Citation
Assessing Stream-Mediated Seed and Shoot Dispersal of Invasive Plants in an Urban Riparian Wetland

Material Information

Title:
Assessing Stream-Mediated Seed and Shoot Dispersal of Invasive Plants in an Urban Riparian Wetland
Creator:
Seitz, Jason C
Place of Publication:
[Gainesville, Fla.]
Florida
Publisher:
University of Florida
Publication Date:
Language:
english
Physical Description:
1 online resource (141 p.)

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Soil and Water Science
Committee Chair:
CLARK,MARK W
Committee Co-Chair:
OSBORNE,TODD Z
Committee Members:
FLORY,STEPHEN
Graduation Date:
12/19/2014

Subjects

Subjects / Keywords:
Buoyancy ( jstor )
Creeks ( jstor )
Floodplains ( jstor )
Flow velocity ( jstor )
Forests ( jstor )
Germination ( jstor )
Seeds ( jstor )
Species ( jstor )
Streams ( jstor )
Tradescantia ( jstor )
Soil and Water Science -- Dissertations, Academic -- UF
alachua -- arm -- bivens -- control -- creek -- dispersal -- florida -- fluminensis -- gainesville -- introduced -- invasive -- management -- non-native -- nonindigenous -- plant -- propagule -- riparian -- ruellia -- seed -- shoot -- simplex -- stream -- tradescantia -- tumblin -- watershed -- wetland
City of Gainesville ( local )
Genre:
bibliography ( marcgt )
theses ( marcgt )
government publication (state, provincial, terriorial, dependent) ( marcgt )
born-digital ( sobekcm )
Electronic Thesis or Dissertation
Soil and Water Science thesis, M.S.

Notes

Abstract:
Ruellia simplex and Tradescantia fluminensis were introduced to Florida and elsewhere and are now considered invasive plants. Millions of dollars are spent annually on the control of invasive plants in Florida alone. The goals of this study were to determine the importance of stream dispersal in invasive plants such as R. simplex (by seed) and T. fluminensis (by shoots) in riparian wetlands. The 13,269 seeds and 179 shoots intercepted during net deployments in a Florida stream included 131 seeds and 20 shoots of R. simplex and 6 shoots of T. fluminensis. Net verification tests showed low retention of R. simplex seeds while retention of T. fluminensis shoots was 100%. The low seed retention values for R. simplex suggested that the actual number of seeds transported downstream during net deployments may have been in the thousands. Numbers of propagules per year transported downstream are estimated at 18,000 to 30,000 R. simplex seeds, 8,000 to 10,000 R. simplex shoots, and 2,000 to 8,000 T. fluminensis shoots. Capture rates of marked propagules showed mobility to be low for R. simplex seeds but high for T. fluminensis shoots. Ruellia simplex seeds are negatively buoyant while T. fluminensis shoots are somewhat positively buoyant but become negatively buoyant within 15 days of submergence. Percent germination of net-captured R. simplex seeds was 6.9% in a germination chamber and 21.8% inside a glasshouse. Survivorship with growth of net-captured shoots amounted to 22.2% for R. simplex and 40.0% for T. fluminensis. Hand-harvested R. simplex seeds remained viable following up to 90 days of submergence and 30 days of burial in sediments. Hand-harvested T. fluminensis shoots survived up to 10 days of submergence. The results of this study reveal the importance of stream transport of propagules as a dispersal mechanism for these species. It is recommended that control efforts be conducted at the watershed scale, beginning at upstream portions of a watershed and working downstream to reduce the potential for recolonization, at least for these two species. ( en )
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.
Thesis:
Thesis (M.S.)--University of Florida, 2014.
Local:
Adviser: CLARK,MARK W.
Local:
Co-adviser: OSBORNE,TODD Z.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2015-12-31
Statement of Responsibility:
by Jason C Seitz.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Embargo Date:
12/31/2015
Resource Identifier:
974373221 ( OCLC )
Classification:
LD1780 2014 ( lcc )

Downloads

This item has the following downloads:


Full Text

PAGE 1

ASSESSING STREAM MEDIATED SEED AND SHOOT DISPERSAL OF INVASIVE PLANTS IN AN URBAN RIPARIAN WETLAND By JASON C. SEITZ A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2014

PAGE 2

2014 Jason C. Seitz

PAGE 3

To my Mom and Dad, for their encouragement

PAGE 4

ACKNOWLEDGMENTS A special thank you is due to my advisor Mark Clark for his excellent advice and good humor , and for some great pig roasts! Luke Flory of my committee provided helpful and timely advice a nd graciously gave me access to his lab and germination chambers . Todd Osborne of my committee also provided helpful advice and guidance. I thank the following people for their help: Carl Gillis (FDAC, Division of Plant Industry ) kindly sterilized hundreds of R. simplex seeds with the Gammacell 1000 irradiator ; Donna Ruhl (Florida Archaeology Collection & Archaeobotany, Florida Museum of Natural History [FLMNH] ) provided her expertise in seed identification and helpful suggestions on proper storage of seeds ; Kent Perkins (Herbarium, FLMNH ) allowed me use of the UF seed collections and the H erbarium library ; Bret Boyd (UF Genetics Institute) allowed me use of the high tech macro camera setup; John Slapcinsky (Division of Invertebrate Zoology, FLMNH ) verified the identity of snails caught in my nets ; Robert Faden (Dept. of Botany, Smithsonian Institution) and Marco Oct vio Pellegrini (Rio de Janeiro Botanical Garden) provided valuable insight into T. fluminensis reproduction. Many thanks go to Joe from the tent camp at Tumblin Creek for his assistance and camaraderie through most of the field effort , for helping to look after my nets , and his tireless efforts at cleaning up the floodplain. I thank N adia Lombardero and Michelle Rau at ANAMAR Environmental Consulting for their support and understanding while I juggled g raduate work with my full time biologist position . I thank Connie Steen of ANAMAR f or her help in formatting this thesis . Most of all , I thank m y wife , Jenn y , for believing in me , providing me with the strength to complete this thesis , all her help in the field, and for her patience and understanding throughout my latein life graduate studies. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES ............................................................................................................ 7 LIST OF FIGURES ........................................................................................................ 10 LIST OF ABBREVIATIONS & ACRONYMS .................................................................. 13 ABSTRACT ................................................................................................................... 14 CHAPTER 1 THE IMPORTANCE OF SEED AND SHOOT DISPERSAL IN INVASIVE PLANTS . 16 Introduction ............................................................................................................. 16 Stream mediated Dispersal Mechanisms ............................................................... 22 Goals and Objectives .............................................................................................. 22 Hypotheses ............................................................................................................. 23 Study Site: Tumbli n Creek at Bivens Arm ............................................................... 24 Past Control Efforts within the Tumblin Creek Floodplain ....................................... 30 2 FREQUENCY AND MOBILITY OF SEEDS AND SHOOTS IN STREAM ................. 32 Introduction ............................................................................................................. 32 Materials and Methods ............................................................................................ 32 Results and Discussion ........................................................................................... 50 3 VIABILITY OF SEEDS AND SHOOTS FOLLOWING INUNDATION ........................ 98 Introduction ............................................................................................................. 98 Materials and Methods ............................................................................................ 98 Results and Discussion ......................................................................................... 104 4 CONCLUSIONS ...................................................................................................... 117 APPENDIX A NONINDIGENOUS PLANTS AND ANIMALS OF THE FLOODPLAIN .................... 119 B STREAM METRICS AND PLANT COVER ALONG STREAM ................................. 121 C NUMBERS OF SEEDS AND SHOOTS INTERCEPTED IN STREAM .................... 126 D VIABILITY OF INVASIVE SEEDS AND SHOOTS INTERCEPTED IN THE STREAM ............................................................................................................... 131 5

PAGE 6

LIST OF REFERENCES ............................................................................................. 134 BIOGRAPHICAL SKETCH .......................................................................................... 141 6

PAGE 7

LIST OF TABLES Table page 1 1 Comparisons of reproduction in Ruellia simplex cul tivars and the naturalized type ................................................................................................................. 19 1 2 Monthly mean flow rates, by baseflow (<0.02 m3/second) and highflow ( 0.02 m3/second), and mean volumes for the period July 2001 through June 2003 at Tumblin Creek approximately 300 m upstream of the floodplain forest ................................................................................................................. 29 2 1 Net deployments by month and by flow regime in Tumblin Creek, May – December 2012 .................................................................................................. 40 2 2 Capture and retention rates for marked Ruellia simplex seeds in Tumblin Creek ................................................................................................................. 65 2 3 Capture and retention rates for marked Tradescantia fluminensis shoots in Tumblin Creek .................................................................................................... 66 2 4 Seeds of Ruellia simplex and other nonindigenous and invasive plants per month and per flow regime in Tumblin Creek ..................................................... 79 2 5 Mean interception rates (individuals/100 m3) for seeds of Ruellia simplex and other nonindigenous and invasive plants per month and per flow regime in Tumblin C reek .................................................................................................... 80 2 6 Shoots of Ruellia simplex , Tradescantia fluminensis , and other nonindigenous and invasive plants per month and per flow regime in Tumblin Creek ................................................................................................................. 81 2 7 Mean interception rates for shoots of Ruellia simplex , Tradescantia fluminensis , and other nonindigenous and invasive plants per month and per flow regime in Tumblin Creek ............................................................................. 82 2 8 Results of a oneway ANOVA for numbers of intercepted and retained Ruellia simplex seeds and shoots, and Tradescantia fluminensis shoots, during the 27 net deployments and the interception rates predicted under the H 1 null hypothesis (no seeds or shoots intercepted) ......................................... 83 2 9 Results of two tailed t tests comparing numbers of Ruellia simplex seeds and shoots, and Tradescantia fluminensis shoots, intercepted and retained during the 27 net deployments to the predicted scenario under the H 1 null hypothesis (no seeds or shoots intercepted) ...................................................... 83 7

PAGE 8

2 10 Results of a oneway ANOVA for rates of interception for Ruellia simplex seeds and shoots, and Tradescantia fluminensis shoots, during the 27 net deployments between baseflow ( < 0.02 m3/second) and highm3/second) conditions ........................................................................................ 96 2 11 In stream testing for buoyancy in Ruellia simplex seeds .................................... 96 2 12 Mean mass, volume, and density of Ruellia simplex seeds hand harvested from an introduced population in Gainesville, Florida ......................................... 97 2 13 In stream testing for buoyancy in Tradescantia fluminensis shoots .................... 97 2 14 Mean mass, volume, and density of Tradescantia fluminensis shoots hand harvested from an introduced population in Gainesville, Florida. ....................... 97 3 1 Results of a oneway ANOVA for germination trials using Ruellia simplex seeds intercepted in Tumblin Creek and tested for viability inside a germination chamber ........................................................................................ 110 3 2 Results of germination trials using Ruellia simplex seeds intercepted in Tumblin Creek .................................................................................................. 110 3 3 Results of a oneway ANOVA for germination trials using Ruellia simplex seeds intercepted in Tumblin Creek and tested for viability inside an unheated glasshouse ....................................................................................... 111 3 4 Results of germination trials using Ruellia simplex seeds intercepted in Tumblin Creek and tested inside an unheated glasshouse .............................. 111 3 5 Results of survival and growth trials for Ruellia simplex and Tradescantia fluminensis shoots intercepted in Tumblin Creek ............................................. 112 3 6 Results of oneway ANO VAs for Ruellia simplex seed germination trials following submergence in the water column for 15, 30 , 90 , and 180 day durations ........................................................................................................... 113 3 7 Results of Ruellia simplex seed germination trials following submergence in the water column for 15, 30 , 90 , and 180day durations ............................... 114 3 8 Results of a oneway ANOVA for Tradescantia fluminensis shoot survival trials following submergence in water for 5to 15 day durations ...................... 114 3 9 Results of Tradescantia fluminensis shoot survival trials following submergence in water for 5to 15 day durations ............................................. 115 3 10 Results of oneway ANOVAs for shoot length, number of leaves, and number of roots in Tradescantia fluminensis shoots following submergence in wat er for 5 to 10 day durations ................................................................................. 115 8

PAGE 9

3 11 Growth of Tradescantia fluminensis shoots following submergence in water for 5 to 15 day durations ................................................................................. 116 A 1 List of nonindigenous plant species observed by the author within the Tumblin Creek floodplain and associated natural areas, 2009 – 2014 ............... 119 A 2 List of nonindigenous invertebrate and vertebrate species observed by the author within the Tumbli n Creek floodplain and associated natural areas, 2009– 2014 ....................................................................................................... 120 B 1 Flow rates, water depths, stream widths, and substrate types measured during baseflow conditions on 26 May, 2012, every 25 m along the 800m long reach of Tumblin Creek between SW 13th Street (US 441) and Bivens Arm ............................................................................................................... 121 B 2 Mean percent aerial coverage of herbaceous and small woody plants within 3 m upslope of the mean high water line along both banks of Tumblin Creek per gro up of transects surveyed on 29 April, and 5 May, 2012 ........................ 122 B 3 Results of a qualitative survey in June 2011 of Ruellia simplex and Tradescantia fluminensis along the main channel of Tumblin Creek from the headwaters to SW 13th Street (US 441) ........................................................... 125 C 1 Seeds intercepted per month and per flow regime in Tumblin Creek ............... 126 C 2 Shoots intercepted per month and per flow regime in Tumblin Creek .............. 129 D 1 Results of seed germination trials for invasive and nonindigenous plant seeds, excluding Ruellia simplex , intercepted by net in Tumblin Creek ........... 131 D 2 Results of seed germination trials for invasive and nonindigenous plant seeds, excluding Ruellia simplex , intercepted by net in Tumblin Creek ........... 132 D 3 Results of shoot survival and growth trials for invasive and nonindigenous plant s hoots, excluding Ruellia simplex and Tradescantia fluminensis , inte rcepted by net in Tumblin Creek ................................................................. 133 9

PAGE 10

LIST OF FIGURES Figure P age 1 1 Ruellia simplex is dominant along many urban waterways and riparian wetlands in Gainesville, Florida. ......................................................................... 18 1 2 Tradescantia fluminensis (a) forms dense monocultures (b) and reduces recruitment of native species (c) in northern Florida and elsewhere ................... 21 1 3 The 30 acre Tumblin Creek floodplain forest is situated primarily on University of Florida land along the north shore of Bivens Arm in Gainesville, Florida ................................................................................................................ 25 1 4 The 6.1km2 Tumblin Creek watershed (delineated with a green line) includes much of southern Gainesville as it flows 3.7 km southwest to Bivens Arm via th e floodplain forest . ........................................................................................... 26 1 5 Monthly mean precipitation in the Tumblin Creek basin July 2000 through June 2003. .......................................................................................................... 28 1 6 Daily flow rates for Tumblin Creek, approximately 300 m upstream of the floodplain forest .................................................................................................. 29 2 1 A total of 33 transects (yellow circles with numbers) were marked every 25 m along the 800m long stream reach between SW 13th Street (US 441) and Bivens Arm ......................................................................................................... 34 2 2 Groups of propagules were marked with fluorescent colors and released into the stream at certain distances upstream of the nets to determine mobility ....... 38 2 3 Rates of flow for each of 27 net deployments conducted May – December 2012 in Tumblin Creek to intercept seeds and shoots ........................................ 40 2 4 A seine net and at least one finemesh net was strung across Tumblin Creek to intercept seeds and shoots as they were transported downstream ................ 43 2 5 Material intercepted in a seine net strung across Tumblin Creek, Gainesville, Florida, during 17– 18 March 2012 ...................................................................... 44 2 6 Ten 400 mL decontaminated beakers with finemesh screen tops (a) were used to submerge Ruellia simplex seeds (15 seeds per beaker, 150 seeds total) to determine buoyancy .............................................................................. 48 2 7 Three cubeshaped cages, measuring 0.6m per side and constructed of hardware cloth on PVC frames (a), were used to submerge Tradescantia fluminensis shoots in Tumblin Creek for determination of buoyancy and viability following submergence .......................................................................... 49 10

PAGE 11

2 8 Percent retention of marked R. simplex seeds for each of six release events (three during baseflow [<0.02 m3/second] and three during hig h m3/second] conditions) in order of increasing rates of flow ................................. 67 2 9 Percent capture of marked R. simplex seed s placed 1 m upstream of the nets for each of six release events (three during baseflow [<0.02 m3/second] and three during high3/second] conditions) in order of increasing rates of flow ....................................................................................... 68 2 10 Percent capture of marked R. simplex seed s placed 10 m upstream of the nets for each of six release events (three during baseflow [<0.02 m3/second] and three during high.02 m3/second] conditions) in order of increasing rates of flow ....................................................................................... 69 2 11 Percent capture of marked R. simplex seed s placed 20 m upstream of the nets for each of six release events (three during baseflow [<0.02 m3/second] and three during high3/second] conditions) in order of increasing rates of flow ....................................................................................... 70 2 12 Percent capture of marked T. fluminensis shoots for each of eight release events (three during baseflow [<0.02 m3/second] and five during highflow 3/second] conditions ) in order of increasing rates of flow ...................... 71 2 13 Percent capture of marked T. fluminensis shoots for each of six release events (thr ee during baseflow [<0.02 m3/second] and three during highflow 3/second] conditions) in order of increasing rates of flow ...................... 72 2 14 A dry seed of Ruellia simplex ............................................................................. 73 2 15 A seed of Ruellia simplex 30 minutes following the addition of one drop of deionized water to the seed coat ........................................................................ 74 2 16 An example of a marked Ruellia simplex seed that was intercepted in a net following the release of seeds into Tumblin Creek ............................................. 75 2 17 A group of Ruellia simplex seeds hanging from a needle following the addition of deionized water ................................................................................. 76 2 18 Intercepted seeds numbered 13,269 and represented at least 34 families ........ 77 2 19 Intercepted shoots numbered 179 and represented at least 14 families ............ 78 2 20 Mean interception rates (number/100 m3 of flow) of Ruellia simplex seeds per month during May – December 2012. .................................................................. 84 2 21 Mean interception rates (number/100 m3 of flow) of Ruellia simplex shoots per month during May – December 2012. ............................................................ 85 11

PAGE 12

2 22 Mean interception rates (number/100 m3 of flow) of Tradescantia fluminensis shoots per month during May – December 2012. ................................................. 86 2 23 Mean interception rates (number/100 m3 of flow) of Ruellia simplex seeds during May – December 2012 by baseflow (<0.02 m3/second), highflow 3/sec ond), and all stream conditions ..................................................... 87 2 24 Mean interception rates (number/100 m3 of flow) of Ruellia simplex shoots during May – December 2012 by baseflow (<0.02 m3/second), highflow 3/second), and all stream conditions ..................................................... 88 2 25 M ean interception rates (number/100 m3 of flow) of Tradescantia fluminensis shoots during May – December 2012 by baseflow (<0.02 m3/second), high3/second), and all stream conditions. ............................................ 89 2 26 Mean interception rates (number/100 m3 of flow) of Ruellia simplex seeds during highflow conditions broken down by storm (basin actively receiving precipitation) and post storm condi tions (basin no longer receiving rain but still experiencing high3/second]) during May – December 2012. ................................................................................................. 90 2 27 Mean interception rates (number/100 m3 of flow) of Ruellia simplex shoots during highflow conditions broken down by storm (basin actively receiving precipitation) and post storm conditions (basin no longer receiving rain but still experiencing highf 3/second]) during May – December 2012. ................................................................................................. 91 2 28 Mean interception rates (number/100 m3 of flow) of Tradescant ia fluminensis shoots during highflow conditions broken down by storm (basin actively receiving precipitation) and post storm conditions (basin no longer receiving rain but still experiencing highflow conditions [<0.02 m3/second]) during May – December 201 2. ........................................................................................ 92 2 29 Ruellia simplex seed interception rates (seeds/100 m3 of flow) for each of 27 net deployments during May – December 2012 in order of increasing rates of flow 93 2 30 Ruellia simplex shoot interception rates (shoots/100 m3 of flow) for each of 27 net deployments during May – December 2012 in order of increasing rates of flow. ................................................................................................................ 94 2 31 Tradescantia flum inensis shoot interception rates (shoots/100 m3 of flow) for each of 27 net deployments during May – December 2012 in order of increasing rates of flow. ...................................................................................... 95 3 1 Examples of materials used in submergence trials of Ruellia simplex seeds in the water column ([a] and inset at upper right) and below the sediment ([b] and inset at left) treatments. ............................................................................. 103 12

PAGE 13

LIST OF ABBREVIATIONS & ACRONYMS cm centimeter g gram(s) Gy gray hrs hours IFAS Institute of Food and Agricultural Sciences FDEP Florida Department of Environmental Protection FLEPPC Florida Exotic Pest Plant Council FLMNH Florida Museum of Natural History FWC Florida Fish and Wildlife Conservation Commission kGy kilogray km kilometer(s) m meter(s) mm millimeter(s) n n umber (of samples) NRCS Natural Resources Conservation Service PTFE lined polytetrafluoroethylene lined PVC polyvinyl chloride sec. second sig. significant sp. species spp. subspecies st. dev. standard deviation undet. undetermined USGS United States Geological Survey var. variety 13

PAGE 14

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science ASSESSING STREAM MEDIATED SEED AND SHOOT DISPERSAL OF INVASIVE PLANTS IN AN URBAN RIPARIAN WETLA ND By Jason C. Seitz December 2014 Chair: Mark W. Clark Major: Soil and Water Science Ruellia simplex and Tradescantia fluminensis were introduced to Florida and elsewhere and are now considered invasive plants . Millions of dollars are spent annually on the control of invasive plants in Florida alone. The goals of this study were to determine the importance of stream dispersal in invasive plants such as R. simplex (by seed) and T. fluminensis (by shoots) in ri parian wetlands. The 13, 269 seeds and 1 79 shoots intercepted during net deployments in a Florida stream included 131 seeds and 20 shoots of R. simplex and 6 shoots of T. fluminensis . Net verificatio n tests showed low retention of R. simplex seeds while r etention of T. fluminensis shoots was 100% . The low s eed retention values for R. simplex suggested that the actual number of seeds transported downstream during net deployments m ay have been in the thousands. Numbers of propagules per year transported downstream are estimated at 18,000– 30,000 R. simplex seeds , 8 ,000 – 10,000 R. simplex shoots , and 2,000 – 8 ,000 T. fluminensis shoots . C apture rates of marked propagules showed mobility to be low for R. simplex seeds but high for T. fluminensis shoots. Ruel lia simplex seeds are negatively buoyant while T. fluminensis shoots are somewhat positively buoyant but become negatively buoyant within 15 days of submergence. Percent germination of 14

PAGE 15

net captured R. simplex seeds was 6.9% in a germination chamber and 21.8% inside a gla sshouse. Survivorship with growth of n et captured shoots amounted to 22.2% for R. simplex and 40.0 % for T. fluminensis . H and harvested R. simplex seeds remained viable following up to 90 days of submergence and 30 days of bur ial in sedi ments. H andharvested T. fluminensis shoots survived up to 10 days of submergence. The results of this study reveal the importance of stream transport of propagules as a dispersal mechanism for these species . It is recommended that control efforts be co nducted at the watershed scale, beginning at upstream portions of a watershed and working downstream to reduce the potential for recolonization, at least for these two species. 15

PAGE 16

CHAPTER 1 THE IMPORTANCE OF SEED AND SHOOT DISPERSAL IN INVASIVE PLANT S Introduction Reducing the effects of invasive nonindigenous species, especially plants, is an important part of restoration and management efforts in natural areas of the southeastern United States, as these species cause significant stress to native ecosy stems (Adams and Steigerwalt 2010) and reduce native plant and animal diversity (Elton 1958). Even as invasive species are known to affect most natural areas of the southeast, little is known about recolonization after removal efforts (Elton 1958 ; Villazo n 2009), and research is needed to understand and predict the success of control efforts. Due to the increasing invasion of Florida’s remaining natural areas by nonindigenous plants, invasive species removal will remain an essential part of restoration ef forts (Villazon 2009). Dispersal traits are perhaps the single most important factor of a species’ propensity to be invasive (Meisenburg 2007), yet dispersal mechanisms are poorly understood for invasive plants. The State of Florida alone spends approxim ately $30 million annually to combat invasive species (Hupp 2007). Thus, it is important to understand invasive species dispersal and recruitment mechanisms in order to best allocate financial resources and provide the best and most cost effective restoration results. The following two species were chosen as test subjects to help invest igate potential dispersal mechanisms of invasive plants . Since the focus of this study was on stream mediated dispersal, the two species were chosen partly for their tendency to live in or along wetlands. They were also chosen for their w ide areas of introduction , not just in Florida but elsewhere in the United States and in other countries, where they are 16

PAGE 17

often known to be pests . Lastly but perhaps most importantly , they were chosen because together they represent a range of very effective reproductive strategies . Ruellia simplex C. Wright 1870 (Mexican Petunia) Ruellia simplex (Mexican petunia, referred to as R. brittoniana or R. tweediana by some authors) ( Figure 11[ a ] ) is a perennial herb reaching a maximum height of 1 m (Langeland et al. 2008). It was introduced to Florida prior to 1940 (Wilson and Mecca 2003) and other southern states, Bermuda, a nd parts of the Antilles including Dominican Republic, Puerto Rico, a nd Trinidad and Tobago ( Ezcurra and Daniel 2007; Liogier and Martorell 1982) . In Florida the species colonizes wetlands and disturbed uplands (Hupp 2007). The native range extends from Mexico and Central Ameri ca into parts of South America although the species was described from Cuba ( Ezcurra and Daniel 2007). Reproduction is by explosive dehiscence ( Witz t um and Schulgasser 1995; Hupp 2007) with an average of 20.6 seeds produced per capsule ( Wilson and Mecca 2003). The s eeds are catapulted as far as 3 m from the parent plant (Witztum and Schulgasser 1995). Ruellia simplex also reproduces vegetatively via rhizomes, stem sprouts, and cuttings (Langeland et al. 2008). The species has been listed as a Category I invasive species by the Florida Exotic Pest Plant Council (FLEPPC) since 2001 based on documented ecological damage ( Hupp et al. 2009; FLEPPC 20 13 ). The latest assessment (May 2010) by the Institute of Food and Agricultural Sciences (IFAS) Inva sive Plant Working Group concluded that R. simplex is n ot recommended for any uses in central and southern Florida and specified uses in northern Florida currently lack IFAS approval ( Cooper 2012; IFAS 2012). Ruellia simplex is a popular landscape plant in Florida due to its attractive flowers and foliage and extreme hardiness in soil types ranging from hydric to near xeric conditions (Hupp 2007 ; Hall and Weber 2011). 17

PAGE 18

Annual sales amount to over $10 million in Florida alone ( Wirth et al. 2004; Wilson et al. 201 2 ). There are 1 1 R. simplex cultivars currently available in Florida, yet only 4 ; ‘purple showers’ , ‘Mayan purple’ , ‘Mayan white’ , and ‘Mayan pink’; are sterile or produce nonviable seeds (Hupp et al. 2009; Freyre et al. 2012 ; Freyre and Wilson 2014) . The remaining seven cultivars are capable of producing viable seeds and are potentially invasive (Table 11) . In Gainesville, established populations of R. simplex produce abundant small seeds (ca. 1.6 – 2.2 m m diameter ) ( Figure 11[ b] ) and occur along many urban creek banks and riparian wetlands (Huey et al. 2007; Hupp 2007; personal observation). Figure 11 . Ruellia simplex is dominant along many urban waterways and riparian wetlands in Gainesville, Florida. Figure 11 (a) shows a mature plant in bloom with ripening seed capsules along Tumblin Cr eek (Gainesville, Florida). Figure 11 (b) shows seeds and a seed capsule harvested from the same creek. Seeds are dispersed up to 3 m from the parent plant via explosive dehiscence ( Witztum and Schulgasser 1995). Photos by author. a b 18

PAGE 19

Table 1 1 . Comparisons of reproduction in Ruellia simplex cultivars and the naturalized type. Sexual reproduction Vegetative reproduction Cultivar Fruit production (yes/no) Seed production (yes/no) Germination rate1 R hizomes (yes/no) Rooting of cuttings (yes/no) ‘Chi Chi’ Yes Yes Trade scantia fluminensis Vellozo 1825 (Small leaf Spiderwort) Tradescantia fluminensis (small leaf spiderwort, Figure 1 2 [ a ] ) is a perennial subsucculent herb native to portions of Brazil and Argentina ( da Concei o Vellozo 1825; Maule et al. 1995) where it produces about 6 seeds per fruit with peak seed production occurring October through February (M.O.O. Pellegrini, Rio de Janerio Botanical Garden, pers. comm., 06/17/14). It is established in at least 15 countries including Australia (Orchard 1994), Chile (Cuevas et al. 2004), the Republic of Nauru in Micronesia (Thaman et al. 1994), Russia (Tolkach et al. 1990), Spain ( Standish 2001a), the Kingdom of Thailand (Holm et al. 1979), Turkey (Tan 1984; 2012), Italy, Japan, Kenya, Portugal (Faden 2000), and the southeastern United States including Puerto Rico (Standish 2001b ) , and Hawaii (Staples et al. 2006). T he species does not produce seeds in Florida (McMillan 1999; Butcher and Kelly 2011) possibly due to a combination of self incompatibility and the original introduction of only very few 19

PAGE 20

individuals (R.B. Faden, Department of Botany, Smithsonian Institution, pers. comm ., 06/09/14). Tradescantia fluminensis has been recognized as a weed in the sout heastern U.S. since 1947 (Langeland et al. 2008) and is now listed as a Category I invasive species by FLEPPC based on documented ecological damage (FLEPPC 201 3 ). The species spreads readily by vegetative means with stem fragments as small as a single nod e successfully rooting into a new plant (Kelly and Skipworth 1984; Hurrell and Lusk 2012). In Florida, the species has colonized much of the stream banks and hydric and mesic hardwood forests of Gainesville (Figure 1 2 [ b ] ) and elsewhere in the northern and central portions of the state (Hall and Weber 2011) . In north ern Florida it forms dense monocultures and reduces recruitment of native species (Schmitz et al. 1997) (Figure 1 2[c]) . The species was found to be invasive in Alachua County floodplain fore sts and mesic hammocks by McMillan (1999), where it reduce d the relative abundance and diversity of native plants in test plots. McMillan (1999) further found the species out competes established native Oplismenus hirtellus (basketgrass) and recommended further research on controlling this invader, concluding that the species should be prohibited from sale in Florida. Tradescantia fluminensis is currently excluded from both the f ederal and Florida Department of Agriculture and Consumer Services noxious weed lists ( http://plants.usda.gov/java/noxious ). An assessment in September 2006 by the IFAS Inva sive Plant Working Group concluded that this invasive species is not recommended for any use in norther n or central Florida (Booth Binczik 2006; IFAS 2012). The species is currently targeted for control by the Florida Department of Environmental Protection 20

PAGE 21

(FDEP) due to its propensity to disturb Florida’s conservation lands (University of Florida 2005). Figure 12 . Tradescantia fluminensis (a) forms dense monocultures (b) and reduces recruitment of native species (c) in northern Florida and elsewhere. Reproduction occurs only by vegetative means in Florida. Photos by author. b a c 21

PAGE 22

Stream mediated Dispersal Mechanisms Water mediated seed and shoot dispersal plays an important role in structuring plant communities in riparian wetlands (Adams et al. 2007), and hydraulic and hydrologic conditions play significant roles in geneti c dispersal, controlling plant colonization and development (Riis and Biggs 2003). Many plants have evolved efficient and successful dispersal and colonization strategies relying on stream dynamics (Riis and SandJensen 2006). Stream mediated plant dispersal was found to be important in studies by SandJenson et al. (1999), Combroux et al. (2001), and Riis (2008), among other studies. Stream conditions combined with high seed or shoot production may allow rapid re colonization of invasive plants such as R. simplex and T. fluminensis following control efforts within the lower reaches of a stream if such efforts do not include the upper watershed. The planting of exotic vegetation by homeowners or business owners upstream of the floodplain could facilitat e the spread of invasive plants by providing sources of genetic material. The seeds or shoots from such plantings could be transported by hydrologic mechanism downstream to the floodplain. If the source of genetic material is not addressed in the upstream portion of the watershed, then there is reason to believe that recolonization of the floodplain is likely to occur. For this reason, control efforts may be more effective if conducted at the watershed scale. Goals and Objectives This study focused on determining the importance of stream dispersal of invasive plant seeds and shoots and the subsequent retention and recolonization of these species within a floodplain. The speci fic goals of this study were to 22

PAGE 23

1. d evelop a technique to intercept and retain R. simplex seeds and shoots and T. fluminensis shoots in the stream; 2. d etermine potential rates of recruitment of invasive plants such as R. simplex and T. fluminensis by quantifying stream dispersed seeds and shoots; 3. e stimate potential stream dispersed seed and shoot viability rates of R. simplex and T. fluminensis ; 4. i dentify possible mechanisms for the dispersal of seeds and shoots and establish the relationship between seed and shoot dispersal and water flow regime. The objectives outlined above were designed to evaluate the relative importance of watershed invasive seed and shoot sources along w ith stream dispersal mechanisms to determine their effect on restoration efforts on a floodplain forest within an urban sett ing. Hypotheses H 1) Null Hypothesis : The technique used here is not able to intercept and retain R. simplex seeds and shoots and T. fluminensis shoots in the stream. Alternative Hypothesis: The technique used here intercepts and retains R. simplex seeds and shoots and T. fluminensis shoots in the stream. H 2) Null Hypothesis : Stream dispersal is not a mechanism for recruitment of R. simplex (by seeds and shoots) and T. fluminensis (by shoots). Alternative Hypothesis: Stream dispersal is a mechanism for recruitment of R. simplex (by seeds and shoots) and T. fluminensis (by shoots). H 3) Null Hypothesis : Stream dispersal occurs mainly from seeds, rather than shoots, in R. simplex . Alternative Hypothesis: Stream dispers al occurs mainly from shoots, rather than seeds, in R. simplex , or both occur at about the same amount. H 4) Null Hypothesis : Stream flow regime will not significantly affect the numbers of stream dispersed R. simplex seeds and shoots. Alternative Hypothesi s : Stream flow regime will significantly affect the numbers of stream dispersed R. simplex seeds and shoots. 23

PAGE 24

H 5) Null Hypothesis : Percent R. simplex seed germination will not be significantly different between stream dispersed and control (handpicked) seeds. Alternative Hypothesis: Percent R. simplex seed germination will be similar between stream dispersed and control (handpicked) seeds. H 6) Null Hypothesis : Ruellia simplex seed germination rates will not be significantly affected by submergence in the st ream for 30 days or more. Alternative Hypothesis: Ruellia simplex seed germination rates will be significantly affected by submergence in the stream for 30 days or more. H 7) Null Hypothesis : Tradescantia fluminensis shoot viability rates will not be significantly affected by floating or submergence in the stream for 1 0 days or more. Alternative Hypothesis: Tradescantia fluminensis shoot viability rates will be significantly affected by floating or submergence in the stream for 1 0 days or more. H 8) Null Hypothesis : Ruellia simplex seeds are not buoyant (they sink to the bottom) when placed in water. Alternative Hypothesis: Ruellia simplex seeds are buoyant (they float at or near the water surface) when placed in water. H 9) Null Hypothesis : Tradescantia fluminensis shoots are not buoyant (they sink to the bottom) when placed in water. Alternative Hypothesis: Tradescantia fluminensis shoots are buoyant (they float at or near the water surface) when placed in water. Study Site: Tumblin Creek at Bivens Arm The Tumblin Creek watershed located in Gainesville, Florida, was used to address these hypotheses and objectives. The Tumblin Creek floodplain (Figure 1 3 ) is a 30acre forested wetland serving as the terminus between T umblin Creek and Bivens Arm (Sc hmidt 2005). Tumblin Creek is a first order perennial stream that originates at a small spring. It drains southern portions of Gainesville as it flows 3.7 k m through backyards, industrial complexes, and parks (Figure 1 4 ) . An estimated 60% o f the creek’ s approximately 6.1km2 watershed is impervious surface (CH2M Hill 1985; Orange Creek Basin Working Group and FDEP 2008). Many of the lands along its 24

PAGE 25

banks contain invasive plants , including R. simplex and T. fluminensis . The presence of dense patches of T. fluminensis at Bivens Arm Nature Park was reported in an unpublished thesis by McMillan (1999). Figure 13 . The 30 acre Tumblin Creek floodplain forest is situated primarily on University of Florida land along the north shore of Bivens Arm in Gainesville, Florida. 2012 Aerial photo courtesy, Google Earth . Nort h 25

PAGE 26

Figure 14 . The 6.1 km2 Tumblin Creek watershed (delineated with a green line) includes much of s outhern Gaines ville as it flows 3.7 km southwest to Bivens Arm via the floodplain forest . 2012 Aerial photo courtesy, Google Earth. The floodplain has been heavily altered by artificial channelization and the creation of berms. These actions have a ffected its hydrology , including allowing heavy sediment loads to enter the floodplain during storm events (Schmidt 2005). Also, considering that hydrologic regime is likely a pr imary factor controlling the establishment North North 26

PAGE 27

of stream side macrophytes (Riis and Biggs 2003), the increased water velocity caused by channelization may limit native plant diversity. Most of the Tumblin Creek floodplain soils are of the Monteocha loamy sand soil series (74.6% of floodplain area) , but also include Bivens s and, 5 to 8 percent slopes (7.4 %), Blichton sand, 2 to 5 percent slopes (7.3%), and Bonneau fine sand, 2 to 5 percent slopes (3.5%), the remaining area being composed of wat er (5. 1%) and urban land (2.1%) ( Natural Resources Conservation Service [ NRCS ] 2010). Although the soil survey of Alachua County by Thomas et al. ( 1985) describe d the vegetation of the floodplain as chiefly Taxodium spp. (cypress), there is no evidence of Taxo dium spp. within the floodplain. Instead, the floodplain canopy is dominated by Nyssa sylvatica var. biflora (swamp tupelo), Fraxinus caroliniana (pop ash) , and Acer rubrum (red maple), with an understory dominated by Sambucus nigra canadensis (elderberry ) and groundcover consisting mainly of invasive herbaceous species (e.g., R. simplex, T. fluminensis, Colocasia esculenta [wild taro], Xanthosoma sagittifolium [elephant ear]) and other nonindigenous species such as Commelina cf. diffusa (common dayflower) and Oxalis debilis (pink woodsorrel). Tables A 1 and A 2 in Appendix A contain lists of invasive and nonindigenous plants and animals, respectively, observed by the author in and around the study site. Directly downstream of the study site is Bivens Arm , a 187.8acre hypereutropic water body with 6.2 km of shoreline, a mean depth of 1 .2 m (Hoyer and Canfield 1994), and a maximum depth of 2 m ( Nordlie 1975). Bivens Arm is hydrologically connected to Paynes Prairie (US GS 1890), a 70.2km2 preserve (Andersen 2001) serving as a premier ecological and historic area hosting more than 750 plant species, about 350 27

PAGE 28

vertebrate species, and countless invertebrates (Nelson 1995). The lake occasionally flows into Paynes Prairie during times of high water , and both Bi vens Arm and the prairie are susceptible to invasion by upstream sources including R. simplex and T. fluminensis . Mean monthly rainfall for the Tumblin Creek basin is lowest in November ( 5.5 cm ) and highest in June ( 17.2 cm ), with most rain occurring betw een May and October (Figure 1 5 ). Flow rates vary significantly , even between consecutive days. Figure 16 shows daily flow rates , and Table 12 shows mean flow rates per month and mean monthly volumes for a threeyear period approximately 300 m upstream of the floodplain forest . Figure 15 . Monthly mean precipitation in the Tu mblin Creek basin July 2000 through June 2003. Source: http://averagerainfall cities.findthedata.org/d/d/Florida. Precipitation (cm) 28

PAGE 29

Figure 16 . Daily flow rates for Tumblin Creek , approximately 300 m upstream of the floodplain forest. The graph represents a threeyear time series from 1 July 2000 to 1 July 2003. A maximum flow of 0.30 m3/second was recorded on 20 June 2001, and a minimum flow of 0.000057 m3/second was recorded on 16 June 2002. Source: St. Johns River Water Management District unpu blished data ( gauge # 1990209) . Table 1 2 . Monthly mean flow rates , by base flow (<0.02 m 3 /second) and high flow ( 0.02 m3/second), and mean volumes for the period July 2001 through June 2003 at Tumblin Creek approximately 300 m upstream of the floodplain forest. Month Percent of time Mean flow rate (m 3 /second) Mean volume (m 3 ) Base flow High flow Base flow High flow January 32 68 0.016 0.033 74,113 19,556 February 30 70 0.015 0.050 95,553 60,473 March 24 76 0.017 0.048 108,694 34,023 April 70 30 0.016 0.046 63,648 31,142 May 67 33 0.011 0.052 66,968 26,623 June 39 61 0.009 0.074 126,427 29,334 July 55 45 0.010 0.068 96,070 16,825 August 37 63 0.013 0.053 101,144 19,907 September 40 60 0.014 0.072 125,990 22,066 October 31 69 0.014 0.042 88,495 22,066 November 27 73 0.011 0.049 100,472 41,317 December 26 74 0.011 0.063 131,643 97,603 TOTAL – – – – 1,179,216 356,048 Source: calculated from St. Johns River Water Management District unpublished data ( gauge # 1990209). 29

PAGE 30

Past Control Efforts within the Tumblin Creek Floodplain Efforts to control invasive plants within the Tumblin Creek floodplain have been ongoing for several years. Kestrel Ecological Services, Inc. workers, including the author, conducted invasive plant control measures beginning February 2009 within the Unive rsity of Florida’s Bivens Rim property , including most of the Tumblin Creek floodplain. By June 2009, much of the floodplain was treated with herbicides using basal bark, girdle, stump treatment, and foliar application methods. Although a wide variety of nonindigenous plants exist within the floodplain, control efforts focused on plant species of particular concern, such as R. simplex and T. fluminensis . Tables A 1 and A 2 in Appendix A list nonindigenous plants and other organisms, respectively, observed by the author during control efforts, investigative walks, and field sampling . The control treatments for R. simplex and T. fluminensis appear to have had unsatisfactory long term impact on th ese species, as both species remain dominant in the floodplain forest. Tradescantia fluminensis , in particular, remains by far the most dominant groundcover species in the floodplain forest. The lack of longterm impact on T. fluminensis densities in the floodplain forest may be partly attributable to the species ’ considerable regrowth potential combined with the difficulties in treatment methods to optimize pesticide transmission rates through the thick waxy c uticle of this semi succulent. Impacts to R. simplex in the floodplain forest were noticeable for the fi rst few years following treatment, and the density of this species appears to be reduced from per treatment levels as of this writing . However, R. simplex was identified as the sixth most dominant species in the floodplain forest based on a field survey conducted in April and May 2012 . Overall, present day conditions in the floodplain forest indicate that the control measures were ineffective at producing satisfactory long term reductions in 30

PAGE 31

densities of these species. This suggests the need for a more comprehensive watershedscale control effort to be applied to both in situ areas of invasive plants and upstream sourc es of genetic material to optimize control efforts in a watershed. 31

PAGE 32

CHAPTER 2 FREQUENCY AND MOBILITY OF SEEDS AND SHOOTS IN STREAM Introduction Th is chapter addresses objectives 1, 2, and 4 . Objective 1 is fully addressed by developing a technique to intercept and retain seeds and shoots in Tumblin Creek . Objective 2 is partially addressed by quantif ying stream dispersed seeds and shoots. Objective 4 is fully addressed by identif ying possible mechanisms for dispersion and relationships between seed and shoot dispersal and water flow regime. Hypotheses H 1 through H 4, H 8, and H 9 are partially or fully addressed in this chapter . Materials and Methods Much of the materials and methods used are based on Riis and Sand Jensen (2006) and Riis (2008), who investigated the dispersal and colonization of plants in lowland streams of Denmark. Description of Stream Reach Tumblin Creek between SW 13th Street (US 441) and Bivens Arm was described by measuring the width, water depth, current velocity, and percent vegetative cover at evenly distributed transects perpendicular to the bank and spaced every 25 m (33 total transects) along the 800m long reach (Figure 2 1) . Within the floo dplain, only the main flow channel was sampled. Water depth was recorded every 10 cm along each transect using a measuring tape and meter stick. Stream velocity was measured once per transect on 26 May 2012 using dye tracer , and flow rates were calculated from velocity, depth, and width measurements using methods presented in McCarthy (2009). Mean flow velocity and volume entering the nets was calculated for each sampling event , including both baseflow and highflow (during or directly following storm event) 32

PAGE 33

conditions. For the purposes of this s tudy, base flow is defined as a flow rate of <0.02 m3/second while high flow is defined 3/second. Percent vegetative cover was recorded on 29 April and 5 May 2012 within 3 m upslope of the mean high water l ine along Tumblin Creek between SW 13th St reet (US 441) and Bivens Arm using a 1 m2 sampling square. Sampled quadrants were spaced every 25 m along both sides of the creek (66 total quadrants) . An emphasis was made to document the pr esence of R. simplex and T. fluminensis that could serve as seed and shoot sources for recolonization of the floodplain. The occurrence of R. simplex and T. fluminensis was assessed using a qualitative survey of the remaining upstream portions of Tumblin Creek from the headwaters to SW 13th St reet (US 441) . 33

PAGE 34

Figure 21 . A total of 33 transects (yellow circles with numbers) were mar ked every 25 m along the 800 m long stream reach between SW 13th St reet (US 441) and Bivens Arm. The blue line represents the m i dline of the main channel at the time of transect placement on 26 May 2012. Net deployments were conducted at Transect 7, 300 m upstream of Bivens Arm and directly above the floodplain forest. 2012 Aerial photo courtesy, Google Earth. North 34

PAGE 35

Evaluation of Net Performance and Mobility of Seeds and Shoots This se ction partially tested hypotheses H 1 and H 2 . A 6 m wide seine net with 8 mm stretch mesh was strung across the width of the main channel to intercept and retain marked shoots of T. fluminensis . One or two smaller (0.5m per side) finemesh (0.8 mm) net s constructed of nylon screening were used to intercept and collect marked R. simplex seeds (Figure 2 4[ a and b]). The location of the net deployments was approximately 300 m upstream of Bivens Arm and 500 m downstream of SW 13th Street (US 441). Effectiveness of the nets to capture and retain target species was evaluated by placing 50 to 100 marked seeds and shoots directly into the mouths of the nets at the initiation of net deployments and recording the numbers retained following each deployment . Three net verification deployments were made during baseflow and thr ee during highflow conditions. Stream velocit ies were measured using dye tracer (velocities were often too low for accurate measurements using a flow meter) . Flow rates were calculated from mean values of velocity, depth, and width measurements recorded at the initiation and termination of each net deployment using methods presented in McCarthy (2009) . Seeds and shoots were handharvested from local wild stock and painted fluorescent colors ( white, yellow, orange, green, or red) to facilitate recognition (Figure 2 2 [a]) . Seeds were sterilized inside a Gammacell 1000 cesium 137 irradiator at 8.16 Gy/minute for 31 hours (15 kGy total) prior to release to prevent additional downstream colonization. Following standard soak times (i.e., 24hours for baseflow events, 1hour or until flow was impeded during highflow events) as stated above, retained seeds and shoots were sorted from the leaves and debris intercepted in the nets and quantified to obtain an estimate of percent retention. An ultravi olet lamp 35

PAGE 36

was used to facilitate the sorting of the small fluorescent painted R. simplex seeds which were otherwise difficult to locate among the nonseed material. The buoyancy of painted and unpainted material was qualitatively compared by placing painted and unpainted propagules together in Tumblin Creek for 24 hours and recording their relative position in the water column at the start and end of the 24hour testing period. Buoyancy of painted versus unpainted R. simplex seeds was compared by submerg ing 15 painted and 15 unpainted seeds in each of 10 400mL decontaminated beakers ( 30 seeds per beaker, 300 seeds total) secured in Tumblin Creek using a plastic crate, wire, fine mesh screening , and rebar stakes. A group of 40 T. fluminensis painted shoots was placed together with 40 unpainted shoots within each of three cube shaped cages constructed of nyloncoated 0.6cm mesh hardware cloth attached to a PVC frame with zip ties. Each cage measured 0.6m per side and was secured to the stream bed using rebar stakes. This was done to ensure that the painted seeds and shoots had similar buoyancy as those occurring in the stream. No differences were observed in the buoyancy of painted and un painted seeds or shoots during the 24hour testing period. Painted propagules appeared to have relative buoyancy similar to that of the unpainted propagules . To evaluate mobility of seeds and shoots, known numbers of marked seeds and shoots were released u pstream of the nets to record capture rates. Ruellia simplex seeds were released 1, 10, and 20 m upstream of the fine mesh nets , while T. fluminensis shoots were released 150 and 300 m upstream of the seine net (Figure 2 2 [ b ]) . Three release events were conducted during baseflow and three during highflow conditions, with the exception of five releases conducted during high36

PAGE 37

flow conditions using T. fluminensis shoots 150 m upstream of the nets. In addition, two releases of R. simplex seeds were conducted 150 m upstream of the nets during highflow conditions to help assess longer distance dispersal scenarios. Seeds and shoots were handharvested from wild stock and painted fluorescent colors as stated above. Seeds were sterilized as stated above pri or to release to prevent additional downstream colonization. Following standard soak times as stated above, retained marked seeds and shoots (Figure 2 2 [ c]) were sorted from the leaves and debris intercepted in the nets and quantified to determine numbers intercepted. An ultraviolet lamp was used to facilitate the sorting of the small fluorescent painted R. simplex seeds , which were otherwise difficult to locate among the nonseed material. 37

PAGE 38

Figure 22 . Groups of propagules were marked with fluorescent colors and released into the stream at certain distances upstream of the nets to determine mobility. Shown here are Tradescantia fluminensis shoots that were painted white (a) and released into Tumblin Creek (b) upstream of the nets. At the end of each deployment , the number s captured of each color (c) were recorded to determine percent captured per release distance. Ruellia simplex seeds (not shown) were also marked and released upstream of the nets . Photos by author. b a c 38

PAGE 39

Interception of Seeds and Shoots To test hypotheses H 1 and H 2, either wholly or in part, nets were deployed three to four times monthly during May – December 2012, for a total of 27 net deployments . Figure 2 3 compares flow rates between net deployments. A 6 m wide seine net with 8 mm stretch mesh was strung across the width of the main channel to intercept and collect large transported shoots of both species. One or two smaller (0.5m per side) finemesh (0.8mm) net s constructed of nylon screening were used to intercept and collect transported R. simplex seeds in a narrow portion of the stream (Figure 2 4 [ a and b ] ). T he location of the net deployments was approximately 300 m upstream of Bivens Arm and 500 m downstream of SW 13th St reet (US 441). This location was upstream and adjacent to the floodplain forest. Stream velocities were measured using dye tracer (velocities were often too low for accurate measurements using a flow meter). Flow rates were calculated from mean values o f velocity, depth, and width measurements recorded at the initiation and termination of each net deployment using methods presented in McCarthy (2009). N ets were soaked for approximately 24 hours during baseflow (<0.02 m3/second) or 1 hour during highfl ow conditions 3/second), or until flow was impeded by buildup of material in the nets. High flow conditions were further broken down into two subcategories: storm and post storm conditions. Storm conditions were defined as occurring while Tumblin Creek basin was actively receiving precipitation. Post storm conditions were defined as occurr ing within a few hours following a precipitation event in the basin and while the creek continued to experience highflow condit ions. Table 21 lists net deploymen t s per month and per flow regime. 39

PAGE 40

Table 2 1 . Net deployments by month and by flow regime in Tumblin Creek, May – December 2012. High flow 1 Month Base flow2 Storm3 Post s torm4 Total May 2 2 0 4 Jun 2 2 0 4 Jul 0 1 2 3 Aug 0 2 1 3 Sep 2 0 1 3 Oct 2 1 0 3 Nov 3 0 0 3 Dec 1 1 2 4 TOTAL 12 9 6 27 1 H igh flow is defined as a flow rate of 0.02 m3/second. 2 Base flow is defined as a flow rate of <0.02 m3/second. 3 Storm conditions is defined as the basin actively receiving precipitation. 4 Post storm conditions is defined as the basin n o longer receiving rain but still experienc ing high flow conditions . Figure 23 . Rates of flow for each of 27 net deployments conducted May – December 2012 in Tumblin Creek to intercept seeds and shoots. The red line denotes the approximate threshold between baseflow (<0.02 m3/second) and highflow ( 0.02 m3/second) regimes. Several deployments had flow rates too low to be seen on this graph. 40

PAGE 41

Nets were emptied and rinsed of material following each deployment. Figure 2 5 shows the organic material and trash retained in a seine net after a 24h our deployment during a high flow storm ev ent. Seeds and shoots were quantified per deployment and per taxon after being carefully separated from the leaves and debris intercepted in the nets. Live shoots were identified, quantified, and measured directly following net deployments. Material was handremoved from the seine net and highpressure hoses were used to remove material from the finemesh nets. Large sticks, g arbage, rotten wood, and fishes were removed and the remaining material from the nets was labeled per net , laid out on a 1 m2 fine mesh frame suspended above the ground in a protected area (a garage or enclosed porch) and left until dry to the touch (up to 2 weeks ). The material was turned over once or twice daily to facilitate drying , to discourage mold development, and to prev ent seeds from germinating while drying. W ith the help of a bright light, a magnifying glass, and fine forceps, seeds were then separated from the leaves and nonseed material. None of the 27 net deployments during May – December 2012 required the net capt ured material to be fieldsplit. However, an exploratory deployment during 17 – 18 March 2012 resulted in an estimated 40 gallons of material in the seine net and required that the material be split in half before sorting ; the results were multiplied by 2 t o represent the entire sample. Species identification was conducted with the aid of a microscope and compar ison with material of known origin. Comparative material included the author’s personal collection of s eeds and fruits and the seed collections at the Florida Museum of Natural History (FLMNH) Herbarium. Along with the use of comparative material, seed and fruit identifications were further assisted by literature such as Harlow (1946), U.S. Forest Service (1948), Musil (1963), 41

PAGE 42

Delorit (19 70), Lander s and Johnson (1976) , and rare titles held at the FLMNH H erbarium library. Identification of shoots was conducted using comparative material, the online Atlas of Florida Vascular Plants ( http://florida.pl antatlas.usf.edu/ ), and regional guides and keys to vascular plants . Specialists from the FLMNH Herbarium and the Florida Archaeology Collection and Archaeobotany were consulted whenever an identity was in question. The statistical significance of numbers of R. simplex seeds and shoots and T. fluminensis shoots was determined by the use of a oneway analysis of variance (ANOVA) prior to the use of t tests. The t test s compar ed actual numbers of seeds or shoots interc epted and retained during the 27 net deployments to predicted numbers intercepted and retained under the H 1 null hypothesis scenario (if none were caught). A p value of less than 0.05 was considered to be significantly different from the H 1 null hypothesis. A p value less than 0.001 was considered statistically highly significant. The p value was not corrected or adjusted because the results of multiple individual tests are important for the purposes of this study , and a general null hypothesis (that all null hypotheses are true simultaneously) is not of interest here ; therefore, the exact p value was used following Perneger (1998) and Armstrong (2014). Microsoft Excel was used for all calculations. 42

PAGE 43

Figure 24 . A seine net and at least one finemesh net was strung across Tumblin Creek to intercept seeds and shoots as they were transported downstream. Photograph (a) shows a deployment during very low baseflow conditions ( 18 June 2011). Photograph (b) shows a deployment during highflow conditions (3 October 2012). Photos by author. a b 43

PAGE 44

Figure 25 . Material intercepted in a seine net strung across Tumblin Creek, Gainesville, Florida, during 17– 18 March 2012. A brief but heavy rain storm resulted in highflow conditions during the night of net deployment, when approximately 40 gallons of organic material and trash were retained in the net. This sample was sorted for seeds and shoots, resulting in 2,092 seeds of at least 20 speci es and 18 shoots of at least 5 species. Photo by author. Effects of Flow on Seed and Shoot Dispersal This section tested hypothesi s H 4. To determine the effects, if any, of flow on seed and shoot dispersal, statistical comparisons were made of seed or shoot count vs. flow per sampling event and per flow condition. Mean values per flow condition were compared and scatter plots were generated including a line of best ‘ fit ’ and an R2 44

PAGE 45

(coefficient of determination) value. Additional comparisons were made of estimated variances from within each of the two major flow regimes and between flow regimes using a oneway ANOVA prior to the use of t tests. Results of t tests having a p value of less than or equal to 0.05 were consi dered to be significantly different between flow regimes. A p value less than 0.001 was considered statistically highly significant. The p value was not corrected or adjusted because the results of multiple individual tests are important for the purposes of this study and a general null hypothesis (that all null hypotheses are true simultaneously) is not of interest here ; therefore, the exact p value was used following Perneger (1998) and Armstrong ( 2014). Microsoft Excel was used for all calculations. Buoyancy To determine if seeds and shoots are suitable for water dispersion, buoyancy of propagules was evaluated to test hypotheses H 8 and H 9 . Hand harvested R. simplex seeds were submerged in Tumblin Creek using 400mL decontaminated beakers (15 seeds per beaker, 150 seeds total) secured in the stream using a plastic crate, wire, fine mesh screening , and rebar stakes (Figure 2 6 ) . Beakers were decontaminated using Liquinox detergent, deionized water, and pesticidegrade isopropanol following FDEP standard operation procedure FC1000 prior to the experiment. Seeds were checked at least weekly and their position in the water column recorded. Because preliminary observations strongly suggested that the s eeds were negatively buoyant , the beakers were allowed to become completely submerged in the stream during buoyancy testing. The seeds were submerged for a total of 32 days during September – October 2012. 45

PAGE 46

The mean density of R. simplex seeds was determined from a series of measurements taken with a laboratory scale (OHaus Adventurer Pro, accurate to 0.0001 g). Seed capsules were handharvested in October 2013 from a small population along Hogtown Creek in Gainesville and allowed to naturally dehisce at room temperature (22 – 25C) for two weeks befor e the seeds were stored for use in this and other experiments . A total of 703 seeds were weighed to determine the mean weight per seed. Twenty five percent of these same seeds ( n = 174) were then used to determine volume by displacement of a known weight and volume of deionized water (at room temperature) within a 2mL plastic vial secured with a water tight ( O ring ) seal. To evaluate the buoyancy of T. fluminensis , a group of 120 shoots was placed within each of three cube shaped cages in May 2013 to test for buoyancy. Each cage was constructed of nyloncoated 0.6cm mesh hardware cloth secured to a polyvinyl chloride ( PVC ) frame with zip ties (Figure 2 7 [a]) . Cages measured 0.6 m per side and w ere secured to the stream bed with rebar stakes . Th e 0.6cm mesh hardware cloth allowed free flow of water through the devices (Figure 2 7 [b]) . The positions of the shoots in the water column were recorded twice weekly throughout the duration of the experiment . Shoots remained in the cages in the water c olumn until all shoots showed advanced stages of decay . The mean density of T. fluminensis shoots was determined from a series of measurements taken with a laboratory scale (Scientech SL5200D , accurate to 0.01 g). Shoots were handharvested the morning of weighing (on 17 June 2014) from the population at the Tumblin Creek floodplain forest. Moisture on the surface of the shoots was removed by a combination of patting dry with towels and evaporation in an air 46

PAGE 47

conditioned room over the course of a few hours . Shoots were cut into sections averaging 76 mm each and weighed to determine the mean weight per length of shoot . The same shoots were then used to determine volume by displacement of a known weight and volume of deionized water (at room temperature) with in a 500mL wide mouth glass jar secured with a polytetrafluor o ethylenelined (PTFE lined) water tight polypropylene closure. 47

PAGE 48

Figure 26 . Ten 400 mL decontaminated beakers with finemesh screen tops (a) were used to submerge Ruellia simplex seeds (15 seeds per beaker, 150 seeds total) to determine buoyancy (temperature logger also shown) . B eakers with seeds (yellow arrow) were completely submerged in Tumblin Creek ( b) and secured with wire mesh and rebar stakes. Photos by author. a b 48

PAGE 49

Figure 27 . Three cub e shaped cages , measuring 0.6m per side and constructed of hardware cloth on PVC frames (a), were used to submerge Trade scantia fluminensis shoots in Tumblin Creek for determination of buoyancy and viability following submergence. Each cage held 120 shoots (b) cut to the mean length of those found in the nets (ca. 18 cm). Shoots were allowed to achieve a natural position in the water column. Photos by author. a b 49

PAGE 50

Res ults and Discussion Description of Stream Reach Approximately 275 m of the 800m long stre am reach between Bivens Arm and SW 13th St reet ( US 441) travels through the floodplain forest, with the remaining 525 m occurring upstream of the forest and south of SW 13th St reet. The stream reach had a mean width of 232 137 cm , a mean depth of 83 71 m m , and a mean flow rate of 0. 012 0.010 m3/second recorded on 26 May 2012 during baseflow conditions. The substrate was primarily sand , with the exception of the area within 100 m downstream (south) of SW 13th St reet, which was concrete. Table B 1 in Appendix B summarizes flow rates, water depths, stream widths, and substrate types every 25 m along the stream reach. At least 61 species of herbaceous and small woody plants, including 19 nonindigenous and invasive species, were recorded within 3 m of the mean highwater line along the stream reach. In the floodplain forest, T. fluminensis was by far the most dominant species , and R. sim plex was the sixth most dominant species, with a mean percent aerial coverage of 19.25% 27.13 % and 2.42 % 10.00% , respectively . Upstream of the floodplain forest, Acmella oppositifolia (oppositeleaf spotflower) was the most dominant species between 300 and 550 m from the lake (10.14% 20.15% ) , while Bidens alba (beggerticks) was the most dominant species between 575 and 800 m from the lake (12.15% 1 6.81 % ). The dominance of R. simplex and T. fluminensis decreased with increasing distance fr om the lake, and T. fluminensis was not present by 575 m from Bivens Arm . Nevertheless, T. fluminensis was the single most dominant plant along the stream reach , with an overall mean of 8.12% 19.81%. Ruellia simplex was the seventh most dominant plant overall, with a mean of 50

PAGE 51

2.56% 7.22 %. Table B 2 in Appendix B summarizes mean percent aerial coverage of plants along the 800m long stream reach. Upstream of SW 13th Street , R . simplex was frequently seen in dense st ands along the banks of the creek from about 50 m downstream of the headwaters (2.8 km upstream of Bivens Arm ) and continuing downstream . It was even found growing through concrete rubble and inside box culverts, where it was observed flowering . Ruellia simplex appeared to have the most flowers and (or) seed capsules when growing in full or nearly full sun. Tradescantia fluminensis was observed beginning about 2.7 km upstream of Bivens Arm , where it formed a monoculture along the concrete armoring, m ostly in shade. The spe cies was seen only occasionally downstream of this area and upstream of SW 13th Street, typically in shady areas . Overall, sources of genetic material for R. simplex and T. fluminensis occurred throughout most of Tumblin Creek. Th us, seeds and shoots intercepted during net deployments may have originated near the headwaters of the creek, directly upstream of the nets, or anywhere inbetween. Table B 3 in Appendix B provides a summary of the qualitative survey conducted along the m ain channel of the creek from the headwaters to SW 13th Street in June 2011. Evaluation of Net Performance and Mobility of Seeds and Shoots The fine mesh nets were found to have low rates of retention of R. simplex seeds . Mean retention rates were 6.3% 6.5 % for baseflow and 35.3% 51.0% for high flow conditions (Table 2 2) . Based on a line of best “fit” and associated R2 value, a weak inverse correlation exists between percent r etention of R. simplex seeds and baseflow conditions ( R2 = 0.0464 [Figure 2 8 ]). However, the low retention rate for R. 51

PAGE 52

simplex seeds may partially reflect the difficulties in recognizing the m arked seeds during sorting. T he circular outline and morphological characteristics of the seeds were often obscured by organic material and sand grains adhered to the mucilaginous seed coat and mucilage hairs (see Figure 21 6 ). In contrast with the low retention results for R. simplex seeds , t he seine was found to retain 100% of T. fluminensis shoots during baseflow and highflow conditions . This strongly suggests that all intercepted shoots were retained in the seine regardless of stream condition. Tables 2 2 and 2 3 include mean percent retention rates for painted R. simplex seeds and T. fluminensis shoots, respectively. M ean percent capture of marked R. simplex seeds ranged from only 0. 3 % 0. 6 % to 13.0% 22.5% when placed 1 to 20 m upstream of the net s (Table 22). There does not appear to be any correlation between percent capture of R. simplex seeds and flow rate for transport distances of 1 to 20 m , based on lines of best “fit” and associated R2 values ( Figures 2 8 to 2 10) . None of 200 marked seeds released 150 m upstream of the nets duri ng supplementary testing were intercepted. All T. fluminensis shoots placed in the nets were retained during both baseflow and highflow regimes. Tradescantia fluminensis shoots were transported as much as 300 m in 0.7 hours during mobility trials, although this distance was achieved only d uring highflow conditions and with a low mean percent capture (6.7% 6.5% [Table 2 3]). Percent capture for marked T. fluminensis shoots released 150 m upstream of the nets was higher during highflow than during base flow conditions, with 28.0% 24.5% and 5.0% 8.7% , respectively. A strong positive correlation exists between percent capture of marked T. fluminensis shoots and highflow conditions for transport 52

PAGE 53

distances of 150 and 300 m based on lines of best “fit” and associated R2 values (Figures 2 11 to 2 1 2 ). Based on these results , and those of the seed and shoot interception testing (below) , the H 1 null hypothesis is rejected. The technique used in this study successfully intercepted and retained at least some R. simplex seeds and all T. fluminensis shoots placed into the nets. The netting methods also intercepted and retained a total of 131 wild R. simplex seeds and 6 wild T. fluminensis shoots. The low percent retention of R. simplex seeds placed in the fine mesh nets may be a result of one or more factors such as low recognition rates amongst the organic material in the nets, the possible movement of seeds out of the nets, or possible consumption of the brightly painted seeds by fishes such as Gambusia holbrooki (eastern mosquitofish [see discussion below on mobility of seeds]). The mobility of R. simplex seeds appears to be limited based on the low percent capture of marked seeds placed up to 20 m upstream of the nets combined with th e n egative buoyancy of the seeds. The fact that none of the 200 marked seeds placed 150 m upstream during the supplementary testing were intercepted in the nets is consistent with the theory that R. simplex seeds have limited mobility. Alternatively, it is possible that R. simplex seeds may be spread by animals such as fishes or wading birds. A brief exploratory test by the author consisted of adding deionized water to a group of handpicked R. simplex seeds and observing the tenacity of the seeds to stic k to each other and to other objects. The author was able to pick up a small group of the wetted seeds using only a needle (see Figure 2 1 7 ). The tendency of the seeds to stick to one another and to other objects is likely due to a combination of 53

PAGE 54

the muc ilage hairs and the mucilaginous content discharged from the seed following the addition of moisture. It is conceivable that wading birds may, on occasion, have seeds stick to their legs, feet, or toes while wading in water and inadvertently transport them to other portions of the same water body or an entirely different water body. Long distance bird mediated dispersal via ectozoochory was suggested by Ezcurra and Daniel (2007) to partially explain the disjunct distributions of R. simplex and other memb e rs of the family Acanthaceae. Although R. simplex lacks fleshy fruit often used to entice birds to consume endozoochoric seeds, i t would be interesting to test the possibility of bir d mediated seed dispersal of R. simplex seeds via ectozoochory . Seeda nd fruit eating fishes such as those of the family Characidae play a similar role in seed dispersal in southern parts of the Amazon River basin as do birds (Goulding 1980; Smith 1981). During periods of high water, Amazonian floodplain forests can be inundated, including portions of the tree canopy, making seeds largely unavailable to most birds yet easily accessible to fishes. Ichthyochory is obviously much less important in North America but seed consumption has been documented for North American fishes such as Ictalurus punctatus (channel catfish). Ictalurus punctatus was found to consume seeds of Ulmus americana (American elm) and Vitis sp. (grape) in the Des Moines River, Iowa (Bailey and Harrison 1948). Seeds of Morus rubra (red mulberry) and Forestiera acuminata (eastern swamp privet) have been found in the stomachs of I. punctatus by Chick et al. (2003) and Adams et al. (2007), respectively. This catfish is common throughout much of the U.S. including Florida ( Page and Burr 2011) . In the present study, the author observed G. holbrooki being attracted to the brightly painted seeds placed into the stream during mobility trials. The 54

PAGE 55

G. holbrooki were seen to strike at the seeds on s everal occasions during the course of the trials , although during most occasions t he fish appeared to eject the seeds from their mouth rather than swallow them . The possible effects of ichthyochory as an additional method of dispersal for invasive plants such as R. simplex should be studied to better understand dispersal mechanisms of invasive plants. Conversely to the apparently limited mobility of R. simplex seeds in water, T. fluminensis shoots are capable of being tran sported at least 300 m during high flow conditions, and therefore appear to be rather mobile. The fact that the shoots were transported 300 m in as little as 0.5 hours in a relatively small stream suggests that the species may be capable of being transported kilometers downstream if given adequate time and flow conditions. Water mediated shoot dispersal may even include estuarine and marine environments . Hurrell and Lusk (2012) found that shoot fragments can survive up to 48 hours in full strength seawater. Clearly, additional research is needed to better estimate the mobility of T. fluminensis shoots in riverine environments. In addition to water mediated dispersal, shoots of T. fluminensis have been documented to be dispersed lodged in the hooves of Bos taurus (cattle) (Ogle and Lovelock 1989). The species may also become lodged in the feet of Gallus gallus (domestic chickens) (Standish 2001 a ). These mechanisms may further increase the chances of dispersal to additional areas. Interception of Seeds and Shoots The volume passing through the nets during the 27 net deployments totaled 27,279 m3 and is only 2% of the average annual volume in the stream reach ( see Table 1 2 for average annual flow rates ). Overall sampling rates were 44% during baseflow 55

PAGE 56

and 56% during high flow conditions . These rates agree well with the average baseflow and highflow conditions for May – December at this stream reach, at 40% and 60%, respectively . A total of 13,269 seeds, representing at least 34 families, were intercepted during the 27 net deployments in Tumblin Creek. Nonindigenous plant seeds totaled 1,785, represented at least 16 species, and amounted to 13.5% of all seeds intercepted. Plants known to be invasive in Florida (FLEPPC categories I or II) totaled 747 and represented 5.6% of all seeds. A total of 131 seeds of R. simplex were intercepted. Figure 21 8 shows the numbers of seeds intercepted by family and as a percentage of total seeds. A total of 179 shoots representing at least 14 families were intercepted by nets in Tumblin Creek. Nonindigenous plants were represented by 59 shoots from six species and amounted to 40.0% of all shoots. Twenty shoots of R.simplex and six shoots of T. fl uminensis were intercepted, representing 11.2% and 3.4% of all shoots, respectively. Other invasive species of shoots intercepted consisted of Cyperus involucratus (umbrella plant [ n = 2 ] ) and Panicum repens ( torpedo grass [ n = 2 ] ). Figure 21 9 shows the numbers of shoots intercepted by family and as a percentage of total shoots intercepted. A one way ANOVA revealed a significant difference between the numbers of R. simplex seeds, R. simplex shoots, and (or) T. fluminensis shoots intercepted and retaine d during the 27 net deployments and the prediction un der the H 1 null hypothesis ( no seeds and [or] shoots intercepted and retained) ( ANOVA p = 0.0063 [Table 28] ) . A significant difference was found between numbers of R. simplex seeds intercepted and retained and the H 1 null hypothesis prediction ( t test p = 0.031 [Table 29] ). No 56

PAGE 57

significant differences were found between numbers of R. simplex shoots or T. fluminensis shoots intercepted and retained versus the H 1 null hypothesis ( t test p = 0. 17 [ R. simplex shoots], t test p = 0.16 [ T. fluminensis shoots]) (Table 29) . The H 1 hypothesis is rejected based on the results of intercepting R. simplex seeds and shoots, as the technique used here intercepted and retained at least some R. simplex seeds and shoots and T. fluminensis shoots in the stream , and the numbers o f R. simplex seeds intercepted and retained were significantly greater than the H 1 null hypothesis prediction . The H 1 alternative hypothesis is supported, at least in terms of R. simplex seeds . Considering the numbers of R. simplex seeds and shoots and T. fluminensis shoots intercepted in the stream, the H 2 null hypothesis is rejected. The H 2 alternative hypothesis is supported as stream dispersal appears to be a mechanism for recruitment in these species. These results failed to reject the H 3 null hypothesis, as 86.8% of dispersed propagules of R. simplex were intercepted in the form of seeds, with only the remaining 13.2% as shoots. The R. simplex seeds often bec a me coated with organic material and sand grains when in the stream due to the mucilaginous seed coat and mucilage hairs (see Figures 21 4 and 21 5 ). This trait along with the small size of the seeds and constraints on seed entrainment and retention in the nets make accurate quantification of these seeds difficult among the organic material intercepted in the nets. For this reason, the number of seeds intercepted in the nets should be considered a minimum value. F igures 2 1 3 through 2 1 5 show a dry seed of R. simplex , the same seed 30 minutes following wetting with a single dr op of deionized water , and an example of a marked seed coated with sand and organic debris following capture in the nets, respectively . 57

PAGE 58

When the retent ion rate found for seeds placed into the nets is applied as a conversion factor for the numbers of wild seeds of R. simplex intercepted and retained during this study, then it becomes clear that the numbers of seeds transported in the stream during net deployments was likely much greater . Applying retention conversion factors to the numbers of seeds intercepted during baseflow and highflow conditions results in the following values. Transported seeds (base flow) = 77 seeds / 6.3% 22.5% = 1,222 5,432 seeds Transported seeds (highflow) = 54 seeds / 35.3% 51.0% = 153 300 seeds Thus, the actual number of R. simplex seeds transported downstream in the vicinity of the nets during deployments may have been in the thousands during baseflow conditions and in the hundreds during highflow conditions. Similarly, applying retention conversion factors to mean interception rates/100 m3 by fl ow regime results in the following values. Seeds/100 m3 (baseflow) = 0.778 / 6.3% 22.5% = 12.3 54.9 seeds/100 m3 Seeds/100 m3 (high flow) = 0.173 / 35.3% 51.0% = 0.490 0.961 seeds/100 m3 Based on the above calculations, interception rates of seeds corrected using retention rates may be much higher than what interception rates alone suggest. Conversely, the numbers of T. fluminensis shoots intercepted in the nets, and rates of interception, appear to represent the true numbers of sho ots transported downstream during net deployments, as the retention rates were 100% during both baseflow and highflow conditions. Numbers of propagules flowing downstream into the floodplain forest can be estimated using mean yearly volume data from St. Johns River Water Management District (see Table 12 in Chapter 1). The number of R. simplex seeds transported 58

PAGE 59

downstream to the floodplain forest is estimated at 18,000 to 30,000 per year , taking into account rates of interception and retention conversi on factors from the present study. Numbers of shoots transported downstream are estimated at 8,000 – 10,000 and 2,000– 8 ,000 per year for R. simplex and T. fluminensis , respectively. Effects of Flow on Seed and Shoot Dispersal Most R. simplex seeds were intercepted during the month of September ( n = 85, 64.9% of all R. simplex seeds) , which also had the highest mean interception rate (2.99 seeds/100 m3 of flow) of the months sampled . This peak during late summer coincides with the highest seed production along with convective storm activity. Table 2 4 shows numbers of seeds per nonindigenous species intercepted by month and by flow regime. Table 2 5 shows mean numbers of seeds/100 m3 of flow for each nonindigenous species by month and by flow regime. F igure 2 20 compares mean interception rates of R. simplex seeds by month. Table C 1 in Appendix C summarizes seeds per species by month and by flow regime. Most R. simplex shoots were intercepted during the month of June , with n = 15 and 75.0% of all R. simplex shoots intercepted. June also had the highest number of T. fluminensis shoots intercepted , with n = 5 and 83.3% of all T. fluminensis shoots intercepted then. June also had the highest mean interception rates for R. simplex and T. fluminensis shoots of any month sampled, with 4.17 and 1.28 shoots/100 m3 of flow , for R. simplex and T. fluminensis shoots , respectively . Table 2 6 shows numbers of shoots intercepted for each nonindigenous species by month and by flow regime. Table 2 7 shows mean numbers of shoots/100 m3 of flow for each nonindigenous species by month and by flow regime. Figures 2 20 and 2 21 compare mean interception rates of 59

PAGE 60

shoots per month for R. simplex and T. fluminensis , respectively. Table C 2 in Appendix C summarizes shoots per species by month and by flow regime. C onsiderable variability was observed in interception rates of R. simplex seeds and shoots and T. fluminensis shoots. S eeds of R. simplex were intercepted with greater mean frequency during bas e flow conditions (0.778 seeds/100 m3 of flow) than during highflow conditions (0.173 seeds/100 m3). Shoots of R. simplex and T. fluminensis were intercepted with a higher mean frequency during highflow conditions, at 1.33 and 0.432 shoots/100 m3, respe ctively, compared to little or none during base3). Figures 2 2 2 and 2 2 3 compare mean interception rates by flow regime for R. simplex seeds and shoots, respectively. Figure 2 2 5 compares mean interception rates of T. fluminensis shoots by flow regime. The considerable variability between samples composing each mean is expressed in the standard error bars in Figures 2 2 2 through 2 2 4 . Of the two sub categories of highflow conditions, storm conditions ( meaning while t he basin was actively receiving precipitation) and post storm conditions ( meaning the basin was no longer receiving rain but was still experiencing highflow conditions 3/second]), the highest mean interception rate of R. simplex seeds was post sto rm , with 0.423 seeds/100 m3 of flow. The opposite was true for mean interception rates of R. simplex and T. fluminensis shoots , as mean interception rates were highest during storm conditions , at 2.17 and 0.720 shoots/100 m3 of flow, respectively . However, the considerable variability observed in the results of each subcategory, combined with somewhat low sample sizes (9 storm and 6 post storm deployments), make the results unclear. Figures 2 2 5 through 2 2 7 show mean interception rates by 60

PAGE 61

storm a nd post storm conditions for R. simplex seeds , R. simplex shoots , and T. fluminensis shoots, respectively. The r ate of interception for R. simplex seeds has a weak inverse correl ation with baseflow conditions based on the line of best ‘fit’ ( R2 = 0.0148) (Figure 2 2 9 ) . Rates of interception for R. simplex and T. fluminensis shoots showed little or no correlation with flow rate based the line of best ‘fit’, at R2 = 0.0065 and R2 = 0.0082, respectively (Figures 2 2 9 and 2 30) . N o significant differences were found between rates of interception for R. simplex or T. fluminensis propagules between baseflow and highflow conditions (ANOVA p = 0.5 0 ) (Table 210 ) . The H 2 null hypothesis is rejected based on the results of the mobilit y testing using marked seeds and shoots. The results of mark and capture testing suggest that stream dispersal is a mechanism for recruitment for the species tested, as propagules were transported up to 20 m for R. simplex seeds and up to 300 m for T. flu minensis shoots during relatively limited time periods and under natural stream conditions. Therefore, the H 2 alternative hypothesis is supported as stream dispersal may be a mechanism for recruitment of R. simplex and T. fluminensis . T he H 4 null hypoth esis was not rejected as stream flow regime does not significantly a ffect the numbers of stream dispersed seeds or shoots of R. simplex or the shoots of T. fluminensis . Capture rates for R. simplex seeds were slightly higher during extremely low and extremely high flow conditions, and capture rates of R. simplex and T. fluminensis shoots were slightly higher during extremely low flow conditions than during moderate to high flow conditions. However, there is considerable 61

PAGE 62

variability in capture rates between deployments and within each of the two main stream flow condition categories , and the results did not differ significantly . Buoyancy Results of the in stream buoyancy experiment using R. simplex seeds suggested that the seeds lack buoyancy . All 150 seeds descended to the bottom of the water column within a minute of being added to the water, and remained at the bottom for the remainder of the 32day trial. The mucilaginous coating rapidly appear ed on the seeds once submerged. Table 2 1 1 summarizes observations recorded during the experiment. R esults of the R. simplex seed mass and volume measurements undertaken in the laboratory indicate a mean density ( 1.39 g/cm3) that exceeds that of pure water (1 g/cm3) at room temperature, making the seeds sink (negative buoyancy) . It is likely that the density of the seeds also exceeds that of stream water, such as at Tumblin Creek, based on the behavior of the seeds during the instream buoyancy experi ment . Table 2 12 summarizes seed mean mass, volume, and density data and indicates sample sizes . Results of the buoyancy experiment using T. fluminensis shoots suggest that shoots are somewhat positively buoyant when first placed in the stream , but become neutrally to negatively buoyant within 15 days of submergence. The reduction in buoyancy appeared to coincide with the beginning stages of decay in the shoots. The results of the buoyancy experiment agree well with T. fluminensis shoot mass and volume measurements undertaken in the laboratory. The somewhat positive buoyancy of the shoots when initially placed in stream water was backed up by the laboratory 62

PAGE 63

results that showed a mean density of 0.958 g/cm3. Table 2 1 3 summarize s observations recorded during the experiment. Table 2 1 4 summarizes mean mass, volume, and density data per 50mm section of shoot . The H 8 null hypothesis was not rejected based on the results of the buoyancy test and density calculation using R. simplex seeds . The seeds of this species are not buoyant in water. Although Hupp’s ( 2007) investigation of R. simplex had a concluding remark that the seeds of this species stay buoyant in water and that the mucilaginous gel coat allows for t he positive buoyancy , the present study indicates that R. simplex seeds do not float in water and therefore having positive buoyancy must not be an advantage relative to having a mucilaginous coating. Past authors have suggested that oily seed coatings or a rough surface texture may aid in enhancing buoyancy in the seeds of some species. For example, Ikeda and Itoh (2001) found that the seeds of Penthorum chinense (Penthoraceae) floated in distilled wa ter but not in river water that contained surfaceactive ingredients or in a solution containing a linear alkylbenzene sulfonate (an anionic surfactant contained in synthetic laundry detergents). The surfactant apparently disrupted the oily seed coat and effectively altered buoyancy. Ikeda and Itoh (2001) also described the seed morphology of P. chinense as allowing an increased surface area and enhancing the interfacial tension and buoyancy of the seeds. This is apparently not the case for R. simplex in the present study, however, as these seeds behaved similarly in deionized water and in Tumblin Creek water (they lacked buoyancy in both instances) and have a relatively smooth surface (see Figure 214 ). Although such a surfactant commonly associated wit h laundry detergents is likely to occur in this urban creek, the fact that the 63

PAGE 64

R. simplex seeds lacked buoyancy in deionized water suggests that if the seeds contained an oily coating, such a coating did not significantly affect buoyancy. The H 9 null hypothesis is rejected based on the results of the buoyancy experiment and density calculation using T. fluminensis shoots. The H 9 alternative hypothesis is supported as the shoots are moderately buoyant during at least the first day following submergence in the stream. However, by the end of the first week, most shoots became negatively buoyant coinciding with tissue decay. 64

PAGE 65

Table 2 2 . C apture and retention rates for marked Ruellia simplex seeds in Tumblin Creek. Three releases per distance category (0 [in net] , 1, 10, and 20 m upstream) were made during baseflow ( <0.02 m3/second) and three during high0.02 m3/second) conditions. Two additional releases at 150 m were conducted during hig h flow conditions. Release distance upstream (m) Number of seeds released Mean flow rate (m 3 /sec.) St. dev. Stream condition Mean soak time (hrs) St. dev. Mean flow volume (m 3 ) St. dev. Mean capture or retention rate (%) St. dev. (%) 0 (in net) 200 0.011 0.008 Base flow 23.8 1. 3 608.3 832.3 6.3 6.5 0 (in net) 150 0. 697 0.939 High flow 0.7 0. 2 1293.5 1665.4 35.3 51.0 1 300 0.011 0.008 Base flow 23.8 1. 3 608.3 832.3 13.0 22.5 1 3 00 0.697 0.939 High flow 0.7 0. 2 1293.5 1665.4 0. 3 0. 6 10 250 0.011 0.00 8 Base flow 23.8 1. 3 608.3 832.3 4.7 5.0 10 300 0. 697 0.939 High flow 0.7 0. 2 1293.5 1665.4 7.7 13.3 20 300 0.011 0.008 Base flow 23.8 1. 3 608.3 832.3 4.7 6. 4 20 300 0.697 0.939 High flow 0.7 0. 2 1293.5 1665.4 3.0 3.6 150 * 200 0.185 0.180 High flow 11.9 16.0 307.6 347.8 0.0 0.0 * Two releases were conducted at 150 m upstream during highflow conditions. These are supplemental to the 0to 20m releases. 65

PAGE 66

Table 2 3 . C apture and retention rates for marked Tradescantia fluminensis shoots in Tumblin Creek. Three releases per distance category (0 [in net] , 150, and 300 m upstream) were made during baseflow ( <0.02 m3/second) and between three and five during high 0.02 m3/second ) conditions. Release di stance upstream (m) Number of shoots released Mean flow rate (m 3 /sec.) St. dev. Stream condition Mean soak time (hrs) St. dev. Mean flow volume (m 3 ) St. dev. Mean capture or retention rate (%) St. dev. (%) 0 (in net) 150 0.011 0.008 Base flow 23.8 1.3 608.3 832.3 100 0.0 0 (in net) 150 0. 697 0.939 High flow 0.7 0.2 1293.5 1665.4 100 0.0 150 300 0.011 0.008 Base flow 23.8 1.3 608.3 832.3 5.0 8.7 150* 463 0. 492 0. 928 High flow 5.1 10.1 1813.7 2089.8 28.0 24.5 300 300 0.011 0.008 Base flow 23.8 1.3 608.3 832.3 0.0 0.0 300 300 0.697 0.939 High flow 0.7 0.2 1293.5 1665.4 6.7 6.5 *A total of five releases were conducted at 150 m upstream during highflow conditions. 66

PAGE 67

Figure 2 8 . Percent retention of marked R. simplex seeds for each of six release events ( three during baseflow [ <0.02 m3/second] and three during highflow [ 3/second] conditions ) in order of increasing rates of flow. Each event consisted of 50 to 100 marked seeds placed in the nets . The line of best fit (curved black line) uses a secondorder polynomial. The R2 value suggests a weak inverse correlation between percent retention and baseflow conditions. Rate of flow (m 3 /second) Percent retention of seeds 67

PAGE 68

Figure 2 9 . Percent capture of marked R. simplex seed s placed 1 m upstream of the nets for each of six release events ( three during baseflow [ < 0.02 m3/second ] and three during highflow [ 3/second] conditions ) in order of increasing rates of flow. Each event consisted of 100 marked seeds . The line of best fit (curved black line) uses a n exponential scale . The R2 value suggests no correlation between percent capture and flow conditions . Percent captured seeds Rate of flow (m 3 /second) 68

PAGE 69

Figure 2 10. Percent capture of marked R. simplex seed s placed 10 m upstream of the nets for each of six release events ( three during baseflow [ <0.02 m3/second ] and three during highflow [ 3/second] conditions ) in order of increasing rates of flow. Each event consisted of 50 to 100 marked seeds . The line of best fit (curved black line) uses a n exponential scale . The R2 value suggests no correlation between percent capture and flow conditions . Rate of flow (m 3 /second) Percent captured seeds 69

PAGE 70

Figure 2 11. Percent capture of marked R. simplex seed s placed 20 m upstream of the nets for each of six release events ( three during baseflow [ <0.02 m3/second ] and three during highflow [ 3/second] conditions ) in order of increasing rates of flow. Each event consisted of 100 marked seeds . The line of best fit (curved black line) uses a n exponential scale. The R2 value suggests no correlation between percent capture and flow conditions . Rate of flow (m 3 /second) Percent captured seeds 70

PAGE 71

Figure 2 12. Percent capture of marked T. fluminensis shoots for each of eight release events ( three during baseflow [ <0.02 m3/second] and five during highflow [ 3/second ] conditions ) in order of increasing rates of flow. Each event consisted of 63 to 100 mark ed shoot s placed 150 m ups tream of the nets . The line of best fit (curved black line) uses a secondorder polynomial . The R2 value suggests a strong positive correlation between percent capture and high flow conditions . Rate of flow (m 3 /second) Percent captured shoots 71

PAGE 72

Figure 2 13. Percent capture of marked T. fluminensis shoots for each of six release events ( three during baseflow [ <0.02 m3/second] and three during highflow [ 3/second] conditions) in order of increasing rates of flow. Each event consisted of 100 marked shoot s placed 300 m upstream of the nets. The line of best fit (curved black line) uses a secondorder polynomial. The R2 value suggests a positive correlation between percent cap ture and highflow conditions . Rate of flow (m 3 /second) Percent captured shoots 72

PAGE 73

Figure 214. A dry seed of Ruellia simplex . The hilum is located at the top of the image. Note the mucilage hairs lying on the surface of the seed. Scale at left represents 1 mm. Photo by author. 73

PAGE 74

Figure 215. A seed of Ruellia simplex 30 minutes following the addition of one drop of deionized water to the seed coat. The hilum is located near the bottom of the image. Note the mucilage hairs radiating out from the wetted seed along with mucilaginous content discharged from the seed and shown as the milky discoloration of the water surrounding the seed. Scale at left represents 1 mm. Photo by author. 7 4

PAGE 75

Figure 2 16. An example of a marked Ruellia simplex seed that was intercepted in a net following the release of seeds into Tumblin Creek. Note the large amount of substrate that adhered to the seed while in the stream, including sand grains, organic detritus, and a wing of an insect (yellow arrow). The presence of orange fluorescent paint, applied to the seed pri or to release, was the main identifying characteristic of the intercepted seed. Scale at left represents 1 mm. Photo by author. 75

PAGE 76

Figure 2 17. A group of Ruellia simplex seeds hanging from a needle following the addition of deionized water. It is possible that R. simplex seeds may stick in similar fashion to the legs, feet, or toes of wading birds and be transported to areas not previously colonized. This method of birdmediated seed dispersal may allow the species to coloniz e new areas of the same basin or entirely new basins. Photo by author. 76

PAGE 77

Figure 2 18. Intercepted seeds numbered 13,269 and represented at least 34 families . Ruellia simplex numbered 131 seeds represented 1.0% of all seeds intercepted. Each pie slice represents the contribution of seeds by a single family of plants relative to the total number of seeds intercepted. 77

PAGE 78

Figure 2 19. Intercepted sh oots numbered 179 and represented at least 14 families . Ruellia simplex numbered 20 and represented 11.2% of all shoots. Tradescantia fluminensis numbered 6 and represented 3.4% of all shoots. Each pie slice represents the contribution of shoots by a single family of plants relative to the total number of shoots intercepted. 78

PAGE 79

Table 2 4 . Seeds of Ruellia simple x and other non indigenous and invasive plants per month and per flow regime in Tumblin Creek. Nets were deployed 27 times during May – December 2012 including base flow (<0.02 m3/second) and high3/second) conditions. Seeds per month Seeds per flow regime Species May Jun Jul Aug Sep Oct Nov Dec Total Base flow High flow Ruellia simplex1 2 22 0 19 85 3 0 0 131 77 54 cf. Butia capitata 0 0 3 0 0 0 0 0 3 0 3 Carya illinoinensis 0 2 0 0 0 0 0 0 2 0 2 Cinnamomum camphora1 66 137 120 43 18 15 4 73 476 28 448 Commelina cf. diffusa 4 16 4 0 3 3 0 0 30 19 11 Cucumis cf. sativus 0 3 0 0 0 0 0 0 3 0 3 Echinochloa crus galli 2 0 0 0 0 0 0 0 2 0 2 Helianthus cf. annuus 1 1 1 0 0 0 0 0 3 0 3 Hypochaeris sp. 3 0 0 0 0 0 0 0 3 2 1 Koelreuteria elegans2 18 27 32 5 3 0 0 3 88 6 82 Lagerstroemia indica 18 25 2 34 51 7 3 23 163 50 113 Macfadyena unguis cati1 0 1 0 4 1 0 0 16 22 0 22 Paspalum cf. notatum 0 91 0 0 0 0 1 2 94 6 88 Paspalum cf. urvillei 185 212 6 17 19 1 1 0 441 183 258 Sapium sebiferum 1 0 0 27 3 0 0 0 0 30 0 30 Ulmus cf. parvifolia 2 3 0 0 31 30 1 227 294 21 273 TOTAL 301 540 195 125 211 59 10 344 1785 392 1393 1 Florida Exotic Pest Plant Council category I invasive species (FLEPPC 20 13) 2 Florida Exotic Pest Plant Council category II invasive species (FLEPPC 20 13) 79

PAGE 80

Table 2 5 . Mean interception rates (individuals/100 m 3 ) for seeds of Ruellia simplex and other no nindigenous and invasive plants per month and per flow regime in Tumblin Creek. Nets were deployed 27 times during May – December 2012 including baseflow (<0.02 m3/second) and high3/second) conditions. Mean seeds /100 m 3 by month Mean seeds /100 m 3 by flow regime Species May Jun Jul Aug Sep Oct Nov Dec Base flow High flow All flow Ruellia simplex1 0.015 0.466 none 0.252 2.99 0.093 none none 0.778 0.173 0.441 cf. Butia capitata none none 0.034 none none none none none none 0.007 0.004 Carya illinoinensis none 0.201 none none none none none none none 0.054 0.030 Cinnamomum camphora1 12.0 15.7 1.96 3.56 0.391 0.782 0.450 5.47 0.609 9.80 5.71 Commelina cf. diffusa 0.690 0.577 0.418 none 0.155 0.177 none none 0.175 0.348 0.271 Cucumis cf. sativus none 0.302 none none none none none none none 0.080 0.045 Echinochloa crus galli 0.675 none none none none none none none none 0.180 0.100 Helianthus cf. annuus 0.008 0.101 0.007 none none none none none none 0.030 0.017 Hypochaeris sp. 1.22 none none none none none none none 0.295 0.090 0.181 Koelreuteria elegans2 2.77 2.88 0.439 0.450 0.054 none none 0.258 0.359 1.48 0.979 Lagerstroemia indica 1.13 2.52 0.014 3.06 1.08 0.273 0.250 1.74 0.400 2.05 1.32 Macfadyena unguis cati1 none 0.101 none 0.053 0.018 none none 1.27 none 0.380 0.211 Paspalum cf. notatum none 26.1 none none none none 0.183 0.322 0.18 6.94 3.94 Paspalum cf. urvillei 62.4 17.7 0.343 2.25 0.882 0.059 0.083 none 1.50 20.9 12.27 Sapium sebiferum1 none none 0.187 0.270 none none none none none 0.091 0.051 Ulmus cf. parvifolia 0.126 0.222 none none 0.559 1.55 0.122 17.3 0.277 4.94 2.86 TOTAL 81.1 66.9 3.40 9.90 6.13 2.94 1.09 26.4 4.58 47.5 28.4 1 Florida Exotic Pest Plant Council category I invasive species (FLEPPC 2013) 2 Florida Exotic Pest Plant Council category II invasive species (FLEPPC 2013) None = no seeds intercepted from that taxon for that month or flow regime. 80

PAGE 81

Table 2 6 . Shoots of Ruellia simplex , Tradescantia fluminensis , and other nonindigenous and invasive plants per month and per flow regime in T umblin Creek . Nets were deployed 27 times during May – December 2012 including baseflow (<0.02 m3/second) and high3/second) conditions . Shoot s per month Shoots per flow regime S pecies May Jun Jul Aug Sep Oct Nov Dec Total Mean length (mm) Mean # of leaves Mean # of roots Base flow High flow Ruellia simplex1 2 15 0 1 1 0 0 1 20 163 9.2 12.8 1 19 Tradescantia fluminensis1 1 5 0 0 0 0 0 0 6 178 5.0 5.0 0 6 Commelina cf. diffusa 3 10 7 3 0 0 0 5 28 137 4.1 3.9 3 25 Cyperus involucratus2 0 2 0 0 0 0 0 0 2 70 9.0 0.0 0 2 Hypochaeris sp. 1 0 0 0 0 0 0 0 1 219 4.0 0.0 0 1 Panicum repens1 0 0 1 0 0 0 0 1 2 403 2.5 6.0 1 1 TOTAL 7 32 8 4 1 0 0 7 59 – – – 5 54 1 Florida Exotic Pest Plant Council category I invasive species (FLEPPC 20 13) 2 Florida Exotic Pest Plant Council category II invasive species (FLEPPC 20 13) 81

PAGE 82

Table 2 7 . Mean interception rates for shoots of Ruellia simplex , Tradescantia fluminensis , and other nonindigenous and invasive plants per month and per flow regime in Tumblin Creek . Nets were deployed 27 times during May – December 2012 including baseflow (<0.02 m3/second) and high3/second ) conditions . Mean shoots/100 m 3 by month Mean shoots /10 0 m 3 by flow regime Species May Jun Jul Aug Sep Oct Nov Dec Base flow High flow All flow Ruellia simplex 1 0.675 4.17 none 0.090 0.018 none none 0.092 0.010 1.33 0.744 Tradescantia fluminensis1 0.338 1.28 none none none none none none none 0.432 0.240 Commelina cf. diffusa 0.463 2.22 0.053 0.270 none none none 0.175 0.057 0.782 0.460 Cyperus involucratus2 none 0.592 none none none none none none none 0.158 0.088 Hypochaeris sp. 0.008 none none none none none none none none 0.002 0.001 Panicum repens1 none none 0.007 none none none none 0.161 0.054 0.001 0.025 TOTAL 6.28 10.6 0.696 0.450 0.333 0.267 0.472 1.31 0.121 2.71 1.56 1 Florida Exotic Pest Plant Council category I invasive species (FLEPPC 20 13) 2 Florida Exotic Pest Plant Council category II invasive species (FLEPPC 20 13) None = no shoots intercepted from that taxon for that month or flow regime. 82

PAGE 83

Table 2 8 . Results of a one way ANOVA for numbers of intercept ed and ret ained Ruellia simplex seeds and shoots, and Tradescantia fluminensis shoots , during the 27 net deployments and the interception rates predicted under the H 1 null hypothesis (no seeds or shoots intercepted) . Source of variation Degrees of freedom Sum of squares f p f critical Significant ly different population means? Between groups 3 423.51 4.34 0.0063 2.69 Yes Within groups 104 3381.26 Total 107 3804.77 Table 2 9 . Results of two tailed t tests comparing numbers of Ruellia simplex seeds and shoots, and Tradescantia fluminensis shoots, intercepted and retained during the 27 net deployments to the predicted scenario under the H 1 null hypothesis (no seeds or shoots intercepted). Population Total number intercepted ( N ) Mean number per deployment St. dev. t test p Significantly different from H 1 null? R. simplex seeds 131 4.9 11.1 0.031 Yes R. simplex shoots 20 0.7 2.7 0.17 No T. fluminensis shoots 6 0.2 0.8 0.16 No H 1 null hypothesis 0 0.0 0.0 – – 83

PAGE 84

Figure 2 20. Mean interception rates (number/100 m3 of flow) of Ruellia simplex seeds per month during May – December 2012. The highest mean rate of interception of R. simplex seeds was in September but there was considerable variability (expressed as standard error with the black bars) . Seeds / 10 0 m 3 of flow 84

PAGE 85

Figure 2 21. Mean interception rates (number/100 m3 of flow) of Ruellia simplex shoot s per month during May – December 2012. The highest mean rate of interception of R. simplex shoots was in June but there was considerable variability (expressed as standard error with the black bars). Seeds / 10 0 m 3 of flow 85

PAGE 86

Figure 2 22. Mean interception rates (number/100 m3 of flow) of Tradescantia fluminensis shoots per month during May – December 2012. The highest mean rate of interception of T. fluminensis shoots was in June but there was considerable variability (expressed as standard error with the black bars). Shoots / 10 0 m 3 of flow 86

PAGE 87

Figure 223. Mean interception rates (number/100 m3 of flow) of Ruellia simplex seeds during May – December 2012 by baseflow (<0.02 m3/second), highflow 3/second), and all stream conditions. The highest mean rate of interception of R. simplex seeds w as during baseflow conditions but there was considerable variability (expressed as standard error with the black bars). Seeds / 10 0 m 3 of flow 87

PAGE 88

Figure 2 24. Mean interception rates (number/100 m3 of flow) of Ruellia simplex shoots during May – December 2012 by baseflow (<0.02 m3/second), highflow 3/second), and all stream conditions. The highest mean rate of interception was during high flow conditions but there was considerable variability (expressed as standard error with the black bars). Shoots / 10 0 m 3 of flow 88

PAGE 89

Figure 2 25. Mean interception rates (number/100 m3 of flow) of Tradescantia fluminensis shoots during May – December 2012 by baseflow (<0.02 m3/second), high3/second), and all stream conditions. The highest mean rate of interception was during high flow conditions but there was considerable variability (expressed as standard error with the black bars). Shoots / 10 0 m 3 of flow 89

PAGE 90

Figure 2 26. Mean interception rates (number/100 m3 of flow) of Ruellia simplex seeds during highflow conditions broken down by storm (basin actively receiving precipitation) and post storm conditions (basin no longer receiving rain but still experiencing high0.02 m3/second ]) during May – December 2012 . The highest mean rate of interception during high flow conditions was post storm . However, the considerable variability (expressed as standard error with the black bar), coupled with the low sample size, make the results unclear. Seeds / 10 0 m 3 of flow 90

PAGE 91

Figure 2 27. Mean interception rates (number/100 m3 of flow) of Ruellia simplex shoots during highflow conditions broken down by storm (basin actively receiving precipitation) and post storm conditions (basin no longer receiving rain but still experiencing high0.02 m3/second ]) during May – December 2012 . The highest mean rate of interception of shoots during high flow conditions was during storm conditions , including the “first flush” . However, the considerable variability (expressed as standard error with the black bar), coupled with the low sample size, make the results unclear. S hoots / 10 0 m 3 of flow 91

PAGE 92

Figure 2 28. Mean interception rates (number/100 m3 of flow) of Tradescantia fluminensis shoots during highflow conditions broken down by storm (basin actively receiving precipitation) and post storm conditions (basin no longer receiving rain but still experiencing highflow conditions [ <0.02 m3/second ]) during May – December 2012. The highest mean rate of interception during highflow conditions was during storm conditions, including the “first flush” . However, the considerable variability (expressed as standard error with the black bar), coupled with the low sample size, make the results unclear. Shoots / 10 0 m 3 of flow 92

PAGE 93

Figure 2 29. Ruellia simplex seed interception rates (seeds/100 m3 of flow) for each of 27 net deployments during May – December 2012 in order of increasing rates of flow . The line of best fit (curved black line) uses a logarithmic scale . The R2 value suggests that the rate of interception has a weak inverse correlation with baseflow conditions . Seeds / 10 0 m 3 of flow Rate of flow (m 3 /second) 93

PAGE 94

Figure 2 30. Ruellia simplex shoot interception rates (shoots/100 m3 of flow) f or each of 27 net deployments during May – December 2012 in order of increasing rates of flow. The line of best fit (curved black line) uses a secondorder polynomial. The R2 value suggests little or no correlation between rate of interception and flow rate. Shoots / 10 0 m 3 of flow Rate of flow (m 3 /second) 94

PAGE 95

Figure 2 31. Tradescantia fluminensis shoot interception rates (shoots/100 m3 of flow) for each of 27 net deployments during May – December 2012 in order of increasing rates of flow. The line of best fit (curved black line) uses a second order polynomial. The R2 value suggests little or no correlation between rate of interception and flow rate. Shoots / 10 0 m 3 of flow Rate of flow (m 3 /second) 95

PAGE 96

Table 2 10 . Results of a one way ANOVA for rate s of interceptio n for R uellia simplex seeds and shoots, and Tradescantia fluminensis shoots, during the 27 net deployments between baseflow ( < 0.02 m3/second) and highflow 0.02 m3/second) conditions . Source of variation Degrees of freedom Sum of squares f p f critical Significant ly different population means? Between groups 5 18.77 0.88 0. 50 2.34 No Within groups 75 319.46 Total 80 338.23 Table 2 11 . In stream testing for buoyancy in Ruellia simplex seeds. A total of 150 seeds were divided into 15 beakers ( n = 15 seeds/beaker) with mesh tops and submerged in Tumblin Creek for 32 days during September – October 2012. Days since initiation Position of seeds in beakers, further notes 0 Submerged crate with 15 beakers in stream. Seeds immediately developed opaque mucilaginous coating and descended to bottom of beakers . 1 Strong storm event during prior night. Seeds are buried in 200– 450 mL of sand at bottom of beakers. One beaker tipped over. Emptied all beakers of sand and placed right side up. Seeds continue to have mucilaginous coating. 3 Seeds mixed with detritus and sand at bottom of beakers. Annelid worms observed in 5 beakers. Emptied detritus and sand from beakers. Seeds have a thick coating of detritus. 14 Strong storms during recent days. Seeds mixed with detritus and sand at bottom of beakers. Stream banks were mowed prior to storms. Crate is partially buried in sand and vegetation and is filled with vegetative debris and sand to 10 cm above beakers. Mucilaginous coating on seeds is now thinner. Cleared log jam from above crate and emptied beakers and crate of sand and debris. 20 Small st orm during previous night buried crate completely under sand. Seeds at bottom of sandfilled beakers. Removed sand from crate and beakers, moved crate 5 m downstream to deeper pool. Annelid worms observed in beakers . 21 Seeds at bottom of beakers. Crate and beakers devoid of sand. 25 Seeds at bottom of beakers along with ca. 5 cm of sand . 26 Seeds buried under sand at bottom of beakers. Emptied all beakers of sand. 32 No change. Removed crate with beakers from stream. A later germination trial with 115 of these seeds in potting soil resulted in 9% germination over a 60day period. 96

PAGE 97

Table 2 12 . Mean mass, volume, and density of Ruellia simplex seeds hand harvested from an introduced population in Gainesville, Florida. Parameter Mean value per seed 1 n M ass 0.00107 g 703 V olume 0.00077 cm3 174 D ensity 1.39 g/cm3 174 1 Calculated from values obtained using a laboratory scale accurate to 0.0001 g . Table 2 13 . In stream testing for buoyancy in Tradescantia fluminensis shoots . A total of 360 18 cm shoots were divided into three 0.6 m3 cages ( n = 120 shoots/ cage ) and partially submerged in Tumblin Creek for 15 days during May 2013. Days following initiation Shoot positions in cages, further observations 0 Deployed 3 cages in stream. 30– 40 cm of water depth in each cage. Added shoots. All shoots floating at or just below water’s surface in all three cages, with most leaves above water . 7 Cage 1: Most shoots submerged below water’s surface. Water depth decreased to 20 cm. Cage 2: Most shoots submerged, resting on bottom, and coated with silt. Water depth ca. 40 cm. Cage 3: All shoots submerged, resting on bottom. Water depth decreased to 12– 13 cm. 15 Cage 1: The few remaining shoot fragments are resting on bottom. Water depth ca. 20 cm. Cage 2: Only 3 shoot fragments remaining, at bottom. Water depth ca. 40 cm. Cage 3: All remaining shoot fragments completely submerged and at bottom. Water depth 15– 20 cm. Most shoots are decayed and are no longer visible in cages. Table 2 14 . Mean mass, volume, and density of Tradescantia fluminensis shoots hand harvested from an introduced population in Gainesville, Florida. Parameter Mean value per 50 mm of shoot 1 Combined length of shoots (mm) Number of 50mm sections Mean number of leaves per section Mass 0.667 g 2356 47.1 3.3 Volume 0.697 cm3 2356 47.1 3.3 Density 0.958 g/cm 3 2356 47.1 3.3 1 Calculated from values obtained using a laboratory scale accurate to 0.0 1 g. 97

PAGE 98

CHAPTER 3 VIABILITY OF SEEDS AND SHOOTS FOLLOWING INUNDATION Introduction This chapter addresses objectives 2 and 3. Object ive 2 is partially addressed by assessing viability of stream dispersed seeds and shoots following net capture. Objective 3 is fully addressed by evaluating germination rates of hand harvested seed s and survivorship of shoot s following submergence in the stream. Hypotheses H 5 through H 7 are fully addressed, and hypothesis H 2 is partially addressed, in this ch apter. Materials and Methods Seeds and shoots recovered from nets and handharvested seeds and shoots were tested for viability by noting germination rates using a combination of methods. Viability of Net captured Seeds and Shoots This section tested hypothesis H 5 by comparing germination rates of stream dispersed R. simplex seeds with control seeds . Hypothesis H 2 is partially addressed. One method of testing viability consisted of placing the net captured R. simplex seeds in a g rowing medium consi sting of a homogenized mixture of 80% sterilized potting soil, 10% verm iculite, and 10% sand by volume (Robbins and Evans [no date] ; A. Moseley, Environmental Horticulture Department, Univers ity of Florida, pers . comm. , 09/05/11), moistened as needed using tap water purified with a Brita filter, and subjected to a natural light and temperature regime in a n unheated glasshouse. Additional R. simplex seeds recovered from the nets were tested for viability using a laboratory based germination chamber (Model CMP6010; Conviron, Pembina, ND) equipped with fluorescent lamps and temperature control following methods in 98

PAGE 99

Wilson et al. (2004) and Hupp (2007). Seeds were placed inside 10 cm diameter petri dishes (typically 20 seeds per dish) above a thin layer of st erilized sand moistened with 5 mL of deionized water. Petri dishes were sealed with Parafilm to prevent desiccation, were marked to identify the sample, and incubated in the germination chamber using a 12hour photoperiod and a temperature regime of 30 and 20C during light and dark periods, respectively. Only a small number of net captured R. simplex and T. fluminensis shoots were available for testing. Upon removal from the nets, shoots were dried for two to four hours and tested for viability. Some shoots were planted in flats filled with a homogenized mixture of equal volumes sterilized peat moss and coarse perlite and moistened with deionized water following methods in Ingram and Yeager (2010). Others were placed in deionized water. Artificial lighting was supplied for 12 hours daily using four 1.2m long 32 watt fluorescent bulbs (Philips T8 Plant and A quarium bulbs) suspended 1 m above the planted flats. Some natural light was also available through nearby windows. Deionized water was added as needed to keep the shoots from drying out . Viability of Hand harvested Seeds and Shoots Following Submergence Th is section tested hypotheses H 6 and H 7 by testing hand harvested R. simplex seeds and T. fluminensis shoots for viability following submer gence. Seed capsules were handharvested in September 2012 from a small population along Hogtown Creek in Gainesville and allowed to naturally dehisce at room temperature (22 – 25C) for two weeks before the seeds were stored for use in 2013. Seeds with o bvious insect or pathogen damage were removed and discarded. A high percentage 99

PAGE 100

of seed capsules were infested with dipteran larvae presumed to be Melanagromyza ruelliae , which commonly uses R. simplex as a host plant in Alachua County (Huey et al. 2007), and these damaged seeds were discarded. Beginning 5 May 2013, groups of 4 0 seeds contained within 0.8mm mesh bags were submerged in Tumblin Creek and secured within the water column just above the sediment surface using PVC, wire, and re bar stakes . Groups of 120 seeds (3 replicates of n = 40 seeds each) were submerged for 15, 30 , 90 , and 180day periods in the water column (Figure 3 1[ a ]) . Additional groups of 120 seeds (3 replicates of n = 40 seeds each) were contained in mesh bags submerged underwater and buried under 10– 20 cm of sediment in the stream bed for 15, 30 , 90 , and 180day periods (Figure 3 1[ b ]) . Each group of seeds was removed from the stream and dried for one week in ambient air prior to testing for viability using a germination chamber. Seeds were placed in plastic 10cm diameter petri dishes ( typically 20 seeds per dish) above a thin layer of st erilized sand moistened with 5 mL of deionized water. Seeds were subjected to ideal conditions following methods in Wilson et al. (2004) and Hupp (2007) and summarized in the above paragraph. Seed g ermination was monitored at least once per week for a period of 30 days. A seed was considered germinated when the radicle protruded at least 2 mm from the seed coat. Percent germination rates were compared betw een test groups and also with control groups of 120 handharvested seeds that were not subjected to submergence or burial . Shoots of T. fluminensis were handharvested from the forested wetland on 9 March 2014 and cut with scissors to the average length of those found in the nets (ca. 18 cm). A group of 120 shoots was placed within each of three cube shaped cages constructed of nyloncoated 0.6cm mesh hardware cloth attached to a PVC frame with 100

PAGE 101

zip ties. Eac h cage measured 0.6m per side and was secured to the stream bed using rebar stakes . Groups of 40 shoots were removed per cage following 5 , 1 0 , and 15day periods, or until remaining shoots showed advanced stages of decay. The cages were open to the at mosphere to allow shoots to achieve natural buoyancy in the wat er column as would be expected during a disper sion scenario. The hardware cloth allowed free flow of water through the devices. Upon removal, shoots were dried for two to four hours and tested for viability. Shoots were planted in flat s filled with a homogenized mixture of equal volumes sterilized peat moss and coars e perlite and moistened with deionized water following methods in Ingram and Yeager (2010). A commercially available root promo ting hormone (Bontone Rooting Powder, Bonide Products, Inc., Oriskany, NY) was applied at the time of planting to enhance rooting. Artificial lighting was supplied for 12 hours daily using four 1.2m long 32 watt fluorescent bulbs (Philips T8 Plant and A quarium bulbs) suspended 1 m above the planted flats. Some natural light was also available through nearby windows. De ionized water was added as needed to keep the planting media moist. Percent germination rates were compared between test groups and al so with a control group of 120 handharvested shoots that were not subjected to stream water. The control group was planted into three flats of 40 shoots each and each flat was considered a pseudoreplicate. Data Analysis G ermination rates and shoot survival rates were quantified and expressed as a percent of the total number of seeds or shoots tested in each replicate sample treatment. T he standard deviation was calculated for each group of treatment 101

PAGE 102

replicates . Statistic al significance o f the results of each treatment was compared to that of the control using a twotailed S tudent’s t test. Comparisons were made of estimated variances from within each treatment and between treatments using a oneway ANOVA prior to the t t ests. Results of the t tests and ANOVAs having a p value of less than 0.05 were considered to be significantly different fr om the control . A p value less than 0.001 was considered statistically highly significant. The p value was not corrected or adjust ed because the results of multiple individual tests are important for the purposes of this study and a general null hypothesis (that all null hypotheses are true simultaneously) is not of interest here; therefore, it was best to use the exact p value ( Perneger 1998; Armstrong 2014). Microsoft Excel was used for all calculations. No statistical analysis could be conducted on net captured shoots due to the low numbers intercepted. 102

PAGE 103

Figure 3 1 . Examples of materials used in submergence trials of Ruellia simplex seeds in the water column ( [ a ] and inset at upper right ) and below the sediment ([b] and inset at left) treatments . Photos by author. a b 103

PAGE 104

Results and Discussion Viability of Net captured Seeds Out of a total of 145 R. simplex seeds intercepted in Tumblin Creek (including 72 intercepted outside the May – December 2012 study) , 6.9% 31.5% germinated within 30 days inside a germination chamber . This overall percent germination for net capture d seeds was not significantly different ( ANOVA p = 0.20, t test p = 0. 2 0) from that of the control (32.5% 24.5% ). The net captured seeds came from both baseflow and highflow conditions. The highest germination rate among those tested in the germinati on chamber was 66.7% of n = 3 seeds from a bas e flow event on 10 October 2012. The secondhighest germination rate was 26.7% of n = 30 seeds from a highflow event on 29 September 2012. Tables 3 1 and 3 2 summarize results of a oneway ANOVA and t test, respectively, on percent germination of intercepted R. simplex seeds in a germ ination chamber compared to that of the control . Out of a total of 32 R. simplex seeds intercepted in Tumblin Creek, 21.8% 40.4% germinated within 70 days inside a glass house. G ermination occurred among 70.0% of n = 10 seeds intercepted on 18 September 2011 . The overall percent ger mination for net captured seeds was not significantly different ( ANOVA p = 0.41, t test p = 0. 41 ) from that of the control (53.0 % 12.7 %). However, because the control in this experiment consisted of only two pseudoreplicates (2 flats of n = 50 seeds each) instead of three as used elsewhere in this study , the statistical significance remains unclear . Tables 3 3 and 3 4 summarize results of a one way ANOVA and t test, respectively, on percent germination of intercepted R. simplex seeds in a glasshouse compared to that of the control . Tables D 1 and D 2 in Appendix D summarize seed 104

PAGE 105

germination results using a germination chamber or soilfilled flats, respectively, for other invasive and nonindigenous plants intercepted in Tumblin Creek. Based on the results of the R. simplex net captured seed germination trials, the H 5 null hypothesis was not rejected. The overall percent germinat ion for net captured seeds did not significantly differ from that of the control. This was true for seeds tested in the germination chamber as well as for those tested in soil inside a glasshouse. Thus, it appears that R. simplex seeds transported downst ream remain viable and are capable of germinating under suitable conditions to establish new populations in downstream portions of watersheds. Further, Hupp (2007) noted that two R. simplex seedlings survived partial submergence for two months in a shallow creek bed in Gainesville until they became completely submerged. Viability of Net captured Shoots Out of 18 R. simplex shoots intercepted in Tumblin Creek, 22.2 % survi ved over a period of at least 21 days inside a glasshouse. Out of 5 T. fluminensis sho ots intercepted in Tumblin Creek, 40.0 % survived at least 16 days inside a glasshouse. Table 3 5 summarize s shoot percent survivorship and growth metrics in a glasshouse by capture event . Table D 3 in Appendix D summarizes shoot survivorship and growth metrics for other invasive and nonindigenous plants in a glasshouse by capture event . The low survivorship among the 18 net captured R. simplex shoots may be due to possible anthropogenic effects. All 14 of the nonsurviving shoots showed diagonal cuts to one or both ends of the shoots, and some also showed additional damage along other portions of the shoots. These diagonal cuts and other damage may have been made by mowing equipment. Considering that both banks of Tumblin Creek are mowed 105

PAGE 106

between SW 13th Street (US 441) and the floodplain forest as part of regular maintenance during the main growing season, some of the shoots found in the nets may have resulted from these mowing events. The damage may have impaired the pote ntial for the shoots to survive and grow during testing in a glasshouse. Although mowing is not a natural event, it is a likely occurrence along waterways that bisect roads or are otherwise maintained as part of flood prevention practices. For example, R iis (2008) collected about twice as many shoots following weed cutting events at his study site in River Aarhus, Denmark, versus weeks when no cutting occurred. That author considered weed cutting to be a “major impacton the ecosystem”. Additional research on the relationship between mowing and shoot dispersal and viability would prove useful in shedding light on this potentially anthropogenically mediated dispersal mechanism. Viability of Hand harvested Seeds Following Submergence Submergence of R. simp lex seeds in the water column for 15 days resulted in 35.7% 14.2% germination in a growth chamber and was not significantly different ( p = 0. 28) from that of the control (32.5% 24.5%) . Submergence and burial under sediment for 15 days resulted in 8.7% 6.2% germination that was significantly different ( p = 0.0 32 ) from that of the control. Submergence in the water col umn for 30, 90, and 180 days resulted in germination rates of 8.8% 12.2%, 0.7% 0.6%, and 0.0% 0.0% , respectively, and all three tr eatments were significantly less ( p ) than that of respective controls (22.5 – 42.5% 3.8 – 7 .5%) . S ubmergence under sedime nt for 30, 90, and 180 days resulted in germination rates of 3.4% 2.9% , 0.0 % 0. 0 %, and 0.0% 0.0% , respectively, and all three treatments were significantly less ( p ) 106

PAGE 107

than that of respective controls (22.5– 42.5% 3.8 – 7.5%). No germination occurred in seeds following 180 days in the water column or 90 to 180 days buried under sediment . Table 3 6 summarizes results of oneway ANOVAs on R. simplex seed germination trials following submergence. Table 3 7 summarize s R. simplex percent germination, standard deviation, and statistical significance following treatments. The H 6 null hypothesis is rej ected based on the results of R. simplex seed submergence and burial trials. Results of the 15to 180 day submergence and burial trials support the H 6 alternative hypothesis due to significant differences between percent germination of these seeds and t hat of the control, with the exception of the 15day submergence testing which was not significantly different from that of the control. The relatively low germination rates of R. simplex seeds used as control populations in this study may have been due to storage conditions prior to testing. All R. simplex seeds were stored in sealed plastic bags under ambient air temperature prior to testing. Unpublished data by E. Barnett discussed in Hupp (2007) suggests that storage in sealed containers may greatly reduce viability of seeds. Ruellia simplex seed viability declined rapidly (from approximately 95% to 0% – 8% viability) for seeds stored under ambient conditions inside sealed containers for approximately four months. Higher viability rates (95%) were fo und when seeds were stored in refrigerated conditions and covered but not completely sealed (E. Barnett unpublished data as discussed in Hupp 2007). The low viability of the control seeds in the present study did not bias the outcome of the germination tests since all R. simplex seeds used in 107

PAGE 108

germination testing (control, submergence, and burial/submergence treatments) were stored using the same techniques and materials prior to testing. Viability of Hand harvested Shoots Following Submergence Submergence of T. fluminensis shoots in the water column for 5 days resulted in 74.2% 22.4% survival in a glasshouse and was not significantly different ( p = 0.12) from survival in the control (100%) . Submergence for 10 days resulted in 29.9% 32.8% su rvival and this was significantly different ( p = 0.0 21) from survival in the control. No shoots survived 15 days of submergence, which was significantly different ( p < 7.4 x 1018) from survival in the control . Table s 3 8 and 3 9 summarize results of a oneway ANOVA and survival and statistical significance, respectively, for T. fluminensis treatment s following 30 days in a n unheated glasshouse. T radescantia fluminensis shoots that survived 5 to 10 days of submergence showed significantly less mean shoot lengths (117.5– 195.5 14.1– 59.6 mm) and numbers of roots (5.5 – 14.8 1.1– 11.1 roots) per plant following 30 days in a glasshouse compared to these metrics in the control . However, the mean number of leaves in shoots following 5 days of submergence (6.5 2.8 leaves) was not significantly different ( p = 0.063) from that of the control (10.0 0.3 leaves). With this exception, all other growth metrics in shoots submerged 5 to 10 days showed a significant difference ( p 0.035 ) from that of the control. Table 3 10 summarizes results of oneway ANOVAs on T. fluminensis shoot lengths, numbers of leaves, and numbers of roots following submergence and growth trials. Table 3 11 compares post submergence T. fluminensis mean shoot lengt hs, mean numbers of leaves, mean 108

PAGE 109

numbers of roots, and significant differences from that of the control following 30 days in a glasshouse. The H 7 null hypothesis is rejected based on the results of the T. fluminensis shoot submergence trials. The results of the 10day submergenc e trials showed significant differences in survivorship from that of the control, and the remaining surviving shoots were significantly shorter and had a significantly fewer mean number of leaves and roots than did the control. Further, none of the shoots submerged for 15 days survived the trial. Thus, the H 7 alternative hypothesis is supported. Although survivorship of T. fluminensis shoots was tested following submergence, survivorship testing of R. simplex shoots was beyond the scope of this study. H upp (2007) observed a 25cm long R. simplex shoot that survived submergence in a creek bed in Gainesville for at least two months. This observation suggests that the species is capable of surviving relatively long periods under water. This trait may help R. simplex colonize downstream areas by stream mediated shoot dispersal. Although considerable strides have been made in establishing cultivars of R. simplex lacking the ability to produce viable seeds (e.g., Wilson and Mecca 2003; Freyre et al. 2012; Fr eyre and Wilson 2014), to the author’s knowledge there has been little or no research on vegetative reproduction as a potential mechanism for dispersal. Considering that all 11 current cultivars and the naturalized type are capable of vegetative reproduct ion, research should be conducted to understand the importance of vegetative dispersal mechanisms in R. simplex . 109

PAGE 110

Table 3 1 . Results of a one way ANOVA for germination trials using Ruellia simplex seeds intercepted in Tumblin Creek and tested for viability inside a germination chamber. Source of variation Degrees of freedom Sum of squares f p f critical Significant ly different population means? Between groups 1 1810.71 2.16 0.20 6.61 No Within groups 5 4181.94 Total 6 5992.66 Table 3 2 . Results of germination trials using Ruellia simplex seeds intercepted in Tumblin Creek. Each trial represents 30 days in a germination chamber with a 12hour photoperiod at 30C (daytime) and 20C (nighttime) . Date of capture or control Flow rate (m 3 /second) Stream condition A n Percent germination St. dev. t test p Significantly different from control? 09Oct 11 Not recorded Not recorded 72B 0.0 – – – 12Sep 12 0.0073 Base flow 40 0.0 – – – 29Sep 12 0.020 High flow 30 26.7 – – – 10Oct 12 0.012 Base flow 3 66.7 – – – Combined – – 145B 6.9 31.5 0.20 No Control n/a n/a 120 32.5 24.5 n/a n/a A B aseflow is defined as a flow rate of <0.02 m3/second and highflow is defined as m3/second. B Results include seeds that were intercepted during a preliminary deployment of the nets on 9 October 2011 and outside of the May – December 2012 interception study. For this reason, the total number of seeds in this table does not match the total number presented in the results of the interception study. 110

PAGE 111

Table 3 3 . R esults of a one way ANOVA for germination trials using Ruellia simplex seeds intercepted in Tumblin Creek and tested for viability inside an unheated glasshouse. Source of variation * Degrees of freedom Sum of squares f p f critical Significant ly different population means? Between groups 1 1056.13 0.92 0. 41 10.13 No Within groups 3 3428.67 Total 4 4484.80 * Control consisted of only two pseudoreplicates (2 flats of 50 seeds each). Table 3 4 . R esults of germination trials using Ruellia simplex seeds intercepted in Tumblin Creek and tested inside an unheated glasshouse. Each trial was conducted in a homogenized soil mixture under artificial grow lights (12 hour photoperiod) and additional natural light . Date of capture or control Flow rate (m 3 /second) Stream condition 1 n Duration of germination trial (days) Percent germination St. dev. t test p Sig. different from control? 18Sep 11 Not recorded Not recorded 10 70 70.0 – – – 16May 12 0.91 High flow 2 192 0.0 – – – 03Jun12 0.014 Base flow 20 181 0.0 – – – Combined – – 32 70– 192 21.8 40.4 0.41 No Control2 n/a n/a 100 42 5 3 .0 12.7 n/a n/a 1 B aseflow is defined as a flow rate of <0.02 m3/second and highflow is defined as m3/second. 2 Control consisted of only two pseudoreplicates (2 flats of 50 seeds each). 111

PAGE 112

Table 3 5 . Results of survival and growth trials for Ruellia simplex and Tradescantia fluminensis shoots intercepted in Tumblin Creek. Following capture, shoots were placed in either water or soil . Each trial lasted at least 17 days (or until mortality) and was conducted under artificial grow lights (12hour photoperiod ) with additional natural light in an unheated glasshouse. Metrics at time of capture Metrics following survival trial Species, date of capture Flow rate (m 3 /sec. ) Stream condition A n Length of S hoot (mm) Number of leaves Number of roots Growth medium , duration Change in shoot length ( % ) Change in number of leaves (%) Change in number of roots (%) Percent survival Ruellia simplex 27 May 12 0.021 High flow 2 305 (mean) 24.0 (mean) 126.0 (mean) Soil, 199 days + 76 (mean) + 340 (mean) + 280 (mean) 100 14 Jun 12 0.025 High flow 1 4 B 158 (mean) 8 .0 (mean) 0. 0 (mean) Water, 80 days – – – 0.0 17 Aug 12 0.31 2 High flow 1 245 9 2 Soil, 117 days +6 +200 +1900 100 29 Sep 12 0.020 High flow 1 152 9 1 Water, 21 days Not measured +44 100 100 Tradescantia fluminensis 27 May 12 0.020 High flow 1 320 6 3 Soil, 199 days 27 +50 33 100 14 Jun 12 0.025 High flow 4 97 (mean) 4.8 (mean) 0.5 (mean) Water , 16 days + 108 (mean) + 53 (mean) + 150 (mean) 25 .0 A Base flow is defined as a flow rate of <0.02 m3/second and high3/second. B All 1 4 R. simplex shoots had diagonal cuts along one or both ends of the shoots that may have been caused by mowing eq uipment prior to capture in the nets . 112

PAGE 113

Table 3 6 . R esults of one way ANOVA s for Ruellia simplex seed germination trials following submergence in the water c olumn for 15, 30 , 90 , and 180 day durations . Some treatments included burial under sediment for the duration of submergence. Source of variation Degrees of freedom Sum of squares f p f critical Significant ly different population means? 15day treatments Between groups 2 100.33 7.70 0.0050 3.68 Yes Within groups 15 97.67 Total 17 198.00 30 day treatments Between groups 2 147.00 24.23 2 .0 x 105 3.68 Yes Within groups 15 45.50 Total 17 192.50 90day treatments * Between groups 4 310. 76 54.28 1.9 x 109 2.96 Yes Within groups 17 24.33 Total 21 335.09 180day treatments * Between groups 5 97.20 62.21 6.0 x 1013 2.62 Yes Within groups 24 7.50 Total 29 104.70 * Replicate 3 of the 90day and replicate 3 of the 180day water column treatment s, and replicate 2 of the 90day under sediment treatment, were lost in the stream. Because of this, these treatments were tested with two replicates instead of three. All other treatments were tested using three replications. 113

PAGE 114

Table 3 7 . Results of Ruellia simplex seed germination trials following submergence in the water column for 15, 30 , 90 , and 180day dur ations. Some treatments included burial under sediment for the duration of submergence. Seeds were tested using a germination chamber with a 12hour photoperiod at 30C (daytime) and 20C (nighttime) . Each test lasted 30 days . Treatment type and duration n Percent germination Standard deviation t test p Significantly different from control? 15day treatments Water column 115 35.7 14.2 0.28 No Under sediment 115 8.7 6.2 0.032 Yes Control 120 32.5 24.5 n/a n/a 30day treatments Water column 113 8.8 12.2 0.021 Yes Under sediment 117 3.4 2.9 0.00030 Yes Control 120 35.8 3.8 n/a n/a 90day treatments W ater column* 153 0.7 0.6 0.0049 Yes Under sediment* 153 0. 0 0. 0 0.0047 Yes Control 120 42.5 7.5 n/a n/a 180day treatments Water column* 268 0.0 0.0 0.0061 Yes Under sediment 193 0.0 0.0 0.0061 Yes Control 120 22.5 4.3 n/a n/a * Replicate 3 of the 90day and replicate 3 of the 180day water column treatment s, and replicate 2 of the 90day under sediment treatment, were lost in the stream. Because of this, these treatments were tested with two replicates instead of three. All other treatments were tested using three replications. Table 3 8 . Results of a one way ANOVA for Tradescantia fluminensis shoot survival trials following submergence in water for 5 to 15 day durations . Shoot survival was determined following 30 days planted in soil under artificial grow lights (12 hour photoperiod) and additional natural light in an unheated glasshouse. Source of variation Degrees of freedom Sum of squares f p f critical Significant ly different population means? Between groups 3 17968.64 15.16 0. 0012 4.07 Yes Within groups 8 3161.49 Total 11 21130.14 114

PAGE 115

Table 3 9 . Results of Tradescantia fluminensis shoot survival trials following submergence in water for 5 to 15 day durations . Shoot survival was determined following 30 days planted in soil under artificial grow lights (12 hour photoper iod) and additional natural light in an unheated glasshouse. Submergence treatment duration n Percent survival Standard deviation t test p Significantly different from control? 5 days 120 74.2 22.4 0.12 No 10 days 117 29.9 32.8 0.021 Yes 15 days 123 0.0 0.0 7.4 x 10 18 Yes Control 120 100 0.0 n/a n/a Table 3 10 . Results of one way ANOVA s for shoot length, number of leaves, and number of roots in Tradescantia fluminensis shoots following submergence in water for 5 to 10 day durations . Growth metrics were measured following 30 days planted in soil under artificial grow lights (12hour photoperiod) and additional natural light in an unheated glasshouse. Source of variation Degrees of freedom Sum of squares f p f critical Significant ly different population means? Shoot length * Between groups 2 1013705 89.15 1.04 x 1029 3.03 Yes Within groups 241 1370116 Total 243 2383821 Number of live leaves * Between groups 2 1790 184.03 3.72 x 1048 3.04 Yes Within groups 226 1099 Total 228 2 889 Number of roots * Between groups 2 7 0204 158.45 1.35 x 1044 3.04 Yes Within groups 240 53167 Total 242 123370 * ANOVAs do not include shoots from the 15day submergence trials because no shoots survived 15 day s of submergence (all leaves had senesced and shoots were completely flaccid and showing advanced stages of decay). 115

PAGE 116

Table 3 11 . Growth of Tradescantia fluminensis shoots following submergence in water for 5 to 15 day durations . Growth metrics were measured following 30 days planted in soil under artificial grow lights (12hour photoperiod) and additional natural light in an unheated glasshouse. Submergence treatment duration n alive Mean shoot length (mm) St. d ev. t test p Sig. different from control? Mean n umber of live leaves per plant St. dev. t test p Sig. different from control? Mean number of roots per plant St. dev. t test p Sig. different from control? 5 days 89 195.5 59.6 0.035 Yes 6.5 2.8 0.063 Yes 14.8 11.1 0.0083 Yes 10 daysA 35 117.5 14.1 0.0047 Yes 4.7 0.1 0.011 Yes 5.5 1.1 1.3 x 10-4 Yes 15 daysB 0 – – – – – – – – – – – – Control 120 292.3 5.8 n/a n/a 10.0 0.3 n/a n/a 45.6 4.0 n/a n/a A By the end of the 10day submergence treatment, leaves of replicate 3 shoots were senescing and portions of stems were visibly flaccid due to decay but were planted in the flats to verify mortality. All shoots of replicate 3 had died by the 30day endpoint. B No shoots survived the 15 day submergence trials (all leaves had senesced and shoots were completely flaccid and showing advanced stages of decay). 116

PAGE 117

CHAPTER 4 CONCLUSIONS This study focused on determining the capability of two invasive plant species to utilize stream mediated dispersal mechanisms for c olonizing downstream habitats. The techniques used here allowed the interception and retention of R. simplex seeds and shoots and T. fluminensis shoots in an urban stream. Stream dispersal of R. simplex propagules occurs mainly from seeds rather than shoots. D iffering stream flow regimes do not appear to significantly affect numbers of stream dispersed genetic material in R. simplex or T. fluminensis . Germination rates of net captured seeds of R. simplex are similar to that of the control (handpicked) seeds. However, handpicked R. simplex seeds subjected to submergence in the stream for 30 day or longer had significantly lower percent germination than did control seeds. Similarly, T. fluminensis shoots subjected to 10 days in stream water had significantly lower survivorship than did control shoots, and had significantly fewer mean numbers o f leaves and roots than did con trol shoots . Further, none of the shoots subjected to stream water for 15 days survived the trial. Seeds of R. simplex are negatively buoyant in water while shoots of T. fluminensis are positively buoyant when initially placed in water but become negativ ely buoyant within 15 days. Ruellia simplex and T. fluminensis are capable of utilizing stream mediated dispersal mechanism s for recruitment and indeed may have used such mechanisms to colonize the Tumblin Creek floodplain. Although R. simplex is somewhat limited by the mobility of its non buoyant seeds , the seeds are nonetheless durable and capable of germinating following up to 90 days submergence (or up to 30 days buried under sediments), and its dispersal capability is probably further enhanced by stream transport 117

PAGE 118

of its shoot s . Tradescantia fluminensis does not produce seeds in Florida but the somewhat positively buoyant shoots are c apable of being transported hundreds of meters downstream and can remain viable following submergence for at least 1 0 days. The resilience and buoyancy of the shoots, coupled with the ability of fragments as small as a single node to produce a new plant, makes T. fluminensis well adapted to stream mediated dispersal. The ability of both species to disperse via flowing water is further enhanced by their propensity to grow along stream banks coupled with their capacity to successfully colonize highly modified riparian wetlands such as those along urban streams . N atural areas land managers should consider these tolerance traits for aquatic habitats and stream dispersal mechanisms when planning control efforts i n riparian wetlands and to plan such efforts at the watershed scale . By b eginning at upstream portions of a watershed and working downstream , genetic mate rial capable of recolonizing downstream habitats will be minimized or eliminated. 118

PAGE 119

APPENDIX A NONINDIGENOUS PLANTS AND ANIMALS OF THE FLOODPLAIN Table A 1 . L ist of nonindigenous plant species observed by the author within the Tumblin Creek floodplain and associated natural areas, 2009 – 201 4 . Scientific n ame Common n ame FLEPPC c ategory * , l ocal a bundance Albizia julibrissin silktree Cat. I, uncommon Aleurites fordii tungoil tree Cat. II, rare Alternanthera philoxeroides alligatorweed Cat. II, common in stream above floodplain Ardisia crenata coral ardisia Cat. I, abundant in uplands Asclepias curassavica scarlet milkweed Isolated population of a few dozen along stream bank Asparagus aethiopicus Sprenger’s asparagus fern Cat. I, isolated population Asparagus setaceus common asparagus fern Isolated population Broussonetia papyrifera paper mulberry Cat. II, uncommon Butia capitata pindo palm Uncommon in floodplain forest Cannabis sp. marijuana Rare, few observed (probably in cultivation) Carya illinoinensis pecan Rare, n = 1 tree on west edge of berm Cinnamomum camphora camphortree Cat. I, rare in floodplain Clematis terniflora Japanese clematis Cat. II, abundant along stream Colocasia esculenta wild taro Cat. I, common in floodplain Commelina cf. diffusa common dayflower Abundant in floodplain forest Cyperus involucratus umbrella plant Cat. II, uncommon in stream Dioscorea bulbifera air potato Cat. I, common Eichhornia crassipes water hyacinth Cat. I, common in lake Elaeagnus sp. thorny olive/autumn olive Cat. II, rare, few observed Hedera helix English ivy Abundant in upper floodplain Hypochaeris sp. catsear Occasional along berm Ipomoea cairica mile a minute vine Uncommon along stream banks and berm Ipomoea quamoclit cypress vine Rare, n = 1 plant observed in bloom along stream Koelreuteria elegans golden raintree Cat. II, uncommon in floodplain Lantana camara lantana Cat. I, occasional along stream banks and berm Ligustrum lucidum glossy privet Cat. I, common throughout area Ligustrum sinense Chinese privet Cat. I, common in floodplain Lonicera japonica Japanese honeysuckle Cat. I, isolated populations in floodplain Ludwigia peruviana Peruvian primrosewillow Cat I, common along stream Lygodium japonicum Japanese climbing fern Cat. I, uncommon in floodplain Macfadyena unguis cati catclaw vine Cat. I, common to abundant throughout area Melia azedarach chinaberrytree Cat. II, rare in floodplain Merremia dissecta noyau vine Common along berm Morus alba white mulberry Common along stream Nephrolepis cordifolia tuberous sword fern Cat. I, isolated population Oxalis debilis pink woodsorrel Abundant in floodplain forest Panicum repens torpedo grass Cat I, abundant, dense stands along stream Paspalum notatum bahiagrass Common along berm Paspalum urvillei vasey grass Undetermined, seeds in nets during May and June Phyllanthus urinaria chamber bitter Common along berm Podocarpus macrophyllus yew plum pine Rare, few observed 119

PAGE 120

Table A 1. Continued Scientific n ame Common n ame FLEPPC c ategory * , l ocal a bundance Pu eraria m ontana var. lobata kudzu Cat I, rare, single vine obs. along stream bank Richardia brasiliensis tropical Mexican clover Occasional along berm Ruellia simplex Mexican petunia Cat. I, abundant in floodplain Sapium sebiferum Chinese tallowtree Cat. I, rare Syngonium podophyllum arrowhead vine Cat. I, isolated populations Tradescantia fluminensis small flowered spiderwort Cat. I, abundant in floodplain Washingtonia robusta Washington fan palm Cat. II, rare, few observed Xanthosoma sagittifolium elephant ear Cat. II, common in floodplain *Florida Exotic Pest Plant Council’s Category I (species documented to cause ecological damage in Florida plant communities) and Category II (species increasing in abundance or frequency but not known to alter Florida plant communities to the extent sho wn in Category I species) (FLEPPC 20 13) . Table A 2 . List of nonindigenous invertebrate and vertebrate species observed by the author within the Tumblin Creek floodplain and associated natural areas, 2009– 201 4 . Scientific name Common name Local abundance INVERTEBRATES Allopeas gracile graceful awlsnail Common, n = 54 found in seed nets Bradybaena similaris Asian tramp snail Abundant along stream and in floodplain Corbicula fluminea Asian clam Common in stream Huttonella bicolor two tone gulella Undetermined, n = 10 found in seed nets Lamellaxis micra tiny awlsnail Undetermined, n = 22 found in seed nets Melanoides cf. tuberculatus red rim melania Common in stream, n Melanoides cf. turriculus fawn melania Uncommon, n Opeas hannense dwarf awlsnail Undetermined, n = 2 found in seed nets Opeas pyrgula sharp awlsnail Undetermined, n = 1 found in seed nets Solenopsis invicta red imported fire ant Common along berm near stream VERTEBRATES Dasypus novemcinctus nine banded armadillo Undetermined, n = 1 obs., also tracks Felis catus domestic cat Undetermined, n = 2 obs. a long stream bank Oreochromis aureus blue tilapia Common in lake, occasional in stream Trachemys scripta elegans red eared slider Rare, n = 1 skeletal remains found along lake Xiphophorus variatus variegated platy Common in stream 120

PAGE 121

APPENDIX B S TREAM METRICS AND PLANT COVER ALONG STREAM Table B 1 . Flow rates, water dept hs, stream widths , and substrate types measured during base flow conditions on 26 May , 2012, eve ry 25 m along the 800m long reach of Tumblin Creek between SW 13th Street (US 441) and Bivens Arm . Transects 1– 6A are in the floodplain forest and transects 7 – 17 are upstream of the forest. Transect number Distance from lake (m) Stream w idth (cm) Mean depth (mm) Flow rate (m 3 /sec.) Substrate 1 0 122 31 0.006 Sand w/ silt 1A 25 127 20 0.006 Sand 2 50 254 146 0.007 Sand 2A 75 116 23 0.002 Sand 3 100 117 26 0.006 Sand w/ gravel 3A 125 142 171 0.044 Sand 4 150 147 29 0.008 Sand 4A 175 135 20 0.004 Sand 5 200 62 28 0.001 Sand 5A 225 213 252 0.014 Sand 6 250 224 83 0.017 Sand 6A 275 211 115 0.003 Sand w/ debris 7 * 300 220 27 0.008 Sand w/ gravel 7A 325 419 252 0.010 Sand 8 350 163 73 0.006 Sand 8A 375 340 51 0.013 Sand 9 400 386 113 0.003 Sand 9A 425 287 20 0.014 Sand 10 450 280 53 0.009 Sand 10A 475 211 32 0.014 Sand 11 500 163 169 0.008 Sand 11A 525 221 32 0.011 Sand w/ gravel 12 550 391 24 0.017 Sand 12A 575 235 69 0.011 Sand 13 600 432 85 0.026 Sand 13A 625 295 197 0.017 Sand 14 650 325 219 0.015 Sand 14A 675 163 150 0.039 Sand 15 700 107 47 0.013 Concrete w/ sand 15A 725 108 56 0.009 Concrete 16 750 124 64 0.005 Concrete 16A 775 150 44 0.012 Concrete 17 800 762 24 0.032 Concrete Overall mean st. dev. – 232 137 83 71 0.012 0.010 – * Net deployments were conducted at Transect 7, just upstream of the floodplain forest . 121

PAGE 122

Table B 2 . Mean p ercent aerial coverage of herbaceous an d small woody plants within 3 m upslope of the mean high water line along both banks of Tumb lin Creek per group of transects surveyed on 29 April , and 5 May , 2012. Transects are spaced 25 m apart along the 800m long reach between SW 13th Street (US 441) and Bivens Arm . Transects 1 – 6A and 7– 17 are in the floodplain forest and upstream of the forest , respectively . Mean percent ( st. dev. ) aerial coverage per group of transects Species or taxon Transects 1 – 6A (floodplain forest, 0 – 275 m from lake) Transects 7 – 12 (upstream of floodplain forest, 300 – 550 m from lake) Transects 12A – 17 (downstream of SW 13th Street, 575 – 800 m from lake) Overall mean Acer negundo 1.42 2.69 1.45 2.50 0.25 1.09 1.08 2.32 Acer rubrum 0.63 0.99 0.68 0.97 0.05 0.22 0.47 0.87 Acmella oppositifolia 0.04 0.20 10.14 20.15 0.00 0.00 3.39 12.58 Alternanthera philoxeroides 0.00 0.00 4.09 16.69 8.00 21.35 3.79 15.55 Apios americana 0.13 0.60 0.00 0.00 0.00 0.00 0.05 0.37 Bacopa caroliniana 0.00 0.00 0.14 0.62 1.25 5.45 0.42 3.07 Bacopa monnieri 0.13 0.60 0.00 0.00 0.05 0.22 0.06 0.38 Bidens alba 3.96 7.23 5.73 6.04 12.15 16.81 7.03 11.34 Bidens bipinnata 0.00 0.00 3.05 7.85 2.00 4.10 1.62 5.23 Campsis radicans 0.29 0.89 0.00 0.00 0.15 0.65 0.15 0.66 Celtis laevigata 0.88 3.03 0.00 0.00 0.10 0.44 0.35 1.89 Centrosema sp. 0.13 0.60 0.00 0.00 0.00 0.00 0.05 0.37 Carex lurida 0.08 0.40 3.23 6.05 0.00 0.00 1.11 3.81 Cinnamomum camphora 0.13 0.33 0.09 0.29 0.00 0.00 0.08 0.26 Clematis terniflora 0.25 1.20 0.00 0.00 0.00 0.00 0.09 0.73 Colocasia esculenta 1.29 2.32 0.00 0.00 0.00 0.00 0.47 1.53 Commelina sp. 0.13 0.60 0.36 1.19 0.00 0.00 0.17 0.79 Commelina cf. diffusa 1.29 5.05 3.36 5.70 0.25 0.77 1.67 4.68 Cyperus haspan 0.00 0.00 0.86 1.79 0.45 1.24 0.42 1.29 Dichondra carolinensis 0.00 0.00 0.55 1.72 0.00 0.00 0.18 1.03 Echinochloa crus galli 0.00 0.00 1.77 3.40 1.30 3.18 0.98 2.74 Eupatorium sp. 0.00 0.00 0.27 0.75 0.15 0.65 0.14 0.57 Eupatorium capillifolium 0.04 0.20 0.00 0.00 0.00 0.00 0.02 0.12 Fabaceae 0.00 0.00 0.00 0.00 0.75 2.26 0.23 1.29 Fraxinus caroliniana 0.08 0.40 0.00 0.00 0.00 0.00 0.03 0.24 Geranium carolinianum 0.17 0.80 0.09 0.42 0.65 1.49 0.29 1.01 Hydrocotyle sp. 0.00 0.00 0.09 0.42 0.25 0.77 0.11 0.50 Hypochaeris sp. 0.00 0.00 0.09 0.42 0.00 0.00 0.03 0.24 Hyptis alata 0.08 0.40 0.00 0.00 0.00 0.00 0.03 0.24 122

PAGE 123

Table B 2 . Continued Mean percent ( st. dev. ) aerial coverage per group of transects Overall mean Species or taxon Transects 1 – 6A (floodplain forest, 0 – 275 m from lake) Transects 7 – 12 (upstream of floodplain forest, 300 – 550 m from lake) Transects 12A – 17 (downstream of SW 13th Street, 575 – 800 m from lake) Ipomoea sp. 0.00 0.00 0.00 0.00 0.10 0.44 0.03 0.24 Lantana camara 0.00 0.00 0.00 0.00 0.25 1.09 0.08 0.61 Lepidium virginicum 0.00 0.00 0.55 1.75 0.85 2.39 0.44 1.70 Ludwigia peruviana 2.92 13.99 2.23 5.18 0.25 1.09 1.88 9.04 Ludwigia repens 0.00 0.00 0.00 0.00 0.35 1.31 0.11 0.74 Macfadyena unguis cati 0.00 0.00 0.18 0.49 0.10 0.44 0.09 0.38 Osmunda cinnamomea 0.00 0.00 0.14 0.62 0.00 0.00 0.05 0.37 Oxalis debilis 0.13 0.44 0.32 0.76 0.10 0.44 0.18 0.57 Panicum sp. 0.63 3.00 7.23 16.56 9.50 24.83 5.52 17.20 Panicum repens 0.00 0.00 0.00 0.00 0.50 2.18 0.15 1.22 Parthenocissus quinquefolia 0.04 0.20 0.09 0.29 0.00 0.00 0.05 0.21 Paspalum sp. 0.00 0.00 1.59 4.09 1.05 3.29 0.85 3.05 Paspalum notatum 0.00 0.00 0.00 0.00 1.50 6.54 0.45 3.66 Paspalum urvillei 0.00 0.00 0.00 0.00 0.25 1.09 0.08 0.61 Phanopyrum gymnocarpon 3.75 17.98 0.00 0.00 0.00 0.00 1.36 10.99 Phyla nodiflora 0.00 0.00 3.59 12.51 0.50 1.77 1.35 7.46 Phyllanthus sp. 0.00 0.00 0.00 0.00 1.05 4.35 0.32 2.44 Phyllanthus urinaria 0.00 0.00 0.05 0.21 0.00 0.00 0.02 0.12 Pinus sp. 0.04 0.20 0.00 0.00 0.00 0.00 0.02 0.12 Polygonum sp. 0.00 0.00 0.55 2.50 0.00 0.00 0.18 1.47 Portulaca sp. 0.00 0.00 0.00 0.00 0.10 0.44 0.03 0.24 Richardia brasiliensis 0.00 0.00 0.00 0.00 0.25 1.09 0.08 0.61 Ruellia simplex 2.42 10.00 3.36 5.14 1.85 4.67 2.56 7.22 Sabal palmetto 0.42 1.53 0.00 0.00 0.00 0.00 0.15 0.93 Sambucus nigra canadensis 0.17 0.55 0.23 1.04 0.00 0.00 0.14 0.69 Sapium sebiferum 0.00 0.00 0.14 0.62 0.00 0.00 0.05 0.37 Saururus cernuus 0.04 0.20 0.00 0.00 0.00 0.00 0.02 0.12 Setaria cf. parviflora 0.00 0.00 0.00 0.00 0.50 2.18 0.15 1.22 Smilax sp. 0.04 0.20 0.00 0.00 0.00 0.00 0.02 0.12 Solanum sp. 0.00 0.00 0.00 0.00 0.25 1.09 0.08 0.61 Solidago sp. 0.00 0.00 2.41 10.40 0.00 0.00 0.80 6.11 Toxicodendron radicans 0.83 4.00 0.00 0.00 0.00 0.00 0.30 2.44 T radescantia fluminensis 19.25 27.13 3.36 12.50 0.00 0.00 8.12 19.81 123

PAGE 124

Table B 2 . Continued Mean percent ( st. dev. ) aerial coverage per group of transects Species or taxon Transects 1 – 6A (floodplain forest, 0 – 275 m from lake) Transects 7 – 12 (upstream of floodplain forest, 300 – 550 m from lake) Transects 12A – 17 (downstream of SW 13th Street, 575 – 800 m from lake) Overall mean Tradescantia ohiensis 0.00 0.00 0.00 0.00 0.50 2.18 0.15 1.22 Trifolium sp. 0.00 0.00 0.00 0.00 0.65 1.53 0.20 0.89 Tripsacum dactyloides 0.00 0.00 0.00 0.00 0.50 1.77 0.15 1.00 Vitis sp. 0.00 0.00 0.09 0.42 0.00 0.00 0.03 0.24 Xanthosoma sagittifolium 2.33 10.01 0.00 0.00 0.00 0.00 0.85 6.14 Undetermined species 0.00 0.00 0.00 0.00 0.65 2.01 0.20 1.14 Bare soil w/ trash 4.17 19.98 0.00 0.00 0.00 0.00 1.52 12.22 Dead vegetation 0.00 0.00 0.00 0.00 4.60 20.05 1.39 11.24 Concrete 0.00 0.00 0.00 0.00 13.50 32.60 4.09 18.99 124

PAGE 125

Table B 3 . Results of a qualitative survey in June 2011 of Ruellia simplex and Tradescantia fluminensis along the main channel of Tumblin Creek from the headwaters to SW 1 3th Street (US 441) . D istance Upstream from Lake (km) Location Occurrence Notes 2.8 South of 4th Ave. SW (creek origin) Creek first emerges as groundwater from a culvert R. simplex obs. 50 m downstream of culvert, along steeply sloping stream banks within 2.4 m of water; healthy but no flowers or capsules T. fluminensis not obs. 2.7 South of 5th Ave. SW Dense patches of R. simplex along sunlit banks, n = 4 plants flowering, others with green capsules Monoculture of T. fluminensis along concrete armoring at culvert pipes, mostly in shade 2.5 Culvert from 7th Ter. SW ( dead end) R. simplex in stands but no flowers or capsules obs. T. fluminensis also obs. 2.4 Tumblin Creek Park R. simplex dominates banks from water’s edge to 3.7 m upslope, sometimes on exposed sandbars T. fluminensis along west bank 2.2 SW Depot Ave. R. simplex along banks but no flowers or capsules obs. T. fluminensis not obs. 2.0 SW 9th St. R. simplex dominant along sunlit portions of banks but no flowers or capsules obs. T. fluminensis not obs. 1.3 SW 14th Ave., east of SW 13th St. R. simplex in patches along banks, no flowers or capsules obs. T. fluminensis not obs. 1.1 South of 16th Ave. SW Most of creek fully armored. R. simplex is dominant, found in patches along banks and along water’s edge, patches in full sun in late bloom and (or) developing seed capsules. T. fluminensis not obs. 0.9 North of SW 13th St. R. simplex is dom inant and is flowering even inside box culverts T. fluminensis not obs. 125

PAGE 126

APPENDIX C NUMBERS OF SEEDS AND SHOOTS INTERCEPTED IN STREAM Table C 1 . Seeds intercepted per month and per flow regime in Tumblin Creek. Nets were deploy ed 27 times during May – December 2012 including baseflow (<0.02 m3/second) and high3/second) events. Seeds per month Mean seeds per flow regime (individuals/10 0 m 3 ) Family, species May Jun Jul Aug Sep Oct Nov Dec Total Base flow High flow All flow ACANTHACEAE Ruellia simplex 2 22 0 19 85 3 0 0 131 0.778 0.173 0.442 ALTINGIACEAE Liquidambar styraciflua 1 1 5 0 0 0 0 2 9 0.007 0.146 0.084 AQUIFOLIACEAE Ilex sp. 2 4 4 0 0 0 0 0 10 none 0.205 0.114 ARECACEAE cf. Butia capitata 0 0 3 0 0 0 0 0 3 none 0.007 0.004 Sabal palmetto 27 42 29 40 6 4 1 10 159 0.188 3.51 2.03 ASTERACEAE Asteraceae 0 0 0 0 2 0 0 0 2 0.011 none 0.005 Bidens cf. alba 0 0 0 0 0 0 0 12 12 none 0.235 0.131 Bidens sp. 0 150 0 0 0 0 0 0 150 none 11.8 6.57 Coreopsis sp. 3 18 0 0 0 1 0 0 22 0.087 0.919 0.549 Helianthus cf. annuus 1 1 1 0 0 0 0 0 3 none 0.030 0.017 Hypochaeris sp. 3 0 0 0 0 0 0 0 3 0.295 0.090 0.181 BIGNONIACEAE cf. Campsis radicans 0 0 0 0 0 2 0 0 2 none 0.030 0.017 Macfadyena unguis cati 0 1 0 4 1 0 0 16 22 none 0.380 0.211 BRASSICACEAE Lepidium virginicum 46 28 3 0 0 0 0 0 77 0.251 4.70 2.72 CELTIDACEAE Celtis laevigata 5 31 2 1 0 1 0 3 43 0.015 1.03 0.577 COMMELINACEAE Commelina cf. diffusa 4 16 4 0 3 3 0 0 30 0.175 0.348 0.271 CUCURBITACEAE Cucumis cf. sativus 0 3 0 0 0 0 0 0 3 none 0.080 0.045 CUPRESSACEAE Juniperus cf. virginiana 0 1 0 0 0 0 0 0 1 0.007 none 0.003 126

PAGE 127

Table C 1. Continued Seeds per month Total Mean seeds per flow regime (individuals/100 m 3 ) Family, species May Jun Jul Aug Sep Oct Nov Dec Base flow High flow All flow CYPERACEAE 0 200 0 0 0 0 0 0 200 none 15.8 8.76 Carex sp. 0 257 1 0 0 0 0 0 258 0.031 20.0 11.1 Cyperus sp. 0 200 0 0 0 0 0 0 200 none 15.8 8.76 ERICAEAE Vaccinium sp. 30 0 0 0 0 0 0 0 30 none 2.70 1.50 EUPHORBIACEAE Sapium sebiferum 0 0 27 3 0 0 0 0 30 none 0.091 0.051 FABACEAE cf. Sesbania sp. 2 0 0 0 0 0 0 0 2 none 0.180 0.100 FAGACEAE Quercus sp. 66 183 345 49 30 24 4 82 783 1.10 11.2 6.74 JUGLANDACEAE Carya sp. 5 13 14 6 2 1 0 2 43 0.269 0.742 0.532 Carya illinoinensis 0 2 0 0 0 0 0 0 2 none 0.054 0.030 LAMIACEAE Callicarpa americana 0 0 0 2 0 0 0 50 52 none 1.27 0.705 LAURACEAE Cinnamomum camphora 66 137 120 43 18 15 4 73 476 0.609 9.80 5.71 LYTHRACEAE Lagerstroemia indica 18 25 2 34 51 7 3 23 163 0.400 2.05 1.32 MAGNOLIACEAE Magnolia grandiflora 0 3 0 0 0 0 0 0 3 none 0.133 0.074 MYRICACEAE Myrica cerifera 0 0 7 0 0 0 0 0 7 none 0.010 0.005 OLEACEAE Fraxinus cf. caroliniana 39 52 366 817 1065 91 17 54 2501 15.2 7.99 11.2 PHYTOLACCACEAE Phytolacca americana 10 0 0 0 0 0 0 1 11 none 0.463 0.257 PINACEAE cf. Pinus sp. 0 5 4 1 2 2 0 6 20 0.082 0.336 0.223 PLATANACEAE Platanus occidentalis 220 105 25 0 5 4 1 24 384 0.191 13.5 7.60 POACEAE Poaceae 0 6 0 8 6 2 2 0 24 0.085 0.724 0.440 cf. Echinochloa sp. 0 0 0 1 0 0 0 0 1 none 0.003 0.001 Echinochloa crus galli 2 0 0 0 0 0 0 0 2 none 0.180 0.100 127

PAGE 128

Table C 1. Continued Seeds per month Total Mean seeds per flow regime (individuals/100 m 3 ) Family, species May Jun Jul Aug Sep Oct Nov Dec Base flow High flow All flow Panicum sp. 0 143 6 1 0 2 0 0 152 0.050 11.3 6.31 Paspalum sp. 63 38 1192 864 20 57 2 26 2262 0.334 25.4 14.3 Paspalum cf. notatum 0 91 0 0 0 0 1 2 94 0.184 6.94 3.94 Paspalum cf. urvillei 185 212 6 17 19 1 1 0 441 1.50 20.9 12.3 RANUNCULACEAE Clematis sp. 0 0 0 0 2 1 0 0 3 0.008 0.007 0.007 ROSACEAE Prunus sp. 15 18 23 2 1 2 0 4 65 0.062 1.55 0.890 Prunus cf. serotina 0 0 2 0 0 0 0 0 2 none 0.003 0.002 SAPINDACEAE Acer negundo 88 213 137 49 32 32 40 1753 2344 4.95 50.8 30.4 Acer rubrum 15 7 2 0 0 0 0 0 24 0.067 0.456 0.283 Koelreuteria elegans 18 27 32 5 3 0 0 3 88 0.359 1.48 0.979 SMILACACEAE Smilax sp. 0 0 2 0 0 0 0 0 2 none 0.042 0.023 SOLANACEAE Solanum sp. 0 0 0 15 0 0 0 0 15 none 0.040 0.022 ULMACEAE Ulmus sp. 1 0 0 0 0 1 0 0 2 0.008 0.090 0.053 Ulmus cf. parvifolia 2 3 0 0 31 30 1 227 294 0.277 4.94 2.86 VITACEAE Vitis sp. / Parthenocissus quinquefolia 14 12 25 20 142 16 3 8 240 1.72 2.09 1.92 UNDETERMINED Undetermined 160 1137 162 24 23 11 0 50 1567 1.60 41.1 23.5 TOTAL 1113 3207 2551 2025 1549 313 80 2431 13269 30.9 278 168 None = no seeds intercepted for that taxon and flow regime. 128

PAGE 129

Table C 2 . Shoots intercepted per month and per flow regime in Tumblin Creek . Nets were deployed 27 times during May – December 2012 including baseflow (<0.02 m3/second) and high3/second) events. Shoot s per month Mean shoots per flow regime (individuals/10 0 m 3 ) Family, species May Jun Jul Aug Sep Oct Nov Dec Total Mean length (mm) Mean # of leaves Mean # of roots Base flow High flow All flow ACANTHACEAE Ruellia simplex 2 15 0 1 1 0 0 1 20 163 9.2 12.8 0.010 1.33 0.744 ALISMATACEAE Sagittaria sp. 0 0 1 0 0 0 0 0 1 205 2.0 0.0 none 0.002 0.001 ARALIACEAE Hydrocotyle sp. 1 3 1 0 1 0 0 0 6 76 1.3 6.7 0.021 0.123 0.077 ARECACEAE Sabal palmetto 0 0 1 0 0 0 0 0 1 33 0.0 1.0 none 0.002 0.001 ASTERACEAE Acmella oppositifolia 12 17 4 1 3 4 1 7 49 115 6.3 12.2 0.280 1.33 0.864 Bidens alba 0 0 9 0 0 0 0 2 11 364 8.3 10.1 none 0.052 0.029 Hypochaeris sp. 1 0 0 0 0 0 0 0 1 219 4.0 0.0 none 0.002 0.001 CELTIDACEAE Celtis laevigata 0 0 0 0 1 0 0 0 1 90 5.0 0.0 0.013 none 0.006 COMMELINACEAE Commelinaceae 0 1 0 0 0 0 0 0 1 102 0.0 3.0 none 0.027 0.015 Commelina cf. diffusa 3 10 7 3 0 0 0 5 28 137 4.1 3.9 0.057 0.782 0.460 Tradescantia fluminensis 1 5 0 0 0 0 0 0 6 178 5.0 5.0 none 0.432 0.240 CYPERACEAE Cyperus sp. 0 1 0 0 0 0 0 0 1 73 7.0 0.0 0.010 none 0.005 Cyperus involucratus 0 2 0 0 0 0 0 0 2 70 9.0 0.0 none 0.158 0.088 FABACEAE Senna sp. 1 0 0 0 0 0 0 0 1 115 2.0 1.0 none 0.002 0.001 ONAGRACEAE Ludwigia cf. repens 0 2 0 0 0 0 0 0 2 85 11.0 2.0 0.021 none 0.009 PLANTAGINACEAE Bacopa cf. monnieri 5 7 7 0 4 0 3 1 27 162 7.4 5.9 0.175 0.254 0.219 POACEAE Poaceae 1 1 1 0 1 0 0 0 4 74 2.5 6.0 none 0.035 0.019 Panicum sp. 0 2 4 0 0 0 0 0 6 143 8.4 1.2 0.021 0.084 0.056 Panicum repens 0 0 1 0 0 0 0 1 2 403 2.5 6.0 0.054 0.001 0.025 129

PAGE 130

Table C 2 . Continued Shoot s per month Mean shoots per flow regime (individuals/10 0 m 3 ) Family, species May Jun Jul Aug Sep Oct Nov Dec Total Mean length (mm) Mean # of leaves Mean # of roots Base flow High flow All flow Paspalum sp. 0 0 0 0 0 0 0 1 1 78 6.0 1.0 none 0.002 0.001 PORTULACACEAE Portulaca sp. 1 0 0 0 0 0 0 0 1 62 25.0 2.0 none 0.090 0.050 SAPINDACEAE Acer rubrum 0 0 1 0 0 0 0 0 1 167 2.0 0.0 none 0.002 0.001 UNDETERMINED Undetermined 2 4 0 0 0 0 0 0 6 188 0.0 22.2 0.031 0.031 0.031 TOTAL 30 70 37 5 11 4 4 18 179 – – – 0.691 4.74 2.94 None = no shoot s intercepted for that taxon and flow regime. 130

PAGE 131

APPENDIX D V IABILITY OF INVASIVE SEEDS AND SHOOTS INTERCEPTED IN THE STREAM Table D 1 . Results of seed germination trials for invasive and nonindigenous plant seeds, excluding Ruellia simplex , intercepted by net in Tumblin Creek. Seeds were tested using a ger mination chamber with a 12hour photoperiod at 30C (daytime) and 20C (nighttime) . Each trial lasted 30 days. Species, date of capture Flow rate (m 3 /second) Stream condition* n Percent germination Arecaceae, cf. Butia capitata 01Jul 12 0.033 High flow 2 0.0% Cinnamomum camphora 09Oct 11 Not measured – 22 0.0% 10Dec 12 1.10 High flow 30 6.7% Combined – – 52 3.8% Koelreuteria elegans 05Jul 12 0.058 High flow 20 0.0% Lagerstroemia indica 09Oct 11 Not measured – 65 1.5% 01Sep 12 0.017 Base flow 30 13.3% Combined – – 95 5.3% Macfadyena unguis cati 10Dec 12 1.10 High flow 7 0.0% Sapium sebiferum 09Oct 11 Not measured – 2 0.0% 05Jul 12 0.058 High flow 25 0.0% Combined – – 27 0.0% Ulmus parvifolia 10Dec 12 1.10 High flow 100 2.0% * Baseflow is defined as a flow rate of <0.02 m3/second and highflow is defined as m3/second. 131

PAGE 132

Table D 2 . Results of se ed germination trials for invasiv e and nonindigenous plant seeds, excluding Ruellia simplex , intercepted by net in Tumblin Creek. Seed germination was determined using soilfilled flats under artificial grow lights (12 hour photoperiod), and some natural light, in an unheated glas shouse. Each trial lasted at least 50 days. Species, date of capture Flow rate (m 3 /second) Stream condition* n Duration of germination test (days) Percent germination Cinnamomum camphora 1 8 Mar 12 0.003 Base flow 50 97 10.0 Koelreuteria elegans 1 8 Mar 12 0.00 3 Base flow 10 112 0.0 Lagerstroemia indica 18 Mar 12 0.00 3 Base flow 32 63 0.0 Macfadyena unguis cati 18 Mar 12 0.00 3 Base flow 5 146 40.0 Commelina cf. diffusa 03 Jun 12 0.014 Base flow 10 50 0.0 * Baseflow is defined as a flow rate of <0.02 m3/second. 132

PAGE 133

Table D 3 . Results of shoot survival and growth trials for invasive and nonindigenous plant shoots, excluding Ruellia simplex and Tradescantia fluminensis , intercepted by net in Tumb lin Creek. S hoots were tested for survival in either water or soil. Each trial lasted at least 17 days (or until mortality) under artificial grow lights (12hour photoperiod) and some natural light in a n unheated glasshouse. Metrics at time of capture Metrics following survival trial Species, date of capture Flow rate (m 3 /sec. ) Stream condition * n Length of Shoot (mm) Number of leaves Number of roots Growth medium, duration Change in shoot length (%) Change in number of leaves (%) Change in number of roots (%) Percent survival Commelina cf. diffusa 18Mar 12 0.003 High flow 3 292 (mean) 3.3 (mean) 10.0 (mean) Water, 61 days +2 (mean) + 615 (mean) + 195 (mean) 67.0 20May 12 0.002 Base flow 1 125 5 1 Water, 42 days +176 +20 +2900 100 14Jun12 0.025 High flow 7 104 (mean) 4.4 (mean) 0.7 (mean) Water, 17days + 115 (mean) + 29 (mean) + 786 (mean) 100 0 1 Jul 12 0.033 High flow 1 37 3 10 Water, 7 days Dead – – 0 .0 Combined – – 12 147 (mean) 4.1 (mean) 3.8 (mean) – +99 (mean) +145 (mean) +879 (mean) 83.0 Panicum cf. repens 18Mar 12 0.00 3 High flow 1 260 0 15 Water, 61 days 1 0 +13 100 * Baseflow is defined as a flow rate of <0.02 m3/second and high3/second. 133

PAGE 134

LIST OF REFERENCES Adams CR and NM Steigerwalt ( 2010) Research Needs and Logistic Impediments in Restoration, Enhancement, and Management Projects: A Survey of Land Managers. Publication ENH1161, Environmental Horticulture Department, Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida, Gainesville, FL. http://edis.ifas.ufl.edu/ ep423 . Accessed 21 November 2010 Adams SB, PB Hamel, K Connor, B Burke, ES Gardiner and D Wise (2007) Potential roles of fish, birds, and water in swamp privet ( Forestiera acuminata) seed dispersal. Southeastern Naturalist 6:669 – 682 Andersen L ( 2001) Paynes Prairie. The Great Savanna: A History and Guide. Pineapple Press, Sarasota, FL Armstrong RA (2014) When to use the Bonferroni correcti on. Ophthalmic & Physiol Opt doi: 10.1111/opo.12131 Bailey RM and HM Harrison (1948) Food habits of the southern channel catfish ( Ictalurus lacustris punctatus ) in the Des Moines River, Iowa. T Am Fish Soc 75:110– 138 BoothBinczik S (2006) Response Form, IFAS Assessment of Nonnative Plants in Florida’s Natural Areas: January 2005 . http://plants.ifas.ufl.edu/assessment/resp_forms_pdf/Tradescantia_fluminensis. pdf . Accessed 24 May 2014 Burns JH (2004) A comparison of invasive and noninvasive dayflowers (Commelinaceae) acros s experimental nutrient and water gradients. Diversity and Distributions 10(5) :387 – 397 Butcher ER and D Kelly (2011) Physical and anthropogenic factors predict distribution of the invasive weed Tradescantia fluminensis . Austral Ecol 36:621– 627 Chick JH, RJ Cosgriff, and LS Gittinger (2003) Fish as potential dispersal agents for flooplain plants: First evidence in North America. Can J Fish Aquat Sci 60:1437– 1439 CH2M Hill ( 1985) Tumblin Creek Stormwater Quality Evaluation. CH2M Hill, Gainesville , FL Combroux I, G Bornette, and C Amoros ( 2001 ) Regenerative strategies of aquatic plants in disturbed habitats: the role of the propagule bank. Arch Hydrobiol 152:215 – 235 134

PAGE 135

Cooper, A (2010) Response Form — Infraspecific Taxon Protocol. http://plants.ifas.ufl.edu/assessment/datasheets/Ruellia_simplex/Purple_Showers _Response_2010.pdf . Accessed 8 June 2014 Cooper, A (2012) Response Form — Infraspec ific Taxon Protocol. http://plants.ifas.ufl.edu/assessment/datasheets/Ruellia_simplex/Response_For m_R_simplexR10108.pdf . Accessed 25 May 2014 Cuevas JG, A Marticorena, and LA Cavieres (2004) New additions to the introduced flora of the Juan Fernndez Islands: origin, distribution, life history traits, and potential of invasion. Revista Chilena de Histora Natural 77(3): 523 – 538 da Conceio Vellozo JM (1825) Florae Fluminensis, seu Descriptionum Plantarum Praefectura Fluminensi Sponte Nascentium Liber Primus ad Systema Sexuale Concinnatus Augustissimae Dominae Nostrae . Rio de Janeiro, Brazil Delorit RJ (1970) Illustrated Taxonomy Manual of Weed Seeds. Agronomy Publications, River Falls, WI Elton CS ( 1958) The Ecology of Invasions by Animals and Plants . Methuen and Co., Ltd., Strand, London Contributions to the leaf and stem anatomy of Tradescantia fluminensis : an alien species new to the flora of Turkey. Artvin oruh niversitesi Orman Fakltesi Dergisi 13(2):270 – 277 Ezcurra C and TF Daniel (2007) Ruellia simplex , an older and overlooked name for Ruellia twe ediana and Ruellia coerulea (Acanthaceae). Darwiniana 42:201– 203 Faden RB (2000) Commelinaceae. Pp. 170 – 197 in: Flora of North America Editorial Committee (ed s.) Flora of North America North of Mexico. Volume 22: Magnoliophyta: Alismatidae, Arecidae, Commelinidae (in part), and Zingiberidae. New York, NY Florida Exotic Pest Plant Council ( 201 3 ) Florida Exotic Pest Plant Council’s 2013 list of Invasive Plant Species. http://www.fleppc.org/list/2013PlantList.pdf . Accessed 27 January 201 4 Freyre R, A Moseley, GW Knox, and SB Wilson (2012) Fruitless Ruellia simplex R10 102 (‘Mayan Purple’) and R10108 (‘Mayan White’). HortScience 47:1808– 1814 Freyre R and SB Wilson (2014) Ruellia simplex R10 105 Q54 (‘Mayan Pink’). HortScience 49:499– 502 Goulding M (1980) The Fishes and the Forest, Explorations in Amazonian Natural History . University of California Press, Berkeley, CA 135

PAGE 136

Guzou A, M Trueman, CE Buddenhagen, S Chamorro, AM Guerrero, P Pozo, and R Atkinson ( 2010) An extensive alien plant inventory from the inhabited areas of Galapagos. PLoS ONE 5(4):e10276 Hall DW and WJ Weber (2011) Wildflowers of Florida and the Southeast. DW Hall and JH Byrd, China Harlow WM (1946) Fruit Key and Twig Key to Trees and Shrubs. Dover Publications, Inc., New York, NY Holm L, JV Pancho, JP Herberger, and DL Plucknett ( 1979) A Geographic Atlas of World Weeds. John Wiley & Sons, New York, NY Hoyer MV and DE Canfield ( 1994) Handbook of Common Freshwater Fish in Florida Lakes . University of Florida/Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, Gainesville, FL Huey LA, GJ Steck, and AM Fox ( 2007) Biological notes on Melanogromyza ruelliae (Diptera: Agromyzidae), a seed feeder on the invasive Mexican petunia, Ruellia tweediana (Acanthaceae). Fla Entomol 90:763 – 765 Hupp KVS ( 2007) Investigating the Determinants of Local Scale Distribution of Ruellia tweediana (Synonym R. brittoniana) in Natural Areas. MS thesis , University of Florida Hupp KVS, AM Fox, SB Wilson, EL Barnett, and RK Stocker ( 2009) Natural Area Weeds: Mexican Petunia ( Ruellia tweediana) . Publication ENH1155, Environmental Horticulture Department, Florida Cooperative Extension Service, Institute of Food and Agricultural Sciences, University of Florida, Gainesville, FL. http://edis.ifas.ufl.edu/pdffiles/ep/ep41500.pdf . Accessed 4 January 2011 Hurrel GA and CS Lusk ( 2012) Survival of Tra descantia fluminensis in sea water. Ecol Manage Restor 13: 314– 316 Ikeda H and K Itoh (2001) Germination and water dispersal of seeds from a threatened plant species Penthorum chinense. Ecol Res 16:99– 106 Ingram DL and TH Yeager ( 2010) Propagation of Landscape Plants . Document CIR579, Environmental Horticulture Department, Florida Cooperative Extension Service, Institute of Food and Agricultural Services, Univer sity of Florida Institute of Food and Agricultural Sciences (2012) Conclusions from the IFAS Assessment of Non native Plants in Florida’s Natural Areas . http://plants.ifas. ufl.edu/assessment/conclusions/ConclusionsTableSortedByGenus.pdf . Accesse d 24 May 2014 Kelly D and JP Skipworth ( 1984) Tradescantia fluminensis in a Manawatu (New Zealand) forest: II. Management by herbicides. New Zeal J Bot 22 :399 – 402 136

PAGE 137

Landers JL and AS Johnson (1976) Bobwhite Quail Food Habits in the Southeastern United States with a Seed Key to Important Foods. Misc. Pub. No. 4, Tall Timbers Re search Station, Tallahassee, FL Langeland KA, HM Cherry, CM McCormick, and KA Craddock Burks ( 2008) Identification and Biology of Nonnative Plants in Florida’s Natural Areas. IFAS Communication Services, Univers ity of Florida Liogier HA and LF Martorell (1982) Flora of Puerto Rico and Adjacent Islands: A Systematic Synopsis. University of Puerto Rico, San Juan, Puerto Rico Maule HG, M Andrews, JD Morton, AV Jones, and GT Daly ( 1995) Sun/shade acclimation and nitrogen nutrition of Tradescantia fluminensis , a problem weed in New Zealand forest remnants. New Zeal J Ecol 19:35 46 McCarthy PM ( 2009) Travel Times, Streamflow Velocities, and Dispersion Rates in the Yellowstone River, Montana. U.S. Geological Survey Scientific Investigations Report 2009– 5261, Reston, VA McMillan BA ( 1999) Tradescantia fluminensis : An Exotic Invasive Species in Florida’s Mesic Hardwood Forests . MS thesis , Univer sity of Florida Meisenburg MJ ( 2007) Reprod uctive and Dispersal Ecology of the Invasive Coral Ardisia ( Ardisia crenata) in Northern Florida. MS thesis, Universit y of Florida Musil AF (1963) Identification of Crop and Weed Seeds. Agriculture Handbook No. 219, Agricultural Marketing Service, U.S. Department of Agriculture, Washington, D.C. Natural Resources Conservation Service ( 2010) Web Soil Survey . http://websoilsurvey.nrcs.usda.gov/app/WebSoilSurvey.aspx . Accessed 4 September 2010 Nelson G ( 1995 ) Exploring Wild North Florida, A Guide to Finding the Natural Areas and Wildlife of North Central and Northeast F lorida. Pineapple Press, Sarasota, FL Nordlie FG ( 1975) Plankton communities of three central Florida lakes. Hydrobiol 48:65 – 78 Ogle C and B Lovelock (1989) Methods for the Control of Wandering Jew ( Tradescantia fluminensis ) at “Rangitawa”, Rangitikei Dist rict, and Notes on Other A spects of Conserving this Forest Remnant. Science and Research Internal Report 56, New Zealand Department of Conser vation, Wellington, New Zealand 137

PAGE 138

Orange Creek Basin Working Group and Florida Department of Environmental Protection ( 2008) 2007 Orange Creek Basin Management Action Plan for the Implementation of Total Maximum Daily Loads Adopted by the Florida Department of Environmental Protection for Newnans Lake, Orange Lake, Lake Waubu rg, Hogtown Creek, Sweetwater Branch, Tumblin Creek, and Alachua Sink. http://www.gainesvillecreeks.org/OCBMAP_Adoptable%20final%20 edited%2052808.pdf . Accessed 22 March 2012 Orchard AE ( 1994) Flora of Australia, Volume 49, Oceanic Islands 1. Australian Government Publishing Service, Canberra, A ustralia Page LM and BM Burr (2011) Peterson Field Guide to Freshwater Fishes of North America North of Mexico. Hought on Mifflin Harcourt, Boston, MA Pe rneger TV (1998) What’s wrong with Bonferroni adjustments. BMJ 316:1236– 1238 Riis T ( 2008) Dispersal and colonization of plants in lowland streams: success and bottlenecks. Hydrobiol 596:341 – 351 Riis T and BJF Biggs ( 2003) Hydrologic and hydraulic control of macrophyte establishment and performance in streams. Limnol Oceanogr 48(4):1488– 1497 Riis T and K SandJenson ( 2006) Dispersal of plant fragments in small streams. Freshwater Biol 51:274– 286 Robbins JA and MR Evans ( No date) Greenhouse and Nursery Series: Growing Media for Container Production in a Greenhouse or Nursery, Part I — Components and Mixes. University of Arkansas Cooperative Extension Service, Fayetteville, AR SandJensen K, K Andersen, and T Andersen ( 1999) Dynamic properties of recruit ment, expansion and mortality of macrophyte patches in streams. Int Rev Hydrobiol 5:497– 508 Schmidt CA ( 2005) Floodplain Impacts from Channelization and Urbanization: A Characterization of the Tumblin Creek Delta Floodplain, Gainesville, FL. MS thesis, Univer sity of Florida Schmitz DC, D Simberloff, RH Hofstetter, W Haller, and D Sutton ( 1997) The ecological impact of nonindigenous plants. Pp. 39– 61 In: D. Simberloff, D.C. Schmitz, and T.C. Brown (eds.) Strangers in Paradise: Impact and Management of Non indigenous Species in Florida. Island Press, Washington, D.C. Smith NJH (1981) Man, Fishes, and the Amazon. Columbia University Press, New York, NY Standish RJ (2001 a ) Prospects for Biological Control of Tradescantia fluminensis Vell. (Commelinaceae). Sci ence Internal Series 9, New Zealand Department of Conser vation, Wellington, New Zealand 138

PAGE 139

Standish RJ ( 2001b ) The Ecological Impact and Control of an Invasive Weed Tradescantia fluminensis in Lowland Forest Remnants. Ph.D. Dissertation, Institute of Natural Resources, Massey University , Palmerston North, New Zealand Staples GW, DR Herbst, and CT Imada ( 2006 ) New Hawaiian plant records for 2004. In: NL Evenhuis and LG Eldredge (eds.) Records of the Hawaii Biological Survey for 2004– 2005, Part 2: Notes. Bishop Museum Occasional Papers 88(1):6 – 9 Tan K ( 1984 ) Commelina L. pp. 554 – 555. In: PH Davis (ed.) Flora of Turkey and the East Aegean Islands . Volume 8. Edinburgh University Press, Edinburgh, Scotland. Thaman RR, FR Fosberg, HI Manner, and DC Hassall ( 1994) The Flora of Nauru. Atoll Research Bulletin No. 392, National Museum of Natural History, Smithsonian Institution, Washington, D.C. Thomas BP, E Cummings, and WH Wittstruck ( 1985) Soil Survey of Alachua County, Florida . U.S. Department of Agriculture, Soil Conservation Service, U.S. Government Printing Office, Washington, D.C. Tolkach VF, AK Chuyan and AV Krylov ( 1990) Characterization of a potyvirus isolated from Tradescantia albiflora in a southern locality of the Soviet far east. Byulleten’ Glavnogo Botanicheskogo Sada 157 :76 – 80 United States Forest Service (1948) Woody p lant Seed Manual. U.S. Department of Agriculture, Misc. Pub. No. 654, U.S. Government Printing Office, Washington, D.C. United States Geological Survey ( 1890) Gainesville East Quadrangle, Florida [map]. 15 minute series. 1:62,500 scale. U.S. Department of the Interior, USGS, Tallahassee, FL University of Florida ( 2005) Invasive and Other NonNative Plants Found in Public Waters and Conservation Lands of Florida and the Southeastern United Stat es . University of Florida Center for Aquatic and Invasive Plants, Gainesville, FL Villazon KA ( 2009) Methods to Restore Native Plant Communities after Invasive Species Removal: Marl Prairie Ponds and an Abandoned Phosphate Mine in Florida. MS thesis, Univer sity of Florida Wilson SB and LA Mecca (2003) Seed production and germination of eight cultivars and the wild type of Ruellia tweediana: A potentially invasive ornamental. J Environ Hort 21:137 – 143 Wilson SB, PC Wilson, and JA Albano ( 2004) Growth and development of the native Ruellia caroliniensis and invasive Ruellia tweediana . HortSci ence 39:1015– 1019 139

PAGE 140

Wilson SB, R Freyre, GW Knox, and Z Deng ( 2012 ) Characterizing the invasive potential of ornamental plants. Acta Hortic 937:1183 – 1192 Wirth FF, KJ Da vis, and SB Wilson (2004) Florida nursery sales and economic impacts of 14 potentially invas ive ornamental plant species. J Environ Hort 22:12 – 16 Witz t um A and K Schulgasser (1995) The mechanics of seed expulsion in Acanthaceae. J Theor Biol 176:531 – 542 140

PAGE 141

BIOGRAPHICAL SKETCH Jason was born and ra ised near Rochester, New York . He received his AAS in Fisheries Technology in 1992 from the State University of New York (SUNY) at Cobleskill and his BS in Biology in 1994 from SUNY Brockport . Following an internship at National Audubon’s Constitution Marsh Sanctuary in New York’s Hudson River Valley in the summer of 1995, he landed another Audubon internship, but this one was at the Corkscrew Swamp Sanctuary in Naples, Florida. He quickly fell in love wit h the biodiversi ty of the state. Jason has worked as a biological scientist for local, state, and federal governmental agencies ; non profit organizations; and private environmental consulting firms . He and his wife Jenny live in Gainesville, Florida, since moving from southwest Florida in 2004. He currently works as a biologist and project manager at a small environmental consulting firm. In his spare time he enjoys hiking, trail running, obstacle course racing, biking, fishing, fossil hunting and collecting, and reading. 141