1 HERBICIDE RELEASE AND PLANT UPTAKE DYNAMICS OF SELECTED GRANULAR AQUATIC HERBICIDES By BRETT WELLS BULTEMEIER A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012
2 201 2 Brett Wells Bultemeier
3 To all those who left my life to o early, you are dearly missed
4 ACKNOWLEDGMENTS First and foremost I am thankful for the grace and love that God has bestowed upon me in this life and for the redeeming power of Jesus Christ I am grateful to my committee for all their help in developing this project and in preparing me for a life away from being a student. I thank Dr. Kane, Dr. MacDonald, Dr. Ferrell and Dr. Haller for not only guiding me in my academic life, but for guidance in many other areas I could not have performed these studies without the gracious funding provided by the Aquatic Ecosystem Restoration Foundation and the Invasive Plant Mini Grant Research Program in the Invasive Plant Bureau of the Florida Fish and Wildlife Conservation Commission. I am thankful to Manchester College for fostering my scientif ic interests and providing a solid academic foundation. They also taught me what it means to be a responsible citizen of this country and of the world. Though small in enrollment, their global impact is large and their graduate s take with them a sense of leadership and responsibility too rare in this world. I am grateful to my church family; your prayers and support have sustained me through some difficult times and have made my life that much better. I am forever grateful to my in laws for entrusting me with their daughter, and truly embracing me into their family. Your love and support never cease to amaze and fill me with joy. My parents have always been a rock of support and encouragement and have instilled in me many of the characteristics that have made my journey to this point possible. You raised me right, and I am forever indebted to you. To my sister, thank you for showing me what real courage is.
5 I am only able to write these words because of my wife. She has been my support in every way imaginable. Thank you for embarking on this journey with me, and keeping me sane throughout. I could never fully express my love for you in words, but know it is there. Finally, I anxiously awai t the birth of our son, I hope his life will be richly blessed, and full of love.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURE S ................................ ................................ ................................ ........ 10 LIST OF ABBREVIATIONS ................................ ................................ ........................... 12 ABSTRACT ................................ ................................ ................................ ................... 13 CHAPTER 1 OVERVIEW ................................ ................................ ................................ ............ 15 2 RELEASE OF FLURIDONE, TRICLOPYR AND ENDOTHALL FROM GRANULES UNDER STATIC AND AERATED CONDITIONS ............................... 22 Introduction ................................ ................................ ................................ ............. 22 Materials and Methods ................................ ................................ ............................ 23 Static Conditions ................................ ................................ .............................. 23 Aerated Conditions ................................ ................................ ........................... 25 Results and Discussion ................................ ................................ ........................... 26 Static Conditions ................................ ................................ .............................. 26 Aerated Conditions ................................ ................................ ........................... 28 3 HERBICID E RELEASE OF FLURIDONE, TRICLOPYR AND ENDOTHALL FROM GRANULES UNDER FLOWING WATER CONDITIONS ............................ 38 Introduction ................................ ................................ ................................ ............. 38 Materials and Methods ................................ ................................ ............................ 39 Low and High Water Flow ................................ ................................ ................ 39 Flur idone Concentration Confirmations ................................ ............................ 40 Results and Discussion ................................ ................................ ........................... 41 Low and High Water Flow ................................ ................................ ................ 41 Fluridone Concentration Confirmations ................................ ............................ 42 4 HERBICIDE RELEASE OF FLURIDONE, TRICLOPYR AND ENDOTHALL UNDER STATIC CONDITIONS IN STERILIZED ORGANIC SEDIMENT ............... 51 Introduction ................................ ................................ ................................ ............. 51 Materials and Methods ................................ ................................ ............................ 53 Soil Organic Matter ................................ ................................ ........................... 53 Herbicid e Release ................................ ................................ ............................ 53 Results and Discussion ................................ ................................ ........................... 55
7 Soil Organic Matter ................................ ................................ ........................... 55 Herbicide Release ................................ ................................ ............................ 55 5 HYDRILLA ROOT AND FOLIAR UPTAKE OF QUINCLORAC, TOPRAMEZONE AND BISPYRIBAC ................................ ................................ ..... 65 Introduction ................................ ................................ ................................ ............. 65 Materials and Methods ................................ ................................ ............................ 68 Results and Discussion ................................ ................................ ........................... 72 6 SUMM ARY ................................ ................................ ................................ ............. 80 Implications ................................ ................................ ................................ ............. 80 Future Research ................................ ................................ ................................ ..... 81 APPENDIX A SEDIMENT STERILIZATION TECHNIQUES ................................ ......................... 84 B ACTIVATED CHARCOAL STUDIES ................................ ................................ ...... 85 LIST OF REFERENCES ................................ ................................ ............................... 86 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 90
8 LIST OF TABLES Table page 1 1 Characteristics of selected aquatic herbicides. ................................ ................... 21 1 2 Gr anular products used in herbicide release studies. ................................ ......... 21 2 1 Estimated time (days) required for 25, 50 and 95 percent of fluridone to release from granules maintained under static conditions ................................ .. 31 2 2 Estimated time (hours) required for 25, 50, and 95 percent of triclopyr and endothall to release from granules maintained under static conditions. ............. 31 2 3 Estimated time (days) required for 25, 50 and 95 percent of fluridone to release from granules maintained under aerated conditions. ............................. 31 2 4 Estimated time (hours) requi red for 25, 50, and 95 percent of triclopyr and endothall to release from granules maintained under aerated conditions. .......... 32 3 1 Es timated time (days) required for 25, 50 and 95 percent of fluridone to release from Q and SRP granules maintained under low and high water flow conditions. ................................ ................................ ................................ .......... 45 3 2 Estimated time (hours) required for 25, 50 and 95 percent of triclopyr and endothall to release from granules maintained under low and high water flow conditions. ................................ ................................ ................................ .......... 45 3 3 Total percent of herbicide recovered from fluridone, triclopyr and endothall treatments after grinding the granules, and liquid fluridone maintained under low and high water flow c onditions.. ................................ ................................ ... 46 4 1 Estimated time (days) required for 25, 50, 95 percent of fluridone, triclopyr and endothall to release from gr anules maintained under static conditions applied over a 5 cm layer of sterilized Bivens Arm Lake sediment. .................... 59 5 1 Concentration of quinclorac recovered from root and shoot material of foliar and root zone treated hydrilla plants 7, 14 and 21 d after treatment with 400 ppb of quinclorac maintained under 16 h photoperiod at 26 C. .......................... 77 5 2 Percent of recovered 14 C quinclorac present in hydrilla, treatment water and sediment barriers 7, 14 and 21 d after treatment with 400 g a.i. L 1 to plan ts maintained under 16 h photoperiod at 26 C. ................................ ...................... 77 5 3 Concentration of topramezone recovered from root and shoot material of folia r and root zone treated hydrilla plants 7, 14 and 21 d after treatment with 50 ppb of topramezone maintained under 16 h photoperiod at 26 C. ................. 78
9 5 4 Percent of recovered 14 C Topramezone present in hydrilla, treatment water and sediment barriers 7, 14 and 21 d after treatment with 50 g a.i. L 1 to plants maintained under 16 h photoperiod at 26 C. ................................ ............ 78 5 5 Concentration of bispyribac recovered from root and shoot material of foliar and root zone treated hydrilla plants 7, 14 and 21 d after treatment with 400 ppb of bispyribac maintained under 16 h photoperiod at 26 C. .......................... 79
10 LIST OF FIGURES Figure page 2 1 Release of fluridone from granules maintained under static conditions for 90 days. Symbols and error bars represent mean standard error of the mean. ... 33 2 2 Fluridone concentration in water of 19 L containers treated with PR and One maintained under static conditions. Granules were crushed at the beginning of the experiment to determine potential degradat ion in the experimental units. ................................ ................................ ................................ ................... 34 2 3 Release of fluridone from granules maintained under aerated conditions for 60 days .. ................................ ................................ ................................ ............. 35 2 4 Release of triclopyr from granules maintained under static and aerated conditions for 25 and 4 days, respectively.. ................................ ........................ 36 2 5 Release of endothall from granules maintained under static and aerated conditions for 30 and 5 days, respectively.. ................................ ........................ 37 3 1 Experimental apparatus for the water movement experiments in which water velocities can be regulated by variable speed ele ctric peristaltic pumps (not shown) from 0.000016 to 0.0016 KPH. 44 3 2 Concentration of fluridone released from Q granules exposed to low (0.000016 KPH) and high (0.0016 KPH) water flow in a darkened room at 20 C.. ................................ ................................ ................................ ....................... 47 3 3 Concentration of fluridone released from SRP granules exposed to low (0.000016 KPH) and high (0.0016 KPH) water flow in a darkened r oom at 20 C.. ................................ ................................ ................................ ....................... 48 3 4 Concentration of triclopyr released from granules exposed to low (0.000016 KPH) and high (0.0016 KPH) water flow in a darkened room at 20 C. ............ 49 3 5 Concentration of endothall released from granules exposed to low (0.000016 KPH) and high (0.0016 KPH) water flow in a darkened room at 20 C. ............... 50 4 1 Concentration of fluridone released from Q granules treated over 1 5 L of water in a plastic pail, maintained under static conditions with a 5 cm thick layer of organic sediment in a darkened room at 20 C.. ................................ ..... 60 4 2 Concentration of fluridone released from SRP granules treated over 15 L of water in a plastic pail, maintained under static conditions with a 5 cm thick layer of organic sedime nt in a darkened room at 20 C. ................................ ...... 61
11 4 3 Concentration of fluridone remaining from Genesis (liquid) in 15 L of treated water in a plastic pail, maintained under static conditions with a 5 cm thick layer of organic sedi ment in a darkened room at 20 C. ................................ ...... 62 4 4 Concentration of triclopyr released from granules treated over 15 L of water in a plastic pail, maintained under static conditions with a 5 cm thick layer of organic sediment in a darkened room at 20 C.. ................................ .................. 63 4 5 Concentration of endothall released from granules treated over 15 L of water in a plastic pail, maintained under static conditions with a 5 cm thick layer of organic sediment in a darkened room at 20 C.. ................................ .................. 64
12 LIST OF ABBREVIATION S a.i. Active ingredient a.e. Acid equivalent DAT Days after treatment ET xx Estimated time for given value (i.e. 25, 50, 75) percent to be achieved HAT Hours after treatment PPB Parts per billion PPM Parts per million
13 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy HERBICIDE RELEASE AND PLANT UPTAKE DYNAMICS OF SELECTED GRANULAR AQUATIC HERBICIDES By Brett Wells Bultemeier May 2012 Chair: William T. Haller Cochair: Greg E. MacDonald Major: Agronomy Granular formulations of aquatic herbicides have been utilized in weed management programs for many years. However, the most basic questions such as how long it takes for a herbicide to be released from granules have been largely unexplored. The release of herbicide from the granule is critical for ensuring that concentrat ions in the water reach the critical CET (concen tration exposure time) required for weed control The cu rrent studies were conducted to determine the rate of herbicide release from selected aquatic granules maintained under static and aerated water condit ions, under known water flow conditions and under static conditions when placed over an organi c sediment Finally, the uptake of herbicides by roots and shoots was compared when herbicides were applied only to the root zone or foliage of hydrilla. Under static conditions the release of herbicide s varied widely. T riclopyr granules requi red the least amount of time (11 hours) to achieve 50% release c ompared to SRP granules which required 72 days to achieve the same release. The aeration of water greatly d ecreased the time of release where only 1.1 hours and 7 days were required to release 50% of triclopyr and fluridone from the granules mentioned above, respectively.
14 A comparison of fluridone re lease from four granular formulations showed that Q released fluridone much faster than SRP. However, there wa s no difference in release from Q, PR and ONE early in the release experiments, and no difference in release from SRP, PR and ONE towards the conclusion of the studies. Under flowing water conditions, the lowest flow tested was not different in release than the static conditions, but the fastest flow was different and more closely resembled the aerated conditions. Endothall and t riclopyr herbicide release from granule s was reduced significantly by placing them over an organic sediment, but 100% release was achieved, whereas fluridone (both formulations) never achieved >25% release. The herbicide s quinclorac topramezone and bispyribac were all absorbed by both hydrilla roots and shoots and were readily t ranslocated throughout the plant, but the highest concentrations for all 3 herbicides occurred when plant foliage was treated The highest amount of absorption was noted for hydrilla plants foliar treated with bispyribac, in which concentrations inside th e plant were >100x higher than the concentration in the water. The use of activated charcoal was also shown to be effective as a barrier between the two water sources, but further adjustments of its use may be necessary. These studies show very different herbicide release profiles between flowing and static water, as well as when granules are applied to high organic matter sediments. Foliar uptake of the herbicides evaluated was much greater than root uptake, but translocation throughout the plant was fo und regardless of treatment location.
15 CHAPTER 1 OVERVIEW Hydrilla ( Hydrilla verticillata (L. F .) Royle ) is a submersed aquatic vascular plant native to Southern Asia that is invasive in the United States and responsible for a multitude of environmental and economic problems in fresh water eco systems. It was discovered in 1959 in Florida and has spread throughout the southeastern U.S. west to California and to several states in the northeast (Blackburn et al. 1969; Lazor 1978 ; Les et al. 19 97 ). Hydrilla has both dioecious and monoecious biotype s The dioecious female biotype is dominant in the southeast U.S. while the monoecious is dominant in the northern half of the country (Cook and Luond 1982; Langeland 1996). Rapid growth, thick cano py formation, and reproduction by vegetative propagules (turions and tubers) make this plant very challenging to control (Langeland 1996; Van and Vandiver 1992). In Florida alone, 40,000 ha of water, representing 43% of public lakes, were infested with hy drilla in the early 1990 s (Langeland 1996). H ydrilla control is limited to only a few registered aquatic herbicides. Management strategies include the use of fast acting, short half life contact herbicides such as diquat (6,7 dihydrodipyrido[1,2 a :2`,1` c ]pyrazinediium) or endothall (7 oxabicyclo[2.2.2]heptanes 2,3 dicarboxylic acid) and slow acting longer half life enzyme inhibiting herbicides such as fluridone (1 methyl 3 phenyl 5 [3 (trifluoromethyl)phenyl] 4(1 H ) pyridinone ), penoxsulam (2 (2,2 difluoroethoxy) N (5,8 dimethoxy[1,2,4]triazolo[1,5 c ]pyrimidin 2 yl) 6 (trifluoromethyl)benzenesulfonamide) and bispyribac sodium (sodium 2,6 bis[(4,6 dimethoxy 2 pyrimidinyl)oxy]benzoate) Herbicides used in aquatic plant management are usually either a liquid formulation or as solid, such as a granule or pellet. Granules are formulated by mixing liquid herbicide
16 to a dr y carrier (most commonly clay but polymers and other carriers are used) Conversely, p ellets are a mixture of liquid herbicid e and a liquid form of the carrier (typically, clay, but some p olymers are utilized) combined to form This is extruded dried, and cut into the desired size. The exact method and specifications of the manufacturing process are proprietary, so information is scarce or largely nonexistent. Hereafter the term granule will be used to generally describe any solid when a more specific designation is required. F luridone is sold in liquid form, and in a variety of granules and pellets which are mar keted as either quick releasing or slow releasing. Although these, and other granular herbicides, have been used successfully in aquatic plant management, little e ffort has been invested in determining the release characteristics and the impact of environmental variables or herbicide movement off of these granular formulations The physical and chemical characteristics of most aquatic herbicides in the environment a re well documented (Table 1 1 ). Previous research has determined herbicide dissipation and residue stability from liquid treatments in the water (Frank and Comes 1967), or focused on the interaction between concentration and exposure time to plants (Nethe rland and Getsinger 1995; Netherland et al. 1991). Previous granular herbicide release studies utilized various methods to stir or mix the water (Mossler et al. 1993; Netherland and Stewart 1994; Van and Steward 1986). For example, Mossler et al. ( 1993 ) used granular tubes that were agitated on a linear shaker. Their results showed that fluridone was
17 b oth products released herbicide the same under vigorous agitation. V an and Steward ( 1986 ) maintained experimental the water was gently stirred just prior to taking samples for analysis Although t he se studies are useful in showing the total amount that can be released, they do not provide quantitative information on the water movement and how it affects the rate at which the herbicide is released. Registrants that produce granular herbicides often m arket their or no independently published information regarding the release rate. T he release characteristics of aquatic herbicides are important because successful pla nt management is dependent upon achieving a lethal dose of herbicide in the water for the proper length of time in order to kill the targeted plant (Netherland et al. 1991). For instance hydrilla requires an exposure of either 2 mg a.e. (acid equivalent) L 1 of endothall for 48 hours or 3, 4, or 5 mg a.e. L 1 endothall for 24 hours to achieve greater than 85% reduction in plant biomass (Netherland et al. 1991). However, in response to the herbicide fluridone, hydrilla requires an exposure of greater than 12 g a.i. (active ingredient) L 1 of fluridone for more than 30 days to achieve a similar reduction in biomass (Netherland et al. 199 5 ). In order to know if this critical level will be reached after treatment with a granular herbicide it is important t o know how quickly the herbicide desorbs from the granule and dilutes into the water column and how environmental factors may affect this process For instance, a fast acting, short half life contact herbicide, would likely need to release quickly in orde r to ensure that the proper concentration is reached for the relatively short amount of exposure time required for control. However, it might b e desirable to have a slower releasing granule for a slow
18 acting, enzyme inhibiting herbicide like fluridone, so that a lower concentration is maintained over a longer period of time. Much is known about current aquatic herbicide dilution and degradation in the water, so an understanding of the release characteristics of these herbicides would make concentration/ex posure predictions more reliable One justification for applying granules in aquatic environments is that the granules sink to the bottom of the lake and thus deliver herbicide to the root zone of submersed plants. Although this might be desirable for herbicides that are absorbed by plant roots and translocated to the rest of the p lant, it introduces another conf ounding factor to herbicide release from granules. Granules that sink to the bottom of a water body will directly contact and interact with the sediment in a different manner than l iquid treatments A granule treatment would potentially create a high concentration of herbicide at the bottom of the lake vs. a high concentration of herbicide in the upper water column associated with surface applied l iquid treatment s Some herbicides have a high affinity and bind ti ghtly to soils, whereas others have less affinity to clay and organic matter. The potential binding of the herbicide to sediments could greatly reduce the amount of herbicide available for plant uptake. Some pr edictions can be made about herbicide/soil interaction based on Koc values. This value indicates how likely a herbicide will be adsorbed based upon the total organic carbon content of that soil. The lower the Koc value, the lower the potential for bindi ng to a soil with high organic carbon and conversely a higher value means a herbicide is more likely to bind to that same soil. Fluridone has a moderate high Koc (350 2500 mL/g) suggesting a higher potential for binding to organic sediments and potential loss of the herbicide (Reinert 1989). The Koc of t riclopyr is
19 intermediate at 25 144 mL/g (Kollman and Segawa 1995) and endothall is moderate as well with a Koc of ~100 mL/g (Senseman 2007) It is possible however, that for those herbicides not binding t o the soil that there is still a sediment/water interface interaction that would slow the release of herbicide from the granule into the water column if the granule sinks into the sediment. The objectives of this research are to quantify the release of h erbicides from various granular formulations under 1) static conditions, 2) flowing water conditions and under 3) static conditions with a sediment layer Additionally the uptake of radioactively lab e led aquatic herbicides by hydrilla will be evaluated t o determine upta ke dynamics by roots and shoots. A list of products used, and the information of the registrants can be found in Table 1 2. From this point forward all herbicides will be referred to by the common name, or as otherwise specified in this t able. The results of th is research will fill a much needed gap in the knowledge of herbicide release from aquatic granules but in reality only provides a base line release profile for these herbicides In addition, soil interaction and water flow impact on herbicide release will allow more informed decisions on how and when to use granular formulations in management strategies. Additionally an understanding of hydrilla herbicide uptake and translocation of herbicides is crucial for the development of the newer herbicides that are currently being developed. Understanding the behavior of these herbicides in the aquatic environment allows managers to avoid treatment strategies that are likely to lead to reduced or no control, and save money by avoiding costly trial and error treatments. The results of this
20 research will allow these decisions to be more informed, and will hopefully result in more effective and cost efficient treatment strategies in the future.
21 Table 1 1. Characteristics of selected aquatic herbicides. Name Mode o f Action Water Half l ife Water Solubility a Koc b Citation Bispyribac ALS inhibitor c <10 d d 73.3 852 1793 Senseman 2007 Endothall Not classified 7 d 100.0 20, 110 138 e Senseman 2007 Fluridone Carotenoid inhibitor 9 20 d 0.012 350 2500 Senseman 2007 Quinclorac Synthetic auxin 21 d 0.062 15 54 Senseman 2007 Miron et al. 2005 Nufarm 2009 Topramezone Carotenoid inhibitor 78 d 15 22 172 Senseman 2007 Triclopyr Synthetic auxin 3 d 0.430 25 144 Senseman 2007 Kollman and Segawa 1995 Woodburn et al. 1993 a g L 1 b mL g 1 c acetolactate synthase d in flooded and upland conditions e pH dependent Table 1 2. Granular products used in herbicide release studies. Common name Trade name As used in this document Fluridone Sonar Q 1 Q Sonar PR 1 PR Sonar ONE 1 ONE Sonar SRP 1 SRP Triclopyr Renovate OTF 1 Triclo p yr Endothall Aquathol Super K 2 Endothall 1 SePRO Corporation. Carmel, IN. 46032. 2 United Phosphorus Inc. King of Prussia, PA. 19406.
22 CHAPTER 2 RELEASE OF FLURIDONE, TRICLOPYR AND ENDOTHALL FROM GRANULES UNDER STATIC AND AERATED CONDITIONS Introduction Granular 2,4 d formulations have been used for aquatic weed control since the early 1950s (Oborn et al. 1954) Surprisingly few published studies have been conducted that describe the impact of water flow on herbicide release characteristics. The release of herbicides from granules depends upon the type of carrier (typically clay) and the manner in which the herbicide is mixed with the carrier. Mossler et al. 1993, cites a personal communicatio n with D. Tarver that states the slower release of of Mossler et al. (1993) showed t hat there were in fact significant differences in release of fluridone between the two pellets under mild agitation, but that the magnitude of difference was decreased under vigorous agitation. Complete release of fluridone from the mild agitation occurre d after ~400 h, compared to the vigorous agitation in which 100% release of the herbicide in both formulations occurred after about 20 h. These results clearly show that water movement (agitation) influences herbicide release from pellet formulations. Se veral other studies have been performed that in some way explored the release of herbicides from aquatic granules. Reinert et al. (1985 ) examined the release of endothall from a granular clay formulation under agitation via a shaker table. Under water, they found that ~50% of the herbicide had released by 3 4 h. Although these studies were performed on a formulation that is not sold today, it does provide insight into the speed at which
23 herbicide can release be noted, that like other studies, these authors defined this mild agitation as being Perhaps the most in depth study of herbicide release from granules to date was that performed by R. E. Wilkinson (1964) on various formulations of 2,4 D. In this study the effect of carrier (clay type), temperature, sediment presence formulation (2,4 D species) and pH were all tested. Wilkinson found that all these factors interacted to create unique release profiles, and in some cases a lack of release ( < 100% recovery). The methods described do not imply any water movement was present, and as such, these could truly be static studies. In these studies 100%, release could be as qui ck as several hours, to more than 1 week, if release was achieved at all. The only possible limitation of this study was the utilization of cucumber root bioassay. It is possible, particularly in the pH study, that 2,4 D released from the granule, but it s form changed due to water chemistry, and the assay could not detect this. This is actually a similar problem experienced in our own preliminary studies, which removed 2,4 D as a candidate for experimentation due to the lack of a quick and reliable metho d to analyze herbicide residues. The objectives of this study were to 1) determine the release of aquatic granular herbicides under static conditions 2) gentle aeration. Materials and Methods Static Conditions Approximately 500 mg of fluridone Q, PR, ONE and SRP, 650 mg of endothall, and 900 mg of triclopyr granules were applied over the top of 15 L of tap water in 19 L
24 plastic bucket s resulting in a theoretical concentration of ~ 2, 19 and 8 mg active ingredien t (a.i.) L 1 were not broken or otherwise altered to achieve exact concentrations, but granules applied to each bucket were weighed. The treatments were each replicated four times. The tops of the buckets were sealed with clear plastic wrap secured with a rubber band (to prevent evaporation) and kept inside a darkened growth room maintained at 20 1 C. The buckets were further covered with black plastic garbage bags to prevent any exposu re to ambient light. Twenty ml water samples were collected periodically over ~30 d for en d o thall and triclopyr and ~ 90 d for fluridone. Samples were taken from the middle of the buckets halfway down the water column and placed in plastic scintillation vi als which were frozen for later chemical analysis T he volume of water that was removed (20 ml) was replaced a fter each sampling period with fresh water in order to maintain a constant volume of 15 L. After the final water sample ( ~ 30 or 90 d) was collected the granules were crushed using a PVC pipe with a metal flange attached to the end. Three separate grindings were performed with a 2 5 minute period in between each grind and a 5 minute period before a final 20 ml water sample was collected. This crushed value was compared to the theoretical value (based upon the exact weight of the granules applied to each bucket) to determine the total amount of herbicide present on or in the granules that were applied to each bucket. All water samples were converted to a percentage found in the water calculated upon the initial we ight applied which represented 100%.
25 Four replications each of PR and ONE were also included, in which the granules were crushed at the beginning of the experiment and sampled over 9 0 d This study was designed to determine what, if any, degradation occurred under these experimental parameters as well as to ascertain if the grinding process resulted in 100% release of the herbicide Water sample collection and all ot her procedures were identical to the other granules. Herbicide concentrations in the water samples were determined by the ELISA method of immunoassay. Endothall and triclopyr were analyzed utilizing kits manufactured by Strategic Diagnostic Inc 1 and fl uridone kits were manufactured by Envirologix 2 The limit of quantification (LOQ), or the lowest concentration of a herbicide that will yield a positive test result, is 7 g a.i. L 1 0.1 g a.i. L 1 and 0.08 g a.i. L 1 for endothall, triclopyr and flur idone ELISA kits respectively. Aerated Conditions In a companion study, an additional factor was introduced which involved gentle water movement Water movement, or flow, was pr ovided by the addition of aquarium pump s delivering a flow rate of 60 mL min ute 1 of air through an airstone secured halfway down the side of the buckets. In preliminary tests, t his flow rate was sufficient to distribute dye throughout the water column within 5 to 10 minutes, but was not strong enough to disturb the granules. Th e precise amount of water movement cause by this aeration was not determined. All methods for this study were identical to the static experiments, except for the addition of the water movement, caused by the aeration. 1 Strategic Diagnostics Inc. Newark, DE. 1 9702. 2 Envirologix. Portland, ME. 04103.
26 Non linear regression analysis was used to analyze herbicide release of all formulations. The equation used is an exponential rise to max which provides the generalized equation is y= a*(1 e bx ). This equation is the most fitting because the maximum rele ase expected from the granules is 100%. A range of ET (estimated time required to release 25, 50, 75 and 90 % of the herbicide) values were calculated for each herbicide from the regression equation. Results and Discussion Static Conditions The fit observed for the regression analysis of all four fluridone granules under static conditions was greater tha n 0.87 indicating that the regression equation is a good prediction of the release of fluridone from these granules (Figure 2 1). The release of flu ridone from the granules was generally in three categories fast intermediate and slow Q (fast) and SRP (slow) were the most different from each other, with the ONE and PR formulations falling between these two extremes. The ET (estimated time) values for 25, 50, 95% release are presented for the four fluridone formulations in Table 2 1 The ET 25 values for all four fluridone formulations were between 12 and 24 d with the only difference being between Q and SRP granules. ONE and PR are similar to each other and Q throughout the study, and only differ from SRP at the ET 50 In some instances 100% release of fluridone was not achieved during the 90 d sampling period, but after being crushed ~ 100% was recovered (with the exception PR) and suggests that the unreleased herbicide was still bound to the granule even after 90 d (Table 2 1 ). The Q and SRP granules have different release profiles under the static conditions ; Q is released much more rapidly than SRP PR and One do not differ from
27 each other and ar e similar to both Q and SRP depending on which point of the regression are being compared The reason for the low amount of recovery for PR is unknown, but is unlikely due to degradation because fluridone is broken down by UV light, and light was excluded from these studies. The two sets of granules (PR and One) that were crushed at the beginning of the studies and sampled over time showed no apparent fluridone degradation (Figure 2 2 ). It is possible, when selecting a limited number of granules from the formulated product that they may not have contained 5% active ingredient, but as noted later, the same low yield occurred in the aerated studies, Triclopyr and endothall granules had a much quicker release compared to that of all fluridone formulations (Table 2 1 ). Triclopyr granules had 50% release by 12 h and 9 5 % at 3 7 h after treatment (Table 2 2) In contrast, release of endothall w as slower, 50% by 3.5 d and 9 5 % by 10 d ( Table 2 2 ). The total recovery of her bicide for triclopyr and endothall gra nules under static conditions was between 95 110% ( Table 2 2 ) These results show the tremendous diversity of release of herbicides from several aquatic granules, under static conditions, where ET 50 values ranged from as short as 12 h to as long as 72 d. The release of aquatic herbicides from granules was expected to differ. Both endothall and triclopyr are rapidly taken up by plants, requiring concentrations of 1 2 ppm or higher in the water for ad equate weed management. Control with these two products requires exposure times of several hours to at most a couple of days. In contrast, fluridone concentration and exposure require ments are ~10 ppb for 40 90 d The aqueous half lives (persistence) of the s e herbicides are also much different, 2 10 d
28 for endothall and triclopyr but 9 20 d for fluridone. Although th ere are few studies investigating granular release of herbicide under static conditions, fluridone has been the most researched. Koschnick et al. (2003) noted that the release of S RP under static, but outdoor mesocosm, conditions to be 25 36% after 36 d. For our study, SRP released 36% by 36 d (calculated from equation in Figure 2 1). The general agreement between these values suggests that both Koshnick et al. (2003) and this study represent the release profile for fluridone that could be expected under static conditions. Reinert et al. (1985) reported that the time to get ~50% of endothall from the granules they tested was between 3 4 h. This is considerably faster than the 86 h observed in our study (Table 2 2). This is likely due to the Reinert et al. (1985) study including a Aerated Conditions The study of Sonar 5P and SRP pellets conducted by Mossler et al. (1993) that water movement may have an impact on herbicide release. Because of this, a similar study with slight water movement, provided by aeration, was undertaken. Adding water movement (aerated studies) compared to the static release, greatly altered the release profiles of all the products Similar to the static studies the release characteristics of Q and SRP were statistically different However, under aerated conditions release from Q was the slowest release and SRP the fastest (Figure 2 3). Fluridone granules changed from ET 50 values of 27 72 d (Table 2 1) under static conditions to values ranging from 4 16 d un der aerated (water movement) conditions (Table 2 3) T riclopyr and endothall granules were similarly affected by water movement Both triclopyr and endothall granules released 100% of herbicide at the
29 conclusion of the studies, but that release was much quicker under aerated than static conditions (Figure 2 4 and 2 5). The time required for 95% release of triclopyr f rom the granule was reduced 9 fold from 37 to 4.2 h and endothall was reduced 70 fold 237.5 to 3.4 h (Table 2 4). Both triclopyr and endoth all treatments were similar to each other (from about 0.5 h to 3 h to go from 25 90% release for both herbicides) under aerated conditions SRP required 16 d for 9 5 % of the fluridone to release under aerated conditions in general agreement with the results of Mossler et al. ( 1993 ) who found that SRP under mild agitation required 20 d for ~100% release of f luridone Reinert et al. (1985) reported that the endothall granules tested under their parameters (flasks on a shaker table) released 50% in ~3 4 h. This is actually slower than the ~1 h required under the aerated conditions (Table 2 4), but is more similar than the time necessary under static conditions, 86 h (Table 2 2). The similarities in these data appear to confirm that the cause of incr eased release in our study is also due to water movement, because both studies show a similar release profile under conditions of water movement. These studies demonstrate the wide range of release profiles found among solid aquatic formulations. To achie ve 50% release required as little as 12 h (triclopyr) or as long as 72 d (SRP) under static conditions and 1 h (endothall) or 16 d (Q) to achieve the same release under aerated conditions. This demonstrates that each herbicide and granule combination hav e unique release profile s into the water column. The addition of minimal water movement greatly increase s the speed of herbicide release from endothall, flu ridone and triclopyr granules. The different fluridone granules divided into three categories of release, fast, intermediate and slow, where the slow and
30 fast were quite different from each other. Each herbicide had a unique release profile, and even the same herbicide (fluridone) differed depending on the type of used for the granule. The total recovery of herbicide at the conclusion of these studies was ~100% of what was theoretically applied and the study where granules were crushed at the beginning of the study maintained concentrations over the duration of the experiment. These data suggest that there is no degradation of herbicide in this experimental design.
31 Table 2 1. Estimated time (days) required for 25, 50 and 95 percent of fluridone to release from granules maintained under static conditions Herbicide ET 25 ET 50 ET 95 Crushed Value a Q 12 (9 15) b 27 (21 33) 68 (48 88) 108 10 c ONE 14 (11 17) 39 (33 46) N/A d 83 4 PR 12 (10 15) 37 (31 42) N/A 78 2 SRP 24 (16 32) 72 (50 93) N/A 100 11 a Mean percent of total concentration recovered after crushing granules. b Mean (95% Confidence Interval). c Mean 95% Confidence Interval d N/A= Value could not be calculated from the regression equation, because total release did not attain 95%. Table 2 2. Estimated time (hours) required for 25, 50, and 95 percent of triclopyr and endothall to release from granules maintained under static conditions. Herbicide ET 25 ET 50 ET 95 Crushed Value a Triclopyr 5.1 (4.5 5.9) b 11.9 (10.5 13.8) 37.0 (32.6 42.9) 104 9 c Endothall 37.7 (31.6 46.7) 86.2 (72.0 106.7) 237.5 (199.1 294.3) 108 15 a Mean percent of total concentration recovered after crushing granules. b Mean (95% Confidence Interval). c Mean 95% Table 2 3. Estimated time (days) required for 25, 50 and 95 percent of fluridone to release from granules maintained under aerated conditions. Herbicide ET 25 ET 50 ET 95 Crushed Value a Q 7 (5 8) b 16 (12 21) N/A c 96 6 d ONE 2 (1 4) 6 (3 9) N/A 90 10 PR 2 (1 2) 4 (2 6) 14 (9 17) 78 8 SRP 3 (2 3) 7 (5 8) 16 (10 23) 99 10 a Mean percent of total concentration recovered after crushing granules. b Mean (95% Confidence Interval). c N/A= Value could not be calculated from the regression equation, because total release did not attain 95%. d Mean 95%
32 Table 2 4. Estimated time (hours) required for 25, 50, and 95 percent of triclopyr and endothall to release from granules maintained under aerated conditions. Herbicide ET 25 ET 50 ET 95 Crushed Value a Triclopyr 0.5 (0.4 0.6) b 1.1 (0.9 1.5) 4.2 (3.4 5.7) 110 10 c Endothall 0.4 (0.3 0.5) 0.9 (0.7 1.2) 3.4 (2.7 4.6) 95 10 a Mean percent of total concentration recovered after crushing granules. b Mean (95% Confidence Interval). c Mean 95%
33 Figure 2 1. Release of fluridone from granules maintained under static conditions for 90 days. Symbols and error bars represent mean standard error of the mean.
34 Figure 2 2. Fluridone concentration in water of 19 L containers treated with PR and One maintained under static conditions G ranules were crushed at the beginning of the experiment to determine potential degradati on in the experimental units Symbols represent mean standard error of the mean.
35 Figure 2 3. Release of fluridone from granules maintained under aerated conditions for 60 days Symbols and error bars represent mean standard error of the mean.
36 Figure 2 4. Release of triclopyr from granules maintained under static and aerated conditions for 25 and 4 da ys, respectively. Symbols and error bars represent mean standard error of the mean.
37 Figure 2 5. Release of endothall from granules maintained under static and aerated conditions for 30 and 5 days, respectively. Symbols and error bars represent mean standard error of the mean.
38 CHAPTER 3 HERBICIDE RELEASE OF FLURIDONE, TRICLOPYR AND ENDOTHALL FROM GRANULES UNDER FLOWI N G WATER CONDITIONS Introduction Although water movement causes much faster herbicide release from solid herbicide formulations no quantification of release due to known rat es of water movement has been reported It is important to determine if water movement alone, or conditions. Ultimately an understanding of which conditions prevail in a lake, static or water movement, is necessary to know what release profile will exist when the water is treated with soli d herbicide formulations Water currents in reservoirs, canals and rivers where aquatic herbicides are used certainly occur, but water movement at the water/soil interface is not easily measured or reported. There is no doubt that water movement, even in closed lake and pond systems, occurs due to wind mixing and thermal gradients. Lemmin and Mortimer (1986) discuss ed the impact that w ind currents can have on water movement in lakes, specifically focusing on the seiche effect. This occurs when sustained wind over a large area (surface of a lake) can actually build a large amount of water on one side of the water body, and as the wind s peed decreases Clearly, this causes a large amount of water movement that is not obvious from the surface of the lake. Furthermore, temperature gradients present in lakes with a thermocline, and any difference in wate r temperature, cause eddy diffusions which cause extensive water movement from upper waters to lower waters and vice versa (Quay et al. 1980). A determination of the minimum water movement necessary to alter
39 herbicide release from solid formulations would be helpful in making release profile predictions in a lake environment. The objective of this experiment was to determine the effect that a 100 fold difference in water velocities have on herbicide release from fluridone, triclopyr and endothall granu lar formulations. Materials and Methods Low and High Water Flow Water flow rates of low (0.000016 KPH (kilometers per hour)) and high (0.0016 KPH) were produced through a 5. 1 cm diameter P VC pipe, a total of 40 cm in length, with a t joint 30 cm from the elbow (Figure 3 1). T he total volume of the pipe was 1 L The pipe was secured such that it was submerged just below (pipe completely filled with approximately 2mm of water overlying the filled pipe) the water surface of a glass aquarium measuring 50 x 2 5.5 cm which was lined with black plastic and filled to a depth of 23.5 cm with deionized water ( 30 L of water) The top of the aquariums were covered with plastic to minimize water loss due to evaporation. All studies were conducted in a darkened growth room maintained at 20 1 C. Fisher Scientific variable speed, low flow, peristaltic pump s 1 were used to achieve the low flow and Greylor pump s with a voltage controlled power pack 2 was used to produce the high flow. The intake tube of the pump was attached to the middle of the bottom of the aquarium and the discharge tube was secured 2 mm below the surface of the water inside the elbow joint of the PVC pipe It was possible to estimate 1 Fisher Scientific Variable Flow Peristaltic Pump, Low Flow Catolog# 13 876 1. Thermo Fisher Scientific. Pittsburg, PA. 15275. 2 Greylor Company Model RF 100 12/24 Peristaltic Pump. Greylor Company. Cape Coral, FL. 33909.
40 water exchange and velocity based upon the volume of water flow and the total volume of the PVC assembly. The flow rates provided a complete exchange of water in the pipe once, or 100 time s in a 24 h period for low and high flow rates, respectively. Water samples (20 ml) were collected at various times from the midd le of the aquarium tank halfway down the water column, frozen and analyzed Approximately 550, 550, 900 and 575 mg of Q, SRP triclopyr and endothall respectively were utilized for these treatments These treatments resulted in a concentration of ~ 0.9, 4.2 and 12.1 mg a.i. L 1 for fluridone, triclopyr and endothall respectively. All treatments were replicated four times and sampled over 70, 6, and 10 d for fluridone, triclopyr and endothall, respectively All samples were analyzed with ELISA and non linear regression was utilized to calculate ET values. Fluridone Concentration Confirmations Since fluridone release from granules occurs over a longer period than triclopyr and endothall and is more likely t o bind to organic material this experiment was established to determine if the components (tubing, plastic liner, aluminum brackets, and pvc pipe) in the test system would adsorb fluridone from the water column. Three replications each of the low and hig h flow systems were included that were treated with liquid fluridone (Sonar Genesis 3 ) at a concentration of 976 g a.i. L 1 These treatments were sampled at 3 and 10 d after treatment and the concentrations analyzed. 3 SePRO Corporation. Carmel, IN. 46032.
41 Results and Discussion Low and High Water Flow As hypothesized, based upon the results of Chapter 2, the release of herbicide from all granules was more rapid under high flow conditions compared to low/no flow. Fluridone released quicker from both Q and SRP granules under the high flow con ditions, compared to the low flow (Figure 3 2 and 3 3) however, f ew comparisons can be made between the two formulations, because the overall release was incomplete ( few ET values could be calculated ) The release of 25% of applied fluridone under high fl ow was the fastest in Q (19.7 d ), and was different from the low flow Q as well as SRP (Table 3 1). No values could be calculated for SRP low flow, due to low release, but the high flow values were not different from low flow values for Q, 81.0 ( 54.3 160) and 43.8 (27 109) d to achi eve 25% release, Q and SRP respectively. Triclopyr granules also had a different release profile under low flow than under high flow (Figure 3 4). Under high flow conditions, the release of triclopyr was much more rapid than under low flow. In contrast to fluridone, the recovery of triclopyr at the conclusion of the study was ~100% (Table 3 3). Similar to fluridone, the release of triclopyr was more rapid under high flow conditions at all times compared (Table 3 1). The release of endothall from granules was very similar under both low and high flow conditions (Figure 3 5). The speed of endothall release in low and high water flow were not significantly different from each other, with the exception of ET 95 values (Ta ble 3 2). The recovery of triclopyr and endothall after grinding and thoroughly mixing at the conclusion of the study was complete, between 95 108%, which is similar to the recovery reported for static and aerated experiments (Table 3 3). There was
42 incomp lete recovery of fluridone however, with the highest value being 57% of the applied concentration (Table 3 3). This low recovery for fluridone was not expected, as recovery in static and aerated conditions was essentially 100% of the applied dose. Flurido ne Concentration Confirmations The recovery of fluridone concentrations from the liquid treatments placed in the water flow system s was also poor Only 51 and 48 % of the applied fluridone was recovered after 10 d under low and high flow conditions, respec tively (Table 3 3) Although the glass aquariums were lined with plastic, to avoid herbicide binding, it is possible that fluridone binds to the PVC, aluminum bracketing, plastic tubing, or even to the rubber tubing in the pumps used to create the flow. Whatever the cause, the loss of fluridone from these studies was high in magnitude (50%) and can not be adequately explained without further research Data on herbicide release from granules are presented for static, aerated, low and high flow rates in Tables 3 1 and 3 2, to support the discussion of the influence water movement has on these granules. The ET 25 for Q in low and high flow conditions were slower, 43.8 and 19.7 d, respectively, than release in both static and aerated conditions, 12.0 and 7. 0 d, respectively. Similarly, the release of SRP under high flow, 81 d, was much longer than in either the static or aerated studies (24.0 and 3.0 d), respectively. The fluridone release results from the low and high flow studies, combined with the low r ecovery values, suggests that some component of the experimental apparatus adsorbed fluridone as it release s from the granule in to the aqueous phase in the aquariums. The ET 25 for Q was faster under high flow than low flow, but beyond this observation, c onclusions for fluridone release in the low and high flow conditions are not possible.
43 Results from the more rapid release of the herbicides triclopyr and endothall however, do allow development of firm conclusions regarding herbicide release from these g ranules. First, comparisons of all the ET values under high and low flow conditions showed that release of herbicide from granules of both products were greater under high flow (Table 3 2). The unknown flow under the aerated studies described in Chapter >0.0016 KPH since herbicide release from both triclopyr and endothall was faster under the aerated conditions, compared to the high flow. The airstone was located on the side o f the buckets, half way down the water column for the aerated studies, and it is possible that after a few hours a circular current was established in the buckets that was much faster than originally suspected. The release of triclopyr under the low flow conditions were not different than the release under static conditions, however, this was not the case with the release from endothall granules (Table 3 2). Herbicide release from endothall was faster under the slow flow conditions (7.2 h) compared to the release under static conditions (37.7 h). These results, in which there is no difference between low flow and static release (triclopyr) and very different release values between those two conditions (endothall) indicate that the composition or matrix of the granule, and very likely the chemistry of the herbicide has a significant impact on herbicide release. Although suspected, these studies show that even minimal water flow also has significant impacts on the release of herbicides from granules in the aquatic environment. Further questions remain about what water exchange or flow conditions prevail around granules after application to the bottom of ponds, canals, reservoirs and lakes.
44 Figure 3 1. Experimental apparatus for the water movement experiments in which water velocities can be regulated by variable speed electric peristaltic pumps (not show n) from 0.000016 to 0.0016 KPH. The inflow water entered the 90 elbow on the left and the granules were placed into the open T joint on the right
45 Table 3 1. Estimated time ( days) required for 25, 50 and 95 percent of fluridone to release from Q a nd SRP granules maintained under low and high water flow conditions. Herbicide Treatment ET 25 ET 50 ET 95 Q Static a 12 (9 15) b 27 (21 33) 68 (48 88) Aerated 7 (5 8) 16 (12 21) N/A c Low 44 ( 27 109) N/A N/A High 20 ( 16 26 ) N/A N/A SRP Static 24 (16 32) 72 (50 93) N/A Aerated 3 (2 3) 7 (5 8) 16 (10 23) Low N/A N/A N/A High 81 ( 54 160 ) N/A N/A a Static and Aerated values from Chapter 2. b Mean (95% confidence interval) c N/A=value could not be calculated from regression equation Table 3 2. Estimated time (hours) required for 25, 50 and 95 percent of triclopyr and endothall to release from granules maintained under low and high water f low conditions. Herbicide Treatment ET 25 ET 50 ET 95 Triclopyr Static a 5.1 (4.5 5.9) b 11.9 (10.5 13.8) 37.0 (32.6 42.9) Aerated 0.5 (0.4 0.6) 1.1 (0.9 1.5) 4.2 (3.4 5.7) Low 3.8 (3.2 4.7) 9.0 (7.6 11.1) 31.8 (26.7 39.1) High 1.3 (1.0 1.8) 3.2 (2.5 4.5) 16.5 (12.8 23.1) Endothall Static 37.7 (31.6 46.7) 86.2 (72.0 106.7) 237.5 (199.1 294.3) Aerated 0.4 (0.3 0.5) 0.9 (0.7 1.2) 3.4 (2.7 4.6) Low 7.2 (4.8 12.0) 16.8 (12.0 28.8) 57.6 (40.8 98.4) High 2.4 (2.4 4.8) 7.2 (7.2 12.0) 26.4 (21.6 38.4) a Static and Aerated values from Chapter 2. b Mean (95% confidence interval)
46 Table 3 3. Total percent of herbicide recovered from fluridone, triclopyr and endothall treatments after grinding the granules, and liquid fluridone maintained under low and high water flow conditions. Percent is based upon 100% representing the exact weight of the granules or liquid concentration applied. Herbicide Treatment Crushed Value a Q Low 57 6 b High 55 5 SRP Low 51 8 High 50 4 Triclopyr Low 108 5 High 107 6 Endothall Low 95 6 High 110 5 Genesis c Low 51 4 High 48 2 a Crushed value= Percent concentration in the water after granules were ground (as described in Chapter 2). Percent is based upon the known weight of granules placed in treatment. b Mean 95% confidence interval c Percent concentration of fluridone remaining 10 days after treatment
47 F igure 3 2. Concentration of fluridone released from Q g ranules exposed to low (0.000016 KPH) and high (0.0016 KPH) water flow in a darkened room at 20 C. Symbols and error bars represent mean standard error of t he mean.
48 Figure 3 3. Concentration of fluridone released from SRP granules exposed to low (0.000016 KPH) and high (0.0016 KPH) wate r flow in a darkened room at 20 C. Symbols and error bars represent mean standard error of the mean.
49 Figure 3 4. Concentration of triclopyr released from granules exposed to low (0.000016 KPH) and high (0.0016 KPH) water flow in a darkened room at 20 C. Symbols and error bars represent mean standard error of the mean.
50 Figure 3 5. Concentration of endothall released from granules exposed to low (0.000016 KPH) and high (0.0016 KPH) water flow in a darkened room at 20 C. Symbols and error bars represent mean standard error of the mea n.
51 CHAPTER 4 HERBICIDE RELEASE OF FLURIDONE, TRICLOPYR AND ENDOTHALL UNDER STATIC CONDITIONS IN STERILIZED ORGANIC SEDIMENT Introduction Granular formulations are often used as carriers for aquatic herbicides, but when they sink to the bottom of a water body, they will directly contact and could interact with the sediment in a way that liquids do not. If a wate r body is thermally stratified it is likely that a higher concentration of herbici de would be in the hypolimnion, compared to a higher concentration in the epilimnion typical with surface applied liquid treatments. The degree of this sediment/herbicide interaction will depend on many factors, the two most important being the chemicals affinity for sediment (Koc) and the characteristics of the sediment (Gawlik et al. 1997) The herbicides tested in our experiments range from a low /moderate soil affinity, triclopyr and e n d othall (Koc 20 200 ml g 1 ) ( Senseman 2007 ), to a very high soil affinity, fluridone (Ko c 350 2500 m l g 1 ) (Reinert 1989). Previous sorption studies performed on fluridone, endothall and triclopyr support the predictions that would be made by the Koc values. Mossler et al. (1993) and Weber ( 198 0 ) reported that fluridone is readily adsorbed to soils, pa rticularly those that are high in organic matter or clay content Mossler et al. (1993) found that the greatest binding of fluridone to soils occurred in bentonite clay, and peat sediments, and that less binding occurred in the presence of kaolinite clay and quartz sand. Triclopyr had moderate sorption to soil (Johnson et al. 1995), and endothall had little to no affinity for soil adsorption (Reinert and Rodgers 1984). These studies all used agitation or mild shaking to ensure herbicide soil interaction, which likely means these results represent question remains as to what happens under field con ditions with herbicide s that are applied in a granular formulation. It is possible
52 however, that even for those herbicides that ha ve a low affinity for soil that there is still a sediment/water interface interaction that would slow the release of herbicide from the granule into the water column One consideration that must be taken into account when working with herbicide soil bindi ng is the potential for microbial degradation. If degradation and binding occur simultaneously, it would be difficult to separate how much herbicide loss was due to each factor. If the soil is sterilized, then the only loss of herbicide should be from bi nding, not degradation. However, any process that sterilizes the soil could change the herbicide binding potential which creates an additional experimental variable Wolf et al. (1989) reported that two cycles in an autoclave (103.4 kPa and 121C for 1 h) was enough to eliminate microbial populations with minimal impact to the soil characteristics that affect herbicide binding. They noted slight changes to available manganese and aluminum, but no change to cation exchange capacity, soil pH, or to surface area of the soils tested. Loux et al. (1989) in their study of imazaquin and imazethapyr state that soil pH and cation exchange capacity are highly important in determining the amount of herbicide that will adsor b to a soil. Due to the lack of impact on these factors by autoclaving, it is unlikely that autoclaving a sediment would drastically alter its capacity to bind herbicide. The sediment used for these studies (Bivens Arm Lake, Gainesville, FL) was used be cause the lake has previously supported hydrilla growth. Currently stocked with grass carp, this lake has not received any herbicide treatments (10+ years), utilizing any of the products tested in these experiments. This makes Bivens Arm soil representat ive
53 of sediment that might be expected to receive herbicide treatments in other situations, yet will not contain any residual herbicide that could interfere with this research. The objective of this study was to determine the static release profile of fl uridone, triclopyr and endothall granules under static conditions when applied over an aquatic sediment. Materials and Methods Soil Organic Matter Soil was collected from Bivens Arm Lake, Gainesville, Florida and dried at 70 C in a forced air oven for 1 we ek. The soil was then crushed and sieved through a 2mm screen to remove large debris. Glass beakers were placed in a muffle furnace at ~500 C and allowed to remain for 3 h to ensure all trace organic matter was oxidized. The organic matter content of th e sediment was determined by placing ~2 g of soil in these weighed glass beakers. The weight of the soil added to the beakers was recorded and the beakers with soil were placed in a muffle furnace at ~500 C for 4 h. After combustion, the beakers were rem oved and allowed to cool for 5 minutes before weighing. The difference in weight was recorded as organic content and converted, based on the initial soil dry weight, as percent organic matter. Herbicide Release Soil was collected from Bivens Arm Lake, kept in covered plastic buckets until treated with sterilization techniques. The surface water was poured from the buckets and then the soil was placed in large glass beakers and autoclaved twice under the conditions described in Wolf et al. (1 989). After the soil was autoclaved, it was placed in plastic pails that stored before use in experiments.
54 The experimental set up consisted of placing a 5 cm layer of sterilized sediment in clean plastic pails (bleach cleaned) then slowly adding deionized water to bring the total volume to 15 L in the 19 L pails The water and sediment in the buckets were allowed a settling period of no less than 24 h to ensure that suspended particles were settled to the bottom of the buckets. In some instances after this settling period, the water was still discolored, so water was poured off the top, deionized water added and allowed to settle again This was repeated until the water was below 10 FTU (F ormazin Turbidity Units) Once this experimental set up was complete all treatment and collection procedures were identical to those described under static release in Chapter 2. Similar weights to the static treatment, of fluridone (Q and SRP), triclopyr and endothall granules were sprinkled over the top of the water in the plastic pails from a height of 0.5 m. In addition, liquid treatments were included for fluridone (Genesis), triclopyr (Renovate 3 1 ) and endothall (Aquathol K 2 ) at equivalent concentra tions to the granular treatments. Water samples were collected in the same manner as described in Chapter 2, and concentrations analyzed via ELISA. At the conclusion of the study, the granules were not crushed because the objective of the study was to de termine herbicide movement from the granules into the overlying water. Grinding or mixing the granules with the sediment would have also bound some of the herbicide thus providing incorrect values on the initial content of the granules. Herbicide releas e curves were generated similar to the static release studies reported in Chapter 2. Non linear regression compared release in organic sediments to 1 SePRO Corporation. Carmel, IN. 46032. 2 United Phosphorus Inc. King of Prussia, PA. 19406.
55 the static release experiments. All statistical comparisons were made based on 95% confidence intervals. R esults and Discussion Soil Organic Matter Bivens Arm Lake sedime nt was determined to have a 401.5% (95% confidence intervaI) organic matter content. Brenner and Binford (1988) studied the organic matter content of the sediment in 97 lakes throughout Flo rida and reported an organic matter range of 0.8 to 84.2%, with a mean of 39.7%. The organic matter content of lake sediment from Bivens Arm Lake was similar to the average lake sediment analyzed by Brenner and Binford (1988) suggesting it to be a reasona ble representation of what sediment a granule might interact with in a Florida lake. Herbicide Release A preliminary study was performed to determine if autoclaving Bivens Arm soil caused any changes in herbicide binding of quinclorac, and no differences w ere found between autoclaved and unautoclaved soil (Appendix A). Total recovery of fluridone from Q and SRP granules was only 23 and 17%, respectively, after 90 d in treatment (Figure 4 1 and 4 2). It is unclear from the release profile if fluridone wou ld continue to release into the water column, or if the values represented are a maximum concentration possible. However, it is also possible that residues would continue to fall in the water column, as evidenced by the continual decline in concentrations from the liquid fluridone treatment (Figure 4 3). By 7 d after treatment fluridone concentrations were only 43% of the applied concentration for the liquid treatments It is clear that the release of fluridone in the presence of sediment is drastically different from the static experiments In these static conditions with sediment
56 an ET 25 value could not be calculated for either Q or SRP even after 90 d (Table 4 1). Under static conditions, 95% of fluridone had released from Q by 70 d, and 50% of fluri done had released from SRP by 72 d. It is unclear as to the exact cause of this limited release of fluridone, but sediment binding is likely occurring. Fluridone is not rapidly degraded by soil microbes (Weber 1980 ), but does readily bind to soil. Furth ermore, the soil was sterilized, and all light was excluded, so the two causes of fluridone degradation were likely eliminated from our experimental design. With these eliminated it is most likely that the rapid loss of fluridone from the liquid treatmen ts, and slower but steady decline thereafter, is due to binding of herbicide to the sediment. This could explain the much lower concentrations of fluridone detected in the granular treatments as herbicide rapidly bound to sediment after release, and a muc h smaller percentage was available to disperse into the water column above the sediment. Mossler et al. (1 993) reported that fluridone mixed with a peat soil with both liquid and SRP formulations that only 10% was still detectable in the water column after 25 d of contact For the liquid fluridone treatments there was a rapid loss and binding, not unlike our experiments, whereas the granular formulations never achieved higher than 10% release. However, recall that Mossler et al. (1993) used agitatio n in their experiments, which could account for the extreme amount of binding in such a short time compared to the results of our studies. However, the trends are still similar, where liquid concentrations rapidly decrease in the first few days and then steadily drop thereafter and granular treatments never reach the theoretically applied concentrations. Triclopyr released almost completely, 96 6% (95% confidence interval) by 11 d after treatment, and might have released 100% if given enough time (Figu re 4 4). The
57 speed at which triclopyr released in the presence of organic sediments was significantly slower than under static conditions (Table 4 1). Under static conditions 95% of triclopyr had released by 1.5 (1.4 1.8 (95% confidence interval)) d, but required 10.8 (8.6 14.6) d in the pr esence of soil, a more than 10 fold difference. This is surprising since triclopyr has a very low Koc value (25 144 mL g 1 ) and would not be expected to readily bind to soils (Senseman 2007). Nearly all (97 % 10) of th e triclopyr applied in the liquid treatments was recovered 12 d after treatment, and further supports the lack of herbicide binding to the soil. Endothall, similarly, released almost completely (82 12%) from the granules, but released slower than triclopyr (Figure 4 5). Much like triclopyr the release of endothall was significantly slower in the presence of sediment compared to the static treatments (Table 4 1). For instance, under static conditions 3.6 (3.0 4.4) d were necessary to achieve 50% release o f herbicide, whereas in the presence of sediments this increased to 8.1 (6.2 11.8) d Even though the study was conducted for longer, 30 d for endothall compared to 11 for triclopyr, 95% release was never detected. This could be due to the slightly highe r Koc of endothall, 110 138 mL g 1 compared to an average Koc of ~50 mL g 1 for triclopyr (Senseman 2007). However, the effect of Koc on endothall is questionable in this experiment because 98 6% of the applied liquid endothall was recovered 30 d after tr eatment. Although it seems clear that sediment binding is likely playing a role in limiting fluridone concentrations in these experiments, a different factor is likely affecting triclopyr and endothall. It is possible that the sediment acts as a physical barrier to herbicide reaching the overlying water Visual observations confirm that many of the
58 triclopyr and endothall granules were completely enveloped by the organic sediment. It is possible that the herbicide has to diffuse into the overlying water through a small sediment/water matrix, compared to the static studies where there was no impediment to herbicide movement. These experiments demonstrate that all the granules tested have a significant interaction with organic sediments Herbicide release is either delayed (endothall and triclopyr) compared to release in static conditions (no sediment) or not possible due to binding in the sedim ent (fluridone). It was also clear that herbicides with high Koc, such as fluridone, are m ore likely to bind to sediments, which greatly reduces the overall herbicide concentration that is ultimately achieved in the water column. These studies demonstrate the impact of sediment on the herbicide release of selected granules and indicates that m uch more research is needed to elucidate the behavior of granular aquatic herbicides.
59 Table 4 1. Estimated time (days) required for 25, 50, 95 percent of fluridone, triclopyr and endothall to release from granules maintained under static conditions appl ied over a 5 cm layer of sterilized Bivens Arm Lake sediment. Herbicide Treatment ET 25 ET 50 ET 95 Q Static a 12 (9 15) b 27 (21 33) 68 (48 88) Sediment N/A c N/A N/A SRP Static a 24 (16 32) 72 (50 93) N/A Sediment N/A N/A N/A Triclopyr Static a 0.2 (0.2 0.3) 0.5 (0.4 0.6) 1.5 (1.4 1.8) Sediment 1.1 (0.8 1.4) 2.6 (2.0 3.5) 10.8 (8.6 14.6) Endothall Static a 1.6 (1.3 1.9) 3.6 (3.0 4.4) 9.9 (8.3 12.3) Sediment 3.2 (2.5 4.7) 8.1 (6.2 11.8) N/A a From Chapter 2 b mean (95% confidence interval) c N/A= value could not be calculated from regression equation
60 Figure 4 1. Concentration of fluridone released from Q granules treated over 15 L of water in a plastic pail, maintained under static conditions with a 5 cm thick layer of organic se diment in a darkened roo m at 20 C. Water samples for residue analysis were collected from the center of the bucket half way down the water column
61 Figure 4 2. Concentration of fluridone released from SRP granules treated over 15 L of water in a plastic pail, maintained under static conditions with a 5 cm thick layer of organic sediment in a darkened room at 20 C. Water samples for residue analysis were collected from the center of the bu cket half way down the water column
62 Figure 4 3. Concentration of fluridone remaining from Genesis (liquid) in 15 L of treated water in a plastic pail, maintained under static conditions with a 5 cm thick layer of organic se diment in a darkened room a t 20 C. Water samples for residue analysis were collected from the center of the bucket half way down the water column.
63 Fi gure 4 4. C oncentration of triclopyr released from granules treated over 15 L of water in a plastic pail, maintained under static conditions with a 5 cm thick layer of organic se diment in a darkened room at 20 C. Water samples for residue analysis were collected from the center of the bucket half way down the water column
64 Figure 4 5. Concentration of endothall released from granules treated over 15 L of water in a plastic pail, maintained under static conditions with a 5 cm thick layer of organic sediment in a darkened room at 20 C. Water samples for residue analysis were collected from the center of the bucket half way down the water column
65 CHAPTER 5 HYDRILLA ROOT AND FOLIAR UPTAKE OF QUINCLORAC, TOPRAMEZONE AND BISPYRIBAC Introduction One justification for the use of solid formulations of herbicides in aquatic plant management is granules and pellet formulations place the herbicide near the base of the plant. This placement could reduce the dilution from water flow, and cause higher herbicide concentrations toward the lower portion of the plant which may be advantageous for the uptake and translocation of systemic herbicides Though some research has been conducted to determine any advantage to the use of granular formulations, this issue remains largely unresolved. The herbicide quinclorac (3,7 dichloro 8 quinlinecarboxylic acid) is primarily used for the control of grasses and selected broadleaf species in rice production and belongs to the auxin mimic family of herbicides (Senseman 2007). Topramezone ([3 (4,5 dih ydro 3 isoxazolyl) 2 methyl 4 (methylsulfonyl)phenyl](5 hydroxy 1 methyl 1 H pyrazol 4 yl)methanone) is an inhibitor of the carotenoid biosynthesis pathway, which is used for post emergence broadleaf and grass weed control in corn ( AMVAC 2012; Senseman 2007 ). Bispyribac sodium (sodium 2,6 bis[(4,6 dimethoxy 2 pyrimidinyl)oxy]benzoate) is an inhibitor of the ALS (acetolactate synthase) pathway and primarily used to control grasses, sedges and broadleaf plants in a variety of rice cropping systems (Senseman 2 007). These three herbicides have been studied under experimental use permits for potential aquatic registration and bispyribac was labeled for aquatic use in 2011 1 1 Haller, W.T. 2012. Personal communication.
66 Quinclorac has been tested on a v ariety of plant species, and a diverse absorption and t ranslocation profile is reported. Qui n clorac was readily absorbed by both roots and shoots of leafy spurge and was translocated throughout (acropetal and basipetal movement) the plant (Lamoureux and Rusness 1995). In torpedo grass it was minimally absorb ed by foliage, with almost no translocation to the roots, yet readily root absorbed with extensive translocation to plant shoot s (Williams et al. 2004). Enloe et al. (1999) found minimal absorption and translocation by both roots and shoots of field bindw eed treated with quinclorac Topramezone was found to translocate both basipetally and acropetally in sorghum and giant foxtail but remain in the treated leaf of black nightshade (Grossman and Ehrhardt 2007). Lycan and Hart (2006) found that in selected cool season turf grasses (creeping bent grass, annual blue grass and Kentucky bluegrass) that bispyribac absorption and translocation was greatest from root treated plants, but that very little herbicide was moved to the roo ts from foliar treated plants. In terrestrial environment s plants must absorb soil applied herbicides through the roots, but in aquatic treatments, herbicides are not limited to the soil, but diffuse from solid formulations into the aqueous envi ronment where uptake dynamics would be different from terrestrial treatments. The lack of clearly developed xylem in most submersed aquatic plants such as hydrilla could also affect uptake and translocation of herbicides (Yeo 1984). Herbicide uptake stu dies in submersed aquatic plants have typically been performed with contact herbicides such as diquat and endothall, which utilized free floating apical segments, and translocation of the herbicide was not tested (Haller and
67 Sutton 1973; Sutton et al. 1972 ). Davies and Seaman (1968) determined that limited translocation of diquat occurred in treated elodea shoots, and that the further away from the treatment site, the less the translocation. The biggest obstacle to investigating herbicide uptake and transl ocation in submersed aquatic plants is separating the sediment (roots) from the overlying water. Terrestrial herbicides can be applied to the soil and absorption and movement can only occur through the roots. In aquatic sediments soil applied herbicide will mix in the soil pore water and ultimately into the impossible to determine. Previous attempts to separate aquatic sediments from the soil have involved the use of bottl es and silicone stopcock grease, or an agar/Teflon sediment barrier to prevent the movement of herbicide from sediment to overlying water and vice versa (Davies and Seaman 1968; Funderburk and Lawrence 1963a and b). These studies did show b oth basipetal a nd acropetal translocation by some herbicides, however the authors do not discuss including any control treatments to prove there was no leaking of herbicide across these barriers. Furthermore, there was no indication what possible effect these barriers h ad on the growth and development of the plant. Another system, proven to freely allow plant growth while totally separating the overlying water from the sediment, would be beneficial for studying plant herbicide interactions. Another problem unique to aqu atic plant research is the ability of algae to grow in plant culture. Algae can negatively affect plant growth, inhibit herbicide uptake absorb herbicides and generally affect the parameters being studied A method for eliminating algal growth is necess ary for studies that will last longer than 7 10 d, due to the
68 confounding impact that algae has on submersed plant growth and response to other environmental variables. Sterile tissue culture is used to grow plants under conditions in which outside contaminants (mold, bacteria, algae) can be limited or eliminated (Thorpe 2007) In general this process involves sterilizing plant tissue in a diluted bleach solution, leaving only a small amount of gree n plant material that is free of contaminants. This tissue is then grown in sterile nutrient/sugar solution that allows plant material to grow Once a stable population is established, large amounts of plants can be created rather rapidly. Hydrilla has been successfully established and maintained in sterile tissue culture, providing an algal free source of plant material f or herbicide uptake experiments (Klaine and Ward 1981). Hinman and Klaine (1992) established sterile populations of hydrilla by expos ing tuber explants to a 1% NaClO solution, but the capacity of these explants to be further propagated was sparsely discussed. Kane and Gilman (1991) and Kane and Albert (1989) were both able to establish sterile cultures of Myriophyllum species using a 12 min wash in 1.05% NaOCl, amended with 0.01% (v/v) Tween 20. They were also able to create great proliferation of these explants. The objective s of this study w ere: 1) Examine root:shoot uptake dynamics of three experimental aquatic herbicides, with three different modes of action. 2) Determine if a layer of activated charcoal can be placed on sediment to prevent herbicide movement across the barrier. Materials and Methods Sterile hydrilla plants were created by exposing two node explants to a 1.05% (v/v) solution of NaOCl containing 2 drops of Tween 20 for 12 min. These explants were rinsed in 3 separate washes of sterilized distilled/deionized water and placed in culture
69 tubes with strength Murashige and Skoog mineral salts amended with 100 mg L 1 myo inositol, 2 mg L 1 glycine, 0.4 mg L 1 thiamine HCL, 0.5 mg L 1 pyridoxle HCL, 0.5 mg L 1 nicotinic acid and 30 g L 1 sucrose. Those explants that had green plant growth, and no signs of contamination were then multiplied in larger flasks containing the media described above. From these cultures ~3 cm long apical sections were excised and planted in fine ng 30 ml of a 1/10x 3 All components of the hydrilla culture, water, glass tube and soil were either autoclaved or surface sterilized. Plants were allowed to establish roots and develop autotrophic growth over a two week period. Larger test tubes containing 59 g of fine builders sand capped with a 1 cm layer of activated charcoal granules (granules were ~3 5 mm in diameter) 2 were autoclaved and filled with 200 ml of sterile 3 The hydrilla plants were then rooted into these tubes so that all root material was below and with the apical shoots above the charcoal layer The tubes with plants were placed in a growth chamber main tained at 26 C with a 16 hour photoperiod. A separate set of tubes were prepared that were identical to those previously described except no plants were placed in the tubes to determine if the charcoal layer was impermeable to herbicide movement. Plants were allowed 1 week to acclimate prior to treatment. On the day of treatment all the water above the charcoal layer was carefully suctioned out of the tubes, so as to not disturb the plant or the charcoal layer, and a cannula (18 cm long needle) was insert ed into the sand below the charcoal. With the 2 A preliminary study was conducted that suggested that activated charcoal could be utilized to separate the root zone from the foliar aqueous solutions without affecting plant growth (Appendix B).
70 cannula in place a second needle attached to a syringe was inserted into the top of the cannula and 1.5 ml of air was forced down to ensure that no sand clogged the cannula. Once cleared, both 14 C labeled and unlabeled herbicide was placed in the syringe ( 6,667 Bq (becquerels) of 14 C labeled quinclorac 3 or 8,333 Bq of 14 C labeled topramezone 1 or 6,667 Bq of 14 C labeled bispyribac 4 ) and then this mixture was injected below the charcoal layer (the total volume below the layer was 14 ml) and was dispensed down the cannula with 1 ml of air. The herbicide concentrations used in these experiments were: 400, 50 and 400 g a.i. (active ingredient) L 1 for quinclorac, topramezone and bispyribac, respectively After injection, 1.5 ml of autoclaved deionized water, followed by 1.5 ml of air was forced down the cannula to ensure all herbicide and 14 C labeled herbicide was flushed from the cannula and into the sediment layer. The cannula was then removed from t he tube and the nutrient solution slowly added back over the plants taking care to avoid any disturbance to the charcoal layer. These treatments were identified as root zone treatments and were replicated 4 times. In addition to treating tubes that conta ined plants, 4 replications were included in which no plants were present. A second set of treatments were performed in which the same activity of 14 C herbicide was used, as well as the same unlab e led concentrations in the water (186 ml) above the charcoal layer. These treatments were also replicated 4 times and indentified as foliar treatments. In addition to treating tubes with plants, 4 replications were included in which no plants were present. 3 BASF Corporation. Florham Park, N.J. 07932. 4 Valent U.S.A Corporation. Walnut Creek, C.A. 94596.
71 At the time of harvest, 7, 14 and 21 d after treatment, a 1 ml sample of the water above the charcoal was taken directly from the tube and identified as t op w ater. After this sample was collected, all the water above the charcoal was suctioned off and the volume recorded. With the water removed, the plant mate rial was removed and separated into roots (material below the charcoal) and shoots (material above the charcoal). For all r oot treatments, the plants were rinsed inside the treatment tube with DI water, whereas all s hoot treatments were rinsed in a separa te container (this water was sampled and tested for r adioactivity and identified as r insewater). All water samples were placed in 15 ml of a scintillation cocktail 5 designed specifically for water analysis, vortexed, and allowed to rest until all bubbles had dissipated (typically 2 3 h ). All plant material was placed in foil packets and d ried in a forced air oven at 70 C for 48 h. Dried plants were ground by hand in a mortar and pestle and placed in a biological oxidizer 6 in which CO 2 was captured in a s cintillation cocktail 7 Water and plants samples radioactivity was determined by light scintillation spectrometry (LSS) 8 The sand, charcoal and remaining water in all the tubes was rinsed from the tube over filter paper 9 in a Buchner funnel. After rinsing the sand (approximately 50 ml) a 1 ml sample was collected from the rinsate and indentified as b ottom water. The sand, charcoal and filter paper were placed in a foil packet and dried in a forced air oven for 96 h at 7 0 C. The dried filter paper was separated from the sand, weighed and then 5 Scintisafe Plus Liquid Scintillation Cocktail. Fisher Scientific Company. Fairlawn, NJ. 07410. 6 Biological Oxidizer, Model OX500. R.J. Harvey Instrument Corporation. Hillsdale, NJ. 07462. 7 Carbon 14 Cocktail. R.J. Harvey Ins trument Corporation. Hillsdale, NJ. 07612. 8 Packard 1600 CA TRI CARB Liquid Scintillation Analyzer. Packard Instrument Company. Downers Grove, IL. 60515. 9 Whatman 3 Filter Paper. Whatman Incorporated. Clifton, NJ. 07013.
72 placed in a biological oxidizer and radioactivity determined by LSS. The oxidizing of the filter paper was only performed for the 7 d quinclorac harvests, all other samples were pl aced in a 20 ml scintillation vial containing 15 ml of scintillation cocktail 10 vortexe d and allowed to rest for 1 h before analysis via LSS. The remaining sand and charcoal mix was separated with a strainer, and the charcoal was ground by mortar and pest le and a sample was placed in a biological oxidizer and radioactivity determined by LSS. A sample of the remaining sand was placed in the same cocktail used for the filter paper samples and radioactivity determined The quantification of all radioactivit y in all the components (soil, water, plant, rinsate, etc) allow ed for an accounting of all radioactivity initially applied to each tube. All radioactivity was expressed as a percent of the total amount of radioactivity recovered. This percent was then u sed to calculate the concentration in shoot and root material by multiplying this percent to the total amount of material applied (ug of active ingredient) and adjusted for the weight of the plant material. This value was recorded as ug a.i. g 1 (dry weig ht). Results and Discussion A significantly higher amount of quinclorac was absorbed by foliar treated hydrilla, than by the root zone treated hydrilla (Table 5 1). The shoots of hydrilla from the root zone treated plants contained 0.2 g (active ingredi ent (a.i.)) g 1 (dry weight of plant material) or 200 parts per billion (ppb) compared to 1400 ppb found in the shoots of the foliar treated plants (Table 5 1). There is also a significantly higher concentration of 10 Scintilene Liquid Scint illation Cocktail. Fisher Scientific Company. Fairlawn, NJ. 07410.
73 quinclorac in the roots of the foliar treated plants compared to the shoots of foliar treated, and the roots of root zone treated plants (Table 5 1). This suggests that in hydrilla foliar absorption is ideal, and that there is a higher degree of basipetal transloca tion, although some acropetal movement occurs as well. This agrees more with the results found by Lamoureux and Rusness (1995) in leafy spurge, where quinclorac translocated throughout the plant from both root and foliar treatments. By contrast, there wa s almost no movement of quinclorac to the roots of torpedo grass from foliar treatments (Williams et al. 2004). No radioactivity was detected in the overlying water of root zone treatments, or under the charcoal layer in foliar treated plants, suggesting that the charcoal layer was an effective barrier for quinclorac (Table 5 2). This is the same trend that was observed for the treatments with quinclorac that had no plants (data not shown) further confirming the efficacy of activated charcoal in separatin g the two treatment zones from quinclorac. Most of the recovered radioactivity was found in the charcoal of both the root zone and foliar treatments (Table 5 2). In both treatment locations the amount of radioactivity in the charcoal raises over time, fr om 31 55% and 85 95% for root zone and foliar treatments, respectively. The water of each treatment area contained the second highest percentage of recovered radioactivity, but especially for the foliar treatments, was much less than recovered from the ch arcoal (Table 5 2). The overall recovery of radioactivity from all sources ranged from 44 93%, and the reason for this low recovery is unknown. Furthermore, the high amount of radioactivity in the filter paper at the 7 d harvest could be due to oxidizing a small portion of the whole sample, and could explain
74 the anomalous values of the 7 d harvest (Table 5 1 and 5 2). Due to this subsequent harvest of the filter paper involved analyzing the entire sample, not a subset. The concentration of herbicide fo und inside hydrilla for the root zone treatments with topramezone, with the exception of 14 DAT, were essentially the same as the treatment concentration (50 ppb) and were slightly higher in the roots than the shoots (Table 5 3). The reason for the higher concentration in the plants 14 DAT from root zone experiments is unknown. An opposite effect occurs in the foliar treated plants, where the shoots have a higher concentration than the roots (Table 5 3 ). The concentrations inside the plant range from a l ow of 430 ppb to a maximum of 5,130 ppb for the foliar treated plants, which is much higher than the treatment concentration of 50 ppb (Table 4). As with the root zone treatments, the highest concentration is found 14 DAT and the reason for this is unknow n. In these experiments, there was translocation of herbicide throughout the plant, but the highest concentration was found in the plant material that corresponded to the site of treatment, or put another way root treated plants had higher concentrations in the roots, and foliar treated plants had higher concentrations found in the shoots. Overall foliar treated plants absorbed more herbicide than did root zone treated plants, which is similar to what was observed in quinclorac treated hydrilla. Grossman and Ehrhardt (2007) found that topramezone moved acro and basopetally in their studies of foliar treated sorghum and giant foxtail but little movement was observed in black nightshade Our study shows translocation in both directions as well but that a higher concentration is found in the plant material that is exposed to the herbicide.
75 Similar to the quinclorac treatments, the topramezone treatments also had a high percentage of the recovered radioactivity found in the charcoal layer (45 93%) with the percent increasing over the duration of the study (Table 5 4 ). Similar to the quinclorac studies overall recovery of radioactivity was less than 100% and ranged from a low of 55% (7 DAT foliar treatments) and a high of 80% (14 DAT foliar treatments). Aga in the reason for this low recovery is unknown. With the exception of 7 DAT the root and shoot concentration of bispyribac for root zone treatments was the same, ~2,000 ppb, and did not increase over the duration of the study (Table 5 5). The concentr ation of bispyribac inside the plant was at equilibrium from root to shoot, but still significantly higher than the treated concentration of 400 ppb. The concentration in both the root and shoot of the foliar treated plants was significantly higher than t hat of any of the root zone treated plants (Table 5 5). The roots of the foliar treated hydrilla 7 DAT was significantly different from all the other plant material over the course of the study, and there were no differences among the other treatments fro m each other (Table 5 5). The concentration inside the roots and shoots of the foliar treatments was between 34,300 77,600 ppb and was equilibrated between root and shoots (Table 5 5). The concentration of bispyribac inside hydrilla was much higher, rega rdless of treatment location, than the original treatment concentration of 400 ppb. Lycan and Hart (2006) found a predominant movement of bispyribac from the roots to the shoots of root treated cool season grasses, and very little shoot to root movement i n foliar treatments. Similar to both quinclorac and topramezone the foliar treatments of bispyribac experienced a higher binding of radioactivity to the charcoal (78 91%) than the root
76 zone treatments (31 47%) with more binding occurring as the study progr essed (Table 5 6 ). The overall recovery of radioactivity was the most consistent for the bispryibac treatments (71% 120%) (Table 5 6 ). Th e s e stud ies demonstrate that the use of activated charcoal can be an effective barrier between the two water locations tested (sediment water and overlying water). Total absorption by all three herbicides into the hydrilla plants was highest when the foliar portions were treate d. However, the roots of quiclorac treated plants contained more herbicide in the roots regardless of the treatment location. Quinclorac therefore preferentially translocated to the roots more than other herbicides, however total bispyribac absorbed was much higher than the other herbicides. The highest level of absorption and translocation occurred in the hydrilla that was foliar treated with bispyribac. For these treatments concentrations inside the plant were 100 fold greater than the concentration of herbicide applied to the water. T he charcoal readily bound a majority of the radioactivity applied for all three herbicides, and was more readily bound from the foliar treatments than the root zone treatments. Future studies should explore the additio n of a thin sand layer on top of the charcoal to determine if this binding could be reduced.
77 Table 5 1 Concentration of quinclorac recovered from root and shoot material of foliar and root zone treated hydrilla plants 7, 14 and 21 d after treatment with 400 ppb of quinclorac maintained under 16 h photoperiod at 26 C. Treatment Days After Treatment Root material Shoot material Root zone 7 4.53.2 a 0.20.1 14 0.30.2 0.20.1 21 0.60.2 0.20.04 Foliar 7 48.734.2 1.30.1 14 4.51.2 1.40.3 21 3.70.6 1.40.1 a mean g a.i./g (dry weight)95% confidence interval Table 5 2 Percent of recovered 14 C quinclorac present in hydrilla, treatment water and sediment barriers 7, 14 and 21 d after treatment with 400 g a.i. L 1 to plants maintained under 16 h photoperiod at 26 C. Treatment DAT a TW b BW c Charcoal FP d Sand Roots Shoots Total p ercent recovered Root zone 7 0 52 31 10 3 3 0.1 93 14 0 57 36 2 5 0.3 0.2 66 21 0 38 55 2 3 0.8 0.3 55 Foliar 7 6 0 85 5 0.8 1 0.5 83 14 4 0 90 1 1.5 0.4 1.9 44 21 2 0 95 0.2 0.1 0.3 1.6 58 a Days after treatment b Water above charcoal/sand layer c Water below charcoal/sand layer d Filter paper e Total percent recovered of the 400,000 DPM that was applied
78 Table 5 3 Concentration of topramezone recovered from root and shoot material of foliar and root zone treated hydrilla plants 7, 14 and 21 d after treatment with 50 ppb of topramezone maintained under 16 h photoperiod at 26 C. Treatment Days After Treatment Root Shoot Root zone 7 0.040.02 a 0.020.01 14 0.140.02 0.110.01 21 0.060.01 0.020.01 Foliar 7 1.160.41 4.020.90 14 2.170.61 5.130.76 21 0.430.27 1.371.02 a mean g a.i./g (dry weight)95% confidence interval Table 5 4 Percent of recovered 14 C Topramezone present in hydrilla, treatment water and sediment barriers 7, 14 and 21 d after treatment with 50 g a.i. L 1 to plants maintained under 16 h photoperiod at 26 C. Treatment DAT a TW b BW c Charcoal FP d Sand Roots Shoots Total percent recovered Root zone 7 0 41 45 8 6 0.1 0.1 68 14 0 15 66 8 9 0.8 2.0 64 21 0 7 72 5 16 0.5 0.2 79 Foliar 7 6 0 80 5 2 0.3 2.0 55 14 1 0 86 5 2 0.7 4.0 80 21 1 0 93 2 3 0.2 2.0 74 a Days after treatment b Water above charcoal/sand layer c Water below charcoal/sand layer d Filter paper e Total percent recovered of the 400,000 DPM that was applied
79 Table 5 5. Concentration of bispyribac recovered from root and shoot material of foliar and root zone treated hydrilla plants 7, 14 and 21 d after treatment with 400 ppb of bispyribac maintained under 16 h photoperiod at 26 C. Treatment Days After Treatment Root Shoot Root zone 7 2.10.6 a 1.00.1 14 2.70.3 2.40.9 21 2.61.3 2.20.8 Foliar 7 34.312.7 53.019.7 14 72.913.8 77.621.9 21 62.915.2 69.812.2 a mean g a.i./g (dry weight)95% confidence interval Table 5 6. Percent of recovered 14C bispyribac present in hydrilla, treatment water and sediment barriers 7, 14 and 21 days after treatment with 400 g a.i. L 1 to plants maintained under 16 hour photoperiod at 26 C. Treatment DAT a TW b BW c Charcoal FP d Sand Roots Shoots Tota l percent recovered Root zone 7 0 60 31 4 4 0.4 1 100 14 0 47 44 3 3 0.8 2 100 21 0 40 47 4 6 0.8 2 120 Foliar 7 15 0 78 2 1 0.6 3 71 14 3 0 89 2 1 2 4 84 21 1 0 91 2 1 1 5 77 a Days after treatment b Water above charcoal/sand layer c Water below charcoal/sand layer d Filter paper e Total percent recovered of the 400,000 DPM that was applied
80 CHAPTER 6 SUMMARY Implications The results of these studies demonstrate a variety of release characteristics of the products used to manage invasive aquatic plants. This variety means that not all granules can be utilized in the exact same manner, while expecting herbicide concentratio ns to behave exactly the same. The release of the tested herbicides ranged from as quick as a few hours, to as long as several months in order to achieve 50% release of product. These profiles could be beneficial, depending on the residue goals, but much more information is required to ensure the goals are achieved. The environmental conditions examined in this research sediment and water movement clearly altered and drastically changed the release of the herbicides. Under experimental conditions these differences were in some cases a >10x difference) and will have a profound impact on the residue profile of these products in the field. A faster release of product could mean that residues are not maintained for a sufficient amount of time to achieve co ntrol, and conversely too slow of a release could mean that a sufficient concentration is never achieved. Furthermore, the impact of highly organic sediments will at the very least slow the release of herbicide, and for some products (those with high Koc) binding will occur ensuring that residues never reach the water column. All of these differences would make predicting release hard enough even under highly controlled conditions, but in the field these predictions may prove impossible These results do successfully, but more information is needed to ensure proper strategies are utilized. At
81 the very least these studies highlight the limited amount of information currently available, and question ho w effectively these products have been utilized. Without further research it is a challenge to determine where and how to best use granular products. The uptake data also suggest that herbicide that is in the water column, and thus exposed to the foliage of hydrilla is absorbed into the plant at a much higher concentration than herbicide that is exposed to the roots only. Although no data was collected on the location of herbicide (water column vs sediment) from granules, at the least these studies sugg est that release of herbicide into the sediment is not a beneficial strategy for maximizing herbicide uptake. Although there was translocation both basipetally and acropetally, the highest level of uptake was from the foliage of hydrilla. Hydrilla may be similarly. Based upon my experience, it appears the roots of hydrilla are much more sparse and exhibit shallower growth when compared to other submersed aquatic plants. Again these st udies are only a small sampling of the types of studies and the sort of information that is direly needed to refine plant management strategies. Future Research These studies, in many ways, raise more questions than answers. Although large in scope, and i nvolving a wide array of conditions, these results should be expanded in order to best make recommendations for aquatic plant managers. There are still several formulations of herbicides, and many more conditions that need to be tested in the lab and in t he field to better understand herbicide release from granules. To better determine the impact that water flow has on granules, more flow rates should be tested to determine exactly how little water flow is necessary to impact
82 release. It is unclear if after a certain flow release is unchanged, or is it that the faster the flow the faster the release. It is also unclear how low a flow is necessary to alter the relea se of herbicide from these, and other products. Ultimately, the conditions that prevail in the field will have to be determined to predict residue profiles, and thus ensure successful management strategies. The studies presented in this docume nt only tested one type of soil and its impact on release. Varied amounts of organic matter, and var ied sediment types such as clay and sand should be tested to determine how these soils may affect herbicide release Further studies should seek to determine if the herbicide that was unaccounted for in these studies was bound to the sediment, and if so h ow strongly. Further uptake studies should be conducted with the products tested in this document, but also with other aquatic herbicides, and should also be tested on other submersed aquatic plant species. Future studies should explore methods that will as a barrier to herbicide leakage. These studies demonstrate that a separation of root water from the overlying water is possible, and these techniques should be utilized to answer questions previously unexplored in aquatic plants. Studies should be designed that seek to explore the relationship between herbicide concentration and exposure time and the effect that has on plant uptake. These types of studies could provide valuable information that could be used to develop better treatment strategies that seek to more efficiently manage invasive aquatic plant species with less herbicide and a reduced cost
83 There are several other factors that could have an impact on how quickly herbicides release from granules. Water chemistry such as pH, temperature and conductivity etc., could all have an impact on the release of herbicides. Furthermore, if these factors prove to impact releas e, then studies should be conducted that seek to combine several of these factors, such as sediment and flow under various temperature, or pH, etc., to better understand how all these parameters could impact performance in the field. Finally, all of these factors could need to be identified in the field, before treatments could take place, in order to ensure successful management occurs. This does not mean it is impossible to be successful with granules, but it does mean more information is needed before managers can have the same c onfidence of success that is present with the use of liquid herbicides. The research presented in this document begins to fill a much needed gap in the knowledge of aquatic herbicides, specifically those formulated on gr anules. At a minimum, these studies raise questions as to how universally granules should be used in management strategies, and certainly warrants further research to determine if the best management strategies are currently being used. Although these st definitively answer all the questions necessary to determine how best to use granules, they should serve as a call to action to conduct further research on these issues
84 APPENDIX A SEDIMENT STERILIZATION TECHNIQUES A preliminary study was perfo rmed to determine if autoclaving aquatic sediments would impact herbicide binding. Autoclaving the soil is necessary to eliminate any microbial populations that might be present in the soil, which could degrade herbicides over the length of a study. Plast ic 50 mL centrifuge tubes (five replications for each treatment) were filled with 2 g of soil (either air dried as a control or the autoclaved samples) and then a 10 mL herbicide solution was added to the tubes. The herbicide mix contained of 1.0 mg L 1 o f cold quinclorac and 5,000 DPM of 14 C labeled quinclorac. After allowing the tubes to incubate for 24 h on a wrist action shaker the tubes were centrifuged at 5000 rpm for 1 min. After centrifugation a 2 mL aliquot of the supernatant was removed and comb ined with 15 mL of scintillation cocktail and then vortexed in a scintillation vial. Counts of 14 C labled herbicide in the vials was achieved via liquid scintillation spectrometry (LSS). To determine the amount of herbicide absorbed to the sediment, the amount of herbicide in solution (determined by the counts in the aliquot) was subtracted from the original concentration applied. Autoclaved and non autoclaved samples were compared, and no significant differences were observed. Autoclaved soil absorbed 63.1% (95% C.I. 61.3 64.9%) and non autoclaved absorbed 64.5% (64.1 64.8%) of the total quinclorac applied.
85 APPENDIX B ACTIVATED CHARCOAL STUDIES Activated charcoal, in a preliminary experiment, was identified as an effective barrier to herbicide mov ement from sediment to the overlying water, with no visual impact on hydrilla plant growth. Plastic cups (350 mL) were filled with sand and treated with quinclorac at a rate of 1 mg a.i. L 1 (water volume enough to saturate the sand), capped with a 2 cm t hick layer of granular activated charcoal, further capped by a 1 cm thick layer of sand, and placed in plastic buckets which contained 15 L of water. No quinclorac was detected, by HPLC analysis, in the overlying water throughout the 37 d experiment and ~ 40% of the applied quinclorac was recovered from the treated bottles at the conclusion of the study suggesting that some of the herbicide either broke down or was bound to the charcoal. This study demonstrated the ability of the charcoal to serve a s a barrier to herbicide movement between the sediment and the overlying water. In a separate preliminary study, 350 mL cups were filled with sand amended with 1g/Kg of osmocote fertilizer and had one 5 cm apical hydrilla explant rooted in the sand. The sand was then capped with charcoal as described above. The plants were observed for 3 weeks, where stem elongation, lateral branching and additional root development was observed. This research suggests that activated charcoal could be used for hydrilla root uptake studies because it blocked herbicide movement from the soil into the overlying water and growth (elongation, lateral development and root development) was observed.
86 LIST OF REFERENCES AMVAC. 2012. Impact Herbicide Label. AMVAC Chemical Corporation. Los Angeles, C.A. 90023. Blackburn, R. D., L. W. Weldon, R. R. Yeo, and T. M. Taylor. 1969. Identification and distribution of similar appearing aquatic weeds in Florida. J Aquat Plant. M anage. 8: 11 21. Brenner, M., and M. W. Binford. 1988. Relationships between concentrations of sedimentary variables and trophic state in Florida lakes. Can. J. Fish. Aquat. Sci. 45: 294 300. Cook, C. D. K. and R. Lnd, 1982. A revision of the gen us Hydrilla (Hydrocharitaceae). Aquat. Bot. 13:485 504. Davies, P. J., and D. E. Seaman. 1968. Uptake and translocation of diquat in elodea. Weed Sci. 16: 293 295. Enloe, S. F., S. J. Nissen, and P. Westra. 1999. Absorption, fate and soil activit y of quinclorac in field bindweed ( Convolvulus arvensis ). Weed Sci. 47: 136 142. Frank, P. A., and R. D. Comes. 1967. Herbicidal residues in pond water and hyrosoil. Weeds. 15: 210 213. Funderburk, H. H. Jr., and J. M. Lawrence. 1963a. Absorpti on and translocation of radioactive herbicides in submersed and emersed aquatic weeds. Weed Res. 3: 304 311. Funderburk, H. H. Jr., and J. M. Lawrence. 1963b. Preliminary studies on the absorption of C 14 labeled herbicides in fish and aquatic plants. Weeds. 11: 217 219. Gawlik, B. M., N. Sotiriou, E. A. Feicht, S. Schulte Hostede, A. Kettrup. 1997. Alternatives for the determination of the soil adsorption coefficient, KOC, of non ionicorganic compou nds a review. Chemos. 34: 2525 2551. Grossmann, K., and T. Ehrhardt. 2007. On the mechanism of action and selectivity of the corn herbicide topramezone: A new inhibitor of 4 hydroxyphenylpyruvate dioxygenase. Pest Manage. Sci. 63: 429 439. Haller, W. T., and D. L. Sutton. 1973. Factors affecting the uptake of endothall 14 C by Hyrdilla. Weed Sci. 21: 446 448. Hinman, M. L. and S. J. Klaine. 1992. Uptake and translocation of selected organic pesticides by the rooted aquatic plant Hydri lla verticillata Royle. Environ. Sci. Technol. 26: 606 613.
87 Johnson, William G., T. L. Lavy, and E. E Gbur. 1995. Sorption, mobility and degradation of triclopyr and 2,4 D on four soils. Weed Sci. 43: 678 684. Kane, M. E. and L. S. Albert. 1989. Abscisic acid induction of aerial leaf development in Myriophyllum and Proserpinaca species cultured in vitro J. Aquat. Plant Manage. 27: 102 111. Kane, M. E. and E. F. Gilman. 1991. In vitro propagation and bioassay systems for evaluating growth re gulator effects on Myriophyllum species. J. Aquat. Plant Manage. 29: 29 32. Klaine, S. J., and C. H. Ward. 1981. Axenic Culture of Hydrilla. J. Aquat. Plant Manage. 19: 59 60. Kollman, W., and R. Segawa. 1995. Interim report of the pesticide chemistry database. Environmental Hazards Assessment Program. Department of Pesticide Regulation. Koschnick, T. K., W. T. Haller, V. V. Vandiver, and U. Santra. 2003. Efficacy and residue comparisons between two s low release formulations of fluridone. J. Aquat. Plant Manage. 41: 25 27. Lamoureux, G. L., and D. G. Rusness. 1995. Quinclorac absorption, translocation, metabolism, and toxicity in leafy spurge ( Euphorbia esula ). Pestic. Biochem. Phys. 53: 210 22 6. Langeland, K. A. 1996. Hydrilla verticillata (L.F.) Royle (Hydrocharitaceae), "The perfect aquatic weed". Castanea. 61:293 304. Lazor, R. L. 1978. The ecology, nomenclature and distribution of hydrilla (Hydrilla verticillata Casp.) and Brazilian e lodea (Egeria densa Planch.). Proc SWSS 28: 269 273. Lemmin, U., and C. H. Mortimer. 1986. Test of an extension to internal seiches of lakes. Limnol. Oceanogr. 31: 1207 1 231. Les, D. H. L. J. Mehrhoff, M. A. Cleland, and J. D. Gabel. 1997. Hydrilla verticillatea ( Hydrocharitaceae) in Connecticut. J. Aquat. Plant Manage. 35: 10 14. Loux, M. M., R. A. Liebl, and F. W. Slife. 1989. Adsorption of imazaquin and imazethapyr on soils, sediments, and selected adsorbents. Weed Sci. 37: 712 718.
88 Lycan, D. W., and S. E. Hart. 2006. Foliar and root absorption and translocation of bispyribac sodium in cool season turfgrass. Weed Technol. 20: 1015 1022. Miron, D S., M. Crestani, M. R. Shettinger V. M. Morsch, B. Baldisserotto, M. A. Tierno, G. Moraes, V. L. P. Viera. 2005. Effects of the herbicides clomazone, quinclorac, and metsulfuron methyl on acetylcholinesterase activity in the silver catfish ( Rhamdia q uelen ) (Heptapteridae). Ecotox. and Environ. Safety. 61: 398 403. Mossler, M. A., D. G. Shilling, K. E. Milgram, and W. T. Haller. 1993. Interaction of formulation and soil components on the aqueous concentration of fluridone. J. Aquat. Plant Manage 31: 257 260. Netherland, M. D., and K. D. Getsinger. 1995. Laboratory evaluation of threshold fluridone concentrations under static conditions for controlling hydrilla and Eurasian watermilfoil. J. Aquat. Plant Manage. 33: 33 36. Netherland, M. D. W. R. Green, and K. D. Getsinger. 1991. Endothall concentration and exposure time relationships for the control of Eurasian watermilfoil and hydrilla. J. Aquat. Plant Manage. 29: 61 67. Netherland, M. D., and R. M. Stewart. 1994. Laboratory and m esocosm evaluations of controlled release matrices as potential herbicide delivery systems. U.S Army Corp of Engineers Technical Report. A 94 3. Nufarm. 2009. Material Safety Data Sheet for Quinclorac SPC 75 DF Herbicide. Nufarm Americas, Inc. Burr Ridge, IL, 60527. Oborn, E. T., W. T. Moran, K. T. Greene, and T. R. Bartley. 1954. Weed control investigations on some important aquatic plants which impede flow on western irrigation waters. Laboratory Report SI 2. Joint Repor t U.S.D.A. and U.S.D.I, Bur. Of Reclamation, Denver, Colo. Pp. 28 29. Quay, P. D., W. S. Broecker, R. H. Hesslein, and D. W. Schindler. 1980. Vertical diffusion rates determined by tritium tracer experiments in the thermocline and hypolimnion of two l akes. Limnol. Oceanogr. 25: 201 218. Reinert, K. H. 1989. Environmental behavior of aquatic herbicides in sediments. Chapter 13 in: Reactions and Movements of Organic Chemicals in Soils. Soil Science Society of American and American Society of Agronomy. Special Publication No. 22. Madison, Wisconsin. Reinert, K. H., and J. H. Rodgers. 1984. Influence of sediment types on the sorption of endothall. Bull. Environ. Contam. Toxicol. 32: 557 564.
89 Reinert, K. H., S. Stewart, M. L. Hinman, J H. Rodgers Jr., and T. J. Leslie. 1985. Release of endothall from aquathol granular aquatic herbicide. Water Res. 19: 805 808. Senseman, S.A., ed. 2007. Herbicide Handbook Ninth ed. Champaign, IL: Weed Science Society of America Pp. 357 358. Sutton, D. L., W. T. Haller, K. K. Steward, and R. D. Blackburn. 1972. Effect of copper on uptake of diquat 14 C by hydrilla. Weed Sci. 20: 581 583. Thorpe, T. A. 2007. History of plant tissue culture. Mol. Biotechmol. 37: 169 180. Van, T. K., an d K. K. Steward. 1986. The use of controlled release fluridone fibers for control of hydrilla ( Hydrilla verticillata ). Weed Sci. 34: 70 76. Van, T. K. and V. V. Vandiver. 1992. Response of monoecious and dioecious hydrilla to bensulfuron methyl. J. Aquat. Plant Manage. 30:41 44. Weber, J. B. 1980. Adsorption of buthidazole, VEL 3510, tebuthiuron, and fluridone by organic matter, montmorillonite clay, exchange resins, and a sandy loam soil. Weed Sci. 28: 478 483. Wilkinson, R. E. 1964. Suba queous release of herbicides from granules. Weeds. 12: 69 76. Williams, W., G. Wehtje, and R. H. Walker. 2004. Quinclorac: soil behavior and foliar vs. root absorption by torpedograss ( Panicum repens ). Weed Technol. 18: 626 633. Woodburn, K. B., W. R. Green, and H. E. Westerdahl. 1993. Aquatic Dissipation of triclopyr in Lake Seminole, Georgia. J. Agric. Food Chem. 41: 2172 2177. Wolf, D. C., T. H. Dao, D. H. Scott, and T. L. Lavy. 1989. Influence of sterilization methods on selected soil, microbiological, physical and chemical properties. J. Eviron. Qual. 18: 39 44. Yeo, R. R., R. H. Falk, and J. R. Thurston. 1984. The morphology of hydrilla ( Hydrilla verticillata (L. F.) Royle). J. Aquat. Plant Manage. 22: 1 17.
90 BIOGRAPHICAL SKETCH Born in 1981 in Merridian, Mississippi, Brett is the son of Craig and Laura Bultemeier. As a child, Brett moved often, as his father was a naval aviator and was redeployed every few years. Finally ending up in Indiana, he pursue d his B.S. degree at Manchester College in North Manchester, Indiana, focusing his studies on biology and environmental studies. While at Manchester, he was a collegiate wrestler, played in the school band, and helped form the environmental club. During hi s undergraduate studies, Brett worked during the summer for Weed Patrol Inc. and it was at this job that a passion for aquatic plant management was fostered. Upon completion of his degree, Brett was married to Megan and promptly moved to Gainesv ille, Flori da, to pursue a m aster agronomy was completed at the University of Florida in the Spring of 2008. Immediately following completion of this degree, work began on a PhD degree in agronomy. Beginning in January of 2012, Brett will be the Water Resource Manager for the Florida region with Clarke (an environmental service company) In May of 20 12, Brett and Megan will be blessed with the birth of their first son.