences may be important considerations when designing natural channels in peninsular Florida. Annual peak flow data for the gaged sites were analyzed to estimate the bankfull discharge return interval using Log Pearson Type III distributions. The bankfull discharge ranged from less than one year to 1.44 years, which is more frequent than the average 1.5-year return interval often cited in the literature. Based on analysis of the flow duration at gaged sites, bankfull discharge for peninsular Florida streams is equaled or exceeded approximately 21% of the time on average, or about 77 days per year. On average, the bankfull discharge is roughly four times that of the mean annual discharge and is 35% of the 1.5-year discharge. Lastly, the regional curves developed for peninsular Florida were compared to regional curves previously developed for other regions of the southeastern United States Coastal Plain. Peninsular Florida bankfull channels were found to have a lower bankfull discharge and smaller bankfull channel (narrower and shallower) than northwest Florida streams, which receives more mean annual precipitation and runoff. These differences indicate that the regional curves developed in the present work are more applicable to peninsular Florida streams than are regional curves developed for other regions of the southeastern Coastal Plain.
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 Kristen Blanton.
Thesis: Thesis (M.S.)--University of Florida, 2008.
Local: Adviser: Wise, William R.
Local: Co-adviser: Mossa, Joann.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0024089:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024089/00001

Material Information

Title: Development of Bankfull Discharge and Channel Geometry Regressions for Peninsular Florida Streams
Physical Description: 1 online resource (326 p.)
Language: english
Creator: Blanton, Kristen
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: bankfull, channel, curves, discharge, geometry, hydraulic, regional, restoration, stream
Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre: Environmental Engineering Sciences thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Regional curves, which relate bankfull discharge and channel geometry (cross-sectional area, width, and mean depth) to drainage area in regions of similar climate, geology, and vegetation, have greatly aided in creating target natural channel designs for stream restoration efforts. Regional curves were developed for peninsular Florida based on cross-sectional and longitudinal survey data collected at 17 gaged and 28 ungaged as near-to-natural streams, ranging in drainage area from 0.2 to 311 square miles and valley slope from 0.02 to 2.27%. Based on an analysis of prevalence among sites, slopes, and hydrologic data, the elevation of the flat floodplain was determined to be the most reliable bankfull indicator at sites with a wetland floodplain, while the elevation of the inflection on the bank was the most reliable indicator at sites with an upland floodplain. Analysis of bankfull indicator slopes further revealed that a water slope threshold of approximately 0.5% exists, above which bankfull indicators appear to more reliable, suggesting that slope-area techniques for calculating the bankfull discharge may be unreliable in peninsular Florida streams with a water slope less than 0.5%. The dataset was further divided based on physiography (flatwoods versus highlands), geography (northern versus southern peninsula), and floodplain types (wetland versus upland and cypress-dominated versus non-cypress-dominated) to determine if significant differences exist in the bankfull regressions and/or various dimensionless ratios (sinuosity, width-to-depth, maximum depth-to-mean depth, valley slope, and maximum discharge-to-mean annual discharge) among various peninsular Florida stream subsets. Streams with wetland floodplains were found to have a significantly greater bankfull area and bankfull width than streams with an upland floodplain. Also, streams with cypress-dominated floodplains had a greater width-to-depth ratio than streams with non-cypress-dominated floodplains. Further, streams draining flatwoods physiographies were found to be flashier. These differences may be important considerations when designing natural channels in peninsular Florida. Annual peak flow data for the gaged sites were analyzed to estimate the bankfull discharge return interval using Log Pearson Type III distributions. The bankfull discharge ranged from less than one year to 1.44 years, which is more frequent than the average 1.5-year return interval often cited in the literature. Based on analysis of the flow duration at gaged sites, bankfull discharge for peninsular Florida streams is equaled or exceeded approximately 21% of the time on average, or about 77 days per year. On average, the bankfull discharge is roughly four times that of the mean annual discharge and is 35% of the 1.5-year discharge. Lastly, the regional curves developed for peninsular Florida were compared to regional curves previously developed for other regions of the southeastern United States Coastal Plain. Peninsular Florida bankfull channels were found to have a lower bankfull discharge and smaller bankfull channel (narrower and shallower) than northwest Florida streams, which receives more mean annual precipitation and runoff. These differences indicate that the regional curves developed in the present work are more applicable to peninsular Florida streams than are regional curves developed for other regions of the southeastern Coastal Plain.
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 Kristen Blanton.
Thesis: Thesis (M.S.)--University of Florida, 2008.
Local: Adviser: Wise, William R.
Local: Co-adviser: Mossa, Joann.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-12-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0024089:00001


This item has the following downloads:


Full Text

PAGE 1

1 DEVELOPMENT OF BANKFULL DISCHARGE AND CHANNEL GEOMETRY REGRESSIONS FOR PENINSULAR FLORIDA STREAMS By KRISTEN M. BLANTON A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Kristen Blanton

PAGE 3

3 To my Mom, Dad, and Eric

PAGE 4

4 ACKNOWLEDGMENTS I thank Jacque Levine, Jessie Taft, and Kory Baxley (a.k.a. "intern") for their countless hours in the field, im pressive fou r-wheel driving and navigation sk ills, and watchful eyes which kept us safe from snakes, gators, and UXO. I thank Cory Catts for the above mentioned reasons, as well as for his endless moral and technical (esp ecially at Photoshop) support. I also thank my advisor, Dr. Wise, for giving me a great deal of independence, but for also always being there to rescue me in times of utter confusion. Thanks go to my committee, Joann Mossa and Tom Crisman who helped me to understand fluvi al geomorphical and ecological concepts, respectively. A very special thanks goes to John Kiefer, the fearless "stream team" leader and my mentor throughout this entire process, and also to his family, Sarah and Nolan. Thanks go to BCI Engineers & Scientists, Inc., and particularly to the GIS, Ad ministrative, and IT personnel. I thank the Florida Instit ute of Phosphate Research (FIPR) fo r funding this important project and the many landowners who allowed us access to their beautiful stream s. Last but not least, I thank my family and friends, whose endless support got me through this challenging, but rewarding endeavor.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4LIST OF TABLES ...........................................................................................................................8LIST OF FIGURES .......................................................................................................................10ABSTRACT ...................................................................................................................... .............15CHAPTER 1 INTRODUCTION .................................................................................................................. 172 LITERATURE REVIEW .......................................................................................................22Introduction .................................................................................................................. ...........22Florida Background ................................................................................................................22Physiography/Geological Context ...................................................................................22Weather and Climate ....................................................................................................... 23Water Resources ..............................................................................................................25Regional Curves ......................................................................................................................28History of Regional Curve Development ........................................................................29Channel-forming Discharge ............................................................................................ 30Indicators of Bankfull Stage ............................................................................................ 32Conclusions .............................................................................................................................343 DETERMINING THE MOST RELIAB LE BANKFULL INDICATOR FOR PENINSULAR FLORI DA STREAMS .................................................................................. 45Introduction .................................................................................................................. ...........45Methods ..................................................................................................................................47Site Selection ...................................................................................................................47Reference Reach Surveys ................................................................................................ 49Data Analysis ...................................................................................................................53Slopes of field indicators ..........................................................................................53Gage analysis ............................................................................................................54Results .....................................................................................................................................56Site Selection ...................................................................................................................57Reference Reach Surveys ................................................................................................ 57Data Analysis ...................................................................................................................58Slopes of field indicators ..........................................................................................59Gage analysis ............................................................................................................61Discussion .................................................................................................................... ...........65Reference Reach Surveys ................................................................................................ 66Data Analysis ...................................................................................................................68

PAGE 6

6 Slopes of field indicators ..........................................................................................68Gage analysis ............................................................................................................72Conclusions .............................................................................................................................754 DEVELOPING REGIONAL CURVES FOR PENINSULAR FLORIDA ............................90Introduction .................................................................................................................. ...........90Methods ..................................................................................................................................91Drainage Area Delineation and Valley Slope Determination ......................................... 92Regional Curve Development .........................................................................................92Dimensionless Ratios ......................................................................................................93Return Interval .................................................................................................................94Comparison to Other Southeastern United States Coastal Plain Regional Curves ......... 94Results .....................................................................................................................................94Drainage Area Delineation and Valley Slope Determination ......................................... 95Regional Curve Development .........................................................................................95Discharge: bankfull, mean annual, 1.5-year .............................................................96Bankfull cross-se ctional area ................................................................................. 102Bankfull width ........................................................................................................105Bankfull depth ........................................................................................................ 107Dimensionless Ratios ....................................................................................................110Sinuosity ................................................................................................................. 110Width-to-depth ....................................................................................................... 110Maximum depth-to-mean depth ............................................................................. 111Valley slope ............................................................................................................111Return Intervals .............................................................................................................111Comparison to Other Southeastern United States Coastal Plain Regional Curves ....... 112Discussion .................................................................................................................... .........113Regional Curve Development .......................................................................................114Discharge: bankfull, mean annual, 1.5-year ...........................................................115Bankfull channel geometry .................................................................................... 116Dimensionless Ratios ....................................................................................................117Return Intervals .............................................................................................................118Comparison to Other Southeastern Un ited States Coastal Plain Studies ...................... 119Conclusions ...........................................................................................................................1205 SYNTHESIS ..................................................................................................................... ....151Objective One: Most Reliable Bankfull Indi cator for Peninsular Florida Streams ..............151Objective Two: Development of Regional Cu rves for Peninsular Florida Streams .............153Objective Three: Comparisons by Physiography, Geography, and Floodplain Types......... 154Objective Four: Estimation of the Ba nkfull Discharge Return Interval ...............................156Objective Five: Comparisons to Other Southeas tern United States Coastal Plain Studies .. 157Conclusions ...........................................................................................................................158

PAGE 7

7 APPENDIX A SAMPLE PERMISSION LETTER AND FORM ................................................................ 160B SITE FIGURES: PLAN FORM, LONGITUDI NAL PROFILE, CROSS-SECTIONS ....... 162C SITE PHOTOGRAPHS ........................................................................................................253D GAGED SITE FIGURES: HYDROGRAPH, STAGE-Q RATING CURVE, FLOW AND STAGE DURATI ON CURVES ................................................................................. 298E STAGE AGAINST WIDTH GRAPHS ................................................................................ 316F SUPPLEMENTAL STAGE DATA .....................................................................................322LIST OF REFERENCES ............................................................................................................. 323BIOGRAPHICAL SKETCH .......................................................................................................326

PAGE 8

8 LIST OF TABLES Table page 1-1 Summary of study sites ......................................................................................................20 3-1 Summary of gaged sites .....................................................................................................76 3-2 Prevalence of field bankfull indicators ..............................................................................77 3-3 Summary of slopes data .....................................................................................................78 3-4 Comparision of various water slope to ba nkfull indicator slope ra tios by water slope .....79 3-5 Gaged sites discharge summary: Reference reach survey results ...................................... 80 3-6 Gaged sites stage summary: Re ference reach survey results ............................................. 81 3-7 Gaged sites discharge summary: Annual m aximum series results ....................................82 3-8 Gaged sites stage summary: A nnual m aximum series results ........................................... 83 3-9 Comparison of various bankfull indicator discharge and stage durations by floodplain type ...................................................................................................................84 4-1 Discharge data used in penisular Florid a regional curve develo pm ent and analysis ....... 123 4-2 Reference reach survey data used in penisular Florida regional curve developm ent and analysis ......................................................................................................................124 4-3 Regression equations for various discharges against drainage ar ea by entire data set and by subsets representing physiography, geography, and floodplain types ................. 125 4-4 Regression equations for bankfull channel geom etry against drainage area by entire data set and by subsets representing physi ography, geography, and floodplain types .... 126 4-5 Comparison of bankfull discharge against drainage area regressions by physiography, geography, floodplain type s, and Coastal Plain regions .......................... 127 4-6 Comparison of various discharge dur ations and ratios by physiography, geography, floodplain types, and Coastal Plain regions ..................................................................... 128 4-7 Comparison of bankfull area against drainage area regressions by physiography, geography, floodplain types, a nd Coastal Plain regions ..................................................129 4-8 Comparision of bankfull width agains t drainage area regr essions by physiography, geography, floodplain types, a nd Coastal Plain regions ..................................................130

PAGE 9

9 4-9 Comparison of bankfull mean depth ag ainst drainage area regressions by physiography, geography, floodplain type s, and Coastal Plain regions .......................... 131 4-10 Summary of dimensionless ratios .................................................................................... 132 4-11 Comparison of various dimensionless ratios by physiography, geography, and floodplain types ................................................................................................................133 4-12 Regression equations for bankfull parame ters against drainage area and bankfull return intervals for studies conducted th roughout the southeastern United States Coastal Plain ....................................................................................................................134

PAGE 10

10 LIST OF FIGURES Figure page 1-1 North and northwest Florid a regional curve study sites .................................................... 21 1-2 Peninsular Florida regi onal curve study sites .................................................................... 21 2-1 Physiographic provinces of the United States ................................................................... 36 2-2 Geologic history of Florida ................................................................................................36 2-3 Pleistocene shorelines in Florida ....................................................................................... 37 2-4 Climate zones in Florida .................................................................................................. ..37 2-5 Florida precipitation map ................................................................................................. ..38 2-6 Floridas water cycle ..........................................................................................................38 2-7 Floridas surface water drainage ........................................................................................ 39 2-8 Floridas watersheds ...................................................................................................... ....40 2-9 Longitudinal, cross-sectional, and plan views of m ajor stream types ............................... 40 2-10 Cross-sectional configuration, composition, and delineativ e criteria of major stream types ......................................................................................................................... ..........41 2-11 Regional curves for four US regions ................................................................................. 41 2-12 Relation of width, depth, and velocity to discharge, Powder River at Arvada, Wyom ing........................................................................................................................ ....42 2-13 Effective discharge determination from sedim ent rating and flow duration curves .......... 42 2-14 Channel cross section iden tifying bankfull param eters ..................................................... 43 2-15 Amount of water in a ri ver channel and frequency with which such an amount occurs. ................................................................................................................................43 2-16 Determination of bankfull stage fr om a stage-discharge rating curve. .............................. 44 2-17 Determination of bankfull stage from a pl ot of width-to-depth ratio against m aximum depth ......................................................................................................................... ..........44 3-1 Various field indicato rs of bankfull stage .......................................................................... 85 3-2 Water slope to various bankfull indica tor slope ratios agai nst water slope.. .....................86

PAGE 11

11 3-3 Width against stage field measurements ............................................................................87 3-4 Example of variability in stage-Q rating curves ................................................................ 88 3-5 Boxplots of stage and discha rge data for gaged sites. .......................................................89 4-1 Drainage area against valley sl ope f or study sites by physiography. ..............................135 4-2 Discharge against drainage area regressions for gaged sites. ..........................................136 4-3 Discharge against drainage area re gressions for gaged sites by physiography (flatwoods versus highlands) ........................................................................................... 137 4-4 Discharge against drainage area regressions for gaged sites by geography (northern versus southern peninsula) ...............................................................................................138 4-5 Discharge against draina ge area regressions for ga ged sites by floodplain type (wetland versus upland) ...................................................................................................139 4-6 Discharge against draina ge area regressions for ga ged sites by floodplain type (cypress-dom inated versus non-cypress-dominated) ....................................................... 140 4-7 Channel geometry against drainage area regressions for all sites.. .................................. 141 4-8 Channel geometry against drainage area regressions for all sites by physiography (flatwoods versus highlands). ..........................................................................................142 4-9 Channel geometry against drainage area regressions for all sites by geography (northern versus southern peninsula). ..............................................................................143 4-10 Channel geometry against drainage area regressions for all sites by floodplain type (wetland versus upland) ...................................................................................................144 4-11 Channel geometry against drainage area regressions for all sites by floodplain type (cypress-dom inated versus non-cypress-dominated). ...................................................... 145 4-12 Boxplots of sinuosity by the entire data set and subsets re presenting physiography, geography, and floodplain types. ..................................................................................... 146 4-13 Boxplots of width-to-depth ratio by the entire data set and subsets representing physiography, geography, and floodplain types .............................................................. 146 4-14 Boxplots of maximum depth-to-mean depth ratio by the entire data set and subsets representing physiography, geogr aphy, and floodplain types. ........................................147 4-15 Boxplots of valley slope by the enti re data set and subsets representing physiography, geography, and floodplain types .............................................................. 147 4-16 Bankfull discharge against drainage area regressions by Coastal Plain study. ................148

PAGE 12

12 4-17 Bankfull area against drainage area regressions by Coas tal Plain study. ........................148 4-18 Bankfull width against drainage area regressions by Coastal P lain study. ...................... 149 4-19 Bankfull depth against drainage area regressions by Coastal P lain study. ...................... 149 4-20 Boxplots of maximum discharge-to-mean a nnual discharge by the entire data set and subsets representing physiography, geography, and floodplain types ............................. 150 4-21 Mean annual runoff in the southeastern US Coastal Plain ..............................................150 B-1 Alexander Springs tributary 2 ..........................................................................................163 B-2 Blackwater Creek near Cassia ......................................................................................... 165 B-3 Blues Creek near Gainesville ........................................................................................... 167 B-4 Bowlegs Creek near Fort Meade ......................................................................................169 B-5 Carter Creek near Sebring ................................................................................................171 B-6 Catfish Creek near Lake Wales ........................................................................................ 173 B-7 Coons Bay Branch ...........................................................................................................175 B-8 Cow Creek ................................................................................................................. ......177 B-9 Cypess Slash tributary......................................................................................................179 B-10 East Fork Manatee River tributary ...................................................................................181 B-11 Fisheating Creek at Palmdale...........................................................................................183 B-12 Gold Head Branch............................................................................................................185 B-13 Hammock Branch ........................................................................................................... .187 B-14 Hickory Creek near Ona ..................................................................................................189 B-15 Hillsborough River tributary ............................................................................................ 191 B-16 Horse Creek near Arcadia ................................................................................................193 B-17 Jack Creek ........................................................................................................................195 B-18 Jumping Gully ............................................................................................................ ......197 B-19 Lake June-in-Winter tributary ..........................................................................................199 B-20 Little Haw Creek near Seville .......................................................................................... 201

PAGE 13

13 B-21 Livingston Creek near Frostproof ....................................................................................203 B-22 Livingston Creek tributary. .............................................................................................. 205 B-23 Lochloosa Creek at Grove Park .......................................................................................207 B-24 Lowry Lake tributary. ......................................................................................................209 B-25 Manatee River near Myakka Head ..................................................................................211 B-26 Manatee River tributary. ..................................................................................................213 B-27 Morgan Hole Creek ..........................................................................................................215 B-28 Moses Creek near Moultrie ..............................................................................................217 B-29 Myakka River tributary 1 ................................................................................................. 219 B-30 Myakka River tributary 2 ................................................................................................. 221 B-31 Nine Mile Creek ...............................................................................................................223 B-32 Rice Creek near Springside ..............................................................................................225 B-33 Santa Fe River near Graham ............................................................................................227 B-34 Shiloh Run near Alachua .................................................................................................229 B-35 Snell Creek .............................................................................................................. .........231 B-36 South Fork Black Creek ...................................................................................................233 B-37 Spoil Bank tributary. ........................................................................................................235 B-38 Ten Mile Creek ................................................................................................................237 B-39 Tiger Creek near Babson Park ......................................................................................... 239 B-40 Tiger Creek tributary..................................................................................................... ...241 B-41 Triple Creek unnamed tributary 1 ....................................................................................243 B-42 Triple Creek unnamed tributary 2 ....................................................................................245 B-43 Tuscawilla Lake tributary. ............................................................................................... 247 B-44 Tyson Creek. ............................................................................................................. .......249 B-45 Unnamed Lower Wekiva tributary. ................................................................................. 251

PAGE 14

14 E-1 Width versus stage: Blac kwater Creek near Cassia. ........................................................ 317 E-2 Width versus stage: Blues Creek near Gainesville. .........................................................317 E-3 Width versus stage: Bowleg s Creek near Fort Meade. .................................................... 317 E-4 Width versus stage: Cart er Creek near Sebring. ..............................................................318 E-5 Width versus stage: Catfish Creek near Lake Wales. ...................................................... 318 E-6 Width versus stage: Fisheating Creek at Palmdale. ......................................................... 318 E-7 Width versus stage: Ho rse Creek near Arcadia. .............................................................. 319 E-8 Width versus stage: Little Haw Creek near Seville. ........................................................319 E-9 Width versus stage: Livings ton Creek near Frostproof. ..................................................319 E-10 Width versus stage: Loch loosa Creek at Grove P ark. ......................................................320 E-11 Width versus stage: Manatee River near Myakka Head. ................................................. 320 E-12 Width versus stage: Moses Creek near Moultrie. ............................................................ 320 E-13 Width versus stage: Rice Creek near Springside. ............................................................ 321 E-14 Width versus stage: Santa Fe River near Graham. .......................................................... 321 E-15 Width versus stage: Tige r Creek near Babson Park. ........................................................321

PAGE 15

15 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science DEVELOPMENT OF BANKFULL DISCHARGE AND CHANNEL GEOMETRY REGRESSIONS FOR PENINSULAR FLORIDA STREAMS By Kristen Blanton December 2008 Chair: William Wise Cochair: Joann Mossa Major: Environmental Engineering Sciences Regional curves, which relate bankfull discha rge and channel geometry (cross-sectional area, width, and mean depth) to drainage area in regions of similar climate, geology, and vegetation, have greatly aided in creating target natural channel designs for stream restoration efforts. Regional curves were developed for peninsular Florida based on cross-sectional and longitudinal survey data collected at 17 gage d and 28 ungaged as near-to-natural streams, ranging in drainage area from 0.2 to 311 square miles and valley slope from 0.02 to 2.27%. Based on an analysis of prevalence among sites, slopes, and hydrologic data the elevation of the flat floodplain was determined to be the most re liable bankfull indicator at sites with a wetland floodplain, while the elevation of the inflection on the bank was th e most reliable indicator at sites with an upland floodplain. Analysis of bankfull indicator slopes fu rther revealed that a water slope threshold of approximately 0.5% exis ts, above which bankfull indicators appear to more reliable, suggesting that slope-area techniques for calcula ting the bankfull discharge may be unreliable in peninsular Florida stream s with a water slope less than 0.5%. The dataset was further divide d based on physiography (flatwoods versus highlands), geography (northern versus southern peninsula), and floodplain types (wetland versus upland and cypress-

PAGE 16

16 dominated versus non-cypress-dominated) to dete rmine if significant differences exist in the bankfull regressions and/or various dimensionles s ratios (sinuosity, width-to-depth, maximum depth-to-mean depth, valley slope, and maximum discharge-to-mean annual discharge) among various peninsular Florida stream subsets. Streams with wetla nd floodplains were found to have a significantly greater bankfull area and bankfull width than str eams with an upland floodplain. Also, streams with cypress-dominated floodplains ha d a greater width-to-depth ratio than streams with non-cypress-dominated floodplains. Furthe r, streams draining flatwoods physiographies were found to be flashier. These differences may be important considerations when designing natural channels in peninsular Florida. Annual peak flow data for the gaged sites were analyzed to estimate the bankfull discharge return interval using Log Pearson Type III dist ributions. The bankfull discharge ranged from less than one year to 1.44 years, which is more fr equent than the average 1.5-year return interval often cited in the literature. Based on analysis of the flow duration at gaged sites, bankfull discharge for peninsular Florida streams is equa led or exceeded approximately 21% of the time on average, or about 77 days per year. On average, the bankfull discharge is rough ly four times that of the mean annual discharge an d is 35% of the 1.5-year discharge. Lastly, the regional curves developed for peni nsular Florida were compared to regional curves previously developed for other regions of the s outheastern United States Coastal Plain. Peninsular Florida bankfull chan nels were found to have a lower bankfull discharge and smaller bankfull channel (narrower and sh allower) than northwest Florida streams, which receives more mean annual precipitation and runoff. These di fferences indicate that the regional curves developed in the present work are more appli cable to peninsular Florida streams than are regional curves developed for other regions of the southeaste rn Coastal Plain.

PAGE 17

17 CHAPTER 1 INTRODUCTION Land use changes (i.e., deforestation, agri culture, m ining, and residential and urban development) and channel and floodplain altera tions (i.e., levees, dams channelization, and dredging) have negatively impacted large num bers of streams across the United States by affecting the amount, location, and timing of water movement through a watershed. These watershed alterations can intr oduce hydraulic instability to a system by altering flow and sediment transport rates, and may ultimately lead to increased deposition (aggradation), increased erosion (degradation), or abandonment of existing channels for new ones (Dunne and Leopold, 1978). Because the physical environmen t largely controls sp ecies composition and abundance of stream-dependent fauna (Allan, 1995; Gordon et al., 2004), restoring streams to a more stable and biologically productive state has become a priority for many government agencies and private organizations, and approxi mately $10 billion has been spent on 30,000 river restoration projects in the United St ates to date (Malakoff, 2004). While traditional stream stabilization practices have relied on harden ing reaches with riprap or concrete, natural channel de signs that take a streams natura l tendencies into account have recently gained popularity and are now commonly practiced in many areas. Regional curves, which relate bankfull discharge and channel ge ometry (cross-sectional area, width, and mean depth) to drainage area in regi ons of similar climate, geology, a nd vegetation, have greatly aided in creating target natural channel designs. Bankfull discharge, or flow that fills a stable alluvial channel to the elevation of th e active floodplain, is a useful pa rameter in developing regional curves because its stage is reasona bly identifiable in the field, a nd it is the flow most often used to estimate the channel-forming discharge. Dunne and Leopold (1978) describe bankfull discharge as the most effective streamflow for moving sediment, forming or removing bars,

PAGE 18

18 forming or changing bends and meanders, and generally doing work that results in the average morphological characteristics of channels. While regional curves provide important information for natural channel structure, they also aid in estimating bankfull discharge and channel geometry in ungaged watersheds where drainage area is known, help confirm field identifications of bankfull st age, and allow for comparisons between regions (Leopold, 1994) Because many Florida streams have been degraded due to land use changes and channel and floodplain alterations, development of regional curves for peninsular Florida will provide the necessary data to implement natural channel designs as a stream restoration technique in Florida. These data will be useful to public agencies su ch as Department of Environmental Protection (DEP), United States Geological Survey (USGS), and Department of Transportation (DOT), as well as to private industries such as the phos phate mining industry. Metcalf (2004) published regional curves for Florida streams, yet his study sites were confined to extreme north Florida and the Panhandle and even included sites in Georgia and Alabama (Figure 1-1). Thus these relationships may not be applicable to streams in peninsular Florida, as it is quite different in physiography, geological context, and rainfall patterns. To develop regional curves for pe ninsular Florida, forty-five as near-to-natural peninsular Florida streams were surveyed, ranging in drainage area from 0.2 to 311 square miles and in valley slope from 0.02% to 2.27% (Table, 1-1, Figur e 1-2). Seventeen of the study sites are or historically have been gaged by the U.S. Geol ogical Survey (USGS), while 29 are ungaged. The sites were further divided into subsets ba sed on their physiography, geography, and floodplain types. Twenty-five sites drain a flatwoods physiography (gen erally with an abundance of wetlands, poorly-drained D-type soils, high wa ter tables, flat topography, and many streams), while 21 drain a highlands physiography (generally with an abundance lakes, relict sand dunes,

PAGE 19

19 well-drained A-type soils, low water tables, roll ing topography, and few streams). Twenty of the sites are located in the northern portion of the peninsula (above the 28.5 degrees north latitude line), while 26 are located in the southern portion of the peninsula (below the 28.5 degrees north latitude). Twenty-three sites had a wetland floodplain (dominat ed by hydrophytic vegetation and hydric soils) and twenty-two had an upland floodplain (dominated by hydrophobic vegetation and non-hydric soils). Of the twenty-three sites with a wetland floodplain, 11 were dominated by cypress ( Taxodium spp.) Research objectives were to 1) determin e the most reliable bankfull indicator for peninsular Florida streams; 2) develop bankfull discharge and channel geometry relationships (regional curves) for peninsular Florida stream s; 3) compare bankfull discharge and channel geometry relationships between streams draini ng different physiographi es (flatwoods versus highlands), geographies (north ern versus southern peninsul a), and floodplain types (wetland versus upland and cypress-dominated versus non-cyrpess-dominated); 4) estimate the recurrence interval associated with the ba nkfull discharge for peninsular Fl orida streams; and 5) compare regional curves developed for peni nsular Florida to those previous ly developed for other regions of the southeastern United States Coastal Plain. Hypotheses were 1) when present, the level of a flat depositional floodplain will be the best bankfull indicator for peninsular Florida streams; 2) bankfull discharge and channel geometry relationships will vary in peninsul ar Florida by physiography and floodplain type, but not by geography; 3) bankfull discharge occurs more frequently in peninsular Florida than the often cited 1.5-year return interval for bankfull discharge; an d 4) regional curves developed for peninsular Florida will be significantly different than regional curves developed for other regions of the southeastern United States Coastal Plain.

PAGE 20

20 Table 1-1. Summar y of study sites Site name Count y LatitudeLongitude USGS station ID Physiograph y Geography Floodplain type Drainage area (sq mi) Valley slope (%) Managed lands Managed lands owner Alexander Springs Creek tributary 2Lake29.100-81.576-HLNUP 1.61.042Ocala National ForestUS Forest Service Blackwater Creek near Cassia Lake28.874-81.49002235200HLNWFC1260.020Seminole State ForestDivision of Forestry Blues Creek near Gainesville Alachua29.728-82.43102322016FWNUP 2.60.206San Felasco Hammock SPFDEP Bowlegs Creek near Ft Meade Polk27.700-81.69502295013FWSWF47.20.050--Carter Creek near Sebring Highlands27.532-81.38802270000HLSUP38.80.237Lake Wales Ridge WMAFWC Catfish Creek near Lake Wales Polk27.961-81.49602267000HLSWFC58.90.050Catfish Creek Preserve SPFDEP Coons Bay Branch Hardee27.594-81.857-FWSWF0.50.348--Cow Creek Levy29.231-82.649-FWNWFC5.30.080Goethe State Forest Division of Forestry Cypress Slash tributary Highlands27.597-81.267-HLSUP 0.51.042Avon Park Air Force RangeUS Air Force East Fork Manatee River tributaryManatee27.523-82.106-FWSUP 0.20.313Duette Preserve Manatee Count y Fisheating Creek at Palmdale Glades26.933-81.31502256500FWSWFC3110.029Fisheating Creek WMAFWC Gold Head Branch Clay29.836-81.951-HLNUP 1.81.316Gold Head Branch SPFDEP Hammock Branch Putnam29.540-81.610-HLNWF3.00.167Dunns Creek SP FDEP Hickory Creek near Ona Hardee27.482-81.88002295755FWSWF3.750.116--Hillsborough River tributary Pasco28.216-82.118-FWSWFC0.70.260Upper Hillsborough SWFWMD Horse Creek near Arcadia De Soto27.199-81.98802297310FWSWF2180.043--Jack Creek Highlands27.364-81.426-HLSWF5.20.286Lake Wales Ridge WMAFWC Jumping Gully Lake29.171-81.598-HLNUP 4.61.111Ocala National ForestUS Forest Service Lake June-In-Winter tributary Highlands27.28781.414-FWSUP 0.40.781Lake June-In-Winter SPFDEP Little Haw Creek near Seville Flagler29.322-81.38502244420FWNWFC930.061--Livingston Creek near Frostproof Polk27.709-81.44602269520HLSUP1200.064Lake Wales Ridge SF Division of Forestr y Livingston Creek tributary Polk27.684-81.459-HLSUP 0.40.250Lake Wales Ridge SF Division of Forestr y Lochloosa Creek at Grove Park Alachua29.600-82.14502241900FWNWFC7.40.116--Lowry Lake tributary Clay29.863-81.982-HLNUP0.250.625Camp Blanding FL Dept. of Military Affairs Manatee River near Myakka HeadManatee27.474-82.21102299950FWSUP65.30.116Duette Preserve Manatee County Manatee River tributary Manatee27.483-82.197-FWSUP 0.31.163Duette Preserve Manatee Count y Morgan Hole Creek Polk27.661-81.303-FWSUP 9.40.091Avon Park Air Force RangeUS Air Force Moses Creek near Moultrie St. Johns29.775-81.31602247027FWNWFC7.40.159--Myakka River tributary 1 Sarasota27.239-82.281-FWSUP 2.60.091Myakka River SP FDEP Myakka River tributary 2 Sarasota27.196-82.309-FWSUP 1.70.129Myakka River SP FDEP Nine Mile Creek Lake29.093-81.610-HLNWF 160.488Goethe State Forest Division of Forestry Rice Creek near Springside Putnam29.688-81.74202244473FWNWFC43.20.041Rice Creek Cons. AreaSJRWMD Santa Fe River near Graham Alachua29.846-82.22002320700FWNUP94.90.058--Shiloh Run near Alachua Alachua29.819-82.47202322050FWNUP0.322.000--Snell Creek Polk28.142-81.572-HLSWF1.70.167--South Fork Black Creek Clay29.930-81.942-HLNWF25.50.110Camp Blanding FL Dept. of Military Affairs Spoil Bank tributary (Highlands)Highlands27.068-81.276-FWSUP 8.60.313Smoak Groves Cons. Ease.FDEP Ten Mile Creek Levy29.144-82.617-FWNWFC250.130Goethe State Forest Division of Forestr y Tiger Creek near Babson Park Polk27.811-81.44402268390HLSUP52.80.081Lake Wales Ridge SF Division of Forestr y Tiger Creek tributary Polk27.858-81.487-HLSWF0.90.139Tiger Creek Preserve TNC Triple Creek unnamed tributary 1Hillsborough27.791-82.252-HL SWF1.70.532Balm Boyette; Triple CreekHillsborough County Triple Creek unnamed tributary 2Hillsborough27.797-82.254-FW SUP 0.20.885Balm Boyette; Triple CreekHillsborough County Tuscawilla Lake tributary Marion29.467-82.285-HLNUP 0.32.273Price's Scrub FDEP Tyson Creek Osceola27.940-81.006-FWSWFC20.50.054Three Lakes WM A FWC Unnamed Lower Wekiva tributaryLake28.919-81.405--HLNWF0.40.769Lower Wekiva River SPFDEP Notes: -= Ungaged site or site located on private lands; FW = F latwoods physiography; HL = Highlands physiography; N = Northe rn peninsula geography; S = Southern peninsula geography; WF = Wetland floodplain; WFC = Wetland floodplain dominated by cypress; UP = Upland floodplain; SP = State Park; FDEP = Florida Department of Environmental Pr otection; SF = State Forest; WMA = Wildlife Management Area; FWC = Florida Fish and Wildlife Conservation Commission; TNC = The Nature Conservancy; Cons. Ease. = Conservation Easement Location Data Subsets Independent Variables Managed Lands

PAGE 21

21 Figure 1-1. North and northwest Florida regional curve study sites. Source: Metcalf, 2004 Figure 1-2. Peninsular Florid a regional curve study sites

PAGE 22

22 CHAPTER 2 LITERATURE REVIEW Introduction Metcalf (2004) published regional curves for Florida streams, yet his study sites were confined to extrem e north Florida and the Pa nhandle and even included sites in Georgia and Alabama (Figure 1-1). Thus, thes e relationships may not be applic able to streams in peninsular Florida, as it is quite different in physiography, geological contex t, and rainfall patterns. The following literature review begins by descri bing Floridas unique physiography, geological context, weather and climate, a nd water resources. A description of regional curves follows, which includes the history of regional curve development, the concept of channel-forming discharge, and methods for the iden tification of the bankfull stage. Florida Background Physiography/Geological Context Florida is located within th e Coastal Plain physiographic pr ovince of the United States, which is a region of low relief underlain by unconsolidated to poorly consolidated sedim ents and hardened carbonate rocks (Berndt et al., 1998) (Figure 2-1). Florida s present configuration is largely a result of sea level fluctuations throughout the Cenozoic Era (the last 65 million years of geologic time). Sea level during the early Cenozoic was significan tly higher than present, and carbonate rocks (limestone and dolomite) formed due to the deposition of marine life fossils. Little siliciclastic material from the eroding A ppalachian Mountains reached the Florida Platform during this time due to a marine current running through the Gulf Trough (Figure 2-2A). However, in the mid-Cenozoic the Appalachians were uplifted, increasing erosional rates, and siliciclastic sediments eventually filled th e Gulf Trough and covered Floridas carbonate foundation with sands, silts, and clay s (Lane, 1994) (Figure 2-2B).

PAGE 23

23 Most landforms characterizing Floridas modern topography, as well as the streams, lakes, springs, and wetlands dotting the state today, fo rmed during the most recent period of geologic time, the Quaternary (1.8 million years ago to present) (Lane, 1994). The Quaternary Period, which is made up of two geologic epochs (the Pleistocene or Ice Age and the Holocene), has been a time of world-wide glac iations and widely fluctuating s ea levels, with seas alternately flooding and retreating from Floridas land area. At peak interglaci al stages, sea level rose to approximately 150 feet above the present level, a nd peninsular Florida likely consisted only of islands (Lane, 1994) (Figure 2-3). As seas retreated, waves and curr ents eroded a series of relict, coast-parallel scarps and construc ted sand ridges spanning the state. Many of these features are found today stranded many miles inland, includ ing the Cody Scarp, Trail Ridge, Brooksville Ridge, and Lake Wales Ridge (Lane, 1994). The de velopment of Pleistocene landforms has also been influenced by the karst nature of Flor idas foundation, as naturally acidic rain and groundwater have flowed through the limestone fo r millions of years dissolving conduits and caverns. Sometimes caverns collapse to create si nkholes, the largest of wh ich can be seen today as lakes (Lane, 1994). Two basic physiographies support peninsular Florida streams: 1) Flatwoodsgenerally with an abundance of wetlands, poorly-drained D-type soils, high water tables, flat topography, and many streams; and 2) Highlandsgenerally w ith an abundance lakes, relict sand dunes, well-drained A-type soils, low water tables, roll ing topography, and few streams. One objective of the present work is to determine what, if a ny, differences exist between streams draining these two physiographies. Weather and Climate Although Florida is located at the sam e latitude as some of the worlds major deserts, it is one of the wettest states in the nation, with an average annual rainfall of 53 inches (Henry, 1998).

PAGE 24

24 Florida has two major climate types: humid subtropical in the northern three quarters of the state and tropical savanna in the sout hern portion of the peninsula and the Keys (Figure 2-4). In the tropical savanna climate, all months average over 64 degrees Fahrenheit, and there are distinct wet (June through September) and dry (winter) seasons In the humid subtr opical climate, some months have an average temperature less than 64 degrees Fahrenheit, and the dry season is not as pronounced (Henry, 1998). Rainfall throughout Florida varies considerably from place to place, season to season, and year to year. It averages from 69 inches (W ewahitchka in the Panhandl e) to 40 inches (Key West) annually (Henry, 1998). The wettest places in Florida are the Panhandle where rain falls abundantly throughout the year and the southeastern part of the state where the Gulf Stream enhances the likelihood of rainfa ll (Henry, 1998). The lowest amounts of rainfall occur in the Keys and in the central portion of the peninsula (Figure 2-5). Seasonally, the Panhandle receives proportionately more winter precipi tation from large-scale frontal sy stems than any other part of the state. The southern portion of the state receives proportionately more summer precipitation, as Floridas peninsular shape, converging sea br eezes of the Atlantic Ocean and the Gulf of Mexico, position relative to the Atlantic high pressure system, and tropical and subtropical location make it an ideal spawning ground for th understorms (Henry, 1998). Rainfall throughout the state also varies from year to year with cycles of drought, the occu rrence of hurricanes that can yield 5 to 12 inches of rain, and the phenom ena of El Nio and La Nia (Henry, 1998). Nearly 70 percent of Floridas rain is returned to the atmo sphere through evaporation and evapotranspiration. The remainder flows to its ri vers and streams or seeps into the ground and recharges aquifers (Figure 2-6). Nearly all of Floridas groundwater origin ates from precipitation

PAGE 25

25 (Berndt et al., 1998). Rainfall contributes to stream-f low in Florida through several pathways, including overland flow, interflow, and baseflow (Mossa, 1998). Water Resources With approximately 10,000 miles of rivers and streams, 7,800 lakes, 33 first-magnitude springs (those that discharge water at a rate of 100 cubic feet per second or more), and millions of acres of wetlands, Florida ha s abundant surfa ce water (Kautz et al., 1998). Even more abundant is Floridas groundwater. With more than a quadrillion gallons flowing beneath the surface through the porous underlying limestone, groundwater comprises 30,000 times the daily flow to the sea of Floridas 13 major rivers (Conover, 1973). Regardless of amounts, Floridas unique karst landscape keeps surface water and groundwater we ll-connected th rough features such as sinkholes and springs. Well-developed karst features are also found in south-central Kentucky, Yucatan peninsula, parts of Cuba a nd Puerto Rico, southern China, and western Malaysia; however, Florida supports more rivers and streams than do thes e other karst areas due to its high water tables a nd flat terrain (Purdum, 2002). Floridas karst terrain and flat topography can make determining watershed boundaries difficult. A watershed is defined as the land area that contributes r unoff, or surface water flow to a water body. Local topography controls drainage direction and patterns, though drainage is also influenced by geology, soil, climate, and ve getation (Mossa, 1998). Networks of channel segments form drainage networks, most of which around the world are dendritic (tree-like). In Florida, however, many drainage networks are best described as either deranged, where numerous depressions (i.e., lakes or wetlands) are interspersed along the channel network, or disjointed, where streams and rivers do not fo rm continuous channels on the land surface and may disappear underground in sinks or depressions (Mossa, 1998). Other areas of Florida, such as the Everglades, are poorly drained with fe w or no streams and water flows across the surface

PAGE 26

26 through swamps and marshes as sheetflow. Florid as major rivers and watersheds are depicted in Figure 2-7 and Figure 2-8, respectively. Several classification systems have been deve loped to categorize Floridas more than 1,700 rivers and streams (Nordlie, 1990). These were developed primarily by ecologists and are based mainly on faunal metrics, water quality, and sedi ment type. Beck (1965) developed the most commonly used classification of Florida wa terways, which includes the following five categories: Sand-bottomed streams: slightly acidic with moderately high color; the most widely distributed and abundant stream type in Florida Calcareous streams: predominantly of spring origin with relatively cool, clear, and alkaline waters Swamp-and-bog streams: very acidic, highly colored, sluggish streams, which originate in swamps, sphagnum bogs, and marshes Large rivers: carry significant sediment loads a nd are always turbid; a category of convenience Canals: also a category of convenience The Florida Natural Areas Inventory (F NAI, 1990) refined Becks work by adding descriptions of landscape settings and water sources to their classi fication system, which includes four categories: Alluvial streams: originate in high uplands and are ty pically turbid due to high sediment loads; typically flood once to tw ice a year, providing an impor tant pulse of nutrient-rich water to the floodplain, as well as sediment for natural levee development; sparsely distributed in Florida and primar ily restricted to the Panhandle Blackwater streams: originate in sandy lowlands where wetlands slowly discharge tannic waters to the channel; genera lly acidic waters; most wide ly distributed and numerous stream type in Florida Seepage streams: originate from an unusual geologic process in which rainwater percolates through deep, sandy upland soils and encounters an impermeable layer causing the water to travel laterally until reaching a surface and prod ucing a seepage face; clear to lightly colored water; generally small in magnitude

PAGE 27

27 Spring-run streams: derive most of their water from artesian vents in the underground aquifer; clear, slightly alka line, cool water; generally have sandy bottoms or exposed limestone. Many Florida rivers are actually a combinati on of stream types. For example, the Suwannee begins as a blackwater river draining the Okefenokee Swamp, but becomes a springfed river as it travels south where many springs contribute to its flow. As the Suwannee approaches the Gulf, it has a low-forested floodpla in more characteristic of an alluvial river (Kautz et al., 1998). Though not specific to Florida streams, Rosgen (1994) developed what is currently the most comprehensive and commonly used stream cl assification system based on the principles of fluvial geomorphology. Rosgen ( 1994) first identified seven major stream types based on differences in geomorphic variables (i.e., entr enchment ratio, width/depth ratio, sinuosity, and channel slope) that can be seen when displayed in the following two-dimensional perspectives: Longitudinal profile: compares the elevation of the wa ter or bed surface with distance downstream; bed features can be inferred from this perspective as these features have consistently been found to relate to channel slope Cross-section: compares the elevation with the width or distance across the channel; width to depth ratio, level of confinement (lateral containment), and level of entrenchment (vertical containment) can be inferred from this perspective Plan form: compares width across the channel wi th distance along the channel; sinuosity, meander width ratio (belt width/bankfull surface width), and radius of curvature can be inferred from this perspective. (Figure 2-9) Rosgen (1994) then identified six additiona l stream types, which were delineated by dominant channel material rangi ng in particle size diameter fr om bedrock to silt/clay. When combined with the previous stre am types, 42 major stream types emerged (Figure 2-10). Metcalf (2004) applied Rosgens shape-based classificati on to streams in extreme north Florida and the Panhandle and identified two ma jor physical classes of str eamsC5, which are broad and shallow sand-bottomed streams, and E5, which are deep and narrow sand-bottomed streams.

PAGE 28

28 Rosgens classification system works on the assu mption that streams are under alluvial control, meaning that their shape is str ongly dictated as a func tion of sediment transport. Because of Floridas unique geology and climate, its fluvia l forms are under variable degrees of alluvial control and may not lend themselves to this type of reach-scale, form-based classification that is now widely used throughout the United States (K iefer, personal communication). For example, Floridas mild humid climate allows for a nearly year-round growi ng season, and vegetation probably exerts significant confinement on ch annel cross-section morphology and planform patterns in Florida compared to other re gions (Kiefer, personal communication). Regional Curves Fluvial geomorphology is often th e most fundam entally important scientific discipline for managing riparian corridors or planning ecological restoration of damaged stream ecosystems. Stream pattern is directly influenced by eight major variables: channel width, depth, velocity, discharge, channel slope, roughness of channel ma terials, sediment load, and sediment size (Leopold et al., 1964). Change in any one of these va riables sets up a series of channel adjustments that can lead to change in the othe rs, resulting in channel pattern alteration (Rosgen, 1994). Land use changes (i.e., deforestation, ag riculture, mining, and residential and urban development) and channel and floodplain altera tions (i.e., levees, dams channelization, and dredging) have impacted large numbers of st reams across the United States by affecting the amount, location, and timing of water moveme nt through a watershed. These watershed alterations can introduce hydraulic in stability to a system by altering flow and sediment transport rates, and may ultimately lead to increased deposition (aggradatio n), increased erosion (degradation), or the abandonment of existing ch annels for new ones (Dune and Leopold, 1978). Because the physical environment largely c ontrols species composition and abundance of stream-dependent fauna (Allan, 1995; Gordon et al., 2004), restoring stream s to a more stable

PAGE 29

29 and biologically productive state has become a priority for many government agencies and private organizations. While traditional stream stabilization practices have relied on harden ing reaches with riprap or concrete, natural channel de signs that take a streams natura l tendencies of adjustment into account have recently gained popularity and are now commonly practiced in many areas. Regional curves, which relate bankfull discharg e and channel geometry (cross-sectional area, width, and mean depth) to drai nage area in regions of simila r climate, geology, and vegetation, have greatly aided in creating ta rget natural channel designs. Ba nkfull discharge, or flow that fills a stable alluvial channel to the elevation of the active flo odplain, is a useful parameter in developing regional curves because its stage is r easonably identifiable in the field, and it is the flow most often used to estimate the channe l-forming discharge. Dunne and Leopold (1978) describe bankfull discharge as the most effec tive stream-flow for movi ng sediment, forming or removing bars, forming or changing bends and mea nders, and generally doi ng work that results in the average morphological characteristics of channels. While regional curves provide important information for natural channel stru cture, they also aid in estimating bankfull discharge and channel geometry in ungaged watersheds when drainage area is known, help confirm field identifications of bankfull stag e, and allow for comparisons between regions (Leopold, 1994). History of Regional Curve Development Dunne and Leopold (1978) are often credited as the pioneers of regional curves. They found that strong correlations ex ist between bankfull discharge a nd drainage area, as well as between bankfull channel geom etry (cross-sectiona l area, width, and mean depth) and drainage area, in regions of similar climate, geol ogy, and vegetation. Dunne and Leopold (1978) developed regional curves for f our regions of the United States (Figure 2-11), including the San

PAGE 30

30 Francisco Bay region, the eastern United States (more specifically, the Brandywine area of Pennsylvania), the Upper Green River in Wyomi ng, and the Upper Salmon River in Idaho (data for which were collected and published by Emme tt, 1975). Their work revealed regional differences between the rainfall -runoff channels of the east and west coasts and the snowmeltrunoff channels of Idaho and Wyoming. Alt hough Dunne and Leopold (1978) are often credited as the pioneers of the regional curve, ol der studies conducted by Nixon (1959) and Emmett (1975) developed similar curves for England and Wales and for the Upper Salmon River in Idaho, respectively. Prior to these studies, Leopold and Maddock ( 1953) developed hydraulic geometry theory, in which channel geometry characteristics such as width, depth, and velocity vary directly with discharge as simple power functions, as shown in Figure 2-12 and Equatio ns 2-1, 2-2, and 2-3. w = aQb (2-1) d = cQf (2-2) v = kQm (2-3) In Equations 2-1 to 2-3, w is the width [L], d is the mean depth [L], v is the mean velocity [LT-1], and Q is the water discharge [L3T-1]. The constants b, f, and m are empirical and represent slopes of the three lines, the sum of which should equal 1.0. The constants a, c, and k are also empirical and represent the intercepts of the three line s, the product of which should equal 1.0. Leopold and Maddock (1953) also post ulated that discharg e is dependent upon drainage area characteristics, which dictate runoff and sediment production. Channel-forming Discharge Several term s are used throughout the literature to describe channel-forming discharge, including dominant discharge, effective discharg e, and bankfull discharge. Although these are often used interchangeably, they have distinct definitions and a brief description of each is thus useful in understanding the concept of channel-forming discharge.

PAGE 31

31 Dominant discharge is defined as the theoretical discharge that if maintained indefinitely in an alluvial stream would produce the same channel geometry as the natural long-term hydrograph (Copeland et al., 2000). Effective discharge, on the other hand, can be derived mathematically and is defined as the discharge th at transports the largest fraction of the average annual bed-material load (Copeland et al., 2000). Effective discharge incorporates the principle prescribed by Wolman and Miller (1960) that ch annel-forming discharge is a function of both the magnitude of the event and the frequency of occurrence. Wolman and Miller found that lowmagnitude, relatively high-frequency events (occu rring at least once each year or two), rather than rare catastrophic floods (occurring once in fifty or a hundred years), are the most effective in transporting sediment and performing work. As shown in Figure 2-13, the effective discharge occurs at the peak of the curve obtained by multiplying the flood frequency curve and the sediment discharge rating curve. Developm ent of a sediment discharge rating curve is difficult; however, because it requires collecting field data of bedload and total suspended sediment coupled with discharge over a wide range of flows (Metcalf, 2004). Effective discharge is thus not often used to develop regional curves. Bankfull discharge is the most commonly used channel-forming discharge in the development of regional curves because it may be reasonably identified in the field by physical indicators (which will be described below). It is defined as flow that fills a stable alluvial channel to the elevation of th e active floodplain (Figure 2-14). Leopold (1994) defines the active floodplain as the flat area adjacent to the river channel, construc ted by the present river in the present climate and frequently subjected to overflow. Bankfull discharge is thus morphologically significant becau se it represents the breakpoint between the processes of channel formation (erosion) and floodplain formation (deposition) (Copeland et al., 2000). Gage

PAGE 32

32 station analysis throughout the United States has shown that bankfull discharge has average recurrence interval of 1.5 years, or a 66.7% a nnual exceedance probab ility (Dunne and Leopold, 1978; Leopold, 1994) (Figure 2-15). However, this widely reported assertion that bankfull discharge occurs on average once every one to two years is now seen as oversimplification (Thorne et al., 1997), with several recent studi es (particularly in the so utheastern United States Coastal Plain) reporting much lower bankfull disc harge recurrence intervals (Table 4-12). One objective of the present work is to estimate th e recurrence interval of bankfull discharge in peninsular Florida streams. Although many hydrologists and river engineers work under the assumption that dominant, effective, and bankfull discharges are approximate ly equal, this is controversialwhile some have found effective and bankfull discharges to be in agreemen t, others have found that the former occurs more frequently than the latt er (Knighton, 1998). It is thus important to understand that a channel is formed by a range of flows, and that bankf ull discharge is but a surrogate of these flows (Knighton, 1998; Emmett, 2004). Indicators of Bankfull Stage Proper identification of bankfull st age, or the elevation at w hic h the stream just begins to overtop its floodplain, is critical to both development of region al curves and calculation of bankfull discharge (Emmett, 2004). Field identi fication of bankfull stag e is the method most often used to estimate channel-forming flow, though its correct iden tification in the field can be difficult and subjective (Knighton, 1998). U.S. Fo rest Service has published a field guide for both determining bankfull stage and conducting a stream channel survey (Harrelson et al., 1994). Videos demonstrating how to identify bankfull st age in the Western and Eastern United States are also available from the U.S.D.A. Fore st Service (1995, 2003). Some common field indicators of bankfull stage include:

PAGE 33

33 Top of bank for non-incised channels Height of depositional featuresespecially the top of the pointbar Position on the bank where the slope first becomes levelthis feature can be identified by facing the stream and draggi ng your foot until it flattens Change in vegetationespecially the lower limit of perennial species Slope or topographic breaks along the bank Change in the particle size of bank materi alsuch as the boundary between coarse cobble or gravel with fine-grained sand or silt Undercuts in the bankwhich usually reach an interior elevation slightly below bankfull stage Stain lines or the lower extent of li chens on boulders or trees. (Harrelson et al., 1994; Leopold, 1994; U.S.D.A. Forest Service, 1995 and 2003) Several analytical, non-field based techniques can also be used to determine bankfull stage, including: Stage-discharge rating curves the inflection point on the rating curv e that corresponds to the point at which the stream overtops its bank and the stage consequently levels off (Figure 2-16) Elevation at which the width-to-depth ratio is at a minimum (Figure 2-17) Flood frequency analysis of available stream gage databankfull discharge has an average recurrence interval of 1.5 years (D unne and Leopold, 1978; Leopold, 1994) Regional curvesalthough the present work focuses on using bankfull indicators to develop regional curves, regional curves, in turn, can be used to confirm field identification of bankfull stage (FISRWG, 1998; Wolman, 1955; Leopold, 1994). Reliable indicators have not b een verified for peninsular Florida, though Metcalf (2004) found that bankfull indicators in extreme north Florida and the Panhandle were most often the top of bank or sometimes a lower bench/bar fe ature. Studies conducted on North Carolina streams found that the top of ba nk or lowest scour or bench was rarely an indicator of bankfull and determined that the highest scour line or the back of the point bar was the most consistent

PAGE 34

34 bankfull indicator (Harman, 1999). One objective of the present work is to determine the most reliable bankfull indicator for peni nsular Florida (Chapter 3). Once bankfull stage has been determined, bank full cross-sectional ar ea, width (width of water surface at bankfull stage), mean depth ( quotient of bankfull cros s-sectional area and bankfull width), and discharge can be determined. In gaged streams, bankfull discharge can be determined from a stage-discharge (Stage-Q) rating curve. In ungaged streams, Mannings equation (Equation 2-4) can be used to calculate bankfull discharge. v = km/n R2/3 S1/2 (2-4) In Equation 2-4, v is the veloci ty, km is a numerical constant (1.49 for units of feet and seconds and 1.0 for units of meters and seconds), n is the roughness coeffi cient (Mannings), R is the hydraulic radius (quotient of cross-sectional area and wetted perimeter) [L], and S is the water slope. Bankfull discharg e and channel geometry (cross-sectional area, width, and mean depth) can then be plotted against drainage area for a population of streams, and a regression can be fit to develop a regional curve (Leopold, 1994). Conclusions There has been a recent surge in regional curve developm ent throughout the United States, which can be attributed to the increased popul arity of natural channe l design as a stream restoration technique. While trad itional stream stabilization pract ices have relied on hardening reaches with rip-rap or concrete, natural channel designs that take a streams natural tendencies of adjustment into account have recently gain ed popularity and are now commonly practiced in many areas. Regional curves, which relate bankfu ll discharge and channel geometry to drainage area in regions of similar climate, geology, and vege tation, have greatly aided in design of stable stream channels. Regional curves also ai d in estimating bankfull discharge and channel

PAGE 35

35 geometry in ungaged watersheds where the dr ainage area is known, help confirm field identifications of bankfull stag e, and allow for comparisons between regions (Leopold, 1994). Metcalf (2004) published regional curves for Flo rida streams, yet his sites were confined to extreme north Florida and the Panhandle, a nd even included sites in Georgia and Alabama (Figure 1-1). Peninsular Florida, however, is quite different in physiography, geological context, and rainfall patterns. For example, the Panhandl e receives abundant rain throughout the year and proportionately more winter pr ecipitation due to large fronta l-based storms coming off the mainland, while the peninsula receives less rain throughout the year a nd proportionately more summer precipitation due to convective storms occurring from the convergence of Gulf of Mexico and Atlantic Ocean sea breezes (Henry, 1998). As a result, streams draining these regions likely have significant differences in their bankfull discha rge and channel geometry for a given drainage area. Development of regional curv es for peninsular Florida is thus justified, and one objective of the present work is to determine what, if any, differences exist between peninsular Florida streams and those of other re gions of the southeastern United States Coastal Plain.

PAGE 36

36 Figure 2-1. Physiographic pr ovinces of the United States Source: Fenneman, 1946. A B Figure 2-2. Geologic history of Florida. A) Through Oligocene time the Florida Platform was a shallow, marine limestone bank envir onment. Currents through the Gulf Trough diverted sands, silts, and clays that were eroding off the Appalachian Mountains to the north. B) Siliciclastic sediments ha d filled the Gulf Trough by Miocene time and encroached down the peninsula, covering th e limestone environments. Source: Lane, 1994.

PAGE 37

37 Figure 2-3. Pleistocene shorelines in Florida. Source: Lane, 1994. Figure 2-4. Climate zones in Florida. Source: Henry, 1998.

PAGE 38

38 Figure 2-5. Florida precipit ation map. Source: Henry, 1998. Figure 2-6. Floridas water cy cle. Source: Purdum, 1998.

PAGE 39

39 Figure 2-7. Floridas surface wate r drainage. Source: Mossa, 1998.

PAGE 40

40 Figure 2-8. Floridas wate rsheds. Source: Mossa, 1998. Figure 2-9. Longitudinal, crosssectional, and plan views of major stream types. Source: Rosgen, 1994.

PAGE 41

41 Figure 2-10. Cross-sectional configuration, composition, and delineative crit eria of major stream types. Source: Rosgen, 1994. A B Figure 2-11. Regional curves for four US regions. A) Bankfull discharge against drainage area. B) Bankfull channel geometry against dr ainage area. Source: Dunne and Leopold, 1978.

PAGE 42

42 Figure 2-12. Relation of width, depth, and veloc ity to discharge, Powder River at Arvada, Wyoming. Source: Leopold and Maddock, 1953. Figure 2-13. Effective discharge determination from sediment rating and flow duration curves. The peak of curve C marks the discharge that is most effective in transporting sediment. Source: FISWRG, 1998 adap tation of Wolman and Miller, 1960.

PAGE 43

43 Figure 2-14. Channel cross section identifyi ng bankfull parameters. Source: FISRWG, 1998. Figure 2-15. Amount of water in a river cha nnel and frequency with which such an amount occurs. Source: Leopold, 1994.

PAGE 44

44 Figure 2-16. Determination of bankfull stage from a stage-discharge rating curve. The inflection point on the rating curve corre sponds to the point at whic h the stream overtops its bank and the stage consequently le vels off. Source: FISRWG, 1998. Figure 2-17. Determination of bankfull stage fr om a plot of width-to-depth ratio against maximum depth. The elevation at which the width-to-depth ratio is at a minimum is the suggested bankfull level. Source: Copeland et al., 2000 citing Knighton, 1984 P.163.

PAGE 45

45 CHAPTER 3 DETERMINING THE MOST RELIABLE BANKF ULL INDICATOR FOR PENINSULAR FLORIDA STREAMS Introduction Proper identification of bankfull st age, or the elevation at w hic h the stream just begins to overflow onto its floodplain, is cri tical to both development of re gional curves and calculation of bankfull discharge (Emmett, 2004). The floodplain is defined as the relatively flat, depositional surface adjacent to the stream that is being built and rebuilt by a st ream in the present hydrologic regime (Emmett, 2004). Bankfull discharge is morphologically significant because it represents the breakpoint between processes of channel formation (erosion) and floodplain formation (deposition) (Copeland et al., 2000). Field identification of ba nkfull stage is the method most often used to estimate the channel-forming flow, though its correct identifi cation in the field can be difficult and subjective (Johns on and Teil, 1996; Knighton, 1998) U.S. Forest Service has published a field guide for both determining ba nkfull stage and conductin g a stream channel survey (Harrelson et al., 1994). Videos demonstrating how to identify bankfull stage in the Western and Eastern United States are also av ailable from the U.S.D.A. Forest Service (1995, 2003). Some common field indicators of bankfull stage include: Top of bank for non-incised channels Height of depositional featuresespecially the top of the pointbar Position on the bank where the slope first becomes levelthis feature can be identified by facing the stream and draggi ng your foot until it flattens Slope or topographic breaks along the bank Change in vegetationespecially the lower limit of perennial species Undercuts in the bankwhich usually reach an interior elevation slightly below bankfull stage

PAGE 46

46 Change in the particle size of bank materi alsuch as the boundary between coarse cobble or gravel with fine-grained sand or silt Stain lines or the lower extent of lichens on boulders or trees (Harrelson et al., 1994; Leopold, 1994; U.S.D.A. Forest Service, 1995 and 2003). Several analytical, non-field based techniques can also be used to determine bankfull stage, including: Stage-discharge (Stage-Q) rating curves inflection point on the ra ting curve corresponds to the point at which the stream overtops its ba nk and stage consequently levels off (Figure 2-16) Elevation at which the width-to-depth ratio is at a minimum (Figure 2-17) Flood frequency analysis of available stream gage datagage station analysis throughout the United States has shown that bankfull disc harge has an average recurrence interval of 1.5 years, or a 66.7% annual exceedance proba bility (Dunne and Leopold, 1978; Leopold, 1994) Regional curvesalthough the current work focuses on using bankfull indicators to develop regional curves, regional curves, in turn, can be used to confirm field identification of bankfull stage (FISRWG, 1998; Wolman, 1955; Leopold, 1994). Reliable indicators have not b een verified for peninsular Florida, though Metcalf (2004) found that bankfull indicators in extreme north Florida and the Panhandle were most often the top of bank or sometimes a lower bench/bar featur e. Studies conducted in other regions of the Coastal Plain, such as North Carolin a, found the top of bank or lowest scour or bench to rarely be an indicator of bankfull and determined the highes t scour line or the back of the point bar to be the most consistent bankfull indicator (Harman, 1999). One objective of the present work is to determine the most reliable bankfull indicator for as near-to-natural peninsular Florida streams. To accomplish this objective, various indicators of bankfull stage were identified, surveyed, and anal yzed individually to determine if there is a single most reliable bankfull indicator for peni nsular Florida streams. The following factors were examined: prevalence of each bankfull indi cator among study sites; how closely the slope

PAGE 47

47 of each bankfull indicator matches the slope of the water; and how frequently and for how long discharge and stage associated with each bankfull indicator occur. This chapter outlines the methods used to reach the objectiv e, including selecti on of study sites, completion of reference reaches surveys, and analysis of both field da ta collected during reference reach surveys and long-term hydrologic data obtai ned from the United States Geological Survey (USGS). The methods are followed by the study results; a disc ussion of the potential errors, trends, and anomalies associated with the data collec tion and analyses; and conclusions. Methods Tasks com pleted to determine the most reliabl e bankfull indicator fo r peninsular Florida streams included: 1) selecting between 40 and 50 gaged and ungaged stream sites that span a variety of physiographies and geographies; 2) co nducting reference reach surveys to measure the plan form, longitudinal profile, a nd cross-sections of the bankfull channel; and 3) analyzing both field data collected during the reference reach surveys and lo ng-term hydrologic data obtained from the USGS. Site Selection Site selections were lim ited to streams located roughly between the Santa Fe River watershed and Lake Okeechobee to assure that the stream population was pe ninsular rather than continental. Only sites with base levels two feet higher than mean high tide were included to assure that systems were palustrine rather than estuarine. The USGS site inventory ( http://fl.waterdata.usgs.gov/nwis/ ) was used to select gaged sites tha t met the initial inclusionary criteria, which included: at least ten years of continuous or peak di scharge measurements (though a two year record was accepted for basin areas betwee n zero and ten square miles) no reaches and/or basins with water c ontrol structures, ditches, or canals

PAGE 48

48 less than 20% of basin is impervious cover less than 20% of basin is ditched or has i nduced discharge (i.e., ag ricultural tail water) less than 10% of basin is mined no major roads no significant land use changes during or si nce the gaging period, which was determined by examining historical aerial photographs at the University of Floridas Map and Imagery Library. Twenty-seven gaged sites were selected using this method. To supplement the gaged sites, areas defined by the Cadastral Sectional grid we re randomly selected to fill the roster with ungaged sites. If the selected Section cont ained more than one stream segment, it was successively quartered, and one of the quarters was then randomly selected until the selected polygon contained just one stream. A stream was then rejected if it did not meet the above inclusionary criteria (minus the minimum gage record criterion). Of the first 100 unaged sites selected in this fashion, 75 str eams were rejected. To select sites more efficiently, Cadastral Sections were restricted to publ ic landholdings, such as state parks, state and national forests, water management district lands, state wildlife lands, military bases, and county preserves, and to large private landholdings not subject to futu re development, such as those owned by the Nature Conservancy and those under conservation easement. Once 70% of the sites had been selected, these were graphically plotted based on their drainage area and valley slope to ensure that the sample was not skewed towards a cluste red regression. Sites continued to be selected randomly, but were rejected if they fit a redundant drainage area to valley slope bin. Fifty-two unaged sites were selected in this manner. Following initial site selection, landowners identifie d using county property appraisal maps were contacted to obtain access to the study sites. They were sent a formal letter requesting permission to access the stream from their property, as well as a permission form that was to be

PAGE 49

49 filled out and mailed back (Appendix A). This method had a surprisingly high response rate, and only 3 landowners denied access to the study site from their property. For sites located on publicly managed lands, the appropriate permits were obtained. Once appropriate permission was obtained to access selected sites, initial field investigations were conducted. Sites were ultimately exclud ed from the study if they had negative local effects (i.e., cattle grazing, ditchin g, evidence of logging, bri dge or road effects), were not single-threaded channels (i.e., braided or anastomosed stream types), did not have a defined channel (i.e., sloughs), had unsafe worki ng conditions (i.e., non-wadeable, presence of large alligators), and/or had uncoope rative landowners. Forty-five of the originally selected sites were ultimately surveyed, 17 of which were gaged, and 28 of which were ungaged (Table 1-1, Figure 1-2). Reference Reach Surveys A reference reach su rvey was conducted at each selected stream site according to Harrelson et al. (1994). Crosssectional and longitudinal surv eys were completed along a minimum reach length of 20 times the channel width (top of bank to top of bank) to determine bankfull width, mean bankfull depth, maximum bankfull depth, bankfull cross-sectional area, slope, and sinuosity of the channel. A Leica To tal Station and a handheld data collector running Carlson SurvCE (Carlson) were used to record measurements to 1000th of a foot, as per accuracy of the equipment. Depth of water at the thalweg was recorded to the nearest 10th of a foot. Plan, longitudinal, and cross-section pr ofiles are provided in Appendix B. Photographs taken in the upstream, downstream, right bank, and left bank di rections at many site s are provided in Appendix C. The survey crew, which generally consisted of two individuals, followed these step-by-step methods to conduct the reference reach surveys:

PAGE 50

50 Step 1: Explore the stream by walking along or in it. Find a representati ve reach that does not cross any obvious breaks in va lley slope and does not span the entry of a tributary. Note indicators of bankfull stage and look for a repr esentative riffle at wh ich to establish the classification riffle cross-section. Step 2: Set a pin flag at a downstream riffle (X S-1). Measure the distance from one bank to the other and extend the survey upstream 20 times this distance, setting flags every channel width distance apart. Upon completion of flagging, there should be 21 flags/longitudinal stations along the reach, each located one channel width dist ance apart. For example, a ten-foot wide stream requires a 200-foot long reach, with flag s set every ten feet apart along the reach. Distances are measured along the thalweg of the ch annel. In smaller streams, this should be done by running a 300-foot long meas uring tape, while in larger streams this can be done by pacing. Step 3: Flag various indicators of bankfull stage at six cros s-sections along the reach, generally at every other odd-numbered flag (X S-1, XS-5, XS-9, XS-13, XS-17, XS-21). Choose one of these cross-sections (gen erally the shallowest riffle) to be the classification riffle. Bankfull indicators include the following: Position on the bank where slope first becomes level (BKF-F): This feature can be identified by facing the stream and dragging your foot along the bank until it flattens. Inflection or break in sl ope of the bank (BKF-I): This feature can be identified by finding the first break in the banks slope as you look or feel from the streambed up the side of the bank. Top of point bar (BKF-TOPB): Bankfull stage is the boundar y between zones of routine sediment transport versus deposition. The top of the point bar repres ents the height of a depositional feature. Top of scour or undercuts in the bank (BKF-S): This feature usually reaches an interior elevation slightly below bankfull stage and may be found around plant roots.

PAGE 51

51 Bottom of moss collars (BKF-M): This feature should only be r ecorded if moss is at least one inch thick. Alluvial break (BKF-A): This feature can be identif ied by finding the break between more easily transported streambed material and less easily transported bank material. This break may be found where roots become denser and prevent movement of sediment from the banks, where sediment texture changes (i.e., bank material may consist of more organics), or where sediment color changes (i .e., bank material may be darker in color due to the presence of organics) (Figure 3-1). Step Four: Establish two temporary benchmarks (TBM-1 and TBM-2) near the reach by driving plastic-capped metal rods into the ground near a feature unlikely to change position or elevation within a few years (i.e., base of large live oak tree, upland terra ce near edge of stable floodplain). Set the tripod up ove r TBM-1, mount the Total Station onto the tripod, and level it using knobs on the unit. Establish a referen ce datum for the site by assigning a reference elevation of 100 feet to TBM-1. Elevations do not n eed to be tied to actual elevations, as all data will be relative to the datum. Elevations may, however, be tied to a known elevation if desired. When using a total station, also assign a refere nce northing (5000 feet) a nd easting (2000 feet) to TBM-1. Backsight to TBM-2 to establish a zer o angle. If a USGS or other permanent benchmark is available near the si te, this can serve as TBM-2. Step 5: Sketch a detailed plan form site map showing any distinctive features (such as secondary channels or backwate r areas), TBM-1 and TBM-2 locati ons, cross-section locations, a direction of flow arrow, a nor th direction arrow, and gage station location (if present). Step 6: Begin the survey. Collect longitudina l survey measurements, including the thalweg, water surface, and two streambed points, at each of the 21 cross-sections. At each thalweg and streambed point, qualit atively classify dominate subs trate/habitat as sand (SAND), mud (MUD), leaf packs (LEAF), fine woody debris (FWD), or large woody debris (LWD). At the six-cross sections along the reach where bank full indicators were flagged, collect additional cross-sectional measurements, including top of bank (TOB) and various bankfull indicators

PAGE 52

52 (BKF). Record the ecosystem type at each top of bank point according to its determined Florida Land Use, Cover and Forms Classi fication System (FLUCCS) (1999) code. At the cross-section selected to be the classifica tion riffle, extend the survey in to the floodplain by at least two channel widths on either side of the channel, making the cross-section at least five channel widths long. Capture unique floodplain features, such as natura l levees and oxbows, and record any changes in FLUCCS. Sketch a detailed cross-sectional view of the classification riffle. Keep the survey error to less than 0.03 feet throughout the su rvey traverse, which is the minimum amount of error preferred fo r the typical distances involved. Step 7: Upon completion of the survey, the fo llowing various field tasks remain: Record locations of TBM-1, TBM-2, and the do wnstream and upstream ends of the reach using a sub-meter GPS. Take four photographs at the classification riffle, one point ing upstream, downstream, to the right bank, and to the left bank. Estimate percent canopy using a densitometer. Estimate base level of the stream by finding the depth at which a penetrometer reaches refusal at the thalweg, on the right bank, and on the left bank. Note dominant bed and bank material. Remove flags. Step 8: Upon returning to the office, download data from the Carlson into a computer and enter it into RIVERMorph 4.0.1 Stream Restor ation Software (RIVERMorph), a program developed by Wildland Hydrology. This summarizes the field methods utilized to perform the 45 reference reach surveys conducted in this study. Additional informa tion on conducting reference reach surveys is provided by Rosgen (1996) and the USDA Forest Service (Harrelson, et al., 1994). It is important to note that for reference reach surv eys conducted at gaged sites, the longitudinal

PAGE 53

53 survey was carried through the gage plate when possible. However, in instances where the gage plate was located at a bridge that had obvious effects on the hydraulics of the stream (as was often the case) or where permissions could not be obtained, the survey was conducted at a sufficient distance upstream or downstream of the bridge (Table 3-1). Data obtained from the reference reach survey s were then used to determine the size (bankfull cross-sectional area, bankfull width, and bankfull mean depth), shape (width-to-depth ratio and maximum depth-to-mean depth ratio), pattern (sinuosity), and slope of each stream. RIVERMorph was used to calculate many of thes e parameters. When calculating the various bankfull parameters, RIVERMorph used the averag e of the left and right bank indicators to determine bankfull elevation at each cross-section. Sinuosity, which is a parameter that describes the meander pattern of a stream, was determined by dividing channel length surveyed in the longitudinal survey by valley length, whic h was calculated from the survey points using Equation 3-1. Pertinent data for each site were th en entered into Microsof t Excel for further data analysis, graphing, and regional curve development, which are discussed in later sections. (3-1) Data Analysis Slopes of field indicators For each site, RIVERMorph was used to plot a b est fit line both through the survey points of each individual field bankfull indicator (BKF -F, BKF-I, BKF-S, BKF-A) and through the top of bank survey points (TOB). Each slope was th en compared to the slope of a line best fit through the water surf ace survey points1. Leopold (1994) used this technique to verify the feature as bankfull if th e two lines were generally parallel a nd consistent over a long reach. To 1 For sites that did not have flowing water on the day of the survey, each bankfull indicator slope was compared to the slope of a line best fit through the channel bed (thalweg) survey points.

PAGE 54

54 determine how parallel the lines were, water slope was divided by slope of each bankfull indicator to determine a water sl ope to bankfull indicator slope ra tio. Theoretically, the closer the ratio is to one, the more parallel the indicator is to the water and thus the more reliable it is. Bankfull indicator slopes within 25% of the water slope, or those with a water slope to bankfull indicator ratio between 0.75 and 1.25, were thus deemed reliable candidate field indicators. Gage analysis Hydrologic data for the 17 surveyed gaged si tes were obtained from the USGS and used to analyze the various bankfull indicators. Spec ifically, daily streamflow (discharge) and gage height (stage) measurements, field measurements, annual peak flow measurements, and drainage area were downloaded off the Internet from the USGSs online National Water Information System (NWIS), while current stage-discharge (s tage-Q) rating tables were obtained from USGS personnel. USGS data were used in conjunction with th e reference reach survey data to determine stage, discharge, return interval, and duration associated with both top of bank and with the various bankfull indicators (BKF-F, BKF-I, BKF-S, BKF-A) at each gaged site. Stream stage and discharge measurements for the specific day the reference reach survey was conducted were downloaded from NWIS (Table 3-1). The stage of each bankfull indicator was then determined by adding the average difference between elevati on of the bankfull indicator and that of the water surface at the time of the reference reach survey to the stage recorded by the USGS on the day of the survey. The most current stage-Q rating table was then us ed to find discharges associated with various determined stages. The di scharges and stages associated with both top of bank and various bankfull indicators were then pl otted graphically onto each gaged sites stageQ rating curve (Appendix D). Stage-Q rati ng curves were developed by plotting the dimensionless discharge (daily mean discha rge divided by mean annual discharge, Q/Qma)

PAGE 55

55 against the adjusted stage (daily mean stage minus the mean annual stage). Dimensionless discharge and adjusted stage were used to facilitate compar isons among gaged sites. Gage station analysis throughout the United St ates has shown that bankfull discharge has an average recurrence interval of 1.5 years, which corresponds to a 66.7% annual exceedance probability (Dunne and Leopold, 1978; Leopold, 1994) (Figure 2-15). However, this widely reported assertion that bankfull discharge occurs on average once every one to two years is now seen as an oversimplification, with several recent studies reporting much lower bankfull discharge recurrence intervals (Thorne et al., 1997) (Table 4-12). One objective of the present work is to estimate the recurrence interval associ ated with the bankfull di scharge in peninsular Florida streams. Annual peak flow data for the gaged sites were thus an alyzed to determine the return intervals associated with the discharges an d stages associated with top of bank and various bankfull indicators using Log Pearson Type III di stributions (skew coefficient of -0.1) in RIVERMorph (USGS, 1982). Discharges and stages associated with th e following set return intervals (in years) were al so determined for each gaged site using RIVERMorph: 1.0101, 1.25, 1.5, 2, 5, 10, 25, 50, and 100. All determined discharges and stages were plotted graphically onto each gaged sites stage-Q rating curve (Appendix D). Long-term continuous discharge data were used to develop flow and stage duration curves for each gaged site. These show the pe rcentage of time a given discharge or stage is equaled or exceeded, by representing the cumulative frequency of daily mean discharges or daily mean stages. Flow and stage duration curves were used to determine the percentage of time that discharges and stages associated with top of bank and various bankfull indicators were equaled or exceeded at each gaged site. Discharges and stages associated with both top of bank and

PAGE 56

56 various bankfull indicators were plotted graphically onto the flow and stage duration curves for visual comparison (Appendix D). Lastly, the USGS data were used to analy ze several analytical, nonfield based techniques to determine or confirm bankf ull stage, including: Stage-discharge (stage-Q) rating curves. Th eoretically, the inflect ion point on the rating curve corresponds to the point at which the stream overtops its bank and stage consequently levels off (Figure 2-16). St age-Q rating curves were developed for each gaged site from the long-term record. The infection point on each gaged sites stage-Q rating curve was then visually compared to field bankfull indicators, which were plotted onto the state-Q rating curve (Appendix D). Elevation at which the width-to-depth ratio is minimal (BKF-W/D) (Figure 2-17). Using the survey data for each sites classificati on riffle (which extended into the floodplain), the elevation of the minimum width-to-m ean depth ratio was determined. The corresponding stage and discharge were then plotted graphically onto each gaged sites stage-Q rating curve and compared visually with other bankfull indicators (Appendix D). Flood frequency analysis of available stream ga ge data. Gage station analysis throughout the United States, has shown that bankfull disc harge has an average recurrence interval of 1.5 years, which corresponds to a 66.7% annual exceedance probability (Dunne and Leopold, 1978; Leopold, 1994) (Figure 2-15). Peak flow data for the gaged sites were analyzed to determine the di scharge and stage th at occurs on average every 1.5 years. The corresponding stage and discharge were th en plotted graphically onto each gaged sites stage-Q rating curve and compared visually with other bankfull indicators (Appendix D). Historical cross-sectional channel geometry data colle cted during routine USGS streamflow measurements. Stage measurements were plotted against width measurements (stage-w graph), as one would expect width to rapidly increase with small changes in stage as the stream overtops its banks. The stage of various bankfull indicators (BKF-F and BKF-I), as well as the stage of the 1.5 year flood, were plotted onto each gaged sites stage-w graph for visual comparison (Appendix E). Results The results of the study are presented below, beginning with a descript ion of the selected study sites and followed by both resu lts of the reference reach surveys and data analyses conducted on both field data co llected during the reference reach surveys and long-term hydrologic data obtained from the USGS.

PAGE 57

57 Site Selection Forty-five peninsular Florida streams were su rveyed, ranging in drai nage area from 0.2 to 311 square miles and in valley slope from 44 to 5,000 feet/feet. Seve nteen sites are or historically have been gaged by the USGS, while 28 sites are ungaged. Twenty-five sites drain a flatwoods physiography (genera lly with abundant wetlands, poor ly-drained D-type soils, high water tables, flat topography, and many streams) while 20 sites drain a highlands physiography (generally with abundant lakes, relict sand dunes, well-drained A-type so ils, low water tables, rolling topography, and few streams) Nineteen sites are located in the northern portion of the peninsula (above the 28.5 degrees no rth latitude line), while 26 ar e in the southern portion of the peninsula (below the 28.5 degrees north latitude line). Twen ty-three had a wetland floodplain (dominated by hydrophytic vegetation and hydric soils), and twenty-two had an upland floodplain (dominated by hydrophobic vegetation a nd non-hydric soils). Of the twenty-three sites with a wetland floodplain, 11 were dominated by cypress (Taxodium spp .). The sites were classified by physiography, geography, and fl oodplain types to determine what, if any, differences exist among and between various stream sets. Table 11 lists the sites and pertinent details such as location (county, la titude/longitude), reference numb er (if gaged), drainage area, valley slope, physiography, geogr aphy, and floodplain type. Sites are located on both private and publicly owned lands in the following counties: Alachua, Bradford, Clay, DeSoto, Flagler, Glades, Hardee, Highlands, Hillsborough, Lake, Levy, Manatee, Marion, Osceola, Pasco, Polk, Pu tnam, Volusia, Sarasota, and St. Johns counties. Figure 1-2 provides a ma p of the study site locations. Reference Reach Surveys The following bankfull indicators were surveyed during reference reach surveys: position on the bank where slope first becom es level (BKF-F ), inflection or break in slope of the bank

PAGE 58

58 (BKF-I), top of point bar (BKF-TOPB), top of scour or undercuts in the bank (BKF-S), bottom of moss collars (BKF-M), and the alluvial break (BKF-A). BKF-F was present at 87% of the sites, BKF-I at 100%, BKF-TOPB at 13%, BKFS at 84%, BKF-M at18%, and BKF-A at 78% of the sites (Table 3-2). Detailed cross-sections that depict the locations of the various bankfull indicators at each site are found in Appendix B. Because of the low number of sites exhibiting BKF-TOPB and BKF-M indicators these two bankfull indicators were excluded from further analyses. In general, bankfull indicators were locat ed in the following order along the bank: BKF-F (highest in elevation), BKF-I, BKF-S, and BKF-A (lowest in elevation) (F igure 3-1). In streams with a wetland floodplain, the BKF-F indicator appeared to be correl ated strongly with the top of bank, while in streams without a wetland floodplain (which were often incised), BKF-F was often absent. In streams with flowing water on the day of the survey, the BKF-S and BKF-A indicators appeared closely asso ciated with water surface elevati on. It was often difficult to find a distinct alluvial break (BKF-A ) as the stream bed and the stream banks at most of the sites were both composed of sand. For streams with high banks (i.e., Manatee River near Myakka Head, Horse Creek near Arcadia, and Livingston Cr eek near Frostproof), there were often two sets of inflection points (BKF-I), a high and a low, as well as two sets of scour lines (BKF-S), also a high and a low. The lower sets of thes e indicators were used in the data analysis. Data Analysis Field data collected during the reference reach surveys and long-term hydrologic data obtained from the USGS were analyzed to determ ine the following: 1) how closely slope of each bankfull indicator matches that of the water, and 2) how frequently/w hat percentage of time discharge and stage associated w ith each bankfull indicator occur.

PAGE 59

59 Slopes of field indicators Water slopes ranged from -0.026% at Blackwate r Creek near Cassia to 1.610% at Gold Head Branch, with an average slope of 0.219% ( 0.336%) and a m edian slope of 0.097%. Channel bed slopes ranged from -0.349 to 16.100%, with an average slope of 0.605% ( 2.393%) and a median slope of 0.164%. Top of bank (TOB) slopes ranged from -0.227 to 1.796%, with an average slope of 0.346% ( 0.48 2%) and a median slope of 0.176%. BKF-F slopes ranged from -0.325 to 1.607%, with an average slope of 0.282% ( 0.443%) and a median slope of 0.122%. BKF-I slopes ranged from 0.268 to 1.518%, with an average slope of 0.300% ( 0.420%) and a median slope of 0.109%. BK F-S slopes ranged from -0.060 to 1.336%, with an average slope of 0.323% ( 0.368%) and a me dian slope of 0.183%. BKF-A slopes ranged from -0.062 to 1.54o%, with an average slope of 0.291% ( 0.369%) and a median slope of 0.121%. (Table 3-3) Appendix B provides the long itudinal profile, which includes slopes of the various bankfull indicators, for each study site. A surprising number of sites had negative wa ter, bed channel, top of bank, or bankfull indicator slopes, meaning that the best fit line through the surveyed points sloped in an upstream direction rather than in a downs tream direction as one would exp ect. More specifically, 7% of sites had a negative water slope, 20% had a nega tive channel bed slope, 22% had a negative top of bank slope, 21% of sites exhibiting the BKF-F indicator had a negative BKF-F slope, 16% of sites exhibiting the BKF-I indicator had a negative BKF-I slope, 16% of sites exhibiting the BKF-S indicator had a negativ e BKF-S slope, and 17% of th e sites exhibiting the BKF-A indicator had a negative BKF-A slope (Table 3-3). Water slope to bankfull indicator slope ratios were calculated to analyze reliability of various bankfull indicators, as ratios close to one suggest that bankfu ll indicator slope runs parallel to water slope over the surveyed reach. Water slope to bankfull indicator ratios ranged

PAGE 60

60 from -2.86 to 7.58 (mean ratio of 0.65 1.36) for the TOB indicator, from -15 to 3.46 (mean ratio 0.24 2.82) for the BKF-F indicator, from -8.00 to 14.57 (mean ratio 1.01 3.01) for the BKF-I indicator, from -7.43 to 3.82 (mean ratio 0.24 2.20) for the BKF-S indicator, and from 9.36 to 3.69 (mean ratio 0.35 2.24) for th e BKF-A indicator. (Table 3-3) When water slope to bankfull indicator slope ratios were plotted against water slope, a distinct break was seen at a water slope of approximately 0.5% for all bankfull indicators (Figures 3-2A-E). The variability of water slope to bankfull indicator slope ratios among sites with a water slope less than 0.5% (less than a 6-inch rise over 100-foot run) appeared to be much greater than that among sites with a water slope gr eater than 0.5% (more than a 6-inch rise over a 100-foot run) for both top of bank and all bankf ull indicators except BKF-I. Assuming unequal variances, t-tests showed that water slope to bank full indicator slope ratios between sites with a water slope greater than 0.5% and sites with a water slope less than 0.5% were indeed significantly different for all bankf ull indicators except BKF-I (Table 3-4). Further, sites with water slopes greater than 0.5% were more likely to have bank full indicator slopes within 25% of the water slope (or a water slope to bankfull i ndicator slope ratio between 0.75 and 1.25). More specifically, for sites with a water slope greater than 0.5%, 75% of sites had a TOB slope within 25% of the water slope, 75% exhibiting the BKF-F indicator had a BKF-F slope within 25% of the water slope, 88% exhibiting the BKF-I indica tor had a BKF-I slope within 25% of the water slope, 88% exhibiting the BKF-S indicator had a BKF-S slope within 25% of the water slope, and 71% exhibiting the BKF-A i ndicator had a BKF-A slope within 25% of the water slope (Table 3-4). In comparison, for sites with a wa ter slope less than 0.5%, only 17% of sites had a TOB slope within 25% of the wa ter slope, 18% of the sites exhi biting the BKF-F indicator had a slope within 25% of the water sl ope, 19% of the sites exhibiting the BKF-I indicator had a BKF-I

PAGE 61

61 slope within 25% of the water sl ope, 23% of the sites exhibiting the BKF-S indicator had a BKFS slope within 25% of the water slope, and 29% of the sites e xhibiting the BKF-A indicator had a BKF-A slope within 25% of the water slope (Tab le 3-4). Additionally, no sites with a water slope greater than 0.5% had negative bankf ull indicator slopes (Table 3-4). Gage analysis Drainage areas for gaged sites ranged from 0.32 square miles (sq mi) at Shiloh Run near Alachua to 311 sq mi at Fisheating Creek at Palmdale, with mean and median values of 75.9 sq mi and 52.8 sq mi, respectively (Table 3-1). Me an annual discharges ranged from 0.29 cubic feet per second (cfs) at Shiloh Run near Alachua to 256 cfs at Fisheating Cr eek at Palmdale, with mean and median values of 59 cfs and 42 cfs, respectively (Table 4-1). The discharges and stages associated with the top of bank (TOB) and various bankfull indicators (BKF-F, BKF-I, BKF-S, BKF-A), as well as their associated retu rn intervals and duration (or percentage of time equaled or exceeded), are provided in Tabl es 3-5 and 3-6 and are detailed below2. The discharges and stages associated with various set return intervals (1.0101-, 1.25-, 1.5-, 2-, 5-, 10-, 25-, 50-, and 100-year), as well as their durations are provided in Tabl es 3-7 and 3-8. The 1.5year event is further detailed be low, because it is the recurrence interval often cited with the bankfull event. QTOB ranged from 25 cubic feet per second (cfs) to 595 cfs, with mean and median values of 156 cfs and 90 cfs, respectively. The return interval associated with QTOB ranged from less than one year to 3.10 years. The percentage of time that QTOB was equaled or exceeded ranged from 0.21% to 41% of the time (or from 0.75 to 150 days per year), with mean and median values of 15% and 8.3% of the time (or 56 and 30 days per year), respectively (Table 3-5). The top of bank stage ranged from 0.36 feet above mean annual stage to 7.53 feet above mean annual stage, wi th mean and median values of 2.13 feet and 2 Note that the results from Hi ckory Creek near Ona, Lochloosa Creek at Grove Park, Moses Creek near Moultrie, and Shiloh Run ne ar Alachua were excluded from the summary statistics, as their period of reco rd was insufficient (less than ten years) for proper peak flo w analysis and/or flow duration curve development. However, rough estimates of the return intervals a nd the durations associated with the top of bank and with the various bankfull indicators can be found in Tables 3-5 and 3-6 a nd the stage-Q rating curves can be found in Appendix D.

PAGE 62

62 1.42 feet above mean annual stage, respectively. The return interval associated with the top of bank stage ranged from less than one year to 1.98 years. The percentage of time that the top of bank stage was equaled or exceeded ranged from 0.06% to 43% of the time (or from 0.24 to 156 days per year), with mean and median values of 15% and 13% of the time (or 55 and 49 days per year), respectively (Table 3-6). TOB durations were not significantly different between stage and discharge measurements, but durations were significantly higher in streams with a wetland floodplain th an in streams with an upland floodplain for both discharge (p<0.01) and stage (p<0.01) (Table 3-9). QBKF-F ranged from 25 cubic feet per second (cfs) to 402 cfs, with mean and median values of 111 cfs and 67 cfs, respectively. The return interval associated with QBKF-F ranged from less than one year to 1.12 years. The percentage of time that QBKF-F was equaled or exceeded ranged from 4.0% to 50% of the tim e (or from 15 to 181 days per year), with mean and median values of 26% and 24% of the time (or 94 and 87 days per year), respectively (Table 3-5). The BKF-F stage ranged from 0.42 feet below mean annual stage to 5.90 feet above mean annual stage, with mean and median values of 1.22 feet and 0.47 feet above mean annual stage, respectively. Th e return interval associated with the BKF-F stage ranged from less than one year to 1.13 year s. The percentage of time that the BKF-F stage was equaled or exceeded ranged from 3.4% to 78% of the time (or from 12 to 283 days per year), with mean and median valu es of 28% and 24% of the time (or 101 and 89 days per year), respectively (T able 3-6). BKF-F durations we re not significantly different between stage and discharge measurements or between sites with a wetland floodplain and those with an upland fl oodplain (Table 3-9). QBKF-I ranged from 18 cfs to 118 cfs, with mean and median values of 64 cfs and 56 cfs, respectively. The return in terval associated with QBKF-I ranged from less than one year to 3.70 years. The percentage of time that QBKF-I was equaled or exceeded ranged from 0.77% to 50% of the time (or from 2.8 to 184 days per year), with mean and median values of 25% and 18% of the time (or 89 and 66 days per year), respectively (Table 3-5). The BKF-I stage ranged from 0.46 feet below m ean annual stage to 2.39 feet above mean annual stage, with mean and median values of 0.76 feet and 0.64 feet above mean annual stage, respectively. The return interval asso ciated with the BKF-I stage ranged from less than one year to 1.50 years. The percentage of time that the BKF-I stage was equaled or exceeded ranged from 0.82% to 52% of the time (or from 3.0 to 191 days per year), with mean and median values of 24% and 25% of the time (or 88 and 91 days per year), respectively (Table 3-6). BKF-I durations we re not significantly different between stage and discharge measurements, but BKF-I durations were found to be significantly higher in streams with a wetland floodplain than in streams with an upland floodplain for both discharge (p<0.01) and stage (0.03) (Table 3-9). QBKF-S ranged from 9.9 cfs to 75 cfs, with mean and median values of 32 cfs and 29 cfs, respectively. The return in terval associated with QBKF-S ranged from less than one year to 1.95 years. The percentage of time that QBKF-S discharge was equaled or exceeded ranged from 6.6% to 71% of the time (or from 24 to 260 days per year), with mean and median values of 43% and 51% of the time (or 157 and 186 days per year), respectively (Table 35). The BKF-S stage ranged from 1.18 feet below mean annual stag e to 0.74 feet above mean annual stage, with mean and median values of 0.25 feet and 0.22 feet below mean

PAGE 63

63 annual stage, respectively. The return interv al associated with the BKF-S stage ranged from less than one year to 1.20 years. The percentage of time that the BKF-S stage was equaled or exceeded ranged from 8.5% to 70% of the time (or from 31 to 257 days per year), with mean and median values of 45% and 48% of the time (or 164 and 175 days per year), respectively (Table 3-6). BKF-S durati ons were not significan tly different between stage and discharge measurements or between sites with a wetland floodplain and those with an upland floodplain. QBKF-A ranged from 5.0 cfs to 38 cfs, with mean and median values of 16 cfs and 13 cfs, respectively. The return in terval associated with QBKF-A ranged from less than one year to 1.08 years. The percentage of time that QBKF-A was equaled or exceeded ranged from 14% to 93% of the time (or from 51 to 338 days per year), with mean and median values of 60% and 57% of the time (or 219 and 208 days per ye ar), respectively (Table 3-5). The BKF-A stage ranged from 1.90 feet below mean annual stage to 0.31 feet above mean annual stage, with mean and median values of 0.69 feet and 0.79 feet below mean annual stage, respectively. The return interv al associated with the BKF-A stage was less than one year. The percentage of time that the BKF-A stage was equaled or exceeded ranged from 18% to 96% of the time (or from 64 to 350 days per year), with mean and median values of 62% and 67% of the time (or 226 and 246 days per year), respectively (Table 3-6). BKF-A durations were not significantly different be tween stage and discharge measurements or between sites with a wetland floodplain and those with an upland floodplain. Q1.5 is a flow event with a 1.5-year return inte rval that has a 66.7% pr obability of occurring in a given year. Q1.5 ranged from 60 cfs to 1,934 cfs, with mean and median values of 523 cfs and 288 cfs, respectively. The percentage of time that Q1.5 was equaled or exceeded ranged from 0.18% to 17% of the time (or from 0.65 to 63 days per year), with mean and median values of 4.0% and 2.3% of the time (or 15 and 8.2 days per year), respectively (Table 3-7). The stage associated with the 1.5 -year return interval ranged from 0.64 feet to 9.66 feet above mean annual stage, with mean and median values of 3.71 feet and 3.18 feet above mean annual stage, respectively. The perc entage of time that the stage associated with the 1.5-year return interval was equale d or exceeded ranged from 0.15% to 12% of the time (or from 0.55 to 45 days per year), with mean and median values of 3.8% and 3.0% of the time (or 14 and 11 days per year), respectively (Table 3-8). All previously mentioned discha rge and stage values were pl otted onto the stage-Q rating curves developed for each gaged site so that th e top of bank and the various bankfull indicators could be compared visually to the set return inte rvals (Appendix D), resul ting in the following: Top of bank points plotted below the 1.0101-year re turn interval points at 29% of the sites, between the 1.0101-year and 1.25-year return interv al points at 35% of the sites, between the 1.25-year and 1.5-year return interval points at 24% of the sites, between the 1.5-year and 2-year interval points at 0% of the sites, and between the 2-year and 5-year interval points at 12% of the sites (Cat fish Creek near Lake Wales an d Shiloh Run near Alachua).

PAGE 64

64 BKF-F points plotted below the 1.0101-year retu rn interval points at 38% of the sites exhibiting the BKF-F indicator, between the 1.0101-year and 1.25-year return interval points at 65% of the sites, and above the 1.25-year interval poin ts at 0% of the sites. BKF-I points plotted below the 1.0101-year return interval points at 59% of the sites exhibiting the BKF-I indicator, between the 1.0101-year and 1.25-year return interval points at 29% of the sites, between the 1.25-year and 1.5-year return interval points at 0% of the sites, between the 1.5-year and 2-year return interval po ints at 6% of the sites, and between the 2-year and 5-year re turn interval points at 6% of the sites (Catfish Creek near Lake Wales). BKF-S points plotted below the 1.0101-year retu rn interval points at 84% of the sites exhibiting the BKF-S indicator, between the 1.0101-year and 1.25-year return interval points at 8% of the sites, betw een the 1.25-year and 1.5-year retu rn interval points at 0% of the sites, between the 1.5-year and 2-year return interval poi nts at 8% of the sites, and above the 2-year return interval points at 0% of the sites. BKF-A points plotted below the 1.0101-year return interval at 93% of the sites exhibiting the BKF-A indicator, between the 1.0101-year an d 1.25-year return inte rval points at 7% of the sites, and above the 1.25-year re turn interval at 0% of the sites. The USGS gage data were also used to analyze several analytical, non-field based techniques of determining or confirming th e bankfull stage, resulting in the following: The inflection point of the Stage-Q rating curv es was found at a point on the stage-Q rating curve well above the field-based bankfull i ndicators at many of the sites (Appendix D), suggesting that bankfull flow occurs more fr equently than the flow at which the stage levels out on the stage-Q rating curve. However, due to the variation found in stage-Q relationships, this method was di fficult and likely unreliable. The elevation and associated discharge at which the width-to-depth ratio was at a minimum (BKF-W/D) at the classification riffle were determined for each gaged site (Tables 3-5 and 3-6). QBKF-W/D ranged from 8.3 cfs to 381 cf s. The return interval associated with QBKF-W/D ranged from less than one year to 1.65 years, with mean and median values of 1.10 years and 1.02 years, re spectively. The percentage of time that QBKF-W/D was equaled or exceeded ranged from 0.33% to 82% of the time (or from 1.2 to 299 days per year), with mean and median va lues of 29% and 27% of the time (or 106 and 97 days per year), respectively (Table 3-5) The BKF-W/D stage ranged from 1.45 feet below mean annual stage to 5.40 feet above m ean annual stage, with mean and median values of 1.20 feet and 0.28 feet above mean annual stage, respectively. The return interval associated with the BKF-W/D stage ranged from less than one year to 1.65 years. The percentage of time that the BKF-W/D stage was equaled or exceeded ranged from 0.31% to 85% of the time (or from 1.1 to 310 da ys per year) with mean and median values of 32% and 30% of the time (or 116 and 108 days per year), respectively (Table 3-6). When plotted on the stage-Q rating curve, BKF-W/D points plotted below the 1.0101-year return interval points at 41% of the sites, between the 1.0101-year and 1.25-year return

PAGE 65

65 interval points at 53 % of the sites, between the 1.25-y ear and 1.5-year return interval points at 0% of the sites, betw een the 1.5-year and 2-year return interval points at 6% of the sites, and above the 2-year interval point s at 0% of the sites (Appendix D). BKF-W/D durations were not significantly different be tween stage and discharge measurements or between sites with a wetland floodplain and those without a wetland floodplain. Most interestingly, though not surprising, BKF-W/D plotted between the BKF-F and BKF-I field indicators on the stage-Q rating curve at many of the sites. Although BKF-W/D is not an indicator found in the field, it is important to note that its de termination does require field survey data. Because gage station analysis throughout th e United States has shown that bankfull discharge has an average recurrence interv al of 1.5 years (Dunne and Leopold, 1978; Leopold, 1994), discharges associated with the 1.5-year return interval were determined and plotted onto each gaged sites stage-Q ra ting curve (Appendix D). As previously mentioned, the majority of bankfull indicator s (93%) plotted below the 1.5-year return interval on the stage-Q curve, suggesting that the bankfull event in peninsular Florida streams occurs more frequently than elsewhere in the United States. Historical cross-sectional channel geometry data colle cted during routine USGS streamflow measurements were used to plot stage against width. For non-incised streams, two distinct clusters were obs erved (an in-the-banks cluste r and an out-of-the-banks cluster), separated by a large increase in width. This occurs because as the stream overtops its banks, its width increases ra pidly with only small increases in stage; however, the water eventually reaches an upland te rrace that confines the lateral extent (or width). When the BKF-F and BKF-I stages were plotted onto this graph, they generally corresponded well with the stage at which the jump in width o ccurs, while when the 1.5-year return interval stage was plotted onto the graph, it generally plotted well above the jump in width, again confirming that bankfull flow in peninsular Fl orida streams occurs more frequently than 1.5 years (Figure 3-3A, Appendix E). For incise d streams, there were no distinct clusters because in incised streams the river valley is largely confined and width thus increases gradually as stage increases. When the BKF-F, BKF-I, and 1.5year return interval stages were plotted onto these graphs, no real distin ctions could be made (Figure 3-3B, Appendix E). For non-incised streams, plotting width against stage can be a good method for determination or confirmation of bankfull stage, while for incised streams it is not as useful. Discussion In this study, various indicator s of bankfull stage were iden tified, surveyed, and analyzed individu ally to determine if there is a single most reliable bankfull indicator for peninsular Florida streams. The following factors were ex amined: how prevalent each bankfull indicator is among study sites; how closely the slope of each bankfull indicator matches that of the water;

PAGE 66

66 and how frequently and for how long discharge a nd stage associated with each bankfull indicator occur. The discussion begins with an examina tion of the potential source s of error involved in conducting the reference reach surveys and implications this could have on interpretation of data. The discussion continues with an examination of analyses conducted on field data collected both during reference reach surveys and long-term hydrologic data obtained from the USGS. Observed trends and anomalies for each data se t are discussed and potential explanations are presented. Interpretation of data is presented as it relates to achieving the objective of this chapter, which is to determine the most reliable bankf ull indicator for peninsular Florida streams. Reference Reach Surveys Common sources of error associ ated with surveying, such as those in transcribing data and in entering data into the com puter, were mi nimized by using a total st ation, which records all the survey points digitally. Rod height read ings were taken carefu lly and double-checked if results were questionable. Extr a care was taken in establishing turning points. When survey data were downloaded into RIVERMorph, they were analyzed for surveying errors, then corrected in Excel. Corrections were highlighted and explaine d in the notes section of each study sites spreadsheet so future users of the raw survey data may be aware of any survey errors. Another potential source of error associated with referen ce reach surveys is the incorrect identification or surveying of ba nkfull stage. As previously described, bankfull stage is the elevation at which the stream just begins to overflow onto its floodplain, wh ich is defined as the relatively flat, depositional surface adjacent to the stream that is being built and rebuilt by the stream in the present hydrologic regime (Emmett, 2004). Field iden tification of bankfull stage is the method most often used to estimate th e channel-forming flow, though its correct identification in the field can be difficult and subjective (Johnson and Teil, 1996; Knighton, 1998). In this study, various indicators of bankfull stage (TOB, BKF-F, BKF-I, BKF-TOPB,

PAGE 67

67 BKF-S, BKF-M, and BKF-A) were surveyed at six cross-sections along a longitudinal profile. Because various indicators of bankfull stage were identified, surveyed, and analyzed separately and consistently, the pot entially subjective nature of choosi ng the bankfull stage was minimized for the most part, with the exception of BKF-A (as expl ained below). Methods of bankfull indicator identifica tion were consistent throughout the study; however, several factors may have le d to the inaccurate reading of a particular bankfull indicator. For example, the alluvial break (BKF-A) was partic ularly difficult to identify as the stream bed and stream banks at all sites were both pre dominantly composed of sand and therefore of uniform particle size. In larg er rivers, such as the Manatee River near Myakka Head and the Santa Fe River near Graham, several distinct breaks in slope (BKF-I) and scour lines (BKF-S) were found. Though all inflections and scour lines were surveyed, only the lowest of each were used in data analysis. Furthe r complicating identification of th e active floodplain is Floridas recent drought conditions, which can lead to floodplain vegetation growing clearly within the channel. Regardless of drought, some floodplain tree species, such as cypress ( Taxodium spp.), can actually grow in the middle of the channel and should be ignor ed when attempting to identify the active floodplain in peninsular Fl orida. Inaccurate readings c ould also be due to the rod not being placed exactly on the bankfull indicator, th e rod sinking into the mud, surveying a relict indicator, surveying a root rather than an actual bank in flection, or survey ing a local deposit resulting from local velocity controls such as vegetation. Inaccu rate readings may affect slope (which will be discusse d in further detail in the following section), width, and depth of the bankfull indicators and ultimately calculation of bankfull discharge. Based on reference reach surveys, several c onclusions can be drawn regarding the most reliable bankfull indicator for peni nsular Florida. Break in slope (BKF-I) appears to be the most

PAGE 68

68 consistent bankfull indicator, as it was found at all of the sites surveye d. The flat floodplain (BKF-F) was a consistent indicator for sites with a relatively flat wetland floodplain (Table 3-2). The scour line (BKF-S) was consistent at most site s, but was generally abse nt at sites dominated by a cypress ( Taxodium spp.) floodplain perhaps due to the pres ence of cypress knees or the low gradient nature of these systems not generating enough stream power to produce a scour line. As previously mentioned, the alluvial break (BKF-A ) was difficult to identify and is thus not a reliable indicator for peninsul ar Florida streams. Furthe rmore, BKF-A and BKF-S were generally located at a lower elevation on the cr oss-section than BKF-I and BKF-F, and they appeared to be more closely associated with the water level on the day of the survey (for those sites with water). Because surveying was c onducted during the dry season, the present water level on the day of the survey would not be expected to be flowing at bankfull stage, and thus these two indicators are likely not the best interp retation of the bankfull st age. Based solely on prevalence and elevation of various bankfull indi cators during reference reach surveys, BKF-I and BKF-F (for streams with relatively flat we tland floodplains) appear to be the most reliable field indicators of bankfull stage for peninsular Florida streams. Data Analysis Slopes of field indicators Slopes of a line best fit through surv ey point s of each individual bankfull indicator (BKFF, BKF-I, BKF-S, BKF-A) and through the top of bank survey points (TOB) were compared to the slope of a line best fit through the water su rface survey points (or the channel bed surface points for those sites that had no flowing water on the day of the survey). Leopold (1994) used this technique to verify the feature as bankfu ll if the two lines were generally parallel and consistent over a long reach. To determine how parallel the lines were, water slope was divided by the slope of each bankfull indicator to determ ine a water slope to ba nkfull indicator slope

PAGE 69

69 ratio. Theoretically, the closer the ratio is to one, the more reliable the indicator. Bankfull indicator slopes within 25% of the water slope, or those with a water slope to bankfull indicator ratio between 0.75 and 1.25, were deemed candidate reliable field indicato rs (Table 3-3). Slopes analysis results show that: 1) variabil ity of water slope to bankfull indicator slope ratios among sites with a water slope less than 0.5% was significantly greater than that among sites with a water slope greater than 0.5% for all indicators exce pt for BKF-I (Table 3-4, Figures 3-2); and 2) sites with a wate r slope greater than 0.5% were more likely to have bankfull indicator slopes within 25% of the water slope (Table 3-4). This suggests that slope-area techniques for calculating the ba nkfull discharge should not be us ed in peninsular Florida for sites with a water slope less than 0.5%, or vice versa, that calc ulating discharge using slope-area techniques is acceptable for sites with a water sl ope greater than 0.5%. Bankfull indicators may be more reliable for streams w ith a water slope greater than 0.5% because the steeper slope can generate more stream power and consequently perhaps the stream can build more consistent morphological features. There may be several explanations why bankfull indicator slopes were unreliable at many of the sites (i.e., were not with in 25% of the water slope or ev en had a negative slope/reverse gradient signature). First, ther e is a certain amount of inherent vertical variability in natural stream systems. A few inches of variability, however, can make a big difference when determining slopes for peninsular Floridas low-gradient stream systems. These low-gradient streams also leave little room for surveying errors that can occur from in correctly identifying or surveying a particular bankfull indicator (as men tioned in the previous section). If a stream drops only a couple of inches in elevation over an entire reach, then any surveying errors may lead to an inaccurate bankfull slope. Solutions to these issues may be to survey a longer

PAGE 70

70 reference reach or to survey more points along the reach to make up for any potential surveying errors. Another solution may be to remove variab ility in bankfull indicator slope before actually surveying. This can be done by picking the best indicator at each crosssection along the reach and making sure that it is within a fixed, small am ount of variability of th e water, rather than by surveying a variety of bankfull indicators at each cross-section, then determining slopes of each indicator individually. This method was tested at Morgan Hole Creek, a site where every bankfull indicators slope was negative. A lthough a more reliable bankfull slope was determined, this method seems to make bankfull stage determination more of an art than a science. Second, the slope of the water en countered on the day of the surv ey may not be an accurate representation of the water slope at bankfull, which would render basing reliability of a bankfull indicator on water slope to bankf ull indicator slope ratio useless. For example, two sites (Blackwater Creek near Cassia and Cow Creek) actually had negative water slopes on the day of the survey. These sites were extremely lowgradient, cypress-dominated systems with muddy streambeds, which may have led to inaccurate pres ent water level readings due to the survey rod sinking into the mud. Solutions to this issue may be to survey wa ter slope when the water is at or near bankfull stage. Third, some sites did not have flowing water on the day of the survey so channel bed slope was used in place of the water slope to calculate water slope to bankfull indicator slope ratio. In these cases, the location of the surveys endpoint s could have significant effects on the resulting channel bed slope and consequently the water slope to bankfull indicator slope ratio. For example, if one endpoint is at a pool and one is at a riffle, this could significantly affect overall slope and could even produce a negative slope. Th e solution to this issue is to be sure to begin

PAGE 71

71 and end the reference reach survey at a riffle. Another solution is to us e another surrogate for water slope when there is no flow ing water on the day of the surve y, such as valley slope divided by sinuosity. Lastly, perhaps peninsular Floridas unique climate, geology, and vegetation prevent its streams from fitting neatly within the concepts of bankfull that were developed in higher gradient piedmont and montane river systems. For exam ple, in peninsular Fl orida, cyclonic storms (versus frontal low pressure systems) lead to pa tchy distributions of inte nse rainfall. Mid-order to high-order streams have a gr eater chance of rainfa ll variation along their lengths than do loworder headwater streams since their drainage areas are larger. This rainfa ll variation may affect water surface profiles, particularly if the dow nstream portion receives more rain and creates backwater effects. In other words, peninsular Florida streams likely do not exhibit a one to one ratio of rain to discharge as do other places in the United States, which may be why the bankfull indicators at many of the site s are smeared. Other hypotheses for reverse gradient bankfull signatures include: backwater effect s (due to Floridas deranged network of wetlands and lakes), drought effects, animal effects (i.e., hogs), ve getative control, or bottom-up wetting (water infiltrating through Floridas sandy soils and enteri ng the stream as groundwater, versus overland flow, and creating a gross movement of water that causes the cha nnel to cut uphill). Based on slope analysis, several conclusions can be drawn regarding the most reliable bankfull indicator for peninsular Florida. Wh en comparing water slop e to bankfull indicator slope ratios, BKF-I was the most reliable bankfull indicator, w ith an average ratio of 1.01. Further, variance in water slope to BKF-I slope ratio between streams with water slope less than 0.5% and those with a water slope greater than 0.5% was not significan tly different (p>0.05) (Table 3-4). Perhaps more importantly, however, slopes analysis suggests that there is a water

PAGE 72

72 slope threshold of approximately 0.5%, above which bankfull indicators become more reliable. It is important to note, however, that the popul ation of streams with water slopes greater than 0.5% was rather small (n=8), thus additional re search is recommended. Findings further suggest that slope-area techniques for calcu lating the bankfull discharge shoul d not be used in peninsular Florida for sites with a water slope less than 0.5% or conversely, that calculating discharge using slope-area techniques is acceptable for sites w ith a water slope greater than 0.5%. Gage analysis Sites with long-term hydrologi c data obtained from the USGS were analyzed to determine frequency and duration of stage and discharge associated with various bankfull indicators. As previously discussed, bankfull stage of each indicator was determined by adding the average difference between elevation of the bankf ull indicator and that of the water surface at the time of the survey to the stage recorded by the USGS on the day of the survey. The most current stage-Q rating table was then used to de termine bankfull discharges associated with the determined bankfull stages. Therefore, any is sues associated with USGS data could have significant effects on bankfull disc harge determination at the gaged sites. An important issue discovered upon analysis of USGS data was the extreme variability found in stage-Q relationships. For example, at Catfish Creek n ear Lake Wales (1947 to present), variation in stage was as much as 1.2 feet fo r a given discharge of 50 cfs and that in discharge was as much as 65 cfs for a given stage of 4.00 f eet (Figure 3-4A). There also a ppeared to be several distinct rating curves within the data. To help discern the data, the daily discharge and stage measurements were separated by de cade (Figure 3-4B). This ex ercise confirmed that stage-Q rating curves for peninsular Florida streams can change over time, sometimes quite drastically. Further, Figure 3-5 provides a visual comparis on of variation within the long-term stage and discharge measurements among gage d sites through use of boxplots.

PAGE 73

73 There may be several explanations for variat ion seen in the stage-Q relationships of peninsular Florida streams. First, USGS only directly measures discharge six to 12 times per year at a cross-section where the velocity can be measured most accurately; therefore, discharge measurements may not always be taken at the same location. Additionally, channel controls such as sand bars, topography, vegetation, and large woody debris can also have a significant effect on discharge measurements. For exampl e, a single large stor m can input large woody debris or cause channel bed adjustments (ma ny of peninsular Floridas streams are sandbottomed and can thus adjust relatively quickly), which can significantly affect discharge. Floridas deranged stream networks of wetlands and lakes may also affect discharge by creating backwater effects. Higher gradient streams system s, however, may be less affected by backwater so their stage-Q relationships may be less va riable, which may explain why their bankfull indicators tend to be more reliable (as discussed in the previous section). Because discharge can be so variable, stage may be a more useful parameter for understanding the concept of bankfull in peninsular Florida streams. Stage may also be more useful because it is the parameter that the USGS actually measures. However, this st udy did not find any signi ficant differences in durations between discharge and stage measuremen ts associated with th e top of bank and with each bankfull indicator (Table 3-9). Gage station analysis throughout the United St ates has shown that bankfull discharge has an average recurrence interval of 1.5 years, which corresponds to a 66.7% annual exceedance probability (Dunne and Leopold, 1978; Leopold, 1994) (Figure 2-15). Frequency analyses of gaged sites found that stage and discharge associat ed with top of the bank and bankfull indicators occurred more frequently than 1.5 years. Freque ncy analyses of gaged sites found that the stage and discharge associated with BK F-A occurred the most frequently on average, while stage and

PAGE 74

74 discharge associated with top of bank and BKF-F generally occurred least frequently (Tables 3-5 and 3-6). Duration analyses found that stage and discharge associated with BKF-A were exceeded the most, while the stage and discharge associated with the t op of bank and with BKFF were generally exceeded the least. This in tuitively makes sense based on observations made during the reference reach survey that BKF-A ge nerally occurred the lowest in elevation on the cross-section, while BKF-F and top of bank were generally highest in elevation (Figure 3-1). Based on gage analysis, it is safe to concl ude that both BKF-A and BKF-S occur far too frequently and are exceeded far too often to be considered the best in dicator of the bankfull discharge, or the most effective discharge in transporting sediment and performing work. Significant differences were then found in durations of discharges and stages associated with top of bank (p<0.01) and the BKF-I indicator (p<0.0 1) between sites with a wetland floodplain and those without a wetland floodplain (Table 3-9). However, significant differences were not found in the durations of discharges and stages asso ciated with BKF-F between sites with a wetland floodplain and those without a wetland floodplain. This is likely due to the nature of the BKF-F indicator itselfa flat floodplain, which is generally found at sites with a wetland floodplain and is generally absent from sites without one as thes e sites are more likely to be incised. Because BKF-I and top of bank were found at every site the fact that signifi cant differences exist between sites with a wetland floodplain and sites w ithout one suggests that a different indicator should be used between these two site types. Because BKF-F is generally found at sites with a wetland floodplain, it is the most reliable bankf ull indicator for peninsular streams Florida streams with a wetland floodplain. For streams w ithout a wetland floodplai n, BKF-I is the most reliable bankfull indicator.

PAGE 75

75 Conclusions In this study, various indicator s of bankfull stage were iden tified, surveyed, and analyzed individu ally to determine if there is a single most reliable bankfull indicator for peninsular Florida streams. The following factors were examined: how prevalent each bankfull indicator among sites; how closely slope of each bankfull indicator matches that of the water; and how frequently and for how long the discharge and stage associated with each bankfull indicator occur. Based on these factors, there is not a sing le most reliable bankfull indicator for peninsular Florida streams, but rather, two: 1) BKF-F, or the position on the bank where the slope first becomes level, should be used for streams w ith a wetland floodplain or those with a broad valley; and 2) BKF-I, or the inflection in bank slope of the bank, should be used for streams without a wetland floodplain or those with a conf ined valley. Another important finding of the study is that bankfull indicators ar e more reliable for streams w ith a water slope greater than 0.5%, suggesting that slope-area te chniques for calculating the bankf ull discharge should not be used in peninsular Florida for sites with a water slope less than 0.5%, or vice versa, that calculating discharge using slope -area techniques is acceptable for sites with a water slope greater than 0.5%.

PAGE 76

76Table 3-1. Summary of gaged sites Site name USGS station ID Period of record (WY) Count y LatitudeLongitude Drainage area (sq mi) Date surveyed Reference reach survey location (in relation to gage) Discharge reported on day of survey (cfs) Adjusted stage reported / observed on day of survey (ft) Blackwater Creek near Cassia 0223520081-07Lake28.874-81.490126 3/3/08Ended reach ~1800 feet US of gage280.25 Blues Creek near Gainesville 0232201685 -94Alachua29.728-82.4312.62 1/10/08Ended reach ~1 mile US of gage 0.37 1-0.43 Bowlegs Creek near Fort Meade0229501365-68/92-07Pol k 27.700-81.69547.2 12/3/07Ended reach ~1375 feet US of gage3.70.12 Carter Creek near Sebring 0227000055-67/92-07Highlands27.532-81.38838.8 12/7/07Ended reach ~1.9 miles US of gage3.2-1.01 Catfish Creek near Lake Wales 0226700048-07Pol k 27.961-81.49658.9 9/27/07Began reach at gage 25-0.64 Fisheating Creek at Palmdale 0225650032-07Glades26.933-81.315311 3/20/08Ended reach ~1.64 miles US of gage11-1.55 Hickory Creek near Ona 0229575582-84*Hardee27.482-81.8803.75 8/9/07Began reach ~550 feet DS of gage13.66 30.94 Horse Creek near Arcadia0229731051-07De Soto27.199-81.9882183/17/08Began reach ~345 feet DS of gage5.8-1.92 Little Haw Creek near Seville0224442052-06Flagler29.322-81.385932/29/08Surveyed through gage 9.3-1.12 Livingston Creek near Frostproof0226952092-07Pol k 27.709-81.446120 12/4-5/07Surveyed through gage 17.5 4-0.56 Lochloosa Creek at Grove Park0224190096-05*Alachua29.600-82.1457.4 1/7/08Began reach ~425 feet DS of gage0.05 3-0.40 Manatee River near Myakka Head0229995067-07Manatee27.474-82.21165.3 11/8-9/07Ended reach ~1150 feet US of gage22.5 4-0.17 Moses Creek near Moultrie 0224702700-02*St. Johns29.775-81.3167.4 1/18/08Began reach ~364 DS of gage 1.3 2,3-0.19 Rice Creek near Springside 0224447374-04Putna m 29.688-81.74243.2 1/11/08Ended reach ~420 feet US of gage 10 2-0.50 Santa Fe River near Graham 0232070057-98Bradford29.846-82.22094.91/15-16/08Ended reach ~550 feet US of gage13.6 2,4-1.36 Shiloh Run near Alachua 0232205084-87*Alachua29.819-82.4720.32 1/8/08Ended reach ~75 feet US of gage 0 1,3-Tiger Creek near Babson Park0226839092-07Pol k 27.811-81.44452.8 3/14/08Ended reach ~1.4 miles US of gage28-0.18 Gage Information Reference Reach Survey Information Notes: WY = Water year; HL = Highlands phys iography; FW = Flatwoods physiography; N = Northern peninsula; S = Southern peninsul a; WF = Wetland floodplain; WFC = Wetland floodplain dominated by cypress; UP = Upland floodplain; US = upstream; DS = downstr eam; Adjusted stage = Reported or observed stage Mean annual stage; -= No stage data, 1 No discharge reported for the day of survey (gage inactive) and sta ff gage no longer at site or no longer accurate-estimated dis charge and then determined the associated stage from the stage-Q rating table; 2 No discharge reported for the day of survey (gage inactive)-us ed gage height observed at the staff gage on the day of the su rvey and then determined the associated discharge from the stage-Q rating table; 3 Period of record for continuous data and/or annual peak data is less than 10 years-gage analysis results are rough estimates and were not included in summary statistics; 4 Discharge averaged over two days; Period of record less than 10 years-data insufficient for proper gage analysis

PAGE 77

77 Table 3-2. Prevalence of field bankfull indicators Site Name Physiography Geography Floodplain type Flat floodplain (BKF-F) Inflection (BKF-I) Top of point bar (BKF-TOPB) Scour (BKF-S) Moss (BKF-M) Alluvial break (BKF-A) Alexander Springs Creek tributary 2HLNUPPresentPresentNot presentPresentNot presentPresent Blackwater Creek near Cassia HLNWFCPresentPresentN ot presentNot presentNot presentNot present Blues Creek near Gainesville FWNUPNot pres entPresentNot presentPresentPresentPresent Bowlegs Creek near Ft Meade FWSWFPresentPrese ntNot presentNot presentNot presentPresent Carter Creek near Sebring HLSUPPresentPre sentNot presentPresentNot presentPresent Catfish Creek near Lake Wales HLSWFCPresentPresentNot presentPresentNot presentPresent Coons Bay Branch FWSWFPresentPresentNot presentPresentNot presentPresent Cow Creek FWNWFCPresentPresentNot presentPresentNot presentPresent Cypress Slash tributary HLSUPPresentPrese ntNot presentPresentNot presentPresent East Fork Manatee River tributaryFWSUPPresen tPresentNot presentPresentPresentNot present Fisheating Creek at Palmdale FWSWFCPresentPre sentPresentPresentNot presentNot present Gold Head Branch HLNUPPresentPresentN ot presentPresentNot presentPresent Hammock Branch HLNWFPresentPresentNot presentPresentNot presentPresent Hickory Creek near Ona FWS WFPresentPresentNot presentPresentNot presentPresent Hillsborough River tributary FWSWFCPresentPresentNot presentNot presentNot presentNot present Horse Creek near Arcadia FWSWFPresentPresentNot presentPresentNot presentPresent Jack Creek HLSWFPresentPresentNot presentPresentNot presentPresent Jumping Gully HLNUPNot presentPresent Not presentPresentNot presentPresent Lake June-In-Winter tributary FWSUPNot presen tPresentNot presentPresentNot presentPresent Little Haw Creek near Seville FWNWFCPresentPre sentNot presentNot presentNot presentPresent Livingston Creek near FrostproofHLSUPPresentPresentNot presentPresentNot presentPresent Livingston Creek tributary HLSUPPresentPresentNot presentPresentNot presentNot present Lochloosa Creek at Grove Park FWNWFCPresentPresentNot presentPresentNot presentPresent Lowry Lake tributary HLNUPPresentPresentNot presentPresentNot presentPresent Manatee River near Myakka HeadFWSUPPrese ntPresentPresentPresentNot presentPresent Manatee River tributary FWSUPPresentPrese ntNot presentPresentPresentNot present Morgan Hole Creek FWSUPPresentPresentN ot presentPresentNot presentPresent Moses Creek near Moultrie FWNWFCPresentPresentNot presentPresentNot presentPresent Myakka River tributary 1 FWSUPPresentPresentNo t presentNot presentNot presentNot present Myakka River tributary 2 FWSUPPresentPresent PresentNot presentNot presentNot present Nine Mile Creek HLNWFPresentPresentNot presentPresentNot presentPresent Rice Creek near Springside FWNWFCPresentPresentPresentPresentNot presentPresent Santa Fe River near Graham FW NUPNot presentPresentNot presentPresentNot presentPresent Shiloh Run near Alachua FWNUPNot presentP resentPresentPresentNot presentPresent Snell Creek HLSWFPresentPresentNot presentPresentNot presentPresent South Fork Black Creek HLNWFPresentPresen tNot presentPresentNot presentPresent Spoil Bank tributary (Highlands)FWSUPPr esentPresentPresentPresentPresentPresent Ten Mile Creek FWNWFCPresentPresentNot presentPresentNot presentPresent Tiger Creek near Babson Park HLSUPPresentPres entNot presentNot pres entNot presentPresent Tiger Creek tributary HLSWFPresentPrese ntNot presentPresentNot presentPresent Triple Creek unnamed tributary 1HLSWFPresentPresentNot presentPresentNot presentNot present Triple Creek unnamed tributary 2FWSUPPresentPresentNot presentPresentPresentNot present Tuscawilla Lake tributary HLNUPNot presen tPresentNot presentPresentPresentPresent Tyson Creek FWSWFCPresentPresentNot presentPresentPresentPresent Unnamed Lower Wekiva tributaryHLNWFPresentPresentNot presentPresentPresentPresent Percentage of sites at which bankfull indicator is present: 87%100%13%84%18%78% Notes: FW = Flatwoods physiography; HL = Highlands physiography; N = Northern peninsula geography; S = Southern peninsula geogr aphy; WF = Wetland floodplain; WFC = Wetland floodplain dominated by cypress; UP = Upland floodplain

PAGE 78

78 Table 3-3. Summar y of slopes data Slope (%) WS : TOB ratio Slope (%) WS : BKF-F ratio Slope (%) WS : BKF-I ratio Slope (%) WS : BKF-S ratio Slope (%) WS : BKF-A ratio Alexander Springs Creek tributary 20.157-0.3470.6910.230.7010.220.5020.310.2800.560.1081.45 Blackwater Creek near Cassia -0.0260.072-0.1450.18-0.1450.18-0.0900.29-------Blues Creek near Gainesville 1No water0.2210.278 0.79 ----0.1571.410.3300.670.262 0.84 Bowlegs Creek near Ft Meade 0.1040.0270.115 0.90 0.115 0.90 0.094 1.11 ----0.117 0.89 Carter Creek near Sebring 0.1730.2240.7190.240.6310.270.4180.410.5600.310.1361.27 Catfish Creek near Lake Wales 0.0510.062-0.050-1.020.055 0.93 -0.069-0.74-0.054-0.940.052 0.98 Coons Bay Branch 1No water0.2530.268 0.94 0.269 0.94 0.4250.600.253 1.00 0.5210.49 Cow Creek -0.0020.2870.002-1.000.049-0.040.157-0.010.072-0.03-0.0180.11 Cypress Slash tributary 1, 2No water1.1401.018 1.12 1.027 1.11 0.921 1.24 1.003 1.14 1.096 1.04 East Fork Manatee River tributary 1No water0.1640.2270.720.2610.63-0.085-1.930.204 0.80 ---Fisheating Creek at Palmdale 0.020-0.1350.1160.170.1220.160.0430.470.0750.27---Gold Head Branch 21.6101.7921.796 0.90 1.607 1.00 1.419 1.13 1.336 1.21 1.540 1.05 Hammock Branch0.1020.357-0.226-0.45-0.102-1.000.00714.570.110 0.93 0.087 1.17 Hickory Creek near Ona 0.155-0.051-0.135-1.15----0.0841.850.0542.870.1211.28 Hillsborough River tributary 1No water-0.1180.481-0.250.392-0.300.271-0.44-------Horse Creek near Arcadia 0.0080.062-0.198-0.04-0.325-0.020.009 0.89 0.0130.62-0.037-0.22 Jack Creek 0.301-0.3490.1591.89-0.078-3.860.02810.750.268 1.12 0.5180.58 Jumping Gully 20.6040.3950.698 0.87 ----0.587 1.03 0.629 0.96 0.550 1.10 Lake June-In-Winter tributary 20.8450.8720.708 1.19 ----0.779 1.08 0.884 0.96 0.742 1.14 Little Haw Creek near Seville0.0400.1480.038 1.05 0.053 0.75 -0.005-8.00----0.0950.42 Livingston Creek near Frostproof0.0610.0960.0096.780.0242.54-0.077-0.790.056 1.09 0.0850.72 Livingston Creek tributary 1, 2No water1.4980.8871.690.5822.570.9571.570.6002.50---Lochloosa Creek at Grove Park 0.0970.4630.0601.620.0641.520.082 1.18 0.1320.730.1990.49 Lowry Lake tributary 0.3510.4701.1780.301.0460.340.6370.550.5370.650.5670.62 Manatee River near Myakka Head0.0620.0360.083 0.75 0.054 1.15 0.0940.660.0990.630.057 1.09 Manatee River tributary 0.0420.8160.8460.050.8460.050.4740.090.2230.19---Morgan Hole Creek 1No water-0.229-0.1002.29-0.1062.16-0.268 0.85 -0.0603.82-0.0623.69 Moses Creek near Moultrie 0.0960.1640.0362.670.0392.460.0761.260.124 0.77 0.092 1.04 Myakka River tributary 1 1No water0.045-0.074-0.610.0133.460.0780.58-------Myakka River tributary 2 1No water0.375-0.131-2.86-0.025-15.000.000**-------Nine Mile Creek 20.7130.6110.595 1.20 0.588 1.21 0.703 1.01 0.747 0.95 0.4071.75 Rice Creek near Springside 0.017-0.0640.2100.080.2090.080.1090.160.0360.470.0360.47 Santa Fe River near Graham 0.068-0.006-0.227-0.30-----0.067-1.01-0.013-5.230.057 1.19 Shiloh Run near Alachua 1, 2No water1.1281.048 1.08 ----1.169 0.96 1.132 1.00 1.033 1.09 Snell Creek0.1030.2450.1540.670.1450.710.114 0.90 0.1610.64-0.011-9.36 South Fork Black Creek 0.0800.1050.1760.450.1710.470.077 1.04 0.0481.67-0.013-6.15 Spoil Bank tributary 1No water0.1440.0197.58-0.111-1.300.0662.18-0.027-5.33-0.056-2.57 Ten Mile Creek 0.097-0.1560.096 1.01 0.096 1.01 0.113 0.86 -0.022-4.410.0731.33 Tiger Creek near Babson Park 0.0580.2110.1790.320.1340.430.0880.66----0.6140.09 Tiger Creek tributary 0.2130.0950.4640.460.4260.500.4970.430.3250.660.221 0.96 Triple Creek unnamed tributary 10.4190.4480.2521.660.2591.620.2731.530.7260.58---Triple Creek unnamed tributary 2 1No water0.4861.5370.321.5390.321.5180.320.7090.69---Tuscawilla Lake tributary 20.8440.7311.2220.69----1.015 0.83 0.738 1.14 0.5841.45 Tyson Creek 0.0080.088-0.033-0.24-0.036-0.220.0061.330.010 0.80 0.1500.05 Unnamed Lower Wekiva tributary0.1560.0610.1231.270.1221.280.1101.42-0.021-7.430.2510.62 Mean 0.2310.290.337 0.80 0.2820.240.300 1.01 0.3230.240.2910.35 Standard deviation 0.3430.450.4881.700.4432.820.4203.010.3682.200.3692.24 Percentage of sites with negative slope7%20%22%N/A21%N/A16%N/A16%N/A17%N/A Percentage of sites with BKF indicator slope within 25% of water slope 3N/AN/AN/A27%N/A24%N/A31%N/A37%N/A37% Flat floodplain (BKF-F) Inflection (BKF-I) Scour (BKF-S) Alluvial break (BKF-A) Site name Water slope (%) Channel bed slope (%) Top of bank (TOB) Notes: 1 Used channel bed slope in place of water slope in the calculation of the water slope to bankfull indicator slope ratio because there was no flowing water on the day of the survey; 2 Site has a water slope >0.5%; 3 Water slope to bankfull indicator ratio between 0.75 and 1.25; -= Bankfull indicator not found at site; WS = Water slope; BKF = Bankfull; N/A = Not applicable; Bold = Water slope to bankfull indicator ratio within 25% of water slope (i.e. has a ratio between 0.75 and 1.25)

PAGE 79

79 Table 3-4. Comparision of various water slope to bankfull indicator slope ratios by water slope Effect Average water slope to BKFindicator slope ratio Standard deviation Percentage of sites with BKF indicator slope within 25% of water slope Percentage of sites with negative slope P-value (t-Test assumming unequal variances) Top of bank (TOB): WS <0.5% 0.551.4817%25%0.03** WS >0.5% 1.090.3075%0% Flat floodplain (BKF-F): WS <0.5% 0.102.9415%24%0.02** WS >0.5% 1.470.7475%0% Inflection (BKF-I): WS <0.5% 0.993.3419%19%0.42 WS >0.5% 1.110.2288%0% Scour (BKF-S): WS <0.5% -0.032.4023%20%0.01** WS >0.5% 1.230.5288%0% Alluvial break (BKF-A): WS <0.5% 0.132.4629%21%0.01** WS >0.5% 1.230.2771%0% Notes: WS = Water slope; BKF = Bankfull; Water slop e to bankfull indicator ratio between 0.75 and 1.25; ** Represents statistical significance (p 0.05)

PAGE 80

80 Table 3-5. Gaged sites discharge summary: Reference reach survey results Discharge (cfs) Duration (% of time) Discharge (cfs) Rl (yrs) Duration (% of time) Discharge (cfs) Rl (yrs) Duration (% of time) Discharge (cfs) Rl (yrs) Duration (% of time) Discharge (cfs) Rl (yrs) Duration (% of time) Discharge (cfs) Rl (yrs) Duration (% of time)Blackwater Creek near Cassia 551.0532551.053233<1.0147N/AN/AN/AN/AN/AN/A42.41.0239 Blues Creek near Gainesville 1861.470.2N/AN/AN/A361.070.812<1.016.66.0<1.011460.01.200.3 Bowlegs Creek near Fort Meade36<1.012135<1.012125<1.0126N/AN/AN/A12<1.014230.6<1.0123 Carter Creek near Sebring 591.106.930<1.012642<1.011411<1.01715.5<1.01938.3<1.0182 Catfish Creek near Lake Wales 903.105.0371.1150973.704.2751.958.8341.0853501.2329 Fisheating Creek near Palmdale 75<1.014175<1.014139<1.015053<1.0146N/AN/AN/A13<1.0165 Hickory Creek near Ona 1, 321<1.016.5N/AN/AN/A4.7<1.01135.8<1.01111.1<1.012522<1.014.0 Horse Creek near Arcadia 289<1.0118280<1.011885<1.013929<1.01586.4<1.01823811.0414 Little Haw Creek near Seville 21081.04251141.052456<1.0137N/AN/AN/A13<1.016460<1.0135 Livingston Creek near Frostproof 41711.428.31001.12191061.141838<1.015138<1.0151771.0627 Lochloosa Creek at Grove Park 2, 313<1.012313<1.01236.3<1.01343.6<1.01440.1<1.01731.9<1.0153 Manatee River near Myakka Head 45951.092.14021.034.0116<1.011416<1.015618<1.0152201<1.018.7 Moses Creek near Moultrie 2, 3431.114.3431.114.3161.029.52.2<1.01291.2<1.0138551.153.3 Rice Creek near Springside 225<1.013325<1.013318<1.01409.9<1.01579.9<1.015721<1.0136 Santa Fe River near Graham 2, 43381.651.8N/AN/AN/A1181.101245<1.013114<1.01583391.651.8 Shiloh Run near Alachua 1, 3163.50IRN/AN/AN/A111.44IR5.51.040.10.2<1.01397.01.13IR Tiger Creek near Babson Park 1041.324.9671.0915641.0716N/AN/AN/A15<1.0193621.0717 Mean 156N/A15111N/A2664N/A2432N/A4316N/A60103N/A29 Standard deviation 162N/A14120N/A1336N/A1623N/A2311N/A23124N/A24 Median 90N/A8.367N/A2456N/A1829N/A5113N/A5760N/A27 Notes: 1 No discharge reported for the day of survey (gage inactive) and staff gage no longer at site or no longer accurate-estimated discharge and then determined the associated stage from the stage-Q rating table; 2 No discharge reported for the day of survey (gage inactive)-used gage height observed at the staff gage on the day of the survey and then determined the associ ated discharge from the stage-Q rating table; 3 Period of record for continuous data and/or annual peak data is less than 10 years, thus gage analysis results are rough estimates-did not include the results for these sites in summary statistics; 4 Discharge averaged over two days; cfs = cubic feet per second; RI = Return interval; yrs = years; W/D = Width-to-depth ratio; IR = Insufficient period of record for gage analysis; N/A = Not applicable (i.e. bankfull indicator not found at site and/or statist ics could not be conducted because results are inconclusive) Minimum W/D (BKF-W/D)Rl (yrs)Flat floodplain (BKF-F) Inflection (BKF-I) Scour (BKF-S) Alluvial break (BKF-A) Site Name Top of bank

PAGE 81

81 Table 3-6. Gaged sites stage summary : Reference reach survey results Site NameAdjusted stage (ft) Duration (% of time) Adjusted stage (ft) Duration (% of time) Adjusted stage (ft) Duration (% of time) Adjusted stage (ft) Duration (% of time) Adjusted stage (ft) Duration (% of time) Adjusted stage (ft) Duration (% of time)Blackwater Creek near Cassia 0.971.13170.971.13170.681.0725N/AN/AN/AN/AN/AN/A0.821.1021 Blues Creek near Gainesville 13.661.980.1N/AN/AN/A2.00<1.010.80.74<1.018.50.31<1.01182.871.450.3 Bowlegs Creek near Fort Meade1.42<1.01131.41<1.01140.98<1.0117N/AN/AN/A0.28<1.01271.23<1.0115 Carter Creek near Sebring 1.50<1.014.60.47<1.01210.94<1.0111-0.46<1.0170-0.92<1.0195-0.67<1.0185 Catfish Creek near Lake Wales 0.551.3316-0.42<1.01780.641.50120.311.2028-0.48<1.0181-0.131.0760 Fisheating Creek near Palmdale0.37<1.01430.37<1.0143-0.46<1.0152-0.09<1.0148N/AN/AN/A-1.45<1.0171 Hickory Creek near Ona 1, 31.09<1.014.5N/AN/AN/A0.77<1.019.90.83<1.018.90.44<1.01211.28<1.012.9 Horse Creek near Arcadia 2.331.04172.251.0417-0.11<1.0135-1.18<1.0153-1.90<1.01723.181.0814 Little Haw Creek near Seville 21.281.07251.371.08240.19<1.0138N/AN/AN/A-1.20<1.01660.28<1.0137 Livingston Creek near Frostproof 41.451.26130.451.08290.541.1027-0.78<1.0167-0.79<1.01670.04<1.0137 Lochloosa Creek at Grove Park 2, 31.16<1.01141.16<1.01140.74<1.01190.55<1.0123-0.17<1.01440.40<1.0126 Manatee River near Myakka Head 47.531.151.65.901.043.42.39<1.0111-0.22<1.0135-0.11<1.01333.63<1.017.6 Moses Creek near Moultrie 2, 32.55<1.013.92.55<1.013.91.37<1.018.70.04<1.0135-0.20<1.01442.91<1.012.8 Rice Creek near Springside 20.36<1.01310.36<1.01310.03<1.0138-0.51<1.0153-0.51<1.01530.20<1.0134 Santa Fe River near Graham 2, 45.381.652.3N/AN/AN/A1.741.1015-0.09<1.0141-1.35<1.01745.401.652.3 Shiloh Run near Alachua 1, 3--3.40--N/AN/AN/A--1.73----<1.01----<1.01----<1.01-Tiger Creek near Babson Park 0.891.1612.30.33<1.01260.27<1.0128N/AN/AN/A-0.94<1.01960.23<1.0130 Mean 2.13N/A151.22N/A280.76N/A24-0.25N/A45-0.69N/A621.20N/A32 Standard deviation 2.16N/A121.71N/A190.85N/A140.57N/A190.68N/A261.98N/A27 Median 1.42N/A130.47N/A240.64N/A25-0.22N/A48-0.79N/A670.28N/A30 Notes: 1 No stage reported for the day of survey (gage inactive) and staff gage no longer at site or no longer accurate-estimated flo w and then determined the associated stage from the stage-Q rating table; 2 No stage reported for the day of survey (gage inactive)-used gage height observed at the staff gage on the day of the survey; 3 Period of record for continuous data and/or annual peak data is less than 10 years, thus gage analysis results are rough estim ates-did not include the results for these sites in summary statistics; 4 Stage averaged over two days; Adusted stage = Bankfull stage Mean annual stage; ft = feet; -= No stage data available for this site; N/A = Not applicable (i.e. bankfull indicator not found at site and/or statistics could not be conducted because results are inconclusive); W/D = Width-to-depth ratioRI (yrs) RI (yrs) RI (yrs) RI (yrs)Inflection (BKF-I) Scour (BKF-S) Alluvial break (BKF-A)Minimum W/D (BKF-W/D) Top of bank Flat floodplain (BKF-F)RI (yrs) RI (yrs)

PAGE 82

82 Table 3-7. Gaged sites discharge su mmary: Annual maximum series results Discharge (cfs) Duration (% of time) Discharge (cfs) Duration (% of time) Discharge (cfs) Duration (% of time) Discharge (cfs) Duration (% of time) Discharge (cfs) Duration (% of time) Discharge (cfs) Duration (% of time) Discharge (cfs) Duration (% of time) Discharge (cfs) Duration (% of time) Discharge (cfs) Duration (% of time)Blackwater Creek near Cassia 3842131121736.92562.84930.46890.1979IR1225IR1493IR Blues Creek near Gainesville 1241.9700.3880.21250.1221IR296IR402IR489IR581IR Bowlegs Creek near Fort Meade83102312.22881.24030.56920.19140.112220.014720.01738IR Carter Creek near Sebring 4313932.71091.71400.52100.02590.03220.0370IR418IR Catfish Creek near Lake Wales 266952256017778.11132.31361.11670.31910.22140.1 Fisheating Creek near Palmdale3751914263.719342.229510.959980.186100.0126180.0160690.0199070.0 Hickory Creek near Ona 1, 3274.6890.71160.51700.33190.34410.1618IR766IR927IR Horse Creek near Arcadia 2771910163.713662.320650.941110.258480.184720.0107400.013243IR Little Haw Creek near Seville 287292668.83415.54902.38870.412020.116560.02028IR2427IR Livingston Creek near Frostproof 46133150101827.32473.34010.25130.16650.0785IR910IR Lochloosa Creek at Grove Park 2, 3368.81551.7521.01870.62500.33760.27430.110520.115210.1 Manatee River near Myakka Head 43534.910540.613410.419140.134200.146030.062950.07674IR9162IR Moses Creek near Moultrie 2, 31311831.81321.02290.66110.11009IR1710IR2388IR3228IR Rice Creek near Springside 21139.63891.05130.67620.314690.12056IR2924IR3656IR4457IR Santa Fe River near Graham 2, 456262154.92932.64481.09140.213180.11936IR2472IR3069IR Shiloh Run near Alachua 1, 35.00.19.0IR10IR13IR19IR24IR29IR33IR37IR Tiger Creek near Babson Park 50271005.51153.91462.02100.62540.23100.1352IR394IR Mean 122233996.25234.07711.714720.420540.229210.13656N/A4463N/A Standard deviation 126184566.86124.79262.218530.626440.338520.14892N/A6046N/A Median 61192153.72882.34030.96920.29140.112220.01472N/A1738N/A Notes: 1 No discharge reported for the day of survey (gage inactive) an d staff gage no longer at site or no longer accurate-estimated discharge and then determined the associated stage from the stage-Q rating table; 2 No discharge reported for the day of survey (gage inactive)-used gage height observed at the staff gage on the day of the survey and then determined the associate d discharge from the stage-Q rating table; 3 Period of record for continuous data and/or annual peak data is less than 10 years, thus gage analysis results are rough estimates-did not include the results for these sites in summary statistics; 4 Discharge averaged over two days; cfs = cubic feet per second; IR = Insufficient period of record for proper gage analysis; N/ A = Not applicable -insufficient period of record for proper gage analysis for the majority of sites Site Name1.01-year event (99% annual exceedance probability) 1.25-year event (80% annual exceedance probability) 1.5-year event (67% annual exceedance probability) 2-year event (50% annual exceedance probability) 5-year event (20% annual exceedance probability) 10-year event (10% annual exceedance probability) 25-year event (4% annual exceedance probability) 50-year event (2% annual exceedance probability) 100-year event (1% annual exceedance probability)

PAGE 83

83 Table 3-8. Gaged sites stage summa ry: Annual maximum series results Site NameAdjusted stage (ft) Duration (% of time) Adjusted stage (ft) Duration (% of time) Adjusted stage (ft) Duration (% of time) Adjusted stage (ft) Duration (% of time) Adjusted stage (ft) Duration (% of time) Adjusted stage (ft) Duration (% of time) Adjusted stage (ft) Duration (% of time) Adjusted stage (ft) Duration (% of time) Adjusted stage (ft) Duration (% of time)Blackwater Creek near Cassia -0.29561.546.41.774.22.231.32.960.43.420.23.79IR4.03IR4.03IR Blues Creek near Gainesville 12.290.52.570.32.940.33.690.14.41IR4.51IR4.53IR4.53IR4.53IR Bowlegs Creek near Fort Meade2.378.03.134.23.432.84.031.04.850.25.390.06.11IR6.11IR6.11IR Carter Creek near Sebring 1.753.42.900.33.140.13.630.14.76IR5.24IR6.09IR6.20IR6.20IR Catfish Creek near Lake Wales-0.36740.49190.64120.945.11.301.71.640.72.230.12.400.022.440.005 Fisheating Creek near Palmdale1.66262.914.73.183.03.721.14.730.15.230.17.260.01 8.350.019.32IR Hickory Creek near Ona 1, 31.203.81.660.81.780.62.030.32.600.22.600.22.600.22.600.22.600.16 Horse Creek near Arcadia1.61217.273.88.102.89.761.011.920.113.120.0214.18IR14.31IR14.33IR Little Haw Creek near Seville 20.68323.097.43.405.14.012.25.190.45.560.26.280.026.530.016.56IR Livingston Creek near Frostproof 40.04371.43142.068.23.322.24.570.35.450.16.57IR6.57IR6.57IR Lochloosa Creek at Grove Park 2, 32.453.33.061.03.420.64.140.35.730.16.610.037.27IR7.28IR7.28IR Manatee River near Myakka Head 45.384.19.110.69.660.510.760.313.060.0414.590.0215.80IR17.75IR17.75IR Moses Creek near Moultrie 2, 32.962.72.962.73.241.83.790.96.41IR6.41IR6.41IR6.41IR6.41IR Rice Creek near Springside 22.038.63.480.73.700.44.150.24.840.05.50IR5.91IR6.12IR6.12IR Santa Fe River near Graham 2, 40.19354.214.64.953.16.420.97.890.28.450.19.56IR9.92IR9.92IR Shiloh Run near Alachua 1, 3-----------------------------------Tiger Creek near Babson Park 0.54191.098.91.247.01.544.32.321.22.960.13.66IR3.66IR3.66IR Mean 1.38253.335.73.713.84.481.55.600.46.240.17.08N/A7.42N/A7.50N/A Standard deviation 1.78242.736.12.874.03.191.73.790.64.160.24.38N/A4.88N/A4.89N/A Median 1.61212.914.63.183.03.721.04.760.25.390.16.11N/A6.20N/A6.20N/A Notes: 1 No stage reported for the day of survey (gage inactive) and staff gage no longer at site or no longer accurate-estimated flo w and then determined the associated stage from the stage-Q rating table; 2 No stage reported for the day of survey (gage inactive)-used gage height observed at the staff gage on the day of the survey; 3 Period of record for continuous data and/or annual peak data is less than 10 years, thus gage analysis results are rough estim ates-did not include the results for these sites in summary statistics; 4 Stage averaged over two days; Adusted stage = Bankfull stage Mean annual stage; ft = feet; -= No stage data available for this site; IR = Insufficient period of record for gage analysis; N/A = Not applicable -insufficient period of record for proper gage analysis for the majority of sites1.01-year event (99% annual exceedance probability) 1.25-year event (80% annual exceedance probability) 1.5-year event (67% annual exceedance probability) 50-year event (2% annual exceedance probability) 100-year event (1% annual exceedance probability) 2-year event (50% annual exceedance probability) 5-year event (20% annual exceedance probability) 10-year event (10% annual exceedance probability) 25-year event (4% annual exceedance probability)

PAGE 84

84 Table 3-9. Comparison of various bankfull indicator discharge a nd stage durations by floodplain type Effect AverageP-valueAverageP-value Top of bank (TOB): Wetland floodplain25<0.01* 23<0.01* Upland floodplain4.0 5.7 Flat floodplain (BKF-F):310.14 320.24 Wetland floodplain16 20 Upland floodplain Inflection (BKF-I): Wetland floodplain35<0.01* 310.03* Upland floodplain13 16 Scour (BKF-S): Wetland floodplain430.97 450.95 Upland floodplain43 44 Alluvial break (BKF-A): Wetland floodplain600.98 600.79 Upland floodplain60 64 Represents statistical significance (p 0.05) Discharge duration (% of time exceeded) Stage duration (% of time exceeded)

PAGE 85

85 Figure 3-1. Various field indicato rs of bankfull stage: flat floodplain (BKF-F), inflection (BKFI), scour (BKF-S), moss (BKF-M), and alluvial break (BKF-A).

PAGE 86

86A B C D E Figure 3-2. Water slope to various bankfull indicator slope ratios ag ainst water slope. A) Top of bank (TOB). B) Flat floodplain (BKF-F). C) Inflection (BKF-I). D) Scour (BKFS). E) Alluvial break (BKF-A). Note: Th e pink parallel lines bracket the range of water slope to bankfull indicator slope ra tios lying between 0.75 and 1.25 (i.e., the ratios for which bankfull indicator slop e is within 25% of the water slope).

PAGE 87

87A 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 02040608010012014016018020 0 Width (ft)Stage (ft) Field Measurment Data Stage 1.5-year event (6.41) Stage BKF-F (4.37) Stage BKF-I (3.24) B 0.00 2.00 4.00 6.00 8.00 10.00 12.00 051015202530354 0 Width (ft)Stage (ft) Field Measurement Data Stage 1.5-year event (8.71) Stage BKF-I (6.51) Figure 3-3. Width against stage field measurements. A) Little Haw Creek near Seville, a nonincised stream with a wetland floodplain. B) Carter Creek near Sebring, an incised stream with an upland floodplain.

PAGE 88

88A B Figure 3-4. Example of variability in stage-Q ra ting curves. A) Catfish Creek near Lake Wales (WY 1947-2007) stage-Q rating curve. B) Catfish Creek near Lake Wales (WY 1947-2007) stage-Q rating curve split into decades.

PAGE 89

89A B Figure 3-5. Boxplots of stage and discharge data for gaged sites. A) Stage Mean annual stage for all gaged sites, except Shiloh Run which had no stage data. B) Discharge/mean annual discharge for all gaged sites. Note : Because Q/Qma is plotted on a log scale there is no zero on the y-axis, and thus fo r sites with a minimum flow of zero, the boxplot line was extended down to the x-axis.

PAGE 90

90 CHAPTER 4 DEVELOPING REGIONAL CURVES FOR PENINSULAR FLORIDA Introduction Regional curves, which relate bankf ull discha rge and channel geometry (cross-sectional area, width, and mean depth) to drainage area in regions of similar climate, geology, and vegetation, have greatly aided in creating target natural channel de signs. Bankfull discharge, or flow that fills a stable alluvi al channel to the el evation of the active floodplain, is a useful parameter in developing regional curves because its stage is reasonably identifiable in the field, and it is the flow most often used to estimate the channel-forming discharge. Dunne and Leopold (1978) describe bankfull discharge as the most effective stream-flow for moving sediment, forming or removing bars, forming or changing bends and meanders, and generally doing work that results in the average morphol ogical characteristics of channels. While regional curves provide important information for natural channel structure, they also aid in estimating bankfull discharge and channel geometry in ungaged watersheds where drainage area is known, help confirm field identifications of bankfull stage, and allow for comparisons between regions (Leopold, 1994). Metcalf (2004) published regional curves for Florida streams, yet his study sites were confined to extreme north Florida and the Panha ndle, and even included sites in Georgia and Alabama (Figure 1-1). Peninsular Florida, however, is quite different in terms of its physiography, geological context, and rainfall patterns, as described in Chapter 2. It is an objective of the present work to develop regional curves for penins ular Florida. To accomplish this objective bankfull discharge and channel geometry (cross-s ectional area, width, and mean depth) were plotted agains t drainage area, and coeffi cients of determination (R2) were determined. Data were analyzed to determin e whether significant differences exist between

PAGE 91

91 streams draining different physiogr aphies (flatwoods versus highl ands), geographies (northern versus southern peninsula), a nd floodplain types (wetland versus upland and cypress-dominated versus non-cypress-dominated), in terms of thei r bankfull parameters and various dimensionless ratios. The return interval associated with ba nkfull discharge was also estimated for peninsular Florida streams. The regional curves developed in this study were also compared to those developed for other regions of the southeastern United States Coastal Plain. This chapter describes the methods used to re ach the objectives; the study result s; a discussion of the potential errors, trends, and anomalies associated with da ta collection and analys is; and conclusions. Methods Tasks com pleted to develop regional curves fo r peninsular Florida in cluded: 1) selecting between 40 and 50 gaged and ungaged stream sites that span a variety of dr ainage area sizes and valley slopes, as well as differe nt physiographies (flatwoods ve rsus highlands) and geographies (northern versus southern peninsula); 2) c onducting reference reach surveys to determine bankfull channel geometry and discharge; 3) ch oosing the most reliable bankfull indicator for peninsular Florida streams; 4) delineating drainage basins and determining valley slopes; 5) developing and analyzing regional cu rves for peninsular Florida base d on the entire data set as well as subsets of the data (physiography, geogr aphy, and floodplain types) ; 6) determining and analyzing various dimensionles s ratios (sinuosity, width-to-depth, maximum depth-to-mean depth, and valley slope); 6) estimating bankfull return intervals; and 7) comparing regional curves developed for peninsular Florida to ot her southeastern United States Coastal Plain regional curve studies. The methods used to co mplete the first three tasks are reported in Chapter 3, while the remaining tasks are presented below.

PAGE 92

92 Drainage Area Delineation a nd Valley Slope D etermination Drainage areas for each site were delin eated by heads-up digitizing in ARCMap GIS using 5-foot contour USGS 1:24000 quads and re fined using high-resolu tion aerials. Valley slopes were also determined in ARCMap GI S using 5-foot contour USGS 1:24000 quads by dividing the change in elevation by the straight line distance betw een the contour lines straddling the reference reach upstream and downstream. Regional Curve Development Data obtained from the reference reach su rveys were used to determine bankfull discharge, bankfull cross-sec tional area, bankfull width, and bankfull mean depth. Bankfull channel geometry parameters were based on the average value of the two smallest cross-sections (based on cross-sectional area) surveyed duri ng the reference reach survey conducted at each study site, while bankfull discharge and stage were estimated for only gaged sites by using reference reach survey data of the field bankfull stage in conjunction with the most current USGS stage-discharge rating table. Regional curves were created in Microsoft Excel by plotting the various bankfull parameters against draina ge area on a log-log scal e. A power function regression was fit to the data, and the coefficient of determination (R2) was determined. Due to the potential inaccuracies of determining bankf ull discharge at gaged sites, mean annual discharge and 1.5-year discharge were also pl otted against drainage area to see if these discharges were better correla ted with drainage area than th e bankfull discharge. More specifically, the 1.5-year return in terval was chosen because it is the return interval most often associated with the bankfull flow (Leopold, 1994). To determine whether peninsular Florida regional curves should be further split by physiography (flatwoods versus highlands), geogr aphy (northern versus southern peninsula), and/or floodplain type (wetla nd versus upland and cypress-dominated versus non-cypress-

PAGE 93

93 dominated), data were sorted by e ach of these subsets, and separate regional curves were created. Analysis of Covariance (ANCOVA) tests were then performed to determine whether significant differences exist in slopes and/or intercepts of bankfull di scharge and channel geometry regressions for each data subset (JMP 7). It is important to note that the slope of the regression gives an indication of how sensitive a parameter is to changes in drainage area, and that if slopes are significantly different, this indicates that bank full parameters in one set of streams change at a different rate with changes in dr ainage area than another set of streams. The intercept of the regression (determined by plotting the log values of the bankfull parameters and drainage area on a linear scale to obtain a linear regression) gi ves an indication of each bankfull parameters starting point, and if the intercepts are signif icantly different, this indicates that one set of streams starts out at a different bankfull parameter than another set of streams. Additionally, durations of various discharges were estimated and several ratios were calculated, including the peak discharge-to-mean annual discharge ratio (Qp/Qma), which is an indication of the flashiness of a stream, the bankfull discharg e-to-mean annual discharge ratio (Qbkf/Qma), and the bankfull discharge-to-1.5-year discharge ratio (Qbkf/Q1.5). Dimensionless Ratios Data obtained from reference reach survey s were also used to determine various dimensionless ratios such as sinuosity, widthto-depth (w/d), maximu m depth-to-mean depth (dmax/d), and valley slope. Sinuosity, which is found by dividing the channel length by the valley length, helps to define a streams pattern. The width-to-depth ra tio, found by dividing the bankfull width by the bankfull mean depth, and th e maximum depth-to-mean depth both help to define a channels shape. Valley slope, as prev iously mentioned, is found by dividing the change in elevation by the straight line distance between the contour lines straddl ing the reference reach upstream and downstream. Comparisons of m eans tests were performed using Excel Data

PAGE 94

94 Analysis ANOVA: Single factor to determine if significant differences exist in the various dimensionless ratios by physiograp hy (highlands versus flatwoods ), geography (northern versus southern peninsula), and/or floodplain types (w etland versus upland and cypressversus noncypress-dominated). Return Interval Annual peak flow data for gaged sites were analyzed to determ ine return intervals associated w ith bankfull discharge using Log Pear son Type III distributions (skew coefficient of -0.1) in RIVERMorph (USGS, 1982). Comparison to Other Southeastern United States Coastal Plain Regional Curves Raw data from both the present work and previous regional curve studies conducted throughout the southeastern United States Coastal Plain were entered into Excel, and regional curves for each bankfull parameter were compiled into one graph for visual comparison. Analysis of Covariance (ANCOVA) tests were then performed to determine whether significant differences exist in slopes and/or intercepts of bankfull di scharge and channel geometry regressions between peninsular Fl orida streams (the baseline regr ession) and other Coastal Plain regional curves (JMP 7). Results Results of the study are presented below, which include description of the drainage areas and valley slopes of the sites, presentation of bankfull discharge and channel geom etry regressions developed for peninsul ar Florida along with analysis of the various data subsets (physiography, geography, and floodplai n type), presentation of va rious dimensionless ratios along with analysis of the various data subsets, presentation of estimated return intervals associated with the bankfull discharge, and co mparison of the regional curves developed for

PAGE 95

95 peninsular Florida to other regional curves st udies conducted throughout the southeastern United States Coastal Plain. Drainage Area Delineation a nd Valley Slope D etermination Drainage areas ranged from 0.2 sq mi at Trip le Creek unnamed tributary 2 to 311 sq mi at Fisheating Creek at Palmdale, with mean and median values of 31.8 sq mi and 4.6 sq mi, respectively (Table 1-1). Valley slopes ranged from a very flat 0.02% at Blackwater Creek near Cassia to a high of 2.27% at Tuscawilla Lake tribut ary, one of the headwater streams, with mean and median values of 0.41% and 0.17%, respectivel y (Table 1-1). Genera lly, sites with smaller drainage areas had steeper slopes than those w ith larger areas, as expected (Figure 4-1). Regional Curve Development Bankfull discharge, m ean annual discharge, and 1.5-year discharge were plotted against drainage area for the 17 gaged sites (Figures 4-2 though 4-6). Bankfull cross-sectional area, bankfull width, and bankfull mean depth were pl otted against drainage area for all 45 sites (Figures 4-7 through 4-11). Thes e relationships are presented and analyzed for the entire dataset and data subsets based on physiography (flatwoods versus highlands), geography (northern versus southern peninsula), a nd floodplain type (wetland versus upland and cypress-dominated versus non-cypress-dominated) in the following subsections. Table 4-1 summarizes discharge data used in peninsular Florid a regional curve development, wh ile Table 4-2 summaries channel geometry data used. Tables 4-3 and 4-4 summarize the power function regression equations, corresponding coefficients of determination, and sa mple sizes for discharge against drainage area and channel geometry against drainage area, respectively.

PAGE 96

96 Discharge: bankfull, mean annual, 1.5-year Relationships for bankfull discharge, m ean a nnual discharge, and 1.5-year discharge as a function of drainage area for gaged sites are shown in Figures 4-2. Power function regression equations, corresponding coefficients of determination (R2), and sample sizes are: Qbkf = 14.26 Aw 0.36 R2 = 0.60 n = 17 (4-1) Qma = 1.36 Aw 0.88 R2 = 0.95 n = 17 (4-2) Q1.5 = 27.85 Aw 0.57 R2 = 0.60 n = 17 (4-3) where Qbkf = bankfull discharge in cubi c feet per second (cfs), Qma = mean annual discharge in cfs, Q1.5 = 1.5-year discharge in cfs, and Aw = watershed drainage area in square miles (sq mi). Bankfull discharge, mean annual discharge, and 1.5-year discharge are all directly related to drainage area across the entire study area, with 60%, 95%, an d 60% of the variability in discharges explained by drai nage area, respectively. On average, bankfull discharge is exceed ed 21% of the time, while mean annual discharge is exceeded 26% of the time and the 1.5 -year discharge is exceeded 3.4% of the time (note that the lower the % duration, the rarer or less frequent th e event). On aver age, the largest flood or peak discharge (Qp) is 52 times greater than mean a nnual discharge. Bankfull discharge is 35% of 1.5-year discharge and is 4.3 times gr eater than mean annual discharge on average. However, at six gaged sites, bankfull discharge is almost equal to or less than mean annual discharge, which is not expected as the bankfu ll discharge is generally higher than the mean annual discharge (Leopold, 1994 see Figure 2-15) (Table 4-1). Bankfull stage, on the other hand, is greater than mean annual stage at all but one gaged site (Catfish Creek near Lake Wales) (Appendix F). Physiography (flatwoods versus highlands) : Relationships for bankfull discharge, mean annual discharge, and 1.5-year discharge as a f unction of drainage area for the gaged sites for

PAGE 97

97 flatwoods (FW) and highlands (HL) physiogr aphies are shown in Figure 4-3. The power function regression equations, corresponding coefficients of determination, and sample sizes are: Qbkf-FW = 14.47 A w 0.38 R2 = 0.64 n = 12 (4-4) Qbkf-HL = 6.97 Aw 0.49 R2 = 0.39 n = 5 (4-5) Qma-FW = 1.35 A w 0.92 R2 = 0.96 n = 12 (4-6) Qma-HL = 2.55 Aw 0.67 R2 = 0.86 n = 5 (4-7) Q1.5-FW = 28.65 A w 0.69 R2 = 0.86 n = 12 (4-8) Q1.5-HL = 10.20 Aw 0.58 R2 = 0.45 n = 5 (4-9) Bankfull discharge is directly related to drainage area, with 64% and 39% of the variability in discharge explained by drainage area for flatwoods and highlands physiographies, respectively. Mean annual discharge is directly related to drai nage area, with 96% and 86% of the variability in discharge explained by drai nage area for flatwoods and highland physiographies, respectively. Discharge associated with the 1.5-year event is di rectly related to draina ge area, with 86% and 45% of variability in discharge explained by drainage area for flatwoods and highlands physiographies, respectively. For a given dr ainage area, bankfull discharge and 1.5-year discharge at the flatwoods sites a ppear to be higher than at the hi ghlands sites, while there is not an obvious trend between physiographies for the mean annual discharge. Additionally, flatwoods streams start out (i.e., if drainage area were to equa l zero) with a larger 1.5-year discharge than highlands streams (p=0.02), while their bankfull and mean annual discharges start out the same. (p>0.05) The various discharges in crease at the same rate with an increase in drainage area for both physiographies (p>0.05) (Table 4-5). No significant difference exis ts (p>0.05) in the duration of bankfull discharge based on physiography, which is equaled or exceeded on av erage 18% of the time at flatwoods sites and 26% at highlands sites. A si gnificant difference exists (p<0.1 ) in duration of mean annual discharge, which is equaled or exceeded on aver age 23% of the time at flatwoods sites and 35%

PAGE 98

98 at highlands sites. A signifi cant difference exists (p=0.01) in duration of 1.5-year discharge, which is equaled or exceeded on average 1.6% an d 7.4% of the time for flatwoods and highlands physiographies, respectively. On average, bankfu ll discharge is 29% of the 1.5-year discharge at flatwoods sites and 49% at highla nds sites. Peak discharge is 69 and 11 times greater than mean annual discharge for flatwoods and highlands physiographies, respectively, which is significantly different (p=0.01) (Table 4-6). Geography (northern versus southern peninsula): Relationships for bankfull discharge as a function of drainage area for gaged sites for northern (NP) and southern peninsula (SP) geographies are shown in Fi gure 4-4. Power function regr ession equations, corresponding coefficients of determina tion, and sample sizes are: Qbkf-NP = 15.98 Aw 0.32 R2 = 0.58 n = 8 (4-10) Qbkf-SP = 9.42 Aw 0.47 R2 = 0.57 n = 9 (4-11) Qma-NP = 1.37 A w 0.88 R2 = 0.94 n = 8 (4-12) Qma-SP = 1.30 Aw 0.88 R2 = 0.95 n = 9 (4-13) Q1.5-NP = 29.79 A w 0.53 R2 = 0.78 n = 8 (4-14) Q1.5-SP = 22.27Aw 0.63 R2 = 0.37 n = 9 (4-15) Bankfull discharge is directly related to drainage area, with 58% and 57% of variability in discharge explained by drainage area for northern and sout hern peninsula geographies, respectively. Mean annual discharge is directly related to drai nage area, with 94% and 95% of variability in discharge expl ained by drainage area for nor thern and southern peninsula geographies, respectively. Discha rge associated with the 1.5-year event is directly related to drainage area, with 78% and 37% of variability in discharge explained by drainage area for northern and southern peninsula geographies, respectivel y. For a given drainage area, bankfull discharge, mean annual discharge, and 1.5-year di scharge appear to be similar at northern and southern peninsula sites. Additionally, northern and southern peninsula streams start out with

PAGE 99

99 the same bankfull, mean annual, and 1.5-year discharges (p>0.05). The various discharges increase at the same rate with an increase in dr ainage area for both geographies (p<0.05) (Table 4-5). No significant difference exis ts (p>0.05) in duration of bankfull discharge based on geography, which is equaled or exceeded on aver age 18% of the time at northern sites and 22% at southern sites. No significant difference exis ts (p>0.05) in duration of mean annual discharge, which is exceeded on average 25% of the time at northern sites and 28% at southern sites. No significant difference exists (p>0.05) in duration of 1.5-year discharge, which is equaled or exceeded on average 2.5% and 4.1% of the time for northern and southern peninsula geographies, respectively. On average, bankfull discharge is 40% of the 1.5-year discharge at northern sites and 31% at southern sites. Peak discharge is 55 and 49 times greater than mean annual discharge for northern and southern pe ninsula sites, respec tively, which is not significantly different (p>0.05) (Table 4-6). Floodplain type (wetland versus upland): Relationships for bankfull discharge as a function of drainage area for gaged sites for wetland (WF) and upland (UP) floodplain types are shown in Figure 4-5. Power function regressi on equations, corresponding coefficients of determination, and sample sizes are: Qbkf-WF = 9.13 Aw 0.44 R2 = 0.55 n = 10 (4-16) Qbkf-UP = 18.64 Aw 0.36 R2 = 0.88 n = 7 (4-17) Qma-WF = 2.16 A w 0.78 R2 = 0.92 n = 10 (4-18) Qma-UP = 1.04 Aw 0.91 R2 = 0.98 n = 7 (4-19) Q1.5-WF = 28.49 A w 0.58 R2 = 0.52 n = 10 (4-20) Q1.5-UP = 27.24Aw 0.53 R2 = 0.65 n = 7 (4-21) Bankfull discharge is directly related to drainage area, with 55% and 88% of variability in discharge explained by drainage area for wetland and upland fl oodplains, respectively. Mean

PAGE 100

100 annual discharge is directly related to draina ge area, with 92% and 98% of variability in discharge explained by draina ge area for wetland and upland floodplains, respectively. Discharge associated with the 1.5-year event is di rectly related to draina ge area, with 52% and 65% of variability in discharge explained by drainage area for wetland and upland floodplains, respectively. For a given drainage area, bankfu ll discharge at sites wi th an upland floodplain appears to be higher than at the sites with a wetland floodplain, while the opposite is true for mean annual discharge, and there is no obvious trend between floodplain types for the 1.5-year discharge. Additionally, str eams with a wetland floodplain and those with an upland floodplain start out with the same bankfull, mean annual, and 1.5-year discharges (p>0.05). The various discharges increase at the same rate with an increase in draina ge area for both floodplain types (p<0.05) (Table 4-5). A nearly significant difference exists (p=0.06) in duration of bankfull discharge based on floodplain type, which is equaled or exceeded on average 25% of the time at sites with a wetland floodplain and 13% at sites with an upland floodplain. No significant difference exists (p>0.05) in duration of mean annual discharge, which is equaled or exceeded on average 25% of the time at sites with a wetland floodplain and 29% at sites with an upland floodplain. No significant difference exists (p>0.05) in dur ation of 1.5-year discharge, wh ich is equaled or exceeded on average 3.8% and 2.7% of the time for wetla nd and upland floodplains, respectively. On average, bankfull discharge is 24% of 1.5-year discharge at sites with a wetland floodplain and 49% at sites with an upland fl oodplain. The largest flood is 66 and 31 times greater than the mean annual discharge for wetland and upla nd floodplains, respectively, which is not significantly different (p>0.05) (Table 4-6).

PAGE 101

101 Floodplain type (cypress-dominated versus non-cypress-dominated): Relationships for bankfull discharge as a function of drainage area for gaged sites for cypress-dominated (CD) and non-cypress dominated (NC) fl oodplain types are shown in Figure 4-6. Power function regression equations, co rresponding coefficients of dete rmination, and sample sizes are: Qbkf-CD = 10.94 Aw 0.35 R2 = 0.48 n = 7 (4-22) Qbkf-NC = 15.49 Aw 0.41 R2 = 0.79 n = 10 (4-23) Qma-CD = 2.87 A w 0.72 R2 = 0.87 n = 7 (4-24) Qma-NC = 1.12 Aw 0.91 R2 = 0.98 n = 10 (4-25) Q1.5-CD = 19.88 A w 0.62 R2 = 0.48 n = 7 (4-26) Q1.5-NC = 30.49Aw 0.56 R2 = 0.68 n = 10 (4-27) Bankfull discharge is directly related to drainage area, with 48% and 79% of variability in discharge explained by draina ge area for cypress-domina ted and non-cypress-dominated floodplains, respectively. Mean a nnual discharge is dir ectly related to drainage area, with 87% and 98% of variability in discharge explained by drainage area for cypress-dominated and noncypress-dominated floodplains, re spectively. Discharge associated with the 1.5-year event is directly related to drainage area, with 48% a nd 68% of variability in discharge explained by drainage area for cypress-dominated and non-cypress-dominated floodplains, respectively. For a given drainage area, bankfull discharge and 1.5year discharge at sites with a floodplain not dominated by cypress appear to be higher than at the sites with a cypress-dominated floodplain, while the opposite is true for the mean annual di scharge. Additionally, streams with a cypressdominated floodplain and those w ith a non-cypress-dominated floodpl ain start out with the same mean annual and 1.5-year discharges (p>0.05), bu t a significantly different bankfull discharge (p=0.05). The various discharges increase at the sa me rate with an increase in drainage area for both floodplain types. (p<0.05) (Table 4-5).

PAGE 102

102 A significant difference exists (p=0.01) in duration of bankfull discharge based on floodplain type, which is equaled or exceeded on average 30% of the time at sites with a floodplain dominated by cypress an d 13% at sites with a floodplai n not dominated by cypress. No significant difference exists (p>0.05) in duration of mean a nnual discharge, which is equaled exceeded on average 26% of the time at sites w ith a floodplain dominated by cypress and 27% at sites with a floodplain not dominated by cyrpess. No significant difference exists (p>0.05) in duration of 1.5-year discharge, which is equale d or exceeded on average 4.9% and 2.2% of the time for cypress-dominated and non-cypress-dominat ed floodplains, respectively. On average, bankfull discharge is 28% of 1.5-year discharge at sites with a floodplai n dominated by cypress and 40% at sites with a floodplain not dominated by cypress. The largest flood is 67 and 41 times greater than mean annual discharge for cypress-dominated and non-cypress-dominated floodplains, respectively, which is not signi ficantly different (p>0.05) (Table 4-6). Bankfull cross-sectional area The rela tionship for bankfull cross-sectional ar ea as a function of drainage area for the entire data set is shown in Figure 4-7A. Power function regressi on equation, corresponding coefficient of determination (R2), and sample size are: Abkf = 6.05 Aw 0.47 R2 = 0.78 n = 45 (4-28) where Abkf = bankfull cross-sec tional area in square feet (sq ft) and Aw = watershed drainage area in square miles (sq mi). Bankfu ll cross-sectional area is directly related to drainage area, with 78% of variability in cr oss-sectional area across the entire stu dy area explained by drainage area. Physiography (flatwoods versus highlands): Relationships for bankfull cross-sectional area as a function of drainage area for flatw oods and highlands physiogr aphies are shown in Figure 4-8A. Power function regr ession equations, corresponding co efficients of determination, and sample sizes are:

PAGE 103

103 Abkf-FW = 6.27 A w 0.46 R2 = 0.82 n = 25 (4-29) Abkf-HL = 5.80 Aw 0.49 R2 = 0.74 n = 20 (4-30) Bankfull cross-sectional area is directly rela ted to drainage area, with 82% and 74% of variability in cross-sectional area explained by drainage area for flatwoods and highlands physiographies, respectively. For a given drainage area, bankfull cr oss-sectional area is similar at flatwoods and highlands sites. Additionally, flatwoods and highlands streams start out with the same bankfull area (p>0.05), and bankfull area incr eases at the same rate with an increase in drainage area for both physiographies (p<0.05) (Table 4-7). Geography (northern versus southern peninsula): Relationships for bankfull crosssectional area as a function of drainage area for northern and southern ge ographies are shown in Figure 4-9A. Power function regr ession equations, corresponding co efficients of determination, and sample sizes are: Abkf-NP = 6.41 Aw 0.49 R2 = 0.80 n = 19 (4-31) Abkf-SP = 5.78 Aw 0.46 R2 = 0.78 n = 26 (4-32) Bankfull cross-sectional area is directly rela ted to drainage area, with 80% and 78% of variability in cross-sectional ar ea explained by drainage area for northern and southern peninsula geographies, respectively. For a given drainage area, bankfull cross-sectional area appears to be similar at northern and southern peninsula sites. Additionally, northern and southern peninsula streams start out with the same bankfull area (p>0.05), and it increas es at the same rate with an increase in drainage area for bot h geographies (p<0.05) (Table 4-7). Floodplain type (wetland versus upland): Relationships for bankfull cross-sectional area as a function of drainage area for wetland a nd upland floodplain types are shown in Figure 410A. Power function regression equations, corre sponding coefficients of determination, and sample sizes are:

PAGE 104

104 Abkf-WF = 8.11 Aw 0.41 R2 = 0.79 n = 23 (4-33) Abkf-UP = 5.13 Aw 0.47 R2 = 0.75 n = 22 (4-34) Bankfull cross-sectional area is directly rela ted to drainage area, with 79% and 75% of variability in cross-sectional ar ea explained by drainage area for northern and southern peninsula geographies, respectively. For a given drainage area, bankfull cross-sectional area appears to be larger at sites with a wetland floodplain than at those without a wetland floodplain. Additionally, streams with a wetland floodplain start out larg er than streams with an upland floodplain (p=0.03). Bankfull area increases at the same ra te with an increase in drainage area for both (p<0.05) (Table 4-7). Note that a cluster of upl and sites occurs between a drainage area of 0.1 and one square mile (Figure 4-10A). Floodplain type (cypress-dominated versus non-cypress-dominated): Relationships for bankfull cross-sectional area as a function of drainage ar ea for cypress-dominated and noncypress-dominated floodplain t ypes are shown in Figure 4-11A. Power function regression equations, corresponding coefficients of determination, and sample sizes are: Abkf-CD = 7.29 Aw 0.46 R2 = 0.84 n = 11 (4-35) Abkf-NC = 5.90 Aw 0.45 R2 = 0.73 n = 34 (4-36) Bankfull cross-sectional area is directly relate d to drainage area, w ith 84% and 73% of the variability in cross-sectional area explained by drainage area for the northern and southern peninsula geographies, respectively. For a given drainage area, th e bankfull cross-sectional area appears to be slightly larger at sites with a floodplain dominated by cypress than at those the floodplain is not dominated by cypress. Add itionally, streams with a cypress-dominated floodplain and those with a non-cypress-dominated floodplain start out with the same bankfull area (p>0.05), and bankfull area incr eases at the same rate with an increase in drainage area for both floodplain types (p<0.05) (Table 4-7). Also note that a cluster of non-cypress-dominated floodplain sites occurs between a drainage area of 0.1 and one square mile (Figure 4-11A).

PAGE 105

105 Bankfull width The relationship for bankfull width as a function of drainage area for the entire data set is shown in Figure 4-7B. Power function regre ssion equation, corresponding coefficient of determ ination (R2), and sample size are: Wbkf = 6.87 Aw 0.30 R2 = 0.81 n = 45 (4-37) where Wbkf = bankfull width in feet (ft), and Aw = watershed drainage area in square miles (sq mi). Bankfull width is directly related to drainage area, with 81% of variability in width across the entire study area explai ned by drainage area. Physiography (flatwoods versus highlands): Relationships for bankfull width as a function of drainage area for flatwoods and high lands physiographies are shown in Figure 4-8B. Power function regression equations, corresponding coefficients of determination, and sample sizes are: Wbkf-FW = 7.28 A w 0.28 R2 = 0.92 n = 25 (4-38) Wbkf-HL = 6.43 Aw 0.33 R2 = 0.72 n = 20 (4-39) Bankfull width is directly relate d to drainage area, with 92% a nd 72% of variability in width explained by drainage area for flatwoods and highlands physiographies, respectively. For a given drainage area, bankfull width appears to be similar at flatwoods and highlands sites. Additionally, flatwood and highland streams start out with the sa me bankfull width (p>0.05), and bankfull width increases at the same rate with an increase in drainage area for both physiographies (p<0.05) (Table 4-8). Geography (northern versus southern peninsula): Relationships for bankfull width as a function of watershed area for northern and southe rn peninsula geographies are shown in Figure 4-9B. Power function regressi on equations, corresponding coefficients of determination, and sample sizes are:

PAGE 106

106 Wbkf-NP = 6.26 Aw 0.30 R2 = 0.76 n = 19 (4-40) Wbkf-SP = 7.32 Aw 0.30 R2 = 0.85 n = 26 (4-41) Bankfull width is directly relate d to drainage area, with 76% a nd 85% of variability in width explained by drainage area for northern and southe rn peninsula geographies respectively. For a given drainage area, bankfull width appears to be slightly wider at northe rn peninsula sites than at southern ones. Additionally, northern and southern peninsula streams start out with the same bankfull width (p>0.05), and it increases at the same rate with an increase in drainage area for both geographies (p<0.05) (Table 4-8). Floodplain type (wetland versus upland): Relationships for bankfull width as a function of watershed area for wetland and upland floodplain types are shown in Figure 4-10B. Power function regression equations, corre sponding coefficients of determination, and sample sizes are: Wbkf-WF = 8.61 Aw 0.26 R2 = 0.77 n = 23 (4-42) Wbkf-UP = 6.04 Aw 0.29 R2 = 0.82 n = 22 (4-43) Bankfull width is directly relate d to drainage area, with 77% a nd 82% of variability in width explained by drainage area for wetland and upland floodplain types, resp ectively. For a given drainage area, bankfull width appe ars to be wider at sites with a wetland floodplain than at those with an upland floodplain. Additionally, streams with a wetland floodplain and those with an upland floodplain start out with the same bankf ull width (p>0.05), and bankfull width increases at the same rate with an increase in drainage area for both floodpl ain types (p<0.05) (Table 4-8). Note a cluster of sites with an upland floodplain occurs between a drainage area of 0.1 and one square mile ( Figure 4-10B). Floodplain type (cypressversus non-cypress-dominated): Relationships for bankfull width as a function of drainage area for cypress-domina ted and non-cypress-dominated

PAGE 107

107 floodplain types are shown in Figure 4-11B. Power function regression equations, corresponding coefficients of determina tion, and sample sizes are: Wbkf-CD = 8.56 Aw 0.28 R2 = 0.86 n = 11 (4-44) Wbkf-NC = 6.67 Aw 0.27 R2 = 0.75 n = 34 (4-45) Bankfull width is directly relate d to drainage area, with 86% a nd 75% of variability in width explained by drainage area for cypress-dominat ed and non-cypress-dominated floodplain types, respectively. For a given drainage area, bankfu ll width appears to be wider at sites with a floodplain dominated by cypre ss than at sites with a non-cypress-dominated floodplain. Additionally, streams with a cypress-domina ted floodplain and thos e with a non-cypressdominated floodplain start out with the same bankfull width (p>0.05), and bankfull width increases at the same rate with an increase in drainage area for both floodplain types (p<0.05) (Table 4-8). Note a cluster of sites with a non-cypress-dominated floodplain occurs between a drainage area of 0.1 and one square mile ( Figure 4-11B). Bankfull depth The relationship for bankfull depth as a function of drainage area for the entire data set is shown in Figure 4-7C. Power function regre ssion equation, corresponding coefficient of determ ination (R2), and sample size are: Dbkf = 0.89 Aw 0.18 R2 = 0.48 n = 45 (4-46) where Dbkf = bankfull depth in feet (ft), and Aw = watershed drainage area in square miles (sq mi). Bankfull depth is directly related to draina ge area, with 48% of va riability in depth across the entire study area explained by dr ainage area. Regressions rela ted to mean depth exhibit the lowest R2 values of all regi onal curves developed in the present study. Physiography (flatwoods versus highlands): Relationships for bankfull width as a function of drainage area for flatwoods and high lands physiographies are shown in Figure 4-8C.

PAGE 108

108 Power function regression equations, corresponding coefficients of determination, and sample sizes are: Dbkf-FW = 0.86 A w 0.18 R2 = 0.49 n = 25 (4-47) Dbkf-HL = 0.91 Aw 0.17 R2 = 0.48 n = 20 (4-48) Bankfull depth is directly related to drainage area, with 49% and 48% of variability in depth explained by drainage area for flatwood and hi ghland physiographies, respectively. For a given drainage area, bankfull mean depth appears to be similar at fl atwoods and highlands sites. Additionally, flatwoods and highlan ds streams start out with the same bankfull depth (p>0.05), and bankfull depth increases at the same rate with an increase in drainage area for both physiographies (p<0.05) (Table 4-9). Geography (northern versus southern peninsula): Relationships for ba nkfull depth as a function of drainage area for the northern and southern peninsula geographies are shown in Figure 4-9C. Power function regression equations corresponding coefficients of determination, and sample sizes are: Dbkf-NP = 1.03 Aw 0.19 R2 = 0.67 n = 19 (4-49) Dbkf-SP = 0.80 Aw 0.17 R2 = 0.44 n = 26 (4-50) Bankfull depth is directly related to drainage area, with 67% and 44% of variability in depth explained by drainage area for northern and southe rn peninsula geographies respectively. For a given drainage area, bankfull mean depth appears to be deeper at northern peninsula sites than at the southern ones. Additionally, northern penins ula streams start out with a deeper bankfull depth than southern peninsula streams (p>0.02). Bankfull depth increases at the same rate with an increase in drainage area for bot h geographies (p<0.05) (Table 4-9).

PAGE 109

109 Floodplain type (wetland versus upland): Relationships for bankfull depth as a function of drainage area for wetland and upland floodplain types are s hown in Figure 4-10C. Power function regression equations, corre sponding coefficients of determination, and sample sizes are: Dbkf-WF = 0.95 Aw 0.16 R2 = 0.52 n = 23 (4-51) Dbkf-UP = 0.85 Aw 0.18 R2 = 0.41 n = 22 (4-52) Bankfull depth is directly related to drainage area, with 52% and 41% of variability in depth explained by drainage area for wetland and upland floodplain types, resp ectively. For a given drainage area, bankfull mean dept h appears to be similar at s ites with a wetland floodplain and those with an upland floodplain. Additionally, st reams with a wetland floodplain and those with an upland floodplain start out with the same bankfull depth (p>0.05), and bankfull depth increases at the same rate with an increase in drainage area for both floodplain types (p<0.05) (Table 4-9). A cluster of sites with an upland floodplain occurs between a drainage area of 0.1 and one square mile ( Figure 4-10C). Floodplain type (cypress-dominated versus non-cypress dominated): Relationships for bankfull depth as a function of drainage area for northern and southern pe ninsula geographies are shown in Figure 4-11C. Power function regre ssion equations, corresponding coefficients of determination, and sample sizes are: Dbkf-CD = 0.85 Aw 0.18 R2 = 0.47 n = 11 (4-53) Dbkf-NC = 0.89 Aw 0.18 R2 = 0.45 n = 34 (4-54) Bankfull depth is directly related to drainage area, with 47% and 45% of variability in depth explained by drainage area for cypress-dominat ed and non-cypress-dominated floodplain types, respectively. For a given drainage area, bankfull mean depth appears to be similar at sites with a cypress-dominated floodplain and those with a non-cypress-dominated floodplain. Additionally, streams with a cypress-dominated floodplain a nd those with a non-cypr ess-dominated floodplain

PAGE 110

110 start out with the same bankfull depth (p>0.05), and bankfull depth increases at the same rate with an increase in drainage area for both floodpl ain types (p<0.05) (Table 4-9). A cluster of sites with an upland floodplain occurs between a drainage area of 0.1 and one square mile (Figure 4-11C). Dimensionless Ratios Di mensionless ratios including sinuosity, width-to-depth, maximum depth-to-mean depth, and valley slope were calculated for the 45 sites (Table 4-10). The results are presented and analyzed for both the entire dataset a nd data subsets based on physiography (flatwoods versus highlands), geography (nor thern versus southern peninsul a), and floodplain type (wetland versus upland and cypress-dominated versus non-cypress-dominated). Boxplots for the various ratios are provided in Figures 4-12 through 4-15. Sinuosity Sinuosity averages 1.32 across a ll the sites. No significant differences exist (p>0.05) in sinuosity based on flatwoods versus physiography, northern ve rsus southern peninsula geography, wetland versus upland floodplain type, or cypress-dom inated versus non-cypressdominated floodplain type (T able 4-11, Figure 4-12). Width-to-depth W idth-to-depth ratio averages 11.11 across all the sites. No significant differences exist (p>0.05) in the width-to-depth ratio based on flatwoods versus highl ands physiography or wetland versus upland floodplain type. However, southern sites and those with a cypressdominated floodplain had significantly greater width-to-depth ratios than northern sites (p=0.01) and sites with a non-cypress-dominated floodplai n (p=0.01), respectively (Table 4-11, Figure 413).

PAGE 111

111 Maximum depth-to-mean depth Maxim um depth-to-mean depth ratio averages 1.62 across all sites. No significant differences exist (p>0.05) in the maximum dept h-to-mean depth ratio based on flatwoods versus highlands physiography, northern versus sout hern peninsula geogra phy, wetland versus upland floodplain, or cypress-dominated versus non-cypr ess-dominated floodplain (Table 4-11, Figure 4-14). Valley slope Valley s lopes average 0.41% across all the si tes. No significant differences exist (p>0.05) in valley slope based on flatwoods versus highlands phys iography or northern versus southern peninsula geography. However, sites with an upland floodplain and those with a noncypress-dominated floodplain had significantly greater valley slopes than sites with a wetland floodplain (p<0.01) and sites with a cypress-dominated floodplain (p =0.02), respectively (Table 4-11, Figure 4-15). Return Intervals Return intervals were estim at ed using Annual Maximum Series from a Log Pearson Type III distribution and ranged from less than one ye ar to 1.44 years (Table 4-2), which is more frequent than the average 1.5-year return interv al often reported in the literature (Dunne and Leopold, 1978; Leopold, 1994), but c onsistent with findings from other southeastern United States Coastal Plain studies (Sw eet and Geratz, 2003) (Table 4-12) An average bankfull return interval could not be determined for the sites most gaged sites had a return interval of less than one year3. Bankfull discharge was equaled or exceeded on average 21% of the time, or 75 days per year, for gaged sites based on flow duration curve analysis. 3 Note that return intervals less than one year cannot be determined when performing an Annual Maximum Series.

PAGE 112

112 Comparison to Other Southeastern United States Coastal Plain Regional Curves Regional curves have recently been developed to estimate bankfull discharge and channel geometry throughout the southeastern United States Costal Plain, including the Georgia Coastal Plain (Buck Engineering, 2004), Virginia and Maryland Coastal Plain (Krstolic and Chaplin, 2007), North Carolina Coastal Plain (Doll et al., 2003; Sweet and Geratz, 2003), Northwest Florida and North Florida Coastal Plain (Metca lf, 2004), and Alabama Co astal Plain (Metcalf, 2005). Results of these studies as well as the present study are compiled in Table 4-12, and the regressions are compiled in Figur es 4-16 through 4-19. The slope of peninsular Florida streams for all bankfull regressions tends to be less steep than the other slopes, indicating that bankfull parameters in peninsular Florida are less sensitiv e to changes in drainage area size, or in other words that the bankfull parameters in peninsular Florida streams increase at a slower rate with drainage area. When other Coastal Plain ba nkfull regressions were compared through ANCOVA testing to peninsular Florida ba nkfull regressions (the baseline regression), the following results were found: Georgia, North Carolina (Sweet and Geratz 2003), and North Florida Coastal Plain streams start out with a significantly lower bankfull discharge than peninsular Florida streams (p<0.01, p<0.01, and p<0.01, respectively), while Alabama, North Carolina (Doll, 2003), Northwest Florida, and Virginia/Marylan d Coastal Plain streams start out with a significantly higher bankfull discharge (p=0.02, p=0.01, p<0.01, and p<0.01, respectively). Bankfull discharge for all stream sets increase s at the same rate with increasing drainage area as peninsular Florida streams (p>0.05) (Table 4-5, Figure 4-16). Georgia and North Florida Coastal Plain stre ams start out with a significantly smaller bankfull area than peninsular Florida st reams (p<0.01 and p=0.02, respectively), while North Carolina (Doll, 2003), Northwest Flor ida, and Virginia/Maryland Coastal Plain streams start out with a significantly larger bankfull area (p<0.01, p<0.01, and p=0.01, respectively). Bankfull area in Alabama stream s increased at a signifi cantly faster rate with increasing drainage area than peninsular Florida streams (p=0.01), while the other regions increased at the same rate (p>0.05) (Table 4-7, Figure 4-17). Both North Carolina stream sets (Sweet and Geratz, 2003; Doll et al., 2003), Northwest Florida, and Virginia/Maryla nd Coastal Plain streams start out significantly wider than peninsular Florida streams (p=0.03, p<0.01, p=0.01, and p=0.01, respectively). Alabama

PAGE 113

113 streams widened at a significantl y faster rate with increasing drainage area than peninsular Florida streams (p=0.03), while North Florida streams widened at a significantly slower rate (p=0.01). Other Coastal Pl ain regions streams wi dened at the same rate as peninsular Florida streams (Table 4-8, Figure 4-18). Georgia Coastal Plain streams start out significantly shallower than peninsular Florida streams (p<0.01), while North Carolina (Doll, 2003) and Northwest Florida streams started out significantly deeper (p=0.01 and p<0.02, respectively). North Florida streams deepened at a faster rate with increasing dr ainage area than penins ular Florida streams (p=0.04) (Table 4-9, Figure 4-19). Discussion In this study, regional curves relating bankfull discharge, bankfull cross-sectional area, bankfull width, and bankfull m ean depth to draina ge area were developed for peninsular Florida streams. Regional curve data, as well as various dimensionless ratios (s inuosity, width-to-depth, maximum depth-to-mean depth, valley slope), were analyzed to determine whether significant differences exist between streams draining different physiographies (flatwoods versus highlands), geographies (north ern versus southern peninsul a), and floodplain types (wetland versus upland and cypress-dominated versus noncypress dominated). The return interval associated with bankfull discharge was also esti mated for peninsular Florida streams. Lastly, regional curves developed in this study were co mpared to those developed for other regions of the southeastern United States Coastal Plain. The discussion begins with an examination of potential sources of error involved in developing regional curves and im plications this could have on interpretation of data. An examination of the analyses conducted on each parameter and dimensionless ratio follows. Observed tre nds and anomalies are discussed, and potential explanations are presented. Inte rpretation of data is presented as it relates to achieving the objective of this chapter, which is to develop the most r obust regional curves for peninsular Florida streams.

PAGE 114

114 Regional Curve Development As previously m entioned in the methods s ection, bankfull channel geometry parameters were based on the average value of the two sma llest cross-sections (b ased on cross-sectional area) surveyed during the referenc e reach survey conducted at each st udy site. It is important to note, however, that six detailed cr oss-sections were surveyed for ea ch stream, and as can be seen in Table 4-2, the range of variability within bankfull indicator paramete rs among cross-sections was highly variable. For example, the range of variability among cross-sections (maximum bankfull measurement minus minimum bankfull m easurement) was as high as 187 square feet for bankfull area at Horse Creek near Arcadia, 50 feet for bankfull width at Tiger Creek near Babson Park, and approximately three feet in bank full depth at Horse Creek near Arcadia. The average range of variability am ong all sites was 24 square feet for bankfull area, 11 feet for bankfull width, and eight tenths of a foot for bankfull mean depth. Further, Wolman (1955) recognized that local variations in cross-sec tional form are a possibl e source of scatter in downstream hydraulic geometry relations. Clearly, the cross-section chosen for development of the regional curves can have a significant effect on the ultimate regression. The two-smallest cross-sections were thus ultimately chosen and their parameters averaged for use in development of peninsular Florida regional curves based on previous work by USGS and based on the notion that the smallest cross-section represents the streams hydraulic cont rol (Chaplin, 2005). As previously mentioned in the methods section, bankfull discharge and stage were estimated for gaged sites by using reference r each survey data of field bankfull stage in conjunction with the most current USGS Stage-Q ra ting table. As discussed in Chapter 3, it is important to note that determina tions of bankfull disc harge and stage are rough estimates, as the reference reach survey was not always conducted ex actly at the USGS gage station due to local

PAGE 115

115 effects of bridges on stream hydraulics. Table 31 gives the location of the gage compared to that of the referen ce reach survey. Discharge: bankfull, mean annual, 1.5-year Several im portant discussion topics arose when examining discharge data, each of which will be briefly discussed herein. First, mean annual discharge had noticeably high R2 values across the board in comparison to R2 values of bankfull discharge and 1.5-year discharge regressions, indicating that draina ge area is a good predictor of mean annual yield but is not as robust a predictor of bankfull discharge (as define d by field indicators) or 1.5-year discharge. Perhaps the concept of return interval is not as important to Florida stream hydrology as to other regions of the United States. Second, bankfull discharge was almost equal to or less than mean annual discharge at six of the gaged sites, which was unexpected as bank full discharge is expected to be higher than mean annual discharge (Leopold, 1994 see Figur e 2-15). A mean annual discharge above bankfull discharge indicates that on average the st ream is over its banks. The bankfull stage, however, was greater than mean annual stage at all but one gaged sites (Catfish Creek near Lake Wales), indicating that on average the water surface elevation is below the banks. It is interesting to note that all six sites had a wetland floodplain, and so perhaps it is not unfathomable for mean annual discharge to, on av erage, actually be over the banks as these wetland systems can withstand flooding. It is also interesting to note th at Catfish Creek near Lake Wales had a bankfull stage less than its m ean annual stage because this stream drains a large lake, which may provide a cons tant source of water to the stre am so it flows at or above the bankfull stage most of the time and is not as dependent on individual storm events. However, the discrepancy between stage and discharge data s till remains, which may be due to issues with USGS gage data and estimates of bankfull st age and discharge (as discussed thoroughly in

PAGE 116

116 Chapter 3). Perhaps stage trumps discharge, or vi ce versa, in peninsular Florida streams, and one or the other is more useful in understanding the concept of bankfull. These issues would be interesting to research further. Third, general trends in bankfull discharge re gional curves show that it is generally higher at: 1) flatwoods sites than at highlands sites, 2) sites wi th an upland floodplain than those with a wetland floodplain; and 3) sites with a non-cypress-dominated th an cypress-dominated floodplain. Sites with an upla nd floodplain and sites with a non-cypress-dominated floodplain have a significantly higher valley slope (as pr esented in the Dimensionl ess Ratios: Valley slope results section) than do sites with a wetland floodplain and sites with a cypress-dominated floodplain, respectively. Perhaps these steeper slopes allow for higher velocities in these streams, and subsequently higher bankfull discha rges. Because no significant differences were found in valley slopes of flatwoods versus highlands sites, perhaps ot her factors such as vegetation and soils are responsib le for differences in bankfull discharge based on physiography. For example, highland soils are less cohesive, which perhaps leads to more erratic stream behavior that may help to explain why R2 values for all highlands regi onal curves are lower than those for flatwoods curves. Lastly, the peak discharge-to -mean annual discharge ratio was significantly higher at flatwoods than at highlands site s (p=0.01), which indicates that flatwoods streams are flashier (Table 4-6). However, it is important to note th at sample size for highlan ds sites was relatively small (n=5). Boxplots of peak discharge-to-mean annual discharge ratios are provided in Figure 4-20. Bankfull channel geometry Several im portant discussion topics arose when examining bankfull channel geometry data (cross-sectional area, width, and mean dept h), each of which will be briefly discussed

PAGE 117

117 herein. First, R2 values of bankfull mean depth regressi ons were noticeably lower than those for bankfull area and bankfull width. This could be due to the fact that the streambed is highly variable, and one may survey a pool versus a riffle, which could significantly affect the results. Second, when examining differences in fl oodplain types (wetland versus upland and cypress-dominated versus non-cypress-dominated), an obvious cluster of sites with an upland floodplain or a non-cypress-dominated floodplain occu rs between a drainage area of 0.1 and one square mile. This is not surpri sing, as sites with smaller drainage areas (i.e., headwater streams) tend to have steeper valley slopes (Figure 4-1). Wetlands, however, tend to occur in areas with lower valley slopes. Third, bankfull area and width appeared to be larger in streams with wetland floodplains and with cypress-dominated floodplains than in streams with upland fl oodplains and with noncypress-dominated floodplains, re spectively. Though not significan tly different, peak dischargeto-mean annual discharge ratios were also higher in streams with wetland floodplains and with cypress-dominated floodplains, i ndicating that these streams are flashier than those with an upland floodplain or a non-cypres s-dominated floodplain (Table 46, Figure 4-20). Flashiness of these streams may help explain why they are wide r, as Osterkamp (1980) found that streams with a flashier regime and relatively high peak flows tend to develop wider channels. Dimensionless Ratios Several im portant discussion topics aros e when examining dimensionless ratios (sinuosity, width-to-depth, maximum depth-to-mean depth, and valley slope), each of which will be briefly discussed herein. First, no significant differences were found in sinuosity based on physiography, geography, or floodplai n type, indicating that stream pattern in peninsular Florida is similar for the stream types studied. Nor were there significant differences in maximum

PAGE 118

118 depth-to-mean depth ratios, indicating that channel shape, in terms of dmax/d, are similar for the stream types studied. Second, southern peninsula sites and thos e with a cypress-do minated floodplain had significantly larger width-to-depth ratios, indicating that they ar e wider for a given depth than northern peninsula sites and sites with a non-c ypress-dominated floodplain. Differences in southern versus northern peninsula streams may be due to land use practices or vegetation. Sites with a cypress-dominated floodplain may have a larg er width-to-depth rati o due to the presence of cypress knees, which may prevent stream banks from becoming as well developed. Third, sites with an upland floodplain and t hose with a non-cypress-dominated floodplain had significantly greater valley slopes than si tes with a wetland floodpl ain and sites with a cypress-dominated floodplain. This is not surprising, as wetlands tend to develop where valley slope is flatter. Return Intervals Return intervals were estim at ed using Annual Maximum Series from a Log Pearson Type III distribution and ranged from less than one ye ar to 1.44 years (Table 4-2), which is more frequent than the average 1.5-year return interv al often reported in the literature (Dunne and Leopold, 1978; Leopold, 1994), but c onsistent with findings from other southeastern United States Coastal Plain studies (S weet and Geratz, 2003) (Table 4-12). These findings have important implications for flood control, as they indicate that peninsular Florida streams are overtopping their banks more fr equently than in other regi ons. Because Annual Maximum Series cannot determine return intervals less th an one year, mean and median bankfull return interval values could not be determined for peninsular Florida streams. It is recommended that future work includes a partial duration series to refine return intervals less than one year.

PAGE 119

119 Comparison to Other Southeastern United States Coastal Plain Stud ies Several important discussion topics arose when comparing regional curves developed for peninsular Florida streams to regional curves de veloped for other region s of the southeastern United States Coastal Plain, each of which will be briefly discussed herein. First, peninsular Florida bankfull channels start out significantl y narrower and shallower than North Carolina (p<0.01; p=0.01) and Northwest Florida (p=0.01; p<0.01) Coastal Plain streams (Table 4-8 and Table 4-9). Perhaps this indi cates that peninsular Florida streams are more efficient at conducting water, as Peninsular Florida streams tend to be lo w gradient with sandy bottoms, which may enable them to conform and conduct water more easily than streams with steeper gradients and rocky streambeds. However, peni nsular Florida streams also start out at a significantly lower bankfull discharge and area than North Carolina (p=0.01; p<0.01) and Northwest Florida (p<0.01; p=0.02) Coastal Plain streams, which could indicate that peninsular Florida streams receive less water overall (Table s 4-6 and 4-7). Peninsular Florida receives approximately ten inches less rain than Northw est Florida (Figure 2-5) and considerably less mean annual runoff (approximately 10 inches) th an North Carolina (approximately 15 inches) and Northwest Florida (approximately 25 inches) (Figure 4-21) (Gerbert et al., 1987). Peninsular Floridas low mean a nnual runoff values are likely attri butable to its sandy soils, flat terrain, and deranged drainage networks. Peni nsular Florida streams also deepen at a significantly faster rate with incr easing drainage area than North Flor ida streams. Perhaps this is because North Florida streams have a steeper grad ient and thus down-cut more. Overall, it is difficult to tweak out exactly why peninsular Fl orida streams are significantly different than other Coastal Plain streams without further research, as there are a variety of variables, including the amount of water these systems receives (w hich depends upon on various factors such as

PAGE 120

120 climate, rainfall patterns, runoff patterns, and baseflow), roughness of the streambed, gradient, level of alluvial co ntrol, and vegetation. Conclusions In this study, regional curves were developed for peninsular F lorida streams. Bankfull discharge and channel geometry (cross-sectional area, width, and mean depth), which were determined from both USGS hydrologic data an d reference reach surveys, mean annual discharge, and the 1.5-year discharge were plotted against drainage area, and coefficients of determination (R2) were determined. Relationships for bankfull discharge, mean annual discharge, 1.5-year discharge, bankfull area, bankfull width, and bankfu ll mean depth are shown in Figures 4-2 through 4-11. Table 4-1 summarizes discharge data used in peninsular Florida regional curve development, while Table 4-2 summa rizes channel geometry data. Table 4-3 and 4-4 summarize power function regression equations, corresponding coefficients of determination, and sample sizes for discharge against drainage area and channel geometry against drainage area, respectively. Bankfull parameters and various discharges vari ed directly with drainage area. Bankfull mean depth had the lowest R2 values, while mean annual discharge had the highest R2 values. Data were further analyzed to determine whether significant differences exist between streams draining different physiographies (flatwoods versus highlands), geographies (northern versus southern peninsula), and floodplai n types (wetland versus upland and cypress-dominated versus non-cypress-dominated), in terms of bankfull parameters and various dimensionless ratios (sinuosity, width-to-depth, maximum depth-to -mean depth, and valley slope). Bankfull discharge appears to be higher at flatwoods sites than highlands, at sites with upland floodplains than wetland floodplains, and at sites with non-cypress-dominated fl oodplains than cypressdominated-floodplains (Figures 4-3, 4-5, and 4-6). Sites with a non-cypress-dominated

PAGE 121

121 floodplain started out with signi ficantly higher bankfull discharg e than those with a cypressdominated floodplain (p=0.05) a nd have a significantly lowe r bankfull discharge duration (p=0.01) (Table 4-6). Flatwoods streams were fl ashier than highlands streams, based on having significantly higher maximum discharge to mean discharge ra tios (p=0.01). Sites with wetland floodplains and cypress-dominated floodplain s appear to have a greater bankfull area and bankfull width than sites with upland floodplains and non-cypress-dominated floodplains, respectively (Figures 4-10 and 4-11). These sites w ith a wetland floodplain started out with a significantly higher bankfull area (p=0. 03) and bankfull width (p<0.01) than sites with an upland floodplain (Tables 4-7 an d 4-8). Bankfull mean depth st arted out significantly higher in northern peninsula streams than in southern peninsula streams (p=0.02) (Table 4-9). No significant differences were found in sinuosity or maximum depth-to-mean depth based on physiography, geography, or floodplai n types. Sites with a cypress-dominated floodplain and sites located in the southern peninsula had sign ificantly higher width-to-depth ratio than sites with either a non-cypress-dominated floodplain (p=0.01) or those located located in the northern peninsula (p=0.01) (Table 4-11). Sites with upland floodplains and non-cypress dominated floodplains had significantly steep er valley slopes than sites w ith wetland floodplains (p<0.01) and cypress-dominated floodplai ns (p=0.02), respectively. The return interval associated with the bankfu ll discharge was also estimated for peninsular Florida streams using the Annual Maximum Series from a Log Pearson Type III distribution and ranged from less than one year to 1.44 years (Table 4-2), whic h is more frequent than the average 1.5-year return interval often reporte d in the literature (Dunne and Leopold, 1978; Leopold, 1994), but consistent with findings from other southeastern United States Coastal Plain studies to which regional curves developed in this study were compared. Bankfull discharge,

PAGE 122

122 area, width, and depth started out significantly smaller in peninsular Florida streams than in Northwest Florida streams and No rth Carolina streams, which receive considerably higher mean annual runoff (Figure 4-21). Further, the slope of peninsular Florida streams for all bankfull regressions tends to be less steep than the other slope s, indicating that bankfull parameters in peninsular Florida are le ss sensitive to changes in drainage ar ea size, or in other words that the bankfull parameters in peninsular Florida streams increase at a slower ra te with drainage area.

PAGE 123

123Table 4-1. Discharge data used in penisular Florida regional curve development and analysis Site name Period of record (WY) Physiography Geography Floodplain type Qbkf(cfs) Qma(cfs) Q1.5(cfs) Qp(cfs) Qbkf(% of time) Qma (% of time) Q1.5(% of time) Q b kf/QmaQbkf/Q1.5 (%) Q p /QmaBlackwater Creek near Cassia 81-07 HLNWFC555817380832316.91.0**3214 Blues Creek near Gainesville 85-94 FWNUP363.5881470.8250.2104142 Bowlegs Creek near Ft Meade65-68/92-07FWSWF3532288145021231.21.1**1245 Carter Creek near Sebring 55-67/92-07HLSUP4224.410935214341.71.73914 Catfish Creek near Lake Wales 48-07 HLSWFC3741602355043170.9**625.8 Fisheating Creek at Palmdale 32-07 FWSWFC7525619343050041242.20.3**4119 Hickory Creek near Ona 82-84*FWSWF215.31164906.5170.54.01893 Horse Creek near Arcadia 51-07 FWSWF28019613661070018242.31.42055 Little Haw Creek near Seville 52-06 FWNWFC11486341181024295.51.33321 Livingston Creek near Frostproof92-07 HLSUP1066418270018327.31.75811 Lochloosa Creek at Grove Park96-05*FWNWFC132152323823181.00.6**25157 Manatee River near Myakka Head67-07 FWSUP116741341644014200.41.6987 Moses Creek near Moultrie 00-02*FWNWFC438.31328614.3141.05.233103 Rice Creek near Springside 74-04 FWNWFC2542513200033230.60.6**547 Santa Fe River near Graham 57-98 FWNUP11852293187012282.62.34036 Shiloh Run near Alachua 84-87*FWNUP110.3105.5IR28IR3811019 Tiger Creek near Babson Park 92-07 HLSUP644311533816353.91.5567.8 Minimum 110.3105.50.8140.20.33.95.8 Maximum 28025619343050050431738110157 Mean 7059418364421263.44.33552 Median 434217386118251.91.53342 Notes: WY = Water year; cfs = cubic feet per second; Qbkf = Bankfull discharge; Qma = Mean annual discharge; Q1.5 = Discharge t hat occurs on average once every 1.5 years; Qp = Peak discharge; IR = Insufficient gage record; FW = Flatwoods physiography; HL = Highlands physiography; N = Northern peninsula geography; S = South ern peninsula geography; WF = Wetland floodplain; WFC = Wetland floodplain dominated by cypress; UP = Upland floodplain; Period of gage record insufficient (less than 10 years) for proper g age analysis, but rough approximations are presented; ** Bankfull discharge approximately equal to or less the mean annual discharge (unexpected result) Discharge Ratios Data Subsets Discharge Duration

PAGE 124

124Table 4-2. Reference reach survey da ta used in penisular Florida regional curve developm ent and analysis Site name Physiography Geography Floodplain type Drainage area (sq mi) Valley slope (%) Bankfull discharge (cfs) Bankfull area (sq ft) Bankfull width (ft) Bankfull mean depth (ft) Bankfull indicator Bankfull area (sq ft) Bankfull width (ft) Bankfull mean depth (ft) Alexander Springs Creek tributary 2HLNUP1.61.042---6.826.631.06I5.083.551.08 Blackwater Creek near Cassia HLNWFC1260.020551.05101.6137.072.75F23.3922.080.74 Blues Creek near Gainesville FWNUP2.60.206361.0712.777.671.67I4.243.730.69 Bowlegs Creek near Ft Meade FWSWF47.20.05035<1.0142.6622.901.87F27.3821.460.54 Carter Creek near Sebring HLSUP38.80.23742<1.0126.6216.061.66I53.1116.491.08 Catfish Creek near Lake Wales HLSWFC58.90.050371.1151.2543.221.19F35.6148.450.58 Coons Bay Branch FWSWF0.50.348---6.926.571.06F5.193.290.39 Cow Creek FWNWFC5.30.080----18.0513.871.31F18.105.840.81 Cypress Slash tributary HLSUP0.51.042---1.254.540.27I2.031.610.30 East Fork Manatee River tributary FWSUP0.20.313---5.015.181.00I3.904.360.64 Fisheating Creek at Palmdale FWSWFC3110.02975<1.0162.5337.641.67F84.8720.801.59 Gold Head Branch HLNUP1.81.316---8.045.811.40I3.555.220.45 Hammock Branch HLNWF3.00.167----15.9110.951.46F25.747.111.75 Hickory Creek near Ona FWSWF3.750.11621<1.018.9210.330.85F19.1322.840.42 Hillsborough River tributary FWSWFC0.70.260---5.108.110.63F13.536.530.70 Horse Creek near Arcadia FWSWF2180.043280<1.0187.2633.232.61F187.0727.322.97 Jack Creek HLSWF5.20.286---7.007.880.89F8.903.000.60 Jumping Gully HLNUP4.61.111---3.584.190.87I4.371.630.66 Lake June-In-Winter tributary FWSUP0.40.781---3.695.320.66I5.233.690.61 Little Haw Creek near Seville FWNWFC930.0611141.0583.8531.342.69F36.7510.791.29 Livingston Creek near Frostproof HLSUP1200.0641061.1444.6927.511.73I69.9019.781.53 Livingston Creek tributary HLSUP0.40.250---3.324.100.81I1.220.690.15 Lochloosa Creek at Grove Park FWNWFC7.40.11613<1.0115.7315.581.02F14.144.780.62 Lowry Lake tributary HLNUP0.250.625---3.654.220.86I2.022.630.28 Manatee River near Myakka Head FWSUP65.30.116116<1.0160.0524.522.45I59.2515.020.92 Manatee River tributary FWSUP0.31.163--8.245.391.61I8.034.711.03 Morgan Hole Creek FWSUP9.40.091----14.219.611.50I10.827.290.56 Moses Creek near Moultrie FWNWFC7.40.159431.1131.5014.592.16F37.856.131.77 Myakka River tributary 1 FWSUP2.60.091---3.609.740.37I7.866.360.36 Myakka River tributary 2 FWSUP1.70.129---1.884.930.39I2.633.390.17 Nine Mile Creek HLNWF160.488----10.289.191.12F6.274.590.41 Rice Creek near Springside FWNWFC43.20.04125<1.0131.9720.471.60F30.798.981.15 Santa Fe River near Graham FWNUP94.90.0581181.1051.9817.643.02I35.4311.221.13 Shiloh Run near Alachua FWNUP0.322.000111.443.045.180.59I4.235.010.22 Snell Creek HLSWF1.70.167----22.6917.801.38F14.7613.610.61 South Fork Black Creek HLNWF25.50.110----45.0917.412.59F23.8111.100.61 Spoil Bank tributary (Highlands) FWSUP8.60.313----14.1114.191.03I10.759.330.55 Ten Mile Creek FWNWFC250.130----26.4015.511.71F23.278.330.51 Tiger Creek near Babson Park HLSUP52.80.081641.0765.8933.911.92I74.4149.990.68 Tiger Creek tributary HLSWF0.90.139---7.2610.130.80F8.0017.430.57 Triple Creek unnamed tributary 1 HLSWF1.70.532---9.028.001.11F10.765.000.64 Triple Creek unnamed tributary 2 FWSUP0.20.885---2.694.870.54I9.323.501.03 Tuscawilla Lake tributary HLNUP0.32.273---3.073.091.00I1.461.640.73 Tyson Creek FWSWFC20.50.054----20.2319.021.06F32.9214.010.63 Unnamed Lower Wekiva tributary HLNWF0.40.769---8.048.300.95F13.526.240.76 Notes: FW = Flatwoods physiography; HL = Highlands physiography; N = Northern peninsula geography; S = Southern peninsula geogr aphy; WF = Wetland floodplain; WFC = Wetland floodplain dominated by cypress; UP = Upland floodplain; cfs = cubic feet per second; ft = feet; sq ft = square feet; yrs = years; sq mi = square miles; -= Ungaged site; IR = Insufficient gage record; F = Flat floodplain bankfull indicator; I = Inflection bankfull indicator; N/A = Not applicable Range of variability within bankfull indicator among cross-sections Return interval (yrs) Data Subsets Independent VariablesBankfull DischargeBankfull Channel Geometry

PAGE 125

125Table 4-3. Regression equations for various discharges against drainage area by entire data se t and by subsets representing physiography, geography, and floodplain types EquationR2nE q u a t i o nR2n EquationR2n Entire data setQ b kf =14.26 Aw 0.360.6017 Qma = 1.36 Aw 0.880.9517Q1.5 = 27.85 Aw 0.570.6017 Physiography: FlatwoodsQ b kf-FW =14.47 Aw 0.380.6412Qma-FW = 1.35 Aw 0.920.9612Q1.5-FW = 28.65 Aw 0.690.8612 HighlandsQ b kf-HL = 6.97 Aw 0.490.395Qma-HL = 2.55 Aw 0.670.865Q1.5-HL = 10.20 Aw 0.580.455 Geography: Northern peninsulaQ b kf-NP = 15.98 Aw 0.320.588Qma-NP = 1.37 Aw 0.880.948Q1.5-NP = 29.79 Aw 0.530.788 Southern peninsulaQ b kf-SP = 9.42 Aw 0.470.579Qma-SP = 1.30 Aw 0.880.959Q1.5-SP = 22.27 Aw 0.630.379 Floodplain Types: WetlandQ b kf-WF = 9.13 Aw 0.440.5510Qma-WF = 2.16 Aw 0.780.9210Q1.5-WF = 28.49 Aw 0.580.5210 UplandQ b kf-UP = 18.64 Aw 0.360.887Qma-UP = 1.04 Aw 0.910.987Q1.5-UP = 27.24 Aw 0.530.657 Cypress-dominated Q b kf-CD = 10.94 Aw 0.350.487Qma-CD = 2.87 Aw 0.720.877Q1.5-CD = 19.88 Aw 0.620.487 Non-cypress-dominatedQ b kf-NC = 15.49 Aw 0.410.7910Qma-NC = 1.12 Aw 0.910.9810Q1.5-NC = 30.49 Aw 0.560.6810 Notes: Qbkf = Bankfull discharge; Qma = Mean annual discharge; Q1.5 = Discharge that occurs on average every 1.5 years; Aw = Watershed drainage area (sq mi); FW = Flatwoods physiography; HL = Highlands physiography; NP = Northern peninsula geography; SP = Southern peninsula geography; WF = Wetland floodplain; UP = Upland floodplain; CD = Cypress-dominated floodplain; NC = Non-cypress-dominated floodplain Bankfull discharge (cfs) Mean annual discharge (cfs) 1.5-year discharge (cfs)

PAGE 126

126Table 4-4. Regression equations for bankfull channel geometry against drainage area by entire data set and by subsets represen ting physiography, geography, and floodplain types EquationR2n EquationR2nE q u a t i o nR2n Entire data setA b kf = 6.05 Aw 0.470.7845W b kf = 6.87 Aw 0.300.8145D b kf = 0.89 Aw 0.180.4845 Physiography: Flatwoods A b kf-FW = 6.27 Aw 0.460.8225W b kf-FW = 7.28 Aw 0.280.9225D b kf-FW = 0.86 Aw 0.180.4925 Highlands A b kf-HL = 5.80 Aw 0.490.7420W b kf-HL = 6.43 Aw 0.330.7220D b kf-HL = 0.91 Aw 0.170.4820 Geography: Northern peninsula A b kf-NP = 6.41 Aw 0.490.8019W b kf-NP = 6.26 Aw 0.300.7619D b kf-NP = 1.03 Aw 0.190.6719 Southern peninsula A b kf-SP = 5.78 Aw 0.460.7826W b kf-SP = 7.32 Aw 0.300.8526D b kf-SP = 0.80 Aw 0.170.4426 Floodplain Types: Wetland A b kf-WF = 8.11 Aw 0.410.7923W b kf-WF = 8.61 Aw 0.260.7723D b kf-WF = 0.95 Aw 0.160.5223 Upland A b kf-UP = 5.13 Aw 0.470.7522W b kf-UP = 6.04 Aw 0.290.8222D b kf-UP = 0.85 Aw 0.180.4122 Cypress-dominated A b kf-CD = 7.29 Aw 0.460.8411W b kf-CD = 8.56 Aw 0.280.8611D b kf-CD = 0.85 Aw 0.180.4711 Non-cypress-dominatedA b kf-NC = 5.90 Aw 0.450.7334W b kf-NC = 6.67 Aw 0.270.7534D b kf-NC = 0.89 Aw 0.180.4534 Notes: Abkf = Bankfull area; Wbkf = Bankfull width; Dbkf = Bankfull mean depth; Aw = Watershed drainage area; FW = Flatwoods physiography; HL = Highlands physiography; NP = Northern peninsula geography; SP = Southern peninsula geography; WF = Wetland floodplain; UP = Upland floodp lain; CD = Cypressdominated floodplain; NC = Non-cypress-dominated floodplain Bankfull Area (sq ft)Bankfull Width (ft)Bankfull Mean Depth (ft)

PAGE 127

127 Table 4-5. Comparison of bankfu ll discharge against drainage ar ea regressions by physiography, geography, floodplain types, a nd Coastal Plain regions Effect EstimateR2P-value Estimate (cfs)R2P-value Physiography: Flatwoods 0.380.640.85 14.470.640.52 Highlands 0.490.39 6.970.39 Geography: Northern peninsula 0.320.580.46 15.980.580.95 Southern peninsula 0.470.57 9.420.57 Floodplain Types: Wetland 0.440.550.40 9.130.550.55 Upland 0.360.88 18.630.88 Cypress-dominated 0.350.480.74 10.940.480.05* Non-cypress-dominated 0.410.79 15.490.79 Coastal Plain Studies: Peninsular FL (Blanton, 2008) 0.360.60Baseline 14.260.60Baseline North FL (Metcalf, 2004) 0.780.920.32 7.510.920.01* Northwest FL (Metcalf, 2004) 0.710.950.88 27.480.95<0.01* AL (Metcalf, 2005) 0.940.930.08 10.950.930.02* GA (Buck Engineering, 2004) 0.780.880.26 6.730.88<0.01* NC (Doll et al., 2003) 0.700.870.97 18.280.900.01* NC (Sweet & Geratz, 2003) 0.710.850.89 9.320.92<0.01* VA & MD (Krstolic & Chaplin, 2007)0.620.790.18 26.650.79<0.01* Represents statistical significance (p 0.05) Slope Intercept

PAGE 128

128 Table 4-6. Comparison of vari ous discharge durations and ratios by physiography, geography, floodplain types, and Coastal Plain regions Effect AverageP-valueAverageP-valueAverageP-valueAverageP-value Physiography: Flatwoods 180.28 23<0.01* 1.60.01* 690.01* Highlands 26 35 7.4 11 Geography: Northern peninsula 180.60 250.35 2.50.50 490.78 Southern peninsula 22 28 4.1 55 Floodplain Types: Wetland 250.06 250.24 3.80.62 660.11 Upland 13 29 2.7 31 Cypress-dominated 300.01* 260.86 4.90.23 670.25 Non-cypress-dominated13 27 2.2 41 Notes: Qbkf = Bankfull discharge; Qma = Mean annual discharge; Q1.5 = Discharge that occurs on average once every 1.5 years; Qp = Peak discharge; Represents statistical significance (p 0.05) Qbkf duration (% of time exceeded) Qma duration (% of time exceeded) Q1.5 duration (% of time exceeded) Qp/Qma

PAGE 129

129 Table 4-7. Comparison of bankfull area agains t drainage area regressions by physiography, geography, floodplain types, a nd Coastal Plain regions Effect EstimateR2P-value Estimate (sq ft)R2P-value Physiography: Flatwoods 0.460.820.68 6.270.820.89 Highlands 0.480.74 5.800.74 Geography: Northern peninsula 0.490.800.75 6.410.800.39 Southern peninsula 0.460.78 5.780.78 Floodplain Types: Wetland 0.410.790.48 8.110.790.03* Upland 0.470.75 5.130.75 Cypress-dominated 0.460.840.99 7.290.840.39 Non-cypress-dominated 0.450.73 5.900.73 Coastal Plain Studies: Peninsular FL (Blanton, 2008) 0.470.78Baseline 6.050.78Baseline North FL (Metcalf, 2004) 0.700.980.96 6.400.980.02* Northwest FL (Metcalf, 2004) 0.640.990.35 17.390.99<0.01* AL (Metcalf, 2005) 1.000.980.01* 4.360.980.88 GA (Buck Engineering, 2004) 0.720.960.63 5.920.96<0.01* NC (Doll et al., 2003) 0.660.880.47 14.330.88<0.01* NC (Sweet & Geratz, 2003) 0.710.960.79 9.570.960.08 VA & MD (Krstolic & Chaplin, 2007)0.660.950.47 11.610.950.01* Represents statistical significance (p 0.05) Slope Intercept

PAGE 130

130 Table 4-8. Comparision of ba nkfull width against drainage area regressions by physiography, geography, floodplain types, a nd Coastal Plain regions Effect EstimateR2P-value Estimate (ft)R2P-value Physiography: Flatwoods 0.280.920.28 7.280.920.66 Highlands 0.330.72 6.430.72 Geography: Northern peninsula 0.300.760.90 6.260.760.13 Southern peninsula 0.300.85 7.320.85 Floodplain Types: Wetland 0.260.770.43 8.610.77<0.01* Upland 0.290.82 6.040.82 Cypress-dominated 0.280.860.91 8.560.860.06 Non-cypress-dominated 0.270.75 6.670.75 Coastal Plain Studies: Peninsular FL (Blanton, 2008) 0.300.81Baseline 6.870.81Baseline North FL (Metcalf, 2004) 0.260.850.01* 9.820.850.06 Northwest FL (Metcalf, 2004) 0.380.960.77 10.810.960.01* AL (Metcalf, 2005) 0.520.940.03* 5.640.940.39 GA (Buck Engineering, 2004) 0.350.840.55 8.580.840.15 NC (Doll et al., 2 003) 0.360.870.89 10.970.87<0.01* NC (Sweet & Geratz, 2003) 0.390.950.40 9.390.950.03* VA & MD (Krstolic & Chaplin, 2007)0.380.890.74 10.550.89<0.01* Represents statistical significance (p 0.05) Slope Intercept

PAGE 131

131 Table 4-9. Comparison of bankfull mean dept h against drainage ar ea regressions by physiography, geography, floodplain t ypes, and Coastal Plain regions Effect EstimateR2P-value Estimate (ft)R2P-value Physiography: Flatwoods 0.180.490.77 0.860.490.83 Highlands 0.170.48 0.910.48 Geography: Northern peninsula 0.190.670.69 1.030.670.02* Southern peninsula 0.170.44 0.800.44 Floodplain Types: Wetland 0.160.520.66 0.950.520.60 Upland 0.180.41 0.850.41 Cypress-dominated 0.180.470.94 0.850.470.77 Non-cypress-dominated 0.180.45 0.890.45 Coastal Plain Studies: Peninsular FL (Blanton, 2008) 0.180.48Baseline 0.890.48Baseline North FL (Metcalf, 2004) 0.430.840.04* 0.660.840.16 Northwest FL (Metcalf, 2004) 0.260.860.16 1.610.86<0.01* AL (Metcalf, 2005) 0.480.960.10 0.770.960.40 GA (Buck Engineering, 2004) 0.380.830.25 0.680.83<0.01* NC (Doll et al., 2003) 0.300.740.46 1.290.740.01* NC (Sweet & Geratz, 2003) 0.310.920.70 1.020.920.68 VA & MD (Krstolic & Chaplin, 2007)0.280.870.26 1.100.870.65 Represents statistical significance (p 0.05) Slope Intercept

PAGE 132

132 Table 4-10. Summary of dimensionless ratios Site name Physiography Geography Floodplain type Drainage area (sq mi) Valley slope (%)SinuosityW/D Dmax/D Alexander Springs Creek tributary 2HLNUP1.61.0421.436.641.34 Blackwater Creek near Cassia HLNWFC1260.0201.0713.501.60 Blues Creek near Gainesville FWNUP2.60.2061.734.611.56 Bowlegs Creek near Ft Meade FWSWF47.20.0501.4412.281.94 Carter Creek near Sebring HLSUP38.80.2371.579.701.62 Catfish Creek near Lake Wales HLSWFC58.90.0501.4736.491.63 Coons Bay Branch FWSWF0.50.3481.256.221.45 Cow Creek FWNWFC5.30.0801.3310.631.60 Cypress Slash tributary HLSUP0.51.0421.0316.731.64 East Fork Manatee River tributary FWSUP0.20.3131.235.531.60 Fisheating Creek at Palmdale FWSWFC3110.0291.4222.631.73 Gold Head Branch HLNUP1.81.3161.074.271.64 Hammock Branch HLNWF3.00.1671.587.521.74 Hickory Creek near Ona FWSWF3.750.1161.1312.091.52 Hillsborough River tributary FWSWFC0.70.2601.4112.971.56 Horse Creek near Arcadia FWSWF2180.0431.0912.821.58 Jack Creek HLSWF5.20.2861.348.891.46 Jumping Gully HLNUP4.61.1111.344.901.63 Lake June-In-Winter tributary FWSUP0.40.7811.248.141.57 Little Haw Creek near Seville FWNWFC930.0611.1811.731.92 Livingston Creek near Frostproof HLSUP1200.0641.3117.991.62 Livingston Creek tributary HLSUP0.40.2501.015.071.74 Lochloosa Creek at Grove Park FWNWFC7.40.1161.0315.581.63 Lowry Lake tributary HLNUP0.250.6251.094.891.83 Manatee River near Myakka Head FWSUP65.30.1161.4710.021.42 Manatee River tributary FWSUP0.31.1631.293.731.79 Morgan Hole Creek FWSUP9.40.0911.336.491.70 Moses Creek near Moultrie FWNWFC7.40.1591.396.941.53 Myakka River tributary 1 FWSUP2.60.0911.0326.311.54 Myakka River tributary 2 FWSUP1.70.1291.2812.781.62 Nine Mile Creek HLNWF160.4881.548.221.46 Rice Creek near Springside FWNWFC43.20.0411.7413.351.45 Santa Fe River near Graham FWNUP94.90.0581.216.091.72 Shiloh Run near Alachua FWNUP0.322.0001.109.081.72 Snell Creek HLSWF1.70.1671.0914.411.71 South Fork Black Creek HLNWF25.50.1101.356.721.44 Spoil Bank tributary (Highlands) FWSUP8.60.3132.0814.882.00 Ten Mile Creek FWNWFC250.1301.229.111.73 Tiger Creek near Babson Park HLSUP52.80.0811.0817.531.58 Tiger Creek tributary HLSWF0.90.1391.3715.321.76 Triple Creek unnamed tributary 1 HLSWF1.70.5321.477.571.39 Triple Creek unnamed tributary 2 FWSUP0.20.8851.779.281.53 Tuscawilla Lake tributary HLNUP0.32.2731.203.111.56 Tyson Creek FWSWFC20.50.0541.0918.331.54 Unnamed Lower Wekiva tributary HLNWF0.40.7691.588.721.58 Minimum 0.20.0201.013.111.34 Maximum 3112.2732.0836.492.00 Mean 31.80.4091.3211.111.62 Median 4.60.1671.319.281.60 Data Subsets Independent Variables Dimensionless Ratios Notes: W/D = Width-to-depth ratio; Dmax/D = Maximum depth-to-mean depth ratio; FW = Flatwoods physiography; HL = Highlands physiography; N = Northern peninsula geography; S = Southern peninsula geography; WF = Wetland floodplain; WFC = Wetland floodplain dominated by cypress; UP = Upland floodplain

PAGE 133

133 Table 4-11. Comparison of vari ous dimensionless ratios by physiography, geography, and floodplain types Effect AverageP-valueAverageP-valueAverageP-valueAverageP-value Physiography: Flatwoods 1.340.56 11.260.86 1.640.36 0.310.14 Highlands 1.30 10.91 1.60 0.54 Geography: Northern peninsula 1.330.93 8.190.01* 1.610.82 0.290.08 Southern peninsula 1.32 13.24 1.62 0.57 Floodplain Types: Wetland 1.330.82 12.700.09 1.610.52 0.18<0.01* Upland 1.31 9.44 1.63 0.64 Cypress-dominated 1.300.79 15.570.01* 1.630.83 0.090.02* Non-cypress-dominated1.33 9.66 1.62 0.51 Notes: W/D = Width-to-depth ratio; Dmax/D = Maximum depth-to-mean depth; Represents statistical significance (p 0.05) Sinuosity W/D Dmax/D Valley Slope

PAGE 134

134 Table 4-12. Regression equations for bankfull parameters against drainage area and bankfull return inte rvals for studies condu cted throughout the southeastern Un ited States Coastal Plain EquationR2nE q u a t i o nR2nE q u a t i o nR2nEquationR2n Peninsular FL (Blanton, 2008)FIPRQ b kf = 14.26 Aw 0.360.6017A b kf = 6.05 Aw 0.470.7845W b kf = 6.87 Aw 0.300.8145D b kf = 0.89 Aw 0.180.4845<1 to 1.44 North FL (Metcalf, 2004) FDOTQ b kf = 7.54 Aw 0.770.9212A b kf = 6.1 Aw 0.710.9812W b kf = 9.2 Aw 0.280.8512D b kf = 0.67 Aw 0.430.84121 to 1.4 Northwest FL (Metcalf, 2004)FDOTQ b kf = 27.7 Aw 0.710.9514A b kf = 17.1 Aw 0.640.9914W b kf = 10.4 Aw 0.390.9614D b kf = 1.64 Aw 0.250.86141 to 1.4 AL (Metcalf, 2005) NOAAQ b kf = 10.94 Aw 0.940.938A b kf = 4.35 Aw 0.990.988W b kf = 5.67 Aw 0.520.948D b kf = 0.78 Aw 0.470.9681 to 1.1 GA (Buck Engineering, 2004)GDOTQ b kf = 6.80 Aw 0.780.8820A b kf = 5.93 Aw 0.720.9620W b kf = 8.59 Aw 0.340.8420D b kf = 0.68 Aw 0.380.83201 to 1.3 NC (Doll et al., 2003) UnknownQ b kf = 16.56 Aw 0.720.9016A b kf = 14.52 Aw 0.660.8816W b kf = 10.97 Aw 0.360.8716D b kf = 1.29 Aw 0.300.74161.0 to 1.25 NC (Sweet & Geratz, 2003) UnknownQ b kf = 8.79 Aw 0.760.9222A b kf = 9.43 Aw 0.740.9622W b kf = 9.64 Aw 0.380.9522D b kf = 0.98 Aw 0.360.92220.11 to 0.31 VA & MD (Krstolic & Chaplin, 2007)NOAAQ b kf = 28.31 Aw 0.600.7920A b kf = 11.99 Aw 0.640.9520W b kf = 10.45 Aw 0.370.8920D b kf = 1.15 Aw 0.270.8720<1 to 2.1 Notes: FIPR = Florida Institute of Phosphate Research; UF = Univ ersity of Florida; FDOT = Florida Department of Transportation; USFWS = United States Fish and Wildlife Service; NOAA = National Oceanic and Atmospheric Administration; GDOT = Georgia Department of Transportation; Qbkf = Bankfull discharge; Abkf = Bankfull area; Wbkf = Bankfull width; Dbkf = Bankfull mean depth; Aw= Drainage area; RI = Return interval; yrs = years Bankfull Width (ft) Bankfull Mean Depth (ft) Bankfull RI (yrs) Coastal Plain Region Funding Agency Bankfull Discharge (cfs) Bankfull Area (sq ft)

PAGE 135

135 Figure 4-1. Drainage area against valle y slope for study sites by physiography.

PAGE 136

136 A y = 14.26x0.36R2 = 0.60 1 10 100 1000 0.1 1 10 100 100 0 Drainage Area (sq mi)Bankfull Discharge (cfs) Gaged sites Power (Gaged sites) B C y = 27.85x0.57R2 = 0.60 1 10 100 1000 10000 0.1 1 10 100 100 0 Drainage Area (sq mi)1.5-year Discharge (cfs) Entire data set Power (Entire data set) Figure 4-2. Discharge agai nst drainage area regressions for gage d sites. A) Bankfull discharge. B) Mean annual discharge. C) 1.5-year discharge.

PAGE 137

137 A y = 14.47x0.38R2 = 0.64 y = 6.97x0.49R2 = 0.39 1 10 100 1000 0.1 1 10 100 100 0 Drainage Area (sq mi)Bankfull Discharge (cfs) Flatwoods Highlands Power (Flatwoods) Power (Highlands) B y = 1.35x0.92R2 = 0.96 y = 2.55x0.67R2 = 0.86 0.1 1.0 10.0 100.0 1000.0 0.1 1 10 100 100 0 Drainage Area (sq mi)Mean Annual Discharge (cfs) Flatwoods Highlands Power (Flatwoods) Power (Highlands) C Figure 4-3. Discharge against drainage area regressions for gaged sites by physiography (flatwoods versus highlands). A) Bankfull discharge. B) Mean annual discharge. C) 1.5-year discharge.

PAGE 138

138 A y = 15.98x0.32R2 = 0.58 y = 9.42x0.47R2 = 0.57 1 10 100 1000 0.1 1 10 100 100 0 Drainage Area (sq mi)Bankfull Discharge (cfs) Northern Southern Power (Northern) Power (Southern) B y = 1.37x0.88R2 = 0.94 y = 1.30x0.88R2 = 0.95 0.1 1 10 100 1000 0.1 1 10 100 100 0 Drainage Area (sq mi)Mean Annual Discharge (cfs) Northern Southern Power (Northern) Power (Southern) C y = 29.79x0.53R2 = 0.78 y = 22.27x0.63R2 = 0.37 1 10 100 1000 10000 0.1110100100 0 Drainage Area (sq mi)1.5-year Discharge (cfs) Northern Southern Power (Northern) Power (Southern) Figure 4-4. Discharge against dr ainage area regressions for ga ged sites by geography (northern versus southern peninsula). A) Bankfull di scharge. B) Mean annual discharge. C) 1.5-year discharge.

PAGE 139

139 A B C y = 28.49x0.58R2 = 0.52 y = 27.24x0.53R2 = 0.65 1 10 100 1000 10000 0.1 1 10 100 100 0 Drainage Area (sq mi)1.5-year Discharge (cfs) Wetland Upland Power (Wetland) Power (Upland) Figure 4-5. Discharge against drainage area regressions for gaged sites by floodplain type (wetland versus upland). A) Bankfull discha rge. B) Mean annual discharge. C) 1.5year discharge.

PAGE 140

140 A y = 10.94x0.35R2 = 0.48 y = 15.49x0.41R2 = 0.79 1 10 100 1000 0.1 1 10 100 100 0 Drainage Area (sq mi)Bankfull Discharge (cfs) Cypress-dominated Non-cypress dominated Power (Cypress-dominated) Power (Non-cypress dominated) B y = 2.87x0.71R2 = 0.87 y = 1.12x0.91R2 = 0.98 0 1 10 100 1000 0.1 1 10 100 100 0 Drainage Area (sq mi)Mean Annual Discharge (cfs) Cypress-dominated Non-cypress dominated Power (Cypress-dominated) Power (Non-cypress dominated) C y = 19.88x0.62R2 = 0.48 y = 30.49x0.56R2 = 0.68 1 10 100 1000 10000 0.1 1 10 100 100 0 Drainage Area (sq mi)1.5-year Discharge (cfs) Cypress-dominated Non-cypress dominated Power (Cypress-dominated) Power (Non-cypress dominated) Figure 4-6. Discharge against drainage area regressions for gaged sites by floodplain type (cypress-dominated versus non-cypress-domin ated). A) Bankfull discharge. B) Mean annual discharge. C) 1.5-year discharge.

PAGE 141

141 A y = 6.05x0.47R2 = 0.78 1 10 100 1000 0.1 1.0 10.0 100.0 1000 Drainage Area (sq mi)Bankfull Area (sq ft) Entire data set Power (Entire data set) B y = 6.87x0.30R2 = 0.81 1 10 100 0.1 1.0 10.0 100.0 1000 Drainage Area (sq mi)Bankfull Width (ft) Entire data set Power (Entire data set) C y = 0.89x0.18R2 = 0.48 0.1 1 10 0.1 1.0 10.0 100.0 100 0 Drainage Area (sq mi)Bankfull Mean Depth (ft) Entire data set Power (Entire data set) Figure 4-7. Channel geometry agai nst drainage area regressions for all sites. A) Bankfull crosssectional area. B) Bankfull width. C) Bankfull mean depth.

PAGE 142

142 A y = 6.27x0.46R2 = 0.82 y = 5.80x0.49R2 = 0.74 1 10 100 1000 0.1 1.0 10.0 100.0 1000. Drainage Area (sq mi)Bankfull Area (sq ft) Flatwoods Highlands Power (Flatwoods) Power (Highlands) B y = 7.28x0.28R2 = 0.92 y = 6.43x0.33R2 = 0.72 1 10 100 0.1 1.0 10.0 100.0 1000. Drainage Area (sq mi)Bankfull Width (ft) Flatwoods Highlands Power (Flatwoods) Power (Highlands) C y = 0.86x0.18R2 = 0.49 y = 0.91x0.17R2 = 0.48 0.1 1 10 0.11.010.0100.0100 0 Drainage Area (sq mi)Bankfull Mean Depth (ft) Flatwoods Highlands Power (Flatwoods) Power (Highlands) Figure 4-8. Channel geometry against drainage area regressi ons for all sites by physiography (flatwoods versus highlands). A) Bankfull cross-sectional area. B) Bankfull width. C) Bankfull mean depth.

PAGE 143

143 A y = 6.41x0.49R2 = 0.80 y = 5.78x0.46R2 = 0.78 1 10 100 1000 0.1 1.0 10.0 100.0 100 0 Drainage Area (sq mi)Bankfull Area (sq ft) Northern Southern Power (Northern) Power (Southern) B y = 6.26x0.30R2 = 0.76 y = 7.32x0.30R2 = 0.85 1 10 100 0.1 1.0 10.0 100.0 1000 Drainage Area (sq mi)Bankfull Width (ft) Northern Southern Power (Northern) Power (Southern) C y = 1.03x0.19R2 = 0.67 y = 0.80x0.17R2 = 0.44 0.1 1 10 0.1 1.0 10.0 100.0 1000 Drainage Area (sq mi)Bankfull Mean Depth (ft) Northern Southern Power (Northern) Power (Southern) Figure 4-9. Channel geometry against draina ge area regressions for all sites by geography (northern versus southern peninsula). A) Bankfull cross-sectional area. B) Bankfull width. C) Bankfull mean depth.

PAGE 144

144 A y = 8.11x0.41R2 = 0.79 y = 5.13x0.47R2 = 0.75 1 10 100 1000 0.1 1 10 100 100 0 Draina g e Area ( sq mi ) Bankfull Area (sq ft) Wetland Upland Power (Wetland) Power (Upland) B y = 8.61x0.26R2 = 0.77 y = 6.04x0.29R2 = 0.82 1 10 100 0.1110100100 0 Drainage Area (sq mi)Bankfull Width (ft) Wetland Upland Power (Wetland) Power (Upland) C y = 0.95x0.16R2 = 0.52 y = 0.85x0.18R2 = 0.41 0.1 1 10 0.1110100100 0 Drainage Area (sq mi)Bankfull Mean Depth (ft) Wetland Upland Power (Wetland) Power (Upland) Figure 4-10. Channel geometry against drainage area regressions for all sites by floodplain type (wetland versus upland). A) Bankfull crosssectional area. B) Bankfull width. C) Bankfull mean depth.

PAGE 145

145 A y = 7.29x0.46R2 = 0.84 y = 5.90x0.45R2 = 0.73 1 10 100 1000 0.1 1 10 100 100 0 Drainage Area (sq mi)Bankfull Area (sq ft) Cypress-dominated Non-cypress dominated Power (Cypress-dominated) Power (Non-cypress dominated) B y = 8.56x0.28R2 = 0.86 y = 6.67x0.27R2 = 0.75 1 10 100 0.1 1 10 100 100 0 Drainage Area (sq mi)Bankfull Width (ft) Cypress-dominated Non-cypress dominated Power (Cypress-dominated) Power (Non-cypress dominated) C Figure 4-11. Channel geometry against drainage area regressions for all sites by floodplain type (cypress-dominated versus non-cypress-domin ated). A) Bankfull cross-sectional area. B) Bankfull width. C) Bankfull mean depth.

PAGE 146

146 Figure 4-12. Boxplots of sinuosity by the entire data set and subsets representing physiography, geography, and floodpl ain types. Figure 4-13. Boxplots of width-to -depth ratio by the entire da ta set and subsets representing physiography, geography, and floodplain types. Indicates stat istical significance (p 0.05)

PAGE 147

147 Figure 4-14. Boxplots of maximu m depth-to-mean depth ratio by the entire data set and subsets representing physiography, geogr aphy, and floodplain types. Figure 4-15. Boxplots of valley slope by the entire data set and subsets representing physiography, geography, and floodplain types. Indicates statis tical significance (p 0.05)

PAGE 148

148 y = 9.13x0.71R2 = 0.85 y = 14.26x0.36R2 = 0.60 y = 18.28x0.70R2 = 0.87 y = 26.65x0.62R2 = 0.81 y = 10.96x0.94R2 = 0.93 y = 6.73x0.78R2 = 0.88 y = 27.48x0.71R2 = 0.95 y = 7.51x0.78R2 = 0.921 10 100 1000 10000 0.1 1 10 100 100 0 Drainage Area (sq mi)Bankfull Discharge (cfs) Power (North Carolina Coastal Plain (Sweet & Geratz, 2003)) Power (Peninsular Florida (Blanton, 2008)) Power (North Carolina Coastal Plain (Doll et al., 2003)) Power (Maryland/Virginia Coastal Plain (Krstolic & Chaplin, 2007)) Power (Alabama Coastal Plain (Metcalf, 2005)) Power (Georgia Coastal Plain (Glickauf et al., 2003)) Power (Northwest Florida Coastal Plain (Metcalf, 2004)) Power (North Florida Coastal Plain (Metcalf, 2004)) Figure 4-16. Bankfull discharge against drai nage area regressions by Coastal Plain study. y = 9.57x0.71R2 = 0.91 y = 6.05x0.47R2 = 0.78 y = 14.33x0.66R2 = 0.88 y = 11.61x0.66R2 = 0.95 y = 4.36x1.00R2 = 0.98 y = 5.92x0.72R2 = 0.96 y = 17.39x0.64R2 = 0.99 y = 6.40x0.70R2 = 0.981 10 100 1000 0.101.0010.00100.001000.00Drainage Area (sq mi)Bankfull Area (sq ft) Power (North Carolina Coastal Plain (Sweet & Geratz, 2003)) Power (Peninsular Florida (Blanton, 2008)) Power (North Carolina Coastal Plain (Doll et al., 2003)) Power (Virgina/Maryland Coastal Plain (Krstolic & Chaplin, 2007)) Power (Alabama Coastal Plain (Metcalf, 2005)) Power (Georgia Coastal Plain (Glickauf et al., 2003)) Power (Northwest Florida Coastal Plain (Metcalf, 2004)) Power (North Florida Coastal Plain (Metcalf, 2004)) Figure 4-17. Bankfull area against drainage area regressions by Coastal Plain study.

PAGE 149

149 y = 9.39x0.39R2 = 0.92 y = 6.87x0.30R2 = 0.81 y = 10.97x0.36R2 = 0.87 y = 10.55x0.38R2 = 0.88 y = 5.64x0.52R2 = 0.94 y = 8.58x0.35R2 = 0.84 y = 10.81x0.38R2 = 0.95 y = 9.82x0.26R2 = 0.771 10 100 1000 0.1 1 10 100 100 0 Drainage Area (sq mi)Bankfull Width (ft) Power (North Carolina Coastal Plain (Sweet & Geratz, 2003)) Power (Peninsular Florida (Blanton, 2008)) Power (North Carolina Coastal Plain (Doll et al., 2003)) Power (Virgina/Maryland Coastal Plain (Krstolic & Chaplin, 2007)) Power (Alabama Coastal Plain (Metcalf, 2005)) Power (Georgia Coastal Plain (Glickauf et al., 2003)) Power (Northwest Florida Coastal Plain (Metcalf, 2004)) Power (North Florida Coastal Plain (Metcalf, 2004)) Figure 4-18. Bankfull width ag ainst drainage area regressions by Coastal Plain study. y = 1.02x0.31R2 = 0.71 y = 0.89x0.18R2 = 0.48 y = 1.29x0.30R2 = 0.74 y = 1.10x0.28R2 = 0.84 y = 0.77x0.48R2 = 0.96 y = 0.68x0.38R2 = 0.83 y = 1.61x0.26R2 = 0.86 y = 0.66x0.43R2 = 0.890.1 1 10 0.1 1 10 100 1000 Drainage Area (sq mi)Bankfull Mean Depth (ft) Power (North Carolina Coastal Plain (Sweet & Geratz, 2003)) Power (Peninsular Florida (Blanton, 2008)) Power (North Carolina Coastal Plain (Doll et al., 2003)) Power (Virgina/Maryland Coastal Plain (Krstolic & Chaplin, 2007)) Power (Alabama Coastal Plain (Metcalf, 2005)) Power (Georgia Coastal Plain (Glickauf et al., 2003)) Power (Northwest Florida Coastal Plain (Metcalf, 2004)) Power (North Florida Coastal Plain (Metcalf, 2004)) Figure 4-19. Bankfull depth ag ainst drainage area regressi ons by Coastal Plain study.

PAGE 150

150 Figure 4-20. Boxplots of maximu m discharge-to-mean annual disc harge by the entire data set and subsets representing physiography, geography, and floodplain types. Indicates statistical significance (p 0.05) Figure 4-21. Mean annual runoff in the sout heastern US Coastal Plain. Source: Gerbert et al., 1987.

PAGE 151

151 CHAPTER 5 SYNTHESIS Objective 1: Most Reliable Bankfull Indicator for Peninsular Florid a Streams Various indicators of bankfull st age, including elevation of the flat floodplain (BKF-F), the inflection point on the bank (BKF-I), scour lin es (BKF-S), moss collars (BKF-M), tops of point bars (BKF-TOPB), and alluvial breaks (BKF-A) were identified, surveyed, and analyzed individually at 45 as ne ar-to-natural peninsular Florida stream s to determine if there is a single most reliable bankfull indicator for peninsular Florida streams. The following factors were examined: how prevalent each bankfull indicator is among study sites; how closely the slope of each bankfull indicator matches th at of the water; and how frequently and for how long the discharge and stage associated with each bankfull indicator occur. Based solely on prevalence of various bankfull indicators during the reference reach surveys, BKF-I and BKF-F (for streams with re latively flat wetland floodplains) were the most reliable field indicators of the bankfull stage for peninsular Florida stream s. The BKF-I indicator was ubiquitous at all study sites, while the BK F-F indicator was predom inantly found at sites with a wetland floodplain. BKF-M and BKF-TOPB were not present at enough sites to be reliable bankfull indicators for peninsular Florida st reams. While present at many sites, the BKF-S indicator was noticeably absent at many s ites with a cypress-dominated floodplain. The BKF-A indicator was too subjective and difficult to identify in the field to be a reliable bankfull indicator. (Table 3-1) Slopes of a line best fit through both the survey points of each individual bankfull indicator (BKF-F, BKF-I, BKF-S, BKF-A) and top of ba nk survey points (TOB) were compared to the slope of a line best fit through the water surface survey points (or the channel bed surface points for those sites that had no flowing water on the day of the survey). Leopold (1994) used this

PAGE 152

152 technique to verify the feature as bankfull if the two lin es were generally para llel and consistent over a long reach. To determine how parallel th e lines were, the water slope was divided by the slope of each bankfull indicator to determine a water slope to bankfull indicator slope ratio. Theoretically, the closer the ratio is to one, the more reliable the indicator. Bankfull indicator slopes within 25% of the water slope, or those with a water slope to bankfull indicator ratio between 0.75 and 1.25, were deemed candidate reliab le field indicators (T able 3-3). Based on this type of slopes analysis, BKF-I was the mo st reliable bankfull indi cator, with an average water slope to bankfull indicator slope ratio of 1.01. Variance in water slope to BKF-I slope ratio between streams with water slope less than 0.5% and streams with a water slope greater than 0.5% was not significantly different (p>0.05) (Table 3-4). Perhaps more importantly, however, slopes analysis suggested that there is a water slope threshold of approximately 0.5%, above which bankfull indicators become more reli able (except in the case of BKF-I) (Figure 32). It is important to note, how ever, that the population of stream s with water slopes greater than 0.5% was rather small (n=8), and thus additi onal research is recommended. These findings further suggest that slope-area techniques for calculating bankfu ll discharge should not be used in peninsular Florida for sites with a water slop e less than 0.5%, or conve rsely, that calculating discharge using slope-area techni ques is acceptable for sites with a water slope greater than 0.5%. Sites with long-term hydrologic data obtained from the USGS were analyzed to determine the frequency and duration of stage and discharge associated with various bankfull indicators. Based on gage analysis, it is safe to concl ude that both BKF-A and BKF-S occur far too frequently and are exceeded for far too much of the time to be considered the best indicator of bankfull discharge, or the most effective disc harge in transporting sediment and performing

PAGE 153

153 work (Tables 3-5 and 3-6). BKF-I and BKFF were thus further examined. Significant differences were found in durati ons of discharges and stages associated with top of bank (p<0.01) and BKF-I indicator (p<0.01) between sites with a wetland floodplain and those without (Table 3-9). Significant differences, however, were not found in durations of discharges and stages associated with BKF-F (p>0.05) be tween sites with a wetland floodplain and sites without. This is likely due to the nature of th e BKF-F indicator itselfa flat floodplain, which is generally found at sites with a we tland floodplain and is generally absent from sites without, as these sites are more likely to be incised. B ecause BKF-I and top of bank were found at every single site, the fact that significant differences exist between s ites with a wetland floodplain and sites without suggest that a different indicator should be used between these two floodplain types. In conclusion, elevation of the flat floodpl ain (BKF-F) is the most reliable bankfull indicator for peninsular Florid a streams with a wetland floodplain, while the inflection point (BKF-I) is the most reliable in dicator for incised streams or streams with an upland floodplain. Objective 2: Development of Regional Curves for Peninsu lar Florida Streams Regional curves, which relate bankfull discha rge and channel geometry (cross-sectional area, width, and mean depth) to drainage area in regions of similar climate, geology, and vegetation, were developed for pe ninsular Florida. Data were collected from 45 as near-tonatural peninsular Florida streams, with drainage areas ranging fr om 0.2 sq mi to 311 sq mi. The data obtained from the reference reach surveys were used to determine bankfull discharge, bankfull cross-sectiona l area, bankfull width, and bankfull mean depth. A power function regression was fit to the data and the coefficient of determination (R2) was determined. Due to potential inaccuracies of determining bankfull di scharge at gaged sites, mean annual discharge and 1.5-year discharge were also plotted against drainage area to see if these were better

PAGE 154

154 correlated with drainage area than the bankfull discharge. More specifically, the 1.5-year return interval was chosen because it is the return inte rval most often associated with bankfull flow (Leopold, 1994). Various discharges and bankfull parameters varied directly with drainage area, as expected. Bankfull mean depth had the lowest R2 values, while mean annual discharge had the highest R2 values. Relationships for bankfull discharge, mean annual discharge, 1.5-year discharge, bankfull area, bankfull width, and bankfull mean depth are shown in Figures 4-2 through 4-11. Table 4-1 summarizes the discharge da ta used in peninsular Florida regional curve development, while Table 4-2 summarizes the cha nnel geometry data used. Table 4-3 and 4-4 summarize the power function regr ession equations, corresponding coe fficients of determination, and sample sizes for discharge against drainage area and channel geometry against drainage area, respectively. Objective 3: Comparisons by Physiogr aphy, Geography, and Floodplain Types Regional curve data were further analyzed to d etermine whether si gnificant differences exist between streams draining different phys iographies (flatwoods versus highlands), geographies (northern versus s outhern peninsula), and floodplain types (wetland versus upland and cypress-dominated versus non-cypress-dominated), in terms of bankfull parameters and various dimensionless ratios (s inuosity, width-to-depth, maxi mum depth-to-mean depth, and valley slope). Analysis of C ovariance (ANCOVA) tests were pe rformed to determine whether significant differences exist in th e slopes and/or intercepts of th e bankfull discharge and channel geometry regressions for each data subset (JMP 7), while comparison of means tests were performed using Excel Data An alysis ANOVA: Single factor to determine if significant differences exist in the va rious dimensionless ratios

PAGE 155

155 Bankfull discharge appears to be higher at fl atwoods than highlands sites, at sites with upland floodplains than wetland fl oodplains, and at sites with non-cypress-dominated floodplains than sites with cypress-dominated-floodplains (Figures 4-3, 4-5, and 46). Sites with a noncypress-dominated floodplain started out with a significantly higher ba nkfull discharge than sites with a cypress-dominated floodplain (p =0.05) and had a significantly lower bankfull discharge duration (p=0.01) (Table 4-6). Flat woods streams were flashier than highlands streams, based on having significantly higher ma ximum discharge to mean discharge ratios (p=0.01). Sites with either wetland floodpl ains or cypress-dominated fl oodplains appear to have a greater bankfull area and bankfull width than sites with upland floodplains or non-cypressdominated floodplains, respectively (Figures 4-10 and 4-11). Th ese wetland floodplain sites also started out with a significantl y higher bankfull area (p=0.03) and bankfull width (p<0.01) than sites with upland floodplains (T ables 4-7 and 4-8). Though not significantly different, peak discharge-to-mean annual discharge ratios were also higher in streams with wetland floodplains and with cypress-dominated floodplains, indicating th at these streams are flas hier than those with an upland floodplain or a non-cypr ess-dominated floodplain (Table 4-6, Figure 4-20). Flashiness of these streams may help to explain why th ey are wider, as Osterkamp (1980) found that streams with a flashier regime and relatively hi gh peak flows tend to develop wider channels. Lastly, sites in the northe rn peninsula started out with a deeper bankfull mean depth than sites in the southern peninsula (p=0.02) (Table 4-9). No significant differences were found in sinuosity or maximum depth-to-mean depth based on physiography, geography, or floodplai n types. Sites with a cypress-dominated floodplain and sites located in the southern peninsula had sign ificantly higher width-to-depth ratio than sites

PAGE 156

156 with a non-cypress-dominated fl oodplain (p=0.01) and sites located located in the northern peninsula (p=0.01) (Table 4-11). Sites with upland floodplains and non-cypress dominated floodplains had significantly steep er valley slopes than sites w ith wetland floodplains (p<0.01) and cypress-dominated floodplai ns (p=0.02), re spectively. In conclusion, some significant differences existed in bankfull discharge and channel geometry of peninsular Florida streams based on geography and floodplain types, including: 1) bankfull depth and width-to-depth ratio were sign ificantly different in northern versus southern peninsula sites; 2) bankfull area and width (si ze), as well as valley slope, were significantly different in sites with wetland versus upland floodplai ns; and 3) bankfull discharge, width-todepth ratio (shape), and valley sl ope were significantly different in sites with cypress-dominated versus non-cypress-dominated floodplains. Sign ificant differences, however, were not found in bankfull discharge and channel size and shape of peninsular Florid a streams based on physiography (flatwoods versus highlands), though flatwoods streams were significantly flashier than highlands streams. Objective 4: Estimation of the Bankfull Discharge Return Interval Return intervals were estim at ed using the Annual Maximum Series from a Log Pearson Type III distribution and ranged from less than one year to 1.44 years (Table 4-2), which is more frequent than the average 1.5-year return interv al often reported in the literature (Dunne and Leopold, 1978; Leopold, 1994), but c onsistent with findings from other southeastern United States Coastal Plain studies (S weet and Geratz, 2003) (Table 4-12). This has important implications for flood control, as it indicates th at peninsular Florida st reams are overtopping their banks more frequently than in other regi ons. Because Annual Maximum Series cannot determine return intervals that are less than one year, mean and median bankfull return interval

PAGE 157

157 values could not be determined for peninsular Florida streams. It is thus recommended that future work includes a partial duration series to refine return intervals that ar e less than one year. Objective 5: Comparisons to Other Southeas tern United States Coastal Plain Studies Regional curves have recently been developed to estim ate ba nkfull discharge and channel geometry throughout the southeastern United Stat es Costal Plain, including Northwest Florida and North Florida Coastal Plain (Metcalf, 2004), Alabama Coastal Plain (Metcalf, 2005), Georgia Coastal Plain (Buck Engineering, 2004), North Carolina Coastal Plain (Doll et al., 2003; Sweet and Geratz, 2003), and Virginia and Maryla nd Coastal Plain (Krsto lic and Chaplin, 2007). Raw data from the present work and from these previous studies conducted throughout the southeastern United States Coastal Plain were en tered into Excel and regional curves for each bankfull parameter were compiled into one grap h for visual comparison (Table 4-12, Figures 416 through 4-19). Analysis of Covariance (ANCOVA) tests were then performed to determine whether significant differences exist in the slopes and/or intercepts of the bankfull discharge and channel geometry regressions between peninsular Florida streams (the baseline regression) and other Coastal Plain region al curves (JMP 7). The slope of peninsular Florid a streams for all bankfull regres sions tends to be less steep than the other slopes, indicating that bankfull para meters in peninsular Florida are less sensitive to changes in drainage area size, or in other words that the ba nkfull parameters in peninsular Florida streams increase at a slower rate with drainage area (Figures 4-16 though 4-19). Only slopes of North Florida bankfull width (p=0.01) and bankfull depth ( 0.04) regressions and Alabama bankfull area (p=0.01) and bankfull wi dth (p=0.01), however, were significantly different than peninsular Flor ida (Tables 4-7, 4-8, and 4-9). When examining intercepts of various Coas tal Plain regressions, peninsular Florida bankfull channels started out significantly narro wer and shallower than North Carolina (p<0.01;

PAGE 158

158 p=0.01) and Northwest Florida (p=0.01; p<0.01) Co astal Plain streams (Table 4-8 and Table 49). Perhaps this indicates that peninsular Florida streams are more efficient at conducting water, as peninsular Florida streams tend to be low gradient with sandy bottoms, which may enable them to conform and conduct water more easily than streams with steeper gradients and rocky streambeds. However, peninsular Florida stre ams start out at a signi ficantly lower bankfull discharge and area than North Carolina (p=0.01; p<0.01) and Northwest Florida (p<0.01; p=0.02) Coastal Plain streams, which could indicate that peninsular Florida streams receive less water overall (Tables 4-6 and 4-7). Figure 25 shows that peninsular Florida receives approximately ten inches less rain than Nort hwest Florida, while Figure 4-21 shows that peninsular Florida receives considerably less mean annual runoff (approximately 10 inches) than North Carolina (approximately 15 inches) and Northwest Florida (approximately 25 inches) (Gerbert et al., 1987). Peninsular Floridas low mean annual runoff values are likely attributable to its sandy soils, flat terrain, a nd deranged drainage networks. Peninsular Florida streams also deepen at a significantly faster rate with increas ing drainage area than No rth Florida streams. Perhaps this is because North Florida streams ha ve a steeper gradient and subsequently down-cut more. In conclusion, it is difficult to tweak out exactly why peninsular Florida streams are significantly different than other Coastal Plain st reams without further re search as there are a variety of variables, including the amount of water these systems receives (which depends upon on various factors such as climate, rainfall patt erns, runoff patterns, and baseflow), roughness of the streambed, gradient, level of alluvial control, and vegetation. Conclusions In conclusion, peninsular Floridas stream s are significantly different than other Coastal Plain regions, thus regional curves presented within the present wo rk should provide useful data

PAGE 159

159 to public agencies such as the Department of Environmental Protecti on (DEP), United States Geological Survey (USGS), and the Department of Transportation (DOT), as well as to private industries such as the phosphate mining industry for implementing natural channel designs as a stream restoration technique in peninsular Fl orida. Though not many significant differences were found within peninsular Florida str eams based on physiography, geography, and floodplain types, there are some important differences that should be cons idered when designing natural channels. For example, streams with wetland fl oodplains had significant ly greater bankfull area and bankfull width than streams with an upl and floodplain. Also, streams with cypressdominated floodplains had a grea ter width-to-depth ratio th an streams with non-cypressdominated floodplains. These size and shape differences based on fl oodplain types may be important restoration considerations.

PAGE 160

160 APPENDIX A SAMPLE PERMISSION LETTER AND FORM July 13, 2007 Florida Department of Agriculture and Consumer Services Division of Forestry ATTN: Joseph A. Bishop 9610 County Road 44 Leesburg, FL 32788 Dear Joseph A. Bishop: Blackwater Creek, which runs through or adjacent to your property located at Parcel # 1710901, Parcel # 1096162, and Parcel # 1096171 in Seminole Springs State Forest near Cassia in Lake County, Florida, has been selected for a publicly funded study to be performed by the University of Florida and BCI Engineers & Scientists, Inc. a Lakeland-based firm. The goal of the study is to assess the physical habitat of a wide variety of intact Florida stream segments. Your segment of the stream has been selected based on its natura l qualities, special contributions to the study requirements, and already available U.S. Geological Survey (USGS) data. We request and value your permission to access Blackwater Creek from your property in order to complete the studys necessary fieldwork. The fieldwork will be performed by a qualified research team, typically comprised of two personnel, and will last between one to two days. The fieldwork will be confined to the stream and floodplain. Accordingly, your property will not be disturbed and you will likely not even notice our presence. Please fill out the enclosed form and return it in the enclosed envelope within three weeks of receiving this letter. You may also fax the completed form to my attention at ( 863) 667-2662 If you have any questions, please feel free to call me at (954) 288-6588 or email me at blantonk@ufl.edu. Blackwater Creek is an integral piece of this study and your cooperation is greatly appreciated! Sincerely, Kristen Blanton, Graduate Research Assistant University of Florida / BCI Engineers & Scientists, Inc. Enclosures: reply envelope site form site map

PAGE 161

161 PHYSICAL HABITAT ASSESSMENT PERMISSION FORM SITE NAME: Blackwater Creek near Cassia, FL COUNTY: Lake ACCESS PROPERTY OWNER NAME: Florida Department of Agriculture and Consumer Services, Division of Forestry, ATTN: Joseph A. Bishop ACCESS PROPERTY ADDRESS: Parcel # 1710901, Parcel # 1096162, and Parcel # 1096171 in Seminole Springs State Forest near Cassia, FL OWNER MAILING ADDRESS: 9610 County Road 44, Leesburg, FL 32788 PREFERRED METHOD OF CONTACT: TELEPHONE: EMAIL: OTHER: HOURS OWNER MAY BE CONTACTED: COMMENTS / SPECIAL INSTRUCTIONS: I HEREBY GIVE THE UNIVERSITY OF FLORIDA AND BCI ENGINEERS & SCIENTISTS, INC. PERMISSION TO ACCESS THE ABOVE STREAM LOCATED ON OR ADJACENT TO MY PROPERTY TO PERFORM A PHYSICAL HABITAT ASSESSMENT. _________________________________ PRINTED NAME OF LANDOWNER _________________________________ ________________________ SIGNATURE OF LANDOWNER DATE

PAGE 162

162 APPENDIX B SITE FIGURES: PLAN FORM, LONGIT UDI NAL PROFILE, CROSS-SECTIONS

PAGE 163

163 A B Figure B-1. Alexander Springs trib utary 2. A) Plan form. B) L ongitudinal profile. C) Crosssections.

PAGE 164

164 C Figure B-1. Alexander Springs trib utary 2. A) Plan form. B) L ongitudinal profile. C) Crosssections.

PAGE 165

165 A B Figure B-2. Blackwater Creek near Cassia. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 166

166 C Figure B-2. Blackwater Creek near Cassia. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 167

167 A B Figure B-3. Blues Creek near Gain esville. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 168

168 C Figure B-3. Blues Creek near Gain esville. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 169

169 A B Figure B-4. Bowlegs Creek near Fort Meade. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 170

170 C Figure B-4. Bowlegs Creek near Fort Meade. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 171

171 A B Figure B-5. Carter Creek near Sebring. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 172

172 C Figure B-5. Carter Creek near Sebring. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 173

173 A B Figure B-6. Catfish Creek near Lake Wales. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 174

174 C Figure B-6. Catfish Creek near Lake Wales. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 175

175 A B Figure B-7. Coons Bay Branch. A) Plan form. B) Longitudinal profile C) Cross-sections.

PAGE 176

176 C Figure B-7. Coons Bay Branch. A) Plan form. B) Longitudinal profile C) Cross-sections.

PAGE 177

177 A B Figure B-8. Cow Creek. A) Pl an form. B) Longitudinal profile. C) Cross-sections.

PAGE 178

178 C Figure B-8. Cow Creek. A) Pl an form. B) Longitudinal profile. C) Cross-sections.

PAGE 179

179 A B Figure B-9. Cypess Slash tributary. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 180

180 C Figure B-9. Cypess Slash tributary. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 181

181 A B Figure B-10. East Fork Manatee River tributary. A) Plan form B) Longitudinal profile. C) Cross-sections.

PAGE 182

182 C Figure B-10. East Fork Manatee River tributary. A) Plan form B) Longitudinal profile. C) Cross-sections.

PAGE 183

183 A B Figure B-11. Fisheating Creek at Palmdale. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 184

184 C Figure B-11. Fisheating Creek at Palmdale. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 185

185 A B Figure B-12. Gold Head Branch. A) Plan form B) Longitudinal profile C) Cross-sections.

PAGE 186

186 C Figure B-12. Gold Head Branch. A) Plan form B) Longitudinal profile C) Cross-sections.

PAGE 187

187 A B Figure B-13. Hammock Branch. A) Plan form. B) Longitudina l profile. C) Cross-sections.

PAGE 188

188 C Figure B-13. Hammock Branch. A) Plan form. B) Longitudina l profile. C) Cross-sections.

PAGE 189

189 A B Figure B-14. Hickory Creek near Ona. A) Pl an form. B) Longitudinal profile. C) Crosssections.

PAGE 190

190 C Figure B-14. Hickory Creek near Ona. A) Pl an form. B) Longitudinal profile. C) Crosssections.

PAGE 191

191 A B Figure B-15. Hillsborough River tributary. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 192

192 C Figure B-15. Hillsborough River tributary. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 193

193 A B Figure B-16. Horse Creek near Arcadia. A) Plan form. B) Longitudi nal profile. C) Crosssections.

PAGE 194

194 C Figure B-16. Horse Creek near Arcadia. A) Plan form. B) Longitudi nal profile. C) Crosssections.

PAGE 195

195 A B Figure B-17. Jack Creek. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 196

196 C Figure B-17. Jack Creek. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 197

197 A B Figure B-18. Jumping Gully. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 198

198 C Figure B-18. Jumping Gully. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 199

199 A B Figure B-19. Lake June-in-Winter tributary. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 200

200 C Figure B-19. Lake June-in-Winter tributary. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 201

201 A B Figure B-20. Little Haw Creek near Seville. A) Plan form. B) Longitu dinal profile. C) Crosssections.

PAGE 202

202 C Figure B-20. Little Haw Creek near Seville. A) Plan form. B) Longitu dinal profile. C) Crosssections.

PAGE 203

203 A B Figure B-21. Livingston Creek near Frostproof. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 204

204 C Figure B-21. Livingston Creek near Frostproof. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 205

205 A B Figure B-22. Livingston Creek tr ibutary. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 206

206 C Figure B-22. Livingston Creek tr ibutary. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 207

207 A B Figure B-23. Lochloosa Creek at Grove Park. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 208

208 C Figure B-23. Lochloosa Creek at Grove Park. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 209

209 A B Figure B-24. Lowry Lake tributar y. A) Plan form. B) Longitudi nal profile. C) Cross-sections.

PAGE 210

210 C Figure B-24. Lowry Lake tributar y. A) Plan form. B) Longitudi nal profile. C) Cross-sections.

PAGE 211

211 A B Figure B-25. Manatee River near Myakka Head. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 212

212 C Figure B-25. Manatee River near Myakka Head. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 213

213 A B Figure B-26. Manatee River trib utary. A) Plan form. B) L ongitudinal profile. C) Crosssections.

PAGE 214

214 C Figure B-26. Manatee River trib utary. A) Plan form. B) L ongitudinal profile. C) Crosssections.

PAGE 215

215 A B Figure B-27. Morgan Hole Creek. A) Plan form B) Longitudinal profile C) Cross-sections.

PAGE 216

216 C Figure B-27. Morgan Hole Creek. A) Plan form B) Longitudinal profile C) Cross-sections.

PAGE 217

217 A B Figure B-28. Moses Creek near Moultrie. A) Pl an form. B) Longitudinal profile. C) Crosssections.

PAGE 218

218 C Figure B-28. Moses Creek near Moultrie. A) Pl an form. B) Longitudinal profile. C) Crosssections.

PAGE 219

219 A B Figure B-29. Myakka River tributary 1. A) Pl an form. B) Longitudinal profile. C) Crosssections.

PAGE 220

220 C Figure B-29. Myakka River tributary 1. A) Pl an form. B) Longitudinal profile. C) Crosssections.

PAGE 221

221 A B Figure B-30. Myakka River tributary 2. A) Pl an form. B) Longitudinal profile. C) Crosssections.

PAGE 222

222 C Figure B-30. Myakka River tributary 2. A) Pl an form. B) Longitudinal profile. C) Crosssections.

PAGE 223

223 A B Figure B-31. Nine Mile Creek. A) Plan form. B) Longitudi nal profile. C) Cross-sections.

PAGE 224

224 C Figure B-31. Nine Mile Creek. A) Plan form. B) Longitudi nal profile. C) Cross-sections.

PAGE 225

225 A B Figure B-32. Rice Creek near Springside. A) Pl an form. B) Longitudinal profile. C) Crosssections.

PAGE 226

226 C Figure B-32. Rice Creek near Springside. A) Pl an form. B) Longitudinal profile. C) Crosssections.

PAGE 227

227 A B Figure B-33. Santa Fe River near Graham. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 228

228 C Figure B-33. Santa Fe River near Graham. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 229

229 A B Figure B-34. Shiloh Run near Alachua. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 230

230 C Figure B-34. Shiloh Run near Alachua. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 231

231 A B Figure B-35. Snell Creek. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 232

232 C Figure B-35. Snell Creek. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 233

233 A B Figure B-36. South Fork Black Creek. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 234

234 C Figure B-36. South Fork Black Creek. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 235

235 A B Figure B-37. Spoil Bank tributar y. A) Plan form. B) Longitudi nal profile. C) Cross-sections.

PAGE 236

236 C Figure B-37. Spoil Bank tributar y. A) Plan form. B) Longitudi nal profile. C) Cross-sections.

PAGE 237

237 A B Figure B-38. Ten Mile Creek. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 238

238 C Figure B-38. Ten Mile Creek. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 239

239 A B Figure B-39. Tiger Creek near Babson Park. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 240

240 C Figure B-39. Tiger Creek near Babson Park. A) Plan form. B) Longitudinal profile. C) Crosssections.

PAGE 241

241 A B Figure B-40. Tiger Creek tributar y. A) Plan form. B) Longitudi nal profile. C) Cross-sections.

PAGE 242

242 C Figure B-40. Tiger Creek tributar y. A) Plan form. B) Longitudi nal profile. C) Cross-sections.

PAGE 243

243 A B Figure B-41. Triple Creek unnamed tributary 1. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 244

244 C Figure B-41. Triple Creek unnamed tributary 1. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 245

245 A B Figure B-42. Triple Creek unnamed tributary 2. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 246

246 C Figure B-42. Triple Creek unnamed tributary 2. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 247

247 Figure B-43. Tuscawilla Lake tributary. A) Pl an form. B) Longitudi nal profile. C) Crosssections.

PAGE 248

248 C Figure B-43. Tuscawilla Lake tributary. A) Pl an form. B) Longitudi nal profile. C) Crosssections.

PAGE 249

249 A B Figure B-44. Tyson Creek. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 250

250 C Figure B-44. Tyson Creek. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 251

251 A B Figure B-45. Unnamed Lower Wekiva tributary. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 252

252 C Figure B-45. Unnamed Lower Wekiva tributary. A) Plan form. B) Longitudinal profile. C) Cross-sections.

PAGE 253

253 APPENDIX C SITE PHOTOGRAPHS

PAGE 254

254 Alexander Springs tributary 2 (February 28, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 255

255 Blackwater Creek near Cassia (March 3, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 256

256 Blues Creek near Gainesville (January 10, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 257

257 Bowlegs Creek near Fort Meade (December 3, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 258

258 Carter Creek near Sebring (December 7, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 259

259 Catfish Creek near Lake Wales (September 27, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 260

260 Coons Bay Branch (November 13, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 261

261 Cow Creek (January 3, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 262

262 Cypress Slash tributary (December 17, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 263

263 East Fork Manatee River tributary (November 5, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 264

264 Fisheating Creek at Palmdale (March 20, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 265

265 Gold Head Branch (March, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 266

266 Hammock Branch (February 18, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 267

267 Hickory Creek near Ona (AugUpstreamt 9, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 268

268 Hillsborough River tributary (November 1, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 269

269 Horse Creek near Arcadia (March 17, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 270

270 Jack Creek (December 13, 2007)

PAGE 271

271 Jumping Gully (February, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 272

272 Lake June-in-Winter tributary (December 10, 2007) DOWNSTREAM LEFT BANK RIGHT BANK

PAGE 273

273 Little Haw Creek near Seville (February 29, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 274

274 Livingston Creek near Frostproof (December 5, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 275

275 Livingston Creek tributary (October 5, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 276

276 Lochloosa Creek at Grove Park (January 7, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 277

277 Lowry Lake tributary (February 14, 2004) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 278

278 Manatee River near Myakka Head (November 9, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 279

279 Manatee River tributary (November 2, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 280

280 Morgan Hole Creek (December 17, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 281

281 Moses Creek near Moultrie (January 18, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 282

282 Myakka River tributary 1 (October 15, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 283

283 Myakka River tributary 2 (October 16, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 284

284 Nine Mile Creek (March 12, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 285

285 Rice Creek near Springside (January 11, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 286

286 Santa Fe River near Graham (January 16, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 287

287 Shiloh Run near Alachua (January 8, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 288

288 Snell Creek (November 12, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 289

289 South Fork Black Creek (February, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 290

290 Ten Mile Creek (March 6, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 291

291 Tiger Creek near Babson Park (March 14, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 292

292 Tiger Creek tributary (December 6, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 293

293 Triple Creek unnamed tributary 1 (October 4, 2007)

PAGE 294

294 Triple Creek unnamed tributary 2 (October 11, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 295

295 TUpstreamcawilla Lake tributary (January 28, 2008) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 296

296 Tyson Creek (December 18, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 297

297 Unnamed Wekiva River tributary (October 30, 2007) DOWNSTREAM UPSTREAM LEFT BANK RIGHT BANK

PAGE 298

298 APPENDIX D GAGED SITE FIGURES: HYDROGRAPH, STAGE-Q RATING CURVE, FLOW AND STAGE DURATION C URVES

PAGE 299

299 Legend:

PAGE 300

300 Legend:

PAGE 301

301 Legend:

PAGE 302

302 Legend:

PAGE 303

303 Legend:

PAGE 304

304 Legend:

PAGE 305

305 Legend:

PAGE 306

306 Legend:

PAGE 307

307 Legend:

PAGE 308

308 Legend:

PAGE 309

309 Legend:

PAGE 310

310 Legend:

PAGE 311

311 Legend:

PAGE 312

312 Legend:

PAGE 313

313 Legend:

PAGE 314

314 Legend:

PAGE 315

315 Legend:

PAGE 316

316 APPENDIX E STAGE AGAINST WIDTH GRAPHS

PAGE 317

317 0.00 2.00 4.00 6.00 8.00 10.00 12.00 02 04 06 08 01 0 01 2 01 4 0 Width (ft)Stage (ft) Field Measurement Data Stage 1.5-year event (7.67) Stage BKF-F (6.87) Stage BKF-I (6.57) Figure E-1. Width versus stage: Blackwater Creek near Cassia. 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 02040608010012014 0 Width (ft)Stage (ft) Field Measurement Data Stage 1.5-year event (4.56) Stage BKF-I (3.24) Figure E-2. Width versus stage: Blues Creek near Gainesville. 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 02040608010012014 0 Width (ft)Stage (ft) Field Measurment Data Stage 1.5-year event (6.99) Stage BKF-F (4.96) Stage BKF-I (4.54) Figure E-3. Width versus stage: Bowlegs Creek near Fort Meade.

PAGE 318

318 0.00 2.00 4.00 6.00 8.00 10.00 12.00 0510152025303540 Width (ft)Stage (ft) Field Measurement Data Stage 1.5-year event (8.71) Stage BKF-I (6.51) Figure E-4. Width versus stage: Carter Creek near Sebring. 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 0 50 100 150 200 25 0 Width (ft)Stage (ft) Field Measurement Data Stage 1.5-year event (4.32) Stage BKF-I (4.32) Stage BKF-F (3.26) Figure E-5. Width versus stage: Catfish Creek near Lake Wales. 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 010020030040050060070080 0 Width (ft)Stage (ft) Field Measurement Data Stage 1.5-year event (6.30) Stage BKF-F (3.49) Stage BKF-I (2.66) Figure E-6. Width versus stage: Fisheating Creek at Palmdale.

PAGE 319

319 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 0 100 200 300400 500 60 0 Width (ft)Stage (ft) Field Measurement Data Stage 1.5-year event (11.79) Stage BKF-F (5.94) Stage BKF-I (3.58) Figure E-7. Width versus stag e: Horse Creek near Arcadia. 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 02040608010012014016018020 0 Width (ft)Stage (ft) Field Measurment Data Stage 1.5-year event (6.41) Stage BKF-F (4.37) Stage BKF-I (3.24) Figure E-8. Width versus stage: Little Haw Creek near Seville. 40.00 41.00 42.00 43.00 44.00 45.00 46.00 47.00 48.00 49.00 0102030405060 Width (ft)Stage (ft) Field Measurement Data Stage 1.5-year event (44.43) Stage BKF-I (43.08) Figure E-9. Width versus stage: Livingston Creek near Frostproof.

PAGE 320

320 g 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 0 50 100 150 200 25 0 Width (ft)Stage (ft) Field Measurement Data Stage 1.5-year event (5.61) Stage BKF-F (3.36) Stage BKF-I (2.94) Figure E-10. Width versus stage: Lochloosa Creek at Grove Park. 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 02040608010012014016018 0 Width (ft)Stage (ft) Field Measurement Data Stage 1.5-year event (12.49) Stage BKF-F (8.73) Stage BKF-I (5.79) Figure E-11. Width versus stage: Manatee River near Myakka Head. 10.00 11.00 12.00 13.00 14.00 15.00 16.00 17.00 18.00 19.00 02040608010012014 0 Width (ft)Stage (ft) Field Measurement Data Stage 1.5-year event (16.41) Stage BKF-F (15.73) Stage BKF-I (14.54) Figure E-12. Width versus stag e: Moses Creek near Moultrie.

PAGE 321

321 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 05010015020025030035 0 Width (ft)Stage (ft) Field Measurement Data Stage 1.5-year event (7.38) Stage BKF-F (4.04) Stage BKF-I (3.71) Figure E-13. Width versus stage: Rice Creek near Springside. 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 02040608010012014016 0 Width (ft)Stage (ft) Field Measurement Data Stage 1.5-year event (10.00) Stage BKF-I (6.79) Figure E-14. Width versus stage: Santa Fe River near Graham. 42.00 43.00 44.00 45.00 46.00 47.00 48.0001 02 03 04 05 06 0 Width (ft)Stage (ft) Field Measurement Data Stage 1.5-year event (45.18) Stage BKF-F (44.27) Stage BKF-I (44.21) Figure E-15. Width versus stage: Tiger Creek near Babson Park.

PAGE 322

322 APPENDIX F SUPPLEMENTAL STAGE DATA Appendix F. Stage data presented to suppl em ent the discharge data used in penisu lar Florida regional curve development and analysisSite name Period of record (WY) Physiography Geography Floodplain type Mean annual stage (ft) BKF stage Mean annual stage (ft) 1.5-year stage Mean annual stage (ft) 1.5-year stage BKF stage (%) Peak stage Mean annual stage (ft) Mean annual stage (% of time) BKF stage (% of time) 1.5-year stage (% of time) Blackwater Creek near Cassia81-07HLNWFC5.900.971.770.803.7245174.2 Blues Creek near Gainesville85-93FWNUP106.622.002.940.943.95320.80.3 Bowlegs Creek near Ft Meade92-07FWSWF3.561.413.432.035.7234142.8 Carter Creek near Sebring 92-07HLSUP5.570.943.142.203.8341110.1 Catfish Creek near Lake Wales48-07HLSWFC3.68-0.42**0.641.062.44507812 Fisheating Creek at Palmdale32-07FWSWFC3.120.373.182.819.1747433.0 Hickory Creek near Ona 82-84*FWSWF12.261.091.780.692.75504.50.6 Horse Creek near Arcadia 74-07FWSWF3.692.258.105.8514.0334172.8 Little Haw Creek near Seville52-06FWNWFC3.011.373.402.046.5141245.1 Livingston Creek near Frostproof92-07HLSUP42.370.542.061.526.4738278.2 Lochloosa Creek at Grove Park99-06*FWNWFC2.191.163.422.256.5539140.6 Manatee River near Myakka Head74-07FWSUP2.832.399.667.2714.8731110.5 Moses Creek near Moultrie 00-02*FWNWFC13.172.553.240.696.09373.91.8 Rice Creek near Springside 74-04FWNWFC3.680.363.703.345.3738310.4 Santa Fe River near Graham 58-93FWNUP108.601.744.953.219.4139153.1 Shiloh Run near Alachua N/AFWNUPN/AN/AN/AN/AN/AN/AN/AN/A Tiger Creek near Babson Park92-07HLSUP43.940.271.240.973.8241287.0 Minimum 2.19-0.420.640.692.44310.80.1 Maximum 108.602.559.667.2714.87507812 Mean 22.761.193.542.356.5440213.3 Median 4.631.133.212.035.9139162.8 Notes: WY = Water year; ft = feet; BKF = Bankfull; N/A = Not applicable -no stage data; Period of gage record insufficient (less than 10 years) for proper gage analysis, but rough approximations are presented; ** Bankfull stag e less than the mean annual stage (unexpected result) Duration Data Subsets Stage

PAGE 323

323 LIST OF REFERENCES Allan, J.D., 1995. Stream Ecology: Structure and Function of Running Waters. Kluwer Academ ic Publishers. Boston, Massachusetts, pp. 388. Beck, W.M., 1965. Streams of Florida. Flor ida State Museum Bulletin 10(3): 91-126. Berndt, M.P., E.T. Oaksford and G.L. Mahon, 1998. Groundwater. In E.A. Fernald and E.D. Purdum, eds. Water Resources Atlas of Florida. Institute of Science and Public Affairs, Florida State University, Tallahassee, Florida. Conover, C.S., 1973. Floridas Water Resources Institute of Food and Agricultural Sciences, University of Florida, Gainesville, Florida. Copeland, R.R., D.S. Biedenharn, and J.C. Fisc henich, 2000. Channel-forming Discharge. US Army Corps of Engineers Technical Note: VIII-5. Doll, B.A., A.D. Dobbins, J. Spooner, D.R. Clinton, and D.A. Bidelspach, 2003. Hydraulic Geom etry Relationships for Rural North Carolina Coastal Plain Streams. NC Stream Restoration Institute, Report to N.C. Division of Water Quality for 319 Grant Project No. EW20011, pp. 11. Dunne, T. and L.B. Leopold, 1978. Water in En vironmental Planning. W.H. Freeman and Company, San Francisco, pp. 818. Emmett, W.W., 1975. Hydrologic Evaluation of th e Upper Salmon River Area, Idaho. U.S. Geological Survey Prof essional Paper 870-A. Emmett, W.W., 2004. A Historical Perspective on Regional Channel Geometry Curves. Stream Notes. Stream Systems Technology Center, U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station, Fort Collins, pp. 2. Fenneman, N. M., 1946. Physical Divisions of th e United States, Map (scale 1:7,000,000). U.S. Geological Survey, Reston, Virginia. FISWRG, 1998. Stream Corridor Restoration: Prin ciples, Processes, and Practices. Springfield (VA): Federal Interagency Stream Corri dor Restoration Working Group. NTIS. pp. 574. FDOT, 1999. Florida Land Use, Cover and Forms Classification System (FLUCCS). FNAI, 1990. Guide to the Natural Communities of Florida. Florida Natural Areas Inventory and Florida Department of Natural Resources, pp. 116. Gerbert, W.A., D.J. Graczyk, and W.R. Krug, 1987. Average Annual Runoff in the United States, 1951-1980, Map (scale 1:7,500,000). U. S. Geological Survey Hydrologic Atlas HA-710.

PAGE 324

324 Gordon N.D., T.A. McMahon, B.L. Finlayson, C.J. Gippel, and R.J. Nathan, 2004. Stream Hydrology: An Introduction for Ecologists. 2nd Ed. Wiley, New Jersey, pp. 429. Harman, W.H., G.D. Jennings, J.M. Patterson, D. R. Clinton, L.O. Slate, A.G. Jessup, J.R. Everhart, and R.E. Smith, 1999. Bankfull Hydraulic Geometry Relationships for North Carolina Streams. In: Wildland Hydrology, D.S. Olsen a nd J.P. Potyondy (Editors). Proceeding of the Wildland Hydrology Symposium, AWRA, Bozeman, Montana, pp. 401-408. Harrelson, C.C., C.L. Rawlins, and J.P. Potyondy, J.P., 1994. Stream Channel Reference Sites: an Illustrated Guide to Field Technique. U.S. De partment of Agriculture Forest Service General Technical Report RM-245, pp. 61. Henry, J. A., 1998. Weather and Climate. In E.A. Fernald and E.D. Purdum, eds. Water Resources Atlas of Florida. Institute of Science and Public Affairs, Florida State University, Tallahassee. Johnson, P.A. and T.M. Teil, 1996. Uncertainty in Estimating Bankfull Conditions. Water Resources Bulletin 32:1283-1291. Kautz, R.S., K. Haddad, T.S. Hoehn, T. Roge rs, E. Estevez, and T. Atkeson, 1998. Natural Systems. In E.A. Fernald and E.D. Purdum, eds. Water Resources Atlas of Florida. Institute of Science and Public Affairs, Florida State University, Tallahassee. Knighton, D., 1998. Fluvial Forms and Proce sses. John Wiley & Sons, New York, pp. 383. Lane, E., 1994. Floridas Geologi cal History and Geological Re sources. Florida Geological Survey Special Publication No. 35, Tallahassee. Leopold, L. B., 1994. A View of the River. Harvard University Press, Cambridge, Massachusetts. Leopold, L.B. and T. Maddock Jr, 1953. The Hydraulic Geometry of Stream Channels and Some Phyisiographic Implications. U.S. Geologi cal Survey Professional Paper 252, pp. 57. Leopold, L.B. and T. Maddock Jr, 1953. The Hydraulic Geometry of Stream Channels and Some Phyisiographic Implications. U.S. Geologi cal Survey Professional Paper 252, pp. 57. Malakoff, D., 2004. The River Doctor. Science 305:937-939. McCandless, T.L. and R.A. Everett, 2003. Maryland Stream Survey: Bankfull Discharge and Channel Characteristics of Streams in the Allegheny Plateau and the Valley and Ridge Hydrologic Region. U.S. Fish and Wildlife Service, Annapolis, Maryland, CBFO-S03-01, pp. 92.

PAGE 325

325 Metcalf, C., 2004. Regional Channel Character istics for Maintaining Natural Fluvial Geomorphology in Florida Streams. U.S. Fish a nd Wildlife Service, Pana ma City, Florida, pp. 45. Mossa, J., 1998. Surface Water. In E.A. Fernald and E.D. Purdum, eds. Water Resources Atlas of Florida. Institute of Science and Public A ffairs, Florida State University, Tallahassee. Nixon, M., 1959. A Study of Bankfull Discharges in England and Wales. In proceedings of the Institution of Civil Engineers, 12:157-175. Nordlie, F.G., 1990. Rivers and Springs. In R.L. Myers and J.J. Ewel, eds. Ecosystems of Florida. University of Centra l Florida Press, Orlando, pp. 392-425. Osterkamp, W.R., 1980. Sediment-morphology Relations of Alluvial Channels. Proceedings of the Symposium on Watershed Management, Americ an Society of Civil Engineers, Boise 1980, pp. 188-99. Rosgen, D.L., 1994. A Classification of Natural Rivers. Catena 22:169-199. Sweet, W.V. and J.W. Geratz, 2003. Bankfull Hydr aulic Geometry Relationships and Recurrence Intervals for North Carolinas Coastal Plai n. Journal of the Amer ican Water Resources Association 39: 861-871. Thorne, C.R., R.D. Hey, and M.D. Newson, 1997. Applied Fluvial Geomorphology for River Engineering and Management: John Wiley & Sons, Chichester, England, pp. 376. U.S.D.A. Forest Service, 1995. A Guide to Fi eld Identification of Bankfull Stage in the Western United States (video), Rocky Mount ain Forest and Range Experiment Station, Stream Systems Tech. Center, Fort Collins, Colorado. USGS, 1982. Guidelines for Determining Fl ood Flow Frequency. Bulletin #17B of the Hydrology Subcommittee Interagency Advisory Committee on Water Data. U.S. Geological Survey, Reston, VA. Wolman, M.G., 1955. The Natural Channel of Brandywine Creek, Pennsylvania. U.S. Geological Survey Pr ofessional Paper 271. Wolman, M.G. and L.B. Leopold, 1957. River Flood Plains: Some Observations on Their Formation. U.S. Geological Survey Professional Paper 282-C, pp. 30. Wolman, M.G. and J.P. Miller, 1960. Magnitude and frequency of forces in geomorphic processes. Journal of Geology 68:54-74.

PAGE 326

326 BIOGRAPHICAL SKETCH Kristen Blanton was born and raised in Fort Lauderdale, Florida, the land of canals. She attend ed Wellesley College in Wellesley, Massachusetts where she pursued a major in environmental studies, graduating in May 2004. During college, she spent a rewarding summer at Archbold Biological Station in Lake Placid, Florida, where she came to appreciate Floridas natural systems. Upon graduation, Kristen work ed as a geologist at Roux Associates, Inc. (Boston, Massachusetts). While at that job, Kristen was given the opportunity to be trained as a wetlands scientist and to work on various wetlan d restoration projects. In August 2006, seeking to trade the cold northeast for her sunny home state, Kristen move d to Gainesville, Florida to begin her masters work with Dr. Wise in the Un iversity of Floridas Environmental Engineering Sciences Department. For her masters wor k, Kristen was lucky enough to join the "stream team" and to explore some of Floridas most beautiful and natural streamsthe opposite of canals. Kristen is now hoping to begin a car eer in land conservation, so she can protect the natural areas she loves.