RESTORATION AND MANAGEMENT OF LAKES AND RESERVOIRS - CHAPTER 16 pdf

16.2 EFFECTIVE CONCENTRATION — DOSE, TIME CONSIDERATIONS, ACTIVE INGREDIENTS, SITE-SPECIFIC FACTORS, AND HERBICIDE FORMULATION Aquatic herbicides were originally developed for terrestr

The decision to use herbicides should be based on the same criteria — efficacy; cost; health,safety, and environmental impacts; regulatory appropriateness; and public acceptability- that areused for other management techniques (Chapter 11) This was not always the case Becauseherbicide (and other pesticide) treatments were fast, relatively cheap, and many times very effective,they were used in inappropriate ways regarding health, safety and environmental impacts Thisinfluenced public perception about the acceptability of using pesticides.

One of the more striking historical cases of overuse of a toxic but very effective aquatic herbicidewas the use of sodium arsenite Between 1950, when the Wisconsin Department of NaturalResources began keeping records and 1970 when it was no longer used, approximately 798,799

kg of sodium arsenite were added to 167 lakes (Lueschow, 1972) The environmental impacts ofthese treatments were not monitored However, the use of sodium arsenite causes long-term prob-lems for further management in some lakes where it was heavily used The sediments in theselakes are a hazardous waste so other lake management options such as dredging become extremelydifficult if not impractical (Dunst, 1982)

Herbicides are a useful technique in a lake manager’s “tool box.” The largest obstacle to usingthem may be public perception Poor public perception can be overcome with good demonstrationprojects, reliable monitoring (Chapter 11), education, full disclosure of known environmentalimpacts, and responsible use by applicators

16.2 EFFECTIVE CONCENTRATION — DOSE, TIME

CONSIDERATIONS, ACTIVE INGREDIENTS, SITE-SPECIFIC

FACTORS, AND HERBICIDE FORMULATION

Aquatic herbicides were originally developed for terrestrial use, mainly for agriculture In terrestrialsystems an effective concentration of active ingredients (a.i.) is applied directly to the plant or thesoil Exposure time is usually not a consideration unless there is a meteorological event like arainstorm that washes the herbicide off the plant Similarly, an effective concentration of herbicidecan be applied directly to emergent and floating-leaf aquatic species For submergent species aneffective dose is delivered through water so dilution and dispersion are considerations The watervolume treated, currents, drift and micro-stratification (Chapter 11) effect dilution and dispersion.The success or failure of treating any species is dependent on an effective dose of activeingredient contacting or being taken up by the plant This is dependent on the concentration/expo-

sure time (CET) relationship for controlling the target plant (Getsinger, 1997) An effective centration can be achieved using a high dose of herbicide and a short contact time or a low dose

con-of herbicide and a long contact time (Figure 16.1) While a low dose con-of material is more desirablefor cost, safety, health, and environmental reasons, an effective CET relationship and thus efficacy

is more difficult to achieve for submersed species because any bulk water movements away fromthe plant affects the CET relationship

This does not imply that an effective dose is always easily achieved for emergent and leaved species Accurate application requires that the equipment be well calibrated and that theboat or other application vehicle is moving at a constant speed This is difficult in heavy vegetationand a boat tends to submerge the plants that it passes over, washing off the herbicide What it doesimply is an effective dose of herbicide is easier to calculate for emergent and floating-leaf species.Application rates are calculated based on the area treated For submergent species, water depth andvelocity also need to be considered

floating-Understanding active ingredient is critical to proper CET calculations Active ingredient is theconcentration of herbicidally active chemical in a formulation It can vary tremendously betweendifferent formulations or different manufactures of the same product It is expressed as weight tovolume (g/L) for liquid formulations and weight to weight (g/kg) for granular formulation, or itmay be represented as a percentage For example in a liquid formulation the active ingredients could

be expressed as 300 g/L or 30% Active ingredient concentration is given on the herbicide label.Site-specific treatment factors affect the choice of herbicide formulation, which affects appli-cation equipment, techniques, and timing For example, a surface application of a liquid formulation

is appropriate in quiescent, isothermal water These conditions allow an even distribution and mixing

of a surface application In a dense plant stand that creates a temperature-stratified environment,

or in areas of great water movement, a granular or pellet formulation, or subsurface injection of aliquid formulation, will more evenly distribute the herbicide

16.3 TYPES OF CHEMICALS

There are only six herbicides: copper (Chapter 10), 2,4-D, diquat, endothall, fluridone, andglyphosate, that are registered and commonly used for lake and reservoir management in the

FIGURE 16.1 Examples of concentration/exposure time (CET) relationships using endothall for

Myriophyl-lum spicatum (A) and Hydrilla verticillata (B) control The shaded area represents CETs that give 85–100%

M spicatum control with very limited regrowth up to 4 weeks post-treatment and 85–100% Hydrilla control

with very limited or no regrowth up to six weeks post-treatment The CET relationship is different for each

species-herbicide combination (After Netherland, M.D et al 1991 In: J Aquatic Plant Manage 29: 61–67.

0

72 66 60 54 48 42 36 30 24 Exposure time, hours 18

12 6 0

United States A seventh herbicide, triclopyr, is used under an experimental use permit Otherherbicides may be approved for use in other countries or approved for aquatic uses that are notappropriate to lake and reservoir management because they have a long use restriction time orthey are toxic to fish or other aquatic organisms.

These herbicides and other chemicals can be categorized in a number of ways depending ontheir use, mode of contact, selectivity, and persistence in the environment (Table 16.1)

16.3.1 CONTACT VS SYSTEMIC

Contact herbicides act quickly and are generally lethal to the plant cells they contact Because oftheir rapid action and other physiological reasons, they do not move extensively within the plantand they kill tissue only where they contact the plant For this reason they are generally moreeffective on annual plants (see Table 12.10 in Chapter 12 for information regarding annual vs.perennial plants) Perennial plants can be defoliated by contact herbicides but they regrow fromunaffected parts, especially parts that are protected beneath the sediment Contact herbicides aremore effective than systemic herbicides on old, slow growing, or senescent plants, so they arepreferred later in the growing season for controlling aquatic nuisances where, for lack of time orfor physiological reason, systemic herbicides are not effective

Method of Disappearance c,d

Complexed

copper

Various complexing

agents with

copper-liquid and granular

Systemic Plant cell toxicant 3 Precipitation

Adsorption 2,4-D Butoxyethel ester — salt

Dimethylamine — liquid

Isooctyl ester — liquid

Systemic Selective plant growth

regulator

7–48 Microbial degradation

Photolysis Plant metabolism Diquat Liquid Contact Disrupts plant cell

membrane integrity

1–7 Adsorption

Photolysis Microbial degradation Endothall Liquid and granular Contact Inactivates plant

protein synthesis

4–7 Plant metabolism

Microbial degradation Fluridone Liquid and granular Systemic Disrupts carotenoid

synthesis, causing bleaching of chlorophyll

20–90 Photolysis

Microbial degradation Adsorption

Glyphosate Liquid Systemic Disrupts synthesis of

phenylalanine

14; Used over but not in water

Adsorption Microbial degradation Triclopyr e Liquid Systemic Selective plant growth

regulator

a Herbicides registered by U.S Environmental Protection Agency.

e Experimental use permit only.

Sources: b After Madsen, J.D 2000 Advantages and Disadvantages of Aquatic Plant Management Tech Rept ERDC/EL MP-00-01 U.S Army Corps of Engineers, Vicksburg, MS c After Langeland, K.A 1997 In: M.V Hoyer and D.E Canfield (Eds.), Aquatic Plant Management in Lakes and Reservoirs, NALMS, Madison, WI and Lehigh, FL pp 46–72 d After Wisconsin Dept Nat Res., 1988 Environmental Assessment Aquatic Nuisance Control (NR 107) Program Wisconsin Dept Nat Res., Madison, WI.

Systemic herbicides are translocated from absorption sites to critical growth points in the plant.They act slowly when compared with contact herbicides, but they are generally more effective forcontrolling perennial and woody plants They are also more selective than contact herbicides.Correct application rates are critical If application rates are too high, systemic herbicides can actlike contact herbicides They stress the plants so much that the herbicides are not translocated tocritical plant growth areas (Nichols, 1991).

16.3.2 BROAD-SPECTRUM VS SELECTIVE HERBICIDES

Broad-spectrum herbicides control all or most of the vegetation they contact Selective herbicidescontrol certain plants but not others Selectivity is based on the different response of differentspecies to the herbicide It is a function of both the plant and the herbicide

Selectivity can be affected by the CET relationship of the herbicide For example, water hyacinth

(Eichhornia crassipes) is selectively controlled amongst spatterdock (Nuphar sp.) using the

rec-ommended rate of 2,4-D, but spatterdock can be controlled by using higher rates and granularformulations (Langeland, 1997)

Systemic herbicides are the most physiologically selective herbicides However, as stated above,they must be translocated to the site where they are active Herbicides may be bound on the outside

of the plant or bound immediately after they enter the plant so they cannot move to the activitysite For other reasons, not all understood, herbicides are more readily translocated in some plantsthan in others, which results in selectivity (Langeland, 1997) Some plants have the ability to alter

or metabolize a herbicide so it is no longer active, and some herbicides affect very specificbiochemical pathways so they only work on plants or groups of plants with those pathways(Langeland, 1997)

Selectivity is also affected by the physiology of perennial species during their growth cycle.During early stages of growth, energy reserves are translocated upward in the plant so an herbicidetaken up by the roots is most effective Late in the growth cycle, material is translocated downward

to the roots so a foliar herbicide is most effective (Langeland, 1997)

16.3.3 PERSISTENT VS NON-PERSISTENT

Persistent herbicides retain their activity in water for a long time, usually measured in weeks ormonths Non-persistent herbicides act only when sprayed directly onto foliage or they lose theirphytotoxicity rapidly on contact with soil, particulate matter in the water, or plant cells Non-persistent herbicides may decay rapidly in water There is no set time that separates persistent fromnon-persistent The half-life of the herbicide in water is a useful measure of persistence (Table 16.1)

16.3.4 TANK MIXES

In addition to single uses, herbicides are mixed to increase efficacy Diquat and copper chelatesare a popular tank mix that provides a broad spectrum of control for aquatic plants plus theconvenience of working with a liquid formulation

16.3.5 PLANT GROWTH REGULATORS (PGRS)

Growth regulators prevent plants from obtaining normal stature They keep plants short but tional by preventing cell division and elongation PGR research on aquatic plants has occurred forover 15 years Unfortunately it has yet to be commercialized so PGRs cannot and have not beenused for management purposes

func-Laboratory and field tests show that Thiadiazuron and Bensulfuron Methyl maintained milfoil

(Myriophyllum spicatum), hydrilla (Hydrilla verticillata) and Potamogeton spp in short stature

(Anderson, 1986, 1987; Anderson and Dechoretz, 1988; Lembi and Netherland, 1990; Nelson and

Van, 1991) Thiadiazuron inhibited tuber and turion production in hydrilla (Klaine, 1986) Bensulfuron

methyl inhibited propagule formation in P nodosus, P pectinatus, and hydrilla (Anderson, 1987).

Growth regulators are a very interesting technology because they have the potential for utilizingthe beneficial aspects of aquatic plants without letting them grow to nuisance proportions Thereare still many questions to answer regarding product delivery, mode of uptake, mode of action,differential plant responses, efficacy, health, safety, and environmental impacts that probably willnot be answered without commercial interest in the technique

16.3.6 ADJUVANTS

Adjuvants are chemicals added to herbicides to increase their effectiveness There are activatoradjuvants, spray-modifier adjuvants, and utility-modifier adjuvants (Thayer, 1998) They includewetting agents and emulsifiers that allow the herbicide to mix more easily Spreaders allow herbi-cides to spread evenly over treated surfaces Stickers, thickeners, invert emulsifiers, and foamingagents increase the adherence of the herbicide to the treated surface and help control herbicidedrift Penetrants enhance absorption of herbicides by decreasing surface tension or by penetratingthrough waxy coatings Many herbicide formulations contain a small percentage of adjuvants andall the categories of adjuvants mentioned may not be used in the aquatic situation Wetting agentsand spreader-stickers are probably the most frequently used adjuvants (Binning et al., 1985)

16.4 INCREASING HERBICIDE SELECTIVITY

Ideally, herbicides should be used to selectively control undesirable species and to change plantcommunity structure to a more desirable type Past control efforts usually did not take selectivityinto consideration and research continues to make herbicides more selective Some tools for usingherbicides selectively are already present and include efficacy information as well as location-selective, time-selective, and dose-selective applications

Using the differential susceptibility of plants to herbicides is one method of selective control

In a mixed plant community, if the undesirable species are controlled by an herbicide and desirablespecies are not, there is a basis for selective herbicide control based on herbicide efficacy An

example is using 2,4-D to control Eurasian watermilfoil or coontail (Ceratophyllum demersum) in

a mixed pondweed (Potamogeton spp.) community 2,4-D effectively controls milfoil and coontail

but not pondweeds As a basis for planning selective management, herbicide efficacy is summarized

in Table 16.2 Label instructions for specific efficacy information should be consulted before usingany herbicide

Applications can be selective by carefully placing the herbicide on target plants and avoidingnon-target plants Experienced personnel for example, using a handgun applicator, can control small

areas of water hyacinth among bulrushes (Scirpus sp.) using 2,4-D and careful placement of the

herbicide on the target plant (Langeland, 1997) Likewise, if diquat were used in the above scenario,although it is a broad-spectrum, contact herbicide, it would only kill bulrush stems above thewaterline The extensive underground bulrush roots and rhizomes are not affected and the plantregrows after the initial effect of the herbicide (Langeland, 1997)

Adjuvants that restrict herbicide movement are a way of selectively treating an area Thismethod is especially appropriate for treating areas that are monotypes of nuisance species whilekeeping the herbicide from drifting into a valuable plant community Another method of restrictingherbicide movement is to treat in conjunction with a drawdown The sediments of Lake Ocklawaha,Florida were treated experimentally with fluridone and other chemicals under drawdown conditions

to test the efficacy of controlling hydrilla plants and tubers (Westerdahl et al., 1988) Herbicidescan be precisely placed in terrestrial areas

Water temperature and light influence macrophyte growth, physiological status, and phenology

Most herbicides work best when plants are actively growing Some species, Elodea canadensis, P.

crispus, and M spicatum for example, grow better at low water temperatures and appear earlier in

the growing season than many other species This provides an opportunity to treat these specieswith a contact or short-lived systemic herbicide before other species are actively growing Refer

to Chapter 11 for a discussion of the importance of phenology and resource allocation patternswhen determining management strategies

A thorough knowledge of CET relationships allows selective management based on varyingdose or contact time of the same herbicide The water hyacinth and spatterdock example was given

above An endothall label suggests that P crispus can be effectively treated at about one-half the concentration needed to control P americanus and many emergent and free-floating species Adams and Schulz (1987) found that M spicatum and E canadensis were highly sensitive to low concen-

trations of diquat “Fine tuning” treatments based on CET relationships constitute a very activearea of research It is not easy because of previously mentioned problems of dispersion and dilutionbut it is an area that holds great promise for selectively managing plant communities with herbicidesand for reducing environmental impacts from herbicide treatments

16.5 ENVIRONMENTAL IMPACTS, SAFETY AND HEALTH

CONSIDERATIONS

16.5.1 HERBICIDE FATE IN THE ENVIRONMENT

Knowing the fate of aquatic herbicides in the environment is important for determining mental impacts, safety and health How long do herbicides persist in the environment, what are thebreakdown products, where do the herbicides or breakdown products go when they “disappear”are all important questions Disappearance refers to the removal of the herbicide from a certainpart of the environment (Langeland, 1997) Aquatic herbicides disappear by dilution, adsorption

Note: C, controlled by the herbicide; CC, conditionally controlled by the herbicide; this could

mean that efficacy depends on specific formulation or application techniques, that it was rated

as only fair or good control by Westerdahl and Getsinger (1988), or that it is labeled only for

partial control —, not controlled by the herbicide, not registered for use with this species, or

information is unknown.

a For use as a general guide; read label instructions for details.

b Can be controlled by copper or copper complexes.

Source: After Lembi, C.A and M Netherland 1988 Category 5, Aquatic Pest Control Dept

Botany, Purdue University, W Lafayette, IN; Westerdahl, H.E and K.D Getsinger 1988.

Aquatic Plant Identification and Herbicide Use Guide, Volume II: Aquatic Plants and

Suscep-tibility to Herbicides Aquatic Plant Cont Res Prog Tech Rept A-88-9 U.S Army Corps

of Engineers, Vicksburg, MS; Binning, L., B Ehart, V Hacker, R.C Dunst, W Gojmerac, R.

Flashinski and K Schmidt 1985 Pest Management Principles for Commercial Applicator:

Aquatic Pest Control University Wisconsin-Ext., Madison; Cooke, G.D 1988 In: The Lake

and Reservoir Guidance Manual USEPA 1440/5-88-02 pp 6-20–6-34.

to bottom sediments, volatilization, absorption by plants and animals, and by dissipation Herbicidesdissipate by photolysis, microbial degradation, or metabolism by plants and animals The rate ofdisappearance (Table 16.1, half-life) depends upon: (1) initial herbicide concentration, (2) watermovement, (3) temperature, (4) amount of plant matter, (5) water chemistry, (6) water volume, (7)the presence of decomposing organisms, and (8) the mode of disappearance.

Table 16.1 summarizes the methods of herbicide disappearance Of the contact herbicides,endothall biodegrades into carbon dioxide and water Diquat is rapidly taken up by plants or bindstightly to particles in the water or bottom sediments When bound to clay mineral particles, diquat

is not biologically available When bound to organic matter, microorganisms slowly degrade diquat

It is photo-degraded to some extent when applied to leaf surfaces Information about the persistence

or biological effects of degradation products of diquat was not found (WDNR, 1988)

Microbial action is the primary mode of degradation of 2,4-D and photolysis may be importantunder alkaline conditions (WDNR, 1988) 2,4-D degrades into naturally occurring compounds 2,4-

D amine for example degrades to carbon dioxide, water, ammonia, and chlorine (Langeland, 1997).Dissipation of fluridone from water occurs mainly by photo-degradation Microbial breakdown

is probably the most important method of breakdown in bottom sediments Degradation rate isvariable and may be related to the time of year of application Applications when days are shorterand sun’s rays less direct result in longer half-lives Fluridone usually disappears from water after

3 to 9 months It usually remains in bottom sediments between 4 months and 1 year (Langeland,1997)

Although glyphosate is not applied directly to water, when it does enter water, binding toparticulate matter and to bottom sediments inactivates it It is degraded to carbon dioxide, water,nitrogen, and phosphorus over a period of several months (Langeland, 1997)

Complexing is the major means of removing soluble copper ions from water The copper ion

is chemically bound by carbonate and hydroxide ions in natural waters as well as by organic humicacids This binding is rapid in high alkalinity, hardness, and pH waters Some lakes received massivedoses of copper over an extended period of time Lakes Kegonsa and Waubesa in Dane County,Wisconsin were treated with 586,750 kg and 692,182 kg, respectively, of copper sulfate between

1950 and 1970 (Lueschow, 1972) Copper sulfate was applied to the five Fairmont Lakes in southernMinnesota at cumulative rates of 1647 kg/ha over a 58-year period (Hanson and Stefan, 1984).Copper concentrations in lake sediments of the Dane County lakes were as high as nearly 1% oftotal sediment weight (WDNR, 1988) In the Dane County lakes the highest concentration of copper

is found in sediments at the greatest water depth and copper concentration decreases toward thetop of the sediment, which indicates the sediments with the highest copper concentration are beingburied There appears to be an annual copper cycle in the lakes with greater copper concentrationsfound in the water during the autumn lake turnover Increased copper levels are largely in thesuspended organic fraction of the water; relatively small increases have been observed in solublecopper (WDNR, 1988) See Chapter 10 for additional details about copper

The active ingredients are not the only chemicals added to the waters Inert ingredients,manufacturing contaminants, and adjuvants are also added The fates of some of these productshave been studied but generally their fate is less well known than the fate of the active ingredients.Modeling is becoming an increasingly important tool for characterizing ecological risks of usingpesticides in aquatic environments at the individual, population and community levels (Bartell etal., 2000)

The USEPA decides whether or not to register an herbicide after considering the ingredients; themanufacturing process; the physical and chemical properties; the mobility, volatility, breakdownrates, and accumulation potential in plants and animals; the toxicity to animals; and the carcinogenic

or mutagenic properties The USEPA can approve or disapprove registration of a new herbicideand may further restrict or cancel the registration of those in use

An herbicide’s capacity to harm fish, plants, and other aquatic life depends on the toxicity ofthe herbicide, the dose rate used, the exposure time of the affected organism, and the persistence

of the herbicide in the environment Toxic effects may be direct or indirect Direct effects impactthe organism of concern Direct effects may be lethal if they kill the organism or they can be sub-lethal Sub-lethal or chronic effects include biomass loss, low resistance to disease, compromisedreproduction rates or sterility, loss of attention, low predator avoidance, and deformed body parts.The short-term indirect effects are the ecological effects caused by the death and decay of the targetplants Long-term effects are changes caused by a restructuring of the plant community or broaderecological changes like the change in stable-state from a macrophyte-dominated lake to an algae-dominated lake, or changes in food webs (Chapter 9) The direct and indirect impacts of herbicideuse are summarized in Figure 16.2

16.5.2.1 Direct Effects

The most obvious direct toxic effect is damage to non-target aquatic plants This can occur to plantspresent in the targeted treatment area or it can affect plants not in the target area by spray drift orresidue movement in water currents The potential for this impact can be calculated knowing theCET relationship between the non-target species, the herbicide, and the herbicide concentrationafter considering dissipating factors

The lethal and sub-lethal effects to invertebrates, fish, and higher animals or humans are not

as easily assessed A variety of tests and extrapolations are performed on aquatic organisms toascertain herbicide toxicity Acute toxicity is usually reported as lethal concentration, effectiveconcentration, or tolerance limit (WDNR, 1988) A lethal concentration (LC) is the concentrationthat kills 50% of the test organisms in a given time period such as 24, 48, or 96 hours It is one

of the most commonly tested and reported parameters for fish and other aquatic organisms It isreported as LC50, 24, 48, or 96 hours The effective concentration is the dosage that immobilizesthe test organism It is often used for insects and crustaceans where determining death is difficult.The tolerance limit is an extrapolated or mathematically determined concentration used to estimatethe point of toxicity The “no observable effect” level is another means of reporting toxicity It isthe highest test concentration that shows no observable impact on the test organisms

Most assays are conducted under laboratory conditions that allow careful control over a widevariety of factors affecting test results Such simplified tests present obvious difficulties interpretingthe impacts of an herbicide on a complicated, dynamic system like a lake There is also a concernover the species and life stages selected for testing (Paul et al., 1994) It is impossible to test allpotentially affected organisms, at all life stages, in all habitat conditions Many of the test speciesmay not occur in the area where the herbicide is used

The bulk of the published data on herbicide toxicity to aquatic biota relates to effects oninvertebrates and fish but there are effects on phytoplankton, micro-organisms, and higher animals.Many higher animals are not obligate aquatic organisms so less attention has been paid to them.However, some higher animals like frogs and toads are obligate aquatic organisms in early life stages.Sub-lethal or chronic effects are probably even more difficult to assess than lethal effects How

do you tell if a bluegill is not feeling well today? The main ways are through population, growth,and life-cycle studies that can be extremely complex in a lake or reservoir ecosystem

The objective of this section is not to review all the toxicological data and do a risk assessmentfor aquatic herbicides, but to give some idea of the complexity of the task The information is toovoluminous and should be done by a professional toxicologist To learn more, the best resources

FIGURE 16.2 Possible effects of a herbicide application on the aquatic ecosystem Main effects are indicated by thick lines (From Murphy, K.J and P.R.F Barrett.

1990 In: A Pieterse and K Murphy (Eds.), Aquatic Weeds, The Ecology and Management of Nuisance Aquatic Vegetation Oxford University Press, Oxford, UK pp.

136–173 With permission of the original author, David Mitchell.)

Increased light penetration

pH levels decrease

Increase in non-susceptible plants

Photosynthesis

in system increased

Increase of certain non-susceptible animals

pH levels increased

Increase in bacteria, fungi and detritivores Autolysis and decomposition of dead material

Release of plant nutrients

Anaerobic conditions

Production of

CH4and H2S

Aerobes killed

Increase in anaerobes

are environmental assessments done by governmental agencies that reviewed the toxicology andassessed risk of aquatic herbicides (Shearer and Halter, 1980; WDNR, 1988) Another excellentresource is the Extension Toxicology Network on the WEB To find it, type Extoxnet in the WEBsearch function Extoxnet provides a pesticide information profile (PIP) that summarizes tradenames, regulatory status, formulations, toxicological effects, ecological effects, environmental fate,physical properties, and manufactures It also provides references to further information.

The two herbicides with the greatest potential for direct toxic effects are the monoamine salt

of endothall (trade name Hydrothol 191) and copper sulfate Due to its toxicity, liquid Hydrothol

191 is not recommended for use in water bodies where fish are an important resource (WDNR,1988) Copper at low levels can produce mortality and sub-lethal toxicity affecting the growth andreproduction of aquatic life on several trophic levels The concentrations of copper used to controlalgae are higher than those that have been shown to produce chronic toxicity in a range of aquaticorganisms and are above those that produce acute toxicity in particularly sensitive organisms(WDNR, 1988) Trout living in soft water are particularly sensitive to copper

16.5.2.2 Indirect Impacts

Indirect impacts of herbicide use (Figure 16.2) include changes to water chemistry; detritus mulation; ecosystem alteration including changes in community structure, food webs, and stablestate; and the possibility of accumulating trace contaminants For managers, Engel (1990) provides

accu-a concise literaccu-ature review of the likely ecosystem impaccu-acts of herbicide use The waccu-ater chemistrychanges are similar to those described in Chapter 11 through natural aquatic plant death and decay.Most water chemistry changes caused by herbicide treatment occur quickly because plant deathoccurs in days or weeks If nuisances are great enough to consider herbicidal control, aquatic plantbiomass is usually high Under natural conditions 33–50% of macrophyte biomass may decompose

in the first three weeks after death (Adams and Prentki, 1982) Decomposition may occur morequickly after an herbicide treatment, especially if the herbicide disrupts plant tissue Therefore,there is a large amount of plant material consuming oxygen for decay, releasing nutrients, andadding to bottom detritus over a short time period Often this occurs during warm months andwarm water temperatures do not hold as much oxygen as cold waters Growing conditions for algaeare optimal so released nutrients stimulate algae “blooms.”

The oxygen demand caused by decomposition is exacerbated by oxygen loss from thesis as plants die The main factors involved in oxygen depletion after herbicide treatments arewater temperature, turnover rate of the water column, water depth, macrophyte biomass and shootnitrogen content, and the rate of external oxygen input Short-term recovery from deoxygenationfollowing an herbicide treatment usually results from a phytoplankton bloom or replacement plantgrowth (Murphy and Barrett, 1990)

photosyn-Respiratory CO2 increases with decay can shift the inorganic carbon equilibrium In poorlybuffered waters this may result in a daytime change of more than one pH unit (Murphy and Barrett,1990) Plant nutrients released from decaying macrophytes to the water column favors the growth

of phytoplankton or free-floating species like Lemna sp (Murphy and Barrett, 1990) If free-floating

species dominate, daytime dissolved oxygen levels may not recover to pre-treatment levels for aprolonged period (Murphy and Barrett, 1990) The plant biomass that ends up on the lake bottom

as detritus continually consumes oxygen in the decay process Low oxygen creates reducingconditions in sediments causing further nutrient releases Loss of canopy foliage can increasesunlight penetration and water temperature Particulate organic matter from macrophyte decay cantemporarily increase turbidity

Long-term studies show the magnitude of some nutrient and detrital inputs In LakeOkeechobee, Florida, an estimated 14,281 metric tons (m.t.) of detritus were produced, and 285m.t of N and 74 m.t of P were returned to the water column over a 24-year period from herbicide-treated, freely floating aquatic vegetation (Grimshaw, 2002) In addition 4,472 m.t of detritus were

produced and 88 m.t of N and 23 m.t of P were returned to the water column in the KissimmeeRiver, the main inflow to Lake Okeechobee, over a 15 year period The nutrient loading fromherbicidal control was estimated at 4–49% for P and 1–17% for N of the external nutrient loading

to the lake In addition, some detritus, N, and P from Kissimmee River treatments likely reached

Lake Okeechobee In Lake Istokpoga, Florida, reduction of hydrilla (H verticillata) through

herbicide treatments during 1988–1992 resulted in significant increases in total P and chlorophyll

a concentrations, and a decrease in Secchi depths (O’Dell et al., 1995) These results were expected

since nutrients bound in the extensive hydrilla mats were released with herbicidal treatment.Decomposition of the hydrilla mats may also have increased sediment resuspension, and changedthe primary producers from macrophytes to algae

Food chains and food webs are changed with loss of macrophyte habitat Plant-dwellinginvertebrates and epiphyton decline from habitat loss but benthic invertebrates can increase withincreased detritus (Hilsenhoff, 1966) Loss of shelter exposes young fish, zooplankton and plantdwelling invertebrates to increased predation Loss of macrophyte cover can increase bank erosionand suspension of bottom sediment Water birds can disperse to quiet waters with protective coverand food When food webs are altered due to loss of macrophytes and associated epiphyton thereare “winner” species and “loser” species For instance, within the carrying capacity for fish in alake, high aquatic plant abundance favors fish species that are adapted to aquatic plants and lowaquatic plant abundance favors fish species that are adapted to open water A major factor deter-mining the value of aquatic plants to fish species is whether the fish is a predator or prey species.The presence of aquatic plants increases the structural complexity of lake ecosystems that providesrefuge for prey species and interferes with the feeding of predator species Even for single speciesthere are “trade-offs.” Herbicides may kill some zooplankton and expose them to increased predationbut phytoplankton blooms after an herbicide treatment increases their food supply

There is some concern that continued use of herbicides will develop herbicide resistant isms In the past there was scant evidence for this occurring (WDNR, 1988) the way herbicideswere normally used However, recent evidence indicates that there is a differential susceptibility

organ-of hydrilla to fluridone in several aquatic systems in Florida (Netherland et al., 2001) This wasunexpected and a significant new development in aquatic plant management Part of the problemmay be related to the low dose rate of fluridone usage Low doses could exert great selectivepressure where there are small differences in susceptibility

Another concern is the development of herbicide resistant plant communities Herbicides areselective so the susceptible species are killed and the tolerant species remain To kill the remainingspecies a different herbicide may be used If this scenario is repeated enough times, only speciesresistant to most herbicides remain This may be beneficial if the species are desirable, but if not,herbicides will no longer be effective and an aquatic plant management tool is lost Over the shortterm, herbicide treatment causes regression to an earlier stage of fresh water plant succession.Opportunistic disturbance-tolerant plants fill the newly vacated niches followed by the seral replace-ment of opportunists by slower-growing, but more competitive, plant species (Murphy and Barrett,

1990; Newbold, 1976) Chara spp., Najas flexilis, and Potamogeton foliosus are often initial pioneering species and Chara spp and Vallisneria americana are persistent species after herbicide

treatments (Brooker and Edwards, 1973; Crawford, 1981; Getsinger et al., 1982; Hestand andCarter, 1977) In the longer term a single herbicide treatment may have little effect on macrophytecommunity structure (Wade, 1981; Wade, 1982 as cited in Murphy and Barrett, 1990) Over theyears following treatment, hydroseral processes lead to the re-establishment of the original plantcommunity but repeated treatments may keep the plant community in a hydroserally early stage(Murphy and Barrett, 1990) Windfall Lake, a 23-ha lake with a maximum depth of 9.2 m innortheastern Wisconsin, was an example of the above scenario (Dunst et al., 1974) Three years ofextensive treatments with a variety of herbicides reduced a mixed aquatic plant community to dense,

monotypic stands of Chara over much of the lakes littoral zone Chara growth reached the water

surface in 2 m of water in some areas of the lake — a perceived macrophyte problem turned into

a real problem for lakeshore residents Within 3 years of a “doing nothing” (see Chapter 12),

Potamogeton amplifolius, a much more desirable species in this case, replaced Chara over large

areas of the lake

In shallow, eutrophic lakes herbicide treatments may shift the “stable state” (Scheffer et al.,1993) from a macrophyte dominated lake to an algae dominated lake (Moss et al., 1996) Herbicidesare not unique in this regard Other management techniques can also cause this shift It is verydifficult to calculate how much management will cause a shift (van Nes et al., 2002) and once theshift occurs it can be difficult to return to a macrophyte dominated state (Scheffer, 1998).The case studies later in this chapter provide some information about both direct and indirectenvironmental effects related to specific treatments More detailed information is often found inthe references related to these treatments

16.5.2.3 What Should a Lake Manager or Concerned Citizen Do?

Ultimately a lake manager, riparian owner, or governmental agency has to make a decision onwhether to use or allow the use of herbicides Are there risks? — some unanswered (such as thepossibility of trace contaminants) — yes As with any management practice, based on the evidenceavailable, the risks need to be balanced with the benefits From a practical point of view, currentlyregistered aquatic herbicides have been used for a long time with no known dire consequences toaquatic ecosystems The majority of the data suggest that the impacts are transient So far, there

is little evidence of any build-up of herbicide residues or chronic toxicity in natural aquatic systemsand fish populations appear not to be adversely affected (Murphy and Barrett, 1990) Most problemscan be traced to inappropriate use Currently, no product can be registered for aquatic use if itposes more than a one in a million chance of causing significant damage to human health, theenvironment, or wildlife resources and, in addition, it may not show evidence of biomagnification,bioavailability, or persistence in the environment (Madsen, 2000) Because of dilution, adsorption

by soil particles and organisms, volatilization, and other means of dissipation, organisms are exposed

to the applied concentration of herbicide for only a short period of time Given an escape route,mobile organisms (mainly fish) show an avoidance reaction to some herbicides (Murphy and Barrett,1990) Can herbicides change aquatic ecosystem functions? The answer again is yes Sometimesthis is the desired result, in other cases the results are known For purposes of this book it should

be noted that there is a big difference between the limited use of herbicides to change aquatic plantcommunity composition or to eradicate an exotic species, and the prolonged use of herbicides tomanage an aquatic nuisance without addressing the cause of the nuisance The Dane County,Wisconsin and Fairmount, Minnesota references given earlier are examples of the latter situation.The next section discusses ways to minimize environmental risks when using herbicides The moreeffective the treatment, the longer lasting the impacts are likely to be or the more environmentalchange that is likely to occur

16.6 WAYS OF MINIMIZING ENVIRONMENTAL RISKS

The most important means of minimizing environmental risk is to follow the label instructions forthe herbicide Herbicides were tested for safety based on labeled conditions Not following labelprocedures is illegal There are restrictions on the use of herbicide treated water for human drinking,swimming, and fish consumption; for animal drinking; and for irrigation of turf, forage, and foodcrops These restrictions are subject to change but are provided on the label so make sure youunderstand and can abide by them before using the herbicide, and follow them after application.Notifying lake users of herbicide applications prevents inadvertent use of restricted waters andmany times is legally required (Figure 16.3) The label also provides information on the efficacy

of the product Applying an herbicide that does not control target species adds unneeded chemicals

to the environment and wastes money and effort

Applying herbicides beginning at the shoreline and working outward provides mobile organisms

an avenue of escape In heavy weed infestations, treat only a portion of the area at one time Allow2–3 weeks between treatments This minimizes dissolved oxygen depletions and nutrient pulsescaused by decomposing vegetation It also allows recruitment of a variety of organisms fromuntreated refuges

Treat only the area that needs to be managed This may seem obvious but, with fluridone awhole lake treatment is recommended Areas can be isolated for treatment by deploying temporary,non-permeable barrier curtains to reduce water exchange with other part of the lake (McNabb,2001) This also reduces herbicide cost

Applicators need to keep current with technology On-board computers, fathometers, globalpositioning (GPS) units, and digital flow meters allow applicators to be much more precise withthe area treated and treatment doses (Figure 16.4) (Kannenberg, 1997) Low-dose applications offluridone and endothall and new formulations of 2,4-D and copper chelates are products or tech-niques that reduce environmental risk (Kannenberg, 1997)

FIGURE 16.3 Posted notice of an herbicide application.

FIGURE 16.4 Typical herbicide application equipment Notice the GPS antennae and the on-board computer.

Maintenance management is another tool to reduce environmental risk A maintenance agement program controls plants at low levels before they become a problem It is used effectively

man-in Florida to control water hyacman-inth By maman-intaman-inman-ing water hyacman-inth to less than 5% coverage,herbicide usage was reduced by a factor as great as 2.6, detritus deposition was reduced by a factor

of 4, and reduced depression of dissolved oxygen occurred beneath vegetation mats (Langeland,1998) By using maintenance management on the St John River, Florida, the U.S Army Corps of

Engineers reduced the area of Pistia stratiotes that needed treatment from 881 ha to 33 ha and the

area of water hyacinth that needed treatment from 649 to 28 ha between 1995 and 2000 (Allen,2001) Maintenance management works well on water hyacinth because it grows rapidly and nearlycontinually, and it is aerially exposed so it is easily targeted Maintenance management wouldprobably work well on other floating or emergent species with similar characteristics Maintenancecontrol of submersed species in lakes is more difficult (Langeland, 1998) Part of the problem isprobably the herbicide dilution factor and part is probably that the plants need to be growing to beeffectively treated Plants cannot be treated if they are not there

Additional governmental regulations may impact the safety of herbicide use Federal courtactions necessitated the issuance of National Pollution Discharge Elimination System (NPDES)

permits for applications of aquatic herbicides used for water hyacinth and egeria (Egeria densa)

control programs in California (Anderson and Thalken, 2001) Permits were issued in 2001 andrequired extensive environmental monitoring and toxicity testing as well as compliance withconditions imposed by the Endangered Species Act

16.7 CASE STUDIES

The literature describing herbicide use to control aquatic plants is voluminous The case studiesselected emphasize using species selective herbicides to change plant community structure and/oreradicate exotic species with minimal damage to native aquatic plants In addition, the herbicidetreatment was done only one to a few times in any water body and there were follow-up plantmonitoring data for at least 1 year after treatment

16.7.1 PLANT MANAGEMENT WITH FLURIDONE IN THE NORTHERN UNITED STATES

16.7.1.1 Minnesota Experiences

In 1992 the Minnesota Department of Natural Resources (MNDNR) initiated an evaluation todetermine whether application of fluridone to whole bays or lakes can control Eurasian watermilfoiland have minimal effects on native vegetation Whole lake applications of herbicides to publicwaters of Minnesota is generally not allowed because it destroys more vegetation than is necessary

to provide lake access Whole lake application of fluridone might be acceptable if it selectivelycontrolled Eurasian watermilfoil This might be possible using low fluridone concentrations andlong contact times Selective milfoil control was defined as removal of milfoil while causing littlereduction in other plants (Welling et al., 1997) Elimination and subsequent re-establishment ofnative plants was not considered selective control Parkers, Zumbra, and Crooked Lakes wereselected for this evaluation (Table 16.3) All were spring treatments, and targeted whole lakefluridone concentrations were 10 μg/L for Parkers and Zumbra Lakes and 15 μg/L for Crooked Lake.Fluridone treatment reduced the percentage of sampling stations with vegetation in both Parkersand Zumbra Lakes (Table 16.4) In Lake Zumbra the average number of vascular plants per samplingstation declined during the year of treatment to one-quarter of the number observed before treatmentand remained at this reduced level through the second year after application (Welling et al., 1997).Eurasian watermilfoil had not reappeared by the second year after application and two native

species, coontail and P zosteriformis disappeared (Table 16.5) Nymphaea sp., P pectinatus, and

curly-leaf pondweed (P crispus) became a more dominant part of the vegetation (Table 16.5),

although based on absolute frequency Nymphaea and P pectinatus both declined.

In Parkers Lake, Eurasian watermilfoil was found at the end of the first year after treatment(Table 16.6) and the frequency of milfoil nearly returned to pre-treatment levels by the end of the

second year (Welling et al., 1997) Coontail and M sibiricum were not found in post treatment surveys (Table 16.6) but they were found at other locations in the lake Sago pondweed, Zosterella dubia, P foliosus, and Chara sp were found at greater frequencies after the fluridone treatment

(Welling et al., 1997) and became more dominant members of the plant community (Table 16.6).Unfortunately, curly-leaf pondweed also became more dominant

Secchi disk transparency decreased after fluridone application in Lake Zumbra and reached aminimal value that was 43% of pre-treatment levels during the first year after treatment Transpar-ency returned to pre-treatment levels the second year after treatment (Welling et al., 1997) Chlo-

Depth (m)

Target conc.

(μg/L)

Parkers, MN Mid-May, 1994 39 11.3 (max.) 10

Zumbra, MN Late May, 1994 66 17.7 (max.) 10

Crooked, MN Early May, 1992 47 8 (max.) 15

Potters, WI Fall, 1997 66 7.9 (max.) 14

Random, WI Fall, 1999 85 6.4 (max.) 12

Big Crooked, MI Mid-May, 1997 65 18.5 (max.) 5 in top 3.05 m

Camp, MI Mid-May, 1997 65 16.7 (max.) 5 in top 3.05 m

Lobdell, MI Mid-May, 1997 221 24.4 (max.) 5 in top 3.05 m

Wolverine, MI Mid-May, 1997 98 17.9 (max.) 5 in top 3.05 m

Burr Pond, VT Early June, 2000 34.5 4.4 (ave.) 6

Hortonia, VT Early June, 2000 195 5.8 (ave.) 6

a Target species for treatment were Myriophyllum spicatum and Potamogeton

cris-pus in all lakes except Potters, Random, and Burr Pond where only M spicatum

First Year after Treatment

Second Year after Treatment

Third Year after Treatment

Source: After Welling, C et al 1997 Evaluation of Fluridone for Selective Control of Eurasian

Water-milfoil: Final Report Minnesota Dept Nat Res., Minneapolis.

rophyll a levels were also higher the first year after treatment than they were pre-treatment or the

year of treatment In Parkers Lake, Secchi disk transparency did not decrease after the fluridonetreatment

Crooked Lake surveys indicated that in the third and fourth years after treatment vegetationcoverage was nearly 100%, values similar to pre-treatment levels (Table 16.4) Eurasian watermilfoil

was not discovered in Crooked Lake until the fourth year after treatment P richardsonii and M sibiricum were not found after the treatment and coontail declined dramatically Najas sp., Z dubia,

Year of Treatment (1994)

First Year After Treatment (1995)

Second Year After Treatment (1996)

a Only species with a frequency more than 24% are included.

bComparisons are made based on August sampling except for P crispus where a May or June sampling are compared for 1994, 1995, and 1996 This could partially explain the large increase in the relative frequency of P crispus between

1993 and the later years.

Source: After Welling, C et al 1997 Evaluation of Fluridone for Selective Control of Eurasian Watermilfoil: Final

Report Minnesota Dept Nat Res., Minneapolis.

Year of Treatment (1994)

First Year After Treatment (1995)

Second Year After Treatment (1996)

a Only species with a frequency more than 24% are included.

bComparisons are made based on August sampling except for P crispus where a May or June sampling are compared for 1994, 1995, and 1996 This could partially explain the large increase in the relative frequency of P crispus between

1993 and later years.

Source: After Welling, C et al 1997 Evaluation of Fluridone for Selective Control of Eurasian Watermilfoil: Final Report Minnesota Dept Nat Res., Minneapolis.

P foliosus, and sago pondweed all became more dominant members of the plant community by

the fourth year after treatment (Table 16.7) Initially curly-leaf pondweed became more dominantbut by the fourth year after treatment, its importance declined

Due to degradation by photolysis, adsorption to hydrosoils, plant uptake, and dilution fluridoneconcentrations are usually less than target values and decrease over time Fluridone concentrationswere equal to or greater than target concentrations for 30 days after application for both Zumbraand Parkers Lakes (Welling et al., 1997) Plant exposure in these lakes was probably more thanneeded to control milfoil (Welling et al., 1997)

Based on these results the MNDNR concluded that the unavoidable damage to non-target plantsand the potential effects on other aspects of the lake ecosystem were great enough so as not togenerally permit whole lake fluridone applications (Welling et al., 1997) Criteria considered topermit an application variance are: (1) high potential to eliminate milfoil from a lake, (2) lowpotential to damage native plants, (3) high potential for the lake to become a source for the spread

of milfoil, and (4) low potential for the reintroduction of milfoil into the lake A hypotheticalsituation where the MNDNR might issue a variance to allow a whole-lake fluridone treatment is

a lake that: (1) has no inlet or outlet, (2) is small (less than 40 ha), and (3) is located in an areawith no other milfoil lakes (Welling et al., 1997)

16.7.1.2 Wisconsin Experiences — Potters and Random Lakes

Potters and Random Lakes (Table 16.3) in southeastern Wisconsin were selected for fall fluridonetreatments Eurasian watermilfoil was confirmed present in Potters Lake in 1975, and by 1997 it

had a 99% frequency Native plants were not diverse or abundant Chara sp., coontail, and Elodea canadensis were the most common native species (Table 16.8) Potters Lake was treated in October,

1997 with an initial target fluridone concentration of 14 μg/L Pre- and post treatment aquatic plant,herbicide residue, and water quality data were collected as part of the permit requirements (Toshner

et al., 2001)

TABLE 16.7

Relative Frequency (%) of Common a Aquatic Plants before and after a Fluridone Treatment

in Crooked Lake, Minnesota

Species

Pre-Treatment (May 1992)

First Year After Treatment (July 1993)

Second Year After Treatment (August 1994)

Third Year After Treatment (August 1995)

Fourth Year After Treatment (August 1996)

a Only species with a frequency more than 24% are included.

Source: After Welling, C et al 1997 Evaluation of Fluridone for Selective Control of Eurasian Watermilfoil: Final Report.

Minnesota Dept Nat Res., Minneapolis.