<%BANNER%>

Biology and Control of the Invasive Aquatic Plant Crested Floating Heart (Nymphoides Cristata)

Permanent Link: http://ufdc.ufl.edu/UFE0045143/00001

Material Information

Title: Biology and Control of the Invasive Aquatic Plant Crested Floating Heart (Nymphoides Cristata)
Physical Description: 1 online resource (109 p.)
Language: english
Creator: Willey, Leif N
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: biology -- competition -- endothall -- herbicide -- imazamox -- imazapyr -- invasive -- light -- management -- nymphoides -- ph -- sediment
Agronomy -- Dissertations, Academic -- UF
Genre: Agronomy thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: First documented in Florida in 1996, the invasive, floating-leaved plant crested floating heart (Nymphoides cristata) has been expanding its range throughout Florida and the Southeastern US. To address broad information gaps on management and biology of N. cristata, research was conducted to evaluate herbicide efficacy and how selected environmental factors influence growth. Herbicide screening suggests N. cristata has broad tolerance to most registered aquatic herbicides; however, submersed applications of liquid endothall (amine salt) and diquat and foliar applied imazamox and imazapyr reduced biomass below a pre-treatment reference. Evaluation of the influence of sediment type and fertility on growth found a significant interaction between these factors. Results further suggest that N. cristata can grow in a wide range of sediments but prefers acidic, fertile sediments. Plants exhibited a negative response to increasing pH. Growth from small submersed ramets showed a positive linear relationship between light availability and biomass, and a non-linear relationship to leaf production. Ramets grew to the water surface at 1% incident light; however, growth was greatest above 25% light. A competition study compared growth of N. cristata from ramets versus hydrilla (Hydrilla verticillata) from apical shoots. N. cristata was able to produce more biomass at planting ratios greater than 1: 5 (N. cristata:hydrilla). Anatomical comparison of N. cristata to yellow waterlily (Nymphaea mexicana) found that N. cristata leaves and stems were thinner and contained fewer supporting structures and vascular tissue. Collectively these data contribute greatly to current knowledge of the biology and management of N. cristata.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Leif N Willey.
Thesis: Thesis (M.S.)--University of Florida, 2012.
Local: Adviser: Netherland, Michael D.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0045143:00001

Permanent Link: http://ufdc.ufl.edu/UFE0045143/00001

Material Information

Title: Biology and Control of the Invasive Aquatic Plant Crested Floating Heart (Nymphoides Cristata)
Physical Description: 1 online resource (109 p.)
Language: english
Creator: Willey, Leif N
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2012

Subjects

Subjects / Keywords: biology -- competition -- endothall -- herbicide -- imazamox -- imazapyr -- invasive -- light -- management -- nymphoides -- ph -- sediment
Agronomy -- Dissertations, Academic -- UF
Genre: Agronomy thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: First documented in Florida in 1996, the invasive, floating-leaved plant crested floating heart (Nymphoides cristata) has been expanding its range throughout Florida and the Southeastern US. To address broad information gaps on management and biology of N. cristata, research was conducted to evaluate herbicide efficacy and how selected environmental factors influence growth. Herbicide screening suggests N. cristata has broad tolerance to most registered aquatic herbicides; however, submersed applications of liquid endothall (amine salt) and diquat and foliar applied imazamox and imazapyr reduced biomass below a pre-treatment reference. Evaluation of the influence of sediment type and fertility on growth found a significant interaction between these factors. Results further suggest that N. cristata can grow in a wide range of sediments but prefers acidic, fertile sediments. Plants exhibited a negative response to increasing pH. Growth from small submersed ramets showed a positive linear relationship between light availability and biomass, and a non-linear relationship to leaf production. Ramets grew to the water surface at 1% incident light; however, growth was greatest above 25% light. A competition study compared growth of N. cristata from ramets versus hydrilla (Hydrilla verticillata) from apical shoots. N. cristata was able to produce more biomass at planting ratios greater than 1: 5 (N. cristata:hydrilla). Anatomical comparison of N. cristata to yellow waterlily (Nymphaea mexicana) found that N. cristata leaves and stems were thinner and contained fewer supporting structures and vascular tissue. Collectively these data contribute greatly to current knowledge of the biology and management of N. cristata.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Leif N Willey.
Thesis: Thesis (M.S.)--University of Florida, 2012.
Local: Adviser: Netherland, Michael D.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2012
System ID: UFE0045143:00001


This item has the following downloads:


Full Text

PAGE 1

1 BIOLOGY AND CONTROL OF THE INVASIVE AQUATIC PLANT CRESTED FLOATING HEART (NYMPHOIDES CRISTATA) By LEIF NEWKIRK WILLEY A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE RE QUIREMENTS FOR THE DE GREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 Leif Newkirk Willey

PAGE 3

3 To my Dad and Mom who have believed in and supported me throughout the pursuit of my goals, I am forever grateful for all the opportunities yo u have given me

PAGE 4

4 ACKNOWLEDGMENTS First and most importantly, I am thankful to God for all of the blessings He has given me in life and also for the creation of this fascinating World in which we all live I would like to thank my major advisor Dr. Mic hael Netherland for all of his time spent going over my writings, presentations, study designs and helping me develop an understanding of proper scientific methods. I am grateful to him further for opening his home to me during Holidays in which I was unab le to return home to be with my own family. I am thankful for the support, guidance and suggestions from my advisory committee: Dr. William Haller, Dr. Greg MacDonald, Dr. Ken Langeland and Dr. Dennis Odero. I am thankful to Garrett Hoyer and Zack Banks, their assistance in harvesting studies, recording data, maintaining cultures and handling tasks while I was away f rom town is greatly appreciated. Roberta Goodwin was also incredibly helpful. It seemed almost every semester she had to go above and beyond in order to track some paperwork that had went missing or straighten out a registration mishap. I am thankful to Purdue University, for instilling a deeper appreciation and interest in aquatic sciences and for providing me with a solid and respected academ ic found ation to build on. D uring my time here my professors were always willing to go that extra mile to help their students and inspired many including myself and are the reason Purdue is a world class institution. I would especially like to than k Dr. C arole Lembi, her passion for teaching and aquatic plant management was very influential in my decision to pursue a Master of Science degree. I am thankful to the employees at Aquatic Control Inc., especially David Isaacs, Nathan Long, Nate Eddy, and Matth e w Johnson. The opportunities, responsibiliti es and challenges you

PAGE 5

5 presented me helped to mold a strong work ethic and understanding of how to handle myself in a professional manner. For all of their support I thank my family. My m om, continually support ed my quest for knowledge and did everything she could to help me follow my dreams. My dad, through his example I developed a love for the outdoors and its responsible management at a young age and learned cou ntless, important life lessons from him. Fina lly I have to thank all the members of Gator Skeet and Trap Club and the Florida Gators Trap Skeet and Sporting Team. You all have been my family and best friends during my time here, each and every one of your friendships are of great value to me.

PAGE 6

6 TA BLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ........................... 8 LIST OF FIGURES ................................ ................................ ................................ ......................... 9 LIST OF ABBREVIATIONS ................................ ................................ ................................ ........ 11 ABSTRACT ................................ ................................ ................................ ................................ ... 12 CHAPTER 1 INTRODUCTION AND LITERATURE REVIEW ................................ .............................. 14 Biology of Two Invasive Species of Nymphoides ................................ ................................ 14 History and Or igin of Crested Floating Heart ( Nymphoides cristata ) ............................ 14 Identification and Biology ................................ ................................ ............................... 14 Habitat ................................ ................................ ................................ ............................. 16 Current Distribution ................................ ................................ ................................ ......... 16 Invasive Potential ................................ ................................ ................................ ............ 17 History and Origin of Yellow Floating Heart ( Nymp hoides peltata ) .............................. 18 Identification and Biology ................................ ................................ ............................... 18 Habitat ................................ ................................ ................................ ............................. 19 Curr ent Distribution ................................ ................................ ................................ ......... 19 Invasive Potential ................................ ................................ ................................ ............ 19 Preventative management of Nymphoides ................................ ................................ ...... 20 Cultural Control ................................ ................................ ................................ ............... 20 Mechanical Control ................................ ................................ ................................ ......... 21 Biological Control ................................ ................................ ................................ ........... 21 Chemical Control ................................ ................................ ................................ ............. 22 Objectives ................................ ................................ ................................ ........................ 23 2 EFFECTS OF AQUATIC HERBICIDES ON CRESTED FLOATING HEART ................. 27 Introduction ................................ ................................ ................................ ............................. 27 Materials and Methods ................................ ................................ ................................ ........... 30 Subsurface herbicide applications ................................ ................................ ................... 30 Foliar herbicide application ................................ ................................ ............................. 33 Results and Discussion ................................ ................................ ................................ ........... 33 Subsur face herbicide application ................................ ................................ ..................... 33 Foliar herbicide application ................................ ................................ ............................. 40 3 THE INFLUENCE OF SEDIMENT TYPE AND FERTILITY ON THE GROWTH OF CR ESTED FLOATING HEART ................................ ................................ ........................... 57

PAGE 7

7 Introduction ................................ ................................ ................................ ............................. 57 Materials and Methods ................................ ................................ ................................ ........... 58 Study 1 ................................ ................................ ................................ ............................. 58 Study 2 ................................ ................................ ................................ ............................. 59 Study 3 ................................ ................................ ................................ ............................. 60 Results and Discussion ................................ ................................ ................................ ........... 60 Study 1 ................................ ................................ ................................ ............................. 60 Study 2 ................................ ................................ ................................ ............................. 60 Study 3 ................................ ................................ ................................ ............................. 62 4 THE INFLUENCE OF LIGHT AVAILABILITY ON GROWTH OF CRESTED FLOATING HEART ESTABLISHED FROM RAMETS ................................ ..................... 69 Introduction ................................ ................................ ................................ ............................. 69 Materials and Methods ................................ ................................ ................................ ........... 69 Number of Floating Leaves ................................ ................................ ............................. 69 Dry Weight ................................ ................................ ................................ ...................... 71 Results and Discussion ................................ ................................ ................................ ........... 71 Number of Floating Leaves ................................ ................................ ............................. 71 Dry Weight ................................ ................................ ................................ ...................... 72 5 COMPETITION BETWEEN CRESTED FLOATING HEART AND HYDRILLA IN A REPLACEMENT SERIES TRIAL ................................ ................................ ........................ 79 Introduction ................................ ................................ ................................ ............................. 79 Materials and Methods ................................ ................................ ................................ ........... 80 Results and Discussion ................................ ................................ ................................ ........... 81 6 A COMPARATIVE ANATOMY OF CRESTED FLOATING HEART AND YELLOW WATER LILY LEAVES AND STEMS ................................ .............................. 87 Introduction ................................ ................................ ................................ ............................. 87 Materials and Methods ................................ ................................ ................................ ........... 88 Le af Sections ................................ ................................ ................................ ................... 88 Stem Sections ................................ ................................ ................................ .................. 89 Results and Discussion ................................ ................................ ................................ ........... 90 Leaf S ections ................................ ................................ ................................ ................... 90 Stem Sections ................................ ................................ ................................ .................. 92 7 CONCLUSIONS AND RECOMMENDATIONS ................................ ................................ 98 APPENDIX: ANOVA TABLES FROM CHAPTER 3 STUDY ................................ ................ 101 LIST OF REFERENCES ................................ ................................ ................................ ............. 102 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ....... 109

PAGE 8

8 LIST OF TABLES Table page 2 1 Subsurface concentrations, exposure times and foliar rates of herbicides used singly ..... 45 2 2 Subsurface concentrations, exposure times and foliar rates of herbicides used in combinations ................................ ................................ ................................ ..................... 46 3 1 Physical characterization of sediments used in sediment type/growth study ................... 64 3 2 Background nutrient and sediment chemistry analysi s from sediments used in sediment type/growth study ................................ ................................ .............................. 64 3 3 Hydrated lime applica tion rates and resulting sediment pH ................................ .............. 65 4 1 Target ambient light levels and the actual light levels measured in light intensity study ................................ ................................ ................................ ................................ .. 75 4 2 N. cristata dry weight at various light intensities at 2 and 4 weeks after planting ........... 75 5 1 Com parison of N. cristata and hydrilla dry weight after 7 weeks of competition ........... 84 6 1 Features measured on N. cristata and Nymphaea mexicana for anatomical and ph ysical comparisons. ................................ ................................ ................................ ....... 95 A 1 2 way ANOVA table for 6 week floating leaf data from trial 1 of Chapter 3 ................ 101 A 2 2 way ANOVA table for 6 week dry weight data from trial 1 of Chapter 3 .................. 101 B 1 2 way ANOVA table for 6 week floating leaf dat a from trial 2 of Chapter 3 ................ 101 B 2 2 way ANOVA table for 6 week dry weight dat a from trial 2 of Chapter 3 .................. 101

PAGE 9

9 LIST OF FIGURES Figure page 1 1 Image of Nymphoides cristata growing in the Santee Cooper Reservoir ......................... 24 1 2 Ramet of N. cristata ................................ ................................ ................................ .......... 24 1 3 Line drawing of a crested floating heart (Nymphoides cristata) flower ........................... 25 1 4 Image comparing leaves of the native Nymphoides aquatica and the exotic Nymphoides cristata ................................ ................................ ................................ .......... 25 1 5 Root system of an established Nymphoides cristata. ................................ ........................ 26 2 1 Dry weight response of N. cristata to triclopyr 4 WAT ................................ .................... 47 2 2 Dry weight response of N. cristata to 2,4 D 4 WAT ................................ ........................ 48 2 3 Dry weight response of N. cristata to a combination of 2,4 D (amine) + triclopyr (amine) 4 WAT. ................................ ................................ ................................ ................ 49 2 4 Dry weight response of N. cristata to endothall (dipotassium salt) 4 WAT ...................... 50 2 5 Dry weight response of N. cristata to 4 week static treatments with endothall (dipotassium salt) ................................ ................................ ................................ .............. 51 2 6 Dry weight response of N cristata to bispyribac sodium 4 WAT ................................ .... 51 2 7 Dry weight response of N. cristata to flumioxazin and carfentrazone ethyl 4 WAT ....... 52 2 8 Dry weight response of N. cristata to diquat 4 WAT ................................ ...................... 53 2 9 Dry weight response of N. cristata to endothall (amine salt) 4 WAT ............................... 54 2 10 Dry weight response of N. cristata to a combination of diquat and dipotassium salt endothall 4 WAT ................................ ................................ ................................ ............... 55 2 11 Dry weight response of N. cristata to foliar aplpied ALS herbicides 4 WAT .................. 56 2 12 Dry weight response of N. cristata tofoliar applications of select herbicide combinations 4 WAT. ................................ ................................ ................................ ....... 56 3 1 Study 1, r esponse of N. cristata grown in sand, muck and potting soil sediments at two fertilizer rates 6 WAT ................................ ................................ ................................ 66 3 2 Study 2, response of N. cristata grown in sand, muck, potting soil and clay sediment s and two fertilizer rates 6 WAT ................................ ................................ ......................... 67

PAGE 10

10 3 3 Linear regression of N. cristata dry weight versus sediment pH ................................ ....... 68 4 1 Study 1, number o f floating leaves of N. cristata in response to light availability 2 and 4 weeks after planting ................................ ................................ ................................ 76 4 2 Study 2, number of floating leaves of N. cristata in response to light availability in t he second study 2 and 4 weeks after planting ................................ ................................ .. 77 4 3 Dry weight of N. cristata in response to light availability at 2 and 4 weeks after planting ................................ ................................ ................................ ............................ 78 5 1 The dry weight of N. cristata, hydrilla, and total plant dry weight after 7 weeks of growth in a replacement series competition study ................................ ............................ 85 5 2 N. cristata and hydrilla growing in replacement series competition for 7 weeks at the 3:3 planting ratio ................................ ................................ ................................ ................ 86 6 1 Leaf tissue samples of N. cristata and Nymphaea mexicana showing thickness, vascular bundles and support ing structures under different magnifications ..................... 96 6 2 Stem cross section of N. cristata and Nymphaea mexicana showing diameter, vascular bundles and supporting structures under different magni fication ....................... 97

PAGE 11

11 LIST OF ABBREVIATION S A E Acid equivalent A I Active ingredient ALS Acetolactate synthase C ET Concentration exposure time DAT Days after treatment HAT Hours after treatment PPO Protoporphyrinogen oxidase WAT Weeks after treatment

PAGE 12

12 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements fo r the Degree of Master of Science BIOLOGY AND CONTROL OF THE INVASIVE AQUAT IC PLANT CRESTED FLOATING HEART (NYMPHOIDES CRISTATA) By Leif Newkirk Willey December 2012 Chair: Michael Netherland Major: Agronomy T he invasive, floating leaved plant crested floating heart ( Nymphoides cristata ) f irst documented in Florida in 1996, ha s been extending its range throughout Florida and the Southeastern US. To address broad information gaps on management and biology of N. cristata research was conducted to evaluate herbicide efficacy and how selected environmental factors influence growth Herbicide screening suggests N. cristata has broad tolerance to most registered aquatic herbicides; however, submersed applications of liquid endothall (amine salt) and diquat and foliar applied imazamox and imazapyr reduced biomass below a pre treatment reference. Evaluation of the influence of sediment type and fertility on growth found a significant interaction betwee n these factors. These data suggest that N. cristata can grow in a wide range of sediments but prefers acidic, fertile sediments. Plants exhibited a negative response to increasing pH. Growth from small submersed ramets showed a positive linear relationship between light availability and biomass, and a non linear relationship to leaf production. Ramets grew to the water surface at 1% incide nt light; however, growth was greatest above 25% light. The growth of N. cristata from ramets versus hydrilla ( Hydrilla verticillata ) from apical shoots in a competition study showed N. cristata was able to produce more biomass at planting ratios greater than 1: 5 ( N. cristata :hydrilla). Anatomical comparison of N. cristata to yellow waterlily ( Nymphaea

PAGE 13

13 mexicana ) found that N. cristata leaves and stems were thinner and contained fewer supporting structures and vascular tissue. Collectively these data contr ibute greatly to current knowledge of the biology and management of N. cristata

PAGE 14

14 CHAPTER 1 INTRODUCTION AND LIT ERATURE REVIEW Biology of Two Invasive Species of Nymphoides History and Origin of Crested Floating Heart ( Nymphoides cristata ) Crested floatin g heart ( Nymphoides cristata (Roxb.)) Kuntze, hereafter referred to as N. cristata is native to Asia where it is found as far south as the island of Sri Lanka and as far north as the Jiangsu Province in China along the East China Sea (Burks, 2002a). The p lant is marketed in the water garden and aquarium trade available throughout North America. Within the United States, N. cristata has escaped from cultivation and become established in seve ral bod ies of water in Florida. N. cristata was first reported in 1996, in Horseshoe Lake in Collier County, Florida (Burks, 2002a) and subsequently, ha s been confirmed in numerous bodies of water, including, water management canals in south Florida, Lake Okeecho bee, and the Big Cypress National Preserve. One of the largest N. cristata infestations has occurred in the 64,750 ha Santee Cooper reservoir system in South Carolina. This infestation covers over 1,416 ha ( Chip Davis, personal communication, 2012) (Figure 1 1 ). Identification and Biology N. cristata is a dicotyledonous aquatic plant of the Menyanthaceae or buckbean family. N. cristata can be identified by the slender bundles of tuberous roots on the underside of its floating leaves (Figure 1 2 ) (Burks, 2002a). The flowers are unique to the species since they bear an erect fold of tissue that runs down the length of the upper side of the petal (Figure 1 3 ) (Burks, 2002a). N. cristata shares the same range as two native species of Nymphoides, N. aquatica a nd N. cordata They are similar in appearance, but can be differentiated by the flowers and the undersides of the leaves. N. cristata has smooth ventral sides unlike the rough, pebble like

PAGE 15

15 underside of N. aquatica leaves (Figure 1 4 ) (Burks, 2002a). N. cor data has a smooth underside N. cristata displays nymphaeid morphology with the plant rooted in the submerged sediment and produces floating leaves with short petioles arising from long petiole like stems (Burks, 2002a; Sculthorpe, 1967). Rooted plants produce many leaves, and while most float on the surface of the water some leaves can remain submersed. The ability of N. cristata to grow in water 3 m deep or deeper is noteworth y because many other nymphaeid plants (rooted in the sediments with floating leaves) cannot grow at this depth (Sculthorpe, 1967) N. cristata is considered to be subdioecious in its native range since it has the ability to produce monoecious plants within a dioecious population (Burks, 2002a). Monoecious plants produce separate female and male flowers on the same plant. Recent evaluations of the floral biology of N. cristata suggest that populations in Florida are gynodioecious since the flowers of the pla nt are either pistillate, only containing female parts or hermaphroditic (perfect), containing both male and female parts (Tippery and Les, 2011). Despite the ability of N. cristata to produce numerous flowers, it has not been observed to produce viable se ed in the US (Burks, 2002a). N. cristata primarily reproduces vegetatively by production of ramets Ramets easily separate from the parent plant and can drift away in the water current to a new location to form a new colony, or sink to the bottom and remai n dormant until conditions are favorable for sprouting (Burks, 2002a). Little is known about the biology or longevity of these ramets; under what environmental conditions (light, sedim ent type or characteristics) these propagules are stimulated to sprout. Furthermore, it is not known what conditions are necessary for these propagules to remain quiescent after separation from the parent plant. Once established, N. cristata forms large

PAGE 16

16 root systems that act as carbohydrate storage organs and anchor the plant into the sediments (Figure 1 5). Habitat N. cristata is typically rooted in sediments, but is also capable of survival in a free floating form for a period of time with tuberous propagules (ramets) attached to the underside of the leaf (Burks, 2002a). The plant grows best in tropical to subtropical climate zones where it typically inhabits shallow areas of lakes, ponds, canals and areas of rivers with low current flow (Burks, 2002a). However, N. cristata similar to other species in this genus, has been ob served growing in water up to 3 m deep ( Chip Davis, personal communication, 2011; Sculthorpe, 1967). Current Distribution In its native range of India, Vietnam, Taiwan and southern provinces of China, N. cristata is found from 6 34 degrees North latitu de. In the western hemisphere this is equivalent to the area between northern South America (Columbia, Venezuela) to Chattanooga, Tennessee (Burks, 2002b). In Florida the plant ranges from the southern part of the State north towards Lake George (CAIP, 201 0). In the U.S. a rapidly expanding population is found as far north as Lake Marion in South Carolina (SC DNR, 2010). Recently, N. cristata has also been confirmed in Texas (Invasive, 2012), North Carolina ( Robert Richardson, personal communication, 2012) and Louisiana ( Alexander Perret, personal communication, 2012). This has implications in Florida, since this suggests that in terms of climate suitability there are not likely to be any regions in Florida where N. cristata could not thrive. A Weed risk ass essment performed by the USDA APHIS (2012) suggests that Plant Hardiness Zones 8 to 13 could provide conditions in which N. cristata could thrive. In addition to climate, N. cristata has been found growing in water 3 meters deep, which is of particular con cern for Florida, since the average depth of ma ny water bodies in the state is less than 3 meters.

PAGE 17

17 Invasive Potential Within its native range, N. cristata is considered a common weed of rice ( Oryza sativa L.) fields (Burks, 2002a). Reproduction and sprea d via vegetative reproduction is very evident, since floating leaves with the attached cluster of tubers easily detach from the branches and establish new stands. In Florida, N. cristata has been observed to spread rapidly in water up to 2 m deep. In Colli er County, FL, a stand was observed to invade and cover 3.6 ha in four weeks (Burks, 2002a). The plant forms dense mats of overlapping leaves and this canopy greatly reduces light penetration into the water column and surface mixing of water and air, reduc ing native submersed plant and algal photosynthesis and therefore reducing dissolved oxygen concentration in the water under the mat (Burks, 2002a). N. cristata was first discovered growing in Lake Marion, South Carolina in 2006 covering 8.1 ha and now cov ers 1,416 ha of the water surface ( Chip Davis, personal communication, June, 2012). N. cristata was recently moved from the Florida Exotic Pest Plant Council (FLEPPC) Category II list, a plant increasing in abundance or frequency but have not altered nati ve plant communities, to Category I, an invasive exotic plant that is altering native plant communities by displacing native species, changing community structures or ecological functions, or hybridizing natives (FLEPPC, 2009). Despite its reclassification to Category I, there is still a lack of peer reviewed literature regarding the invasion ecology, biology and management of the plant. In 2012, United States Department of Agriculture Animal and Plant Health Inspection Service (USDA APHIS) performed a weed risk assessment for N. cristata and determined that the plant is high risk, ranking especially high for both impact and establishment/spread potential (USDA APHIS, 2012)

PAGE 18

18 History and Origin of Yellow Floating Heart ( Nymphoides peltata ) The invasive yello w floating heart ( Nymphoides peltata (S.G. Gmel.) Kuntze) hereafter referred to as N. peltata is a native of Europe and temperate Asia (Smits et al. 1992). N. peltata has been present in the United States since at least 1882, when it was collected from a p ond in Massachusetts (Les and Mehrhoff, 1999). N. peltata has been offered for sale in the United States since 1891 (Countryman, 1970). Similar to N. cristata N. peltata was also introduced as an escape from cultivation (Les and Mehrhoff, 1999; Stuckey, 1 973). Identification and Biology N. peltata is a perennial aquatic plant with dark green, heart shaped floating leaves that rise from creeping, underwater rhizomes (Washington Noxious Weed Control Board, 2007). Flowers of the plant are bright yellow with 5 petals. N. peltata exhibits optimal growth, and is present at highest frequencies in well buffered (high pH) water systems (Smits et al, 1992). Studies conducted on plant growth in various sediments showed that when grown in both basic and acidic soils but in buffered water, plants could successfully grow and produce numerous leaves. However, when grown in the same sediment conditions but acidic water, plants were unable to produce leaves. These results suggest that the presence or absence of the plant i n different bodies of water is dependent on the water alkalinity not sediment alkalinity (Smits, 1992). It is not known how these conditions of water, sediment pH and quality effect the growth of N. cristata Floral structure of N. peltata is perfect (cont aining both male and female parts). However, unlike N. cristata N peltata is capable of self pollination and producing viable seed (Ornduff, 1970; Van Der Velde and Van Der Heijden, 1981). It is noteworthy that both of these species have very long flower ing seasons. A long flowering season is advantageous to N. peltata because of the ephemeral nature of the flower (once a flower opens it withers within one day). In order to maximize the possibility of pollination, flowers have to open individually over a long period of

PAGE 19

19 time rather than all at once (Van Der Velde and Van Der Heijden, 1981). Even though N. cristata does not self pollinate it produces flowers in a similar manner and N. peltata is also capable of clonal propagation via fragmentation (Larson, 2006). Habitat N. peltata grows in shallow areas of lakes, ponds and rivers where currents are slow (Nature Conservancy of Vermont, 2003). It grows best in water 1 m deep, but has occasionally been observed growing well on the muddy shores of lakes where changing water levels have left plants above the water surface (Countryman, 1970). Climatically, this plant was not thought to persist in warmer subtropical or tropical climates (USGS, 2001). Current Distribution N. peltata was initially considered to ha ve a low potential as an invasive in the southeast United States due to its pref erence of temperate zones (USGS 2001). However, a weed risk assessment conducted by USDA APHIS in 2012 suggests that N. peltata could persist in Plant Hardiness Zones 4 throug h 11, which includes Florida. Invasive Potential N. peltata has been shown to be competitive and aggressive when grown in competition with other species. N. peltata allocated most of its resources to producing above ground tissue and floating leaves when g rown in competition with a species of water chestnut ( Trapa bispinosa ) and Eurasian watermilfoil ( Myriophyllum spicatum ) (Wu et al, 2006). In this study it formed a canopy over the Eurasian watermilfoil and crowded out the water chestnut. Similarly, three native species, coontail ( Ceratophyllum demersum ), elodea ( Elodea canadensis ) and fan leaved crowfoot ( Ranunculus circinatus ) were grown under varying densities of N. peltata (Larson, 2006). At all coverage levels of N. peltata present (33 100%), native sp ecies experienced significant reductions in growth rate (Larson, 2006). These studies suggest that if N.

PAGE 20

20 peltata is introduced into a location where submersed vegetation grows without other floating vegetation, alterations to the community structure and bi odiversity are likely (Larson, 2006). Studies have not been conducted to assess the competitiveness of N. cristata or to evaluate its direct impact on native species. Preventative management of Nymphoides Prevention typically include measures such as quara ntines, bans on sale and importation, and creating public awareness in stopping the spread of a species or new infestations. Invasive species of Nymphoides became established after escape from cultivation, yet there are several preventive measures that can still be implemented to slow the spread. Another method used to prevent spread of invasive plant species is by creating boat and trailer cleaning stations at boat ramps. Since vegetative fragments often adhere to boat hulls and trailers, aquatic plants ar e easily transported and introduced to new locations in this manner. The best way to prevent further infestations is to limit Nymphoides use as a cultivated water garden plant (Les and Mehrhoff, 1999). This would require significant changes to current aqua rium trade practices since N. cristata is readily available from numerous on line sources. Cultural Control Cultural control is manipulating the habitat to achieve control of nuisance weeds. Some cultural methods include winter time water level drawdowns and altering light levels or other environmental factors (Bellaud, 2009). Cultural methods have not been successful on the Santee Cooper population of N. cristata W hen lake levels were lowered in the winter to expose plants to drying and brief periods of freezing temperatures in 2007 2008 rapid recovery was noted in the spring and no observable impact on the N. cristata populations was noted ( Chip Davis, personal communication, June 2011). Other options could include benthic barriers on the sediment surfa ce early in the growing season to control new regrowth of invasive species by

PAGE 21

21 blocking light and creating an anoxic zone beneath the mat (Bellaud, 2009). Nonetheless, b enthic ba rriers are very costly are difficult to deploy in large areas. Mechanical Cont rol Mechanical techniques are used to physically remove plant biomass from the impacted areas. Some examples of mechanical control are dredges, harvesters and cutter/shredder boats (Haller, 2009). Mechanical control is typically expensive, generally nonse lective and access to equipment is limited (Burks, 2002a). One attempted method of mechanical control for Nymphoides control is cutting the plant below the water surface; however this method proved ineffective (Middleton, 1990). Also, N. cristata can easi ly survive and recover from underwater fragments (Burks, 2002a). For N. peltata control mechanical harvesting has also failed with many locations becoming worse as a result of mechanical harvesting (Larson, 2006). W hile time and labor intensive h and remov al of the entire plant for newly established infestations has been proposed (Nature Conservancy of Vermont, 2003). This method of control is currently being attempted on Lake Okeechobee as the infestation is fairly localized and various chemical control s trategies have been ineffective. H and pulling may be effective in early detection and rapid response programs, or as a follow up to herbicide programs. Biological Control Biological weed control is implementing the use of a natural predator of the targe t plant to reduce populations of the weed. In North America, aquatic larvae of the native moth Paraponyx seminealis have been collected while feeding on the native N. aquatica Impacts of this moth on N. cristata are not known (Burks, 2002a). This moth has been observed to cut large pieces of leaf tissue from floating leaves for both food and shelter (Habeck, 1974). The young larvae feed as miners, eating holes in the leaf tissue. Where present the larvae are found in large numbers for most of the year, and can cause severe injury to floating heart reducing its factor as

PAGE 22

22 a weed (Habeck, 1974). To date, there is no evidence that Paraponyx seminealis will feed on N. cristata ; however, culture plants used for trials at the University of Florida Center for Aqua tic and Invasive Plants (CAIP) required consistent insect management to prevent loss of the culture due to insect damage from aquatic caterpillars ( Leif Willey; personal observation). The triploid grass carp ( Ctenopharyngodon idella ) is widely used for aq uatic plant control. Grass carp are general herbivores that consume almost any plant material with a few exceptions (Colle, 2009). Within its native range, grass carp will not consume N. cristata (Van Dyke et al. 1984) Stud ies performed in the early 1980 s in Florida, showed after grass carp had completely eliminated bladderwort ( Utricularia spp. ), umbrella grass ( Cyperus alternifolius ), and water shield ( Brasenia schreberi ) some decline in the native N. aquatica coverage was apparent (Van Dyke, Leslie, & N all, 1984) In the Santee Cooper reservoir system, over 850,000 grass carp have been stocked prior to 2012 ( Chip Davis, personal communication, 2012) and N.cristata continues to expand aggressively. Lake Fairview, near Orlando, Florida is also stocked with grass carp, yet intense chemical and physical management of N. cristata is regularly required since the plant was beginning to show signs of rapid expansion. Water bodies stocked with grass carp could be highly susceptible to N. cristata invasion due to t he grass carp facilitating an open niche through the consumption of preferred vegetation. N. cristata could spread rapidly due to lack of competition from other species of aquatic vegetation. The lack of host specific insect species for invasive Nymphoides control suggests that biological co ntrol is unlikely to be a management option in the near future. Chemical Control To date, reported success with registered aquatic herbicides on N. cristata has been limited, often anecdotal, and sometimes conflicting. Vermont reported in the 1970 s success using applications of 2,4 D in various formulations to control N. peltata in Lake Champlain

PAGE 23

23 (Countryman, 1970). In other areas herbicides have proven ineffective in controlling this plant in the long term. For N. cris tata control, various formulations and combinations of herbicides have been used, but with no long term success (Burks, 2002b). In Florida, Collier County Storm Water Management has obtained up to 4 months control using 2% glyphosate solution with a surfac tant, but numerous re treatments were required (Burks, 2002b). Herbicide treatments in South Carolina have resulted in control of the surface mats, but recovery is generally noted (SC DNR, 2010). A recent study on herbicide efficacy on N. cristata in south Florida showed endothall (dipotassium salt) was most effective, giving 98 100% control at 1.5 2.5 ppm (Puri and Haller, 2010). Field trials with endothall at concentrations of 2 3 ppm resulted in 80 90% control at 8 weeks post treatment (Puri and Haller, 2010). The longer term impacts of these treatments were not reported. Objectives Currently, 2 publications by Burks (2002a, b) constitute a large portion of the published literature pertaining to N. cristata Furthermore, t he information in these publica tions is mainly based on field observations and reports from managers. The placement of N. cristata on the FLEPPC Category I list given this lack of peer reviewed literature was a major influence for this thesis research. The objectives of this research we re to determine the growth responses of N. cristata from propagules exposed to various environmental conditions and to evaluate chemical control optio ns for management of this plant.

PAGE 24

24 Figure 1 1. Image of Nymphoides cristata growing in the Santee Coop er Reservoir. Photo from Chip Davis, Santee Cooper Analytical and Biological Services 2009. Figure 1 2 Bundle of tapered, tuberous roots of the ramet developing beneath floating leaves of crested floating heart ( Nymphoides cristata ). From University o f Florida Center for Aquatic and Invasive Plants (University of Florida, 2001). Available at http://plants.ifas.ufl.edu/node/291.

PAGE 25

25 Figure 1 3 Line drawing of a crested floating heart ( Nymphoides cristata ) flower, showing the erect fold of tissue that oc curs longitudinally along the center of each petal. From University of Florida Center for Aquatic and Invasive Plants (University of Florida, 2002). Available at http://plants.ifas.ufl.edu/node/291. Figure 1 4 Image comparing the native Nymphoides aquat ica leaf underside (right) showing rough texture and the exotic Nymphoides cristata leaf underside (left). From Chris Page, South Carolina Department of Natural Resources, Early Detection and Distribution Mapping System (2010). Available at http://invasive plantatlas.org/subject.html?sub=11616.

PAGE 26

26 Figure 1 5. Root system of an established Nymphoides cristata plant collected in January, 2011 from a South Florida Water Management District pond. Photo by Leif Willey, 2011.

PAGE 27

27 CHAPTER 2 EFFECTS OF AQUATIC H ERBICI DES ON CRESTED FLOAT ING HEART Introduction Crested floating heart ( Nymphoides cristata (Roxb.) Kuntze), hereafter referred to as N. cristata is an invasive aquatic weed native to Southeast Asia and introduced to North America through the water garden trade It is not known how long the plant may have been cultivated in the United States (Burks, 2002a). N. cristata is valued as an ornamental plant because of the expression of accessory pigments in the foliage, numerous small, white flowers, and long flowerin g season. It is often marketed as water snowflake because it can cover a water surface in tiny white flowers giving the appearance of snow. N. cristata has escaped from cultivation and exists in thriving populations in many waterways in Florida, South Caro lina, Texas (Center for Invasive Species and Ecosystem Health, 2012 ), North Carolina (Robert Richardson personal communication, August 2 012) and Louisiana (Alexander Perret personal communication, July 2012). N. cristata was first reported in Florida ou tside of ornamental culture in 1996 in Horseshoe Lake, Collier County, FL (Burks, 2002a). One of the largest N. cristata infestations has occurred in the 64,750 ha Santee Cooper reservoir system in South Carolina where the i nfestation now covers over 2,428 ha (6,000 ac) (Westbrooks, et al. 2012). The Santee Cooper SC infestation suggests that all water bodies in Florida are with a climate zone favorable for sustaining N. cristata The magnitude of an invasion is likely to be dependent on numerous water qua lity variables and lake morphology N. cristata has been observed growing in water from 0.6 to 3 m deep i n the Santee Cooper system (Chip Davis personal communication, July 2012). The plant can also survive extended periods in moist soil (Willey and Lang eland, 2011). Lakes in Florida

PAGE 28

28 generally tend to be shallow. Lake Okeechobee for example, despite its size (189,000 ha), has an average depth of only 3 m (SFWMD, 2012 ). In areas where it has become established, floating leaves of the plant form dense mats of overlapping leaves on the surface of the water, which interfere with boat traffic and recreational uses of the water. The mats shade the water column below reducing light availability to submersed native vegetation, lowering dissolved oxygen levels, a nd reducing water flow and aeration (Burks, 2002a). Studies on similar mat forming vegetation have shown that submersed macrophyte growth beneath these mats is significantly reduced (Janes et al. 1996). N. cristata reproduces and spreads by fragmentation, which can be caused by contact with boat motors, wave action and mechanical harvesting (Burks, 2002a). Spread is also facilitated through the production of clonal reproductive structures called ramets. Ramets develop beneath the floating leaves and protru de from the stems of the plant as a tuber cluster with several small leaves. These propagules easily separate from the parent plant and drift away or sink to the bottom (Burks, 2002a). In terms of potential management options for N. cristata triploid gra ss carp ( Ctenopharyngodon idella ) is a widely used biological control for many aquatic plants. However, previous studies have reported that grass carp do not consume N. cristata even when provided no other option (Van Dyke et al. 1984). A study conducted i n the native range of N. cristata also found that grass carp would not consume the plant (Singh et al. 1966). This presents a troublesome management issue in the Santee Coo per system because which is stocked with grass carp (109,000 in 2012 in addition to 750,000 previously stocked) to control hydrilla ( Hydrilla verticillata ) (Chip Davis personal communication, July 2012). Many lakes in Florida and other Southeastern reservoirs are stocked with grass carp for the purpose of hydrilla control; this could

PAGE 29

29 in turn create a niche for N. cristata colonization with reduced competition from hydrilla. Managers at the Santee Cooper system have also tried winter season drawdowns to expose plant parts to desiccation and potentially freezing temperatures but the plants re grew the following spring (Page, 2010). A study conducted in the native range of N. cristata suggests that mechanical harvesting would be ineffective because the plant was able to recover quickly from underwater clipping (Burks, 2002a; Middleton, 1990). Effective management plans are needed to prevent further spread and reduce current populations yet there has been very little published information on the use of herbicides as a management technique for N. cristata Reports from managers tend to be anecd otal and often conflicting. Current literature from Burks (2002b), states that a maximum of 4 months of control has been achieved when using foliar applications of a 2 % glyphosate solution with a surfactant Herbicide activity on N. cristata is not well u nderstood. It is often hypothesized that different formulations or methods of application may also influence efficacy of aquatic herbicides (Wersal and Madsen, 2010). Foliar applications are typically easier to make than subsurface applications and may als o be more economical (Wersal and Madsen, 2010). In an open system, when a herbicide is applied via a subsurface application, it immediately begins to diffuse away from the treatment zone to areas with lower herbicide concentration (Sprecher et al. 2002; Fo x et al. 2002). Due to diffusion and other forces such as water flow and herbicide degradation, it is not likely to remain in the treatment zone for long periods of time, which could require higher concentrations to be applied and repeat applications to ma intain specific concentrations. These issues illustrate the importance of assessing efficacy as a function of application method and potential exposure time for submersed applications The lack of a standard management option suggests that research is need ed to evaluate registered aquatic

PAGE 30

30 herbicides in order to improve management recommendations for control of N. cristata To address this deficiency 12 r egistered aquatic herbicides were screened to evaluate and compare activity on N. cristata Materials an d Methods Subsurface herbicide applications Experiments were conducted at the University of Florida, Center for Aquatic and Invasive Plants (CAIP) in Gainesville, FL in 2011 to 2012. N. cristata plants were collected in January 2011 from the Storm Water Tr eatment Areas (STA) in South Florida. Ramets were established in 1 L plastic containers that were filled with Margo Professional Topsoil 1 (92% sand, 4% silt, 4% clay) amended with fertilizer (Osmocote 15 9 12) 2 at 1g kg 1 of soil. Plants were cultured in 95 L tanks in a greenhouse from January until early March then moved into outdoor 1,000 L mesocosm tanks until ramets were produced. Ramets were collected from this culture and planted in 1 L pots to conduct herbicide studies. Ramets were allowed to grow u ntil leaves emerged at the surface and flower production was observed. When all plants had produced at least one flower treatments began. All registered aquatic herbicides were screened using both submersed and foliar applications as well as liquid and gra nular formulations (Table 2 1, 2 2). For the submersed applications, exposure times of 24 and 96 hours were evaluated. Current literature suggests that, depending on the rate of water exchange, size of treatment area in relation to water body size and othe r characteristics in a natural system the half life of the herbicide concentration may range from as low as a few hours (Poovey et al. 2004) to as long as a few weeks (Green et al. 1989; Simsiman and Chesters, 1975; Langeland and Warner, 1986). 1 Margo Garden Products. Folkston, GA 31537 2 The Scotts Company. Marysville, OH 43041

PAGE 31

31 Since the o bjective of this screening was to evaluate comparative efficacy, it was decided to focus on one short exposure time (24 h) and one long exposure (96 h) for contact herbicides. In contrast, f luridone and ALS inhibiting herbicides were tested for 3, 6 and 12 weeks under static exposure conditions in a separate greenhouse trial All trials were conducted in 95 L tanks. Eight herbicides and 3 combinations of herbicides were evaluated using submersed application techniques at 24 and 96 hour exposures (Table 2 1 2 2). Herbicides were teste d at maximum and half maximum label concentrations with the exception of carfentrazone which was tested at a single concentration. Plants were treated at the time of flowering. Submersed herbicide applications were made using a n adjustable pipette to inject the liquid herbicide into the water column. Granular herbicides were weighed to within 0.02 gram of the calculated amount using a digital scale (Denver Instrument APX 203) 3 then placed into the water, avoiding the floating leaves. Water samples were collected 1 day after treatment (DAT) from all treatments of 2,4 D, triclopyr, and endothall and analyzed using an ELISA test (SDIX RaPID Assay) 4 to confirm the herbicide concentration matched the target treatment rate. A small e lectric pump was used 24 and 96 hours after treatment (HAT) to remove the treated water from each tank and then the tanks were refilled with untreated well water as described by Wersal and Madsen (2010). Visual observations of phytotoxicity were recorded w eekly. Entire plants, including all live roots and foliage, were harvested 4 WAT and rinsed in untreated water to remove algae, sediment and dead tissue from the roots and foliage. Harvested plants were placed in labeled paper bags and dried in a drying ov en (76 C) for 1 week. 3 Denver Instruments. Arvada, CO 80004 4 Strategic Diagnostics Inc. Newark, DE 19713

PAGE 32

32 The first trial for 2,4 D, triclopyr, and a combination of the 2 herbicides was initiated June 17, 2011 and repeated July 12, 2011. The first trial for endothall, diquat and combinations of the herbicides as well as bispyribac was i nitiated on July 7, 2011 and repeated August 5, 2011. An additional evaluation of endothall was conducted on March 3, 2012 to evaluate the influence of cooler water temperatures (less than 20 C) on possible control. Treatment dry weight was compared to a pre treatment dry weight. P re treatment weights represent plant biomass the time of application and served as a reference to evaluate control for the herbicides. Th e pre treatment stage of growth was selected as the base for herbicide activity because the untreated control is continually growing. A herbicide could result in a reduction from the untreated control, but still have higher dry weight than the time of treatment. If at 4 WAT, in two separate trials, a herbicide resulted in a significantly lower dr y weight than the pre treatment sample, it was identified as having activity on the plant (net decrease in dry weight from tim e of treatment) and was recommended for additional evaluation. If a herbicide did not reduce dry weight below the pre treatment we ight (no change or net increase in dry weight from time of treatment), that herbicide was determined to have limited activity on N. cristata All studies were arranged as a complete randomized design with 3 replications of each treatment. Treatment effect s on dry weight were analyzed using ANOVA ( with 95% confidence intervals to determine differences among the treatments. Statistical analysis and graphical presentations of data were performed using SigmaPlot 11.0 5 Data were pooled when no differences were detected between trial s; however, data are presented separately for each trial in cases where data were different between trials. 5 Systat software, Inc. San Jose, CA 95110

PAGE 33

33 Foliar herbicide application Fourteen herbicides and combinations of herbicides were screened for efficacy following foliar applications (Table 2 1 2 2). Foliar applications were made using a CO 2 pressurized, single nozzle spray system at the time of flowering. A spray volume equivalent to 934 L ha 1 (100 gal ac 1 ) was used for all foliar treatments over an area of 0.185 m 2 Output pressure was regu lated at 83 to 103 kPa, which allowed for a consistent spray with minimal drift. Foliar application use rates are listed in Table 2 1. Tanks were drained following the foliar treatments and re filled at 24 HAT to remove any confounding issues associated wi th herbicide concentrations in the water column. Foliage was not rinsed. Triclopyr, 2,4 D, diquat and endothall applications were performed at the same time as the submersed applications. The first trials for the ALS herbicides were initiated on June 21, 2011 and repeated on July 19, 2011. Combinations of selected herbicides were initiated on July 3, 2011 and repeated August 5, 2011. Visual observations were recorded weekly after the herbicide treatments were applied. Whole plants, including all live roots and foliage, were harvested 4 WAT and rinsed in untreated water to remove any dead tissue and debris. Live tissue was placed in paper bags and dried at 76 C for 1 week. Statistical analyses were conducted as previously described for submersed treatments. Results and Discussion Subsurface herbicide application Triclopyr, 2, 4 D, 2, 4 D + triclopyr, endothall (dipotassium salt), bispyribac, carfentrazone, and flumioxazin did not reduce dry weight be low the pre treatment dry values at 4 WAT regardless of ex posure time, rate and formulation. Dry weight either remained constant or increased above the pre treatment reference levels at 4 WAT following treatment (Figure 2 1 to 2 4, 2 6, 2 7). For these 7 herbicide treatments, visual observations of injury symptom s were not

PAGE 34

34 indicative of herbicide efficacy because symptoms were transient and while noticeable within days of treatment, injury symptoms were generally short lived. ELISA tests performed 24 HAT of the auxin mimic herbicides showed that actual mean conce ntrations were within the target concentration. 2,4 D ranged from 2.25 to 2.60 mg ae L 1 for 2.5 mg ae L 1 target concentrations and 1.10 to 1.40 mg ae L 1 for 1.25 mg ae L 1 target concentrations. Triclopyr concentrations tested from 2.15 to 2.53 mg ae L 1 for target concentrations of 2.5 mg ae L 1 and 1.09 to 1.42 mg ae L 1 for target concentrations of 1.25 mg ae L 1 An ANOVA test of the triclopyr treatments found differences among the treatments (p<0.001) in both trials. No treatment reduced dry weigh t below the pre treatment reference (Figure 2 1 A, B). Furthermore, there were no differences found between subsurface and foliar applications or between liquid and granular formulations (Figures 2 1 A, B). No differences were found among the treatments of 2,4 D in the first trial (p=0.329) (Figure 2 2 A). Differences among treatments were found in the second trial (p<0.001) in which concentrations of 2.5 mg ae L 1 as both liquid and granular formulations outgrew the pre treatment reference (Figure 2 2 B), however, no treatments reduced dry weight below the pre treatment reference. Similarly, a combination of 2, 4 D and triclopyr did not reduce dry weight below the pretreatment reference (Figure 2 3 A, B). In the first trial no differences were found among t he treatments (p=0.429). Differences among treatments were found in the second trial (p<0.001), in which a concentration of 1.25 mg ae L 1 outgrew the pre treatment reference (Figure 2 3 B). There were also no differences found between the 24 and 96 hour e xposure times for these compounds. Auxin mimic type symptoms began to develop by 1 WAT, with noticeable epinasty of the stems, elongated flower stalks and leaf curling. These symptoms were transient and did not persist, by 2 WAT and 3 WAT the plants had co mpletely recovered (no visible symptoms).

PAGE 35

35 Cessation of growth by these compounds was not observed during either of the trials. Typically broadleaved aquatic plants are highly susceptible to herbicides with auxin mimic mode of action. Previous studies have found that subsurface applications of 2,4 D ester at 1.5 and 2.5 mg a.e. L 1 to fragrant waterlily ( Nymphaea odorata ) resulted in significantly less dry weig ht than the untreated controls (Glomski and Nelson, 2008). Spatterdock ( Nuphar advena ) dry weight w as also significantly reduced by the same rates of 2,4 D ester as well as 2.0 mg a.e. L 1 triclopyr amine by 6 WAT (Glomski and Nelson, 2008). Studies evaluating 2,4 D and triclopyr at rates of 0.5, 1, and 2 mg ai L 1 using 24 and 48 hour exposure times re ported that all rates controlled water chestnut ( Trapa natans ) (Poovey and Getsinger, 2007). Furthermore these compounds control submersed dicots such as Eurasian watermilfoil ( Myriophyllum spicatum ) (Netherland and Getsinger, 1992). It should be noted how ever, there are other dicotyledons such as cabomba ( Cabomba caroliniana ) that are more tolerant to auxin mimic herbicides (Bultemeier et al. 2009). Endothall is a contact herbicide that causes defoliation and tissue necrosis leading to plant death (Sprech er et al. 2002). Actual treatment concentrations were within the target concentration according to water samples analyzed at 24 HAT with an ELISA. 3.0 mg ae L 1 target concentrations ranged from 2.82 to 3.11 mg ae L 1 1.5 mg ae L 1 target concentrations r anged from 1.15 to 1.77 mg ae L 1 Differences were found among the treatments in both trials (p<0.001), however, neither of these formulations caused a reduction in dry weight equal to or below the pre treatment reference (Figure 2 4 A, B). In trial 1 (Fi gure 2 4 A) all 24 hour exposures outgrew the pre treatment reference, while all treatments outgrew the pre treatment reference in the second trial (Figure 2 4 B). There was a difference between exposure times in which the 96 hour exposure time caused an a verage of 55% and 50% greater reduction in dry

PAGE 36

36 weight than the 24 hour exposure time of the liquid submersed or granular treatments respectively (Figure 2 4 A, B). This suggests that increasing ex posure times beyond 96 hours would improve endothall efficac y. Symptoms developed slowly after treatment. One WAT isolated spots of foliar desiccation were observed. Two WAT foliage and stems began to show signs of necrosis and dropped off the water surface, however regrowth began between 2 and 3 WAT. Endothall has been used successfully in managing several submersed invasive aquatic plants such as hydrilla and Eurasian watermilfoil (Skogerboe and Getsinger, 2001). Previous studies have shown that plants with a similar growth form to N. cristata, such as spatterdock and fragrant waterlily were susceptible to endothall following 5 day exposure times at concentrations of 2 and 5 mg ai L 1 (Skogerboe & Getsinger, 2001). Recent work on N. cristata has shown that endothall (dipotassium salt) was most effective in static t ests resulting in 98 to 100% control using rates of 1.5 and 2.5 mg ai L 1 at 6 WAT (Puri and Haller, 2010). In the field, endothall (dipotassium salt) at rates of 2.0 to 3.0 mg ai L 1 provided 80 to 90 % control at 8WAT (Puri and Haller, 2010). To address the discrepancies between these findings and the findings of the current endothall (dipotassium salt) trials a n ad ditional trial was initiated August 5, 2011 to assess the efficacy of endothall under a 4 wk static exposure. Dry weight of the plant was red uced below that of a pre treatment reference; however, healthy regrowth was observed by 3WAT (Figure 2 5). These results suggest that endothall (dipotassium salt) can be effective in systems where extended exposures are possible. Bispyribac sodium is an ac etolactate synthesis (ALS) inhibiting herbicide that was recently registered for aquatic use. While d ifferences were notes among the treatments (p<0.001) b ispyribac sodium did not reduce dry weight below pre treatment levels 4 WAT regardless of exposure t ime (Figure 2 6). There was an exposure response observed with the 96 hour exposure

PAGE 37

37 time resulting in a 30% greater reduction in dry weight than the 24 hour exposure time (p<0.001) (Figure 2 6). Flower production ceased at 1 WAT, but by 2 WAT foliage began to turn red in color and become necrotic, however plants quickly recovered and new growth was visible by 3 WAT. Bispyribac sodium has been found to have activity on several emergent aquatic plants (Koschnick et al. 2007). The ALS herbicides typically have very slow activity and will likely require much longer exposure to evaluate efficacy; however, I wanted to evaluate at least one ALS herbicide using a protocol similar to the auxin mimic and contact herbicides. Carfentrazone and flumioxazin are protoporp hyrinogen oxidase (PPO) inhibitors that impact chlorophyll synthesis and cause cell membrane leakage. Both are fast acting, contact herbicides (Senseman, 2007) The half life of flumioxazin ranges from several hours to a few days depending on pH (Mudge and Haller, 2010). The pH of the water at the time of application was 6.8 and this should have provided the range of a 24 hour product half life based on estimation of pH dependent hydrolysis Differences were found among the treatments (p<0.001) although no treatment reduced dry weight below the pre treatment reference. Carfentrazone treated plants outgrew the pre treatment reference as well as the 0.2 mg ai L 1 concentration of flumioxazin (Figure 2 7) There was an observed rate response noted between the tw o concentrations of flumioxazin in which the 0.4 mg ai L 1 concentration resulted in a 66 % reduction in dry weight from the 0.2 mg ai L 1 concentration (Figure 2 7). One WAT the foliage of the flumioxazin treated plants had turned necrotic and had begun t o drop away from the surface, however rapid regrowth was observed 2 WAT. Similar results were observ ed with carfentrazone. B oth herbic ides have provided good activity on several floating aquatic plants (Mudge and Haller, 2010; Koschnick et al. 2 004), yet a ctivity on N. cristata was limited to initial injury followed by rapid regrowth.

PAGE 38

38 Diquat is a cell membrane disruptor that interrupts photosynthetic electron transport and diverts electrons to oxygen resulting in the formation of radical oxygen which caus es cell membrane damage and death (WSSA 2007) Diquat readily binds to organic matter and sediment particles which can impact activity in turbid water and also likely explains the lack of a granular formulation. Water clarity was optimal for diquat activi ty in these trials. Differences were found among treatments in both trials (p<0.001). None of the treatments reduced dry weight below the pre treatment reference in trial 1, but all treatments were less than the untr eated control (Figure 2 8 A, B) A rat e response was observed during the second trial in which the maximum labeled concentration (0.38 mg L 1 ) yielded 47% less dry weight than the half maximum concentration (Figure 2 8 B). There was no observed response between the 24 and 96 hour exposure time s at the maximum label rate. Diquat treated plants rapidly developed symptoms. One WAT, foliage was necrotic and stems had dropped away from the water surface, however new leaves began to re grow by 2 WAT. During re growth, new foliage would open on the wa ter surface and would become chlorotic, and then would either die or very slowly recover. Plants began to recover between 3 and 4 WAT, but re growth remained limited with some new foliage remaining chlorotic near the midvei n of the leaf. Diquat applied as a submersed application strategy was shown to be effective in controlling other invasive aquatic species with emergent foliage such as parrotfeather even when exposed to short exposure times (Wersal and Madsen, 2010). The amine salt formulation of endotha ll applied into the water column was the most effective herbicide evaluated. ANOVA determined there were differences among the treatments (p<0.001). Liquid subsurface treatments caused a 100% reduction in dry weight compared to the pre treatment reference (Figure 2 9 A, B). Actual treatment concentrations for the target

PAGE 39

39 concentration of 0.5 mg ae L 1 ranged from 0.30 to 0.62 mg ae L 1 and target concentrations of 0.25 mg ae L 1 ranged from 0.11 to 0.35 mg ae L 1 Treated plants showed necrosis by 24 HAT. On e WAT all plant material had dropped away from the water surface and by 4 WAT there was no observed regrowth. There was no observed rate or time response si nce both concentrations and exposures tested resulted in 100% control of the plant. Granular applica tions of endothall were less effective than liquid applications and were highly variable. There was a rate response to the granular herbicide in the first trial (Figure 2 9 A) in which the maximum concentration resulted in 50% greater reduction than the ha lf maximum concentration with a 24 hour CET. The first round of Endothall (amine salt) treatments occurred in June and July when the water temperatures were high. It is thought that uptake of this herbicide may be increased under higher temperatures (Hall er and Sutton, 1973). While endothall (amine salt) was the only herbicide to result in 100% control at both rates tested, there are some issues associated with use of this herbicide. While the amine salt is 2 to 3 times more active than the dipotassium sal t on submersed plants, activity on algae, potential rapid drops in dissolved oxygen, and toxicity to fish are all significant concerns (Sprecher et al. 2002). Combining herbicides could have a synergistic effect (two herbicides together result in greater e fficacy than either of the herbicides used singly). Other advantages of herbicide combinations may be reduced herbicide quantity used and greater selectivity (Pennington et al. 2001). Analysis of a combination of diquat + endothall (dipotassium salt) found differences among treatments (P<0.001). Dry weight was reduced below that of the pre treatment reference (Figure 2 10 ) and exposure time did not affect efficacy of the treatment. The combination of 2,4 D + triclopyr, did not result in dry weight less than the pre treatment and results were similar to 2,4 D and triclopyr used alone (Figures 2 1 A,B; 2 2 A, B; 2 3 A, B).

PAGE 40

40 In summary, the screening procedure identified endothall amine salt as the most active compound followed by diquat. The lack of activity of the dipotassium salt of endothall was unexpected since previous research suggested good activity for this herbicide (Puri and Haller, 2010). We were able to achieve improved control with the dipotassium salt of endothall under static conditions; however, many sites and treatment scenarios will not provide this level of exposure time. Foliar herbicide application Generally, foliar herbicide applications offer a more economical treatment option when compared to subsurface applications (Wersal & Madsen, 20 10), since the herbicides are placed directly on the foliage and not subject to dilution. Foliar herbicide trials were performed simultaneously with the sub surface applications to determine any differences between application strategies. Foliar applicati ons of triclopyr di d not reduce dry weight (Figure 2 1A, B) and similar results were observed with 2,4 D (Figure 2 2 A, B). Typical auxin mimic symptoms began to develop within 1 WAT but the plants quickly recovered by 2 WAT and continued growing until the study was harvested. There were also no differences found between foliar and subsurface applications of these herbicides. Studies performed on other broadleaf, emersed plants like parrotfeather ( Myriophyllum aquaticum ), have shown that 2, 4 D is more effe ctive when foliar applied (Wersal and Madsen, 2010). Other studies examining effects of foliar applications have shown that rates of 2,4 D amine up to 4.48 kg ha 1 cause symptom development in spatterdock, but did not result in death of the plant tissue (H anlon and Haller, 1990). Waterlily has also been reported to be sensitive to triclopyr amine via foliar application, but no details were given pertaining to rates (Glomski and Nelson, 2008; Langeland et al. 1993; Gettys et al. 2009).

PAGE 41

41 Foliar applications of diquat and endothall (dipotassium salt) did not reduce dry weight. Endothall (dipotassium salt) showed similar results (Figure 2 4 A, B and 2 8 A, B). Diquat treated foliage died back from the surface quickly and regrowth began 2 WAT. Regrown foliage w ould reach the water surface and become chlorotic. This chlorotic regrowth suggests that diquat may be having a temporary impact on the new foliage. While the growth was initially slow, normal growth was noted by 3 WAT. When applied to foliage at rates of 2.3 L ha 1 diquat reduced biomass of duckweed ( Lemna minor ), water hyacinth ( Eichhornia crassipes ) and water lettuce ( Pistia stratiotes ) greater than 90% (Langeland et al. 2002). Alligator weed ( Alternanthera philoxeroides ) has also shown similar results to N. cristata When sprayed with diquat at rates of 5.6 and 22.4 kg ha 1 alligator weed showed complete foliar control within the first 2 weeks, but by 4 WAT only achieved 80 and 89 % control and by 8 WAT 45 and 60 % control respectively (Blackburn, 1963 ). Endothall (dipotassium salt) treated plants began to show symptoms 1 WAT as bronzed spots on the foliage. These damaged areas did not expand and the foliage never became totally necrotic or fell from the water surface. Diquat applied to the foliage was less effective than sub surface applications, while endothall was more effective when foliar applied than 24 hour sub surface applications but less effective than 96 hour applications (Figures 2 4 A and B and 2 8 A and B). Diquat and endothall have also be en evaluated against other species of floating vegetation. Giant salvinia ( Salvinia molesta ), when treated with endothall at 5.04 kg ha 1 and a surfactant resulted in 80% control 4 WAT and when treated with diquat at 2.24 and 1.12 kg ha 1 resulted in 100 a nd 98% control respectively (Nelson et al. 2001). Four amino acid inhibiting herbicides were also screened for foliar application. An ANOVA test of the treatments determined that there were differences among the treatments (p<0.001). Three of the herbicid es (imazamox, imazapyr and penoxsulam) are ALS inhibitors,

PAGE 42

42 while the fourth (glyphosate) is an EPSP synthase inhibitor. Glyphosate and penoxsulam were largely ineffective compared to imazamox and imazapyr which reduced dry weight below the pre treatment re ference by 83% and 84% respectively (Figure 2 11 ). All herbicides resulted in cessation of growth by 1 WAT. Two WAT foliage from imazamox and imazapyr treated plants was not present on the surface while live foliage was still present on glyphosate and peno xsulam treated plants. No regrowth had occurred 4 WAT on imazamox and imazapyr treated plants suggesting the herbicide had translocated into the root system; however, the root systems were still intact upon harvesting. Penoxsulam resulted in no reduction i n dry weight compared to the pre treatment reference. However, when plants were harvested it was noticed that even though shoot material did not die off, the root system sustained severe damage and was almost entirely dead tissue. Glyphosate treated plants began to show signs of recovery 4 WAT with appearance of new flowers. The result for glyphosate is interesting since it is considered a broad spectrum herbicide, yet seemed to have little impact on this plant. Other emergent plants such as water lily, spa tterdock, and lotus ( Nelumbo lutea ) are effectively controlled with glyphosate. Spatterdock and water lilies have been reported to be controlled at rates of 2.5 kg ai ha 1 (Baird et al., 1983; Seddon, 1981). Lotus is a third emergent plant that is controlle d with glyphosate using rates of 2.4 kg ai ha 1 with little to no regrowth 8 WAT ( Nathan Eddy, personal communication, 2012). Numerous applicators include glyphosate in combination with imazapyr as part of their current treatment regime for N. cristata (Bu rks, 2002). Combinations of foliar applied herbi cides (Table 2 2 and Figure 2 12 ) were analyze d via ANOVA and no differences were detected among treatments (P<0.001). All combinations tested reduced dry weight below the pre treatment reference however r esu lts were comparable to at least one of the two herbicides used. The combination of triclopyr + penoxsulam reduced dry

PAGE 43

43 weight 43% bel ow the pre treatment reference and t his effect was not observed when triclopyr or penoxsulam were tested alone (Figure 2 1, 2 11, 2 12). Foliar combinations have been used in the field on several other emergent plants. A combination of triclopyr and imazapyr 1.75 and 0.15 kg ai ha 1 WAT ( Nathan E ddy, personal communication, 2012). This same combination was used in a pond treatment of yellow floating heart ( Nymphoides peltata ), but only provided 3 weeks of control ( Nathan Eddy, personal communication, 2012). In deep water treatments at the Santee C ooper Reservoir a combination of glyphosate (4.8 kg ai ha 1 ) and imazamox (2.4 kg ai ha 1 ) has resulted in good control with limited regrowth at 8 WAT ( Chip Davis, personal communication, 2012). The screening of foliar applie d herbicides has identified im azamox and imazapyr as having good activity on N. cristata When making foliar applications of herbicides, translocation plays a critical role in the efficacy of a treatment. In the Santee Cooper system, it has been noted that post treatment regrowth is le ss at depths less than1.5 m ( Chip Davis, personal communication, 2011). Furthermore, managers in south Florida have reported greater control in shallower, quiescent areas (Burks, 2002). As a result of these observations it is hypothesized that water depth may impact the ability of herbicides to translocate from foliage to roots ( Chip Davis, personal communication 2011). N. cristata has been found to grow in up to 3 m of water, if water depth (stem length) has an impact on the ability of a herbicide to trans locate, it may be very difficult to manage this plant in deep water when using foliar treatment strategies This screening has identified 4 herbicides; liquid endothall (amine salt), imazamox, imazapyr and diquat (high concentration and long CET) that wer e able to reduce dry weight of N. cristata below the pre treatment reference. In addition, there were no differences between liquid and granular formulations with the exception of endothall (amine salt), in which the

PAGE 44

44 granular formulation was less effective than the liquid. These findings also suggest that in most cases foliar or sub surface applications have no influence on treatment efficacy and that foliar applications of a herbicide result in similar efficacy as would be observed using subsurface applica tions of the same herbicide. This result is similar to that reported by Wersal and Madsen, (2010). Management of N. cristata will continue to remain a challenge, due to the small number of herbicides that show good activity on this invasive plant.

PAGE 45

45 Table 2 1. Single herbicide subsurface treatment concentrations and foliar rates, exposure times and application method used in the herbicide screening on N. cristata Spray volume of foliar treatments was equivalent to 934 L ha 1. All foliar appl ications made with methylated seed oil as surfactant at a rate of 0.25 % vol : vol. Herbicide treatment Exposure time (hr) Application method a 2,4 D (amine) 2.5 and 1.25 mg ae L 1 24,96 L Diquat 0.370 and 0.185 mg ai L 1 24,96 L Flumioxazin b 0.4 and 0. 2 mg ai L 1 24,96 L Bispyribac sodium b 0.03 mg ai L 1 24,96 L Triclopyr (amine) 2.5 and 1.25 mg ae L 1 24,96 L,G Endothall (dipotassium) 3.0 and 1.5 mg ae L 1 24,96 L,G Endothall (amine) 0.5 and 0.25 mg ae L 1 24,96 L,G 2,4 D (ester) 2.5 and 1.25 mg a e L 1 24,96 G Imazapyr 1.2 kg ae ha 1 F Penoxsulam 0.1 kg ai ha 1 F 2,4 D (amine) 2.2 kg ae ha 1 F Diquat 2.2 kg ai ha 1 F Glyphosate 2.4 kg ae ha 1 F Endothall (dipotassium) 2.5 kg ae ha 1 F Imazamox 1.2 kg ae ha 1 F Triclopyr (amine ) 3.5 kg ae ha 1 F a Abbreviations: L, liquid subsurface; G, granular subsurface; F, foliar. b Liquid stock formulations were prepared from granular and wettable powder formulations for submersed liquid applications.

PAGE 46

46 Table 2 2. Herbicide combina tion subsurface treatment concentrations and foliar rates, exposure times and application method used in the herbicide screening on N. cristata. Spray volume of foliar treatments was equivalent to 934 L ha 1. All foliar applications made with methylated se ed oil as surfactant at a rate of 0.25% vol : vol. Herbicide treatment Exposure time (hr) Application method a 2,4 D (amine)1.94 and 0.98 mg ae L 1 + triclopyr (amine) 0.55 and 0.27 mg ae L 1 24,96 L,G Diquat 0.2 mg ai L 1 + endothall (dipotassium salt) 0.8 mg ae L 1 24,96 L Glyphosate 2.4 kg ae ha 1 + imazapyr 0.6 kg ae ha 1 F Glyphosate 2.4 kg ae ha 1 + imazamox 0.6 kg ae ha 1 F Triclopyr (amine) 1.7 kg ae ha 1 + imazapyr 0.6 kg ae ha 1 F Triclopyr (amine) 0.44 kg ae ha 1 + penoxsulam 0.07 kg ae ha 1 F Diquat 0.5 kg ai kg ha 1 + imazamox 0.9 kg ae ha 1 F Endothall 0.6 kg ae ha 1 + imazamox 0.9 kg ae ha 1 F a Abbreviations: L, liquid subsurface; G, granular subsurface; F, foliar.

PAGE 47

47 Figure 2 1. Combined dry weight of live roots and foliage of N. cristata at 4 WAT in response to triclopyr at 24 and 96 h exposures in trial 1 (A) and in trial 2 (B). PT = pre treatment reference, CON = untreated control, L = liquid subsurface, G = granular subsurf ace and F = foliar rate in kg ae ha 1 Numbers in front of L and G are conce ntrations in mg ae L 1 Error bars represent 95% confidence interval s

PAGE 48

48 Figure 2 2. Combined dry weight of live roots and foliage of N. cr istata at 4 WAT in response to 2,4 D for 24 and 96 h exposures in trial 1 (A) and in trial 2 (B). PT = pre treatment reference, CON = untreated control, L = liquid subsurface, G = granular subsurface and F = foliar rate in kg ae ha 1 Numbers in front of L and G are concentrations in mg ae L 1 Error bars represent 95% confidence interval.

PAGE 49

49 Figure 2 3. Combined dry weight of live roots and foliage of N. cristata 4 WAT in response to a combination of 2,4 D (amine) + triclopyr (amine) at 24 and 96 h exposures in trial 1 (A) and in trial 2 (B). PT = pre treatment reference, CON = untreated control, L = liquid subsurface, G = granular subsurface. Numbers in front of L and G are concentrations in mg ae L 1 2.5 concentrat ion is 1.94 mg ae L 1 2,4 D + 0.55 mg ae L 1 triclopyr. 1.25 concentration is 0.98 mg ae L 1 2,4 D + 0.27 mg ae L 1 triclopyr. Error bars represent 95% confidence interval.

PAGE 50

50 Figure 2 4. Combined dry weight of live roots and foliage of N. cristata 4 WAT in response to endothall (dipotassium salt) at 24 and 96 h exposures in trial 1 (A) and in trial 2 (B). PT = pre treatment reference, CON = untreated control, L = liquid subsurface, G = granular subsurface and F = fol iar rate in kg ae ha 1 Numbers in front of L and G are concentrations in mg ae L 1 Error bars repr esent 95% confidence interval

PAGE 51

51 Figure 2 5. Combined dry weight of live roots and foliage of N. cristata 4 WAT in re sponse to dipotassium salt endothall for 4 weeks. PT= pre treatment reference, CON = untreated control, G = granular subsurface, L = liquid subsurface. Concentrations are 4 mg L 1 ae for both liquid and granular. Different letters indicate significant diff erences. Error bars represent 95% confidence interval. Figure 2 6. Combined dry weight of live roots and foliage of N. cristata 4 WAT in response to bispyribac sodium (0.03 mg ai L 1 at 24 and 96 h exposures, both trials combined. PT = pre treatment reference, CON = untreated control, L = liquid subsurface. Different letters indicate differences between herbicide treatments. Error bars represent 95% confidence interval.

PAGE 52

52 Figu re 2 7. Combined dry weight of live roots and foliage of N. cristata 4 WAT in response to carfentrazone ethyl (C) and flumioxazin (F) at 24 and 96 h exposures, both trials combined. PT = pre treatment reference, CON = untreated control, L = liquid subsurf ace. C oncentrations are in mg ai L 1 Different letters indicate differences between herbicide treatments. Error bars represent 95% confidence interval.

PAGE 53

53 Figure 2 8. Combined dry weight of live roots and foliage of N. cristata 4 WAT in response to diquat at 24 and 96 h exposures in trial 1 (A) and in trial 2 (B). PT = pre treatment reference, CON = untreated control, L = liquid subsurface and F = foliar rate in kg ae ha 1 Liquid subsurface concentrations are in mg ai L 1 Error bars represent 95% confidence interval.

PAGE 54

54 Figure 2 9. Combined dry weight of live roots and foliage of N. cristata 4 WAT in response to endothall (amine salt) for 24 and 96 h exposures in tri al 1 (A) and in trial 2 (B). PT = pre treatment reference, CON = untreated control, L = liquid subsurface, G = granular subsurface. L and G subsurface concentrations are in mg ae L 1 Error bars represent 95% confidence interval.

PAGE 55

55 Figure 2 10 Combined dry weight of live roots and foliage of N. cristata 4 WAT in response to a combination of diquat (0.2 mg ai L 1 ) + dipotassium salt endothall (0.8 mg ae L 1 ) at 24 and 96 hours. PT = pre treatment reference, CON = untre ated control, Diq/endo = diquat + endothall. Error bars represent 95% confidence interval.

PAGE 56

56 Figure 2 11 Combined dry weight of live roots and foliage of N. cristata 4 WAT in response to foliar applied amino ac id inhibiting herbicides. GLY = glyphosate, IM = imazamox, IP = imazapyr, PEN = penoxsulam. Numbers represent treatment rate in kg ai ha 1 Error bars represent 95% confidence intervals. Figure 2 1 2 Combined dry we ight of live roots and foliage of N. cristata 4 WAT in response to foliar applications of selected herbicide combinations. GLY = glyphosate, IP = imazapyr, IM = imazamox, TRI = triclopyr, PEN = penoxsulam, DIQ = diquat, END = endothall (dipotassium salt). Error bars represent 95% confidence intervals. Herbicide rates are listed in Table 2 2.

PAGE 57

57 CHAPTER 3 THE INFLUENCE OF SED IMENT TYPE AND FERTI LITY ON THE GROWTH O F CRESTED FLOATING HEA RT Introduction Crested floating heart ( Nymphoides cristata (Roxb.) Ku ntze), hereafter referred to as N. cristata is an invasive aquatic plant native to Southeast Asia that has recently become established in the southeastern United States including Florida (Burks, 2002). N. cristata grows rooted in the submersed sediments in water up to 3 m deep in lakes, ponds, reservoirs and canals; however it can persist in the absence of standing water given sufficient soil moisture (Willey and Langeland, 2011). Plants of the nymphaeid growth form (rooted in sediments with floating or eme rgent leaves) such as N. cristata rely on the submersed sediments to provide many of the nutrients required for growth (Twilley et al. 1977). Nutrients are only absorbed from the water column when leaves are in the submerged stage; even then nutrient absor ption is highest in the roots (Twilley et al. 1977). It is unknown what characteristics of the sediments allow N. cristata to thrive or if there are certain conditions that may inhibit growth and spread in some sediments, but s ome aquatic plant nurseries suggest planting in enriched sediments that are mildly acidic to mildly basic (Anonymous, 2012). It is known that many plants grow better in certain substrates than others and nutrient availability can play a major role in plant invasions (Van et al. 1999) Research on the invasive plant yellow floating heart ( Nymphoides peltata ) suggests that the sediment pH was not as important to its growth compared to water pH, due to the observation that N. peltata grew equally well in basic and acidic sediments when t he pH of the water was high (Smits et al. 1992). The objective of this study was to determine the influences of sediment type, nutrient availability and the interaction of these two factors on the growth of N. cristata when established from ramets.

PAGE 58

58 Materi als and Methods Study 1 M ature N. cristata plants were collected from the storm water treatment areas (STAs) in South Florida in January 2011 Sediment/growth studies were conducted at the University of Florida Center for Aquatic and Invasive Plants (CAIP) Ramets and root crowns were planted in 1 L pots filled with a commercial potting soil mixture with 7% organic matter and a mineral content of 92% sand, 4% silt and 4% clay. Pots were placed into 95 L tanks inside a closed greenhouse for 2 months to initi ate growth in January through February 2011. In March 2011, plants were then moved into outdoor 1,000 L mesocosm tanks at the time of prolific production of ramets. Ramets were collected and used fo r these studies. The first study was conducted May 15 thro ugh July 12, 2011 and repeated August 23 through October 4, 2011. In the first study the effects of 3 sediment types (sand, muck and potting soil) on the growth of N. cristata ramets were evaluated. A sand and muck sedim ent were used because they are repr esentative of the two ends of the substrate spectrum common lake sediments in Florida lakes The potting soil was used because it had been used to grow N. cristata stock for other studies and it has been widely used in other mesocosm trials Sediments were collected from lakes with the exception of the commercial potting soil. Sand sediment was collected from West Lake Tohopekaliga (Toho) near Kissimmee, FL and muck (organic) sediment was collected from Orange Lake (Orange) located Southwest of Gainesville, FL (Table 3 1). Each sediment type was amended with 2 rates of fertilizer (Osmocote 15 9 12); 1, or 4 g kg 1 dry sediment (Mony et al. 2007) and a non amended control for each sediment type was also used. One L pots were filled with sediment and then fer tilizer was mixed into the soil and a single ramet was planted 0.6 to 1.2 cm into the sediment. Nine pots per sediment type were used as unplanted sediment controls (3 per fertilizer rate) to compare nutrient content of planted and unplanted sediments

PAGE 59

59 Pot s were randomly placed in 12 uncovered 1,000 L mesocosms and whole plants including roots and foliage were harvested and rinsed with untreated well water to remove sediments and dead plant tissue at 6 weeks after planting. The number of floating leaves was counted at harvest and plants were bagged and placed in a drying oven for one week at 76 C to determine dry weights of each plant. Sediment was saved for additional nutrient analysis and after harvesting samples were dried and sent to the Analytical Resea rch Lab (ARL) at the University of Florida in Gainesville, FL to be evaluated for organic matter content (loss on ignition), total Kjedahl nitrogen, phosphorus (Mehlich 1 test) and particle size composition. All treatments were arranged as a completely ran dom ized design with 5 replicates. SigmaPlot 11.0 statistical analysis software was used to analyze data. Data were analyzed using a 2 and the interaction of these factors on the number o f floating leaves produced and dry weight. To test for differences among fertilizer treatments within a soil type and for differences among the hoc Fishers protected LSD test was u sed to separate the treatment means. For the 6 week harvest, dry weight data were log transformed to meet the equal var iance assumption for ANOVA. Non transformed data are presented in the results. ANOVA tables from this study can be found in the appendix Study 2 In addition to the 3 sediments described above a clay based sediment collected near Lewisville, TX was included in the repeat of the study (Table 3 1). The clay sediment has been used in numerous mesocosm trials with a wide range of aquatic speci es. Methods of planting, harvest and data analysis remained the same as those described in study 1.

PAGE 60

60 Study 3 An third study was performed to evaluate the effects of pH on growth using builders sand with a base pH ranging from 6.6 to 6.8. Hydrated lime wi th a calcium carbonate equivalent (CCE) of 110% was added at 6 rates in addition to a control (Table 3 3). Lime was thoroughly mixed with the sand and fertilized with Osmocote 15 9 12 at 1g kg 1 dry sediment. Ramets were planted and allowed to grow for 4 weeks then whole plants were harvested and dried in an oven for 1 week at 76 C then weighed. Sediment was saved for pH analysis following the procedure described by Batjes (1995). Data were analyzed using linear regression (Y= a +bx); where Y is the depend ent (response) variable, a= Y intercept (value of Y when x=0), b= slope of the line, x= independent (explanatory) variable. Results and Discussion Study 1 The 6 week harvest data showed a significant interaction between soil type and increased fertility w ith respect to the number of floating leaves and increase in dry we ight (p<0.001; Table A 1, A 2 ). Leaf production and dry weight increased with increasing fertilizer rates across all sediment types (Figure 3 1 A, B) at the 6 week harvest. Growth, as measu red by changes in dry weight remained very limited in the non fertilized sand and muck sediments through 6 WAT. Plants grown in sand at the highest fertilizer rate produced 2 times more dry weight than muck and potting soil (Figure 3 1 B). Furthermore, wit hin the sand sediment biomass increased 16 fold from fertilizer rates of 0 to 1 g kg 1 and an additional 2.7 fold from the 1 to 4 g kg 1 fertilizer rates demonstrating the importance of nutrients in the sand sediment (Figure 3 1 B). Study 2 The 6 week har vest data showed a significant interaction between soil type and increased fertility with respect to the number of float ing leaves (p<0.001; Table B 1 ) and increase in dry

PAGE 61

61 weight (p<0.001; Table B 2 ). Plants at this harvest had prolific growth and exhibite d a clear and strong response to increasing fertilizer rates in the sand and muck sediments (Figure 3 2 A, B). Plants grown in sand sediment increased dry weight by 20 fold compared to t he non amended treatment and were double the dry weight of the 1 g kg 1 treatment. Increased dry weight through increasing fertilizer treatments was also observed in the muck sediment. In contrast plants grown in potting soil and clay sediments exhibited poor growth and a weak but varied response to increasing fertilizer (F igure 3 2 A, B). The first 2 studies provides important initial information regarding the potential role of substrate type and fertility in supporting growth of N. cristata The results suggest that nutrient availability plays a key role in the ability of this plant to grow in two different sediment types (sand and muck) but were not a factor in other sediment types (potting soil and clay). This finding seems intuitive because a plant is expected to grow better with additional nutrients ; however, t he compar atively poor growth observed in the clay sediment and potting soil suggests factors other than nutrients play a key role in N. cristata growth. This study has shown that sand and muck sediments were capable of supporting dense growth of N. cristata althoug h both of these sediments differ greatly in organic matter content and mineral composition (Table 3 1). The ability of N. cristata to grow well in these two sediment types, despite their physical differences suggests that a wide range of sediment textures can support the growth of this plant. It is important to note that growth in clay potting soil sediment, while reduced produced healthy, but less robust plants in these trials. There were no differences in macronutrient content between blank (no plants g rown) and 6 week harvest sediment samples within sand, potting soil and clay sediments indicating ample macronutrients remained available in all sediments at 6 weeks to continue to support plant

PAGE 62

62 growth. An exception to this is phosphorus (P) in the muck se diment in which P content was 16 to 342 times low er in the blank pots than in those with plants (Table 3 2). A possible explanation for this is that the presence of N. cristata actually enhances P retention in the sediment. One study conducted on Lake Mich igan found that total P was higher in areas where vascular aquatic plants were present and when tested in situ found that in the absence of vascular plants P was released in greater amounts to the water (Jaynes and Carpenter, 1986). In aquatic sediments P is generally not limiting since the sediment is the primary source for this macronutrient (Barko et al. 1991). While P detection was very low in the clay sediment, it should be noted that the clay sediments received the same fertilizer regime as the other sediments and that P detection can be very difficult in clay sediments (Gatiboni et al. 2010). Given that N.cristata growth was also greatly reduced in potting soil (with much higher concentrations of P), a significant role for P limitation on N. cristata growth is not indicated (Table 3 2). It is interesting to note that nitrogen (N) levels were much higher in clay sediments tha n sand sediments (Table 3 2) but this did not result in increased plant growth in the clay sediment. Study 3 The pH of sand and muck sediments ranged from 5.8 to 6.4 while the pH of potting soil and clay sediments ranged from 7.8 to 8.3. In the follow up study using builders sand p lant dry weight regressed against sediment pH showed a negative growth response (slope coefficient 0 .334; p<0.001) of N. cristata to increasing pH (Figure 3 3). Some plants are very sensitive to pH and have specific requirements for optimum growth. Another plant in the Nymphoides genus, N. peltata shows little apparent pH preference and instead growth is more dependent on water pH (Smits et al. 1992). In Southeast Asia, the native range of N. cristata a predominant wetland soil type is the gleysol, an acidic, fertile soil (FAO, 2012).

PAGE 63

63 These data suggest N. cristata may grow best in acidic, fertile sedim ents which support rapid growth and establishment; however, it is shown to be capable of growth in a wide variety of sediment types and in various levels of nutrient availability. This information could be useful to predict which lakes may be more or less susceptible to rapid colonization and spread of N. cristata based on the sediment profiles. Additional studies on the sediment pH, sediment type and nutrient av ailability need to be conducted confirm the prel iminary results obtained in these studies

PAGE 64

64 Table 3 1. Characterization of sediments used in study to determine effects of sediment type, nutrient addition and their interaction on growth of N. cristata Sediment type Collection location Mineral content Organic matter Sand S ilt Clay -----------------------% -----------------------Sand Lake Toho, FL 97 1.5 1.5 0.9 Muck Orange Lake, FL 21.8 0.5 0.5 77 Clay Lewisville, TX 22.5 33.7 33.7 10 Potting soil a CAIP, b FL 85.5 3.7 3.7 7 a Margo Garden Products. Folkston, GA 315 37 b University of Florida Center for Aquatic and Invasive Plants Table 3 2. Background nutrient and sediment chemistry analysis with standard error from sediment samples collected during the sediment type by nutrient addition study to determine their impa ct on growth of N. cristata. Soil Fertilizer Sample(n) TKN a SE P b SE pH SE Sand 0g kg 1 6 386 63 0.38 0.13 6.4 0.02 0 g NP 6 467 29 1.05 0.68 5.8 0.06 1 g kg 1 6 363 53 11.40 0.32 6.1 0.08 1 g NP 6 443 45 24.97 7.37 5.6 0.05 4g kg 1 6 549 17 22.70 2.54 5.8 0.08 4g NP 6 565 31 45.36 4.7 5.6 0.10 Muck 0g kg 1 6 19535 1401 5482.50 54.93 6.0 0.07 0 g NP 6 20008 1217 16.71 12.47 5.8 0.18 1g kg 1 6 15124 2783 5327.90 23.15 5.8 0.12 1 g NP 6 8939 4643 210.71 21.60 5.6 0.04 4g kg 1 6 18140 1610 5666.10 16.59 5.7 0.09 4 g NP 6 19070 2429 354.73 83.85 5.5 0.1 Potting 0g kg 1 6 1002 109 20.79 2.26 7.8 0.05 0 g NP 6 1110 41 20.56 1.29 7. 8 0.01 1g kg 1 6 909 86 24.47 1.9 8.0 0.06 1 g NP 6 1182 64 22.78 0.71 8.0 0.08 4g kg 1 6 1195 34 28.80 1.50 7.8 0.1 4 g NP 6 1174 33 31.80 2.46 7.8 0.1 Clay 0g kg 1 3 1403 63 0 0 8.3 0.01 0 g NP 3 1411 105 0 0 8.4 0.02 1g kg 1 3 1435 85 0 0 8.3 0.10 1 g NP 3 1233 67 0 0 8.3 0.03 4g kg 1 3 1716 52 0 0 8.2 0.05 4 g NP 3 1666 30 0 0 8.2 0.06 a TKN is Total Kjedahl Nitrogen in mg kg 1 b P is phos phorus from Melich 1 test in mg kg 1 NP No Plant grown in sediment

PAGE 65

65 Table 3 3. Strudy 3. Impact of hydrated lime (CCE 110%) application rates on sediment pH and growth of N. cristata grown in sand sediment amended with 1 g kg 1 fertilizer (15 9 12) N=5 Lime rate (kg m 3 ) Mean pHSE 0 6.7 0.04 0.02 7.8 0.11 0.04 8.7 0.15 0.12 9.6 0.13 0.24 9.7 0.11 0.32 9.6 0.30 0.36 10.3 0.04

PAGE 66

66 Figu re 3 1. Study 1, response of N cristata grown in sand, muck and potting soil sediments at two fertilizer rates. A) Six week number of floating leaves produced. B) Six week total plant dry weight. PS = potting soil. Different letters indicate significant d ifferences N=5.

PAGE 67

67 Figure 3 2. Study 2, response of N. cristata grown in sand, muck, potting soil and clay sediments and two fertilizer rates. A) Six week number of floating leaves produced. B) Six week total plant dry weight. PS = potting soil. Different letters indicate significant are SE. N=5.

PAGE 68

68 Figure 3 3. Study 3 l inear regression of N. cristata dry weight versus sediment pH. Sand sediment amended with hydrated lime (CCE 110%).

PAGE 69

69 CHAPTER 4 THE INFLUENCE OF LIG HT AVAILABILITY ON G ROWTH OF CRESTED FLOATING HEART ESTA BLISHED FROM RAMETS Introduction Many aquatic plants are adapted to life in low light environments. Some aquatic invasive plants such as hydrilla ( Hydrilla verticillata ) have a particularly low light compensation point and can thrive and dominate in deepe r or more turbid systems where light is limited (Bowes et al. 1977) which provides the plant with the ability to grow to the water surface Crested floating heart ( Nymphoides cristata ), hereafter referred to as N. cristata is an invasive aqua tic plant that has been expanding throughout Florida and the Southeast US. While N. cristata exhibits nymphaeid mo r p hology (plants rooted in submersed sediments with floating l eaves at the end of long stems) clonal reproductive structures called ramets represent a fully submersed form of the plant. These ramets are also the initial propagules that must establish i n order for the plant to spread. N. cristata has been found growing in water up to 3 m deep sugg esting that the small ramets are likely tolerant to low light co nditions. Impacts of light availability on the growth and establishment of N. cristata ramets have not been reported and this lack of information on such a basic environmental factor and the impact on growth suggests research is needed to determine how N. cristata ramets respond to varying degrees of reduced light availability. The objective of this study was to assess the growth response of N. cristata ramets to different light levels. Materials and Methods Number of Floating Leaves In January 2011, mature N. cristata plants were collected from the storm water treatment areas (STA) in Southern Florida and planted in 1 L pots filled with Margo commercial potting soil with a mineral composition of 85.5% sand, 3.7% silt and 3.7 % clay and organic matter conte nt of 7 %. Pots were placed in 95 L tanks filled with well water inside of a greenhouse for 2

PAGE 70

70 months to initiate growth and then moved into outdoor, 1,000 L tanks until prolific production of ramets was observed. A sample of 5 ramets was collected to deter mine initial mean dry weight prior to the beginning of each study. The first 4 week trial was conducted from June 9, 2011 through July 7, 2011 and repeated July 5, 2011 through August 4, 2011. To evaluate the e ffect of light on ramet growth, five target li ght levels were selected. These were 1, 5, 25, 50 and 100% ambient light availability at the water surface. A total of fifteen, 1,000 L mesocosm tanks were filled with well water to a depth of 50 cm. Frames were constructed to fit over the top of each meso cosm and shade cloth was pulled over the frames and layered until the desired amount of light was reached. Shade cloth was secured to the frame using 15.25 cm plastic cable ties. Light measurements were taken at the water surface with a LI COR model LI 250 A light meter 1 Table 4 1 lists the desired light levels with the range of light availability that was actually achieved. Ramets were planted 0.6 to 1.2 cm deep in 1 L pots filled with potting soil and amended with Osmocote 15 9 12. A thin layer of sand w as added over the potting soil to prevent any turbidity that may be caused by floating or suspended organic matter from the potting soil. Light treatments were applied randomly to the mesocosms and each treatment was replicated 5 times. One set of plants w as harvested at 2 weeks after planting and the second set was harvested 4 weeks after planting. Data were collected for the number of floatin g leaves at the time of harvest. Nonlinear regression analysis was performed on data with light as the independent variable and the number of floating leaves as the dependent variable using SigmaPlot 11.0 st atistical analysis software. A 3 parameter rectangular hyperbola model was fit 1 LI COR Biosciences. Lincoln, NE 68504

PAGE 71

71 to data ( ). The treatment by experimental run interaction was signific ant (p<0.001) therefore data were analyzed separately. Data from both trials are presented. Dry Weight After the number of floating leaves had been counted at 2 and 4 WAT, entire plants including all live roots, stems and foliage were cleaned of debris a nd placed in a drying oven at 76 C for 1 week. Dry weight was recorded, and data were analyzed using a linear regression with light as the independent variable and dry weight as the dependent variable (y = y 0 + ax); where y 0 is the value of y (dependent v ariable) when x = 0, a is the slope of the line and x is the independent variable. Differences between 2 and 4 weeks dry weight were also determined using a t test (p<0.05). An additional t test was used to test for differences in initial mean dry weight a nd the mean dry weight at 1% light. The treatment by experimental run interaction was not significant (p=0.279) therefore data were pooled. Results and Discussion Number of Floating Leaves The leaves of ramets grown in 25 to 100% light reached the water s urface in 3 to 4 days after planting. Plants grown at 1 and 5% light had their first leaves reach the surface in 8 and 6 days after planting respectively. In the first study, the number of leaves did not increase across light levels above 25% available li ght during the 2 week harvest (Figure 4 1 A). A similar trend was observed in the secon d study (Figures 4 2 A). The R 2 values (0.65 and 0.51 in study 1 and 2, respectively) suggest that initial l ight availability was not strongly predictive of the ability of ramets to produce floating leaves. During the first study leaf production increased as light availability increased from 1 to 25% (p=0.03) but no increase in leaf production was noted as light levels were increased beyond 25%

PAGE 72

72 during the 4 week harvest (p=0.44) (Figure 4 1 B). Results from the second trial showed leaf production after 4 weeks responded to light availability in a slightly different manner. The regression line showed a gradual but nonlinear increase (p=0.0009) across all light levels. In t he second study (Figure 4 2 B) the R 2 value was much higher (0.90), suggesting that light played a more important role in the plants ability to produce floating leaves at 4 weeks compared to the first study. Plants grown at 100% available light produced g reen foliage with red violet pigment expression in the leaf margins after 4 weeks. In contrast, plants grown under lower light levels (1 to 50%) did not exhibit a similar response, suggesting the appearance of red violet coloration in the foliage is a resp onse to exposure to the higher light intensities. Variegation has been shown to be a heritable, nonlethal genetic mutation in some plants (Orakwue and Crowder, 1983). Since variegation is a heritable trait, the red violet pigmentation observed in N. crista ta grown in full sunlight is believed to be a response to high light conditions. This accessory pigment expression likely protect s the chlorophyll inside the thin leaves and potentially produce biologically active compounds against herbivores (Schaefer and Rolshausen, 2005; Zhang et al., 2010). Plants grown under the lowest light levels produced the fewest number of leaves and also never produced flowers suggesting these plants were not utilizing carbohydrates to produce leaves or reproductive structures in production was observed at 2 weeks in the 50 and 100% light levels. Ramet production was observed 2 weeks after planting at 100% light and after 4 weeks in 50% light. Dry Weight The initia l dry weight of ramets was 0.06 grams ( 0.009) g. Two weeks after planting, there was a positive linear response (slope coefficient 0.0049; p<0.0001) of dry weight to light availability (Figure 4 3 A). From 2 to 4 weeks after planting, dry weight increase d as light

PAGE 73

73 increased from 5 to 100%. Linear increases in dry weight were also noted at the 4 week harvest (slope coefficient 0.044; p<0.0001). In contrast, no change in dry weight was noted at the 1% light level between the 2 harvest dates (Table 4 2). Ho wever, it was found that plants were able to grow at 1% light compared to the initial sample dry weight (p<0.001). The relationship based on linear regression shows that light availability plays an important role in growth of N. cristata (Figure 4 3 A, B). The fact that dry weight and floating leaf production showed different responses to light intensity suggests that leaf production is related to the ability of the ramet to produce leaves independent of light intensity, while subsequent growth is influence d by light intensity to a much larger degree (Figure 4 3 A, B). In general, increases in leaf production were less pronounced as light intensities increased above 25% light; suggesting leaf production was likely a response of carbohydrate content in the r amets. Light levels at which the cessation of additional leaf production occurred was within the range of 20 to 40% incident light that has been estimated in previous studies to support active growth of many submersed aquatic plants (Abernethy et al. 1996; Middelboe and Markage, 1997; Van et al. 1976). This light response only applies to the submersed stage of growth for N. cristata since it is unlikely that floating leaves would be subject to a sustained low light environment upon emergence. While the res ults of this study show a clear relationship between higher light availability and dry matter production by N. cristata a key finding was that even at ~1% available light (99% shade) ramets wer e able to produce stems that rapidly elongated until small lea ves emerged on the water surface in approximately 1 week. Rapid stem elongation by ramets under low light conditions suggests a mechanism that allows N. cristata to become established in deep or turbid

PAGE 74

74 water which may allow the plant to grow where other su bmersed aquatic plants or nymphaeid species would be unable to survive, further enhancing the invasive potential of N. cristata Rapid stem elongation by N. cristata under low light conditions may be somewhat analogous to rapid shoot elongation observed f or the submersed invasive species hydrilla (Van et al. 1976, Bowes et al. 1979, Steward, 1991). Similarities between N.cristata and hydrilla include rapid elongation under low light, formation of dense canopies near the water surface, and prolific producti on of vegetative propagules. All of these factors contribute to the success of these invasive species within the introduced range. While the ramets are quite small, they serve as a source of carbohydrate for the establishment of new plants. Due to their ability to tolerate and overcome low light environments, it is likely that many water bodies in Florida will provide a light environment that favors establishment and growth of N. cristata

PAGE 75

75 Table 4 1. Target ambient light levels and the actual light le vels measured at the water surface. Light readings were taken between noon and 1pm the day of planting. Target ambient light (%) Actual achieved ambient light range (%) 1 1 2 5 5 7 25 23 30 50 48 55 100 100 Table 4 2. N. cristata dry weight was measured at 2 and 4 weeks after planting. A t test was used to compare 2 and 4 week means (n=10). Mean dry weight reported in grams. Ambient light Mean dry weight (SE) a P value % 2 weeks 4 weeks -------------------------g -----------------------1 0.15(0.007)a 0.18(0.014)ab 0.173 5 0.19(0.015)b 0.44(0.023)cd 0.001 25 0.32(0.015)c 2.09(0.113)e 0.001 50 0.47(0.022)cd 2.37(0.167)e 0.001 100 0.65( 0.022)d 5.23(0.427)f 0.001 a Means followed by the same letter within a row or co lumn are not different (p = 0.05)

PAGE 76

76 Figure 4 1. Number of floating leaves of N. cristata in response to light availability in the first study at. A) 2 week harvest. B) 4 week harvest.

PAGE 77

77 Figure 4 2. Number of floating leaves of N. cristata in response to light availability in the second study at. A) 2 week harvest. B) 4 week harvest.

PAGE 78

78 Figure 4 3. Dry weight of N. cristata in response to light availability at A) 2 week harvest and B) 4 week harvest. Data were combined from both trials (n=10).

PAGE 79

79 CHAPTER 5 COMPETITION B ETWEEN CRESTED FLOAT ING HEART AND HYDRIL LA IN A REPLACEMENT SERIES TR IAL Introduction Evaluating the competitive ability of an exotic plant is an important step in assessing the community and ecosystem level impacts that may result if the plant becomes est ablished in an area. In the last 20 years there has been a significant increase in the interest of understanding the competitive strategies of exotic species (Mony et al., 2007; Lonsdale, 1999; Williamson, 19 99). However less of this effort has been on ev aluating the interactions between exotic invasive species (Mony et al., 2007; Simberloff and Von Holle, 1999; Richardson, 2000). A previous study (Chapter 4), demonstrated that crested floating heart ( Nymphoides cristata (Roxb.) Kuntze), hereafter referre d to as N. cristata is tolerant of and responds to low light by producing rapid stem elongation to produce floating leaves at the surface as quickly as possible. This ability to tolerate and respond to low light conditions is an important factor in the suc cess of an invasive aquatic plant. N. cristata also does not reproduce sexually, but spreads prolifically through production of ramets. These features suggest that N. cristata could be a highly competitive species; however, studies are needed to confirm th is. One way to evaluate the competitive ability of two different species is evaluating a replacement series in which two species are grown at a constant density but the ratio of each species changes. While there are many limitiations to the replacement se ries model, it is a valid way to compare species yield as a function of dry matter production and the ability of one species to accumulate greater mass than another (Joliffe, 2000). Similarities between N.cristata and hydrilla ( Hydrilla verticillata ) incl ude rapid elongation under low light, formation of dense canopies at the water surface, and production of copious small vegetative propagules that can promote spread and allow survival The objective of this

PAGE 80

80 study was to assess the competitive abilities of N. cristata established from ramets and hydrilla ( Hydrilla verticillata ) established from apical sections as a function of dry weight accumulation in a replacement series study. Materials and Methods In January 2011, mature N. cristata plants were colle cted from the storm water treatment areas (STA) in South Florida. Ramets and root crowns were planted in 1 L pots filled with Margo commercial potting soil (Chapter 1) with a mineral composition of 85.5% sand, 3.7% silt and 3.7 % clay and a n organic materi al content of 7 % at the University of Florida, Center for Aquatic and Invasive Plants (CAIP) in Gainesville, FL. Pots were placed in 95 L tanks in a greenhouse from January 2011 through March 2011 for establishment and were then moved into outdoor 1,000 L culture tanks until prolific production of ramets was observed. Ramets were collected and used to perform the planned studies. Hydrilla tips (4 cm) were collected from plants growing in 1,000 L culture tanks. The first study was conducted from July 30, 201 1 through September 10, 2011 and was repeated from August 10, 2012 through September28, 2012. Studies were conducted outside under a shade canopy with 70% incident light. Twenty one 95 L plas tic tanks were filled with a 7.6 cm deep layer of potting soil a mended with Osmocote 15 9 12 at a rate of 1g kg 1 of dry soil. A layer of sand 0.5 1 cm deep was added over the potting soil as a cap to prevent turbidity and floating organic matter. Tanks were filled with well water and a small skim net was used to remo ve any debris after the tanks were filled. Tanks were allowed to set for 24 hours to allow the sediments to settle out of the water. The desired plant density was 23 plants m 2 or a total of 6 plants per tank. Sixty three ramets of N. cristata of approxim ately equal size (0.52 0.02 g fresh weight) and 63, apical tip sections of hydrilla were collected for planting. Hydrilla tips were clipped with scissors until the

PAGE 81

81 fresh weight was similar to that of the N. cristata ramets (0.490.04 g fresh weight). Eac h tank was randomly assigned a different planting ratio of N. cristata : hydrilla ranging from 0:6 to 6:0 changing the number of each plant in the mixture by 1 each time. Visual observations and photographs were taken on a weekly basis to document each spec ratios thr oughout the 7 week experiment Planting ratios were assigned randomly to the 21 tanks and each ratio was replicated 3 times. Plants were harvested 7 weeks after planting. Plant shoots were cut at the surface of the s ediment, removed and separated by species. Each species was then placed in a drying oven for 1 week at 76 C. Mean dry weight for each species and mean total combined plant dry weight were calculated for each planting ratio and plotted to determine comparat ive growth A n ANOVA test ring the mean dry weights in the first and second study. There were no treatment by experiment interactions found (p=0.15) and data were combined. Data were further analyzed with a significant differences in each species mean shoot dry weight for each planting ratio. Results and Discussion N. cristata was the dominant plant 7 weeks after planting (> 60% total plant biomass) in all trea tments above a 1:5 ( N. cristata: hydrilla ) planting ratio (Figure 5 1). Dry weight and standard errors are listed in Table 5 1. Total plant shoot dry weight remaine d near 30 g across all planting ratios. Dry weight of both species was equal at a planting r atio of 1:5 ( N. cristata : hydrilla ). These results show that N. cristata was able to accumulate more biomass even at lower initial density than hydrilla and contribute more overall biomass, suggesting that N. cristata was the more competitive of the two sp ecies under these experimental conditions. Nonetheless, hydrilla was never e xcluded from the study system at any planting ratio. This finding is similar to the results reported by Wu et al. (2006), in which yellow floating heart ( Nymphoides peltata ),

PAGE 82

82 herea fter referred to as N. peltata and produced more biomass than water chestnut ( Trapa bispinosa ). T. bispinosa was not excluded from the study system. Given the similar size of the initial propagules used for this study, the results were somewhat surprising based on reports of rapid summer growth rates of hydrilla (Glomski and Netherland, 2012). It could be argued that the ramets of N. cristata had an advantage in growth due to the possibility of a dense carbohydrate reserve in tuber cluster at its base to initiate rapid growth, while the hydrilla tip fragment lac ked this type of structure. H owever, it is important to note that the starting weights of the two species were the same. It was also observed that by 2 weeks after planting N. cristata had not crea ted a full canopy in the tanks and stems of hydrilla had grown to the water surface and begun to produce lateral branches (Figure 5 2 B). By 4 weeks after planting, N. cristata had formed a solid canopy over much of the water surface in the tanks and by 6 weeks had completely covered the water surface in all tanks with a layer of overlapping leaves (Figure 5 2 C and D), except those planted at the 1:5 ratio. H ydrilla was not visible at the water surface a fter canopy closure by N. cristata leaves But a t the time of harvest it was noted that hydrilla stems had apparently elongated under the canopy of N. cristata leaves. Stems of hydrilla reached lengths of 76 cm although the depth of the water in the tank was only 45 cm. This continued growth under the canopy demonstrates that even though the N. cristata had reached its capacity and shaded the water column beneath it, hydrilla, which is very tolerant of low light, was still able to grow and elongate below the floating leaves. The ability of hydrilla to continu e to grow contributed to a synergistic effect of total plant biomass in the study system. This synergistic effect of plant biomass is evident because the total plant biomass ex ceeds the predicted biomass (Figure 5 1) ( Gregory MacDonald, personal communicat ion, 2012)

PAGE 83

83 Studies reporting the competitiveness of N. peltata with the submersed invasive Eurasian watermilfoil ( Myriophyllum spicatum ) have shown that N. peltata ac cumulates most of its biomass in above ground parts first (Wu et al. 2006), then begin s to partition mass to below ground structures. Give n the short term nature of the current competition trials, there was no focus on belowground biomass allocation. Larson (2006) reported that the effects of N. peltata on the submersed aquatic plants coon tail ( Ceratophyllum demersum ) elodea ( Elodea canadensis ) and fan leaved crow foot ( Ranunculus circinatus ) caused all three of these species to produce reduced growth when under any canopy coverage. However, these submersed species showed elongation of the ir nodes and stems in response to a limited light environment (Larson, 2006). While node length was not measured in hydrilla, it is possible the same response could have been present in the present studies. Longer term studies on the competitiveness of N. cristata will be useful in furthering our understanding of how this highly invasive species interacts with other species. In summary, N. cristata is a highly competitive, invasive plant that can compete well with aggressive, low light tolerant, submersed i nvasives such as hydrilla. This information suggests that N. cristata ramets could serve as highly competitive structures in the presence or absence of submersed vegetation. The ability of N. cristata to become established in water up to 3 m deep suggests an ability to withstand low light conditions (Chapter 5) and this typically suggests a strong competitive ability for an aquatic species.

PAGE 84

84 Table 5 1. Comparison of dry weight ( SE) of N. cristata and hydrilla after 7 weeks of competition. N=6. Differe the plant species. Ratio Mean dry weight (SE) a N. cristata hydrilla -------------------------g ----------------------------6:0 32.2(3.6) a 0.0(0) i 5:1 30.3(1.4) ab 1.5(0.3) i 4:2 27.9(1.8) abc 4.2(0.9) hi 3:3 23.9(2.0) cd 8.9(1.3) gh 2:4 20.9(3.0) de 10.3(1.7) g 1:5 15.6(1.9) f 17.2(1.5) ef 0:6 0.0(0) i 26.4(2.5) bc a Different letters represent differences among the treatments determined by Fishers pr otected LSD (

PAGE 85

85 Figure 5 1. The dry weight of foliage of N. cristata, hydrilla, and total plant dry weight after 7 weeks of growth in a replacement series competition study.

PAGE 86

86 A B C D Figure 5 2. Picture s of study tanks at the 3 :3 ( N. cristata: hydrilla ) planting ratio over the course of the study. A) Initial planting, B) 2 weeks after planting, C) 4 weeks after planting, D) 7 weeks after planting

PAGE 87

87 CHAPTER 6 A COMPARATIVE ANATOM Y OF CRESTED FLOATIN G HE ART AND YELLOW WATER LILY LEAVES AN D STEMS Introduction Crested floating heart ( Nymphoides cristata (Roxb.) Kuntze), hereafter referred to as N. cristata a member of the Menyanthaceae family, is a recently expanding exotic invasive aquatic plant from South east Asia. Since it was positively identified in 1996, this plant has been observed completely blocking canals in areas throughout south Florida (Burks, 2002) and covering 2,428 ha of water surface on the Santee Cooper reservoir in South C arolina (Wesrbroo ks et al. 2012 ). Aquatic invasive species have been reported to cause several ecological problems including formation of dense surface mats, displacement of native vegetation and reduction of biodiversity (Janes et al. 1996). N. cristata grows rooted in s ubmersed sediments and produces floating leaves at the end of l ong stems (a nymphaeid morphology ), and has been observed to form large monocultures. The overlapping leaves can shade out microscopic algae and other aquatic plants which can cause reductions in dissolved oxygen due to the reduction in photosynthetic productivity below the water surface. Despite an increased emphasis on management of this species, control measures have proven inconsistent. Currently there are no known species specific insect herbivores that will consume N. cristata the plant is not consumed by the herbivorous grass carp ( Ctenopharyngodon idella ), and mechanical and cultural control methods have also proven ineffective (Van Dyke et al. 1984; Singh et al. 1966; Middleton 1990). This suggests that chemical control may be the only current option for slowing the spread of this rapidly expanding invasive plant. Many chemical control options and techniques have been ineffective for controlling N. cristata The lack of reported effi cacy could potentially be due to anatomical and physical

PAGE 88

88 features of the plant. The leaves and stems are very flaccid, and are easily disturbed by light wind and wave action. This makes herbicide contact time requirements difficult to achieve especially du ring foliar herbicide applications from a boat or helicopter as any disturbance Yellow water lily ( Nymphaea mexicana ), hereafter referred to as N. mexicana is very similar to N. cristata in appearance. Both plants have small surface leaves and long slender stems. However, N mexicana has much more rigid and stiff leaves and stems and may not be affected as much by wind and wave action. The objecti ve of this st udy was to compare anatomical and physical differences between N. cristata and the native plant N. mexicana. While these species have numerous similar morphological characteristics, it is possible that more detailed anatomical differences could explain pot ential differences in herbicide efficacy. Materials and Methods Leaf Sections Plant material from both species used in this study was collected from two different sources. N. cristata plants were collected from South Florida Water Management District (SFWM D) ponds in south Florida in January, 2011. Populations established at the University of Florida Center for Aquatic and Invasive Plants (CAIP) in Gainesville, FL. N. cristata plants were planted in Margo potting soil with a mineral composition of 85.5% sa nd, 3.7% silt and 3.7 % clay and an organic content of 7% that was amended with Osmocote (15 9 12) at 1 g kg 1 dry soil then set in 1,000 L culture tanks filled with well water. Tissue samples were taken at the time of flower production. Nymphaea mexicana samples were collected from plant cultures growing in 95 L tanks at the CAIP. Four stems and 4 leaves were collected from each species. Leaves were left whole, and loosely rolled and placed in vials of formalin acetic acid (FAA) for

PAGE 89

89 preservation for 72 hou rs. Tissue samples were then moved into vials of 70% ethanol for storage (Stern et al., 1997). Half of the leaf tissue for each species was then cut into 1cm squares and placed in new vials for dehydration in a tert iary butyl alcohol series, embedded in p araffin and sectioned on a rotary microtome (Kuo Huang et al. 1994) Microtome sections of 15m were permanently mounted on glass slides. Slides were then emersed in a safranin solution to stain sclerified and lignified tissues. Once prepared, slides were examined under light microscopes of magnif ication varying from 4x to 40x to count t he number of vascular bundles and scler eids in each species Leaf thickness was measured using an ocular micrometer calibrated to 4x magnification. Leaf sections for analys is were selected randomly from the preserved sample and sections differences in the means between the 2 species. Stem Sections Stems were cut into 4 cm sections and placed in vials of FAA for 72 hours and then transferred to vials of 70% ethanol for storage. A notch was cut out in one half of a 6.4 cm 2 block cut from a carrot to hold the stem without crushing it when the other half was placed o ver top of it and locked into the sliding microtome. Blocks of carrot were used because the orange colored tissue of the carrot provided a color contrast against the green stem section (compared to using two cork sections to hold the sample) which made recovery of sections easier after cutting with a sliding microtome. The block of carrot containing the stem was placed in a sliding microtome and set to cut 30m sections and then prepared for permanent mounting following the procedure used by Stern and Carlsward (2006). In additio n to the sliding microtome sections, hand sections were cut using disposable razor blades to cut thin cross sections of tissue. For some species of aquatic plants these thicker hand sections are preferable to the thinner microtome sections

PAGE 90

90 because they can display larger portions of vascular tissues, and can minimize damage to the microstructure during handling of the tissue sample (Schneider and Carlquist, 2009) Stem tissue was evaluated for diameter, arrangement of vascular tissues and number of sclere ids. Tissue samples were photographed under the microscope using a Canon PowerShot CD1400 IS 1 digital camera. Stem sections for analysis were selected randomly from the preserved sample and sections were replicated 12 times. Data were analyzed using a stud test to detect differences between the two species in each measured parameter. Results and Discussion Leaf Sections Leaf thickness was greater in N. mexicana than in N. cristata (450 26 m and 289 5 m, respectively) (Table 6 1). The palisade (upper) layer contributed to approximately half of the total leaf thickness in N. cristata while the aerenchymatous (lower) leaf of N. mexicana made up over half of its leaf thickness (Figure 6 1 A, B). The mean number of vascular bundles present in le af tissue samples was not different between species (Table 6 1). The vascular bundles of N. cristata were located in the very center of the leaf (Figure 6 1 C) whereas the vascular bundles in N. mexicana leaves were located in the lower portions of the lea f surrounded by the aerenchyma (Figure 6 1 D). Asterosclereids are crystal like structures founds in plants. In aquatic plants they are strengthening structures (Huang et al. 2000) (Figure 6 1 E, F). The mean number of asterosclereids differed between th e two species with N. Mexicana having more than 2 times the number that were present i n N. cristata (Table 6 1). Sclereids were arranged randomly throughout the leaf tissue of both species. 1 Canon U.S.A., Inc. Lake Success, NY 11042

PAGE 91

91 Leaves of N. mexicana were 1.5 times thicker than the leaves of N. cristata (Table 6 1). The leaves of N. mexicana also stained with safranin more readily than those of N. cristata While this could have been due to the time exposure of the sample to safranin not being long enough, it could also indicate a lack of scle rified tissue as well as less lignified tissue in N. cristata This is important to note because with these structures and compounds lacking, it could be difficult for foliar herbicide application to be effective due to the resulting flaccid nature of the leaf. While not measured, the spongy mesophyll layer of the leaf appeared to account for a greater portion of leaf thickness in N. cristata In the lower aerenchymatous layer of the leaf, N. cristata had numerous small aerenchyma while N. mexicana had few er, larger aerenchyma. It was observed that the aerenchymatous layer contained most of the sclereids in the leaf of both species Since N. mexicana had a larger aerenchymatous layer this could account for the greater number of sclereids and therefore the l eaf being more rigid than the leaf of N. cristata The lack of sclerified and lignified leaf tissue in N. cristata may also play a role in the vegetative reproduction of the plant. The leaf composed of mostly soft tissue could allow for the development of roots from leaf fragments ( Kenneth Langeland, personal communication, 2011). This type of vegetative growth was observed during this research. Cuticle thickness was not measured since it has been reported that species with floating leaves do not vary great ly in cuticle thickness but rather more in the length of carbon chains that make up the waxy layer (Amaral et al. 1990) The number of vascular bundles present in each leaf sample did not differ between the 2 species which could be due to the small physic al size of each sample which limited the sampling region to about a 1cm or smaller piece of leaf tissue. In the past, however, most studies on the

PAGE 92

92 vasculature of the order Nymphaeaceae have concentrated on the stem rather than the leaves (Weidlich, 1976) N. mexicana leaf samples contained 2.3 times more sclereids per leaf section than N. cristata These sclereids are arranged randomly, but occur mostly around air channels (Kuo Huang et al. 1994) These sclereids in N. mexicana are known to have cell walls consisting of prismatic crystals of calcium oxalate (Kuo Huang et al. 1994) but this composition has not been reported in N. cristata Leaf sclereids of N. cristata averaged 148.34m long, while in blue water lily ( Nymphaea nouchali ) sclereid length avera ged 302.15m (Khatun and Mondal, 2011) Fewer and smaller sclereids would likely contribute to in the flaccid nature of N. cristata leaves. Stem Sections S tems of N. mexicana were 1.7 times greater in diameter than N. cristata (Table 6 1) and t here were al so visible differences in the size and arrangement of the gas tubes in the center of the stems (Figure 6 5 A, B). The gas channels in N. cristata were arranged radially around a central vascular bundle with 1 or 2 smaller vascular bundles located to the ou tside of the stem. In N. mexicana 2 gas channels occupied most of the center of the stem with the other cells and with 6 8 vascular bundles located on the outside of the gas channels. There were also areas of highly sclerified tissue located in the epiderm is of N. mexicana which were absent in N. cristata N. mexicana had twice the number of vascular bundles compared to N. cristata (Table 6 1). In N. mexicana the number o f vascular bundles in a stem was 8, with the phloem in the middle surrounded by a ring of xylem (Weidlich, 1976) The arrangement of these bundles was also different between the species. In N. mexicana the bundles were arranged toward the outside of the stem, while in N. cristata they were in the center of the stem (Figure 6 5 C, D). The ind ividual bundles in N. cristata were larger than the individual bundles in N. mexicana

PAGE 93

93 Sclereid counts were higher in N. mexicana and ranged from 11 to 25 while in N. cristata these counts ranged from 5 to 17 (Table 6 1). In general, for both species the se structures tended to be distributed along bands of connecting cells (Figure 6 5 E, F). Similar to the leaves, these structures were arranged around the air channels. While types of different sclereids were not identified, previous work has found three d ifferent types of sclereids present in N. cristata i.e. astero, osteo, and trichosclereids (Khatun and Mondal, 2011) The differences between these species in terms of supporting structures and cell wall composition provide strong evidence that anatomical and physical differences may play a role in the efficacy of foliar applied herbicides on each of these plants. The lack of lignified and sclerified structures in leaves of N. cristata may be a factor that results in the flaccid nature of the leaf which all ows the leaf to easily submerge below the water surface in the slightest boat wake or ripple on the water surface. This would likely wash any herbicide off of the leaf surface before sufficient contact for control has been achieved. In contrast, the presen ce of greater numbers of those structures and compounds may allow a longer contact time to be achieved in N. mexicana The difference in vascular bundle numbers in the stem could also play a role in the apparent difference in herbicide efficacy. The numbe r of vascular bundles was found to be the same in the leaves of the two species but less in the stems of N. cristata compared to N. mexicana This reduction in the number of vascular bundles present from leaf to stem could cause a bottleneck effect for her bicide translocation. A bottleneck effect could cause a build up of herbicide in the leaves before it can be loaded and moved into the stems and translocated to the roots. This could explain why many foliar treatment results in quick necrosis of the foliag e but do not damage the root systems (see Chapter 2). Water depth (stem length) also may play a role in the ability of a herbicide to translocate from leaf to root. It has been noted from field

PAGE 94

94 applications that when plants are in shallow water (<1 m), gre ater control has been achieved than when plants in deeper water receive the same treatment ( Chip Davis, personal communication, 2011). This difference could be because, even though a bottleneck effect may exist, in shallow water (short stems) the herbicide has a shorter distance to travel f rom the leaf to the root system, where it may provide longer and more complete control.

PAGE 95

95 Table 6 1. Features measured on N. cristata and N. mexicana for anatomical and ph ysical comparisons. Data report ed as mean value standard error. Feature measured Mean (SE) a N. cristata N. mexicana Leaf thickness (m) 289.7(5.48)a 450.6(26.46)b Vascular bundles (leaf) 2.6(0.28) 2.5(0.39) Sclereids (leaf) 3.5(0.28)a 8.1(1.10)b Stem diameter (m m) 1.5(0.04)a 2.4(0.08)b Vascular bundles (stem) 3.3(0.13)a 6.8(0.16)b Sclereids (stem) 12.1(0.96)a 18.8(1.19)b a Different letters next to the means indicate differences between the 2 species within the specific feature evaluated. A studen test was used to compare means.

PAGE 96

96 A B C D E F Figure 6 1. A) Leaf tissue sample under 4x magnification of N. cristata B) N. mexicana leaf section under 4x magnification, C) Leaf vascular bundles in N. c ristata under 10x magnification, D) N. mexicana leaf vascular bundle under 10x magnification, E) Asterosclereid in N. cristata under 40x magnification, F) Asterosclereid in N. mexicana under 40x magnification.

PAGE 97

97 A B C D E F Figure 6 2. A) Stem section of N cristata under 4x magnification, B) N. mexicana stem section under 4x magnification, C) Vascular bundle in stem of N. cristata under 10x magnification, D) vascular bundle of N. mexicana under 10 x magnification, E) asterosclereids in stem of N. cristata under 40x magnification, F) asterosclereids in N. mexicana stem under 40x magnification.

PAGE 98

98 CHAPTER 7 CONCLUSIONS AND RECO MMENDATIONS The information gained from these studies is important sinc e there is very limited peer reviewed literature pertaining to the basic biology and management of N. cristata N. cristata was found to be tolerant of most aquatic herbicides; however the herbicides diquat, endothall (amine salt), imazamox, imazapyr and s everal combinations of these herbicides reduced dry weight of N. cristata below their respective pre treatment reference samples. It was also determined that application method and formulation did not have an impact on the efficacy of most herbicides. Exce ptions to this were imazamox which had greater efficacy as a foliar application compared to subsurface application and endothall (amine salt) was more effective in a liquid formulation than a granular formulation. The lack of strong endothall (dipotassium salt) activity on N.cristata in these assays is in contrast to findings from previous research; however, multiple trials with the dipotassium salt of endothall suggested that prolonged exposure periods (essentially days of static exposure) were required to reduce dry weight. Results of these herbicide efficacy studies require further field verification. N. cristata ramets are tolerant of low light conditions. When exposed to 1% available light at the water surface the small propagules responded with rapid p etiole elongatio n and quickly developed a canopy of leaves on the surface of the water. Once at the water surface these leaves are in full sunlight and photosynthetic output is no longer restricted. There was a linear relationship between biomass and ligh t availability, but a non linear relationship with regard to leaf production. Leaf variegation was noted in full sunlight and believed to be a response to high light intensity. N. cristata exhibited the ability to grow and establish in a wide range of sed iment types as well as different levels of sediment fertility. The plant grew best in sand and muck sediments

PAGE 99

99 compared to clay or commercial potting soil. Growth was generally enhanced in the more fertile sediments ; however, increased fertility did not gre atly enhance growth in the clay or commercial potting soil Sediment pH was linked to N. cristata growth and biomass accumulation. Plants grew best in low pH (acidic) sediments compared to high pH (basic) sediments and growth showed a strong, negative line ar response to increasing the pH of a sand based sediment. N. cristata demonstrated that it is a competitive plant by accumulating more dry weight than hydrilla in a replacement series competition study. Even when there was 5 times more hydrilla than N. cristata planted, N. cristata still accumulated the same dry weight as when planted alone N. cristata petioles quickly elongate from submersed ramets and reach the water surface before hydrilla. N. cristata formed a complete canopy over the hydrilla with i n a few weeks ; however hydrilla continued to elongate beneath the surface mat. The copious production of ramets, rapid stem extension to form a canopy in water up to 3 m in depth, ability to grow under a wide range of sediment types, and tolerance to low l ight conditions suggest N. cristata has several invasive characteristics. The anatomy of the leaves and petioles of N. cristata offer some insight as to possible mechanisms for the plants foliar applied herbicides when compa red to the more easily controlled N. mexicana A reduced number of vascular bundles from the leaf to the petiole could cause a bottleneck effect for herbicide translocation. N. cristata also contains less lignified and sclerified structures than several ot her p lants of similar life form which translocate to the roots. This lack of supporting structure in the leaf also offers a mechanism of reduced contact time from foliar ap plications since the flaccid nature of the leaf allows it to

PAGE 100

100 easily dip beneath the water surface in the wake of a spray boat or air disturbance from aerial application. Additional studies to evaluate translocation of foliar applied herbicides and the pot ential impact of water depth (petiole length) would be helpful to management efforts since it is currently thought that the plant is more susceptible to herbicides in shallower (<1.5m) water. The large numbers of ramets that are produced by each plant also present a challenge to management efforts. Research to better understand the factors that induce ramet sprouting and quiescence could be important to future control programs. The wide availability of N. cristata through the ornamental trade and well estab lished populations in south Florida and North and South Carolina (as well as nascent populations in Louisiana and Texas) suggest high potential for continued spread of this plant Management options are currently limited and further emphasis on herbicide a pplication timing and response of ramets to herbicides would be valuable to resource managers.

PAGE 101

101 APPENDIX ANOVA TABLES FROM CHAPTER 3 STUDY Table A 1. 2 way ANOVA table for 6 week floating leaf data from trial 1 of the sediment type x nutrient availability study. S ource of variation DF SS MS F P soil 2 628.133 314.067 8.212 < 0.001 F ertilizer 2 14358.933 7179.467 187.726 <0.001 S oil x fertilizer 4 1909.333 477.333 12.481 <0.001 R esidual 36 1376.8 38.244 T otal 44 18273.2 415.3 Table A 2 2 way ANOVA table for 6 week dry weight data from trial 1 of the sediment type x nutrient availability study. S ource of variation DF SS MS F P S oil 2 208.780 104.39 10.082 <0.001 F ertilizer 2 2429.420 1214.71 117.322 <0.001 S oil x fertilizer 4 700.378 175.095 16.911 <0.001 R esidual 36 372.732 10.354 T otal 44 3711.311 84.348 Table B 1 2 way ANOVA table for 6 week floating leaf data from trial 2 of the sediment type x nutrient availability study with addition of clay sediment. S ource of variation DF SS MS F P S oil 3 6916.078 2305.359 33.001 <0.001 F ertilizer 2 11367.394 5683.697 81.362 <0.001 S oil x fertilizer 6 10383.089 1730.515 24.772 <0.001 R esidual 42 2934.00 69.857 T otal 53 30825.5 581.613 Table B 2 2 way ANOVA table for 6 week dry weight data from trial 2 of the sediment type x nutrient availability study with addition of clay sediment. S ource of variation DF SS MS F P S oil 3 0.159 0.0529 2.154 0.108 F ertilizer 2 8.973 4.487 182.698 <0.001 S oil x fertilizer 6 4.53 0.755 30.745 <0.001 R esidual 42 1.031 0.0246 T otal 53 14.865 0 .28

PAGE 102

102 LIST OF REFERENCES Abernathy VJ, Sabbatini MR, Murphy KJ. 1996. Response of Elodea canadensis (Michx) and Myriophyllum spicatum (L.) to shade cutting and competition in experimental culture. Hydrobiologie. 340:219 224. Amaral M, Silva A, Salatino A. 1990. Alkanes of surface waxes from eight species of aquatic angiosperms. Aquatic Botany. 36:281 286. Baird D, Baker G, Brown H, Urrutia V. 1983. Aquatic weed control with glyphosate in s outh Florida. Southern Weed Science Society. pp.430 435. Barko JW, Gunnison D, Carpenter SR. 1991. Sediment interactions with submersed macrophyte growth and community dynamics. Aquatic Botany. 41:41 65. Batjes NH, 1995. A global data set of soil pH proper ties. International Soil Reference and Information Center. Technical paper 27. Bellaud M. 2009. Chapter 6: cultural and physical control of aquatic weeds, pp. 35 40. In: Biology and control of aquatic plants: a best management practices handbook (Gettys L A, WT Haller and M Bellaud, eds.). Aquatic Ecosystem Restoration Foundation, Marietta GA. 210 pages. Bowes G, Van TK, Garrard LA, Haller WT. 1977. Adaptation to low light levels by hydrilla. J. Aquat. Plant Manage. 15:32 35. Bultemeier BW, Netherland MD, F errell JA, Haller WT. 2009. Differential herbicide response among three phenotypes of Cabomba caroliniana. Invas. Plant Sci and Manage. 2:352 359. Burks KC(a), 2002. Nymphoides cristata (Roxb.) Kuntze, a recent adventive expanding pest in Florida. Castanea 67:206 211. Burks KC(b), 2002. Nymphoides cristata Snowflakes in sunny Florida. Aquatics 24 :8 10. Carlsward BS, Stern WL, Judd WS, Lucansky TW. 1997. Comparative leaf anatomy and systematic in Dendrobium sections aporum and Rhizobium (Orchidaceae). Inte rnational J. of Plant Sci. 158:332 342. Center for Invasive Species and Ecosystem Health. 2010. Crested floating heart, Nymphoides cristata (Roxb.) (Ktze.). Accessed June 21, 2012. www.invasive.org/browse/subinfo.cfm?sub=11616 Colle D. 2009. Chapter 10: gr ass carp for biocontrol of aquatic weeds, pp. 61 64. In: Biology and control of aquatic plants: a best management practices handbook. Aquatic Ecosystem Restoration Foundation, Marietta GA. 210 pages.

PAGE 103

103 Countryman WD. 1970. The History, spread and present dis tribution of some immigrant aquatic weeds in new England. J. Aquat. Plant Manage. 8:50 52. Crested Floating Heart 2010, October 23. Accessed on January 14, 2011, from Center for Aquatic and Invasive Plants: http://plants.ifas.ufl.edu/node/291 Florida Exot ic Pest Plant Council, 2009. 2009 List of Invasive Plant Species. Florida: FLEPPC. Food and Agriculture Organization. Natural Resource Management and Environmental Department. 2001. Lecture notes on major soils of the world. Set #4. Mineral soils condition ed by topography. Accessed on May 17, 2012. www.fao.org/DOCREP/003/Y1899E/y1899e07.htm Fox A, Haller W, Getsinger K, Petty D. 2002. Dissipation of triclopyr herbicide applied in Lake Minnetonka, MN concurrently with rhodamine WT dye. Pest Manage. Sci. 58:6 77 686. Gatiboni LC, Oliveira CMB, Ernani PR, Almeida JA, Miquelluit DJ. 2010. Soil available phosphorus by Mehlich 1, Mehlich 3 and anion exchange resin and its sensitivity to soil clay content. Soil Science and Natural Resources Dept. Santa Catarina Stat e Univ. Lages, SC, Brazil. 19 th World congress of soil science. Brisbane, Australia. Glomski LM, Nelson L. 2008. Evaluation of 2,4 D ester and triclopyr amine against waterlily and spatterdock. US Army Corps of Engineers, ERDC/TN APCRP. Glomski LM, Netherl and MD. 2012. Does hydrilla grow an inch per day? Measuring short term changes in shoot length to describe invasive potential. J. Aquat. Plant Manage. 50:54 57. Green WR, Westerdahl HE, Joyce JC, Haller, WT. 1989. Triclopyr (Garlon 3A) dissipation on Lake Seminole, Gerogia. Miscelaneous Paper A 89 2. US Army Engineer Waterways Experiment Station, Vicksburg, MS. 19pp. Habeck DH, 1974. Caterpillars of Parapoynx in relation to aquatic plants in Florida. Hyacinth Control Journal. 12:15 18. Haller WT. 2009. Chap ter 7: mechanical control of aquatic weeds, pp. 41 46. In: Biology and control of aquatic plants: a best management practices handbook (Gettys LA, WT Haller and M Bellaud, eds.). Aquatic Ecosystem Restoration Foundation, Marietta, GA. 210 pages. Haller WT, Sutton DL. 1973. Factors affecting uptake of endothall 14 C by hydrilla. Weed Science. 21:446 448. Hanlon C, Haller WT. 1990. The impact of 2,4 D used in water hyacinth control programs on the growth of non target spatterdock. Aquatics 12:14 16. Janes RA, Eaton JW, Hardwick K. 1996. The effects of floating mats of Azolla filiculades Lam. and Lemna minuta Kunth. on the growth of submerged macrophytes. Hydrobiologia. 340:23 26.

PAGE 104

104 Jolliffe PA, 2000. The replacement series. J. Ecol. 88:371 385. Khatun M, Mondal K 2011. Studies on the sclereids diversity and distribution pattern in the different plant organs (leaves, stems and fruits) of some selected medicinally viable angiospermic taxa in eastern India: a systematic approach. Advances in Bioresearch 2:111 121. K irk JP, Henderson JE. 2006. Management of hydrilla in the Santee Cooper Reservoirs, South Carolina: Experiences from 1982 2004. J. Aquat. Plant Manage. 44:98 103. Kirk JP, Morrow JV, Killgore KJ, DeKozlowski SJ, Preacher JW. 2000. Population response of tr iploid grass carp to declining levels of hydrilla in the Santee Cooper Reservoirs, South Carolina. J. Aquat. Plant Manage. 38:14 17. Koschnick TJ, Netherland MD, Haller WT. 2007. Effects of three ALS inhibitors on five emergent native plant species in Flo rida. J. Aquat. Plant Manage. 45:47 51. Koschnick TJ, Haller WT, Chen AW. 2004. Carfentrazone ethyl pond dissipation and efficacy on floating plants. J. Aquat. Plant Manage. 42:103 108. Kuo Huang L, Chen S, Chen S. (2000). Ultrastructural study on the form ation of sclereids in floating leaves of Nymphoides coreana and Nuphar shimadai Botanical Bulletin of Academia Sinica 41:283 291. Kuo Huang L, Sheue CR, Yang YC. 1994. Calcium oxalate crystals in some aquatic angiosperms of Taiwan. Botanical Bulletin of A cademia Sinica 34: 179 188. Langeland KA, Bowes G, Rattray M, Shilling D. 1993. Evaluation of bensulfuron methyl and triclopyr as aquatic herbicides. In Integrated management of aquatic weeds, ed. W.T. Haller, Final Report, Oct. 1, 1988 to Sept. 30, 1993, 1 70 188. Gainesville, FL: Center for Aquatic Plants, IFAS, University of Florida. Langeland KA, Hill ON, Koschnick TJ, Haller WT. 2002. Evaluation of a new formulation of Reward landscape and aquatic herbicide for control of duckweed, water hyacinth, water lettuce and hydrilla. J. Aquat. Plant Manage. 40:51 53. Langeland KA, Warner JP. 1986. Persistence of diquat, endothall and fluridone in ponds. J. Aquat. Plant Manage. 24:43 46. Larson D, 2006. Growth of three submerged plants below different densities of Nymphoides peltata (S.G. Gmel.) Kuntze. Aquat. Botany. 86:280 284. Larson D, 2006. Reproduction strategies in introduced Nymphoides peltata populations revealed by genetic markers. Aquat. Botany. 86:402 406. Les DH, Mehrhoff LJ. 1999. Introduction of noni ndigenous aquatic vascular plants in southern New England: a historical perspective. Biol. Invasions. 1:281 300.

PAGE 105

105 Lonsdale WM, 1999. Global patterns of plant invasions and the concept of invasibility. Ecology. 80:1522 1536. Middelboe AL, Markager S. 1997. D epth limits and minimum light requirements of freshwater macrophytes. Freshwater Biol. 37:553 568. Middleton BA, 1990. Effect of water depth and clipping frequency on growth and survival of four wetland plant species. Aquat. Botany. 37:189 196. Mony C, Kos chnick TJ, Haller WT, Muller S. 2007. Competition between two invasive Hydrocharitaceae ( Hydrilla verticillata (L.f.)(Royle) and Egeria densa (Planch)) as influenced by sediment fertility and season. Aquat. Botany. 86:236 242. Mudge C, Haller WT. 2010. Eff ects of pH on submersed aquatic plant response to flumioxazin. J. Aquat. Plant Manage. 48:30 34. Nelson LS, Skogerboe JG, Getsinger KD. 2001. Herbicide evaluation against giant salvinia. J. Aquat. Plant Manage. 39:48 53. Netherland MD. 2009. Chapter 11: ch emical control of aquatic weeds, pp. 65 78. In: Biology and control of aquatic plants: a best management practices handbook (Gettys LA, WT Haller and M Bellaud, eds.). Aquatic Ecosystem Restoration Foundation, Marietta GA. 210 pages. Netherland MD, Getsing er KD. 1992. Efficacy of triclopyr on Eurasian watermilfoil: concentration and exposure time effects. J. Aquat. Plant Manage. 30:1 5. Orakwue FC, Crowder LV. 1983. Genetic inheritance of variegation in Brassica campestri L. Journal of Heredity. 74:65 67. O rnduff R, 1970. Cytogeography of Nymphoides (Menyanthaceae). Taxon. 19:715 719. Page C, 2010. Crested floating heart Clarendon County, South Carolina: South Carolina Department of Natural Resources. Pennington TG, Skogerboe JG, Getsinger KD. 2001. Herbic ide/copper combinations for improved control of Hydrilla verticillata J. Aquat. Plant Manage. 39:56 58. Anonymous. 2012. Plant this Club. Nymphoides cristata, Crested floating heart, white water snowflake, Menyanthaceae. Accessed, July 31, 2012. www.plant this.com/plant information.asp?gardener=25751&plantSpot=3 Poovey AG, Getsinger KD. 2007. Subsurface applications of triclopyr and 2,4 D for control of water chestnut. J. Aquat. Plant Manage. 45:63 66.

PAGE 106

106 Poovey AG, Getsinger KD, Skogerboe JG, Koschnick TJ, Ma dsen JD, Stewart RM. 2004. Small plot, low dose treatment of triclopyr for selective control of Eurasian watermilfoil. Lake and Reservoir Manage. 20:322 332. control. UF/I FAS Center for Aquatic and Invasive Plants. FWC Task 155, Annual report. Richards JH, Troxler TG, Lee DW, Zimmerman MS. 2011. Experimental determination of effects of water depth on Nymphaea odorata growth, morphology and biomass allocation. Aquat. Botany 95:9 16. Richardson DM, Pysek P, Rejmanek M, Barbour MG, Panetta FD, West CJ. 2000. Naturalization and invasion of alien plants: concepts and definitions. Divers. Distrib. 45:93 107. Sayre, J, 2002. Propagation protocol for American lotus ( Nelumbo lutea ) Willd. USDA NRCS. Schaefer HM, Rolshausen G. 2005. Plants on red alert. Bio. Essays. 28:65 71. Schneider EL, Carlquist S. 2009. Xylem of early angiosperms: Novel microstructure in stem tracheids of Barclaya (Nymphaceae). Aquat. Botany. 96:61 66. Sculthor pe CD. The Biology of Aquatic Vascular Plants. 1967. Edward Arnold Publishers. London, England. Seddon JC, 1981. The control of aquatic weeds with the isopropylamine salt of N phosphonomethyl glycine. Conference on Aquatic Weeds and their Control. pp. 141 148. Senseman SA. Herbicide Handbook 9th edition. 2007. Weed Science Society of America. Lawrence, KS. Simberloff D, Von Holle B. 1999. Positive interactions of nonindigenous species: invasional meltdown? Biol. Invasions. 1:21 32. Simsiman G, Chesters G. 1 975. Perisitence of diquat in the aquatic environment. Water Research. 10:105 112. Simsiman GV, Chesters G. 1975. Persistence of endothall in the aquatic environment. Water, Air and Soil Pollution 4:399 413. Singh SB, Sukumarah KK, Pillai KK, Chakrabarti D C. 1966. Obeservation of efficacy of grass carp in controlling and utilizing aquatic weeds in ponds in Asia. Indio Pacific Fish Council Proceedings, 220 235. Skogerboe JG, Getsinger KD. 2001. Endothall species selectivity evaluations. J. Aquat. Plant Manag e. 39:129 135.

PAGE 107

107 Smits AJ, Van Avesaath PH, Van Der Velde G. 1990. Germination requirements and seed banks of some nympaheid macrophytes: Nymphaea alba L., Nuphar lutea (L.) Sm. and Nymphoides peltata (Gmel.) O. Kuntze. Freshwater Biol. 24:315 326. Smits A, Schmitz G, Van Der Velde G. 1992. Calcium dependent lamina production of Nymphoides peltata (Gmel.) O. Kuntze (Menyanthaceae): implications for distribution. J. Expt. Bot. 43:1273 1281. South Florida Water Management District. Restoring Lake Okeechobee. Ac cessed June 21, 2012. www.sfwmd.gov/portal/page/portal/xweb%20protecting%20and%20restoring/lake%20okeechob ee Sprecher S, Getsinger K, Sharp J. 2002. Review of USACE generated efficacy and dissipation data for the aquatic herbicide formulations Aquathol a nd Hydrothol Vicksburg, MS: US Army Engineer research and Development Center. Steward K, 1991. Light requirements for growth of monoecious hydrilla from the Potomac River. Florida Scientist 54:204 214. Tippery NP, Les DH. 2011. Phylogenetic relationships and morphological evolution in Nymphoides (Menyanthaceae). Systemic Botany. 36:1101 1113. Twilley RR, Brinson MM, Davis GJ. 1977. Phosphorus absorption, translocation and secretion in Nuphar luteum Limnology and Oceanography. 22:1022 1032. USDA. 2010. Ny mphoides peltata (S.G. Gmel.) Kuntze yellow floating heart Retrieved January 5, 2011, from USDA PLANTS profile: http://plants.usda.gov/java/profile?symbol=NYPE USDA. 2012. Weed Risk Assessment for Nymphoides cristata (Roxb.) Kuntze (Menyanthaceae) Crested floating heart. Animal and Plant Health Inspection Service. Raleigh, NC. USDA. 2012. Weed Risk Assessment for Nymphoides peltata (S.G. Gmel.) Kuntze (Menyanthaceae) Yellow floating heart. Animal and Plant Health Inspection Service. Raleigh, NC. USGS. 2001. USGS summary report of nonindigenous aquatic species in U.S. FWS Region 4. USGS. Van TK, Haller WT, Bomes G. 1976. Comparison of the photosynthetic characteristics of three submersed aquatic plants. Plant Physiology. 58:761 768. Van TK, Wheeler GS, Center TD. 1999. Competition between Hydrilla verticillata and Vallisneria americana as influenced by soil fertility. Aquat. Botany. 62:225 233.

PAGE 108

108 Van Der Velde G, Van Der Heijden L. 1981. The floral biology and seed production of Nymphoides peltata (Gmel.) O. Kuntze (menyanthaceae). Aquat. Botany. 10 :261 293 Van Dyke JM, Leslie AJ, Nall LE. 1984. The effects of the grass carp on aquatic macrophytes of four Florida lakes. J. Aquat. Plant Manage. 22:87 95. Weidlich W, 1976. The organization of the vascu lar system in the stems of the Nymphaceae. Am. J. of Bot. 63:499 509. Wersal RM, Madsen JD. 2010. Comparison of subsurface and foliar herbicide applications for control of parrotfeather ( Myriophyllum aquaticum ). Invasive Plant Sci. and Manage. 3:262 267. W estbrooks R, McCord L, Manning S. 2012. EDRR fact sheet on crested floating heart. United States National Early Defense and Rapid Response for Invasive Species. Willey LN, Langeland KA. 2011. Aquatic weeds: Crested floating heart ( Nymphoides cristata ). U niversity of Florida, IFAS Extension. Agronomy Department. Williamson M, 1999. Invasions. Ecography. 22:5 12. Wu Z, Yu W, Tu M, Wang J, Li Z. 2006. Competitive performance of Nymphoides peltata (Gmel.) O. Kuntze growing in microcosm. Hydrobiologia. 527: 4 1 49. Yellow Floating Heart. 2003. Vermont invasive Exotic plant fact Sheet Vermont, USA: Nature Conservancy of Vermont. Zhang KM, Kai S, Yu HJ, Zhan YH, Yu JQ, Xia XJ. 2010. Photoprotective roles of anthocyanins in Begonia semperflorens Plant Sci. 179: 202 208.

PAGE 109

109 BIOGRAPHICAL SKETCH Born in 1988 in Seymour, Indiana, Leif is the son of Michael and Lora Willey. Leif grew up in Seymour, where his father owned a nursery and landscape construction company. He attended Seymour Senior High School and graduated with Academic H onors in June, 2006. After graduating, he pursued his Bachelor of Science degree at Purdue University in West Lafayette, Indiana majoring in fisheries and aquatic sciences. While at Purdue, he was a member of FarmHouse Fraternity and was a ctive in various student organizations. Leif was a member and later president of the Recreational Fishing Club as well as the Purdue University Trap and Skeet Club; he also was a member of the Purdue University Student Chapter of the American Fisheries Soc iety. During his undergraduate career, Leif worked for Aquatic Control, Inc. during the summers assisting the applicato rs and leading vegetation survey crews. Through his summer work experiences here, he developed an interest in aquatic plant management. A fter completion of his degree, Leif moved to Gainesvi lle, Florida to pursue a Master of Science degree in the field of aquatic plant management. In his second year at the University of Florida, Leif co founded the Florida Gators Trap Skeet and S porting Team which became largely successful during their first competitive season. Upon completion of his degree, he plans on continuing to work in the field of aqua tic plant management in the private sector