Milfoil Ecology, Control and Implications for Drinking Water Supplies

Transcription

1 Milfoil Ecology, Control and Implications for Drinking Water Supplies Subject Area: Environmental Leadership

2

3 Milfoil Ecology, Control and Implications for Drinking Water Supplies

4 About the Awwa Research Foundation The Awwa Research Foundation (AwwaRF) is a member-supported, international, nonprofit organization that sponsors research to enable water utilities, public health agencies, and other professionals to provide safe and affordable drinking water to consumers. The Foundation s mission is to advance the science of water to improve the quality of life. To achieve this mission, the Foundation sponsors studies on all aspects of drinking water, including supply and resources, treatment, monitoring and analysis, distribution, management, and health effects. Funding for research is provided primarily by subscription payments from approximately 1,000 utilities, consulting firms, and manufacturers in North America and abroad. Additional funding comes from collaborative partnerships with other national and international organizations, allowing for resources to be leveraged, expertise to be shared, and broad-based knowledge to be developed and disseminated. Government funding serves as a third source of research dollars. From its headquarters in Denver, Colorado, the Foundation s staff directs and supports the efforts of more than 800 volunteers who serve on the board of trustees and various committees. These volunteers represent many facets of the water industry, and contribute their expertise to select and monitor research studies that benefit the entire drinking water community. The results of research are disseminated through a number of channels, including reports, the Web site, conferences, and periodicals. For subscribers, the Foundation serves as a cooperative program in which water suppliers unite to pool their resources. By applying Foundation research findings, these water suppliers can save substantial costs and stay on the leading edge of drinking water science and technology. Since its inception, AwwaRF has supplied the water community with more than $300 million in applied research. More information about the Foundation and how to become a subscriber is available on the Web at

5 Milfoil Ecology, Control and Implications for Drinking Water Supplies Prepared by: K.J. Wagner, D.F. Mitchell, J.J. Berg, and W.C. Gendron ENSR Corporation 2 Technology Park Drive, Westford, MA Jointly sponsored by: Awwa Research Foundation 6666 West Quincy Avenue, Denver, CO and U.S. Environmental Protection Agency Washington, D.C. Published by Distributed by:

6 DISCLAIMER This study was jointly funded by the Awwa Research Foundation (AwwaRF) and the U.S. Environmental Protection Agency (USEPA) under Cooperative Agreement No. CR AwwaRF and USEPA assume no responsibility for the content of the research study reported in this publication or for the opinions or statements of fact expressed in the report. The mention of trade names for commercial products does not represent or imply the approval or endorsement of either AwwaRF or USEPA. This report is presented solely for informational purposes. Copyright 2008 by Awwa Research Foundation ALL RIGHTS RESERVED. No part of this publication may be copied, reproduced or otherwise utilized without permission. ISBN Printed in the U.S.A.

7 CONTENTS LIST OF TABLES... ix LIST OF FIGURES... xi FOREWORD... xiii ACKNOWLEDGMENTS...xv EXECUTIVE SUMMARY... xvii CHAPTER 1: INTRODUCTION...1 Project Background... 1 Methods Used for Literature Review... 3 Report Organization... 3 CHAPTER 2: GENERAL INTRODUCTION TO AQUATIC VEGETATION...5 Fundamentals of Aquatic Plant Communities... 5 Taxonomy Diversity... 5 Functional Habitat Classification... 6 Factors Important for Growth of Aquatic Macrophytes... 7 Nuisance Aquatic Vegetation Characteristics of Nuisance Species Exotic Species CHAPTER 3: MYRIOPHYLLUM TAXONOMY, ECOLOGY AND DISTRIBUTION...13 The Genus Myriophyllum North American Species of Myriophyllum Class I Milfoil Species Myriophyllum spicatum Myriophyllum heterophyllum Myriophyllum aquaticum Class II Milfoil Species Myriophyllum humile Myriophyllum sibiricum Myriophyllum hippuroides Class III Milfoil Species Myriophyllum verticillatum Myriophyllum pinnatum Myriophyllum laxum Myriophyllum farwellii Myriophyllum quitense Myriophyllum ussuriense Myriophyllum tenellum Myriophyllum alterniflorum Hybridization of Myriophyllum Species v

8 CHAPTER 4: POTENTIAL ISSUES FOR DRINKING WATER RESERVOIRS...45 Water Quality Issues Disinfection By-Product Precursors Taste and Odor Dissolved Oxygen, ph, Light, and Temperature Other Aesthetic Standards Biomass Issues Intake Clogging with Plant Material Associated Periphyton Clogging Sediment/Particulates Recreational Impacts Benthic Algae Relation to Milfoil Growths Impact on Water Quality Associated Ecological Issues Nutrient Releases Interactions with Phytoplankton Community Effects on Fish, Macroinvertebrates and Zooplankton CHAPTER 5: POTENTIAL CONTROL METHODS FOR NUISANCE AQUATIC VEGETATION, WITH EMPHASIS ON MILFOIL SPECIES...57 Factors for Evaluating Control Methods Prevention Approaches Physical/Mechanical Controls Benthic Barriers Dredging Reverse Layering Dyes and Covers Harvesting Hand Pulling Suction Harvesting Mechanical Harvesting Rotovation Hydroraking Water Level Control Chemical Controls Overview Diquat Endothall ,4-D Fluridone Triclopyr Biological Controls Overview Herbivorous Fish Herbivorous Invertebrates Plant Competition vi

9 No Management Alternative Overview CHAPTER 6: CONSIDERATIONS FOR APPLICATION OF TECHNIQUES TO DRINKING WATER SUPPLIES Reservoir Characteristics Morphometry and Bathymetry Sediment Features Nature and Depth of Intake Pattern of Water Withdrawal Additional Water Uses Summary of Key Reservoir Features Affecting Management of Milfoil Aquatic Plant Community Invasive Milfoil Species Ecology Native plant assemblage Rare, Threatened, or Endangered (RTE) species Summary of Key Biological Features Affecting Management of Milfoil Watershed Characteristics Upstream Milfoil Presence Inflow From Watershed Downstream Water Uses Watershed Stakeholders Summary of Key Watershed Features Affecting Management of Milfoil Identification of Candidate Control Options Preventive Measures Threshold Control Criteria Screening of Control Options Criteria for Selection of Preferred Control Methods Technical Feasibility Duration of Effect Effectiveness Against Target Species Potential Collateral Impacts Integration with Other Methods Costs Regulatory Response Stakeholder Response CHAPTER 7: SURVEY OF DRINKING WATER SUPPLIERS AND RESERVOIRS Introduction Methodology Results Discussion CHAPTER 8: CASE STUDIES IN MILFOIL MANAGEMENT Introduction Wachusett Reservoir Milfoil Assessment vii

10 Management Efforts Overall Result of Management Lessons Learned and Recommendations Reservoir A Milfoil Assessment and Control Lessons Learned and Recommendations Sebago Lake Milfoil Assessment Management Efforts Overall Result of Management Lessons Learned and Recommendations Lake Massabesic Milfoil Assessment Management Effort Overall Result of Management Lessons Learned and Recommendations Lake Auburn Milfoil Assessment Management Efforts Overall Result of Management Lessons Learned and Recommendations Lake Auburn Rapid Response Plan for EWM General Observations from Case studies CHAPTER 9: CONCLUSIONS AND RECOMMENDATIONS APPENDIX A: SPECIES KEY AND PICTURES APPENDIX B: QUESTIONNAIRE APPENDIX C: GLOSSARY REFERENCES ABBREVIATIONS viii

11 LIST OF TABLES Table 3-1. Table 3-2. Table 4-1. Table 5-1. Table 5-2. Table 6-1. Table 7-1. Table 7-2. Current species names, synonyms, and common names for selected species of Myriophyllum...14 Myriophyllum species grouped into three classes according to potential for invasive and/or nuisance status, with distribution by state or province...17 Comparison of biomass for plant assemblages with and without milfoil Management options for control of aquatic plants...59 Herbicides used to control milfoil and information relevant to use in potable water supplies...97 Threat analysis matrix for milfoil in reservoirs Total water volume and water volume that potentially has milfoil in billion gallons (BG) for both milfoil-infested and un-infested Reservoirs Assessed and impaired waters from based on Section 305b reports Table 8-1. Threat analysis matrix for EWM in Wachusett Reservoir Table 8-2. Table 8-3. Table 8-4. Threat analysis matrix for VWM in Reservoir A Threat analysis matrix for VWM in Sebago Lake Threat analysis matrix for VWM in Lake Massabesic Table 8-5. Threat analysis matrix for VWM in Lake Auburn Table 8-6. Threat analysis matrix for EWM in Lake Auburn ix

12 x

13 LIST OF FIGURES Figure 2.1 Figure 2-2 Typical aquatic plant zones in lakes and ponds...6 Conceptual model of scales of environmental factors affecting macrophytes...10 Figure 3-1a Distribution of Myriophyllum spicatum in North America as of Figure 3-1b Distribution of Myriophyllum spicatum in the USA as of Figure 3-2 Distribution of Myriophyllum heterophyllum in North America as of Figure 3-3 Distribution of Myriophyllum aquaticum in North America as of Figure 3-4 Distribution of Myriophyllum humile in North America as of Figure 3-5 Distribution of Myriophyllum sibiricum in North America as of Figure 3-6 Distribution of Myriophyllum hippuroides in North America as of Figure 3-7 Distribution of Myriophyllum verticillatum in North America as of Figure 3-8 Distribution of Myriophyllum pinnatum in North America as of Figure 3-9 Distribution of Myriophyllum laxum in North America as of Figure 3-10 Distribution of Myriophyllum farwellii in North America as of Figure 3-11 Distribution of Myriophyllum quitense in North America as of Figure 3-12 Distribution of Myriophyllum ussuriense in North America as of Figure 3-13 Distribution of Myriophyllum tenellum in North America as of Figure 3-14 Distribution of Myriophyllum alterniflorum in North America as of Figure 4-1 Distribution of biomasses of plant assemblages without invasive milfoil (upper panel), with some milfoil (middle panel) and dominated by milfoil (lower panel) Figure 6-1 Decision tree for the control of invasive milfoil (Myriophyllum spp.) Figure 7-1 Figure 7-2 Figure 7-3 Percent of questions answered in each section of the questionnaire Types of reservoirs with and without milfoil infestation Residence time in milfoil-infested and un-infested reservoirs xi

14 Figure 7-4 Figure 7-5 Figure 7-6 Figure 7-7 Sediment types in the 0-15 ft contour in both milfoil-infested and un-infested reservoirs Presence or absence of upstream milfoil-infested waterbodies for both milfoil infested and un-infested reservoirs Percent of the water column occupied by plants in milfoil-infested and un-infested waterbodies Plant controls used in both milfoil-infested and un-infested reservoirs Figure 8-1 Site map of Wachusett Reservoir Figure 8-2 Areas in Upper Thomas and Thomas Basins managed in Figure 8-3 Variable water milfoil distribution in Sebago Lake as of Figure 8-4 Lake Massabesic and its immediate watershed in New Hampshire Figure 8-5 Bathymetry of Lake Massabesic Figure 8-6 Figure 8-7 Distribution of variable water milfoil in Lake Massabesic Milfoil susceptibility in Lake Auburn and the Basin, with known VWM distribution as of late xii

15 FOREWORD The Awwa Research Foundation is a nonprofit corporation that is dedicated to the implementation of a research effort to help utilities respond to regulatory requirements and traditional high-priority concerns of the industry. The research agenda is developed through a process of consultation with subscribers and drinking water professionals. Under the umbrella of a Strategic Research Plan, the Research Advisory Council prioritizes the suggested projects based upon current and future needs, applicability, and past work; the recommendations are forwarded to the Board of Trustees for final selection. The foundation also sponsors research projects through the unsolicited proposal process; the Collaborative Research, Research Applications, and Tailored Collaboration programs; and various joint research efforts with organizations such as the U.S. Environmental Protection Agency, the U.S. Bureau of Reclamation, and the Association of California Water Agencies. This publication is a result of one of these sponsored studies, and it is hoped that its findings will be applied in communities throughout the world. The following report serves not only as a means of communicating the results of the water industry's centralized research program but also as a tool to enlist the further support of the nonmember utilities and individuals. Projects are managed closely from their inception to the final report by the foundation's staff and large cadre of volunteers who willingly contribute their time and expertise. The foundation serves a planning and management function and awards contracts to other institutions such as water utilities, universities, and engineering films. The funding for this research effort comes primarily from the Subscription Program, through which water utilities subscribe to the research program and make an annual payment proportionate to the volume of water they deliver and consultants and manufacturers subscribe based on their annual billings. The program offers a cost-effective and fair method for funding research in the public interest. A broad spectrum of water supply issues is addressed by the foundation's research agenda: resources, treatment and operations, distribution and storage, water quality and analysis, toxicology, economics, and management. The ultimate purpose of the coordinated effort is to assist water suppliers to provide the highest possible quality of water economically and reliably. The true benefits are realized when the results are implemented at the utility level. The foundation's trustees are pleased to offer this publication as a contribution toward that end. David E. Rager Chair, Board of Trustees Awwa Research Foundation Robert C. Renner, P.E. Executive Director Awwa Research Foundation xiii

16 xiv

17 ACKNOWLEDGMENTS The authors of this report are indebted to the members of the Project Advisory Committee for the Awwa Research Foundation: Dr. G. Dennis Cooke, Kent State University, Ohio Dr. Stanley A. Nichols, Wisconsin Geologic and Natural History Survey Dr. Robert C. Hoehn, Virginia Polytechnic Institute and State University Mr. Steven Lohman, Denver Water Company, Colorado We also acknowledge the efforts of the AwwaRF Project Officers, Mr. Jason Allen at the start of the project, Mr. Ryan Ulrich during the middle portion of the project, Ms. Jennifer Warner for later stages of the project effort, and Ms. Linda Reekie for the final stages of the process. A large number of individuals assisted through the provision of data and discussions of various aspects of milfoil ecology and impacts on water supply, recreation and aquatic ecology. The generosity of the scientific and management communities addressing lake and reservoir problems made this report possible, and we would like to thank the many members of the American Water Works Association, North American Lake Management Society, and Aquatic Plant Management Society who gave of their time and data to improve this project. xv

18 xvi

19 EXECUTIVE SUMMARY The overall objective of the Milfoil Scoping Project is to improve the understanding of drinking water managers with regard to detecting, mitigating, and controlling the potential impacts of invasive water milfoil on their reservoirs. The review of the current literature and experience presented in this report provides these managers with important and practical knowledge regarding Myriophyllum species, including taxonomy, distribution, ecology, and current methods of control of forms with potentially negative impacts on drinking water supplies. General principles of aquatic plant ecology and control are presented in a manner that links plant management to overall reservoir operation for maximum beneficial use. Factors that influence raw water quality, such as light availability, water temperature, nutrients, dissolved carbon, ph, and water velocity, also affect the potential for plant nuisances. Invasive species, including several species of milfoil, can outcompete native plants and achieve higher densities and greater impact on water quality. Where light penetrates to a hospitable bottom substrate, rooted plants will grow. Control of plants, where necessary, therefore involves either directly attacking the plants or altering light or substrate conditions to reduce plants or favor one type over another. There are 14 species of Myriophyllum found in the USA, and these have been divided among three classes. Class I milfoils are those milfoils, both native and non-native to North America, considered to be highly invasive. Class II milfoils are those milfoils that have proven to be or have the potential to be problematic in specific cases, but are not likely to generate widespread concern. Class III milfoils are not known to present management issues, and some are actually protected under state laws as rare plants. Emphasis is placed on the milfoils known to become dominant and potentially cause problems in drinking water reservoirs, including M. spicatum (Eurasian water milfoil), M. heterophyllum (variable water milfoil), and M. aquaticum (parrotfeather). These species have somewhat different water quality and substrate preferences, such that they are not often found together, but they are not mutually exclusive and each can become dominant in a reservoir, limiting other growths and impairing uses. Potential issues associated with aquatic vegetation in drinking water reservoirs include water quality issues, such as disinfection by-product precursors (DBPPs) and alteration of oxygen or ph; biomass issues, such as clogging or creation of excessive suspended solids upon decay of plants; potential encouragement of benthic or attached algae responsible for taste and odor; and overall ecological issues, such as shifts in biological structure that favor higher algal biomass in the reservoir. Milfoil can become a problem for drinking water supplies where it is dominant and covers an extensive area, either in the reservoir overall or in the general vicinity of the intake. Methods currently used for control of aquatic vegetation in general and milfoils specifically are outlined to provide an overview of the potential milfoil management tools that a drinking water manager might consider. The various methods are divided between physical/mechanical (benthic barriers, dredging, reverse layering, dyes and covers, multiple forms of harvesting, and water level control), chemical (herbicides), and biological (herbivorous fish and invertebrates, plant pathogens, competition with other plants) controls. There is no one technique that will meet all needs, and selection of the most appropriate approach for a given reservoir will depend on which milfoil is involved and site-specific factors that affect the efficacy of each possible control technique. Often more than one method will be needed to gain and maintain control. Eradication is rarely achieved, except where an invasion is detected early and promptly addressed. In drinking water supplies, physical/mechanical methods are generally xvii

20 preferred to chemical approaches, although the latter can be more effective in bringing a problem under control under a wider range of circumstances. Biological controls are viewed as having great future potential, but remain less reliable than physical or chemical measures. Information and issues that drinking water managers will need to consider when developing a site-specific aquatic vegetation control program for their reservoir are described. These include reservoir and watershed characteristics, composition and coverage of the aquatic plant community, reservoir operations and other water uses, monitoring programs, preventive measures, and the advantages and limitations of each possible control technique with regard to the water supply. Features that either minimize the need for milfoil control or constrain which techniques can be applied are identified. A process for performing a threat analysis is outlined, and this can be instrumental in determining the level of priority to be assigned in a Rapid Response Plan. Protocols for detecting and quantifying milfoil invasions are also provided. Based on site-specific information, candidate control options can be identified and subjected to critical evaluation, resulting in a suite of preferred control methods to be implemented. While management decisions are necessarily reservoir specific, the need for control will generally be a function of the extent of potential milfoil coverage and the probability of invasion. The preferred controls will depend on the extent of actual coverage, proximity to intakes, presence of other sensitive species, regulations governing possible methods, and consumer or public perceptions. Preventive measures are stressed, but the range of mitigative measures is discussed and a graphic framework for reaching management decisions is provided. The extent of actual impacts across the USA is not known, but was the subject of a questionnaire survey and additional investigation of state records conducted as part of this Milfoil Scoping Project. There is a low level of awareness of potential milfoil threats, and about the potential impacts of aquatic vegetation in general on water supplies. There is also a surprising lack of knowledge of reservoir ecology and even basic physical and chemical features among water suppliers as a whole. Up to about 30% of drinking water reservoirs in the USA may be impacted to some degree by nuisance milfoils, although the value seems more likely to be 10-15%. However, it is not 10-15% of each reservoir that is threatened, but rather a large portion of a few reservoirs. Additionally, the actual volume of all drinking water affected is probably <10%, based on the best available data. This may explain why water suppliers do not perceive a great threat overall, but indicates that awareness should be raised among those with susceptible reservoirs. Case studies of milfoil invasions and management responses are provided to illustrate the range of susceptibility, use impacts, response planning, and level of management. In general, utilities tend not to have complete Rapid Response Plans, do not monitor with any regularity for invasive species, and delay management actions after discovery of milfoil invasions to the point where typical rapid response actions cannot lead to eradication. There is a strong aversion to chemical controls and biological controls are almost never applied, but physical methods are often applied to minimize the spread of milfoil and preventive actions limit further introductions. The need for threat analysis and response planning prior to invasion is apparent. xviii

21 CHAPTER 1 INTRODUCTION ENSR Corporation (ENSR) was selected by the Awwa Research Foundation (AwwaRF) to conduct a study of the impacts and management of water milfoil, a plant genus including potential nuisance aquatic species, on the water supply industry (AwwaRF Project #03024). The overall objective of the Water Milfoil Scoping Study: Impacts on the Water Supply Industry (the Milfoil Scoping Project) is to improve the understanding of drinking water managers with regard to detecting, mitigating and controlling the potential impacts of invasive water milfoil on their reservoirs. Accordingly, the Milfoil Scoping Project will summarize relevant knowledge regarding Myriophyllum spp. taxonomy, distribution, ecology and control, and establish a framework for using this knowledge to manage milfoil where necessary in drinking water supplies. The Principal investigators for ENSR are Drs. Kenneth Wagner ( ; kwagner@ensr.aecom.com) and David Mitchell ( ; dmitchell@ensr.aecom.com). Specifically, the project consists of four major activities: 1. A review of the current literature on milfoil ecology, distribution and control, with focus on the patterns of invasion/infestation in North America by problem species and assessment of the methods for their control; 2. Survey of drinking water utilities, state agencies, lake associations, volunteer monitoring networks, and reservoir management professionals to assess occurrence of potential problems with milfoil and preferred control approaches; 3. Review past and current case studies of milfoil management and develop several to represent the impact range of invasive milfoils (as suggested by Tasks 1 and 2); 4. Development of an interactive guide to assessment of potential milfoil problems and evaluation of feasible management methods, facilitating analysis and informed decision-making by the water utilities. This report represents the summary of the first three tasks. It details Myriophyllum ecology and distribution for selected species as well as providing information on current aquatic vegetation control methodologies, and their potential application to milfoil in drinking water supplies. The results of the questionnaire survey and follow-up contact with state agencies are provided. A set of case histories representing milfoil issues that are likely to be encountered by drinking water suppliers is also offered. PROJECT BACKGROUND Water milfoil is the term commonly given to a number of aquatic macrophyte species in the genus Myriophyllum, including both native and non-native (introduced) species. Water milfoil has received considerable attention over the last 20 years due to the invasive nature of some of the non-native species. These invasive species often form dense monoculture stands and out-compete native plant assemblages. Many studies have shown that invasive milfoil can reduce plant and wildlife biodiversity, aesthetic value, habitat quality, water quality and real estate value 1

22 of waterfront property. The potential for impacts on water supplies is known but largely unquantified. Species of milfoil are varied in their morphology and role in the aquatic environment, but at least three introduced species have invasive tendencies and can cause problems for water supply reservoirs and recreational lakes (Mattson et al., 2004). At least one native species sometimes causes resource impairment, but many of the native species are on protected lists in multiple states. Discerning species and understanding the ecology of the invasive forms are important to determining potential impacts and selecting control strategies, if needed. Some of this is already known (e.g., Crow and Hellquist, 1983; 2000; Mattson et al., 2004) for regional areas, but needs extension to a national scale. One of the major objectives of the Literature Review task is to provide updated maps of current milfoil species distribution and relative importance on a national scale. Milfoil abundance appears independent of watershed features; no amount of watershed management is likely to control milfoil, making this an in-lake management challenge. This is unlike the situation with algal blooms, where watershed management is almost always an essential component of a control strategy applied by many utilities. Water quality may affect the type of milfoil present, but whether milfoil has an impact on water supply is more a matter of reservoir features such as morphometry, sediment types, and intake location. Linkages between milfoil and water supply problems are often perceived as indirect effects. These include odorous algal mats often associated with milfoil growths, generation of disinfection by-product precursors (DBPPs) from milfoil, and fluctuations in ph or oxygen induced by photosynthesis and decay. Clogging of intake structures or screens with plant material is a more direct potential consequence. Discerning the role of milfoil in these impacts may be complicated by other plants in the system and the spatial separation between rooted plant growths and most intake structures. A major objective of the Literature Review task is to review and evaluate existing research in this regard, and to provide accurate statements of what is and is not known of linkages between milfoil and water supply problems. Many water supplies are multi-use waterbodies that are threatened by these invasive species, particularly in the northern half of the United States. The problems caused by milfoil in lakes, where the primary uses are recreation and fish and wildlife propagation, include physical interference with swimmers and boaters, loss of biodiversity, and undesirable alteration of habitat features for many species (Mattson et al., 2004). These same issues may be important in some water supplies, but where drinking water is the top priority use of a waterbody, issues relating to water quality are more critical. Invasive milfoil can present significant direct or indirect challenges in drinking water reservoirs with regard to water quality, taste and odor problems, DBPP production, and intake and reservoir operation. Most aquatic vegetation assessment and control effort has been focused on recreational lakes, but much of what has been learned is applicable to reservoirs as well. Considerable knowledge of control techniques has been summarized in the MA DCR/DEP (2004) publication (Mattson et al., 2004) and accompanying practical guide (Wagner, 2004), in the in-lake chapter (Wagner, 2001) of Managing Lakes and Reservoirs, and in the third edition of Restoration and Management of Lakes and Reservoirs (Cooke et al., 2005). However, management options to control these plants in reservoirs are often subject to greater restriction than in conventional recreational lake settings due to regulation of drinking water, consumer concerns and public scrutiny. One of the major objectives of the Milfoil Scoping Study is to provide guidance for 2

23 utilities to assess potential milfoil problems and evaluate feasible management alternatives where warranted. METHODS USED FOR LITERATURE REVIEW The literature review was based on relevant information and articles selected from scientific journals, existing compilation of lake management information, aquatic plant taxonomy reference books, and interviews with experts in the field. Articles were identified by review of appropriate scientific journals, database search, citation lists, and research team familiarity with much of this literature through years of experience. For our initial approach we referred to lake management and restoration manuals including Wagner (2001; 2004), Mattson et al. (2004), and Cooke et al. (2005) to encompass the North American experience with milfoils. To expand this literature review, papers were identified using two approaches: (1) examination of reference lists in current review articles and relevant scientific papers; and (2) a search of electronic databases of scientific literature. Reference lists in the articles obtained during this literature review were examined for additional relevant articles. Several useful articles and additional information were provided to us directly by contacted researchers and experts. The Cambridge Scientific Abstracts (CSA) database was the primary database used to search electronically for current literature. The CSA database includes the following databases: Aquatic Sciences and Fisheries Abstracts (1978-present), Biological Sciences (1982-present), Biology Digest (1989-present), MEDLINE (1989-present), Oceanic Abstracts (1981-present), Plant Science (1994-present) and Water Resources Abstracts (1967-present). JSTOR, an online repository for older scientific literature was queried for relevant scientific articles. The popular non-scientific Google search engine provided a few references, mostly in the form of so-called grey literature reports. These databases were searched by looking for keywords contained in article abstracts, titles, and keyword lists. While considerable effort was spent in identifying candidate articles for this review, the search should rightfully be considered representative and not comprehensive in scope. The literature search focused on identifying references referring to the taxonomy, biology, ecology, and control of Myriophyllum species. During the search phase, many papers discussing general taxonomy and ecology for North American species of Myriophyllum were identified. ENSR obtained information on 14 Myriophyllum species (both native and introduced) and their taxonomy, distribution, ecology, regulatory status, control and linkage to drinking water problems. Taxonomic and distribution information was derived from botanical summaries, including the Crow and Hellquist (2000) update of Fassett s Manual of Aquatic Plants, as well as relevant websites. ENSR focused on the primary problem species, Myriophyllum spicatum (Eurasian water milfoil), M. heterophyllum (variable milfoil), and M. aquaticum (parrotfeather), but all other species of milfoil in the USA are covered. REPORT ORGANIZATION The results and findings were summarized into the project deliverables in using the format and content following the AwwaRF s Format-Style Guide for Preparing Research Reports, Fifth Edition (AwwaRF, 2003a). The report is organized as follows: 3

24 Chapter 1: Introduction Chapter 2: General Introduction to Aquatic Vegetation Chapter 3: Myriophyllum Ecology and Distribution Chapter 4: Potential Issues for Drinking Water Reservoirs Chapter 5: Potential Control Methods Chapter 6: Considerations for Application of Techniques to Drinking Water Supplies Chapter 7: Survey of Drinking Water Suppliers and Reservoirs Chapter 8: Case Studies in Milfoil Management Chapter 9: Conclusions and Recommendations Appendix A. Milfoil Species Pictures and Keys Appendix B. Questionnaire Appendix C. Glossary References Abbreviations 4

25 CHAPTER 2 GENERAL INTRODUCTION TO AQUATIC VEGETATION Excessive aquatic plant growth can interfere with protected reservoir uses such as drinking water, boating, swimming and fishing. It can also impede navigation, block pumps and sluices and cause encroachment, silting and flooding (Seagrave, 1988), and can affect the overall health and aesthetics of a reservoir. In some states, bathing beaches must meet a four foot minimum Secchi depth requirement to provide adequate visibility and prevent swimmer entanglement, although more recent public health statutes have adopted a less objective narrative standard. In addition to impeding recreational uses, excessive aquatic plant growth can also result in the degradation of fish and wildlife habitat, strong fluctuations in dissolved oxygen concentrations and ph, clogging of water intakes, associated taste and odor concerns, an increase of phosphorus release from sediments and an acceleration of sediment deposition. In many waterbodies, the bulk of the excessive plant biomass is due to a limited number of invasive species (including Myriophyllum). In order to understand the potential causes of milfoil success as well as potential means of control, it is first necessary to briefly introduce some of the natural factors which control macrophyte growth. The ability of invasive species, particularly exotic or non-native species, to exploit these conditions is then considered. FUNDAMENTALS OF AQUATIC PLANT COMMUNITIES Taxonomy Diversity One of the first and most important issues when contemplating control and management of excessive aquatic vegetation involves identifying precisely what species are present. When considering management of any aquatic plant species it is important that it and all members of the plant community be correctly identified and mapped at sufficient detail. This can be difficult as there are sometimes discrepancies between experts as to which plants are present in the state and which are in need of management, and methods for plant mapping are not standardized. Correct identification is essential in order to prevent the eradication of rare and endangered species and to document the plant population so that it can be monitored over time (Hellquist, 1993; Crow and Hellquist, 2000). Management techniques often target specific plant species and an incorrect identification of the target plant or non-target plants could result in eradication or negative impacts to a desirable species or ineffective control of the target species. The term aquatic macrophytes is a very inclusive and poorly-defined term analogous to referring to non-microscopic terrestrial plants, which could include everything from a moss to an oak tree. Aquatic macrophytes are phylogenetically diverse, including aquatic angiosperms (cattails, rushes, pondweeds), bryophytes (mosses), quillworts, ferns, and charophytes (macroscopic algae). This document focuses on milfoil, but some understanding of the wide range of potential plant species found in drinking water reservoirs is needed by reservoir managers. 5

26 Functional Habitat Classification One useful way of classifying aquatic macrophytes conceptually is based on their habitat and location relative to the waterbody surface. There are four growth forms of aquatic plants that are commonly recognized (Figure 2-1): floating unattached, floating attached, submergent and emergent (Riemer, 1984; Kishbaugh et al. 1990). Some plants consist of both submerged and floating leaves, and some have different growth forms under different abiotic conditions (submergent and emergent forms), so the groupings are not quite so distinct. There are many taxonomic groups, but the above categories are often the most useful for understanding the causes of a macrophyte problem and determining an appropriate management strategy. In fact, within each category, many species may look very similar as their growth habit responds to common reservoir conditions. However, even though many macrophyte species appear similar, their propensity to cause problems in reservoirs varies. Effective management of macrophytes requires identification, usually to the species level. For example, the herbicidal active ingredient fluridone is much more effective on Eurasian water milfoil than on variable water milfoil. A drawdown targeting any species of milfoil may reduce densities of Potamogeton robbinsii but may increase densities of Potamogeton amplifolius, based on their overwintering strategies (vegetative vs. seeds). Figure 2-1 Typical aquatic plant zones in lakes and ponds (from Kishbaugh et al., 1990). 6

27 Rooted aquatic plants typically grow from a root system embedded in the bottom sediment. Unlike algae, they derive most of their nutrients from the sediments just like terrestrial plants, but they may be able to absorb nutrients from the water column as well. Because they need light to grow, they cannot exist where the reservoir bottom is not exposed to sufficient light. The part of a reservoir where light reaches the bottom is called the photic zone. For many such plants, nutrients in the sediments may be in excess and growth is limited by light, particularly during early growth when the plant is small and close to the bottom. Emergent plants solve the light problem by growing out of the water, but that limits them to fairly shallow depths. Freefloating plants also are not limited by light, except in cases of self-shading when growths are dense, but cannot use the sediments as a source of nutrients. Finally, floating-leaf plants have attempted to achieve the best of all worlds by having their roots in the sediment and leaves at the surface. Although less limited by water depth than emergent plants, floating-leaf plants still have depth limits. Factors Important for Growth of Aquatic Macrophytes The factors which influence the growth of aquatic macrophytes have been well characterized (Barendregt and Bio, 2003; Carr et al., 1997; Wetzel, 2001), and are briefly summarized here. A generalized equation for describing aquatic macrophyte production is given by: S = P R L where: S = standing crop or biomass over time; P = biomass produced by gross photosynthesis; R = biomass lost to respiration; and L = plant loss by washout, decay, or grazing. Primary productivity, or the rate of biomass generation, is influenced by environmental factors that include light availability, water temperature, nutrients, dissolved inorganic carbon, ph, and water velocity. Light availability is determined by a number factors including incident radiation, surface reflectance, light attenuation, bottom reflectance, shading by other macrophytes, encrusting periphyton, or overhanging riparian vegetation (Carr et al., 1997; Wetzel, 2001). In high density rooted macrophyte beds, light availability, particularly near the bottom of the plant layer, is likely to be a significant growth limiting factor. It should be recognized that such light limitation does not apply to floating macrophytes such as duckweed (Wolffia, Lemna spp.) or water lilies (Nymphaea, Nuphar spp.) or to emergent plants such as cattail (Typha spp) or bulrushes (Scirpus spp). However, some milfoil species can create canopies that shade out other submergent forms. Water temperature influences macrophyte growth primarily through its effect on the rate of chemical or enzymatic reactions that can control the rate of photosynthesis and therefore productivity. It is generally assumed that net photosynthesis is sluggish at low temperatures, increases with temperature until a species-specific optimum range is met and then declines 7

28 sharply once this optimum has been exceeded. Except in rare circumstances, it is unlikely that the optimum temperature will be exceeded in most northern USA reservoirs. Nutrients (phosphorus and nitrogen) are important to macrophyte growth, but there is disagreement as to whether they are typically the major limiting factor for macrophytes in the field. Dense macrophyte beds have been often observed downstream of known point sources of nutrients (e.g., wastewater treatment facility discharges) in rivers, indicating a positive response from nutrient enrichment. However, several investigators maintain that the ambient levels of phosphorus and nitrogen in the water are secondary to light or even carbon availability as limiting factors (Carr et al., 1997). This uncertainty regarding nutrient-controlled growth is primarily due to the ability of many macrophyte to obtain nutrients either through uptake into the leaf from the water column or by the root from the sediments or interstitial pore water (Barendregt and Bio, 2003; Wetzel, 2001). Therefore, it may be argued that the macrophytes are responding more to the nutrients in the enriched sediments rather than those in the water column. This would not be true for the true floating macrophytes such as duckweed or watermeal, which would rely on assimilation of nutrients from the water column. However there are certainly many examples of dense macrophytes in waterbodies where nutrient concentrations are below detection in the water column, suggesting that the rate of supply produced by water movement can be important in keeping nutrient limitation from occurring in macrophytes. For milfoil species, nitrogen and phosphorus are obtained mainly from the sediment (Barko and Smart, 1981). Carbon availability, in the form of dissolved inorganic carbon (DIC), has been shown to be a limiting factor for submerged macrophytes under certain circumstances. This is one of the reasons given for the lower annual production of freshwater wetlands relative to coastal marine systems (Stevenson, 1988, cited in Carr et al., 1997). While DIC includes various forms including CO 2, H 2 CO 3, HCO 3 -, and CO 3-2, CO 2 is generally considered the preferred form of carbon for photosynthesis for aquatic plants (Wetzel, 2001). Based on the ph level of the water, CO 2 may be less available and many plants can alternatively use HCO 3 -, but not as efficiently (Wetzel, 2001). Some macrophytes are able to take up CO 2 from the interstitial water of sediments, thus providing an additional source. Finally, floating or emergent macrophytes are relatively immune to carbon limitation due to their ability to draw directly upon CO 2 stores in the atmosphere. Water velocity (current) has been shown to be important with regard to replenishment of both nutrients and carbon for macrophytes due to a decreased boundary layer thickness at the plant surface (see reviews in Carr et al., 1997). However, a water velocity exceeding 1 meter per second has been associated with reduced macrophyte growth; (Carr et al., 1997; Barendregt and Bio, 2003), with a reported decrease in biomass and species numbers (Barendregt and Bio, 2003). As current velocity increases, the potential for scour and loss of plant fragments starts to increase and is an important factor in delineating the location of macrophyte beds within a river channel or in parts of some reservoirs. This is evidenced by the absence or poor growth of macrophytes in high-energy environments (e.g., main thalweg of river, wave-impacted shoreline, navigation channels subject to high amounts of propeller wash disturbance). Substrate quality or stability is a major factor in plant biomass, as evidenced by reduced macrophyte abundance on stones, cobble and gravelly environments as opposed to the often dense development on sandy, clay, or highly organic materials. In general, the substrate composition is largely determined by flow and energy regime, including stream velocity, 8

29 scouring potential, and wave generation and interaction with shallow sediments. As many reservoirs are created by damming streams or rivers, large portions of them become depositional zones for sediment that can support dense plant growths. Additional human influence can lead to localized areas of macrophyte abundance, such as in the depositional delta formed below a stormwater pipe or stream, or in the highly organic sediments often found downstream of a wastewater treatment facility. Yet highly organic sediments are often unstable and can also contain high concentrations of heavy metals or herbicides, resulting in growth inhibiting or even lethal conditions (Barendregt and Bio, 2003). Variability of plant growth can therefore be high. Grazing of aquatic macrophytes can be an important factor in controlling macrophytes in some parts of the United States. There are many species, over many trophic levels ranging from aquatic invertebrates (e.g., aquatic weevils, crayfish) and fish (e.g., grass carp) to waterfowl (e.g., many duck species) and mammals (e.g., muskrat) which consume aquatic macrophyte biomass In some cases, grazing and accompanying bottom disturbance may lead to increased turbidity, which may have a more direct negative impact on macrophytes by sharply reducing water transparency. In the northern states, the duration of the growing season may limit availability of aquatic plants to support large grazer populations on a permanent basis. Barendregt and Bio (2003), for purposes of ecological modeling, make an attempt to integrate these potential controlling factors and the scale at which they act on macrophyte species in the landscape, as shown in Figure 2-2. The authors distinguish between factors operating at different levels, each influenced by the level above it and culminating in a distinct plant community determined by collective influences. 9

30 Figure 2-2. Conceptual model of scales of environmental factors affecting macrophytes (from Barendregt A. and A.M.F. Bio Relevant Variables to Predict Macrophyte Communities in Running Waters. Ecol. Modeling 160: with permission from Elsevier.) NUISANCE AQUATIC VEGETATION Characteristics of Nuisance Species Nuisance species is a generic term given to organisms (both fauna and flora) that display a marked ability to colonize or exploit a particular environment resulting in an impairment of some designated use, such as swimming or water supply. Typically, their replacement of the existing community members is considered detrimental in terms of reducing biodiversity or in more specific biological ways (e.g., loss of habitat structure, reduced wildlife function). Some species may act as nuisance species in some environmental settings but not in others, while other species rarely or never exhibit nuisance tendencies. The ability of an aquatic plant to become a nuisance is dependent on the interactions of many factors, among them reproductive and dispersal mechanisms, growth rate, competitive abilities for light and nutrients, presence of natural biological controls, resistance to and presence 10

31 of pathogens, and favorable abiotic conditions. Favorable abiotic conditions for a particular plant can include nutrient abundance, preferred water depth and sediment type, hardness or softness of water and ph. Occasionally a cycle of invasion and decline is observed in aquatic plants, attributable to the presence of pathogens (Shearer, 1994), the presence of herbivorous insects (Sheldon, 1994), competition between plant species (Titus, 1994, Madsen et al., 1991), or a change in abiotic conditions (Shearer, 1994). Exotic Species A large number of nuisance species are not native to the geographic area in which they are problematic. In some cases these non-native (or exotic) species have expanded their range through natural means, but it the large majority of such cases, it is through human introduction (either intended or inadvertent). The introduction of these species constitutes a major ecological disturbance that is often detrimental to native species of plants and to wildlife that rely on those native species. Accordingly, there is a great need for control of rooted exotic or non-native plants. The native plant communities in the ecosystem have evolved under long-term ecological conditions and relationships including inter-specific and intra-specific competition for nutrients, space and sunlight; presence of natural enemies like insects, waterfowl and fish; and a range of environmental conditions such as temperature, ph and mineral content. These relationships tend to keep any one native species from dominating and encourage a diverse plant community. Introduced species are often able to out-compete native vegetation because of the absence of natural enemies and competitive pressures and, in some instances such as several submergent milfoil species (Myriphyllum spicatum, M. hetrophyllum or M. aquaticum), the floating-leaf water chestnut (Trapa natans), or the emergent purple loosestrife (Lythrum salicaria), can result in a near monoculture of the introduced species. Non-native species, unlike the natural biota and even the non-native biota introduced more than a hundred years ago, have few or no enemies, and are often invasive pests that can totally dominate or even eliminate native populations. They are easily introduced in a variety of ways, such as through the aquarium and horticulture trades. Once established in a reservoir, most nuisance species are extremely difficult to eradicate, and waterfowl, boats and water flow may facilitate their spread to other locations. Introduced non-native species can displace a healthy and desirable aquatic community and produce economically and recreationally severe impacts even though no other change has occurred in the watershed. The introduction of a non-native and undesirable species can result from the actions of a single person who does not realize the eventual impact and/or may not be aware that he/she has introduced the non-native species. There are many examples of non-native species which have successfully invaded new ecosystems. Introductions of Eurasian milfoil (M. spicatum) in Lake Champlain (Vermont/New York), Lake George (New York), Okanagan Lake (British Columbia) and many lakes in southern New England threaten otherwise healthy lakes. Within just a few years, a small patch of the introduced species can grow to fill the littoral zone of a lake or reservoir, top to bottom. Another nuisance species, fanwort (Cabomba caroliniana), is a popular aquarium plant. Many believe it was introduced from freshwater aquaria (Les, 2002). Hydrilla (Hydrilla verticillata) and Brazilian elodea (Egeria densa) are other invasive species spread by the aquarium trade. 11

32 Purple loosestrife, a beautiful non-native wetland plant, completely crowds out native species and creates stands so dense that wildlife habitat is degraded. It was introduced by horticulturists and gardeners desiring the beauty of the plant for their area (Les, 2002). Water hyacinth (Eichornia crassipes) and parrotfeather (Myriophyllum aquaticum) are other species commonly sold in garden centers. Many states have or are considering laws to prevent the introduction and spread of invasive species, but in many cases much damage has already been done. In most cases, introduced species demand special attention. While an overabundance of native species and diminution of desired uses can be managed over time, introduced species generally require quick action if eradication is to be achieved. The environmental cost of delay is usually higher than the risk of non-target damage from immediate use of most control options. The quicker the response, the smaller the degree of intervention needed to protect the environment. It may be difficult to impossible to actually eradicate an invasive species, but the probability of achieving and maintaining control is maximized through early detection and rapid response. 12

33 CHAPTER 3 MYRIOPHYLLUM TAXONOMY, ECOLOGY AND DISTRIBUTION Several species of milfoil are among the greatest aquatic nuisance species, particularly in the northern USA. As a critical first step in the management of water milfoil, it is important to appreciate the taxonomic diversity of this plant family, their current geographical distribution and habitat preferences. This section provides a brief review of the distribution and ecology of important North American Myriophyllum species, including both exotic invasive species. as well as those unlikely to grow to nuisance populations. A taxonomic key and photographic guide to the species described below are provided in Appendix A. THE GENUS MYRIOPHYLLUM The genus Myriophyllum, water milfoil, is almost cosmopolitan in nature. Approximately 60 species occur world-wide in three main geographic centers: Australia, North America, and India/Indo-China (Orchard 1981). To date, species in the genus Myriophyllum are found on every continent, except Antarctica. For nearly all introduced species, introductions are the result of the aquaria and aquatic gardening industries. Marketing of Myriophyllum species is wide-spread in these markets due to their feather-like appearance and hearty nature. All species in the Myriophyllum genus are flowering perennial or annual herbs (Aiken, 1981). While most species in this genus are aquatic, some take residence on muddy shores of aquatic habitats. All species except for M. tenellum are hydrophytes with many finely divided leaves, giving most species a feather-like appearance. Water milfoil reproduction is mainly through plant fragmentation and lateral stems (stolons). Seed production does occur in most species, but seeds account for a limited amount of overall reproduction. The name Myriophyllum has Greek roots, and literally means too many leaves to count. Overlapping physical and vegetative characteristics, as well as some hybridization, make species within the genus Myriophyllum difficult to distinguish from one another. Vegetative similarities within the genus have led to difficulty in species differentiation, and confusion in the available scientific literature. This confusion has led to numerous taxonomic revisions and name changes (Table 3-1). Presence/absence information for individual states is also conflicting, and casts doubt on the accuracy of distribution maps and reported ranges. Often, species within this genus cannot be separated using individual specimens or specimens without flowers. Both aquatic and semi-terrestrial specimens may be found in the same location for some species, with markedly different physical characteristics. Appendix A provides a species key, photographs, and diagrams to aid in species identification, but genetic techniques may be necessary in some cases to achieve certainty. However, where a milfoil species achieves dominance and impairs water supply functions it may be sufficient to determine that it is not one of the protected species. 13

34 Table 3-1. Current species names, synonyms, and common names for selected species of Myriophyllum. Species Name Synonyms Common Name Myriophyllum spicatum Eurasian watermilfoil European watermilfoil Spike watermilfoil Myriophyllum aquaticum Enydria aquatica Parrotfeather Myriophyllum brasiliense Watermilfoil Myriophyllum proserpinacoides Myriophyllum quitense Myriophyllum elatinoides Andean watermilfoil Myriophyllum pinnatum Myriophyllum scabratum Cutleaf watermilfoil Potamogeton pinnatum Pinnate watermilfoil Green parrot's feather Myriophyllum laxum Lax watermilfoil Loose watermilfoil Myriophyllum farwellii Farwell's watermilfoil Myriophyllum tenellum Leafless watermilfoil Slender watermilfoil Dwarf watermilfoil Myriophyllum alterniflorum Slender watermilfoil Alternate-flowered watermilfoil Watermilfoil Myriophyllum heterophyllum Broadleaf watermilfoil Two-leaved watermilfoil Variable-leaved watermilfoil Various-leaved watermilfoil Myriophyllum verticillatum Myriophyllum verticillatum var. intermedium Whorled watermilfoil Myriophyllum verticillatum var. pectinatum Comb watermilfoil Myriophyllum verticillatum var. pinnatifidum Green watermilfoil Myriophyllum verticillatum var cheneyi Myriophyllum sibiricum Myriophyllum exalbescens Common watermilfoil Myriophyllum exalbescens var. magdalenense American watermilfoil Myriophyllum magdalenense Northern watermilfoil Myriophyllum spicatum var. capillaceum Short-spike watermilfoil Myriophyllum spicatum var. exalbescens Myriophyllum spicatum ssp. exalbescens Myriophyllum spicatum ssp. squamosum Myriophyllum spicatum var. squamosum Myriophyllum humile Burshia humilis Low watermilfoil Myriophyllum ambiguum var. limosum Myriophyllum procumbens Myriophyllum hippuroides Western watermilfoil Myriophyllum ussuriense 14

35 The following are general characteristics for species within the genus Myriophyllum: Habit Annual or perennial herbs, aquatic or emergent Stem Weak flexuous stems, often glabrous Root Adventitious roots often present at basal nodes Leaves Submerged leaves whorled, opposite, alternate or irregularly arranged around the stem, deeply pinnated. Emergent leaves whorled, opposite or alternate, and entire or toothed. Flowers Mostly sessile, usually unisexual, sometimes hermaphroditic, usually solitary in axils of emergent leaves or on flowering spike. 2-4 petals in male flowers, minute or absent in female flowers. Fruit Nut-like, 4-lobed (carpels), splitting into 4 mericarpals Habitat Fresh and brackish waters of marshes, swamps, lakes, ponds, ditches, kettles, sinkholes, streams, rivers, and springs. NORTH AMERICAN SPECIES OF MYRIOPHYLLUM The genus Myriophyllum has undergone taxonomic reclassification due to larger herbarium collections for study and comparison, as well as advances in identification tools such as DNA technology. Presently, 14 species of Myriophyllum are found in North American waters. Of these, two are non-native to the North American continent, Myriophyllum spicatum and Myriophyllum aquaticum. Considering only the United States distributions, four species native to some portion of the USA are wide-spread (Myriophyllum heterophyllum, Myriophyllum pinnatum, Myriophyllum sibiricum, Myriophyllum verticillatum), seven are found within relatively small geographic areas (Myriophyllum ussuriense, Myriophyllum alterniflorum, Myriophyllum farwellii, Myriophyllum humile, Myriophyllum laxum, Myriophyllum quitense, and Myriophyllum tenellum), and one is found in three small non-contiguous areas (Myriophyllum hippuroides). To facilitate research and focus efforts on species likely to be problematic to drinking water facilities, we empirically separated North American milfoil species into three distinct categories based on invasiveness and potential for becoming a management issue (Table 3-2). Species were assigned to Class I, II, or III depending on their behavior and ecology. Scientific literature and personal communication with regional experts on Myriophyllum were the basis for species classification. Class I milfoils are those milfoils, both native and non-native to North America, considered to be highly invasive in new territories. These milfoils are persistent and aggressive throughout most of their present range in the United States. When introduced, these plants can rapidly spread into native plant communities and displace the native plant species. In addition, they are prolific, can quickly spread throughout a water system, and are known to create large monocultures. North American milfoil species classified as Class I are M. spicatum, M. aquaticum and M. heterophyllum. The first two have been introduced to North America from other continents, while the last species is native to some southern areas, but has moved 15

36 northward and is suspected to have become particularly invasive as a consequence of hybridization (Moody and Les, 2002a). 16

37 Table 3-2. Myriophyllum species grouped into three classes according to potential for invasive and/or nuisance status, with distribution by state or province. 17 Class I Class II Class III State M. spicatum M. aquaticum M. heterophyllum M. sibiricum M. humile M. hippuroides M. verticillatum M. pinnatum M. laxum M. farwellii M. quitense M. ussuriense M. tenellum M. alterniflorum Alabama NX NX x x x Alaska x x x Arizona x x x x Arkansas x x x x x California x x x x x x Colorado NX x x Connecticut x x x x x x SC x SC E Deleware x x x x x x x x Florida PR x x x x x Georgia x x x x x T Hawaii x x Idaho NX x x x x Illinois x x x x x x Indiana x x x T E x Iowa x x x x x Kansas x x x x x Kentucky x x SC H Louisiana x x x x x Maine I I I x x x x x x Maryland x x x x E, EX Massachusetts x x x x E SC E x T Michigan x x x x T x x Minnesota x x x x x x x x Mississippi x x x x x Missouri x x x x x Montana NX x x Nebraska x x x x Nevada NX x x New Hampshire x x x x x T x x New Jersey x x x E x E E E New Mexico NX x x x x x New York x x x x x E T x T North Carolina NX x x x T x North Dakota x x x x x Ohio x x E T x E Oklahoma x x x x x x (continued) 17

38 Table 3-2. (Continued) 18 Class I Class II Class III State M. spicatum M. aquaticum M. heterophyllum M. sibiricum M. humile M. hippuroides M. verticillatum M. pinnatum M. laxum M. farwellii M. quitense M. ussuriense M. tenellum M. alterniflorum Orgeon W x x x x x x Pennsylvania x x E E x E E T Rhode Island x x x x x x x H South Carolina I, P x x x x South Dakota R x x x x Tennessee x x x T Texas NX x x x x Utah x x x Vermont NX NX NX x x x x x x Virginia x x x x x x x x x Washington NX NX x x x x x West Virginia x x x x x x Wisconsin x x x x x x x x x Wyoming x x x Alberta x x x British Columbia x x x x x x x x Labrador x x x Manitoba x x x x x New Brunswick x x x x x x x Newfoundland Island x x x x x Northwest Territories x x x Nova Scotia x x x x x x Nunavut x Ontario x x x x x x x Prince Edward Island x x x x Quebec x x x x x x x x Saskatchewan x x x x Yukon Territory x x E=Endangered T=Threatened SC= Special Concern NX=Noxious I=Invasive P=Pest EX=Extirpated PR=Prohibited W=Weed requiring quarantine R=Regulated (Letters other than x denote special designations by the corresponding state or province, in accordance with the legend above. As new specimens may be identified or discredited on a frequent basis, verify the designation of the target species with the natural heritage program in each respective state or province when planning management programs.) 18

39 Impacts caused by dense beds of Class I species may include but are not limited to: Decreased biological diversity and related ecological disruption Clogging of irrigation canals, hydroelectric systems and navigation passages Reduction or elimination of swimming and boating Alteration of water chemistry, including oxygen, ph and organic content Possible taste, odor and color issues Class II milfoils are those milfoils, all native to North America, that have been or have the potential to be problematic in specific cases, but are not likely to be a widespread concern. These species will typically persist in a given environment for an extended period of time, and may become invasive and problematic under certain optimal environmental conditions. In the event of an infestation of a Class II species, the potential impacts to the infested waterbody are similar to those caused by Class I species. North American milfoil species classified as Class II are M. humile, M. sibiricum, and M. hippuroides. Class III milfoils are those milfoils, all native to North America, not known to develop high biomass infestations and present management issues and therefore do not fall into either Class I or Class II. North American milfoil species classified as Class III are M. pinnatum, M. verticillatum, M. ussuriense, M. alterniflorum, M. farwellii, M. laxum, M. quitense, and M. tenellum. Milfoil species are described in the following sections, with information on taxonomy, geographic range, habitat preferences, reproduction, and indications of the regulatory status of each species as known. CLASS I MILFOIL SPECIES Myriophyllum spicatum Taxonomic Description Eurasian water milfoil, M. spicatum, or EWM, is a submersed perennial herb that attaches to the substrate with fibrous roots. The stems of EWM are slender, reddish-brown, and can reach 6 meters in length, typically branching near the surface of the water. The leaves are green, less than 5 centimeters in length, and contain at least 12 segments. When removed from the water, the leaves of EWM tend to collapse around the stem. Mature leaves are typically arranged in whorls of 4 around the stem, ranging from 3 to 6 on rare occasions. Flowers of EWM are located on a spike protruding from the water. Flowers are reddish to pink in color, each containing four petals, and are most often observed in August and September, although flowering is possible in June as well in some locations. The fruit of EWM is four-lobed and splits into four separate one-seeded nutlets. Pigment or DNA analysis is sometimes needed for species identification as a consequence of morphological variability and possible hybridization. Other milfoils share some of these characteristics. Reproductive parts are the most definitive character. In the absence of flowers and/or seeds, EWM does not produce turions and the most distinctive characteristics are the normally reddish stem tips, the 12 or more filaments on each side of the central axis of each leaf, and the truncated leaf tips. This latter feature gives leaf ends 19

40 the appearance of having been trimmed with scissors. EWM is sometimes confused with other species of milfoils, most notably the native northern water milfoil (M. sibiricum). Geographic Range and History of Invasion Eurasian water milfoil is native to Europe, Asia and northern Africa. It is believed to have been introduced to the Chesapeake Bay area in the 1880 s (Aiken et al., 1979); the first known sample of EWM was collected in a Washington, DC waterbody in 1942 (Couch and Nelson, 1985). By 2002, EWM had been reported in 45 of the 50 United States, and in the southern portions of Canada from Quebec to British Columbia (Madsen and Welling, 2002; see also Figure 3-1a). EWM has great potential for expansion due to an adaptive life history strategy, rapid vegetative growth, and carbohydrate storage in the root crowns, allowing for overwintering in cold climates (Adams and Prentki, 1982; Madsen, 1991; Madsen and Welling, 2002). Plant fragments are easily transported to new waterbodies by boats, trailers, fishing gear, wind, animals and currents (Aiken et al., 1979). In one study, Minnesota authorities found aquatic plants on 23% of all boats inspected (Bratager et al., 1996). Plant fragments transported to new waterbodies can become rooted and form new shoots. Current North American Distribution Figure 3-1a. Distribution of Myriophyllum spicatum in North America as of 2003 Source: 20

41 The distribution of M. spicatum in the United States as of 2003 is provided in Figure 3-1a. However, it is known that this distribution map is not up to date. A slightly more recent map, generated by the United States Geological Survey, shows the U.S. distribution (Figure 3-1b). Figure 3-1b. Distribution of Myriophyllum spicatum in the USA as of 2004 Source: Even in the brief period since this map was created, the State of Maine has reported one occurrence of EWM and claims to be the last of the 48 contiguous states to get it; suggesting that both Montana and Wyoming (shown without EWM in the figure above) have this plant already. Habitat Requirements/Preferences Eurasian water milfoil is a hearty plant with the ability to survive a wide range of environmental conditions. EWM prefers alkaline waters, but occasionally grows in acidic waters of lakes, ponds and rivers. Specimens have been reported in waters ranging in ph from 5.4 to 11, and can tolerate salinities up to 10 or 15 parts per thousand (ppt) (Iversen, 1929; Spence, 1964; Hutchinson, 1970; Seddon, 1972; Reed, 1977; Aiken et al., 1979; Crow and Hellquist, 1983; Gerber and Les, 1996; Crow and Hellquist, 2000). The preferred depth range for EWM is 21

42 1.5 to 11.5 feet (Aiken et al., 1979), but plants grow to depths up to 30 feet. Fine-textured, inorganic sediments appear to be the most suitable for EWM, as growth on highly organic sediments is relatively poor (Barko, 1983; Barko and Smart, 1986; Smith and Barko, 1990). EWM inhabits lakes and ponds of all trophic states, from oligotrophic to eutrophic, but achieves dominance most often in mesotrophic to moderately eutrophic lakes (Madsen, 1998). EWM can survive in lakes and rivers that are too turbid for many native species (Engel, 1995), and also forms a canopy that shades out other species even if water clarity is adequate for their growth. The life history characteristics of EWM can lead to rapid distribution and infestation. Open substrates can be colonized at a high density within two years, while EWM can become the dominant species in an existing plant community in as little as five years. However, very dense, healthy, native plant assemblages have resisted colonization for more than a decade. Reproduction EWM produces seeds in the late summer, but seeds are not a significant source of reproduction for the species (Coble and Vance, 1987), and seeds may not be viable in many cases. The majority of reproduction is through the spread of vegetative fragments. In late summer, EWM becomes brittle and plant fragments are easily broken off through physical contact with boat props or currents. EWM also forms autofragments allowing for rapid invasion of new areas. Root formation often occurs at the nodes of fragments prior to abscising. The adventitious roots allow rapid establishment upon settling. Fragments drift, settle, root, survive the winter, and begin growing in the spring (Madsen and Welling, 2002). Along with plant fragments, basal shoots and roots of the EWM plant survive the winter, aided by carbohydrates formed in spring and summer (Madsen 1991; Madsen and Welling, 2002). Root crowns are the primary means of local expansion, while fragments facilitate relocation to more distant areas (Madsen and Smith, 1997). EWM does not produce turions. Species Threat Summary New shoots grow rapidly in the spring and branch repeatedly as they approach the surface of the water. Leaf canopies created by fast growing shoots can shade out the germinating seeds or vegetative propagules of understory plants. This reduces growth by other plants and potentially eliminates native plants, reducing species diversity (Smith and Barko, 1990; Madsen et al., 1991; Madsen, 1994). Monospecific stands of EWM negatively affect wildlife, and can alter the predator/prey relationship as well as the overall ecology of an aquatic ecosystem (Aiken et al., 1979; Engel, 1990; Engel, 1995; Madsen et al., 1995; Madsen and Welling, 2002). Small amounts of EWM may be consumed by aquatic waterfowl, but is not considered a valuable food source (Aiken et al., 1979). Aquatic macrophytes can provide food, shelter and spawning habitat for a wide variety of fishes (Lillie and Budd, 1992). Intermediate densities of aquatic macrophytes, even including EWM, enhance fish diversity, feeding, growth and reproduction (Dibble et al., 1996). However, dominance by EWM tends to replace native macrophytes, altering the food web and creating food shortages for many fishes (Engel, 1995). For further discussion of vegetation impacts refer to Section 4. Dense mats of EWM limit human uses of the waterbody and can lead to decreased water quality. Dense mats choke channels, clog water intakes, encourage mosquito reproduction, and 22

43 restrict aquatic activities such as fishing, swimming and boating (Orr and Resh, 1989). The mass of large mats can cause flooding in some waterbodies, and increase sedimentation by trapping sediment and detritus (Adams and Prentki, 1982). Oxygen levels can be reduced underneath large EWM mats due to a decrease in wind mixing, and decaying mats decrease oxygen and increase the internal nutrient load to the waterbody (Honnell, et al., 1992; Engel, 1995). Regulatory Status M. spicatum is listed as regulated, prohibited, invasive or noxious in at least 15 different states. In addition, EWM is on lists generated by government agencies or pest plant councils as an undesirable species in at least 21 different states. It is not accorded protected status anywhere in the USA. Myriophyllum heterophyllum Taxonomic Description Variable, or two leaf milfoil, M. heterophyllum (VWM) is a submersed perennial herb. The stem of VWM is stout, robust, and typically red in color, with white, thread-like roots which are not always present. VWM can have two distinct leaf types, emergent and submerged. Submerged leaves are delicate, 5-30 centimeters long, and usually in whorls of 4-6 around the stem, each with 6-12 segments. The emergent leaves of VWM are small, oval and typically bright green in color. Teeth are sometimes present along the edges of the leaf, and are arranged in whorls around the stem. Emergent leaves are associated with mature plants and may be present during summer months. Winter buds are formed at the base of the plants, and not along the erect stem (Aiken, 1981). The flowers of VWM are green to reddish in color, and grow on 2-12 inch spikes above the surface of the water. The floral bracts are whorled, and smaller than the foliage leaves. The fruit is four-lobed, and splits into four distinct mericarps. VWM is commonly described as having a raccoon-tail or pipe cleaner appearance due to its finely dissected, thread-like leaves. Geographic Range and History of Invasion Variable water milfoil is native to the southern portions of the United States from Florida to Texas. Today, it ranges throughout the eastern United States, westward to North Dakota, south to New Mexico, and as far north as Maine and Quebec. However, to date, VWM has not been located in Vermont. The current North American distribution is shown in Figure 3-2. VWM is highly capable of transport to new water bodies due to vegetative growth and reproduction. Plant fragments are easily transported to new waterbodies by boats, trailers, fishing gear, wind, animals and currents (Aiken et al., 1979). Plant fragments transported to new waterbodies can become rooted and begin forming new shoots. 23

44 Current North American Distribution Figure 3-2. Distribution of Myriophyllum heterophyllum in North America as of 2004 Source: Habitat Requirements/Preferences VWM is a hearty plant with the capabilities to survive a wide range of environmental conditions. VWM prefers slow-moving, still water of ponds, lakes, streams, ditches, spring-fed swamps and sloughs, growing in depths up to 3 meters. It thrives in warm waters, but can also over-winter in the colder northern temperatures. VWM grows in acidic and alkaline waters, can tolerate a wide range of calcium concentrations, and prefers silty-clay sediments with high ammonium nitrogen levels (Crow and Hellquist, 1983). It seems to be most abundant in acidic waters, however, including waters with natural color from humic substances. Reproduction VWM produces seeds in the summer, although re-growth from seeds is not common. Similar to EWM, the majority of reproduction is through the spread of vegetative fragments and 24

45 rhizome division. VWM also reproduces asexually, through the formation of buds. Pieces of plant fragments are easily broken off through physical contact with boat props or currents. These fragments can survive the winter, root, and begin growing in the spring. New plants grow stolons and begin forming new shoots, each eventually fragmenting. Natural autofragmentaton may occur twice during the growing season, typically June and September in northern states. Species Threat Summary Threats due to infestations of VWM are the same as those from EWM. However, this species does not appear to cause the same level of problem in southern states, and is not so widely distributed in northern states as to be recognized as a threat in all areas. There is mounting evidence that where VWM has become a problem, this is actually a hybrid species recognizable only through genetic studies. Regulatory Status VWM is listed as an invasive or noxious weed in at least two states. However, it is listed as endangered in two states, and as a species of special concern in another. Uncertain taxonomy may be a major factor in this variability in status. Waterbodies with VWM at high densities tend to have been invaded by this species, possibly a hybrid of M. heterophyllum and a native species such as M. pinnatum, over the last few decades. Myriophyllum aquaticum Taxonomic Description Parrotfeather, M. aquaticum, is an aquatic perennial herb. The stems of parrotfeather are long and unbranched, with roots extending from the lower nodes, and reach 2 meters or more in length. Submersed leaves are arranged in whorls of 4-6 around the stem, are cm in length, and pectinate. The emergent leaves are arranged in whorls of 4-6 and oblong in appearance. Emergent leaves are typically cm in length with linear divisions (Aiken, 1981). The emergent leaves are darker and stiffer than submerged leaves. The emergent flowering spike and leaves of parrotfeather can reach 20 cm in length, and are the most distinctive feature of the plant (Orchard, 1981). Flowers are singly located in the axils of the upper emergent leaves (Aiken, 1981). In North America, only female flowers are present, and these are white in color (Aiken, 1981). No fruiting or seed producing plants are known to occur outside of South America (Aiken, 1981; Orchard, 1981). Geographic Range and History of Invasion Parrotfeather is native to South America but has been introduced to other countries worldwide. Parrotfeather is found throughout the southern United States, and extends north along both coasts. The current North American distribution is shown in Figure 3-3. Introduction to the United States occurred in the late 1800 s for use in the aquarium and ornamental pond trades (Kane et al., 1991). The first specimens of parrotfeather were collected near Washington D.C. in Parrotfeather, like EWM and VWM, is highly capable of transport to new water 25

46 bodies due to vegetative growth and reproduction. Plant fragments are easily transported to new waterbodies by boats, trailers, fishing gear, wind, animals and currents (Aiken et al., 1979). Due to its popularity in the ornamental trade, introductions in some areas have been intentional. After transport to new areas, plant fragments transported become rooted and begin forming new shoots. Current North American Distribution Figure 3-3. Distribution of Myriophyllum aquaticum in North America as of 2004 Source: In addition to the shaded reas in Figure 3-3 indicating the presence of M. aquaticum, the United States Department of Agriculture (USDA) reports M. aquaticum in Washington and Ohio as of M. aquaticum was also found in a lake in Massachusetts in 2004 (J. Straub, MA DCR, pers. comm.) and was observed for sale at a commercial nursery in Vermont in 2002 (Wagner, pers. obs.). 26

47 Habitat Requirements/Preferences Parrotfeather is found in freshwater lakes, ponds, streams, canals, and ditches (Crow and Hellquist, 2000). It prefers slow moving and shallow water, but can also be a floating plant in deep water lakes. Plants can survive on the wet, muddy banks of streams and rivers, so it is well adapted to changes in water level. Parrotfeather is found in a range of environmental conditions, but appears to prefer nutrient-rich waters. According to the Federation of New Zealand Aquatic Societies Inc. website ( parrotfeather prefers a ph range of 6-8, temperatures between C, and a water hardness of ppm. Reproduction No fruits or seeds are produced in North American specimens of parrotfeather, and this species does not contain tubers or turions (Aiken, 1981). Parrotfeather reproduction is asexual through creeping rhizomes and plant fragmentation. Rhizomes give rise to multiple stems, and stems branch and root at the nodes, which allows for the formation of fragments. Stems are brittle and fragment easily through mechanical and physical disturbances. Fragments are spread by boats, trailers, dumping of aquarium plants, waterfowl, animals and water currents. New fragments settle, root, and produce new plants. Species Threat Summary The threats and impacts of a parrotfeather invasion are the same as those from an invasion of EWM or VWM. Resultant conditions in northern states tend not to be as severe as for EWM or VWM, but very high densities with all the expected physical, chemical and biological complications have been observed in more southern states, including the Carolinas. Regulatory Status Parrotfeather is classified as an invasive or noxious weed in at least 4 states in the United States. It is not accorded protected status anywhere in the USA. It does not appear to be as well known to management agencies, utilities, and the public in general as EWM or VWM. CLASS II MILFOIL SPECIES Myriophyllum humile Taxonomic Description Low or humble water milfoil, M, humile, is a native, perennial, submersed aquatic or amphibious plant. The stem of M. humile is shorter than that of EWM or VWM, often less than 2 m in length. The leaves are closely spaced, scattered, and finely pinnately divided with less than 14 pairs. The internodal distance is usually short, giving M. humile a bushy appearance. The leaves are strictly arranged alternately. Flowers and fruits are produced in the axils of submerged leaves (Aiken, 1981). The fruit is a smooth or slightly rough, 4-parted nutlet, lacking a dorsal ridge (Aiken, 1981). No winter buds are produced by M. humile, but root crowns and some stems can survive the winter. Identification of M. humile can be extremely difficult due to 27

48 three ecological variations of the plant, depending on environmental conditions (Aiken, 1981; Crow and Hellquist, 2000). Geographic Range Low water milfoil is native to North America and is not considered invasive, although it has been found in new areas in recent years. It is primarily a northeastern species, but has been found in the upper Midwest as well. The current North American distribution is shown in Figure 3-4. Current North American Distribution Figure 3-4. Distribution of Myriophyllum humile in North America as of 2004 Source: In addition to the shaded areas on this map, the USDA reports the presence of M. humile in Ohio and Illinois as of

49 Habitat Requirements/Preferences M.humile is found in shallow acidic waters of lakes, ponds and streams (Crow and Hellquist, 2000). M. humile can be locally abundant, and is found in waters with a moderate Dissolved Organic Carbon (2-5 mg/l). It prefers shallow water and fine sand/silt sediments with high organic content. M. humile tends to support less periphyton growth compared to other milfoils. Species Threat Summary Growths can get quite dense in shallow waters, potentially altering water quality and affecting water uses in localized areas. Lowered summer water levels represent the greatest potential for this species to achieve nuisance levels, as it normally would not grow to the surface on its own in most reservoirs. Regulatory Status Low milfoil is protected in Vermont and Virginia, and is considered vulnerable or imperiled to some degree in several Canadian provinces, but is not under any regulatory status elsewhere. Myriophyllum sibiricum Taxonomic Description Northern water milfoil, M. sibiricum, is a native, submersed, perennial aquatic herb. The majority of the submersed stem appears erect and whitish in appearance, rarely branches near the surface, and may reach 3 m in length. The stem of M. sibiricum is not thickened below the flower spike. Roots are white and thread-like but not always present. M. sibiricum has both submersed and emergent leaves. Submersed leaves are 1-4 cm in length, arranged in whorls of 3-4, pinnately dissected, and have 5-14 filiform segments on each side of the midrib (Aiken, 1981). Leaves are stiff, projecting at right angles when removed from the water. Emergent leaves are located above the water on reddish flowering spikes, 2-15 cm in length (Crow and Hellquist, 2000). Emergent leaves are tiny, 1-3 mm in length, smooth toothed to coarse, and located beneath the flowers. The tiny flowers of M. sibiricum are reddish-purple and imperfect. The male flowers have 4 petals, while female flowers lack petals altogether. M. sibiricum produces a nut-like, olive colored fruit with four chambers, each producing one seed. During the later portion of the growing season, cylindrical turions consisting of smaller, darker green leaves are often present at the lower nodes (Aiken and Walz, 1979). The majority of M. sibiricum reproduction is vegetative from turions or fragmentation, although seeds are produced. Geographic Range Northern water milfoil is native to North America and is primarily a far northern species. It has been losing range since the introduction of M. spicatum. It also hybridizes with M. spicatum, creating taxonomic difficulty for correct identification, especially in the upper 29

50 Midwest. The current North American distribution is shown in Figure 3-5. The hybrid has been characterized as more invasive and persistent than either M. spicatum or M. sibiricum, but confirmatory scientific studies are lacking (D. Pullman, personal communication). Current North American Distribution Figure 3-5. Distribution of Myriophyllum sibiricum in North America as of 2004 Source: Habitat Requirements/Preferences Myriophyllum sibiricum is often found in shallow or deep water of lakes, ponds, ditches, marshes and slow moving streams. It can tolerate a wide range of ph and calcium levels, and prefers sandy-muck sediments with high ammonium-nitrogen levels (Muenscher, 1944; Swindale and Curtis, 1957; Correll and Correll, 1972; Aiken et al., 1979; Crow and Hellquist, 1983; Voss, 1985; Gerber and Les, 1996; Crow and Hellquist, 2000). 30

51 Species Threat Summary Northern water milfoil is not considered a threat by itself, but hybridization with M. spicatum may represent a very hardy and aggressive invader. Regulatory Status M. sibiricum is listed as threatened in Ohio and endangered in New Jersey and Pennsylvania. It is generally on the decline in most states, but is not accorded protected status in most areas. Myriophyllum hippuroides Taxonomic Description Western milfoil, M. hippuroides, is a submersed native perennial herb. The stem is yellowish green, and highly branched, reaching 1m or more in length. Western milfoil has two types of leaves, submersed and emergent. The submersed leaves are feather-like in appearance and 2-5 cm in length. They are arranged around the stem in whorls, each whorl containing 4-5 leaves. Each leaf contains 6-30 pairs of thread-like leaflets. The emergent leaves are not feather-like, and are located on a stout flower spike that protrudes from the surface of the water (Aiken, 1981). Each leaf is 5-10 mm in length, 1-2 mm wide, and has smooth to toothed edges. The flowers of M. hippuroides are tiny, with 4 petals per flower and located on a flowering spike that ranges in size from 3-12 cm (Aiken, 1981). Propagation of western milfoil is accomplished through fragmentation, rhizomes and seeds. Geographic Range Western water milfoil is native to North America and is primarily a far northern species. It has been losing range since the introduction of EWM. The current North American distribution is shown in Figure 3-6. Habitat Requirements/Preferences M. hippuroides inhabits lakes, ponds, ditches and small streams on the west coast of the United States. Species Threat Summary Western water milfoil is not considered a threat in most reservoirs, but has been known to clog irrigation ditches. Regulatory Status M. hippuroides is not accorded regulatory status in any state or province. 31

52 Current North American Distribution Figure 3-6. Distribution of Myriophyllum hippuroides in North America as of 2004 Source: CLASS III MILFOIL SPECIES Myriophyllum verticillatum Taxonomic Description The comb water milfoil, M. verticillatum, is a perennial aquatic herb. The stem of the comb water milfoil is stout, can be branched or unbranched, and reaches 3 m in length. White, thread-like roots are not always present. Submerged leaves are arranged in whorls of 4-5, are 1-5 cm in length and have 9-13 filiform segments on each side of the midrib (Aiken, 1981). Emergent leaves are deeply divided and arranged in whorls around a 4-12 cm flowering spike. The emergent leaves can be 2-10 mm in length, and are typically longer than the flowers. The flowers are sometimes perfect, but typically imperfect, with males on the top of the spike and female flowers below. M. verticillatum produces a small brown fruit, 2-3 mm in length, and four chambered, with each chamber producing one seed. The comb water milfoil produces yellow- 32

53 green turions for overwintering that are present from fall to early spring. The turions of comb water milfoil are wider at the tips than at the base, sometimes referred to as club-shaped. Current North American Distribution Figure 3-7. Distribution of Myriophyllum verticillatum in North America as of 2004 Source: The current North American distribution is shown in Figure 3-7. In addition to the shaded areas on the map, the USDA reports M. verticillatum in Nevada and New Mexico. Habitat Requirements/Preferences Comb water milfoil is found in shallow, still waters of lakes and ponds ranging from slightly acidic to slightly alkaline, with a wide range of calcium concentrations. Preferred sediments are silty-mucks with high ammonium-nitrogen levels (Almestrand and Lundh, 1951; Hutchinson, 1970; Hutchinson, 1975; Aiken et al., 1979). 33

54 Regulatory Status M. verticillatum is listed as threatened in Indiana, and endangered in Massachusetts, New Jersey, Ohio and Pennsylvania. It is not considered invasive anywhere. Myriophyllum pinnatum Taxonomic Description M. pinnatum is an aquatic or amphibious perennial plant. It can be found submersed with elongate stems, or on shore, with branching stems rooted in the muddy substrate. Leaves are 1-3 cm in length and arranged in whorls of 3-5 or sometimes scattered on the stem (Aiken, 1981). The flowers are perfect or imperfect in the axils of emersed leaves with long purplish petals above. Each flower typically contains 4 stamens, but on rare occasions may contain 6 stamens. Emersed leaves are cm long, and pinnatifid to shallowly toothed. The fruit of M. pinnatum is pale in color and cubic-ovoid in shape. Current North American Distribution Figure 3-8. Distribution of Myriophyllum pinnatum in North America as of 2004 Source: 34

55 Habitat Requirements/Preferences M. pinnatum is found in shallow water or mud of coastal salt pond communities, marshes, ponds, lakes, streams, swamps, and ditches (Crow and Hellquist, 2000). Quite often, M. pinnatum is associated with Phragmites, growing on the exposed shores of salt ponds and pondholes. The current North American distribution is shown in Figure 3-8. Regulatory Status M, pinnatum is listed as a species of special concern in Connecticut and Massachusetts. It is not considered invasive anywhere. Myriophyllum laxum Taxonomic Description Loose water milfoil, Myriophyllum laxum, is a native perennial aquatic herb. The stem of loose water milfoil is elongate, slender, and has conspicuous black hydathodes (Aiken, 1981). Submersed leaves are cm in length and usually arranged in whorls of 4 or 5, but are sometimes conspicuously alternate. Leaf divisions are few, ranging from 8-16, and as a result, M. laxum does not have a feathery appearance (Aiken, 1981; Godfrey and Wooten, 1981). The flowering spike is slender, and appears naked and interrupted (Godfrey and Wooten, 1981). Flowers are minute, and often arranged alternately. At the base of the inflorescence, bracts are longer than the flower, while the upper bracts are shorter than the flowers (Aiken, 1981). Pinkish petals are present on bisexual and staminate flowers (Godfrey and Wooten, 1981). M. laxum produces a fruit approximately 1 mm in length, with a smooth to warty surface. 35

56 Current North American Distribution Figure 3-9. Distribution of Myriophyllum laxum in North America as of 2004 Source: Habitat Requirements/Preferences The current North American distribution is shown in Figure 3-9. M. laxum is found in lakes, swamps, streams, ditches and sinkholes. Regulatory Status In Georgia and North Carolina, M. laxum is listed as a threatened species. It is not considered invasive anywhere. Myriophyllum farwellii Taxonomic Description M. farwellii is a native perennial aquatic herb that grows entirely under the surface of the water. The stem of M. farwellii is submerged and sparsely branched. At the lower nodes, the stem roots freely. The roots are thread-like and white in color. Leaves are long, thread-like, and 0.25 to 1.25 inches in length. Most leaves of M. farwellii are alternate or irregularly scattered 36

57 along the stem (Aiken. 1981). M. farwellii produces single flowers in the axils of leaves. The flowers of M. farwellii do not appear above the surface of the water. The female flower has four purple petals and inconspicuous sepals. The fruit is nut-like with four lobes, each producing a seed. A longitudinal ridge is present on each segment of the fruit. A positive identification of M. farwellii requires presence of fruit or winter buds (Crow and Hellquist. 2000). The current North American distribution is shown in Figure Current North American Distribution Figure Distribution of Myriophyllum farwellii in North America as of 2004 Source: Habitat Requirements/Preferences M. farwellii inhabits ponds and small lakes from Minnesota to Maine. In the northeast it is found in low alkalinity, slightly acidic waters and grows in low nutrient sediments consisting of coarse sand and gravel. Other studies have shown that M. farwelli is found in soft acidic waters, ranging in color from clear to tannic, and grows in sandy sediments with high organic content and low ammonium-nitrogen levels (Muenscher, 1944; Almestrand and Lundh, 1951; Crow and Hellquist, 1983; Gerber and Les, 1996; Crow and Hellquist, 2000). 37

58 Regulatory Status M. farwellii is given threatened status in Michigan, New Hampshire, and New York. In Massachusetts and Pennsylvania, M. farwellii is listed as endangered. It is not considered invasive anywhere. Myriophyllum quitense Taxonomic Description Andean milfoil, Myriophyllum quitense, is a native perennial aquatic herb. The stem of Andean milfoil is cylindrical and 1-4 m in length. Roots are white and very branched. Andean milfoil has two types of leaves, submersed and emergent. The submersed leaves are feather-like in appearance, cm in length, and 2 cm wide (Aiken, 1981). They are arranged around the stem in whorls, each whorl containing 2-4 leaves, but occasionally 5. Each leaf contains 5-10 pairs of thread-like leaflets. The emergent leaves are blue-green to reddish in color, not featherlike, and are arranged in whorls on a flower spike. Each leaf is cm in length with teeth along the edges halfway to the midrib, where it becomes less toothed. Unlike other milfoil species, Andean milfoil can produce multiple emergent flower spikes (Aiken, 1981). The flowers of Andean milfoil are tiny, only mm in length, and each flower has 4 petals located at the base of the emergent leaves. Compared to other milfoil species, Andean milfoil flowers late in the season, often in August and September. Olive brown fruits are square, 1.7 mm in length, and rarely found. Propagation of M. quitense is accomplished primarily through seeds and plant fragmentation, although they also produce winter buds. 38

59 Current North American Distribution Figure Distribution of Myriophyllum quitense in North America as of 2004 Source: Habitat Requirements/Preferences M. quitense is primarily found in oligotrophic lakes and flowing rivers, but is also found in hot-spring fed rivers. The current North American distribution is shown in Figure In large lakes it is typically growing near the shoreline, in areas with wind driven wave action (Ceska et al., 1986; Couch and Nelson, 1988). It can grow in pure stands or with other species of aquatic macrophytes, but rarely takes a terrestrial form (Ceska et al., 1986). In South America, M. quitense is found in lakes and streams ranging in altitude from meters (Orchard, 1981). Regulatory Status Unknown and generally unregulated, but may be protected in Wyoming. M. quitense is not considered invasive anywhere. 39

60 Myriophyllum ussuriense Taxonomic Distribution M. ussuriense is a native, aquatic perennial herb. This species is semiterrestrial, and grows in shallow water or on exposed mud. Stems are light green in color, and plants reach a height of 25 cm. Leaves are 1-8 mm in length and arranged oppositely or in whorls of three. Each leaf has 1-3 pairs of segments. Submerged leaves are in whorls of three, and have 2-4 segments on each side arranged alternately (Ceska et al., 1986). Within the leaf axils, M. ussuriense forms slender filiform winter buds. M. ussuriense is a dioecious plant, occurring in both female and male populations (Ceska et al., 1986). Flowers are located within the axils of floral bracts, and flowers are typically entire. Floral bracts differ between submerged and terrestrial plants. The floral bracts of terrestrial plants are similar to the leaves, while submerged plants experience a gradual transition between leaves and entire bracts (Ceska et al,. 1986). The current North American distribution is shown in Figure Current North American Distribution Figure Distribution of Myriophyllum ussuriense in North America as of 2004 Source: 40

61 Habitat Requirements/Preferences M. ussuriense grows on sandy or muddy substrates of lakes and rivers, and prefers areas with fluctuating water levels (Ceska et al., 1986). M. ussuriense prefers temperatures ranging from C and a ph of 6-8. Regulatory Status M. ussuriense is not accorded any regulatory status, and is not considered invasive anywhere. Myriophyllum tenellum Taxonomic Description Slender milfoil, M. tenellum, is a native perennial aquatic herb. The stem of M. tenellum is slender and largely unbranched, reaching 30 cm in length. Roots are white, thread-like, unbranched, and not always present. Unlike other milfoils, leaves are absent from M. tenellum, making it easy to distinguish from other milfoils but difficult to identify as a milfoil in the first place. Flowers are present on a flower spike above the water s surface. The floral bracts and flowers are similar in size, with the bracts mostly alternate. The sterile submersed form of M. tenellum is the most common variety, so flowers are rare (Crow and Hellquist, 2000). M. tenellum produces a 4-lobed, nut-like fruit with one seed produced per lobe. Habitat Requirements/Preferences Slender milfoil prefers acidic to slightly alkaline waters, and sandy sediments with low ammonium-nitrogen levels (Muenscher, 1944; Swindale and Curtis, 1957; Voss, 1965; Aiken, 1981; Voss, 1985; Gerber and Les, 1996). The current North American distribution is shown in Figure It tends to grow peripherally in very shallow water, along with plants not typically associated with other milfoils, such as quillwort, pipewort and lobelia. 41

62 Current North American Distribution Figure Distribution of Myriophyllum tenellum in North America as of 2004 Source: Regulatory Status In Connecticut, M, tenellum is a species of special concern. Pennsylvania has listed the plant as threatened, and in Maryland and New Jersey it is listed as endangered. It is not considered invasive anywhere. Myriophyllum alterniflorum Taxonomic Description Alternate flowered milfoil or little milfoil, M. alterniflorum, is a submersed, native, perennial aquatic herb with a very slender stem, and white, unbranched roots. The leaves of M. alterniflorum are submersed, and no emergent leaves are present on the flower spike. Submersed leaves are finely dissected, thread-like, and arranged in whorls of 3-5 around the stem. One distinguishing characteristic of M. alterniflorum is the short leaves, usually less than 3/8 in length, and shorter than the stem internodes (Aiken, 1981). The male flowers of M. alterniflorum are arranged alternately near the top of a leafless flowering spike protruding from the surface of the water. Female flowers are arranged in whorls of 2-4 at the bottom of the flowering spike, with hermaphrodite flowers above. M. alterniflorum produces a 4-lobed nut- 42

63 like fruit, with one with seed produced per lobe. The current North American distribution is shown in Figure Habitat Requirements/Preferences Alternate flowered milfoil prefers neutral to slightly acidic lakes with low alkalinity, and can tolerate a wide range of calcium concentration. It is often found in glacial lakes and kettleholes, and prefers sandy sediments with low ammonium-nitrogen levels (Iversen, 1929 as stated in Gerber and Les 1996; Swindale and Curtis 1957; Spence, 1964 as stated in Gerber and Les, 1996; Seddon, 1972; Hutchinson, 1975; Aiken, 1981; Crow and Hellquist 1983; Voss, 1985; Gerber and Les, 1996). Current North American Distribution Figure Distribution of Myriophyllum alterniflorum in North America as of 2004 Source: Species status M. alterniflorum is listed as threatened in Massachusetts and New York, and endangered in Connecticut. It is not considered invasive anywhere. 43

64 HYBRIDIZATION OF MYRIOPHYLLUM SPECIES Researchers have suspected the hybridization of species within the genus Myriophyllum for years (Patten, 1954). Hybridization between M. spicatum and M. sibiricum has been documented in the upper Midwest (Moody and Les, 2002a) and a hybrid cross of M. heterophyllum and M. pinnatum has been found in New England (Moody and Les, 2002b). A number of papers about hybrid milfoils are in review at the time of this writing, and advances in genetic techniques are enabling greater detection of such hybrids. M. spicatum hybrids may be more resistant to control with herbicides. M. heterophyllum hybrids form large, aggressive, monoculture populations of management concern, while at least some non-hybrid M. heterophyllum grows among native plants and does not exhibit invasive characteristics with such severity. What are recognized as different strains of some species may actually be hybrids, but thorough studies have not been completed. Hybridization poses identification and management problems. Hybrids are difficult or impossible to identify without the use of DNA, and may be more aggressive and invasive compared to the non-hybrid parental plants. As some milfoil species are native, and some native forms are protected, knowing which species is which is important in milfoil management. Thum (2005) has developed a relatively simple genetic technique for identifying M. heterophyllum hybrids in New Hampshire, but practical application of such methods for identification of hybrids is still in its infancy. 44

65 CHAPTER 4 POTENTIAL ISSUES FOR DRINKING WATER RESERVOIRS The presence of a large or dominant population of water milfoil in the aquatic macrophyte community of a drinking water reservoir can lead to variety of potential impacts, including water quality concerns, elevated biomass, ecosystem alterations, and recreational impacts. While these impacts could exist to a lesser extent for reservoirs with dense coverage by any macrophytes, regardless of the species, they are often more pronounced at the high density and heavy biomass accumulation characteristic of the nuisance invasive species of water milfoil. One of the primary concerns regarding invasive water milfoil on drinking water reservoirs is the potential for water quality impacts. Potential impacts can affect water quality standard attainment as well as reduce the potability or treatability of the water. High levels of nuisance aquatic vegetation can lead to direct impairment of water quality and uses, including violations of numeric water quality criteria (e.g., dissolved oxygen or ph levels) or impairment of designated water quality uses (e.g., impairment of Aquatic Life support function). Impacts may be indirect, as with the production of disinfection by-products precursors (DBPPs) that become contaminants of concern during water treatment processes. Unaesthetic taste and odor may also result, directly or indirectly. Interaction of water milfoil with the other ecological components of the reservoir (notably phytoplankton, periphyton, zooplankton, and fish communities) can indirectly affect water quality, usually through shifts in the algal community. The physical biomass and seasonal breakdown of macrophyte beds can also clog intake structures or impede outlet water flow. The large biomass can negatively affect any allowed recreational uses of a reservoir, such as boating or swimming. WATER QUALITY ISSUES The drinking water quality issues associated with nuisance levels of milfoil include (1) the production of DBPPs; (2) development of taste and odor; (3) impacts on general water quality and physical variables such as temperature, dissolved oxygen (DO) and ph that affect treatment needs; and (4) violations of other water quality standards. These are discussed below. Disinfection By-Product Precursors Some forms of dissolved organic carbon (DOC) and particulate organic carbon (POC) are important in the formation of so-called disinfection by-products (DBPs). The DBPs are generated in drinking water plants during the disinfection process by the reaction of the carbonaceous compounds with the disinfectant (typically halogenated products and most often chlorine). The produced DBPs pose potential human heath risks in drinking water. The three classes of DBPs regulated by the U.S. EPA include trihalomethanes (THMs), haloacetic acids (HAA), and bromate ions (U.S. EPA, 1998), and levels of these DBPs in the finished water are regularly monitored by drinking water facilities. However, it should be noted that the precursors represented by DOC and POC are not inherently harmful until they react with chlorine or other disinfectants inside the water treatment plant. Lowering the concentrations of DBPPs in the source water therefore simply minimizes risk. 45

66 DBPPs include natural organic carbon derived from plant detritus, algae, peaty sediments, terrestrial vegetation and other watershed inputs (Brown, 2003). Work by Hoehn and co-workers (Hoehn et al., 1980) indicated that the presence of algal blooms provided a significant source of DBPPs in a Virginia reservoir. Nuisance aquatic macrophytes also have the potential for being significant seasonal contributors to the DOC and POC fractions in the organic carbon budget of a waterbody. In macrophyte-dominated waterbodies, these primary producers may be responsible for a considerable proportion of the in-lake production during the growing season. For example, the annual organic budget for Lake Wingra, (WI) indicated that macrophytes were 17.4% of the annual production in the lake, and they represented 67% of carbon in Lawrence Lake (MI) (reported in Wetzel, 2001). The inputs of macrophyte organic carbon pose a variety of challenges for a drinking water reservoir, including entrainment of DOC or fine POC into water intakes with potential DBP generation, consumption of oxygen due to bacterial breakdown of organic material resulting in localized anaerobic conditions and acid build-up that affects ph, or accumulation of decaying material on shorelines, leading to unaesthetic conditions. As the nuisance milfoils (Class I milfoils from Chapter 3) can produce very high biomass, they represent a distinct threat in this regard. Many submerged macrophyte species have been shown to release inorganic nutrients and DOC during the active growth phase (Barko and Smart, 1980; 1981; Smith and Adams, 1986), with milfoils demonstrated to be among the leakiest forms. During the growth phase, much of these released nutrients and DOC are rapidly and effectively assimilated by attached algae (i.e., epiphytes) and bacteria living on the surface of the plants such that little enters the water column. However, those algae and bacteria can themselves release objectionable compounds. During the seasonal decline and breakdown (i.e., senescence) of the macrophytes, decaying tissue material can release large amounts of inorganic nutrients and DOC to the waterbody (Carpenter, 1980; Smith and Adams, 1986). The release rates are generally determined by the macrophyte population growth dynamics, the rapidity of decline, and lake conditions at the time or death and resulting decomposition (Wetzel, 2001). The initial stages of decomposition are marked by a leaching of soluble organic material (transitory-labile) from tissue material followed by a more gradual microbial breakdown of the structural carbon. Decomposition and release of DOC and nutrients is considered to occur more rapidly from submerged and floating-leaved macrophytes than from emergent types or terrestrial vegetation. For example, M. heterophyllum was shown to lose 30-40% of its dry weight within 1-2 weeks of senescence, with an 88% loss within 12 weeks (Wetzel, 2001). This rapid loss is probably partially due to the lack of lignin (a highly refractory carbon compound) in submergent aquatic plants. Accordingly, aquatic plants tend to have an initial high release of DBPPs, but further microbial transformations produce decay products with less DBP-formation potential (Lee and Jones, 1991). Eventually, the fine organic detritus particles of decaying macrophytes will be transferred from the shallow peripheral (littoral) zone to deeper open water (pelagic) areas through wind and water movement. The movement of organic carbon is usually accompanied by nutrients, especially phosphorus, which can represent a significant source of nutrients for phytoplankton (Moore et al., 1984), therefore representing an indirect influence on water quality. 46

67 Taste and Odor Taste and odor are often seasonal concerns for drinking water reservoir managers due to the presence of so-called nuisance algal growths (Spencer, 2003). It has been shown that certain algal species produce and release compounds (e.g., 2-methylisoborneol, geosmin) which cause unpleasant taste and odor that may require additional treatment at the drinking water plant to make the water acceptable to the public. While some macrophytes (e.g., Chara) have a distinctive musky odor, there is little evidence to indicate that normal populations of macrophytes in general and milfoils in particular are a direct cause of taste and odor problems, other than possibly strong vegetation and decay odors associated with mass senescence of those plants. However, it is probable that nuisance milfoils indirectly lead to taste and odor problems, due to their localized modification of available light, temperature and water chemistry (see below). These modified environmental conditions can lead to encouragement of problematic species of algae known to produce taste and odor compounds. Establishment of attached (epiphytic) growths or bottom (benthic) mats of algae have been observed by the authors in multiple New England reservoirs in association with dense milfoil growths, with similar observations reported for the Pacific Northwest (Gibbons, personal communication). These nonplanktonic growths can impart taste and odor to the water without actually showing up in intake samples screened for problem algae. Dissolved Oxygen, ph, Light, and Temperature Aquatic macrophytes in general and nuisance milfoils in particular have the potential to impact water quality and uses within the littoral zone, with some effects extending throughout the reservoir (Honnell et al., 1992). For example, photosynthetic and respiratory activities of dense vegetation can lead to wide diel (i.e., daily) fluctuations of DO measured within a macrophyte bed. This is due to the high levels of photosynthesis achieved by growing macrophytes, leading to elevated (i.e., supersaturated) DO conditions during the daylight with an associated steep decrease in DO during the night or cloudy days due to high levels of aquatic biological community respiration. Madsen (1997) recorded hourly DO readings over 48 days in replicate ponds planted with invasive aquatic macrophyte species, including M. spicatum. Daily minimum values were lower (below 5 mg/l, a level at which some fish experience oxygen stress and a common state standard) for ponds with invasives than for those with a mixture of native aquatic plants. Daily mean values were also decreased for ponds dominated by M. spicatum (7.9 mg/l) compared to native plants (13.8 mg/l). Values <2 mg/l have been observed in some M. spicatum beds (Wagner, pers. Obs.) If low oxygen conditions persist sufficiently long and no refuges (e.g., other accessible areas of the reservoir or tributary streams) are available, localized fishkills can result. If DO oscillations are persistent, both the magnitude of the diel swings and the absolute DO minima can have adverse effects on other biota and water chemistry as well. Reduced compounds, such as sulfides and ammonium, may accumulate, and are themselves toxic to many aquatic biota. Even if the oscillations are not severe enough to induce toxic conditions, the accompanying changes in water quality can create a need for treatment adjustments in potable water supply operations over the course of a day. By managing aquatic plants at a moderate abundance level, 47

68 the probability of having oxygen depletions and influential water chemistry oscillations due to aquatic plants is sharply decreased (Hoyer and Canfield, 1997). As noted above, the seasonal pulse of organic material due to macrophyte population decline and decay can also lead to decreased DO levels as well. Since macrophyte decline and decay may occur over a several month period, particularly with the asynchronous growth and decay pattern inherent to mixed species assemblages, it is likely that levels of decomposition would only moderately impact the metabolism and water column DO of most reservoirs. This may not be true, however, for waterbodies dominated by a single invasive species, where the decline in water column DO may be particularly pronounced following a widespread crash of the dominant macrophyte species. The probability of a resultant major water chemistry change would partly depend on the size, shape, and depth characteristics of the reservoir, but the potential for water quality changes that affect treatment needs and treatability exists. Another environmental factor which is related to high levels of primary production is the effect that high rates of photosynthesis can have on dissolved carbon dioxide [CO 2 ] concentrations and, secondarily, ph. Diffusion of CO 2 in water is relatively slow compared to that in air. High rates of photosynthesis can rapidly deplete water column dissolved inorganic carbon pools, lowering dissolved [CO 2 ] and increasing the ph (Wetzel, 2001). Since ph is a master variable for most chemical reactions, this change is likely to have secondary implications for local water chemistry and the treatability of the water. This ph elevation is a transient condition, however. While increased ph may be observed at mid-day during peak photosynthetic activities, it will not persist as night shuts off photosynthesis and the water column [CO 2 ] is replenished by biotic respiration and diffusion from the atmosphere. This may itself be a problem for water supplies run on a 24 hour basis, necessitating treatment adjustment over the course of a day. Submerged plant canopies can also lead to significant alterations in the local ambient light environment in the littoral zone, compressing the depth of the photic zone and changing spectral quality (Barko and Smart, 1981). Myriophyllum species can grow to form dense canopies of intertwined branches that reduce light penetration, resulting in reduced photosynthesis and growth by adjacent macrophyte species. Madsen followed invasion of M. spicatum in a series of permanent plots located in Lake George (NY) that led to an increase from 25% to 97% canopy cover within 3 years, while the number of submerged macrophyte species present in the plots declined from 21 to 9, a greater than 50% reduction (Madsen, 1991). It should be noted, however, that the alteration in light may lead to the establishment of selected plants (algae and macrophytes) which are able to tolerate the low light environment. Associated with the macrophyte canopy modification of the depth of light penetration is a potential modification in temperature (Barko and Smart, 1981). As light energy entering the water column is either absorbed or reflected by the dense stands of macrophyte, it would be predicted that water temperature will be decreased beneath the shade. Since water movements in the littoral would also likely be damped by the large amount of suspended biomass in the water column, thereby reducing exchange with open water (pelagic), a small microclimate can be generated. The resulting conditions of low light and reduced temperature may be attractive to some fish species. They may also foster variable water chemistry, which with advective transport of water out of the littoral zone can affect water quality at intakes and induce variability in source water that changes treatment needs with limited predictability. 48

69 In summary, dense growths of macrophytes tend to induce undesirable and variable water quality conditions, such that treatment needs may change on hourly, daily or seasonal scales. Invasive species, including the Class I milfoils, tend to produce some of the highest density growths observed in northern reservoirs, thus increasing the risk of water quality impacts and adverse impacts to the drinking water supply. Other Aesthetic Standards Many states water quality standards include narrative criteria for unaesthetic conditions, with language prohibiting actions resulting in unnatural surface scums or bottom deposits, etc. While macrophytes may be considered natural in origin, waterbodies with dense macrophyte concentrations will often have unappealing aesthetic conditions due to their presence. Dense populations of macrophytes present an unpleasant physical appearance and can alter the apparent color of the waterbody. Physical disturbance, either natural (wind and waves) or anthropogenic (motorboat propellers and wakes) can lead to large amounts of floating, detached plants which are likely to accumulate in downwind shoreline areas. Accumulated (and decaying) vegetation is unsightly, can create odor problems and provides breeding location for unwanted insects (Hoyer and Canfield, 1997). These conditions may lead to both recreational impacts (treated elsewhere in this study) and economic detriments to shoreline residents and local real estate. BIOMASS ISSUES The biomass of milfoil species is highly variable by species and environmental conditions, especially light and sediment features. Some species of milfoil are relatively innocuous (e.g., M. tenellum, M. alterniflorum), generating only minor amounts of biomass in most aquatic systems. However, nuisance forms such as M. spicatum, M. heterophyllum and M. aquaticum have generated biomasses of between 0.5 and 1.0 kg dry weight per square meter (Madsen et al., 1989; Johnson et al., 2000). While native plant assemblages can get dense as well, the nuisance milfoils attain a higher density than native assemblages on average, and the biomass turnover rate is higher for milfoil than most other species (Madsen, 1991), resulting in greater input of organic matter to the aquatic system with milfoil presence in a reservoir. To investigate this issue further, data from numerous published (Davis and Carey, 1981; Getsinger et al., 1997; Johnson et al., 2000; Lillie et al., 2003; Newman and Biesboer, 2000; Parsons et al., 2001; Wilson, 1937; Wetzel, 2001) and unpublished (Eichler, ENSR, Horsburgh, Johnson, Madsen, Newman and Rusak) were combined into a database for the three major nuisance (Category I) milfoil species and compared over a gradient represented by classes of no milfoil, some milfoil, and dominant milfoil growths for biomass per unit area. Considering Eurasian watermilfoil, perhaps the best studied of these species, the average biomass for northern lakes without it is not appreciably different than for those lakes having some of this species, or even for those lakes dominated by Eurasian watermilfoil (Table 4-1). The comparison is more striking for variable watermilfoil and parrotfeather, based on less data from Florida. 49

70 Table 4-1. Comparison of biomass (as grams dry weight per square meter) for plant assemblages with and without milfoil. Florida Lakes - Milfoil Status Northern Lakes - M. spicatum M. heterophyllum or M. aquaticum Average Average g DW/m2 n g DW/m2 n No Milfoil Some Milfoil Milfoil Dominant Examination of the distribution of Eurasian watermilfoil biomass over 100 g/m 2 increments (Figure 4-1) shows that while native assemblages can reach higher biomasses, they have a distribution of biomass skewed toward low values. Assemblages with milfoil present or dominant have fewer low biomass values and more intermediate biomass values, consistent with greater density for this one species than many native plants. However, the creation of canopies and shading out of other species limits vegetative layers and actually appears to limit the maximum biomass that can be achieved, compared to at least some native assemblages. The change in distributions is statistically significant by the Kruskal-Wallis test (p <0.01); milfoil presence signals a move away from a low biomass assemblage, if one exists in the reservoir. The variable watermilfoil and parrotfeather data from Florida suggests a similarly significant shift in the biomass distribution, but with achievement of more high biomasses and a resultant increase in the average biomass (Table 4-1). Since nuisance milfoil tends to exhibit productivity to biomass (P/B) ratios that are at least 50% higher than most native species and as much as 500% higher than some native species (Madsen, 1991), the impact can be substantial even without an increase in average biomass. Milfoil produces more organic matter that is released from plants over the growing season, including leakage of compounds and loss of lower leaves and branches as the plant grows longer and/or canopies form. Even the same biomass with a 100% higher P/B ratio for a milfoil dominated assemblage could result in a doubling of the organic material generated during the growing season over what a native assemblage might be expected to produce. Intake Clogging with Plant Material Where free-floating plant material becomes abundant, the potential for clogging of screened intakes is high. Problems have been frequently reported for irrigation systems and recreational lake outlets in summer and fall, but documentation of offending species and overall impacts is scant. Clogging has been a problem for some drinking water supplies, particularly those with shallow intakes in weedy reservoirs (ENSR, 2003), but there appear to be few published reports of such clogging issues. Recent emphasis on enforcing Section 316b of the federal Clean Water Act (regarding intake features) and the presence of thriving businesses in screening design and cleaning devices (e.g., attests to the reality of this issue, but its extent in potable water reservoirs is unknown. The questionnaire survey conducted as part of this AwwaRF project is intended to advance our knowledge of the extent of clogging problems for water supplies. 50

71 Frequency Distribution of Plant Biomass in Northern Lakes without Myriophyllum spicatum # of Occurrences More Biomass (DW g m -2 ) Frequency Distribution of Plant Biomass in Northern Lakes with some Myriophyllum spicatum More # of Occurrences Biomass (DW g m -2 ) Frequency Distribution of Plant Biomass in Northern Lakes with Myriophyllum spicatum dominant # of Occurrences More Biomass (DW g m -2 ) Figure 4-1 Distribution of biomasses of plant assemblages without invasive milfoil (upper panel), with some milfoil (middle panel) and dominated by milfoil (lower panel). 51

72 Associated Periphyton Clogging Beyond clogging by macrophyte biomass itself, clogging with periphytic growths associated with rooted plants is known to have occurred at intakes with finer screens (e.g., ENSR, 2000), but clear documentation is scant. Even where currents are too weak to move heavier macrophyte masses, periphytic growths may slough off from late summer into winter and accumulate on intake screens. Filamentous green algae and diatoms are the most common problem groups, along with the filamentous blue-green Lyngbya, and all are associated with rooted plant growths in general and invasive milfoil species in particular (Elser, 1966; Wagner, personal observation). Note that attached Lyngbya can produce taste and odor compounds sufficient to affect supply water; it may show up on screens while not detected in the water column by routine sampling. Periphytic growths appear to be maximal at moderate trophic status (i.e., moderate nutrient loading); periphyton make use of nutrients leaked from macrophytes and respond strongly to light availability (Vymazal, 1994). While most macrophytes will collect on intake screens, some periphyton will break up in response to pressure increases on clogging screens, creating the potential for filter medium clogging in the treatment process and potentially releasing dissolved substances into the raw water. Sediment/Particulates Rooted plants reduce water velocity and trap sediment, and dense milfoil growths are among the more effective sediment trapping plants (Adams and Prentki, 1982). This might be a valued function near inlets for water supplies, but not near intakes. Senescence of milfoils may add considerable particulate material to the water column during fall in northern climates, but little is known of the individual effect of this decay on raw water supply quality. Impacts could include increased turbidity, increased organic content, increased oxygen demand, and fluctuating ph and oxygen levels. Recreational Impacts Swimming and boating are impeded by dense plant growths in general, and invasive milfoils can achieve high densities. A discussion of impacts and a number of case histories can be found in Mattson et al. (2004), and the effects of milfoil invasion are well known to lake and reservoir managers across the northern USA. Economic impact of milfoil infestations has received more recent attention (Mongin, 2005), with related control efforts possessing a favorable economic balance on recreational and property value bases alone. Where a water supply is also a recreational resource, control of invasive milfoil species may be justifiable on recreational grounds alone. BENTHIC ALGAE The inhibitory power of macrophytes over benthic algae as a function of shading is well known (Vymazal, 1994; Stephenson et al., 1996), but the potential for rooted plants in general and milfoil in particular to foster benthic growths has been noted but minimally studied. Taste and odor producing blue-green mats have been detected by divers and controlled by copper 52

73 treatments (McGuire et al., 1984; Means et al., 1984), and have been observed to grew in close association with M. spicatum beds in a Connecticut reservoir (ENSR, 2005a) and in waterbodies in Washington state (Gibbons, pers. comm.). It may be that these blue-greens are more tolerant of low light, but it could also suggest that the sediments are conditioned by macrophyte growths in a way that favors these algae. Relation to Milfoil Growths The relationship of benthic algal mats to milfoil growths is largely uninvestigated, other than to document the effect of low light on algal growth in general. It has been noted that different species of vascular plants tend to be associated with different species of periphytic algae (Stephenson et al., 1996), so it is reasonable to suspect that milfoils may foster a distinct benthic algal flora. Observation of blue-green algal mats (mainly members of the Oscillatoriaceae, which include taste and odor and toxin producers) in association with M. spicatum and M. heterophyllum (Wagner, pers. obs.) warrants scrutiny in this regard. Impact on Water Quality The potential impact of benthic algal mats on water quality features of interest to water suppliers includes both taste and odor production and toxin generation. In most cases these objectionable compounds are associated with blue-green algae (AWWA, 2001; 2003b), or more properly, cyanobacteria. The extent of coverage necessary to cause problems remains uncertain, but must be sufficient to impart enough of the offending compounds to the overlying water to affect raw water at the intake, and is therefore a function of water depth as well as areal mat coverage. In a study of a Connecticut reservoir, taste and odor were detected in raw water when the 17 acre littoral zone associated with a shallow intake was at least 25% covered by blue-green mats (ENSR, 2005a). Rooted plants grew in a similar pattern, with milfoil as one of the two most abundant plant species, but any correlation between milfoil growths and cyanobacterial mats is speculative at this time. ASSOCIATED ECOLOGICAL ISSUES The information provided below reflects the potential direct and indirect ecosystem issues associated with infestation of milfoil species. While not directly related to water supply management, it is important that reservoir managers be aware of these topics when considering selection of a macrophyte control system. Nutrient Releases As noted with regard to the breakdown of macrophyte biomass and DBPP production, decay of macrophytes can lead to a release of nutrients into the waterbody. The respective nitrogen and phosphorus content of M. spicatum has been measured at 2-4% and % of dry weight, which is about 10% of wet weight (Elser, 1966; Madsen, 1991). With dry weight biomasses at 500 g/m 2 (an achievable level), this equates to 0.5 to 2.0 g P and 10 to 20 g N for every square meter of coverage by dense M. spicatum growths. Decaying stands of Myriophyllum species in a reservoir in Indiana contributed an amount equivalent to 2% and 18% 53

74 of the annual nitrogen and phosphorus watershed loadings, respectively (Landers, 1982). It is possible that a decline associated with a reservoir-wide chemical treatment could produce similar amounts of nutrients, although not all would be in an immediately available form. The timing of the nutrient release associated with natural milfoil declines (typically during autumn) suggests that most of these nutrients will not be available during the peak algal bloom season, but if the detention time of the reservoir is more than about 3 months, carryover to the following spring is possible. Overall, nutrients leaching from decaying macrophyte biomass are likely to be responsible for small local blooms, but are considered unlikely to significantly alter nutrient cycles in large water bodies (Hoyer and Canfield, 1997). Interactions with Phytoplankton Community Considerable additional information is available regarding competition between phytoplankton and macrophytes. The brief review presented below is intended to reinforce the concept that selection of control and management techniques for reduction of macrophyte populations in drinking water reservoirs need to consider whether application of such techniques will inadvertently result in a shift to higher levels of phytoplankton production. The factors controlling phytoplankton and macrophyte growth indicate that these two communities are often competing for similar resources, especially light and to some extent nutrients, although rooted plants typically use nutrients from the sediment while algae use nutrients from the water column in most cases. The relative success or abundance of one community may have a negative impact on the other. For example, high levels of phytoplankton may significantly reduce the water transparency and depth of light penetration (i.e., the photic zone) and decrease the density, diversity and coverage of macrophytes. Conversely, large amounts of littoral macrophytes may alter water chemistry and affect phytoplankton in ways that result in reduced algal blooms and greater light penetration (i.e., deeper Secchi disk transparency depth). Experimental manipulation of environmental factors in enclosed aquatic systems has allowed much better understanding of the interactions between phytoplankton and macrophytes (e.g., Stephan et al., 1998; Jones et al., 2002). The limnological literature is well versed with examples of waterbodies either dominated by phytoplankton in the water column or by rooted aquatic macrophytes and associated periphyton (Wetzel, 2001). Observations on shallow lakes under moderate nutrient regimes indicate that such systems may shift between two stable states, or alternative equilibria: a turbid state dominated by phytoplankton versus a contrasting clear water state dominated by submerged macrophytes (Hosper, 1998; Coops and Hosper, 2002; Jones et al., 2002). It has been postulated that these states exhibit so-called hysteresis effects that is when the ecosystem state shifts from one state to the other (e.g., clear to turbid), it is highly resistant to switching back (Harris, 2001). The presence of several positive feedback mechanisms are thought to be key to the stability of these systems (Beklioglu and Moss, 1996). However, it is possible to shift these states if environmental conditions are changed sufficiently to overcome these buffering factors. One of the more common ecosystem shifts can occur when nutrient loadings to a waterbody are either significantly increased or decreased. Several researchers have found a relationship between nutrients and macrophyte biomass or community structure (Lougheed et al., 2001; Barendregt and Bio, 2003; Carr et al., 2003). These results indicate that macrophytes respond to eutrophication and changes in water quality (nutrients, turbidity) and sediment quality 54

75 (nutrients, inorganic content) (Lougheed et al., 2001). Barendregt and Bio (2003) reviewed seven macrophyte studies in European waterbodies, all of which had nutrients as an a priori variable of interest. This principle is the basis for some current ecosystem restoration strategies which focus on the control and reduction of nutrient inputs (Cooke et al., 2005). However, the response of macrophytes to increased nutrients is more complex than phytoplankton due to the two routes of potential uptake, either from the water column or from the sediments (Barko and Smart, 1980). Localized responses of macrophytes to nutrient enrichment may also be complicated by the influence of increased organic inputs or sediment loads which may accompany the nutrient loadings and which also enhance the substrate s suitability for macrophyte colonization. While phosphorus is considered the usual limiting nutrient for freshwater phytoplankton in a majority of waterbodies, nitrogen may also be important to explain macrophyte abundance. Carr et al. (2003) conducted a two-year empirical survey of macrophyte biomass from 28 rivers in southwestern Ontario to quantify the relationship between biomass and instream nutrient levels. Multiple regression models indicated that mean and maximum biomass levels were weakly, but significantly, related to nutrients; TN and TP explained about 27% of the variability in biomass. Based on their models, Carr and co-workers suggested that nutrient abatement programs, focused primarily on nitrogen control, may be successful in reducing nuisance macrophyte biomass (Carr et al., 2003). Physical factors can also play a role in determining whether macrophytes or phytoplankton dominate a system. For example, Soballe and Kimmel (1987) found that rapid water renewal (i.e., flushing rate) had significant and negative impact on the abundance (per unit phosphorus) and predictability of phytoplankton and that in short-residence impoundments, abiotic factors unrelated to nutrient availability (e.g., temperature, turbidity, flow variations) may be important determinants of algal biomass (Soballe and Kimmel, 1987). Not surprisingly, variability in algal yield was greater in short-residence systems. Accordingly, rapid flushing rates (if unaccompanied by scouring flows) could favor establishment of macrophytes, if suitable substrate is available for rooting. For milfoil species with the potential to achieve high densities, the preference appears to be for at least moderate light and sandy substrates with substantial organic content. The alternative stable states hypothesis appears applicable to those plants, which could both suppress algae and be suppressed by high algal density. However, the potential problem milfoil species are likely to at least survive low light conditions, allowing resurgence when conditions are suitable. Effects on Fish, Macroinvertebrates and Zooplankton Aquatic macrophytes can provide food, shelter and spawning habitat for a wide variety of fishes (Lillie and Budd, 1992). Intermediate densities of aquatic macrophytes, even including milfoil, enhance fish diversity, feeding, growth and reproduction (Dibble et al., 1996). The optimal amount of macrophyte coverage is highly dependent on the fish species, its diet, preferred feeding habitat, spawning locations, and tolerance for macrophyte-mediated environmental change. However, dominance by milfoil tends to replace native macrophytes, altering the food web and creating food shortages for many fishes (Engel, 1995). Beds of milfoil can also impede predation, shelter overly dense assemblages of panfishes, and cover spawning areas, leading to 55

76 potential decreases in sportfish abundance (Engel, 1995). Large piscivorous fishes spend more time foraging for prey as plant density increases, thus reducing growth rates through unfavorable energetics (Savino and Stein, 1982). Milfoil beds have been shown to decrease fish abundance compared to native vegetation, and Keast (1983) found that beds of native vegetation supported up to four times as many fish and up to seven times as many macroinvertebrates. Decreases in macroinvertebrate abundance were observed as milfoil coverage increased in a Michigan study examining six lakes (Cheruvelil et al. 2001). In another Michigan study involving 13 lakes (Schneider 2000), milfoil was implicated in undesirable population features for centrarchid fishes (bass and sunfish). The direct impact of low DO levels caused by macrophyte overabundance has been previously described. The depletion of oxygen near dense milfoil beds can result in fish avoidance and, in extreme cases, fish kills (Holland and Huston, 1984; Lillie and Budd, 1992; Engel, 1995). Submerged macrophytes also influence fish populations through effects on habitat and predator-prey relationships. Several studies indicate that the growth rate of prey fish is stunted (reduced) when they are confined to dense macrophyte beds (Spencer, 2003). Adverse impacts on game fish (e.g., largemouth bass) have also been noted, including the reduced probability that a visual-based piscivorous fish will locate its prey (Savino and Stein, 1982). High densities of blue-gill can also reduce bass numbers through ingestion of eggs in shoreline areas or due to increased competition with young-of-year bass for their food sources (Petr, 2000). Macrophytes provide a structural refuge for the zooplankton against predation by zooplanktivorous fish, as well as decreasing the chances for their washout from a rapidlyflushing impoundment (Stephen et al., 1998; Jones et al., 2002), and milfoil can support this function. This refuge can help maintain zooplankton populations as well as increase the average zooplankton size (and accompanying algal grazing efficiency). By maintaining higher abundance of grazing populations and shifting the size spectrum of zooplankton towards larger species, control of phytoplankton is enhanced. This, in turn, can increase light transparency, reduce organic particulates, and reduce DBPP concentrations. However, the types of zooplankton that inhabit dense milfoil beds tend not to be the ones valued for open water grazing capacity. This may be a function of altered water chemistry, predation by the panfish that also hide in those weed beds, or alternative feeding strategies associated with the periphytic algae growths common to milfoil beds. In most cases, expansive milfoil beds do not represent an appropriate refuge for large-bodied species of Daphnia, the preferred grazing zooplankter for algae control. 56

77 CHAPTER 5 POTENTIAL CONTROL METHODS FOR NUISANCE AQUATIC VEGETATION, WITH EMPHASIS ON MILFOIL SPECIES A plethora of information is available regarding potential control methods for nuisance aquatic vegetation. A compendium of lake and reservoir management techniques which includes vegetation management has been published by the Massachusetts Office of Environmental Affairs in two documents: Eutrophication and Aquatic Plant Management in Massachusetts Final Generic Environmental Impact Report (hereon referred to as the GEIR; Mattson et al., 2004) and a companion guide, The Practical Guide to Lake Management in Massachusetts (hereon referred to as The Practical Guide; Wagner, 2004). These documents are available through the Massachusetts Department of Conservation and Recreation on the worldwide web at Much of the information presented herein was edited from these documents, as they represent a comprehensive and up to date perspective on plant management. During the preparation of this review, two more national publications became available and were consulted. These are Restoration and Management of Lakes and Reservoirs by Cooke et al. (2005) and Aquatic Plant Management: Best Management Practices for the Control of Aquatic Plants by a consortium of authors under the banner of the Aquatic Ecosystem Restoration Foundation (2005). The basic information contained in the documents generated for the Commonwealth of Massachusetts was augmented where warranted from these other two publications, but there was overall agreement among all resources on most issues. Nuisance aquatic plants, or weeds, are defined as any unwanted plant in excessive densities such that they degrade the utility of a waterbody. These plants can limit contact recreation, boating, fishing, impair habitat and aesthetics, or clog water supply intake structures. They can also impact water quality, as with increases in organic carbon concentrations, decreases in dissolved oxygen that encourage elevated sulfide, iron or manganese levels, or fluctuations in ph, each of which can affect the water treatment necessary for potable water. Nuisance vegetation is not limited to non-native species; many native plant species are known to form dense nuisance monocultures. However, invasive species, including three species of milfoil, are among the worst nuisance creators. This review will focus on milfoil, the target of this document, but many principles and techniques are applicable to many other nuisance species as well. Control of nuisance vegetation is often necessary to maintain designated uses of the waterbody. Preventing the introduction of non-native species is the first and often most critical management option. Controlling introduction is difficult and not always effective, but minimizes the number of infestations that must be addressed and can greatly decrease the cost of control. This is especially true for the Category I milfoil species, EWM, VWM and parrotfeather. Once an infestation has occurred, plants can be controlled in a variety of ways. The following sections provide specific control techniques for milfoil. They include: Prevention Physical/Mechanical Controls Chemical Controls 57

78 Biological Controls No Management Alternative Methods to control aquatic plants are summarized in Table 5-1 after Wagner (2001), as expanded by Wagner (2004) and revised again after reviewing AERF (2005) and Cooke et al. (2005). FACTORS FOR EVALUATING CONTROL METHODS A critical step in controlling nuisance aquatic vegetation is setting an appropriate goal for control, as factors for evaluating control measures will be highly dependant on these goals. For water supply purposes, having no macrophytes may seem desirable, but a low, dense cover in shallow reservoirs with silty bottoms can minimize turbidity, and a healthy plant community can be integral to overall stability of water quality during the growing season. As noted in Mattson et al. (2004), perhaps the simplest axiom for plant management is that if light penetrates to the bottom and the substrate is not rock or cobble, plants will grow. There may be a choice between types of vascular plants and algae, but growth by primary producers appears inevitable. A program intended to eliminate all plants is both unnatural and maintenance intensive, if possible at all. A program to structure the plant community to meet clear goals in an ecologically and ethically sound manner is more appropriate, although potentially still quite expensive. For water supplies, it will usually be sufficient to maintain a plant community that minimizes fluctuations in water quality and maintains desirable water quality overall. However, there may be cases where this is not enough, especially where multiple uses are being supported or treatment options after intake are limited. Factors for evaluating control measures include: Effectiveness short and long term Impacts to Non-Target Organisms short and long term Impacts to Water Quality short and long term Logistics and Mitigative Measures application limitations and needs Cost long and short term Regulations acceptability and constraints on use Maintenance and Monitoring repetitive use needs and tracking effects How a technique fares under review for application in a specific circumstance will depend on management goals and constraints. For instance, long term efficacy may be sacrificed for a less expensive technique if the goal is for temporary relief. A highly effective technique may be rejected if impacts to non-target organisms are considered too great. Because of this, a priority list of management options has not been established in this report. Evaluation factors will be discussed in association with the descriptions of control measures to allow for independent assessment based on management goals. Chapter 6 of this report provides more discussion of the considerations needed for the application of milfoil control techniques specific to drinking water supplies. 58

79 59 Table 5-1. Management options for control of aquatic plants (modified from Wagner 2001; 2004). OPTION MODE OF ACTION ADVANTAGES DISADVANTAGES APPLICABILITY TO WATER SUPPLIES Physical Controls 1) Benthic barriers Mat of variable composition laid on bottom of target area, preventing growth 1.a) Porous or looseweave synthetic materials 1.b) Non-porous or sheet synthetic materials Can cover area for as little as several months or permanently Maintenance improves effectiveness Laid on bottom and usually anchored by weights or stakes Removed and cleaned or flipped and repositioned at least once per year for maximum effect Laid on bottom and anchored by many stakes, anchors or weights, or by layer of sand Not typically removed, but may be swept or blown clean periodically Highly flexible control Reduces turbidity from soft bottoms Can cover undesirable substrate Can improve fish habitat by creating edge effects Allows some escape of gases which may build up underneath Panels may be flipped in place or removed for relatively easy cleaning or repositioning Prevents all plant growth until buried by sediment Minimizes interaction of sediment and water column May cause anoxia at under barrier May limit benthic invertebrates Non-selective interference with plants in target area May inhibit spawning/feeding by some fish species Allows some growth through pores Gas may still build up underneath in some cases, lifting barrier from bottom Gas build up may cause barrier to float upwards Strong anchoring makes removal difficult and can hinder maintenance High Localized control can be achieved and maintained Not typically used on larger areas (>2 acres) High Practical to apply and maintain in small areas Must maintain, however, to keep effective High More effort to install than porous forms, must vent trapped gases Less maintenance needed, but must avoid sediment accumulation on top (continued) 59

80 60 1.c) Sediments of a desirable composition Table 5-1. (Continued) OPTION MODE OF ACTION ADVANTAGES DISADVANTAGES APPLICABILITY TO WATER SUPPLIES Sediments may be Plant biomass buried Sediments may sink Low added on top of Seed banks can be into or mix with Regulatory existing sediments or buried deeper underlying muck restrictions on adding plants. Sediment can be made Permitting for added fill Use of sand or clay less hospitable to plant sediment may be Loss of storage can limit plant growths difficult volume growths and alter Nutrient release from Addition of sediment sediment-water sediments may be may cause initial interactions. reduced turbidity increase Sediments can be applied from the surface or suction dredged from below muck layer (reverse layering technique) 2) Dredging Physical sediment removal, with deposition in a containment area Can be applied on a limited basis, but is most often a major restructuring of an impacted system Plants and seed beds are removed and regrowth can be limited by light and/or substrate limitation Surface sediment can be made more appealing to humans Reverse layering requires no addition or removal of sediment Plant removal with some flexibility Increases water depth Can reduce pollutant reserves Can reduce sediment oxygen demand Can improve spawning habitat for many fish species Allows complete renovation of aquatic ecosystem New sediment may contain nutrients or other contaminants Generally too expensive for large scale application Temporarily removes benthic invertebrates May create turbidity Threat to fish Possible impacts from dredged material disposal Interference with recreation or other uses during dredging Usually expensive High Provides additional benefits of added volume and changing substrate Will not usually prevent regrowth; must apply additional maintenance techniques Expense is limiting factor (continued) 60

81 61 Table 5-1. (Continued) OPTION MODE OF ACTION ADVANTAGES DISADVANTAGES APPLICABILITY TO WATER SUPPLIES 2.a) Dry excavation Lake lowered to Tends to facilitate a Eliminates most Moderate maximum extent very thorough effort aquatic biota unless a Necessary drawdown practical May allow drying of portion left undrained impacts water supply, Target material dried to maximum extent sediments prior to removal Eliminates lake use during dredging unless areas are sequestered and possible Allows use of less pumped Conventional excavation equipment used to remove sediments specialized equipment May be achieved under seasonal low water conditions 2.b) Wet excavation Lake level may be lowered, but sediments not substantially dewatered 2.c) Hydraulic (or pneumatic) removal Draglines, bucket dredges, or longreach backhoes used to remove sediment Lake level not reduced Suction or cutterhead dredges create slurry which is hydraulically pumped to containment area Slurry is dewatered; sediment retained, water discharged Requires least preparation time or effort, tends to be least cost dredging approach May allow use of easily acquired equipment May preserve most aquatic biota Creates minimal turbidity and limits impact on biota Can allow some lake uses during dredging Allows removal with limited access or shoreline disturbance Usually creates extreme turbidity Messy removal operation May need to dry sediments prior to hauling Impairs most lake uses during dredging Often leaves some sediment behind Cannot handle extremely coarse or debris-laden materials Requires advanced and more expensive containment area Requires overflow discharge from containment area Low Creates too much turbidity, possible impacts in other areas of reservoir and downstream High Minimizes impacts to water supply and downstream Allows pumping of sediment slurry to location off site Requires substantial engineering and disposal arrangement 61 (continued)

82 62 Table 5-1. (Continued) OPTION MODE OF ACTION ADVANTAGES DISADVANTAGES APPLICABILITY TO WATER SUPPLIES 3) Dyes and surface covers 4) Mechanical removal ( harvesting ) Water-soluble dye is mixed with lake water, thereby limiting light penetration and inhibiting plant growth Dyes remain in solution until washed out of system. Opaque sheet material applied to water surface Plants reduced by mechanical means, possibly with disturbance of soils Collected plants may be placed on shore for composting or other disposal Wide range of techniques employed, from manual to highly mechanized Application once or twice per year usually needed Light limit on plant growth without high turbidity or great depth May achieve some control of algae as well May achieve some selectivity for species tolerant of low light Highly flexible control May remove other debris Can balance habitat and recreational needs while avoiding sensitive areas May not control peripheral or shallow water rooted plants May cause thermal stratification in shallow ponds May facilitate anoxia at sediment interface with water Covers inhibit gas exchange with atmosphere Possible impacts on aquatic fauna Non-selective removal of plants in treated area Possible spread of undesirable species by fragmentation Possible generation of turbidity Low Undesirable color to be avoided in water supplies Covers for very large areas are impractical Some possible applicability for small areas (covers) or offline, back-up systems (dyes) Moderate Will not prevent regrowth, but can maintain desired biomass level Applicable to moderate but not very large areas May spread species that reproduce by fragments, such as problem milfoil species 4.a) Hand pulling Plants uprooted by hand ( weeding ) and bagged for removal Highly selective technique Minimal disruption Labor intensive Difficult to perform in dense stands High Localized technique for low density growths (continued) 62

83 63 Table 5-1. (Continued) OPTION MODE OF ACTION ADVANTAGES DISADVANTAGES APPLICABILITY TO WATER SUPPLIES 4.b) Cutting (without Plants cut in place Generally efficient and Leaves root systems Low collection) above roots without being harvested less expensive than complete harvesting and part of plant for re-growth Water quality impacts of large Leaves cut vegetation to decay or to re-root quantities of plant biomass too great for Not selective within applied area most water supplies 4.c) Harvesting (with collection) Plants cut at depth of 2-10 ft and collected for removal from lake 4.d) Rotovation Plants, root systems, and surrounding sediment disturbed with mechanical blades Allows plant removal on greater scale Can thoroughly disrupt entire plant Limited depth of operation Usually leaves fragments which may re-root May impact fauna Not selective within applied area More expensive than cutting Usually leaves fragments which may re-root May impact fauna Not selective within applied area Creates turbidity More expensive than harvesting Moderate With adequate equipment, can maintain biomass consistent with use goals Not used for early infestations, due to risk of spread Low Turbidity impacts too high for most water supplies Not used for early infestations, due to risk of spread (continued) 63

84 64 Table 5-1. (Continued) OPTION MODE OF ACTION ADVANTAGES DISADVANTAGES APPLICABILITY TO WATER SUPPLIES 4.e) Hydroraking Plants, root systems Can thoroughly disrupt Usually leaves Moderate and surrounding entire plant fragments which may Applicable on very sediment and debris Also allows removal of re-root small scale disturbed with stumps or other May impact fauna Creates high turbidity mechanical rake, obstructions Not selective within Not used for early material usually applied area infestations, due to collected and Creates turbidity risk of spread removed More expensive than harvesting 5) Water level control Lowering or raising the water level to create an inhospitable environment for some or all aquatic plants Disrupts plant life cycle 5.a) Drawdown Lowering of water over winter period allows desiccation, freezing, and physical disruption Exposure and dewatering are critical aspects Variable species tolerance to drawdown Most effective on annual to once/3 yr. basis Outlet control can affect large area Provides widespread control in increments of water depth Complements certain other techniques (dredging, flushing) Control with some flexibility Opportunity for shoreline cleanup/structure repair Flood control utility Impacts vegetative propagation species with limited impact to seed producing populations 64 Potential issues with water supply Potential issues with flooding Potential impacts to non-target flora and fauna Possible impacts on linked wetlands Possible effects on reptiles/amphibians Reduction in potential water supply and fire fighting capacity Alteration of downstream flows May result in greater nutrient availability for algae Highly variable Negative impacts on storage Potential to alter peripheral sediments over many years Highly variable Drawdown to effective level may severely impair water supply Drawdown as a matter of normal withdrawal may control plants (continued)

85 65 Table 5-1. (Continued) OPTION MODE OF ACTION ADVANTAGES DISADVANTAGES APPLICABILITY TO WATER SUPPLIES 5.b) Flooding Higher water level in the spring can inhibit seed germination and Where water is available, this can be an inexpensive technique Water for raising the level may not be available Highly variable Water level rise could cause flooding, plant growth Plant growth need not Potential peripheral possible downstream Higher flows which be eliminated, merely flooding impacts are normally retarded or delayed Possible downstream Increased storage associated with Timing of water level impacts may result while elevated water levels control can selectively Non-target impacts limiting plant can flush seeds and favor certain desirable growths by depth plant fragments from species limitation on light system availability Chemical controls 6) Herbicides Liquid or pelletized herbicides applied to target area or to plants directly Contact or systemic poisons kill plants or limit growth Typically requires application every 1-5 yrs Wide range of control is possible May be able to selectively eliminate species May achieve some algae control as well Algal nuisances may increase where nutrients are available Possible toxicity to non-target species Possible downstream impacts Restrictions of water use for varying time after treatment Increased oxygen demand from decaying vegetation Possible recycling of nutrients to allow other growths Moderate Most herbicides can be used in water supplies with restrictions Will not prevent eventual regrowth, but can be used to reset plant community Social acceptability issues exist (continued) 65

86 66 Table 5-1. (Continued) OPTION MODE OF ACTION ADVANTAGES DISADVANTAGES APPLICABILITY TO WATER SUPPLIES 6.a) Forms of copper Contact herbicide Moderately effective Toxic to aquatic Moderate Cellular toxicant, suspected membrane control of some submersed plant species fauna as a function of concentration and Used for algae, but not rooted plants in transport disruption More often an algal ambient chemistry most cases Applied as wide variety of liquid or control agent Ineffective at colder temperatures Sometimes used in conjunction with granular forms, often in conjunction with polymers or other herbicides Copper ion persists; accumulates in sediments or moves downstream other herbicides to enhance their effectiveness 6.b) Forms of endothall (7-oxabicyclo [2.2.1] heptane-2,3- dicarboxylic acid) Contact herbicide with limited translocation potential Membrane-active chemical which inhibits protein synthesis Causes structural deterioration Applied as liquid or granules Moderate control of some emersed plant species, moderately to highly effective control of floating and submersed species Limited toxicity to fish at recommended dosages Rapid action Non-selective in treated area Toxic to aquatic fauna (varying degrees by formulation) Time delays on use for water supply, agriculture and recreation Safety hazards for applicators Low Restriction on use in drinking water supply limits but does not eliminate applicability Used as a spot treatment for localized control, but only rarely in potable supplies (continued) 66

87 67 6.c) Forms of diquat (6,7-dihydropyrido [1,2-2,1 -c] pyrazinediium dibromide) Table 5-1. (Continued) OPTION MODE OF ACTION ADVANTAGES DISADVANTAGES APPLICABILITY TO WATER SUPPLIES Contact herbicide Moderate control of Non-selective in Low Absorbed by foliage some emersed plant treated area Restriction on use in but not roots species, moderately to Toxic to zooplankton drinking water supply Strong oxidant; highly effective control at recommended limits but does not disrupts most cellular of floating or dosage eliminate functions submersed species Inactivated by applicability Applied as a liquid, Limited toxicity to fish suspended particles; Used as a spot sometimes in at recommended ineffective in muddy treatment for conjunction with dosages waters localized control, but copper Rapid action only rarely in potable supplies 6.d) Forms of 2,4-D (2,4-dichlorophenoxyl acetic acid) Systemic herbicide Readily absorbed and translocated throughout plant Inhibits cell division in new tissue, stimulates growth in older tissue, resulting in cell disruption Applied as liquid or granules, frequently as part of complex formulations, preferably during early growth phase Moderately to highly effective control of a variety of emersed, floating and submersed plants Can achieve some selectivity through application timing and concentration Fairly fast action Time delays on use for water supply, agriculture and recreation Variable toxicity to aquatic fauna, depending upon formulation and ambient water chemistry Time delays for use of treated water for agriculture and recreation Necessary concentration limits use in water supplies to locations very far from intakes Low Highly effective on milfoils and many other species, but restriction on concentration in drinking water supplies limits effective use Public perception issues often prevent use even where distance to intake is sufficient (continued) 67

88 68 Table 5-1. (Continued) OPTION MODE OF ACTION ADVANTAGES DISADVANTAGES APPLICABILITY TO WATER SUPPLIES 6.e) Forms of fluridone Systemic herbicide Can be used selectively, Impacts on non-target Moderate (1-methyl-3-phenyl-5- Inhibits carotenoid based on concentration plant species possible Effective on milfoils [-3-{trifluoromethyl} pigment synthesis Gradual deterioration of at higher doses and some other phenyl]-4[ih]- pyridinone) and impacts photosynthesis affected plants limits impact on oxygen level Extremely soluble and mixable; difficult submergent problem species Best applied as liquid (BOD) to perform partial Can be applied in or granules during Effective against lake treatments isolated areas early growth phase of plants several difficult-tocontrol species Requires extended contact time Can be applied near intakes at lower but Low toxicity to aquatic fauna potentially effective doses Social acceptability issues exist 6.g Amine salt of triclopyr Systemic herbicide, Effectively controls Impacts on non-target Moderate (3,5,6-trichloro-2- registered for aquatic many floating and plant species possible Limited experience, pyridinyloxyacetic use by USEPA in submersed plant species at higher doses but effective on many acid) 2004 Can be used selectively, Current time delay of problem species, Readily absorbed by more effective against 30 days on including milfoils foliage, translocated throughout plant dicot plant species, including many consumption of fish from treated areas Potential for more complete kill of Disrupts enzyme nuisance species target vegetation with systems specific to plants Effective against several difficult-tocontrol spot treatments than for contact herbicides Applied as liquid species Social acceptability spray or subsurface injected liquid Low toxicity to aquatic fauna issues exist Fast action (continued) 68

89 69 Table 5-1. (Continued) OPTION MODE OF ACTION ADVANTAGES DISADVANTAGES APPLICABILITY TO WATER SUPPLIES Biological Controls 7) Biological introductions Fish, insects or pathogens which feed on or parasitize plants are added to system to affect control 7.a) Herbivorous fish (mainly grass carp, Ctenopharyngodon idella, but other similar species possible) Most commonly used organism is the grass carp, but the larvae of several insects have been used more recently, and viruses are being tested Sterile juveniles stocked at density which allows control over multiple years Growth of individuals offsets losses or may increase herbivorous pressure Provides potentially continuing control with one treatment Harnesses biological interactions to produce desired conditions May produce potentially useful fish biomass as an end product May greatly reduce plant biomass in single season May provide multiple years of control from single stocking Sterility intended to prevent population perpetuation and allow later adjustments Typically involves introduction of nonnative species Effects may not be controllable or reliable Plant selectivity may not match desired target species May adversely affect indigenous species May eliminate all plant biomass, or impact non-target species Funnels energy into algae Alters habitat May escape upstream or downstream Population control issues Low to Moderate Legality and effectiveness issues exist Success may improve with more experience, but variability of biological responses is to be expected Low Grass carp are not legal in many states Conversion of plant biomass to available nutrients may promote algal blooms Loss of all plants generally undesirable (continued) 69

90 70 Table 5-1. (Continued) OPTION MODE OF ACTION ADVANTAGES DISADVANTAGES APPLICABILITY TO WATER SUPPLIES 7.b) Herbivorous insects Larvae or adults stocked at density intended to allow Can involve species native to region, or even targeted lake Population ecology suggests incomplete control likely Moderate Milfoil weevil has provided variable control after growth Expected to have no Oscillating cycle of results for Eurasian Intended to selectively control negative effect on nontarget species control and re-growth expected water milfoil Success achieved for target species May facilitate longer Predation by fish may some other problem May be selfsustaining term control with limit control species (purple or require periodic augmentation limited management Other lake management actions may interfere with loosestrife, water hyacinth), but not for other milfoil species 7.c) Fungal/bacterial/ viral pathogens Inoculum used to seed lake or target plant patch Growth of pathogen population expected to achieve control over target species 7.d) Selective plantings Establishment of plant assemblage resistant to undesirable species Plants introduced as seeds, cuttings or whole plants May be highly species specific May provide substantial control after minimal inoculation effort Can restore native assemblage Can encourage assemblage most suitable to lake uses Supplements targeted species removal effort success Effectiveness and longevity of control not well known Infection ecology suggests incomplete control likely Largely experimental May not prevent nuisance species from returning Introduced species may become nuisances Low No commercially available forms; some under development for algae, but not milfoil Moderate Not sufficient by itself, but supplements control techniques that reset the plant community, if natural colonization is inadequate 70

91 PREVENTION APPROACHES Prevention of introduction and infestation is the preferred method of avoiding nuisance milfoils. This method is not always successful, but is clearly the most cost-effective approach. Prevention activities include education and awareness, monitoring and early detection, and boat, bird and inlet related controls. Education and awareness may be fostered through the dissemination of information regarding targeted species to the general public. Newsletters, flyers, websites, courses, public seminars, and direct contact at points of access can all be used to educate people about the impacts associated with dense vegetation in general and introducing problem milfoil species Education and awareness go hand in hand with monitoring for early detection. Active stewards are more likely to be proactive in the monitoring of conditions and identifying infestations. Early detection is a must for cost effective management. Early detection generally results in less impacted area which in turn results in lower management cost. Many states have developed or are developing Early Detection and Rapid Response Plans. These plans describe recommended approaches to locate and identify infestations of aggressive species as early after introduction as possible. This is the time when eradication is most feasible, as the area of impact is usually small. The plan may also include targeted areas for surveillance, including areas where introduction is most likely (i.e., boat ramps or an inlet from a waterbody with an infestation) or areas that deserve special attention (i.e., areas near intakes or with protected species). Rapid Response Plans can be found on the internet for some states (e.g., Maine at: Monitoring at access points and areas where birds congregate will be the most efficient means of detecting invasions without a more intensive, lakewide assessment, although periodic assessment of the entire littoral zone is advised. For more comprehensive assessments, there are multiple methods of plant survey, with no truly standard technique. The object is to be as thorough as time and trained manpower allow, maximizing detection probability. To detect a suspected invasion, or simply to monitor for possible invasion, the water supplier should develop an effective monitoring plan. More information on what should be included in such a plan is provided in Chapter 6.0. Aside from education, monitoring, and early detection, specific controls can be placed on potential transport vectors such as boating, selling of exotic plants within the aquarium and gardening trades, and accidental or intentional release into the wild. Laws and ordinances exist in some states and can be passed in others, and road side and boat launch inspections can be instituted with fines enforced for those found to be transporting nuisance species. Some states have been working closely with retailers in the aquarium and landscape trades, educating the suppliers about the plants and animals they sell and performing inspection of retail establishments to be sure they are not selling invasive species to the public. Controlling the sale of nuisance species is becoming a more difficult challenge, however, with the ability to purchase plants and animals over the internet. Natural vectors of transport, especially birds and mammals, are also difficult to control. Animals can transport plant fragments and seeds on their bodies and through ingestion and defecation. A careful analysis of vectors and possible means of control is recommended for each waterbody. 71

92 For the problem milfoils, the aquarium trade has been a source of EWM, VWM and parrotfeather, aquatic garden outlets have sold parrotfeather as an ornamental plant, and boat transport of EWM and VWM has been a major concern in most states. Birds appear to be the primary non-human vector, other than downstream transport of viable fragments by flow from infested waterbodies. Posting access areas with pictures of problem species is desirable, with information on why potential transport of aquatic plant fragments to or from the reservoir is undesirable and who to contact for more information. It is also paramount to encourage reservoir users to clean boats, trailers, fishing equipment and any other possible transport means before and after using the reservoir, as milfoil fragments are the primary means of new infestations. PHYSICAL/MECHANICAL CONTROLS Physical and mechanical controls involve the direct alteration of the plant itself, the substrate, water column or other environmental features upon which milfoil depends for survival. Some texts (AERF, 2005; Cooke et al., 2005) define physical controls as those that affect plants by altering the environment (e.g., drawdown, dyes) and mechanical controls as those techniques that directly attack the plants (e.g., various forms of harvesting, benthic barriers), with methods like dredging representing a combination of the two. This review treats all of these techniques as physical controls for simplicity. Physical controls for milfoil include benthic barriers, dredging, dyes, surface covers, harvesting, and water level controls. Each of these techniques is described in the following sections. Benthic Barriers How it Works The use of benthic barriers, or bottom covers, is predicated upon the principles that rooted plants require light and cannot grow through physical barriers. Applications of clay, silt, sand, and gravel have been used for many years, although plants often root in these covers eventually, and current environmental regulations make it difficult to gain approval for such deposition of fill. Artificial sediment covering materials, including polyethylene, polypropylene, fiberglass, and nylon, have been developed over the last three decades. A variety of solid and porous forms have been used. Manufactured benthic barriers are negatively buoyant materials, usually in sheet form, which can be applied on top of plants to limit light, physically disrupt growth, and allow unfavorable chemical reactions to interfere with further development of plants. Various plastics and burlap have also been used, but are not nearly as durable or effective in most cases. In theory, benthic barriers should be a highly effective plant control technique, at least on a localized, area-selective scale. In practice, however, there have been difficulties with the deployment and maintenance of benthic barriers, limiting their utility over the broad range of field conditions. Benthic barriers can be effectively used in small areas such as dock spaces and swimming beaches to completely terminate plant growth. The creation of access lanes and structural habitat diversity is also practical. Large areas are not often treated, however, because the cost of materials, application and maintenance is high. Benthic barrier problems of prime concern include long-term integrity of the barrier, billowing caused by trapped gases, accumulation of sediment on top of barriers, and growth of 72

93 plants on porous barriers. Successful use is related to selection of materials and the quality of the installation. As a result of field experience with benthic barriers, several guidelines can be offered: Porous barriers will be subject to less billowing, but will allow settling plant fragments to root and grow; annual maintenance is therefore essential Solid barriers will generally prevent rooting in the absence of sediment accumulations, but will billow after enough gases accumulate; venting and strong anchoring are essential in most cases Plants under the barrier will usually die completely after one to two months, with solid barriers more effective than porous ones in killing the whole plant; barriers of sufficient tensile strength can then be moved to a new location, although continued presence of solid barriers restricts recolonization Proper application requires that the barriers be placed on the sediment surface and staked or securely anchored. This may be difficult to accomplish over dense plant growth, and a winter drawdown can provide an ideal opportunity for application in exposed areas. Late spring application has also been effective, despite the presence of plant growths at that time, and barriers applied in early May have been removed in mid-june with no substantial plant growth through the summer. Scuba divers normally apply the covers in deeper water, which greatly increases labor costs. Bottom barriers will accumulate sediment deposits in most cases, which allow plant fragments to root. Barriers must then be cleaned, necessitating either removal or laborious in-place maintenance. Despite application and maintenance issues, a benthic barrier can be a very effective tool. Benthic barriers are capable of providing control of rooted plants on at least a localized basis, and have such desirable side benefits as creating more edge habitat within dense plant assemblages and minimizing turbidity generation from fine bottom sediments. Considerations for the installation of benthic barriers include the size of the area to be treated, bottom features and possible obstructions, the cost of the product, application and maintenance costs, and possible impacts to non-target organisms in the installation area. Sheeting materials come in a variety of dimensions, from about 20 ft by 50 ft to 7 ft by 100 ft, although custom sizes of a wider range are possible. Deployment is therefore a function of manpower and cleverness by the installer. Careful consideration of site conditions is essential to maximizing effectiveness, as barriers must remain in place for at least a month and possibly two months to kill the target plants. There are many ways to install barriers, ranging from spreading them out with the reservoir drawn down to underwater positioning by divers. In water less than about 10 ft. deep, snorkeling may be sufficient to get the barrier properly positioned. One aid to application involves rolling the barrier onto PVC pipe with a slightly longer wooden or metal pole inside the PVC pipe, allowing the barrier to be rolled out like paper towels. Anchoring systems vary with barrier type, but most forms do require staking or weighting. Sleeves can be sewn into sheet materials to allow rebar to be inserted, pieces of chain can be attached to edges, or patio blocks can be dropped onto the barrier to hold it in place. Burial under sandy sediments has been tried, but may allow more rapid plant recolonization. Where removal at a later date is desired, the weighting system should be simple and reversible (patio block weights are very convenient in this regard). 73

94 One way to extend the benefits of benthic barrier involves flipping the barrier over into the adjacent area after one to two months. Plants are killed over that time period, and the barrier can be redeployed to the adjacent plot as part of normal maintenance. In this manner, two or three times the area of the benthic barrier can be treated in a single growing season. If plant elimination is not necessary, and simply reducing plant biomass is acceptable, it may be possible to move the barrier on a biweekly schedule. This could allow a linear band of nuisance vegetation to be managed over the first few months of the growing season, creating acceptable conditions over a larger area with a smaller barrier. Manpower is the primary limiting factor in this approach, although not all barriers can be moved once installed. Effectiveness Benefits Detriments There are several examples where benthic barriers have proven to be a rapid and effective method over the short and long term (Mattson et al., 2004). Efficacy is highly dependant on proper installation and maintenance, both of which can be labor intensive Complete elimination of plants in target area with proper application and maintenance Some barrier materials are re-useable, allowing coverage of multiple areas over time with the same material Creates edge effect and habitat enhancement when portions of dense assemblages are covered May foster improved assemblage after removal, by seeds or selective planting Non-selective technique; all plants under barrier will be killed Effectiveness declines without labor-intensive maintenance Impacts to Non-Target Species Non-selective technique; all plants under barrier will be killed Spawning and/or feeding of fish may be inhibited Benthic invertebrates may be killed or harmed Factors Favoring the Use of this Technique The target area has dense plant growths of undesirable species The target area is small (<1 acre) and relatively free of obstructions (stumps, logs, boulders, pilings and moorings) The target area represents only a small portion of the whole reservoir (<10%) Long-term control is sought over a small area with recognition of necessary maintenance needs Inexpensive labor is available 74

95 No significant shellfish resources are present in the target area A favorable plant assemblage is expected to develop (or can be encouraged by planting) after barrier removal Regulatory Requirements May need State and/or local Wetlands Protection permit for installation May need permit associated with Endangered Species Act from U.S. Fish and Wildlife, State, or local authority Cost Considerations The most commonly used materials for benthic barriers and the cost (material only) include Texel at $0.25/sq.ft, Palco at $0.40/sq.ft, and Aquatic Weed Net at $0.60/sq.ft. Less expensive substitutes can be found, but usually lack the properties that make these barriers as effective as they are. Such substitution will save initial material costs, but may require more material over the long-term and may increase labor costs to achieve the same effectiveness. Cost per acre is estimated at $20,000 to $50,000 for benthic barrier installation, including design, permitting, materials and labor for a year. The initial capital cost is substantial, but the annual cost diminishes greatly after original installation, as material costs are minimal after initial purchase. Dredging How it Works Dredging involves the removal of sediment. Conventional dry, conventional wet, and hydraulic/pneumatic dredging are each addressed separately here, as planning and impact considerations vary substantially. Dredging is perhaps best known for increasing depth, but dredging can be an effective reservoir management technique for the control of excessive algae and invasive growth of macrophytes. The management objectives of a sediment removal project are usually to deepen a shallow reservoir for boating and fishing, or to remove nutrient rich sediments that can cause algal blooms or support dense growths of rooted macrophytes. Control of rooted aquatic vascular plants is achieved by either the removal of substrate hospitable for their growth or by deepening the area enough to create a light limitation on plant growth. Yet invasives such as milfoil can grow in waters up to 30 feet deep with adequate light, and removal of hospitable substrate must be very complete; both forms of limitation are therefore difficult to achieve with milfoil. Removal of root systems and any seeds may greatly retard growth, but true eradication of milfoil by dredging is elusive. Benefits of dredging can extend beyond rooted plant control. Removal of algaestimulating nutrients from reservoir sediments can reduce internal loading and suppress algal production if internal sources are the dominant nutrient source. Even where incoming nutrient loads remain high, dredging can reduce benthic mat formation and related problems with filamentous green and blue-green algae, as these forms may initially depend on nutrient-rich substrates for nutrition. Dredging also removes the accumulated resting cysts deposited by a variety of algae. 75

96 Each dredging technique has its own unique benefits, detriments, impacts, favorable factors and regulatory requirements, but key characteristics associated with the three techniques are summarized below. A properly conducted dredging program removes accumulated sediment from a reservoir and effectively sets it back in time, to a point prior to significant sedimentation. Partial dredging projects are possible and may be appropriate depending upon management goals, but for maximum benefit it is far better to remove all soft sediment. Failed dredging projects are common, and failure can almost always be traced to insufficient consideration of the many factors that govern dredging success (Mattson et al., 2004; Cooke et al., 2005). Effectiveness Benefits Detriments Successful short and long term control provided final depth is deep enough to limit light penetration or all hospitable substrate is removed Failures are associated with lack of thorough pre-dredging feasibility assessments and planning, or application of dredging where it is not an appropriate technique Deepening of the reservoir for many purposes, including increased flood or water supply storage, improved recreational uses, enhanced pollutant trapping effectiveness and dilution of nutrient Control of rooted plants if a depth (light) or substrate limitation is imposed Removal of all plant parts, including roots and seeds, will at least temporarily limit growths Reduced algal mat formation by reduced nutrient supply and elimination of resting cysts Reduced planktonic algal abundance if internal loading is an important nutrient source and enough sediment is removed Removal of toxic substances or other unwanted materials accumulated in the sediment Reduced sediment-water interactions, with potential improvement in water quality Loss of most biological components of any dredged portion of the reservoir through physical disturbance Loss of most biological components of any drained portion of the reservoir through desiccation Creation of open area that may be colonized by invasive species Upland area must be provided for sediment disposal, with temporary alteration Contaminated sediments potentially subject to many restrictions on disposal 76

97 Impacts to Non-Target Species Widespread impacts to benthic invertebrates, non-target plants, and other non-mobile organisms. Organisms may be physically removed (or allowed to dry out in the case of dry dredging) Change in species community; complete recolonization may take years Decrease in food for fish Habitat loss for fish and benthic species Burial of non-target species at disposal site Possible contamination of soil and groundwater at disposal site Factors Favoring the Use of this Technique There is a distinct need for increased depth or storage capacity in the reservoir Rooted plants and algal mats dependent on the soft sediments in shallow water are impairing water supply or other valued uses Studies have demonstrated the impact of internal loading on the reservoir Studies have demonstrated the presence of contaminants that are impacting reservoir biota or uses Sediments are clean, based on State and Federal regulatory thresholds Suitable and sufficient containment and disposal areas are available close to the reservoir Regulatory Requirements May need State and/or local Wetlands Protection permit May need permit associated with Endangered Species Act from U.S. Fish and Wildlife, State, or local authority 404 permit through the Corps of Engineers 401 Water Quality permit through State Agencies Solid Waste permit for sediment disposal through State Agencies May need permit associated with Endangered Species Act from U.S. Fish and Wildlife, State, or local authority Conventional Dry Dredging Dry dredging involves partially or completely draining the reservoir and removing the exposed bottom sediments with a bulldozer or other conventional excavation equipment and trucking it away. Projects involving silts, sands, gravel and larger obstructions where water level can be controlled favor conventional, dry methodology. Although ponds rarely dry to the point where equipment can be used without some form of support (e.g., railroad tie mats or gravel placed to form a road), excavating under dry conditions allows very thorough sediment removal and a complete restructuring of the pond bottom. The term dry may be a misnomer in many cases, as organic sediments will not dewater sufficiently to be moved like upland soils. 77

98 Dry dredging may resemble a large-scale excavation of pudding, and the more the material is handled, the more liquid it becomes. Control of inflow to the reservoir is critical during dry excavation. For dry excavation, water can often be routed through the reservoir in a sequestered channel or pipe, limiting interaction with disturbed sediments. Water added from upstream or directly from precipitation will result in solids content rarely in excess of 50% and often as low as 30%. Consequently, some form of containment area is needed before material can be used productively in upland projects. Where there is an old gravel pit or similar area to be filled, one-step disposal is facilitated, but most projects involve temporary and permanent disposal steps. Conventional Wet Dredging Wet dredging may involve a partial drawdown, especially to avoid downstream flow of turbid water, but sediment will be excavated from areas overlain by water. Sediment will be very wet, often only 10 to 30% solids unless sand and gravel deposits are being removed. Clamshell dredges, draglines, and other specialized excavation equipment are used in what most people would consider a very messy operation. Excavated sediment must usually be deposited in a bermed area adjacent to the pond or into other water-holding structures until dewatering can occur. This approach is most often practiced when water level control is limited. Aside from small ponds, this technique is applicable to ocean harbors. Conventional wet dredging methods create considerable turbidity, and steps must be taken to prevent downstream mobilization of sediments and associated contaminants. For wet excavation projects, inflows must normally be routed around the reservoir, as each increment of inflow must be balanced by an equal amount of outflow, and the in-lake waters may be very turbid. It should be noted, however, that more recent bucket dredge designs greatly limit the release of turbid water and have been approved for use in potentially sensitive aquatic settings. Hydraulic or Pneumatic Dredging A more advanced form of wet dredging, hydraulic dredging usually involves a suction type of dredge that has a cutter head. Agitation combined with suction removes the sediments as a slurry containing approximately 15-20% solids by volume, although this may increase to as high as 30 to 40% in some cases or be as low as 5% with especially watery sediments in difficult areas. This slurry is typically pumped to a containment area in an upland setting where the excess water can be separated from the solids by settling (with or without augmentation). The supernatant water can be released back to the reservoir or some other waterway. The containment area for a hydraulic dredging project is usually a shallow diked area that is used as a settling basin. The clarified water may be treated with flocculation and coagulation techniques to further reduce the suspended solids in the return water. Hydraulic dredging is normally favored for removal of large amounts of highly organic sediments with few rocks, stumps or other obstructions and where water level control is limited. This type of project does require a containment area to be available where removed sediments are separated from water, and may involve secondary removal of the dried sediment from the containment area for ultimate disposal elsewhere. Usually the containment area is not far from the reservoir, but a slurry can be pumped multiple miles along a suitable route with booster pumps. 78

99 Innovations in polymers and belt presses for sediment dewatering have reached the point where hydraulically dredged slurry can be treated as it leaves the reservoir to the extent necessary to load it directly onto trucks for transport to more remote sites. Solids content of the resultant material is still too low for many uses without further drying or mixing with sand, but the need for a large containment area can be avoided with this technology. The cost of coagulation and mechanical dewatering may be at least partially offset by savings in containment area construction and ultimate material disposal. Likewise, pumping the slurry into geo-tubes (engineered filter bags) can also enhance dewatering in a limited space. Pneumatic dredging, in which air pressure is used to pump sediments out of the reservoir at a higher solids content (50 to 70%) has not yet been widely performed in the United States. This would seem to be a highly desirable approach, given containment area limitation in many cases and more rapid drying with higher solids content. However, few of these dredges are operating within North America, and there is little freshwater experience upon which to base a review. Considerations are much like those for hydraulic dredging. Cost Considerations Because the cost varies depending on the volume of material removed, costs are usually expressed per cubic yard (cy) of material removed. In general, larger projects incur a smaller cost per cubic yard. The proper way to estimate dredging costs is to consider each element of the project, which may vary dramatically among projects. The total cost can be divided by the total yardage to get a cost per cubic yard, but this may not be especially meaningful in estimating other dredging projects. With that caveat in mind, a typical range of costs for dredging projects in recent years is $7 to $25/cy, with $15/cy suggested as a rough estimator for considering the general magnitude of a project under initial consideration. It is important, however, to develop a more careful estimate during further project planning, and many smaller projects (<50,000 cy) have incurred costs in excess of $30/cy. Total cost can be reduced if the dredged material is clean enough to be sold as a soil amendment. Recovery of more than $1/cy is unusual, however. In some cases, contractors have wanted material from the reservoir, reducing cost. Income from material sale should not be assumed, however, unless a firm agreement is in hand. Cost factors of major importance include: Volume of material Cost per unit volume generally declines with increasing volume Distance to containment area Trucking costs or the need for booster pumps in hydraulic dredging increases cost Size of containment area Cost will escalate if dredging must cease periodically to allow sediment settling or containment area clean-out Obstructions and clogging agents Dredging may be impeded by rocks, stumps, structures and dense plant growths Reverse Layering How it Works An alternative method to dredging that is believed to provide some of the same benefits is the reverse layering of sediments. While dredging involves the removal of sediment, reverse 79

100 layering simply reorganizes sediment layers. It is still a largely experimental procedure that has been tested in small areas of Red Lily Pond in Barnstable, Massachusetts. It is believed to be especially applicable to the glacial "kettle hole" ponds that are common to Cape Cod and Southeastern Massachusetts because of a layer of glacial sand that lies beneath the accumulated muck layer. The purpose is to extract glacial sand that underlies the nutrient-rich, anaerobic, organic sediments of a eutrophic lake and place it on top of those less desirable sediments. Reverse layering is accomplished by hydraulic jetting. Water is pumped down below the muck and/or peat layer to the layer of glacial sand. The glacial sand is forced up through pipes and spread over the bottom sediments. A cavity is created by the removal of glacial sand, which causes the bottom sediments to subside and fill the cavity. The purpose of this method is to restore the lake bottom to the original sediment type, potentially promoting a more diverse plant and animal community. This method does not require disposal of dredged materials, nor does it deepen the lake. It simply switches the location of existing sediment layers. Reverse layering is not considered dredging by some groups, most notably the US Army Corps of Engineers, which therefore does not require a permit under Section 404 of the Clean Water Act as it normally does for dredging projects. It is certainly a very different technique than the other methods of actual sediment removal, but the underlying goal is the same with regard to nutrient and algae control; limit the availability of nutrients from accumulated muck sediments. The goal for plant control will not involve light limitation as with some dredging projects, but creation of a less hospitable substrate may be possible, thereby reducing rooted plant density or at least shifting community composition. In this regard, reverse layering is more like a benthic barrier treatment. Effectiveness Benefits Detriments Short term success can be achieved through plant substrate disturbance Long term control is not expected since no light limitation has been achieved Control of rooted plants if a substrate limitation is imposed by the new sand layer, but this is unlikely with milfoils Reduced algal mat formation by reduced nutrient supply and burial of resting cysts Reduced planktonic algal abundance if internal loading is an important nutrient source and enough organic sediment is covered Reduced sediment-water interactions, with potential improvement in water quality Fine sediment mixed with sand may be dispersed during the jetting and layering operation, along with any associated contaminants The new surficial sediment layer may support plants with equal or greater nuisance potential, and represents a colonization opportunity for invasive species 80

101 Impacts to Non-Target Species Widespread impacts to benthic invertebrates, non-target plants, and other non-mobile organisms. Organisms may be buried. Possible habitat loss for fish and benthic species Possible habitat impairment due to increased turbidity in waterbody and downstream resources Factors Favoring the Use of this Technique There is a layer of coarse, clean sand under the surficial muck Studies have demonstrated the impact of internal loading on the lake Studies have demonstrated the presence of contaminants in surficial sediments that are impacting lake biota or uses Rooted plants and algal mats dependent on the soft sediments are impairing water supply or other valued uses Regulatory Requirements May need State and/or local Wetlands Protection permit May need permit associated with Endangered Species Act from U.S. Fish and Wildlife, State, or local authority Cost Considerations Costs for reverse layering of sediments were estimated at $10,000/acre in This technique has not been used enough to provide a reliable estimate of costs, however. Dyes and Covers How it Works The use of dyes as algal or vascular plant control agents is often grouped with herbicides in lake management evaluations, but this can be very misleading with regard to how dyes work. Dyes are used to limit light penetration and therefore restrict the depth at which rooted plants can grow or the total amount of light available for algal growth. They are only selective in the sense that they favor species tolerant of low light or with sufficient food reserves to support an extended growth period (during which a stem could reach the lighted zone). Dyes are generally non-toxic to all aquatic species, including the target species of plants. In lakes with high transparency but only moderate depth and ample soft sediment accumulations, dyes may provide open water where little would otherwise exist. Repeated treatment will be necessary, as the dye eventually flushes out of the system. Dyes are typically permitted under the same process as herbicides, despite their radically different mode of action. Given this association with herbicides and the color imparted to the water, they are rarely used in drinking water situations, but might be applicable in some off-line reservoirs where weed control is desired and water is not regularly drawn into the supply system. 81

102 Surface shading has received little attention as a rooted plant control technique, probably as a function of potential interference with recreational pursuits, the enhancement of which is a goal of most rooted plant control programs. This procedure should be a useful and inexpensive alternative to traditional methods of weed control in small areas such as docks and beaches, and could be timed to yield results acceptable to summer human users with minimal negative impacts to system ecology. The shading effect of bottom barriers is well known, and would be at work with surface covers. However, such covers are impractical on a large scale, so use in drinking water reservoirs is likely to be limited to smaller storage reservoirs. As covered storage is required in finished water reservoirs to minimize impacts from birds and human sources, covers that limit light and plants may be applicable in some cases. Effectiveness Benefits Detriments Success can be achieved through light limitation by either dyes or covers Long term control is not expected with dyes because of dilution Long term control may be possible with covers if they are not removed Can cause shifts in plant community without physical disruption or toxic reactions Can be used for control on a temporary basis Altered color is rarely acceptable in drinking water reservoirs Increased heat absorption may cause stratification of shallow lakes and possible loss of oxygen in the bottom waters Surface covers limit atmospheric interaction and may affect water quality Impacts to Non-Target Species Covers impede waterfowl and human use, but that may be desirable in storage reservoirs May cause shift of plant community to low light tolerant species May cause shift of faunal communities May impact visual fish feeding Factors Favoring the Use of this Technique There is normally no surface outflow from reservoir if dyes are being considered Access for humans or waterfowl is not needed during the time surface covers will be in place Regulatory Requirements May need State and/or local Wetlands Protection permit 82

103 May need permit associated with Endangered Species Act from U.S. Fish and Wildlife, State, or local authority Cost Considerations The cost of dye is about $70 per gallon or $250 for a 4 x 1 gallon case. One gallon treats about 4 acre-feet. A cost of $100 to $500 per acre, including planning, permitting, materials and labor, might be expected. Costs have not been reported for any surface cover installations, but assuming the use of bottom barrier materials, the cost would be at least $20,000/acre for materials. Assuming the use of simple black plastic sheeting, material costs would be largely a function of frame and anchoring materials. It seems likely that a cost of $2,000 to $5,000 per acre could be achieved, but such installations may not be sturdy enough for covered storage reservoirs. While costs for plant control by light limitation may appear favorable, these techniques are not applicable to most potable water reservoirs. Harvesting There are several methods of harvesting with varying degrees of scale, mechanization, and cost, with the choice of which to apply based largely on the plant community and size of area to be addressed. These techniques include hand pulling, suction harvesting, mechanical harvesting (cutting with and without collection), rotovation, and hydroraking. Each of these harvesting methods is described separately below, and those considering harvesting methods should be specific about which method is intended. Hand Pulling How it Works Hand pulling is exactly what it sounds like; a snorkeler or diver surveys an area and selectively pulls out unwanted plants on an individual basis. This is a highly selective technique, and a labor intensive one. It is well suited to vigilant efforts to keep out invasive species that have not yet become established in the reservoir or area of concern. Hand pulling can also effectively address non-dominant growths of undesirable species in mixed assemblages, or small patches of plants targeted for removal. This technique is not well suited to large-scale efforts, especially when the target species or assemblage occurs in dense or expansive beds. Hand pulling can be augmented by various tools, including a wide assortment of rakes, cutting tools, water jetting devices, nets and other collection devices. McComas (1993) provides an extensive review of options. Suction dredging is also used to augment hand pulling, allowing a higher rate of pulling in a targeted area, as the diver/snorkeler does not have to bag pulled plants and carry them to a disposal point. Use of these tools transitions into more mechanized forms of harvesting. Effectiveness Annual monitoring and pulling recommended for newly arriving invasive species 83

104 Benefits Detriments Extremely effective for low density (<1 target plant per 10 square feet), localized populations Impractical for large areas (>10 acres) unless a large workforce is available Successful at minimizing regrowth after other methods are applied Highly selective plant control Limited impact to non-target organisms Can prevent infestations before they become problems Incomplete harvesting may foster regrowth or dispersal of plants Turbidity generation may be substantial Highly labor intensive, especially over large areas Impacts to Non-Target Species No long term effects to non-target species from well designed pulling program Temporary impacts from increased turbidity Factors Favoring the Use of this Technique Nuisance species are not yet established; target plant density is low (<500 stems/acre) Target species are in shallow water, or dive crew is readily available Target species are not strongly rooted or prone to fragmentation Regulatory Requirements May need State and/or local Wetlands Protection permit May need permit associated with Endangered Species Act from U.S. Fish and Wildlife, State, or local authority Cost Considerations Many hand harvesting efforts are volunteer programs, so costs are difficult to estimate. For programs where cost accounting is possible, the cost of hand harvesting when targeted plant density is sparse is estimated at $150-$300/acre, with most of these representing control of new and sparse milfoil growths. A range of $100 to $500/acre for sparse to moderate growths is suggested; the cost for hand harvesting dense stands would be much higher. 84

105 Suction Harvesting How it Works Suction harvesting, or suction dredging, is mechanically augmented hand pulling. The diver hand pulls the unwanted plants and allows them to be transported through a vacuum hose to the surface into a mesh bag or other collection device. This technique accelerates the hand pulling process allowing pulling for denser assemblages but generally does not increase the area of control (Mattson et al., 2004). See Hand Pulling for more information regarding this technique. Effectiveness Benefits Detriments Effective for localized populations (<2 acres) of varying density Impractical for large areas (>10 acres) Improves diver efficiency Moderately selective plant control Limited impact to non-target organisms Can prevent infestations before they become problems Incomplete harvesting may foster regrowth or dispersal of plants Turbidity generation is substantial Impacts to Non-Target Species No long term effects to non-target species from well designed pulling program Some loss of non-target species is possible in dense assemblages Temporary impacts from increased turbidity Factors Favoring the Use of this Technique Nuisance species are not yet established; target plant patches cover <10 acres Target species are not strongly rooted or prone to fragmentation Plant densities are too high for hand pulling alone Regulatory Requirements May need State and/or local Wetlands Protection permit May need permit associated with Endangered Species Act from U.S. Fish and Wildlife, State, or local authority 85

106 Cost Considerations Costs for augmentation of hand pulling through suction harvesting are estimated at >$5,000/acre, with some estimates approaching $15,000/acre. This may be worthwhile for small areas, but will limit the utility of this technique on a lake-wide basis. Mechanical Harvesting How it Works Mechanical harvesting is most often associated with large machines on pontoons that cut and collect vegetation, but encompasses a range of techniques from simply cutting the vegetation in place to cutting, collecting, and grinding the plants, to collection and disposal outside the reservoir. In its simplest form, cutting, a blade of some kind is applied to plants, severing the active apical meristem (location of growth) and possibly much more of the plant from the remaining rooted portion. Regrowth is expected, and in some species regrowth is so rapid that it negates the benefits of the cutting in only a few weeks. If the plant can be cut close enough to the bottom, or repeatedly, it will sometimes die, but this is more the exception than the rule. Cutting is defined here as an operation that does not involve collecting the plants once they are cut, so impacts to dissolved oxygen and nutrient release are possible in large-scale cutting operations. Advanced technology cutting techniques involve the use of mechanized barges normally associated with harvesting operations, in which plants are collected for out-of-lake disposal. In its use as a cutting technology, the harvester cuts the plants but does not collect them. A modification in this technique employs a grinding apparatus that ensures that viable plant fragments are minimized after processing. There is a distinct potential for dissolved oxygen impacts and nutrient release as the plant biomass decays, much like what would be expected from many herbicide treatments. Harvesting may involve collection in nets or small boats towed by the person cutting the plants, or can employ smaller boat-mounted cutting tools that haul the cut biomass into the boat for eventual disposal on land. It can also be accomplished with larger, commercial machines with numerous blades, a conveyor system, and a substantial storage area for cut plants. Offloading accessories are available, allowing easy transfer of plants from the harvester to trucks that haul the plants to a composting area. Choice of equipment is really a question of scale, with larger harvesting operations usually employing commercially manufactured machines built to specifications suited to the job. Some organizations choose to purchase and operate harvesters, while others prefer to contract harvesting services to a firm that specializes in lake management efforts. Cutting rates for commercial harvesters tend to range from about 0.2 to 0.6 acres per hour, depending on machine size and operator ability, but the range of possible rates is larger and is often dependent upon distance to the offloading location when out-of-lake disposal is planned. Even at the highest conceivable rate, harvesting is a slow process that may leave some lake users dissatisfied with progress in controlling aquatic plants. Plant disposal is not usually a problem, in part because area farmers will often use the plants as mulch and fertilizer, towns often provide mulching facilities for vegetative wastes, and land for disposal is often available near the reservoir. Also, since aquatic plants are about 90 percent water, their dry bulk is comparatively small. Key issues in choosing a harvester include depth of operation, volume and weight of 86

107 plants that can be stored, reliability and ease of maintenance, along with a host of details regarding the hydraulic system and other mechanical design features. Effectiveness Benefits Detriments Effective temporary relief; regrowth may be rapid Cutting into soft sediments and destroying roots can provide longer control Frequent and continued harvesting may gradually alter the plant assemblage Clears target area of plant biomass to selected depth (usually up to 7 ft) Does not kill most plants through single cutting Repeated harvest can control seed producing species Harvesting at the sediment level may disrupt plants and provide greater longevity of results or a shift to more desirable species Minimally selective; only depth of harvest is controlled, Although this may be adequate to favor desirable low-growing plants, it may also open areas for colonization by invasive species May not control species that propagate vegetatively, and may help expand their populations Regrowth may overrun ability of harvester to keep target area clear of plants in larger areas Harvesting with collection tends to collect many small fish and other aquatic life forms Cutting without collection may affect water quality through plant decay Impacts to Non-Target Species Decreased oxygen levels from plant fragment decomposition may impact non-target biota Non-selective technique; can harm many non-target plant species Collection of plant matter with cutting will remove non-target organisms (fish, invertebrates and non-target plants) Reduced cover for fish and invertebrates Factors Favoring the Use of this Technique The reservoir is dominated by undesirable annual species that propagate by seeds Overall density of macrophytes is excessive throughout the littoral zone Surficial and underwater obstructions in targeted areas are minimal Suspended sediments resettle quickly and leave minimal residual turbidity Access for equipment and trucks and a nearby location for plant disposal are available 87

108 Regulatory Requirements May need State and/or local Wetlands Protection permit May need permit associated with Endangered Species Act from U.S. Fish and Wildlife, State, or local authority Cost Considerations Commercial harvesting costs vary depending on the target plant(s), the density of growth, travel distance for disposal of harvested plants and the number of obstructions present. The harvesting cost per acre usually ranges from $350 to $550, excluding trucking and disposal. An exception to this range is very dense growths, such as water chestnut (Trapa natans), where costs have ranged from $1,000 to $1,500/acre. Lower costs are possible where cut vegetation is left in the reservoir. The cost per acre of harvesting is inversely proportional to the size of the area harvested; there is an efficiency of scale for larger projects. A cost range of $200 to $600 per acre for mechanical harvesting at typical densities and $1,000 to $2,000 per acre for very high densities of plants is suggested, including disposal in a nearby location. Rotovation How it Works Rotovation is basically the application of an underwater rototiller to an area of sediment, typically one with dense growths of an unwanted rooted aquatic plant. A rotovator is a hydraulically operated tillage device mounted on a barge. The tiller can be lowered to depths of 10 to 12 feet for the purpose of tearing up roots. On a much simpler scale, cultivation equipment or even old bed springs pulled behind tractors can accomplish much root disturbance. Rototilling and the use of cultivation equipment are highly disruptive procedures normally applied on a small scale. Rotovation has a limited track record, mostly in British Columbia. Use of a variety of cultivation equipment has been practiced in New England for many years, but is rarely documented. Potential impacts to non-target organisms and water quality are substantial, but where severe plant infestations exist, this technique could be appropriate. Effectiveness Benefits Detriments Effective short term control of rooting plants in soft sediment Long term control is not likely without repeated rotovation Disrupts the entire plant, especially the roots Very disruptive in areas applied; may generate high turbidity and drastically alter habitat 88

109 May spread plants that reproduce by fragmentation Decay of damaged plants may affect water quality Impacts to Non-Target Species Benthic fauna are disturbed and possibly buried Decreased oxygen levels from plant fragment decomposition may impact non-target biota Non-selective technique; may harm non-target plant species Reduced cover for fish and invertebrates Temporary impacts from increased turbidity Factors Favoring the Use of this Technique The target species depends on root masses for expansion or overwintering but occupies only a small part of the reservoir Underwater obstructions are minimal Suspended sediments resettle quickly and leave minimal residual turbidity or oxygen demand Regulatory Requirements May need State and/or local Wetlands Protection permit May need permit associated with Endangered Species Act from U.S. Fish and Wildlife, State, or local authority Cost Considerations Rotovating costs have been on the order of $500 to $2,000/acre. Use of cultivation equipment has not been documented, but may be very inexpensive if readily available materials are used and access for pulling equipment is available. Hydroraking How it Works Hydroraking involves the equivalent of a floating backhoe, usually outfitted with a york rake that looks like a farm implement for tilling or moving silage. The tines of the rake attachment are moved through the sediment, ripping out thick root masses and associated sediment and debris. A hydrorake can be a very effective tool for removing submerged stumps, water lily root masses, or floating islands. Use of a hydrorake is not a delicate operation, however, and will create substantial turbidity and plant fragments. Hydroraking in combination with a harvester can remove most forms of vegetation encountered in reservoirs. Hydroraking is effective in the short-term in that it removes plants immediately. It is not an especially thorough or selective technique, and is therefore not well suited to submergent species that can reroot from fragments (e.g., milfoil) or mixed assemblages with desirable 89

110 species present at substantial densities. It is particularly effective for water lilies (white or yellow) and other species with dense root masses. Hydroraking is also often used to remove subsurface obstructions such as stumps or logs. Hydroraking can kill and remove some benthic invertebrates during operation, and nontarget plants will also be impacted in treated areas. High turbidity is usually generated by hydroraking operations, as there is extensive sediment disturbance. This technique is applied on a very limited areal scale in the vast majority of cases, however, and is not expected to have a lakewide effect on non-target organisms or water quality. Effectiveness Benefits Detriments Effective short term control of plants Best used against species with large root masses like water lilies May provide 2-5 years of relief where appropriately applied Removes vegetation difficult to harvest by other means Allows removal of stumps or other obstructions Very disruptive in areas applied; may generate high turbidity and drastically alter habitat May spread plants that reproduce by fragmentation Impacts to Non-Target Species Benthic fauna can be killed or removed Non-selective technique; may harm non-target plant species Factors Favoring the Use of this Technique The target species has dense root masses but occupies only a small part of the reservoir Underwater obstructions are targeted for removal Suspended sediments resettle quickly and leave minimal residual turbidity or oxygen demand Access for equipment and trucks and a nearby location for plant disposal are available. Regulatory Requirements May need State and/or local Wetlands Protection permit May need permit associated with Endangered Species Act from U.S. Fish and Wildlife, State, or local authority 90

111 Cost Considerations Costs for hydroraking range from about $1,500 to $4,000 per acre for typical submergent operations and $6,000 to $10,000 per acre for emergent growths, large floating mats and dense root masses. This cost may be worthwhile for small areas where other techniques are ineffective, but will limit lakewide feasibility. Water Level Control Control of rooted aquatic plants can be achieved through water level control. Two methods can be used, flooding and drawdown. Flooding, increasing water depth in an effort to achieve light limitation for plant control, is rarely used since water quantity and potential flooding impacts limit the utility of this technique. Drawdown is often used, however, and is described below. How it Works Drawdown is a process whereby the water level is lowered by gravity, pumping or siphoning and held at that reduced level for some period of time, typically several months and usually over the winter. Drawdown can provide control of plant species that overwinter in a vegetative state, and oxidation of sediments may result in lower nutrient levels with adequate flushing. Drawdown also provides flood control and allows access for nearshore clean ups and repairs to structures. The ability to control the water level in a reservoir is affected by area precipitation pattern, system hydrology, reservoir morphometry, and the outlet structure. Withdrawal from water supplies may create a drawdown over the summer and into the fall as a matter of greater withdrawals than natural inputs. The base elevation of the outlet or associated subsurface pipe(s) will usually set the maximum drawdown level, while the capacity of the outlet (to downstream) or intake (for water supply) to pass water and the pattern of water inflow to the reservoir will determine if that base elevation can be achieved and maintained. In some cases, sedimentation of an outlet channel or other obstructions may control the maximum drawdown level, and water suppliers must be mindful of having enough capacity to meet supply needs at all times. Several factors affect the success of drawdown with respect to plant control. While drying of plants during drawdowns may provide some control, the additional impact of freezing can be substantial, making drawdown a more effective strategy during late fall and winter in northern areas. However, a mild winter or one with early and persistent snow may not provide the necessary level of drying and freezing. Water suppliers will want to achieve maximum water level by early spring, minimizing the time at maximum drawdown during winter. The presence of high levels of groundwater seepage into the reservoir may mitigate or negate destructive effects on target submergent species by keeping the area moist and unfrozen. The response of macrophyte species to drawdown is quite variable, but many invasive species, including the milfoils, are impacted. Aside from direct impact on target plants, drawdown can also indirectly and gradually affect the plant community by changing the substrate composition in the drawdown zone. If there is sufficient slope, finer sediments will be transported to deeper waters, leaving behind a coarser substrate. If there is a thick muck layer present in the drawdown zone, there is probably not 91

112 adequate slope to allow its movement. However, where light sediment has accumulated over sand, gravel or rock, repetitive drawdowns can restore the coarse substrate and limit plant growths. This is a typical process in many reservoirs, such that the exposed zone is inhospitable to rooted plant growth and invasions are minimized. The actual conduct of an intentional drawdown for plant control involves facilitating more outflow than inflow for several weeks or months. After the target water level is reached, outflow is roughly matched to inflow to maintain the drawdown for the desired period, usually at least a month and often up to 3 months, usually over the winter months. At a time picked to allow refill before any undesirable spring impacts can occur, outflow is reduced (although it should not be eliminated) and excess inflow causes the water level to rise. In some cases, refill is commenced after an inch or two of ice forms, ripping up plants and bottom material. This extreme disturbance approach has been applied where sediments will not dewater sufficiently to provide the level of freezing and desiccation desired, but impacts have not been studied extensively. Effectiveness Benefits Detriments Drawdown can alter the sediment to be less hospitable to all plant species Drawdown can an effective plant control technique for sensitive submergent species, mainly those that do not reproduce by seeds, by exposure Increases plant species richness in many cases Allows sediment oxidation and compaction, with potential reduction of sediment oxygen demand, sediment volume, and available nutrient content May reduce fine sediments in drawdown zone, creating coarser peripheral substrate Provides protection from ice damage to shoreline and associated structures Facilitates access for shoreline clean up, sediment removal, and structural maintenance Provides flood storage capacity Will not kill seeds or other non-vegetative overwintering propagules, and may stimulate increased seed germination Nutrient release during exposed sediment oxidation may fuel increased algal production if not flushed from system before the next growing season Will reduce available water for supplies, and may impair nearby shallow well production May strand and harm minimally mobile aquatic fauna (such as molluscs) Concentration of fish in smaller volume may harm some populations through predation or oxygen stress particularly in warmer months Fish may not be able to reach spawning areas during drawdown May expose and harm hibernating reptiles and amphibians May restrict access and cover for aquatic mammals and birds 92

113 May limit human access via boat ramps or where peripheral sediments are soft Hydrologically connected wetlands may experience some changes Impacts to Non-Target Species Reduction of water supply during drawdown Facilitation of drawdown resistant species Reduced attraction to waterfowl Altered littoral habitat for fish and invertebrates Potential mortality to hibernating reptiles and amphibians Water loss in hydrologically connected wetlands Erosion of shoreline Possible temporary increase in algae Recreational access impairment during drawdown Possible downstream flow impacts Factors Favoring the Use of this Technique Water supply impairment will not result from drawdown Peripheral domination by undesirable species that are susceptible to drying and freezing Drawdown can be achieved by gravity outflow via an existing outlet or normal withdrawal Drawdown can reach a depth that impacts enough of the targeted plants to make a difference Areas to be exposed have sediments and slopes that promote dewatering Drawdown and refill can be accomplished within a few weeks under typical flow conditions and without causing downstream flows outside the natural range Drawdown can be timed to avoid key migration and spawning periods for non-target organisms Populations of molluscs or other nearshore-dwelling organisms of limited mobility are not significant Flood storage capacity generated by drawdown is needed Shoreline structures are prone to ice damage Regulatory Requirements May need State and/or local Wetlands Protection permit May need permit associated with Endangered Species Act from U.S. Fish and Wildlife, State, or local authority 93

114 Cost Considerations Drawdown is a relatively inexpensive reservoir management technique, if the means to conduct a drawdown are present. Where an outlet or intake structure facilitates drawdown, the cost may be as little as what is required to obtain permits, open and close the discharge structure, and monitor. If pumps are required to lower the water level, the drawdown will be expensive. It is unusual to alter a dam for less than $100,000, but if the structure already supports water level control, costs of $3,000 to $10,000 per year would be a reasonable expectation for permitting and monitoring. Where protected species are present, permitting may be difficult and monitoring and mitigation costs can escalate. CHEMICAL CONTROLS Overview Chemical treatment is one of the oldest methods used to manage nuisance aquatic plants, and is still the most frequently applied approach. Other than perhaps drawdown, few alternatives to herbicides were widely practiced until relatively recently. Herbicides are used less commonly in drinking water supplies, either because of regulatory restrictions or as a function of negative public perception of such use. However, multiple herbicides are approved for use in drinking water reservoirs, with restrictions, and can be an effective tool in the control of aquatic vegetation in general and milfoil specifically. Herbicides contain active ingredients that are toxic to target plants. Mattson et al. (2004) provides a full technical discussion of most of the available herbicides, and both labels and Material Safety Data Sheets are available on-line for commercially available herbicides. Herbicides are typically classified as contact or systemic herbicides based on the action mode of the active ingredient. Contact herbicides are toxic to plants by uptake in the immediate vicinity of external contact, while systemic herbicides are taken up by the plant and are translocated throughout the plant. In general, contact herbicides are more effective against annuals than perennials because they may not kill the roots, allowing perennials (including milfoil species) to grow back. Seeds are also not likely to be affected, but with proper timing and perhaps several treatments, growths can be eliminated much the same way harvesting can eliminate annual plants. Systemic herbicides tend to work more slowly than contact herbicides because they take time to be translocated throughout the plant. Systemic herbicides generally provide more effective control of perennial plants than contact herbicides, as they kill the entire plant under favorable application circumstances. Systemic herbicides will also kill susceptible annual species, but regrowth from seeds is usually substantial. If annual species are the target of control, additional treatment will be required, normally a year after initial treatment and for as long as the seed bank facilitates new growths. Another way to classify herbicides is by whether the active ingredients are selective or broad spectrum. Selective herbicides are more effective on certain plant species than others, with control of that selectivity normally dependent on variability in the features of the plant affected by the herbicide. Plant factors that influence selectivity include plant morphology, physiology and the stage of growth. Even a selective herbicide can kill most plants if applied at high rates. In 94

115 some cases, selectivity can be achieved with a broad spectrum herbicide by altering dose and exposure duration. Herbicides may also contain adjuvants. An adjuvant is any chemical added to the herbicide to increase the effectiveness of the application. There are different classes of adjuvants, which generally function to increase the uptake of the herbicide by the plant, spread the herbicide through the water column, or help the herbicide adhere to the plant. Adjuvants are not expected to be toxic to the target species, but increase the toxicity of the herbicide or otherwise aid herbicide effectiveness. Copper is widely used in water supplies to control algae, and is sometimes used alone as a herbicide, but is more often mixed with other herbicides to increase the effectiveness of those herbicides. Copper tends to rapidly kill algae growing on plants that might interfere with herbicide effectiveness, and may also weaken plant cell walls and increase exposure to herbicides. Aquatic herbicides must be registered by the EPA and many individual state agencies for legal use in the state of application. The criteria addressed in the registration process include data on forms of toxicity, impacts to non-target organisms, environmental persistence, breakdown products and fate of the herbicide constituents in the aquatic environment. The nature and variability in toxicity reporting can create confusion and ambiguity in herbicide evaluations. Risk is a function of toxicity and exposure, and expressions of risk should address both of these key elements. The choice of herbicide to manage an undesirable plant population depends on the properties of the herbicide, the relative sensitivity of the target and non-target plants and other organisms that will be exposed, water use restrictions after herbicide use, and cost. Effectiveness in controlling the target plant species is a critical consideration, although water use restrictions and general public perception will play a major role in herbicide treatments in drinking water supplies. Aside from copper and adjuvents, there are nine active ingredients currently approved for aquatic use by the United States Protection Agency (AERF, 2005). However, not all are approved for use in all states, and two are recent enough additions to have very limited experience upon which to base any practical review of field effectiveness. Furthermore, only five herbicidal ingredients are directly effective against one or more species of milfoil, plus copper as an associated algal periphyton control agent, so the choice of herbicides is limited. Key information for the five herbicides used against milfoil is provided in Table 5-2. The local permitting authority should be contacted before planning any herbicide application to obtain state specific details regarding any additional use restrictions. Herbicide effectiveness may be influenced by such factors as timing, rate and method of application, species present and weather conditions. Additionally, dose determination should consider detention time, morphometry and water hardness to maximize effectiveness. Herbicide treatment can be an effective short-term management procedure to produce a rapid reduction in algae or vascular plants for periods of weeks to months. Although long-term effectiveness of herbicide treatments is possible, in most cases herbicide use is considered a short-term control technique. Herbicides are generally applied seasonally to every four years to achieve effective control. Systemic herbicides, which kill the entire plant including the roots, generally provide results with greater longevity than contact herbicides, which can leave roots alive to regrow. 95

116 The use of herbicides to get a major plant nuisance under control is a valid early step in long-term management, but other means of keeping plant growths under control should be applied. However, failure to apply alternative techniques on a smaller scale once the nuisance has been abated can allow regrowth of problem plants and necessitate repeated herbicide applications. This is unlikely to be an acceptable approach in water supplies; use of herbicides should be focused on reducing nuisance growths to the point at which other techniques can be effectively and economically applied to maintain control. In general, the likelihood of undesirable impacts decreases as the applied concentration decreases. Each herbicide is evaluated individually based on the formulation, and the expected concentration as a function of the percent active ingredient, application rate and depth of water. Concentrations allowed as application rates are much higher than those to which the public would be exposed under normal circumstances. The product may be applied at lower rates, and often is. Granular products are intended to slowly dissolve in water over time and dissipate. Many of the compounds are rapidly removed from the water. Use in accordance with label instructions and restrictions is therefore not expected to result in toxicity to non-target fauna, including humans, other mammals, waterfowl, fish and invertebrates. There have been toxic reactions in rare cases, and the chronic effects of frequent exposure are not truly known in many cases. Additional information for individual herbicidal active ingredients in use today is provided below. Copper is not generally used to directly control milfoil growth, but rather to kill 96

117 Table 5-2. Herbicides used to control milfoil and information relevant to use in potable water supplies. * Potable Water Tolerance as reported by Plant Management in Florida Waters ( This does not reflect individual States water quality standards and is used for relative comparison purposes only 97 Compound Trade Name(s) Company Formulations Mode of Action Complexed Cutrine-Plus Applied Copper Biochemists 2,4-D Algimycin Komeen Nautique Aqua-Kleen Navigate Applied Biochemists Griffin Corp. SePRO Corp. Cerexagri, Inc. Applied Biochemists Various complexing agents with copper DMA liquid BEE salt Plant cell toxicant, weakens cell wall Contact action Selective plantgrowth regulator Systemic action Typical Dose of Active Ingredient ppm ppm Distance to Potable Water Intake None required Distance not specified Potable Water Tolerance Additional Potable * Water Restrictions 1.0 ppm No other restrictions, used mainly for algae control; useful to remove algae encrustations from target plants before use of other herbicides 0.03 ppm Do not use treated water for drinking until the concentration falls below 0.07 ppm 21 day use limitation (continued) 97

118 Table 5-2. (Continued) Compound Trade Name(s) Company Formulations Mode of Action Diquat Reward Syngenta Liquid Disrupts plant cell membrane integrity Weedtrine-D Applied Biochemists Liquid Contact action Typical Dose of Active Ingredient ppm Distance to Potable Water Intake Need to post within ¼ mile (1,320 ft) in non-flowing water, 1,600 ft downstream in flowing water Potable Water Tolerance Additional Potable * Water Restrictions 0.05 ppm Potable water use restrictions based on application rates with 1-3 day use limitation 98 Endothall Fluridone Aquathol K Aquathol Super K Hydrothol 191 Sonar Avast Cerexagri, Inc SePRO Corp. Triclopyr Renovate SePRO Corp. Liquid or granular Liquid or granular Liquid Inactivates plant protein synthesis Contact action Disrupts carotenoid synthesis, causing bleaching of chlorophyll Systemic action Selective plant growth regulator Systemic action ppm ppm ppm Not specific ppm Need to restrict dosage to <0.020 ppm within ¼ mile (1,320 ft) of potable water intake Variable setback distances associated with application rates from ft Potable water use restrictions based on concentrations ranging from ppm with 7-25 day use limitation ppm May be applied <0.020 ppm where intakes are present 0.40 ppm Intake must be taken off-line or until concentration is < 0.40 ppm 98

119 associated periphyton that might interfere with the effectiveness of treatment with other herbicides. Copper is therefore not addressed further. Diquat Diquat is a fast acting contact herbicide, producing results within 2 weeks of application through disruption of photosynthesis. It is a broad-spectrum herbicide with potential risks to aquatic fauna, but laboratory indications of invertebrate toxicity have not been clearly documented in the field. A domestic water use restriction of 3 days is normally applied. Irrigation restrictions of 2 to 5 days are applied, depending on dose and crop to be irrigated. Regrowth of some species has been rapid (often within the same year) after treatment with diquat, but two years of control have been achieved in some instances. Diquat is used as a general purpose aquatic herbicide, both as a primary control agent for a broad range of macrophytes and as a follow-up treatment chemical for control of plants (especially milfoil) missed by other herbicides or physical control techniques, on a localized (spot) basis. Treatment with diquat is recommended early in the season to impact early growth stages, but can be applied any time. Diquat is less effective in turbid, muddy water due to adsorbance onto sediments and other particles. Diquat treatments typically cost $200 to $500 per acre. Since diquat is a broad spectrum herbicide, it can be expected to impact non-target plants when they are present. Loss of vegetative cover may have some impact on aquatic animals, but short-term effects are not expected. The acute toxicity of diquat for fish is highly variable depending on species, age, and hardness of water. Young fish are more sensitive than older fish. Toxicity is decreased as water hardness increases. As the potable water tolerance for diquat is lower than normally applied doses and multiple day use restrictions are required, diquat is not often used in drinking water reservoirs, but might be applied for spot treatments in areas remote from the intake area. Endothall Endothall is a contact herbicide, attacking a wide range of plants. The method of action of endothall is suspected to inhibit the use of oxygen for respiration. Only portions of the plant with which the herbicide can come into contact are killed. There are two forms of the active ingredient; the inorganic potassium salt that is found in the products Aquathol Granular and Aquathol K and the alkylamine salt formulations of Hydrothol 191 Granular and Hydrothol 191. Effective control can range from weeks to months. Most endothall compounds break down readily and are not persistent in the aquatic environment, disappearing from the water column in under 10 days and from the sediments in under 3 weeks. Endothall acts quickly on susceptible plants, but does not kill roots with which it cannot come into contact, and recovery of many plants occurs. Rapid death of susceptible plants can cause oxygen depletion if decomposition exceeds re-aeration in the treated area, but this can be mitigated by conducting successive partial treatments. Toxicity to invertebrates, fish or humans is possible but not expected at typical doses. The potable water tolerance for endothall matches the range of normally applied doses, but longer duration use restrictions preclude use in most drinking water reservoirs. Endothall is sometimes applied for spot treatments in areas remote from any intake. Endothall treatments typically cost $400 to $700 per acre. 99

120 2,4-D 2,4-D, the active ingredient in a variety of commercial herbicide products, has been in use for over 30 years. This is a systemic herbicide; it is absorbed by roots, leaves and shoots and disrupts cell division throughout the plant. Vegetative propagules such as winter buds, if not connected to the circulatory system of the plant at the time of treatment, are generally unaffected and can grow into new plants. Seeds are also not affected. It is therefore important to treat plants early in the season, after growth has become active but before such propagules form. 2,4-D is sold in liquid or granular forms as sodium and potassium salts, as ammonia or amine salts, and as an ester. Lower doses are more selective but require more contact time; a range of one to three days of contact time is typically needed at the range of doses normally applied. 2,4-D has a short persistence in water but can be detected in the mud for months. The BEE granular formula is easy to apply for spot treatments and the active ingredient is slowly released near the root zone. The BEE form is typically more toxic to both plants and fish than the DMA salt, but toxicity is rarely observed at normal application rates of any formulation. Experience with granular 2,4-D in the control of nuisance macrophytes has generally been positive, with careful dosage management providing control of such non-native nuisance species as Eurasian water milfoil with only sublethal damage to many native species. 2,4-D is the most effective herbicide for all Category I milfoil species. However, the 2,4-D label does not permit use of this herbicide in water used for drinking, other domestic purposes, or irrigation until the concentration is less than 0.1 ppm, much less than normally applied doses. Except for spot treatments well away from intakes, this limits use in water supply reservoirs. 2,4-D treatments typically cost $300 to $800 per acre. Fluridone Fluridone is a systemic herbicide that comes in two general formulations, an aqueous suspension and a slow release pellet, although several forms of pellets are now on the market. This chemical inhibits carotene synthesis, which in turn exposes the chlorophyll to photodegradation. Most plants can be damaged by sunlight in the absence of protective carotenes, resulting in chlorosis of tissue and death of the entire plant with prolonged exposure to a sufficient concentration of fluridone. When carotene is absent the plant is unable to produce the carbohydrates necessary to sustain life. Some plants, including Eurasian water milfoil, are more sensitive to fluridone than others, allowing selective control at low doses. Control of variable water milfoil and parrotfeather, the other two Category I milfoils, is less reliable with fluridone and requires higher doses. For susceptible plants, lethal effects are expressed slowly in response to treatment with fluridone. Existing carotenes must degrade and chlorosis must set in before plants die off; this takes several weeks to several months, with days given as the observed range of time for die off to occur after treatment. The slow rate of plant die-off minimizes the risk of oxygen depletion. Fluridone concentrations should be maintained in the lethal range for the target species for at least 6 weeks, preferably 9 weeks or longer. This presents some difficulty for treatment in areas of substantial water exchange and limits use in spot treatments unless a curtain or other sequestration approach can be applied. Treatment protocols have been developed to address a variety of field conditions, including single or sequential treatments, sequestered applications, continuous inlet inputs, and 100

121 application at various times of the year. Early treatment (April/early May) with fluridone effectively controls overwintering perennials like the milfoils before some of the beneficial species of pondweed and naiad begin to grow. Variability in response has also been observed as a function of dose, with lower doses causing less impact on non-target species. However, lesser impact on target plants has also been noted in some cases, so dose selection involves balancing risk of failure to control target plants with risk of impact to non-target species. Fluridone is considered to have low toxicity to invertebrates, fish, other aquatic wildlife, and mammals, including humans. The USEPA has set a tolerance limit of 0.15 ppm for fluridone or its degradation products in potable water supplies. While some state restrictions are lower, typically applied doses are allowable in the vast majority of cases. Public perception issues may limit whole-lake treatments in drinking water reservoirs, but such treatments are acceptable in the regulatory framework. Fluridone treatments typically cost $500 to $1000 per acre for single treatments with the liquid form. Costs rise to $1000 to $2000 per acre for sequential treatments. For partial treatments in which a portion of the reservoir is sequestered, an additional cost of about $10 to $20 per linear foot of sequestering curtain is to be expected. The cost of application with the pelletized form is usually $800 to $1200 per acre. Triclopyr Triclopyr is a systemic herbicide. Its mode of action is to stimulate growth (auxin mimic) while preventing synthesis of essential plant enzymes, resulting in disruption of growth processes. Uptake of the herbicide generally occurs within 6 to 12 hours of exposure, with effects becoming observable in about a day. Plants sink from the reservoir surface in 3 to 5 days and death of the plant is complete in 1 to 3 weeks. Various studies have shown triclopyr to be an effective herbicide for control of certain macrophytes. It is highly selective and effective against water milfoil and other dicotyledonous plants. Effectiveness increases as both concentration and exposure time increase. Treatments to date have revealed little or no effect on most monocotyledonous naiads and pondweeds, which are mostly valued native species. This herbicide is most effective when applied during the active growth phase of young plants. The half-life for triclopyr can range from 12 hours to 29 days, with degradation into nontoxic forms. Lethal or chronic effects on fauna have not been observed within the recommended dosage range. It should be noted that one of the terrestrial formulations, Garlon 4, has much higher toxicity, but aquatic formulations are derived from the chemical family of the less toxic Garlon 3A (triethylamines). Triclopyr can be applied to waters used as potable water supply, but with a setback distance from any functioning intake that is determined by dose and size of the area treated. There are no federal label restrictions for recreational use of treated waters or for use in livestock watering. Crop irrigation use is prohibited for 120 days or until the triclopyr concentration is undetectable by immunoassay testing. There is no restriction on use for irrigating established turfgrass. Triclopyr treatments are expected to cost $600 to $800 per acre for single treatments with the liquid form, but there is limited experience upon which to base cost estimates. 101

122 BIOLOGICAL CONTROLS Overview Any introduction of organisms may have impacts on the aquatic community structure and food web, however imperceptible. Greater impact occurs when the introduced species becomes abundant or affects another species that is or was abundant. Understanding the nature of these interactions can allow manipulation of system biology to produce a desired effect. However, the biggest pitfall of biomanipulation is that we seldom fully understand all of the relevant interactions. Interest has grown in biological control methods over the last two to three decades. Most methods are still experimental and have a limited degree of achieved effectiveness. Most methods have the potential to inflict negative impacts on the environment. Biological methods differ from other plant control methods in that there are more variables to consider and usually a longer time span needed to evaluate effectiveness. These methods are unusual in that the treatments consist of either altering conditions to favor certain organisms or introducing live organisms that may be difficult or impossible to control or recall once introduced. For this reason non-native introductions are restricted in most cases. Biological control has the advantage that it is perceived as a more natural or organic plant control option, but it still represents human interference within an ecological system. The potential for long-term effectiveness with limited maintenance is attractive, but has been largely elusive with biological controls. Herbivorous Fish How it Works There are several species of fish that consume macrophytes, including the African cichlids (Tilapia spp.). Tilapia species can only survive in water temperatures greater than 10 C and are therefore unlikely candidates for macrophyte control in colder climates. The introduction of herbivorous fish therefore generally centers on grass carp (Ctenopharyngodon idella). Grass carp are not approved for introduction in many states, however, so local restrictions should be investigated if such an introduction is being considered. The grass carp (Ctenopharyngodon idella), also known as the white amur, has a native range that includes the Pacific slope of Asia. They are typically found in low gradient reaches of large river systems. Grass carp can grow to 4 feet long and attain weights of over 100 pounds, making them the largest member of the cyprinid family. They have a very high growth rate, with a maximum at about 6 pounds per year. They typically grow to a size of pounds in North American waters and have adapted quite well to life in reservoirs where they are stocked for aquatic vegetation control. As with other carp species, they are tolerant of wide fluctuations in water quality including water temperatures from 0 to 35 C, salinities up to 10 ppt, and oxygen concentrations approaching 0 mg/l. Grass carp do not feed when water temperatures drop below 11 C (52 F) and feed heavily when water temperatures are between 20 C and 30 C (68 F and 86 F). Dietary preference is an important aspect of grass carp, as pertains to their use as a plant control mechanism. Grass carp have exhibited a wide variety of food choices from study to study. In 102

123 some cases grass carp have been reported to have a low feeding preference for milfoils, but in some recent studies milfoils were consumed. The majority of grass carp currently stocked in North America are sterile triploids. Fish are usually stocked in the size range of 200 mm to 300 mm (8 to 12 inch). The most common stocking rates are at 80 to 100 fish per acre for plant eradication and 12 to 80 fish per acre for plant control, although stable control is not typical. The major difficulty in using grass carp to control aquatic plants is determining what rate will be effective and yet not so high as to eradicate the plants completely. The fish usually live ten or more years but the typical plant control period is reported to be 3 to 4 years with some restocking often required. They are difficult to capture and remove unless the reservoir is treated with rotenone that will kill other fish species as well and is unlikely to be allowable in drinking water supplies. Algal blooms resulting from nutrients being converted from plant biomass by the grass carp have been common, even without elimination of vascular plants. Effectiveness Benefits Detriments Highly dependant on stocking rate; impact observable from one to several years after stocking Controls new growth better than established mature vegetation Grass carp control option of water milfoil has been less reliable than for other submergent species Potential control of aquatic plants from a single introduction of an appropriate density of fish for perhaps 5 years Grass carp can decimate native plant communities, resulting in severe impacts to waterfowl, invertebrate, and fish habitats Grass carp stocking can result in major impacts to water quality, including algae blooms, increased turbidity, and fluctuating dissolved oxygen and ph Grass carp are highly migratory and can easily escape over spillways or through bar grates to impact waters other than those intended; escape barriers may be required and may interfere with reservoir operation The impacts and effectiveness of grass carp are highly variable and unpredictable Impacts to Non-Target Species Loss or eradication of non-target plant species Increase in algal biomass; change in algal community structure Change in fish community structure 103

124 Factors Favoring the Use of this Technique Target species (especially milfoils) are so abundant that they are likely to be consumed Undesirable algal growths are not expected, despite increased nutrient availability (factors such as light, mixing or flushing control algal biomass or the increase in nutrient availability is somehow small) Regulatory Requirements May need State and/or local Wetlands Protection permit May need stocking permit from Fish and Wildlife authorities May need permit associated with Endangered Species Act from U.S. Fish and Wildlife, State, or local authority Grass carp may be illegal for use; check specific state requirements Cost Considerations Costs for 8-10 inch grass carp vary widely between $4 and $13 depending on the source. At stocking rates of 12 to 100 fish per acre this would amount to a cost of $48 - $1300 per acre, and the treatment effects typically last about five years. A rather wide cost range of $100 to $1500/acre for grass carp stocking might be expected, including planning and monitoring. Herbivorous Invertebrates How it Works Significant improvement in our future ability to achieve lasting control of nuisance aquatic vegetation may come from plant-eating biocontrol organisms, or from a combination of current procedures such as harvesting, drawdown, and herbicides with these organisms. Biological control has the objective of achieving control of plants without introducing toxic chemicals or using machinery. Yet it suffers from an ecological drawback; in predator-prey (or parasite-host) relationships, it is rare for the predator to completely eliminate the prey. Consequently, population cycles or oscillations are typically induced for both predator and prey. It is not certain that the magnitude of the upside oscillations in plant populations will be acceptable to human users, and it seems likely that a combination of other techniques with biocontrols may be necessary to achieve lasting, predictable results. Biological control using invertebrates (mainly insects) from the same region as the introduced target plant species include the root boring weevil (Hylobius transversovittatus) and two leaf beetles (Galerucella calmariensis and G. pusilla) for the control of purple loosestrife (Lythrum salicaria). Such efforts also include the tuber feeding weevil (Bagous affinis) and the leaf-mining fly (Hydrellia pakistanae), both for the control of Hydrilla verticillata in Florida. However, as introduced non-native species have sometimes caused bigger problems than they solved, native species are preferred. Augmentation of a native insect population has been studied with the milfoil midge (Cricotopus myriophylli), a moth (Acentria ephemerella) and the milfoil weevil (Euhrychiopsis lecontei). Releases of the native weevil (Euhrychiopsis lecontei) for the control of Eurasian 104

125 milfoil have occurred since 1995, with variable results, but this is the only commercially available biocontrol agent for any milfoil species. Euhrychiopsis lecontei is a native North American insect species believed to have been associated with northern water milfoil (Myriophyllum sibericum), a species largely replaced by non-native, Eurasian water milfoil (M. spicatum) since the 1940s. It does not utilize non-milfoil species. In controlled trials, the weevil clearly has the ability to impact milfoil plants through structural damage to apical meristems (growth points) and basal stems (plant support). Adults and larvae feed on milfoil, eggs are laid on it, and pupation occurs in burrows in the stem. Field observations linked the weevil to natural milfoil declines in nine Vermont lakes and additional lakes in other states. Lakewide crashes of milfoil populations have generally not been observed in cases where the weevil has been introduced into only part of the lake, although localized damage has been substantial. Widespread control may require more time than current research and monitoring has allowed. As with experience with introduced insect species in the south, the population growth rate of the weevil is usually slower than that of its host plant, necessitating supplemental stocking of weevils for more immediate results. Just what allows the weevil to overtake the milfoil population in the cases where natural control has been observed is still unknown. It is known that fish (especially sunfish) prey heavily on the weevils, limiting their effectiveness in many cases. Densities of 1-3 weevils per stem appear to collapse milfoil plants, and raising the necessary weevils is a major operation. Weevils are marketed commercially as a milfoil control, with a recommended stocking rate of 3000 adults per acre. Release is often from cages or onto individual stems; early research involved attaching a stem fragment with a weevil from the lab onto a milfoil plant in the target lake, a highly labor-intensive endeavor. Although weevils may be amenable to use within an integrated milfoil management approach, interference from competing control techniques has been suggested as a cause for suboptimal control by weevils. Harvesting may directly remove weevils and reduce their density during the growing season. Also, adults are believed to overwinter in debris along the edge of the lake, and techniques such as drawdown, bottom barriers, or sediment removal could negatively impact the weevil population. Weevils could be stocked in drinking water supplies with no adverse impact to water quality, but this technique has not yet proven reliably effective. Crayfish have been documented to control vegetation (Lodge and Lorman, 1987; Olsen et al., 1991), but the most effective species (Orconectes rusticus) is considered invasive, can destroy all vegetation, and is not recommended for transfer into new waters. Additionally, even native species tend to impact the benthic fauna and may upset the existing food web, possibly with undesirable consequences for water supplies. Effectiveness Benefits Weevils have exhibited highly variable effectiveness, and only for Eurasian water milfoil Some other control agents for non-milfoils have demonstrated effectiveness Potential control with native (or carefully researched and approved non-native) species that may be self-perpetuating 105

126 Detriments Harnesses natural processes to control nuisance or invasive species High variability in results; not especially reliable Generally slow in achieving desired results Impacts to Non-Target Species No non-target impacts observed within the year of treatment and no long term impacts are expected Factors Favoring the Use of this Technique The biocontrol agent is a native species that is highly host-specific for the target species. Relationships between the introduced species and the reservoir are understood from studies at other reservoirs. The biocontrol agent can be removed from the reservoir if necessary, or has limited powers of reproduction, migration, or longevity. Small-scale field tests can be run to examine likely effectiveness and non-target impacts before moving to full-scale introduction. Regulatory Requirements May need State and/or local Wetlands Protection permit May need permit associated with Endangered Species Act from U.S. Fish and Wildlife, State, or local authority Cost Considerations Milfoil weevils are sold for $1 each, with a recommended stocking density of 3,000 per acre. Several stocking events over a 3 5 year period may be needed. No other biological agents for milfoil control are currently readily available. Plant Competition How it Works Although invasive nuisance plant species are just what the name implies, there is evidence that the presence of a healthy, desirable plant community can minimize or slow infestation rates. Most invasive species (and definitely the problem milfoils) are favored by disturbance, so a stable plant community should provide a significant defense. Unfortunately, natural disturbances abound, and almost all common plant control techniques constitute disturbances. Therefore, if native and desirable species are to regain dominance after disturbance, it may be necessary to supplement their natural dissemination and growth with 106

127 seeding and planting. The use of seeding or planting of vegetation is still a highly experimental procedure, but if native species are employed, it should yield minimal controversy. Experiments indicate that the addition of dried seeds to an exposed area of sediment will result in rapid germination of virtually all viable seeds and rapid cover of the previously exposed area. However, if this is not done early enough in the growing season to allow annual plants to mature and produce seeds of their own, the population will not sustain itself into the second growing season. Transplanting mature growths into exposed areas has generally been found to be a more successful means of establishing a seed producing population. More research is needed, but establishment of desired vegetation is entirely consistent with the primary plant management axiom: if light and substrate are adequate, plants will grow. Rooted plant control should extend beyond the limitation of undesirable species to the encouragement of desirable plants. The basic premise of plant competition as a management technique is therefore to maximize spatial resource use by desirable species to keep out undesirable plants. Effectiveness Benefits Detriments Not an effective short term control Potentially effective at slowing the invasion of water milfoil if dense coverage of nonnuisance plant species is established Harnesses natural processes to develop desired conditions May be self-perpetuating Augments other techniques for plant control May not prevent invasions over a long time period Requires ongoing effort to keep up with natural disturbances Indigenous species may become nuisances in some cases Likely to require application of a major control technique prior to planting Impacts to Non-Target Species No substantial impacts are expected if native vegetation is used, but there are insufficient data to evaluate possible impacts Factors Favoring the Use of this Technique Portions of the reservoir support healthy growths of desirable species Other sources of desirable species are available Control methods for undesirable species are applicable and supported 107

128 Regulatory Requirements May need State and/or local Wetlands Protection permit May need permit associated with Endangered Species Act from U.S. Fish and Wildlife, State, or local authority Cost Considerations Costs are largely unknown and unpredictable, but will be determined by the need for removal of unwanted vegetation, the source of desired vegetation, planting method, and monitoring approach. Most costs will be associated with labor. Based on limited experience, a cost of $5000 to $10,000 per acre is suggested, but more efficient and lower cost methods will be needed to make this approach practical. NO MANAGEMENT ALTERNATIVE Overview The no management alternative for nuisance aquatic plants in general and invasive milfoil species in particular would exclude all active reservoir management programs, but would include normal monitoring and any normal operations such as drawdowns. The no management alternative would provide neither prevention nor remediation efforts other than those required by current laws; although there is federal precedent upon which to base strong statutes, only some states have provisions intended to stop the spread of invasive species or preserve the desirable features of reservoirs. The normal tendency for reservoirs is to gradually accumulate sediments and associated nutrients and to generally become more productive. In cases where there is development in the watershed leading to increased erosion and sediment transport to the reservoir, the rate of infilling and expansion of macrophyte beds would be expected to increase more rapidly. Impacts to drinking water reservoirs associated with excessive rooted plant densities, notably water milfoil, are extensive and varied (see section 4.0 Potential Issues for Drinking water Reservoirs). It is expected that these impacts would be exacerbated with no management. Yet many reservoirs have no apparent problems with milfoil or other plants, so the no management alternative is not necessarily a guarantee of adverse impact, although it does constitute an invitation to problems. Reservoirs with steep sides, rocky peripheral substrates, substantially fluctuating water levels, or especially turbid water may simply not have problems with plants or invasive species, including milfoils. Where a problem milfoil species has become established, but the potential area of colonization is small, it may not be necessary to take any action to prevent problems relating to drinking water supply. The no action alternative could actually be a viable strategy, depending upon local conditions. It is essential that those conditions be understood, however, and in advance of any invasion, to determine if a lack of action will have potential impacts that could be far more expensive to mitigate at a future date. 108

129 Effectiveness Benefits Detriments Lack of management may allow invasive plants such as milfoil to reach maximum coverage and density, but just what those maxima will be is reservoir-specific No financial expenditures on management if plant densities are naturally controlled Possible increased plant density Possible decreased plant diversity Possible alteration of water quality Possible decreased recreational utility Impacts to Non-Target Species Taking not action will not directly impact non-target species, but growth of invasive species may have major impacts on flora, fauna, and human uses Factors Favoring the Use of this Technique The reservoir is in an acceptable condition for designated and desired uses and is expected to remain so There are no apparent threats to reservoir condition. Existing compliance with all federal and state laws relating to vascular plants and algae Regulatory Requirements There are no regulations associated with the no management alternative Cost Considerations Direct costs do not apply to the no management alternative, although there may be hidden or opportunity costs associated with the impacts to non-target organisms and water quality. Such costs are difficult to estimate and would vary on a case by case basis. Substantial costs for water treatment may be associated with a lack of action if invasive milfoil becomes abundant over a large portion of the reservoir or near the intake. Intake screens may be clogged on a more frequent basis and of the risk of disinfectant by-products may increase from dissolved organic carbon increases. 109

130 110

131 CHAPTER 6 CONSIDERATIONS FOR APPLICATION OF TECHNIQUES TO DRINKING WATER SUPPLIES RESERVOIR CHARACTERISTICS Morphometry and Bathymetry The general shape (morphometry) and depth characteristics (bathymetry) of a water supply reservoir may determine whether control of milfoil is necessary, and, if so, will affect the feasibility of control options. Depth contours will determine what portion of the reservoir might be colonized by nuisance milfoil species, and the shape will often influence the rate and pattern of spread. Large, deep reservoirs created by damming river valleys tend to have steep slopes and a central channel with multiple side arms and smaller inlets, and will be less susceptible to lakewide colonization than less dendritic waterbodies with lesser side slopes and less inlet to outlet orientation of flow. At the same time, reservoirs created by damming rivers usually have larger watersheds and higher flows, creating the potential for rapid spread of invasive milfoils along the axis of flow. On a local level, shallow coves and backwaters are more susceptible to colonization if exposed to invasive milfoils, but may be more amenable to rapid response and successful eradication or control. Where water levels fluctuate greatly, milfoils will have more difficulty becoming established. Sediment Features Potential problem milfoils can grow on a wide range of substrates, but reach maximum densities on a mixture of sand and organic silts (sandy muck sediments). Underwater slope, water level fluctuation, currents, external sediment and organic matter inputs, and internal productivity over a period of decades will determine sediment properties. Slope is probably the most important determinant of substrate particle size in most reservoirs, with steep sided waterbodies having the least accumulation of fine sands and silts in water shallow enough for milfoil to grow. Nature and Depth of Intake Intakes almost always have some kind of attached grate, to prevent entrainment of large debris that could damage pumps and transport piping between the source and the treatment or distribution facility. Finer screens are often used to minimize the intake of smaller debris and biota, both for the protection of biota and the treatment/distribution system. If the intake structure allows the passage of milfoil plant fragments, control to prevent adverse impacts on water quality may be necessary. If the intake is screened to prevent plant intake, anti-clogging actions may be needed. The depth of intake will affect the interaction of milfoil with the intake, with deeper (below the depth of plant growth) intakes having less potential to entrain milfoil or interact with water conditioned by milfoil. However, deeper waters are often thermally stratified and anoxic for part of the year; this tends to favor use of a shallower intake during the summer, when milfoil growths and potential effects are at a maximum. 111

132 Pattern of Water Withdrawal For a reservoir which is nearly always on-line, conditions created by milfoil must be addressed, either in the reservoir or through the treatment process. However, where multiple reservoirs are used with some temporal variability, it may be possible to avoid many potential milfoil effects by changing water sources. Whether this is practical will depend on the ability of alternative water sources to meet demand during the summer and early fall period of milfoil dominance and maximum potential effects. Additional Water Uses Although potable water supply is generally considered to be the most sensitive use when planning any plant control activity, other water uses may affect the choice of management action for milfoil in a drinking water supply. Where the resource is not entirely controlled by the water utility, or use is shared by agreement, the presence of swimming and boating facilities may restrict the choice of options or the location of application. Additionally, use of the reservoir by protected species may carry substantial regulatory weight where management actions require permits. In general, reduction of nuisance milfoils will be considered desirable for all uses, but the method by which that reduction is achieved may be restricted. One particularly contentious area is between water supply and fishing uses, as loss of plant cover of any kind is often viewed as a negative influence on the fish community and fishing success (Horton, 2005). While the utility may not be concerned over what species replaces milfoil upon control, anglers may have a strong interest and may organize to oppose plant control unless there is a clear plan that addresses their interests. Summary of Key Reservoir Features Affecting Management of Milfoil Reservoir features that minimize the need for control of milfoil: Small area with water depth 2-15 ft Water level that fluctuates by 10 ft or more Very coarse or very mucky sediment beneath water of 2-15 ft depth Reservoir not used as a primary source in summer and fall Intake is very far from milfoil bed with high dilution potential in between Reservoir features that tend to restrict which techniques can be applied: Shallow intake near targeted milfoil eliminates most herbicide use, increases difficulty of other techniques without impacting water supply Short detention time affects herbicide effectiveness and downstream impact potential Long detention time limits winter drawdown (may not refill fast enough to meet demand) Other uses (especially fishing) with high priority non-target impacts may be critical 112

133 AQUATIC PLANT COMMUNITY Invasive Milfoil Species Ecology It is likely, although not inevitable, that an invasive milfoil will become the dominant plant in a reservoir which it invades over a period of five to ten years. Invasive milfoils share some common features that make them successful invaders and help create nuisance conditions warranting control. The ability to regrow entire plants from vegetative fragments is particularly relevant, and suggests that mechanical harvesting operations of virtually any kind are ill-advised in the early stages of invasion, as they tend to create fragments. Even hand pulling should be conducted with some form of fragment control plan in place. Likewise, the ability to grow and expand coverage from root crowns necessitates killing the whole plant to achieve lasting control. Systemic herbicides, drawdown, dredging, bottom barriers, and hand/suction harvesting offer the greatest potential to achieve this level of control, but not consistently. Other techniques are more likely to leave some portion of the plant alive. For systemic herbicides, all nuisance milfoils are susceptible to the active ingredients 2,4-D and triclopyr, but only Eurasian water milfoil (M. spicatum) is consistently susceptible to fluridone. Mechanical techniques are likely to affect all nuisance milfoil species similarly, but the milfoil weevil (Euhrychiopsis lecontei) is very hostspecific for Eurasian water milfoil. Native plant assemblage Most ecologically sound plant management efforts seek to replace invasive species with a native assemblage more conducive to intended water uses. From a drinking water perspective, there may be plants less desirable than invasive milfoils, but most candidates are also non-native species. Native forms can become nuisances in some cases, however, so control of milfoil must consider what is likely to replace it. In plant management, the cardinal rule is that where light penetrates to a suitable substrate, plants will grow. Unless light or substrate will be limiting to plant growth as a result of management actions, consideration should be given to the likely make-up of the plant community after milfoil control. Ideally, plants with water quality benefits (or at least no major water quality drawbacks) would be desired, and the choice of management method may affect which species dominate after milfoil control. Many species are susceptible to the use of various herbicides (with differences based on herbicide, dose and duration of exposure), and virtually all vegetative forms will be killed by benthic barriers, creating colonization opportunity for seed producers or other invasive species if the barrier is removed or sediment accumulates on it. Harvesting favors species that spread by vegetative fragments over seed producers. Dredging removes vegetative forms and seed banks; unless a light or substrate limitation is created, conditions are likely to favor opportunistic species, especially invasive forms. Rare, Threatened, or Endangered (RTE) species Water utility management activities in reservoirs are usually subject to state and federal regulations addressing rare, threatened or endangered species (collectively termed protected species). While exemptions exist in some cases, consideration should be given to how any proposed milfoil control program might affect protected species. The extent or timing of control 113

134 actions can often be adjusted to minimize impacts, and some techniques have a higher probability of non-target impacts than others. Dredging, while potentially very effective, is likely to create a major ecological disruption in the areas that are dredged. Herbicides can affect nontarget plants and animals, although effects upon the latter are normally minimized by label restrictions on dose and other application protocols. Drawdown can affect some invertebrates (especially mollusks) directly, and may affect many invertebrates indirectly by altering substrate features over time. Knowing which protected species make use of a waterbody is essential to management planning. Summary of Key Biological Features Affecting Management of Milfoil Biological features that minimize the need for control of milfoil: Healthy, dense, native community that restricts milfoil colonization Milfoil occurs only in isolated areas with minimal expansion opportunity Abundant herbivores consume milfoil (Note: None of the above is likely to provide adequate protection from expansion over time; if water supply may be impacted, action is urged even under the above conditions) Biological features that tend to restrict which techniques can be applied: Presence of RTE species that would be impacted by specific techniques Healthy native community limits extent of actions that harm all plants Milfoil growths widespread limits applicability of local techniques such as benthic barriers or hand pulling Milfoil growths very dense limits applicability of hand pulling Milfoil growths in only small portion of reservoir limits efficacy of mechanical harvesting Native community dominated by vegetative reproducers limits herbicide types, doses, duration of exposure, and/or areal extent of treatment if community is to be maintained; also affects impact of drawdown WATERSHED CHARACTERISTICS Upstream Milfoil Presence If invasive milfoil species are present in an upstream waterbody, the potential is high that milfoil will reach the reservoir under consideration. Protective management actions would be highly appropriate if the risk of supply-related problems from milfoil is significant. Inflow From Watershed The rate of inflow, and the related detention time in portions of the reservoir, will determine the efficacy of some potential management techniques. Herbicides require more than fleeting contact time to be maximally effective, and fluridone requires several months of contact time for best effect. Dyes, although rarely used in water supplies, also require higher detention 114

135 times to be effective. Low inflow will affect the rate of refill after a drawdown and may limit drawdown if adequate water supply is threatened. Downstream Water Uses Management actions that can affect downstream uses may have additional permitting requirements or at least notification obligations. Use of herbicides where active ingredients can exit the reservoir at a concentration that could impact downstream resources can be problematic, and any action that creates viable milfoil fragments that can drift downstream may constitute a threat to downstream uses. Grass carp are illegal in some states, and their effect on water quality usually makes them a poor choice for plant control in a drinking water supply, but where they can be used, it is usually required that the outlet be structured to prevent their escape. Watershed Stakeholders Interest groups may exert some degree of control over the choice of management actions, depending upon their actual control over the reservoir or the public/political pressure they can bring to bear on the utility. Recreational interests, where such activities are allowed on the reservoir, are likely to be most vocal, although some organizations devoted to habitat protection for one or more species of interest may also get very involved. In most cases, control of invasive milfoil will be in the best interest of all uses, but the means used to achieve that control can have non-target impacts of concern to select individuals or groups. Some groups may be outside the watershed, with either a regional or even national presence. It is advisable to be aware of stakeholder groups and to involve them in the management process early and to the maximum practical extent. Summary of Key Watershed Features Affecting Management of Milfoil Watershed features that minimize the need for control of milfoil: No milfoil present in upstream areas Inflow from surface sources is minimal Watershed features that tend to restrict which techniques can be applied: Frequent input of milfoil from upstream sources creates ongoing need for actions, limiting more costly options High inflow limits effective herbicide use and increases potential for downstream impact Low inflow limits utility of drawdown without water supply impacts Sensitive downstream uses exist limits use of herbicides, drawdown, possibly mechanical harvesting and biological introductions Stakeholder groups have sufficient power and opinion on specific techniques most often applies to herbicides and biological introductions, but any technique that could affect fishing or valued species may be scrutinized 115

136 IDENTIFICATION OF CANDIDATE CONTROL OPTIONS Preventive Measures Prevention involves threat awareness, education of potentially influential parties, a few specific actions focused on common milfoil vectors, and vigilant monitoring with follow up rapid response as needed. If there is no access to the reservoir other than for utility staff for water supply management purposes, then only internal actions may be needed. If access to the water supply reservoir is open to some portion of the public, education of the range of potential users is needed. A variety of informational approaches can be taken, but direct interaction at access points is usually most efficient. This may involve simple postings, but will be most effective through individual interaction with knowledgeable staff. It may be necessary to extend education and awareness efforts to upstream areas that constitute sources of milfoil to initiate control of those sources. Actual preventive actions mostly involve controlling possible vectors of milfoil. Key steps in preventing milfoil introduction include: Post all access points to the waterbody (e.g., boat launches, swimming areas, fishing piers or obvious shoreline fishing points) with a photograph or drawing of any milfoil of concern and wording to indicate why it is a threat and who to contact if someone thinks they have found it in the reservoir Have personnel at any staffed locations approach all users to draw their attention to the postings or to personally explain why there is concern over milfoil invasion and what steps users can take to minimize the threat Inspect all boats and trailers with access to the reservoir for milfoil fragments (or other signs of possible invasive species introductions) Inspect all bait buckets or other equipment brought to the reservoir by users that could result in milfoil (or other species) introduction Provide a wash station for use before and after using a transported boat in the reservoir Restrict access by boats and use of outside equipment where the threat of introduction is high or ability to inspect is limited; provide boats and other equipment as needed to support desired uses (e.g., a utility operated boat livery) Use scare tactics, habitat manipulation, fencing, overhead wires, or other exclusionary devices to minimize migratory bird use of the reservoir Survey upstream waterbodies for milfoil and encourage control activities where they are found Monitor the reservoir for milfoil invasion, and take an appropriate rapid response action if it is found Monitoring can be tailored to fit available resources and specific features of the reservoir, but the following step by step approach is recommended: 1. Acquire a suitable map of the waterbody, showing shoreline features and reference points, and preferably with water depth contours. 2. Use the information from taxonomic guides, plant keys, or herbarium sheets to identify target milfoil species. 116

137 3. As nuisance milfoils overwinter in a vegetative state, they can be surveyed any time, but are most easily detected and identified in spring as one of the earliest plants to begin growth or in late summer when milfoil may reach the surface. 4. Ideally, space transects around the waterbody, extending from shore to the end of plant growth, with one transect per defined shoreline segment, determining transect location with GPS or readily identified shoreline features. Segments should represent roughly equal areas of the potential plant growth zone within the waterbody. Be sure to cover all boat launch, swimming, inlet, bird congregation, key habitat and intake areas, and any other key access points. 5. Priority can be given to transects of key concern, either based on likely invasion points (access points) or potentially threatened resources (intakes, swimming areas, key habitat) if the number of transects is too great for the manpower and time resources available, but recognize the limitations this will impose on invasion detection. 6. Using a boat with a viewing tube or underwater videocamera, or employing snorkeling or SCUBA gear, examine the plant community along transects between the shore and the maximum depth of plant growth (typically <20 ft, usually <15 ft). Note presence/absence of milfoil and extent of coverage and density where milfoil is encountered. Record observations for 2 ft water depth intervals, with each observation representing either a defined area within the depth range or the length of the transect between depth intervals (typically 0-2 ft, 2-4 ft, 4-6 ft, and so on). GPS is particularly useful for both transect and point location for future reference. 7. Send specimens to a taxonomic expert for confirmation. Note in writing the presumed identification, the name of the waterbody, the approximate location in the waterbody (a map is helpful) with water depth and any other site-specific observations, and contact information for the collector or sender. 8. Tabulate all data in a manner that facilitates future comparisons, typically in a spreadsheet or GIS format. Evaluate presence of any milfoil, extent of coverage and density, and pattern of occurrence. Map the distribution of milfoil in the waterbody for visual reference. 9. Repeat the survey at least once every 3 years (about the time for an invasion to have a detectable impact), and preferably every year to allow the earliest possible detection. Where an invasion is detected, useful steps in quantifying the invasion include: 1. Use the data generated by the transect method above to get a first impression of the extent of invasion, preferably in mapped format. Where milfoil is discovered in multiple locations, look for spatial patterns that suggest either transport from the earliest infestation or invasion from multiple sources. 2. If a discovered growth is in a definable cove, examine the entire cove, or at least that portion with a water depth <10 ft. 3. If a discovered growth is associated with a boat ramp, check a suitable area (typically 1-2 acres) associated with that ramp, and check other ramps if present. 117

138 4. Where growths occur near a tributary mouth, check area maps for upstream ponds or impoundments on the offending stream and any other tributary and investigate where possible milfoil sources seem most likely. 5. When the new growth appears associated with areas of bird congregation, check all such areas in the waterbody. 6. In all cases, note which areas have established beds vs. scattered plants vs. a single plant or just a few stems. 7. Identify all other plants in association with milfoil growths, to the limit of areas likely to be targeted for control. Follow the protocols for species confirmation where specimens of unknown identity are encountered, paying particular attention to possible protected species or other invasive species. Responses to newly detected invasions will vary with the extent and nature of coverage. Possible responses are covered under Screening of control options. Threshold Control Criteria There is no clear threshold above which action must be taken against milfoil to protect water supply interests. It is clear that control is greatly facilitated by early action, so if a threat is perceived, do not wait to act. It is not yet clear that milfoil represents a definite and persistent threat to potable supply, although the potential certainly exists. There are, however, a number of considerations that allow threat evaluation, and from which the need to act can be assessed. Of primary interest are how great an infestation may become, how readily it may be transmitted to new areas (both inside and outside the infested waterbody), what resources may be impacted to what degree, and what the potential is for eradication or control through rapid response to detection of an invasion. In evaluating the potential threat from a new milfoil infestation, consider the following: What portion of the waterbody could be colonized (estimate as the area with water depth <20 ft)? What is the potential for dense bed formation (estimate as the area with stiff muck or sandy silt substrate)? What is the potential for rapid (<3 years) spread (estimate as the common area from #1 and #2 above and not densely covered by native plants)? What is the potential strength of vectors of internal spread (boat traffic, flow, currents, open expanses vs. isolated coves)? What is the potential strength of vectors of external spread (trailered day-use boats, daily or seasonally mobile bird populations, outlets without screening)? What resources and uses are potentially threatened (water supply, swimming, boating, fishing, aesthetics, sensitive or protected populations)? What is the potential for eradication (based on extent and density of coverage, vectors of spread)? What is the potential for confinement (based on extent and density of coverage, physical isolation of area affected, vectors of spread)? 118

139 By answering these questions, one can characterize the threat according to the matrix in Table 6-1, which can then govern the response to detection of an invasion. 119

140 Table 6-1. Threat analysis matrix for milfoil in reservoirs. FACTOR YES NO THREAT EVALUATION A large area Extent and speed could be of possible affected infestation Plant density could be high Spread could be rapid Water supply may be impacted Swimming may be impacted Boating may be impacted Fishing may be impacted Aesthetics may be impacted Sensitive species may be impacted Protected species may be impacted Spread by water flow likely Spread by birds likely Spread by boating likely Spread by other human activities likely Eradication is possible Confinement is possible Annual macrophyte assessments Nature of possible impacts Ability to spread Potential success of rapid response HIGH MEDIUM LOW 120

141 Where an invasion is detected and consideration of control options is underway, it is often wise to quarantine the affected area to the extent possible. Quarantine may involve physical separation by some form or barrier or restrictions on activities that might help spread the milfoil. Options include: Screen the surface outlet of the waterbody to minimize downstream movement of milfoil, maintaining the screen as necessary to facilitate outflow. Lower the water level to prevent surface outflow; a subsurface drain may be used to continue outflow, but milfoil may escape through this exit if not screened, and such screening will require cleaning. Post access points with warnings to avoid the plant and/or certain areas of the waterbody; use marker buoys to identify infested areas. Surround smaller infested areas with sequestration curtain or other enclosing materials that prevent spread and limit access. Curtain off coves or other isolated areas to prevent milfoil spread and limit access. Use scare tactics or other approaches to limit bird use of the waterbody. Set up a washing station and inspection point for boats taken out of the waterbody; require inspection and cleaning where needed. Close any access point (e.g., boat ramp, beach, other points of active contact) in close proximity to milfoil, where the potential for internal or external spread is considered high. Close the waterbody to human use. Screening of Control Options Background information for all possible milfoil control options is provided in Chapter 5.0. Here we address which options appear most useful in typical water supply situations, and suggest key features of each that will determine utility of method for control of invasive milfoil growths. The most commonly recommended early actions are hand harvesting and bottom barriers, each of which has a high potential for success, low cost on a localized basis, and limited permitting needs. Where growths are too dense for effective hand harvesting and too extensive for cost-effective bottom barrier placement, suction harvesting is often considered. Suction harvesting can create considerable turbidity and affect raw water quality, however, so areas larger than about 10 acres are not typically handled with this technique, and not if the target area is close to the intake. If growths are too extensive for effective suction harvesting, herbicides are most often employed, but there will be limitations on application in water supplies and particularly near intakes. On a localized basis, the herbicide triclopyr has great potential for control of milfoil where exposure times are limited and with limited impacts on other native species; more experience is needed to make a more definitive recommendation, however, as this herbicide was registered with EPA and most states between 2002 and The herbicide fluridone is very effective for control of Eurasian water milfoil but is difficult to use on a small scale. However, where Eurasian water milfoil growths are too widespread or dense for other methods, or where detection and removal of all growths in an area is impeded by obstructions or other vegetation, fluridone may represent the preferred approach. Fluridone can be used on variable water milfoil 121

142 and parrotfeather as well, but at higher doses and with less reliable control. Both fluridone and triclopyr are approved for use in potable water supplies with restrictions that do not seriously compromise effectiveness, except perhaps near any intakes. The use of copper to kill any algae associated with milfoil growths can improve effectiveness of these herbicides where such algal growths encrust the target plants. Other herbicides carry label restrictions that minimize their effective use in potable water supplies. Diquat, endothall and 2,4-D can be used in some cases, but not commonly. Drawdown, where applicable, is perhaps the most widely effective preventive control in cases where repeated invasion is expected or documented, but is not applicable in most water supplies as a typical winter measure. However, as a function of repeated water level fluctuations as a consequence of withdrawal and refill cycles in most reservoirs, the sediment alteration effect of drawdown is realized in many reservoirs and serves as a deterrent to milfoil colonization. Commercial biological milfoil controls are limited at this time to grass carp (affects most plants) and the milfoil weevil (affects only Eurasian water milfoil). At a density necessary to achieve plant control, grass carp tend to promote algal blooms, an undesirable effect in a water supply. Use of grass carp is also illegal in many states. Use of the milfoil weevil in combating Eurasian water milfoil has produced mixed results and never a rapid lakewide response; this approach has potential, but is not yet reliable. Dredging is a fairly extreme approach, in terms of cost and impact to non-target organisms, and is usually not employed specifically to control milfoil unless there is a perceived need to increase water depth in the target area or reservoir overall. To achieve control, either a light limitation (through increased depth) or a substrate limitation (by removal of all sediment suitable for milfoil growth) must be created, and this usually proves to be very difficult. Dyes are rarely used in water supply situations, as they are difficult to remove in the treatment process. The degree of control that can be obtained in a reservoir from dyes is also uncertain, and repeated treatment is essential. Milfoil is unlikely to be eliminated by this method, and would only be reduced if light was limiting to nearly all aquatic plants. Mechanical harvesting is a recognized maintenance technique, and repeated harvest of milfoil can encourage growth by some other plants (mainly low growing forms), but the creation of fragments and incomplete collection makes this technique unsuitable where milfoil has not already expanded to most of the potential occupation area. A graphic summary of common milfoil management options is provided in Figure 6-1. Most early responses will involve sparse growths over a limited area or small, dense beds in a confined area. While the listed techniques remain applicable after growths have become widespread, addressing them may involve additional considerations (e.g., impacts to non-target organisms on a lakewide basis). The selection pathways shown in Figure 6-1 represent logical choices based on general features of the aquatic system, and are not intended to provide infallible rules or inflexible options. 122

143 Contiguous Acres of Infected Area Milfoil Stems per 100 Square Feet 123 Significant and Sensitive Protected or Desirable Species Significant Dilution and Flushing Management Options Figure 6-1. Decision tree for the control of invasive milfoil (Myriophyllum spp.) Notes: Hand harvesting and suction harvesting must include root system removal. Benthic barrier should remain in place for 30 to 60 days. Herbivorous insect use is limited by fish predation; where appropriate, expect a 5 year process with multiple stockings. Fluridone use may include liquid, pellets, sequestration and repeat (boost or bump) treatments to maximize exposure; concentrations may have to be increased for maximum impact on milfoil species other than EWM. Triclopyr approved for use only recently; experience is limited. Choice of 2,4-D, diquat or endothall may not be permitted in all potable water supplies; fluridone may work in these circumstances if the area can be sequestered. Drawdown use is dependent on many factors, including hydrology and annual water supply cycle. Moderate to dense growth over an extensive area (>10 acres) may require additional impact considerations. 123

144 Practitioners should use a careful process of option review based on site specific data when selecting a rapid response for milfoil in any target reservoir. Figure 6-1 addresses mainly scientific issues of choice, but economic, regulatory and social factors cannot be ignored. For example, if a one-acre area of invasive milfoil infestation is discovered in a reservoir that could be susceptible to widespread colonization, the scientific choice of techniques would depend mainly on the density of the milfoil, the presence or absence of sensitive non-target species, and the degree of flushing in the target area (Figure 6-1). If the infestation is not dense, but sensitive non-target species are present, hand pulling or suction dredging would be preferable. If the infestation is dense, use of herbicides would be the most cost-effective way to achieve initial control, but application of benthic barriers could also be effective at greater cost but less perceived risk to users of the water supply. The importance of non-scientific factors, such as consumer confidence and cost, should be considered in such decisions. The presence of sensitive non-target species may require the use of more intensive and repeated hand pulling or suction dredging, although the herbicide fluridone may be effective with limited non-target damage if flushing is low or the area can be sequestered. With higher flushing, the herbicide triclopyr might still be effective, and tends to have limited non-target impacts, although consumer acceptance of a herbicide in their drinking water supply remains an important factor in the decision. If less rapid results can be tolerated, use of the milfoil weevil may be appropriate if the target is Eurasian water milfoil. In contrast, if the infested area is large (>10 acres in Figure 6-1), herbicide use is clearly the most effective way to gain control, but may not be allowable or advisable in drinking water supplies, depending on the herbicide of choice, consumer reaction, and various site-specific factors such as flushing rate, sensitive non-target species, and cost. In such cases, more localized physical techniques may have to be applied, attacking the problem piecemeal over an extended time period. Drawdown may offer at least peripheral control, as milfoil species are susceptible to this technique, but unless the water level can be lowered to the maximum depth at which milfoil can grow (usually >20 ft), complete control is unlikely. Techniques other than those to which the decision tree leads might be considered, but may have feasibility issues when applied to cases for which they are not considered a logical choice. Examples include applying hand harvesting to dense milfoil growths and stocking of grass carp, which are not even included in Figure 6-1 as a function of unintended impacts and illegality in many states. CRITERIA FOR SELECTION OF PREFERRED CONTROL METHODS Technical Feasibility Whether or not a technique will work in the specific circumstances under which it is to be applied is a critical consideration for its possible application. No method comes with an unconditional guarantee, but each has strengths and weaknesses that make it more or less applicable for specific situations (see Chapter 5.0). Careful attention to detail in planning for the use of a milfoil control technique is advised. Duration of Effect Ultimately, the objective of milfoil control is eradication of the target species. Ecologically, it is expected that the milfoil will be replaced in the plant community with a more desirable plant. If the goal really is to have few or no plants in some portion of a reservoir, light 124

145 or substrate limitations must be imposed or ongoing maintenance will surely be required. Where it is accepted that some plants will grow in the managed area, the duration of milfoil control will be a function of the completeness of initial control, the proximity of other milfoil growths and immigration from those sites, and the success in establishing a native assemblage that limits colonization. Incomplete control usually results in milfoil becoming dominant within four years and often in only two years, if milfoil growths were extensive before control. Where invasive milfoil growths nearby have not been controlled, they are as likely to recolonize the area as native species, and do so within three to five years in most cases. However, if a native assemblage can regrow before milfoil colonization can proceed, the duration of control is typically more than five years. With follow up monitoring and rapid response to new growths, the duration of effectiveness should be indefinite, but it is rare to keep milfoil under control for a decade or more. For the listed techniques, those that kill the whole plant and offer the best opportunity for lasting control include proper hand or suction harvesting, benthic barriers, and the herbicides fluridone and triclopyr. Dredging and drawdown may also eliminate the whole plant, although some milfoil frequently survives these approaches. Effectiveness Against Target Species All the techniques described in Chapter 5.0 have some potential to be effective against one or more species of invasive milfoil. Some techniques are more selective than others. The milfoil weevil is fairly specific for Eurasian water milfoil, while benthic barriers kill all plants under the mat and hand harvesting can be as selective as the training of participants allows. Dredging tends to eliminate all plants from an area, at least temporarily. Fluridone can be used selectively for Eurasian water milfoil, but may have to be used at higher doses that are nonselective when attacking variable water milfoil or parrotfeather. Triclopyr targets milfoils and other dicotyledonous plants, with lesser impact on monocotyledons. Drawdown, diquat, endothall, dredging, grass carp and dyes target a broad range of plants, including milfoil, and results are somewhat site dependent. Harvesting also targets a range of plants, based mainly on height of growth, but different results can arise based on ability of cut plants to regrow and reproductive effects of harvesting (e.g., seed head removal, prevention of food reserve build up in roots). Potential Collateral Impacts Many collateral impacts relate to the selectivity discussed above. Selectivity is usually desired but not always possible. Where control efforts focus on small areas within a reservoir, damage to local non-target resources is usually accepted and is not expected to have long-term or lakewide impacts. Where a substantial part of the reservoir is involved, as with a whole reservoir treatment with herbicide or drawdown, potential collateral impacts may require consideration for possible impacts on water supply or within a regulatory framework. Integration with Other Methods It may be possible to control milfoil with a single technique, especially where an introduction of an undesirable species is detected early, is confined to a small area, and is addressed rapidly. It is rare, however, to achieve control and maintain a desirable plant 125

146 community with only one technique. Where something more than a localized control technique is required, at least one localized technique should be ready for implementation to address any inefficiencies of the larger scale technique. For example, hand harvesting and benthic barrier placement are often applied to prolong the control expected from larger herbicide treatments, dredging, or drawdown. Costs Costs are best evaluated on a case by case basis, as site specific features often have a substantial impact on the expense of a given technique. However, it is reasonable for those considering possible options to have some approximation of probable unit costs, or at least a range of unit costs, for general evaluation and comparison purposes. Cost estimates are provided for each technique in Chapter 5.0. Consider cost over an expanded timeframe, not just the year of initial application. Five- and 20-year plans are recommended for plant management. Regulatory Response Many activities of water suppliers that would be regulated in a recreational lake are exempt from at least some environmental regulations in many states. However, not all activities are exempt and utilities are not exempt from federal regulations such as the label restrictions on herbicide use or conditions to be maintained under the Safe Drinking Water Act. The actual control of invasive milfoil is unlikely to trigger any negative regulatory response, but any nontarget impacts to other aquatic biota, water quality, or downstream resources is likely to be subject to regulation. Control activities should be planned with a complete knowledge of the federal and state requirements related to each option. Consultation with state agencies responsible for invasive species management and permitting under environmental laws is encouraged early in the process. Stakeholder Response As noted in the section on watershed characteristics to be considered, stakeholder groups may have an interest in control options that will affect the outcome of the selection process. Rate payers of the utility can be considered to be one stakeholder group, whether or not they are in the watershed, and will exert some pressure relating to both water quality and cost. Recreational groups (notably fishermen) and environmental organizations (especially those devoted to habitat preservation) can provide helpful alliance or formidable opposition, depending upon their viewpoint on proposed control actions. Where control of the reservoir is not complete, or where downstream resources are potentially affected, involving these groups early in the planning process is strongly recommended. 126

147 CHAPTER 7 SURVEY OF DRINKING WATER SUPPLIERS AND RESERVOIRS INTRODUCTION A questionnaire was developed for distribution to utilities managing water supplies to assess the extent of impacts and mitigation measures associated with milfoil infestations. General notices were posted in several newsletters and on-line by AwwaRF prior to distribution to targeted and/or interested water company executives or related personnel. This report summarizes survey results from 43 utilities managing 57 reservoirs, plus an assessment of potential drinking water sources impaired by nuisance plants obtained by checking records with state agencies. METHODOLOGY ENSR developed the questionnaire with assistance from the Project Advisory Committee (PAC) via and during a meeting in Georgia in February The goals of the questionnaire were to characterize the extent of known problems in water supplies and gather background information, such as geographic information and waterbody characteristics, that might aid in the assessment of reservoir susceptibility. Representatives from eight utilities betatested the questionnaire for clarity, relevance, and required completion time. The questionnaire was revised based on the beta-testing. Following revision, a request for participants was sent out by through AwwaRF to 5,000 water utility executives and notices were posted on-line. Additionally, ENSR used the New England Water Works Association (NEWWA) directory, with NEWWA permission, to contact utilities in New England by phone with a request to participate in the survey. In December 2005, 314 questionnaires were distributed to drinking water utilities and state agencies that responded to the request for participants. In an attempt to increase the number of respondents, three follow up phone calls were administered to the utilities that had not responded after several weeks following having been sent the questionnaire. As it became apparent that the return rate would be low, ENSR contacted state agencies in 32 states to determine what records they might have for the distribution of milfoil in drinking water reservoirs and any assessment of impairment as a result. These data were used to further our assessment of the threat of milfoil to drinking water supplies. RESULTS Of the 314 questionnaires distributed to utilities, less than 15% were returned and an average of 71% of the questions on each questionnaire were completed. Utilities managing milfoil-infested reservoirs completed an average of 80% of the questions on each questionnaire; utilities managing reservoirs without milfoil filled out an average of 76% of the questionnaire; and utilities that did not know if their reservoir was infested with milfoil answered an average of 58% of the questions. The questionnaire was divided into several sections: waterbody characteristics, watershed characteristics, raw water characteristics, and milfoil related questions. Some sections of the questionnaire were answered more completely than others. For example, 82% of the 127

148 utilities completed the section regarding raw water characteristics, while only 45% provided answers to the questions relating to milfoil. Regarding the milfoil section, utilities reporting milfoil infestation in their reservoirs completed an average of 69% of the milfoil related questions; utilities reporting no milfoil infestation in their reservoirs completed an average of 64% of those questions. Utilities that did not know if their reservoirs were infested with milfoil answered only an average of 5% of the questions related to milfoil (Figure 7-1). Overall, 88% of the 43 utilities were public, regional suppliers, 7% were municipal entities and 5% were private organizations. 100% 92% 100% 80% 80% 76% 83% 82% 80% 78% 78% 76% 75% 70% 70% 67% 69% 64% % Complete 60% 40% 58% 56% 46% 20% 5% 0% All Questions Raw Water Characterstics Watershed and Recreational Uses Physical Characteristics Milfoil Information All Questionnaires Yes-Milfoil No-Milfoil Unknown-Milfoil Figure 7-1. Percent of questions answered in each section of the questionnaire. Out of the 57 reservoirs, 15 (26%) were reported to have milfoil, 22 (39%) were reported not to have milfoil, and the remaining 20 (35%) were managed by utilities that either did not know if milfoil was present or did not answer the question for some other reason. Of those reservoirs with milfoil, two had Eurasian watermilfoil only, five had variable watermilfoil only, five had both Eurasian and variable watermilfoil, one had parrotfeather only, and in two reservoirs the species was unknown. The total water volume of the 21 reservoirs for which water volume estimates were provided (7 with milfoil, and 14 without), was 1,164 billion gallons (BG), and of that, 99 BG were at depths less than 15 ft (Table 7-1). The 0-15 ft depth is the most conducive to milfoil growth. While milfoil can grow at depths greater than 15 ft, it is unlikely to be very dense or represent a great threat to the waterbody. Most of the volume associated with water depths <15 ft, or 79 BG, is in milfoil-infested reservoirs. This is only 7% of the total water volume of all the reservoirs, but is 79% of the shallow water volume of the reservoir set examined. However, most of the total volume (1088 BG, or 93%) was also in the infested reservoirs. The level of boating activity as well as the extent of residential and agricultural development varied only slightly between milfoil-infested and un-infested reservoirs. Of the 128

149 respondents with milfoil-infested reservoirs, 47% allowed boating, 40% prohibited boating, and 13% did not answer the question. Of the utilities managing reservoirs without milfoil, 55% Table 7-1. Total water volume and water volume that potentially has milfoil in billion gallons (BG) for both milfoil-infested and un-infested reservoirs. Infested Reservoirs (n=7) Un-infested Reservoirs (n=14) Total Water Volume (n=21) Water Volume (BG) 1, ,164 Water Volume in 0-15 ft. Contour (BG) Water Volume that Potentially Has Milfoil (BG) permitted boating and 45% prohibited boating. Concerning watershed land use, 40% of the infested and 40% of the un-infested reservoirs did not have residential areas in their watersheds. Of the utilities managing milfoil-infested reservoirs, 27% reported the presence of livestock in the reservoir s watershed and 33% reported having no livestock in the watershed; 40% of the utilities did not answer the question. Of the utilities managing reservoirs without milfoil, 23% reported the presence of livestock in the reservoir s watershed and 77% reported having no livestock in the watershed. Utilities were asked various questions regarding raw waterbody characteristics. Parameters including ranges for ph, hardness, turbidity, color, dissolved and total organic carbon and total phosphorus did not seem to differ greatly between reservoirs with and without milfoil. In reservoirs with milfoil, 80% had an alkalinity concentration of less than 20 mg/l, whereas only 13% of reservoirs without milfoil had an alkalinity concentration in that same range. The survey included questions regarding physical characteristics of the waterbodies, including reservoir type, residence time, sediment type and percent of the waterbody in the littoral zone. The primary difference between waterbodies infested with milfoil and waterbodies without milfoil was that un-infested waterbodies included more tributary reservoirs, waterbodies fed by many small tributaries. Proportions of other types of reservoirs were similar among the two groups (Figure 7-2). Utilities reported that 40% of the milfoil-infested reservoirs and 59% of the un-infested reservoirs had a residence time of less than one year (Figure 7-3). Sediment type in the 0-15 ft contour and average depth were related to milfoil presence. Almost half (46%) of the utilities with milfoil-infested reservoirs reported that the majority of sediment in the 0-15 ft range was sandy, 27% reported silt/muck sediment, 7% reported rocky sediment, and 20% did not answer the question. Of the un-infested reservoirs, 5% had sandy sediment, 54% had silt/muck, 9% had rocky/sandy sediment, and 32% of the utilities did not answer the question (Figure 7-4). Only 20% of milfoil-infested reservoirs had at least half of their total area at depths between 0 and 15 ft, although 53% did not respond to the question, suggesting a lack of knowledge about reservoir conditions. Of the un-infested reservoirs, 55% 129

150 had over half of their total area at less than 15 ft deep, while 32% of the utilities managing uninfested reservoirs did not answer the question. 10 Number of Reservoirs Milfoil (n=15) No Milfoil (n=22) Reservoirs Mainstem Reservoir Natural Lake Storage Reservoir Tributary Reservoir Figure 7-2. Types of reservoirs with and without milfoil infestation. Reservoirs with Milfoil Un-Infested Reservoirs 27% 7% 13% Less than 1 month 1-3 months 27% 14% 3-6 months 18% 33% 20% 6 months to 1 year Greater than 1 year Unknown 14% 9% 18% Figure 7-3. Residence time in milfoil-infested and un-infested reservoirs. Utilities were asked if milfoil-infested waterbodies were present upstream or within 0-50 miles of the reservoir. Of the 15 milfoil-infested reservoirs, 27% had upstream infested waterbodies, another 27% did not, and the remainder either did not know or did not answer the question for other reasons (Figure 7-5). For the milfoil-infested reservoirs, 13% had infested waterbodies within one mile and 40% had infested waterbodies within 5-50 miles. Almost half (47%) of the utilities, however, were unaware of nearby infested waterbodies or did not answer the question. Of the 22 un-infested reservoirs, 9% had upstream infested waterbodies, 45% did 130

151 not have upstream milfoil-infested waterbodies, and no response was given for the remaining 46%. For the un-infested reservoirs, utilities reported that 14% had infested waterbodies within a 5 mile radius, 9% had infested waterbodies within 5-50 miles, 14% had infested waterbodies farther than 50 miles away, and 63% did not know if there were infested waterbodies anywhere in the region. Milfoil Infested Reservoirs Un-Infested Reservoirs 20% 7% Rocky and Sandy Sandy 32% 9% 5% Silt and Muck 46% Unknown 27% Rocky 54% Figure 7-4. Sediment types in the 0-15 ft contour in both milfoil-infested and un-infested reservoirs. Milfoil Infested Reservoirs Un-Infested Reservoirs Yes 27% Yes 9% Unknow n 47% Unknow n 45% No 45% No 27% Figure 7-5. Presence or absence of upstream milfoil-infested waterbodies for both milfoilinfested and un-infested reservoirs. Nuisance factors were assessed with questions regarding percentage of water column occupied by plants and the extent of clogging issues at water intakes. Of the utilities with 131

152 milfoil-infested reservoirs, 13% reported that plants occupied less than 5 percent of the total water column in the reservoir, 33% had 5-25 percent occupation, 13% experienced percent of the volume filled, and 41% did not answer the question (Figure 7-6). Of the utilities managing un-infested reservoirs, 27% reported plant water column occupation as less than 5 percent, 9% reported that 5-25 percent of the water column was occupied by plants, and 64% did not answer the question. Of the utilities with milfoil-infested reservoirs, 33% sometimes experience clogging of their intakes while the remainder reported that clogging rarely or never takes place. Of the un-infested reservoirs, 5% reported clogging takes place often, 5% reported clogging occurring sometimes, 87% reported that clogging rarely or never takes place, and 5% did not answer the question. Milfoil Infested Reservoirs Un-Infested Reservoirs 13% <5% 27% 41% 5-25% 33% 50-75% Unknown 64% 9% 13% Figure 7-6. Percent of the water column occupied by plants in milfoil-infested and uninfested reservoirs. Utilities managing reservoirs with milfoil tended to implement some method of plant control (87%), while those managing un-infested reservoirs applied some form of plant control much less often (23%). The primary methods for plant control were benthic barriers and handharvesting (Figure 7-7). In both milfoil-infested and un-infested reservoirs, descriptions of effectiveness of each plant control method varied greatly, ranging from no change, to an increase in plant growth, to complete elimination of the targeted plants. The survey of state agencies involved in assessing use impairments in waterbodies generated some amount of useable data for 24 out of the 32 more northerly states contacted (Table 7-2). The data are almost entirely from the Section 305b reports prepared every other year by states to document the level of impairment of waters in their respective jurisdiction. Reporting is not consistent across states, is incomplete from the perspective of water supply in many, and is often not as specific as desired for those looking for detailed breakdowns by use category and type of impairment, but it does provide some measure of the extent of invasive plant problems for water suppliers. Almost two-thirds of the acreage of lakes, ponds and reservoirs in 24 states has been assessed for use impairment, although 9 out of the 24 states did not specify invasive plants as 132

153 causes of impairment. Almost half a million acres have been declared impaired by invasive plant species, although this includes species other than milfoil. Milfoil is the dominant cause of plant impairment in northern lakes; the most southerly lakes were not included in this assessment because, while they have even more impairment from invasive plants, species other than milfoil often cause the impairment. The waters impaired by invasive plants comprise 17.2% of the total acres surveyed for invasive plant impairments. If we assume that drinking water surface supplies are no more or less susceptible to impairment than the overall population of lakes, this could represent a reasonable estimate of the threat level from invasive plants (primarily but not exclusively milfoil) as of % % of Reservoirs Practicing Each Method 80% 60% 40% 20% 0% Benthic Barriers Hand Harvesting Plant-eating Insects Floating Fragment Barriers Drawdown Herbicides Type of Plant Control Mechanical Harvesting Plant-eating Fish Other None Milfoil infested reservoirs Un-infested Reservoirs Figure 7-7. Plant controls used in both milfoil-infested and un-infested reservoirs. 133

154 Table 7-2. Assessed and impaired waters from Section 305b reports. Lake, Pond and Reservoir Acres in State Total Acres Assessed Percent of Total Acres Assessed Acres Impaired by Non-native Aquatic Plants % of Total Acres Impaired by Nonnative Aquatic Plants Public Water Supply Acres Assessed Public Water Supply acres impaired % Public Water Supply Acres Impaired Listed Impairment by Nonnative Plants State Connecticut Yes Delaware Unknown No Idaho Unknown Unknown Unknown No Illinois Yes Indiana Yes Iowa Yes Kansas Unknown Yes Kentucky Unknown No Maryland Unknown Unknown No Massachusetts Unknown Yes Michigan Unknown Yes Montana Yes Missouri Unknown No Nebraska Unknown No New Hampshire Yes North Dakota Unknown No Rhode Island Unknown Yes South Dakota Yes Tennessee Yes Vermont Unknown Yes Virginia Unknown No West Virginia Yes Wisconsin Unknown Yes Wyoming Unknown Unknown No Total Assessed for Invasive Plants Drinking water supplies identified as having been surveyed for impairment account for just over 2.1 million acres in the 24-state database, but those surveyed for impairment by invasive plants total only slightly more than half that amount, at 1.13 million acres. Approximately 22% of the surveyed drinking water supplies were found to be impaired by some factor, invasive plants being one possibility. Considering only those reservoirs surveyed for invasive plant impairments, 29.2% were found to be impaired. However, factors other than invasive plants could still be responsible for that impairment, and of those that are impaired by invasive plants, plants other than milfoil may be involved. This estimate is therefore a maximum value. Recall that only 7% of the volume of reservoirs for which questionnaires were received was found to be potentially impaired as a consequence of milfoil infestation, and that the total volume in <15 ft of water and therefore potentially directly affected was only 8.5% of the total volume (Table 7-1). Indirect effects in the deeper portions of these reservoirs could increase the portion of the volume impaired to some degree by milfoil presence. At the same time, the data from Table 7-2 suggest that if all acres in a reservoir are included in the estimate of impaired area, as much as 29.2% of the resource could be affected. As factors other than milfoil may cause some of this impairment, and not all portions of the reservoirs deemed impaired may actually be 134

155 impaired, the actual portion of the resource that is impaired is undoubtedly lower. The same calculation for all lakes included in the database represented by Table 7-2 yields an estimate of 17.2% of the resource being impaired. The actual potential level of impairment of drinking water reservoirs by milfoil is probably somewhere between 10 and 15%, applying all available data and professional judgment. DISCUSSION The low questionnaire response rate is an issue in the interpretation of results. Relatively few utilities filled out the questionnaire and most that did fill it out did not answer all the questions. The low response rate overall is attributed to several factors including: 1. Lack of awareness of any problem relating to milfoil 2. Inability to address many questions as a function of incomplete or absent data for reservoirs 3. Security concerns 4. Regulatory concerns The low response rate to specific questions can be attributed to many of the same factors plus fragmentation of knowledge within utilities that increased the effort necessary to address all questions. It is apparent that the utilities with known milfoil infestation in their reservoirs recognize that it is a potential issue, since not only do the overwhelming majority of utilities with infested reservoirs use plant controls, but most use several methods simultaneously. While plant controls were implemented by a few utilities managing un-infested reservoirs, the vast majority did not use any plant control methods, and those that did tended to use only one method. The results revealed that only 7% of the total water volume of the reservoirs for which responses were received is likely to support significant growths of milfoil. Utilities were not asked where milfoil was present or the percentage of their littoral zone occupied by milfoil, therefore making it difficult to draw a conclusion on the exact percentage of water volume that had milfoil. However, it is apparent that the portion of overall reservoir water volume directly impacted by milfoil is low. Indirect impacts may still be significant in some cases, mainly as changes in water quality relating to organic carbon content, ph, or dissolved oxygen. While milfoil is most likely to be abundant at water depths less than 15 ft, the survey indicates that shallow depth (0-15 ft) does not necessarily promote or increase the probability of milfoil growth; infested reservoirs for which depth distributions were reported actually had less shallow water than reservoirs without milfoil. Water level fluctuations are likely to be very important in these shallow areas, with periods of exposure limiting milfoil growth. Because of the relatively small sample size and due to the fact that a large percentage of the utilities did not answer the questions regarding depth, it is more difficult to draw conclusions on a link between depth and milfoil infestation. While milfoil does not grow at great water depths, it is clear that shallow depth alone is not a primary determinant of milfoil presence. Even though plant biomass appears to be higher in milfoil-infested reservoirs, as evidenced by the portion of water column filled, the majority of utilities with infested reservoirs did not report clogging of their intakes. This is most likely because intakes are typically located in deeper sections of reservoirs, and since milfoil tends to thrive in the 0-15 ft depth range, intake clogging for potable water supplies may be less of a threat than for irrigation systems. 135

156 Based on the available data, it appears that there is a link between low alkalinity and milfoil growth. The vast majority of milfoil from this survey grows in waterbodies that have an alkalinity of less than 20 mg/l, while reservoirs without milfoil tended to have a much higher alkalinity. As Eurasian water milfoil is quite tolerant of high alkalinity, it is difficult to conclude that reservoirs with a high alkalinity are less susceptible to milfoil infestation, but low alkalinity lakes may be more susceptible to more nuisance forms of milfoil. Variable water milfoil was the most common problem species noted in this survey, and is known to prefer low alkalinity lakes. While boating provides an easy method for milfoil to be transferred between waterbodies, the surveys did not reveal that boating greatly influenced milfoil presence, since in both infested and un-infested reservoirs about half allowed boating and the other half prohibited the use of boats. The lack of correlation may be due to the small sample size, or because other routes of invasion may be more prevalent in water supplies, in contrast to many recreational lakes. It is also possible that other features of water supply reservoirs, most notably water level fluctuations, limit colonization by milfoil from any vector. Watershed characteristics including the presence of residential and agricultural development did not show a direct correlation with milfoil growth. Concerning residential development, responses from utilities managing milfoil-infested reservoirs did not vary greatly from the responses from utilities managing un-infested reservoirs. Regarding livestock, almost half of the respondents managing milfoil-infested reservoirs did not answer the question, and the remaining utilities were almost equally divided between those having livestock and those not having livestock, making it difficult to conclude that agricultural development influences milfoil growth. Regarding flow rate and residence time, the questionnaires suggest that a reservoir is more susceptible to milfoil infestation if there is less flow and a higher residence time. Rivers, tributaries and other flowing waterbodies tend to have a lower residence time than natural lakes and may have a lower risk of milfoil infestation. This is not intuitively obvious, as flowing water represents a route of invasion, but it may be that instability relating to higher flows limits colonization potential. The presence of upstream milfoil-infested waterbodies is potentially important as evidenced in the questionnaire responses, and would suggest that flow is an encouragement to milfoil invasion, although it is even more evident that most respondents are unaware of the distribution of milfoil in their region. The available data suggest that sediment type in the 0-15 ft depth range may influence milfoil growth. Results from the survey show that sandy substrates are preferred, and silt and muck sediments do not strongly favor milfoil. As with flow issues, this may be a matter of stability; milfoil can grow on a wide range of substrates, but sandy substrates with substantial organic content do seem to be preferred. Overall it is difficult to draw conclusions on the optimum sediment type for milfoil growth because so many utilities did not know the dominant sediment type in the 0-15 ft contour or did not answer the question for some other reason. The high percentage of unknown responses to milfoil questions indicates a lack of knowledge and/or interest in the subject among utilities. Through communication with representatives from utilities it was clear that many had never heard of milfoil and were not aware of its potential impacts. Representatives also commented that they did not have the time or interest to fill out the survey; several mentioned that a one or two page survey would be feasible, but any more would be too long. Based on the available data, it is apparent that milfoil is affecting a small percentage of the total water volume of water supplies, although some individual supplies may be heavily 136

157 impacted. Certain factors, including low alkalinity, sandy substrates, low flow rate and the presence of upstream milfoil-infested waterbodies, seem to promote milfoil growth. Waterbody and watershed characteristics including depth, the use of boats, and the presence of residential and agricultural development in the watershed did not necessarily promote milfoil growth. Because of the low questionnaire response rate from utilities and the high percent of unknowns or unanswered questions on each questionnaire, it is apparent that there is a general lack of interest in or knowledge about milfoil. Overall, characterization of the extent of problems is needed before more pressure can be exerted to react to any threat from milfoil. The results of queries to state agencies responsible for assessing aquatic conditions relative to designated water uses indicate that the actual level of impact by milfoil on drinking water supplies may be considerably higher than estimated from the surveys, but the approach taken by state agencies in evaluating impairment tends to overestimate the specific impact of milfoil. Based on all data available at this time, it is reasonable to project that 10-15% of the volume of all drinking water supplies in the northern USA is potentially threatened by invasive species of milfoil. This level of threat is worthy of consideration in management planning, but does not represent a crisis for the drinking water industry as a whole. The critical caveat is that it is not 10-15% of each reservoir that is threatened, but rather a range of 0-100% of any reservoir that averages to 10-15% overall. If a susceptible reservoir is invaded by milfoil, this may indeed represent a crisis for that water supply. Consequently, education as to the potential threat and the features of a reservoir that make it susceptible to impairment from milfoil is warranted. 137

158 138

159 INTRODUCTION CHAPTER 8 CASE STUDIES IN MILFOIL MANAGEMENT As discussed in Chapter 6 of this document, there are numerous considerations for the susceptibility of reservoirs to nuisance milfoil invasion and application of milfoil control techniques to drinking water supplies. Physical characteristics of the infested waterbody (i.e., morphometry, bathymetry and sediment features) play a major role in determining the necessity and applicability of milfoil control techniques. Reservoirs with large littoral (shallow, peripheral) zones consisting of sandy/organic silt sediments represent a high risk for invasion and colonization by nuisance milfoil species. In general, increased slope will decrease accumulation of fine sands and silts preferred by milfoil. Also, as milfoil is susceptible to drying, large fluctuations in water level will limit survival of invading milfoil. In addition, the location of the drinking water intake(s), the pattern of water withdrawal, and other water uses are all factors when determining what measures are suitable to control an infestation. Consideration of a wide range of factors is therefore necessary in assessing risk and developing a response plan for a specific reservoir. Identification and assessment of a milfoil infestation must occur in order to initiate actions to prevent its spread within the waterbody. Assessment protocols and general techniques for performing macrophyte surveys have been described in Chapter 5. Identifying and mapping milfoil occurrence is important to understanding the rate of expansion, identifying areas for implementation of control measures, identifying possible sources of introduction, and establishing a protocol to prevent additional spread of plants within and outside of the infested waterbody. In many cases, early identification of pioneering colonies and implementation of a Rapid Response Plan (RRP) may eliminate the milfoil altogether. Once the plant becomes established, however, eradication is rarely accomplished; at that point, control and minimization of spread become the most realistic goals in most cases, and an evaluation of the extent of possible damage becomes important to the economics of control efforts. Evaluation of site specific data is necessary to select the most appropriate control technique(s) for a given waterbody. Effectiveness against target species, technical feasibility, duration of effect, potential collateral impacts, integration with other methods, costs, regulatory restrictions, and stakeholder response are criteria that must be considered during the selection of preferred control methods for milfoil. Each of these criteria is discussed in Chapter 6. Case studies of drinking water reservoirs with varying degrees of milfoil presence and management activity are presented here as illustrations of the assessment and management process. The individual responses of drinking water utilities appear to be related to the degree of infestation and perception of milfoil threat and potential for control. The level of management activity in the following examples varies considerably and illustrates much of the range of concerns and responses to invasive milfoil. 139

160 WACHUSETT RESERVOIR Wachusett Reservoir is located in the towns of Boylston, West Boylston, Clinton, and Sterling, Massachusetts, in the Nashua River Basin, and is used year round for water supply. Diversion of water (Quabbin Reservoir and occasionally the Ware River) accounts for 67% of the raw water in Wachusett Reservoir, with the balance coming from direct flow from tributaries (33%). The primary direct inputs come from the west, which is also where the milfoil invasion appears to have begun (Figure 8-1). Wachusett Reservoir consists of 3,904 surface acres with maximum and mean depths of 128 and 51 feet respectively. The maximum reservoir volume is 66 billion gallons. Drinking water from Wachusett Reservoir supplies approximately 2.2 million people in 43 communities. The Massachusetts Water Resources Authority (MWRA) is the managing utility, but local reservoir management activities are carried out by the Massachusetts Department of Conservation and Recreation, Division of Water Supply Protection (DWSP). Figure 8-1. Site map of Wachusett Reservoir. The box on the left side indicates basins where milfoil control has been focused. 140

161 The watershed of Wachusett Reservoir covers 114 square miles. The majority of the watershed area is forested (71%), with 10% in wetlands and lakes, 11% residential/urbanized, 7% cultivated land, and less than 1% used for livestock. There are no residential developments along the shoreline. Public access to the reservoir is restricted to shoreline fishing only in designated areas; no boat launches or beach areas are available. DWSP, MWRA and contracted vendors, when necessary, are the only people with boat access. Water uses and associated potential transport vectors within Wachusett Reservoir are therefore limited. The only significant transport vectors for aquatic macrophytes appear to be water flow and waterfowl. A substantial amount of water is relayed to Wachusett Reservoir from Quabbin Reservoir, which does allow boating in a very limited area, but there is no report of EWM in Quabbin Reservoir. VWM has been documented in Quabbin Reservoir since 1973, but its distribution has remained limited to small isolated patches in coves and tributary inlets. Approximately 20% of the total surface area of Wachusett Reservoir is contained in the 0-15 foot contour, suitable for substantial growths of aquatic macrophytes. Bottom type consists of 55% sandy sediment, 40% rock, and 5% silt and muck. This equates to approximately 430 surface acres in the 0-15 foot contour with sandy sediment preferred by EWM. This is a rough estimate, but provides an indication of the potential for colonization at a density high enough to affect at least local conditions. VWM prefers silt/muck sediments which would correspond to 39 acres of optimal habitat. Again this is a rough estimate, but suggests that a much greater area could be colonized by EWM than VWM in Wachusett Reservoir. Parrotfeather is the third invasive milfoil known in the USA, but prefers nutrient-rich waters not characteristic of Wachusett Reservoir and is known only from one infestation in Massachusetts, far from Wachusett Reservoir. While vigilance is urged with regard to all possible invasive species, parrotfeather does not appear to represent a major threat at this time in this case. Aquatic macrophytes in Wachusett Reservoir can be found growing at depths approaching 30 feet, but growths are sparse at deeper than 15 ft and are not commonly dense even in the shallower areas. The 0-15 foot contour experiences 1-25% coverage by aquatic plants with a corresponding biovolume of 5-25%. The drinking water intakes at Wachusett Reservoir are located at feet and >30 feet of water depth and are not within the preferred depth range for milfoil species; direct impacts are therefore likely to be limited. Wachusett Reservoir has documented populations of both VWM and EWM. VWM was first discovered in 1994 and EWM was discovered in VWM is widespread in Wachusett Reservoir but is not dense, does not appear to have an extensive area of potential colonization, and is not viewed as a major threat by the water utility. EWM was first detected in the upper reaches of the reservoir in Stillwater Basin, an impounded portion of the Stillwater River. Stillwater Basin is three sub-basins removed from the main body of Wachusett Reservoir in the upstream direction. At the time of its discovery in 1999, EWM had a spotty distribution among a variety of native plants. Controlling EWM in Stillwater Basin was not considered feasible due to the physical characteristics of this basin (generally shallow and with thick deposits of fine organic sediments) and EWM distribution among dense beds of native plants. It is assumed that viable fragments moved downstream in the Stillwater River to this impoundment. In 2000, there was a minor increase in EWM density, within an assemblage that included a mix of native plant species. In 2001, a rapid expansion of the infestation was observed, with EWM colonizing Upper Thomas Basin, the 2 nd sub-basin from the main basin of Wachusett Reservoir. Planning for management was initiated in 2001, with actual controls beginning in

162 Milfoil Assessment Detailed assessment of the EWM problem and management planning occurred in 2001 in response to the rapid expansion of the infestation observed that summer. This expansion represented a significant threat to the main reservoir because of the rapid and overwhelming colonization potential of the plant. DWSP and MWRA were concerned with water quality implications of an infestation of EWM in the main basin, especially since water treatment is limited in this generally very clean and unfiltered water supply system. During the fall and winter of , management techniques were considered, selected, and a scope of work was developed. The procurement process took place in late winter/early spring of 2002 and inwater work began in the spring of Five management techniques were considered to control the EWM in Wachusett Reservoir. Herbicides were immediately rejected as inconsistent with public opinion of appropriate management of a drinking water supply. Suction harvesting of the milfoil was considered but rejected due to lack of availability of the necessary equipment, potential for fragment creation, and expected high turbidity creation. Bottom barriers, floating fragment barriers and hand-harvesting were all considered and ultimately implemented in a multi-process control plan in Management Efforts Management efforts have continued through 2006 and are expected to continue indefinitely. Management has focused on Stillwater, Upper Thomas and Thomas Basins (Figures 8-1 and 8-2), but has also extended into the main basin of the reservoir wherever EWM has been found EWM Management An aquatic macrophyte control company was contracted in 2002 to complete the work associated with controlling EWM in Upper Thomas and Thomas Basins. An aquatic macrophyte survey using GPS was conducted to create accurate maps of EWM distribution in Thomas Basin and Stillwater Basin to provide a useful baseline. The goal of the management program was to actively control plants in Upper Thomas and Thomas Basins before they were able to migrate to the main reservoir. The macrophyte survey revealed increased abundance of EWM in Upper Thomas Basin and Thomas Basin proper. In addition, a small colony of EWM was identified in the main reservoir. Management techniques first suggested by DWSP personnel were confirmed as appropriate by the control company and implemented in Benthic barriers were deemed the most appropriate control measure for the largest area of EWM in Upper Thomas Basin. This area was approximately 2 acres in size and had greater than 50% coverage by EWM plants. Seventy two panels of 20 mil thick polyvinyl chloride (PVC) liner were installed between June and July to control existing vegetation and prevent future vegetation growth. Floating fragment barriers were installed in three locations within Thomas Basin. One was installed between Stillwater Basin and Upper Thomas Basin, one between Upper Thomas Basin and Thomas Basin proper, and the last one spanned the open end boundary of the area covered by bottom barriers, at the northern tip of Upper Thomas Basin (Figure 8-2). The goal of 142

163 Figure 8-2. Areas in Upper Thomas and Thomas Basins managed in From 2006 summary report by Aquatic Control Technology, Inc. the fragment barriers is to intercept and accumulate any autofragments or pieces of EWM on the surface of the water. Autofragments are stem segments with adventitious roots that float after abscission from the main plant, capable of colonizing new areas. The final control technique applied in 2002 was hand-harvesting of the EWM. The initial hand-harvesting in Upper Thomas and Thomas Basins resulted in certified divers spending

164 hours removing 60,250 to 72,350 individual plants. A second hand-harvesting effort occurred in early October after follow-up inspections revealed a fair amount of new growth in Upper Thomas Basin and a small pioneering infestation in the main reservoir. The second effort resulted in diver hours and the removal of 18,250 to 25,550 individual plants. Only 14 individual plants were removed from the main reservoir. Total cost for the 2002 control techniques was $151,000, which included the purchase of benthic barrier and floating fragment barrier materials. In addition to the management techniques implemented in 2002, DWSP employees examined the feasibility of biological control of the EWM in Stillwater Basin using the milfoil weevil. Based on the status of the Stillwater Basin infestation and the recommendations of a milfoil weevil specialist, plans were made to introduce the milfoil weevil (Euhrychiopsis lecontei) into Stillwater Basin in The plan called for the introduction of 10,000 weevils as eggs and larvae to help control the infestation of EWM in Stillwater Basin. The success of milfoil weevils varies greatly with individual waterbody, but reproducing populations may have the ability to control an infestation EWM Management Preliminary macrophyte surveys of Thomas Basin in May enabled DWSP staff to target areas for hand-harvesting of EWM in Upper Thomas Basin and Thomas Basin proper. In 2003, a total of diver hours were logged and an estimated 3,251 individual plants were removed. The post-harvesting surveys indicated nearly total control of EWM in Thomas Basin proper. In addition, no EWM was identified in the main basin of the reservoir. DWSP staff re-deployed the floating fragment barriers in the same locations as 2002 to reduce autofragment dispersal, although barriers were not deployed immediately downstream of the benthic barriers. The total cost for milfoil control in 2003 was $12,500, mainly labor cost, as materials from 2002 were still useable. A survey of Stillwater Basin was conducted in June to select a milfoil weevil stocking site and an unstocked, or reference site. Approximately 10,000 milfoil weevils were stocked in Stillwater Basin in June 2003, mostly in the form of eggs and larvae. A post-stocking survey was conducted in August and lab analysis of milfoil stem samples indicated the presence of a pre-existing milfoil weevil population. Analysis at the lab indicated 100% stem damage at the stocking site, up from 43%, while the reference site showed 30% stem damage, up from 13% pre-stocking. Total costs for purchase and stocking of the weevils in 2003 was $20, EWM Management Preliminary macrophyte surveys of Thomas Basin in May enabled DWSP staff to target areas for hand-harvesting of EWM in Upper Thomas Basin and Thomas Basin proper. In 2004, a total of diver hours were logged and an estimated 7,424 individual plants were removed. In 2004, the non-native plant fanwort, Cabomba caroliniana had also grown in abundance and was removed along with the EWM. Similar to 2003, post-harvest surveys in 2004 indicated nearly total control of the milfoil, and no EWM was identified in the main basin of the reservoir. DWSP staff re-deployed the floating fragment barriers in the same locations as 2003 to reduce autofragment dispersal. Total cost for these control techniques in 2004 was $15,500. A survey of Stillwater Basin was conducted in July to assess the milfoil weevil stocking results. Qualitative observations of the stocking site indicated a decline in milfoil abundance. 144

165 Observations of the control site were similar to those in In 2004, the lab analysis of milfoil stems from the stocking site showed increased presence of weevil life stages compared to However, analysis of stems from the control site was inconsistent with data from 2003; in 2004, weevil damage and life stages were absent from the control site EWM Management Preliminary macrophyte surveys of Thomas Basin in June enabled DWSP staff to target areas for hand-harvesting of EWM in Upper Thomas Basin and Thomas Basin proper. An additional hand-harvesting event occurred in August in Upper Thomas Basin, Thomas Basin proper, and the main reservoir in Powerline Cove. In 2005, a total of 97 diver hours were logged and an estimated 4,847 individual EWM plants were removed, with only 21 individual EWM plants removed from the main reservoir. Again in 2005, the non-native plant fanwort was removed along with the EWM. Post-harvest surveys indicated effective control of EWM along the shoreline in Upper Thomas Basin, but significant regrowth in the central area. No additional plants were identified in the main reservoir. DWSP staff re-deployed the floating fragment barriers in the same locations as 2003 to reduce autofragment dispersal. Total cost for these 2005 control techniques was $13,200. In 2005, the bottom barriers were vacuumed to remove accumulated sediments. Six panels were removed to expose new sediment so it could be re-colonized in 2006, facilitating monitoring of native vs. non-native re-vegetation. The concern was that accumulated sediments on top of the bottom barrier might act as suitable substrate for re-colonization by the EWM. After cleaning, the bottom barriers were inspected and secured as necessary. The cost associated with cleaning and reinstalling the bottom barriers was $30,000. In addition to the control techniques described above, a dredging feasibility assessment was performed for Stillwater Basin. A private consultant concluded that the dredging would be expensive, and would involve a lengthy permitting process which could take between 2 and 10 years to complete at a cost of around $200,000. Dredging would require construction of a 10,000 square foot stone staging area for heavy machinery on the shoreline with an associated cost of $60,000. The cost of the actual dredging was estimated at $2.6 to $4.2 million dollars. Due to the costs and timeframe associated with dredging and the uncertainty of re-growth, dredging Stillwater Basin was deemed inappropriate EWM Management Preliminary macrophyte surveys of Thomas Basin in May enabled DWSP staff to target areas for hand-harvesting of EWM in Upper Thomas Basin and Thomas Basin proper in July. An additional hand-harvesting event occurred in August in Upper Thomas Basin and Thomas Basin proper. In 2006, a total of 174 diver hours were logged and an estimated 6,937 individual EWM plants were removed. Again in 2006, the non-native plant fanwort was removed along with the EWM. In 2006 the number of fanwort plants removed was greater than EWM for the first time, although EWM abundance did not appear to be negatively impacted. This suggests these plants are not competing for the same resource space. Post-harvest surveys in September indicated no significant amounts of EWM or fanwort. No additional plants were identified in the main reservoir. DWSP staff re-deployed the floating fragment barriers in the same locations as 2003 to reduce autofragment dispersal. 145

166 The area of bottom barrier removal in 2006 was re-colonized by an abundance of native plants with scattered invasive species. The initial result of this experiment is promising because it does not appear that non-native plants have an edge in the re-colonization of newly exposed sediments in this area. The total cost of the 2006 control effort was $15,500. Overall Result of Management The control program for Wachusett Reservoir, consisting of hand-harvesting, benthic barriers and floating fragment barriers, appears to be adequate to prevent the spread of EWM from the upstream infested basins into the main body of the reservoir downgradient. The DWSP staff is continuously monitoring the basin for EWM plants while completing other in-lake tasks in the reservoir. The commitment of time and money by DWSP and MWRA staff and state agencies allows the current control program to limit the spread of the EWM. Successful colonization of the milfoil weevils in Stillwater Basin may reduce the amount of new infestations in Upper Thomas Basin and Thomas Basin proper. One major concern of the parties involved with controlling the spread of the EWM is that one day hand harvesting will not be sufficient to control the spread of EWM and additional control techniques will be needed. As acceptable options are limited, the next phase of possible action is uncertain. Lessons Learned and Recommendations Threat Analysis A threat analysis matrix for Wachusett Reservoir is supplied in Table 8-1. Overall, the threat of infestation by EWM based on suitable habitat is high. The total area potentially covered is only about 11% of the reservoir s surface area, but 430 acres is a substantial area that could affect at least localized water quality. Additionally, limited treatment of water from this generally clean and unfiltered system means that added organic matter represents a threat of formation of undesirable compounds (e.g., trihalomethanes, halo-acetic acids) upon disinfection with chlorine. Fortunately, the potential for successful control through rapid response activities is also high. Limited human activities on the reservoir reduce the threat of introduction or spread by many vectors, leaving downstream movement with flow as the primary means of transport to be addressed. Additional Data Needs The management group has properly assessed the situation and remained involved in monitoring to an appropriate degree. Additional data needs for this reservoir, as relates to EWM infestation and control, revolve around the presence and nature of any sensitive or protected species. In Massachusetts, lake management techniques can be severely limited by the presence of protected species that might be impacted by a chosen technique. As the current approach is providing reasonable control, no additional study may be needed. Rapid Response Assessment In the early stages of a milfoil infestation the array of viable control techniques available is much greater than subsequent years when the density and abundance of the milfoil increases. 146

167 In general, as infestations increase in size and density, aquatic herbicides, drawdown and dredging become the only three viable options for eradication or extreme control on a large scale. Action was taken in this case early enough to prevent rapid expansion into the main basin of Wachusett Reservoir, but not soon enough to eradicate EWM in the pre-reservoir basins. Table 8-1. Threat analysis matrix for EWM in Wachusett Reservoir. FACTOR YES NO THREAT EVALUATION HIGH MEDIUM LOW A large area could Extent and speed of possible X be affected infestation Plant density could be high X X Spread could be rapid X Water supply may be impacted X Swimming may be impacted Boating may be impacted Fishing may be impacted Aesthetics may be impacted Sensitive species may be impacted Protected species may be impacted Spread by water flow likely Spread by birds likely Spread by boating likely Spread by other human activities likely Eradication is possible Confinement is possible X X?? X X X X X X X? Nature of possible impacts No filtration system, Disinfection only Limited treatment, EWM effects felt in water Sensitive species may be impacted Minor impacts to other human activities Ability to spread No human contact No boating allowed Potential success of rapid response X X X Annual macrophyte assessments X Typically, procurement of funding, bidding projects, and obtaining permits (where necessary) can take a significant amount of time to accomplish. Through the exercise of creating a RRP, drinking water managers should have identified funding limitations, permit issues, and 147

168 other problems associated with various control techniques specific to the waterbody in question. If an infestation is identified and a RRP is already in place, these issues should not hinder the ability to implement selected control techniques. To ensure EWM does not become established in the main reservoir, DWSP personnel have implemented a pseudo-rrp in conjunction with the annual removal of milfoil from Upper Thomas Basin and Thomas Basin proper. DWSP employees are continuously inspecting the main reservoir for EWM whenever they are performing duties on the water. Any plants identified by DWSP employees are hand-harvested when harvesting is occurring in Upper Thomas Basin and Thomas Basin proper. Hand-harvesting of EWM plants has occurred in the main reservoir, but the quick response after identification of new plants has resulted in no known established growths within the main reservoir area. Continuation of the current plans appears to be adequate as a preventative measure to protect the main reservoir from EWM colonization. Longer Term Response Assessment DWSP and MWRA should continue to monitor the occurrence of EWM in Wachusett Reservoir and the western basins of the Stillwater River leading into it. In the event that an infestation is identified in the main body of the reservoir, DWSP employees should respond quickly with control techniques. DWSP and MWRA should examine and be ready to implement additional, more severe control techniques if the current infestation becomes unmanageable with hand-harvesting or a large scale infestation is discovered in the main reservoir. RESERVOIR A Reservoir A is a natural lake of slightly more than 400 acres that serves as a drinking water reservoir. It located in the northeast United States, but exact details of location have been withheld at the request of the utility, which has strict security policies. Consequently, there are no maps in this case study. Raw water entering the lake consists of 74% diversion from a river or transfer from other supplies, 17% from groundwater and direct overland flow, and 9% as backflow from a downstream waterbody at high downstream water levels. Reservoir A is less than 500 surface acres in size with maximum and mean depths of 115 and 48 feet, respectively. The maximum reservoir volume is less than 6 billion gallons with a daily withdrawal of <10 million gallons per day. The watershed of Reservoir A is 15.1 square miles, including areas from which diversions are made, and there is no residential development along the shoreline. About 23% of the watershed is owned by the operating utility and completely forested. The exact breakdown of land use in the watershed is unknown, but residential/urbanized land, cultivated land, livestock use areas, wetlands and lakes, and less human-impacted land uses all exist within the watershed area. Public access to the reservoir shoreline is restricted to fishing from shore, hiking and biking by permit. One boat ramp exists where the public may rent boats and electric motors to fish. Utility employees and state agencies may use small horse power gasoline outboard motors for project work. Approximately 64 surface acres, or 17% of Reservoir A, is within the 0-15 foot contour, and could host significant milfoil growths. Bottom type within the 0-15 foot contour consists of 75% silt/muck, 20% sandy sediment, and 5% rock. Therefore, 13 acres possess ideal depth and sediment type for colonization by Eurasian water milfoil (EWM), and 48 acres appear ideal for colonization by variable water milfoil (VWM), although some overlap is certainly possible. 148

169 Alkalinity is moderate, however, so VWM would not be expected in this aquatic system, based on its water quality preference. Aquatic macrophytes in Reservoir A can be found growing out to depths of 20 feet, although dense growths are rare at deeper than 12 ft. Within the 0-15 foot contour, aquatic macrophytes cover 50-75% of the bottom and occupy 50-75% of the water column. Additional water uses and potential transport vectors within Reservoir A are limited. Direct contact (i.e., swimming) by the public in any form is not allowed. Shoreline access is granted to allow for fishing, hiking and special permit biking, but all well away from the intake area. The one boat ramp is operated by utility employees, private boats are not permitted in the reservoir, and access is restricted; boats are not allowed in the southern portion of the lake, where the intake is located in ft of water. The use of private boats on the reservoir was eliminated in 1990 due to growing concern over possible zebra mussel introduction into the reservoir. Due to the limited number of additional uses within the reservoir, the only transport vectors for aquatic macrophytes are water flow, waterfowl, and boating. Milfoil Assessment and Control Reservoir A currently has one species of milfoil, EWM. Although the actual date of initial colonization in the reservoir is unknown, the presence of EWM is acknowledged as far back as the early 1980s according to utility personnel. In 1983 the reservoir was opened to the public at the request of the state, to allow for fishing. When fishing access was allowed, there was a perceived need for EWM control around the access point to limit transport out of and within the lake, although no actual control techniques were implemented at that time. An assessment of the macrophyte community around the public fishing access and boat launch area was completed in 1988, along with an evaluation of possible control techniques. One concern at the time, which is still a concern today, is clogging of the intake and pumping mechanism between the reservoir and the treatment facility; the intake is not in deep water. One recommendation of the 1988 report that was implemented was the extension of a floating fishing dock beyond the depth covered by the milfoil. No additional alternatives or control techniques were implemented at that time. Continued issues with EWM at the boat access area prompted additional consideration of control at that site in An initial quote was obtained from an outside contractor to cut and harvest up to one acre of EWM for $2,650. The utility rejected this bid and instead decided to complete the cutting and harvesting with in-house personnel. In the summer of 2003 a hand cutter was purchased for $300, and the first cutting operation occurred in August of In 2003, a semi-annual hand cutting program began and has continued through 2006 with plans to continue every year hereafter. Each cutting day consists of two utility employees for 3 hours each day, for a total of 6 man hours per cutting day. Removed plants are not quantified, but each cutting day results in enough removed plants to fill a small pick-up truck. This technique is adequate to keep the boat dock area clear, but does not eliminate the problem or prevent its expansion. A secondary concern of utility personnel is the apparent association of blue-green algal mats with colonies of EWM and the resulting taste and odor issues. The utility hired an environmental consultant to survey for algal mats in the south end of the lake, an area of about 17 acres near the drinking water intake and less than 25 ft deep. Taste and odor causing benthic algal mats were found in association with EWM and coontail (Ceratophyllum demersum) growths in Reservoir A, which tend to occupy a peripheral band in 5-12 ft of water depth. When 149

170 coverage by blue-green algae mats (mainly several species of Oscillatoria) in this area exceeds about 20%, taste and odor becomes detectable in the raw water at the treatment facility. Most mats are found in the areas of densest plant growth, but not exclusively in association with EWM. EWM also produces a large amount of periphytic algal growth, but these growths have not been implicated in the taste and odor incidents. Between 1998 and 2004, approximately $50,000 was spent on assessment and on treatment of the areas of potential algal mat growth with aluminum sulfate to bind up phosphorus in surficial sediments and limit mat development. The first treatment, in 1999, provided tangible results for about two years. Testing of the sediments in 2001 indicated that sufficient phosphorus had either been deposited in the target area since the treatment, or had not been completely inactivated (both are distinct possibilities). A second treatment in 2002 limited algal mat development through at least Coupled with enhanced treatment at the associated water treatment facility, the utility believes that the taste and odor problem can be handled without any direct action against rooted aquatic plants. A survey of the aquatic macrophyte community by a state extension agency in 2004 indicated that EWM was the dominant plant in the reservoir at depths less than 10 feet. The east side of the reservoir can experience 100% coverage by EWM in the 2-10 foot contour. The west side tends to be steeper, with very coarse sediments and fewer plants. EWM can be found in up to 60 acres of the lake, but is abundant in no more than 16 acres, based on assessment from the map generated in This matches fairly well with the estimates based on water depth and substrate type offered previously. Another 12 species of aquatic macrophytes were present in 2004, but of these only coontail achieved potential nuisance densities. The abundance of EWM decreased in the area of the intake pipes where steep slopes and sediment features are less favorable to EWM growth. It appears that EWM is growing everywhere that it can in Reservoir A, with minimal potential for further expansion. It is limited in shallow water by fluctuating water levels and in water >12 ft deep by low light, as this reservoir is productive and subject to algal blooms. EWM is abundant over <4% of the lake area. The volume of Reservoir A that is potentially directly affected represents <1% of the total volume. Lessons Learned and Recommendations Threat Analysis There is only a limited perception that EWM has to be controlled or eradicated in Reservoir A. This perception results mainly from recreational interference with limited potential impact of the current infestation on drinking water quality and utility operation. The nature of the reservoir banks and sediments limit colonization by EWM. If the nature of benthic habitat in the intake pipe area was conducive to EWM colonization, the perceived need for control might be much greater, but this is not the case. Recently enhanced treatment capabilities in the water treatment plant, including activated carbon injection, add a measure of protection as well. A threat analysis matrix for Reservoir A is supplied in Table 8-2. Overall rate of expansion beyond current conditions and extent of the possible infestation for the reservoir are low. Optimal conditions for EWM exist in only 13 acres by prediction, and the actual measurement of area dominated by EWM in 2004 is estimated at 16 acres; depth will limit growth in many areas. The potential for possible impacts and the ability of the milfoil to spread are both low, although not negligible. The primary concerns from a water supply standpoint are 150

171 the potential for the milfoil to clog the pumping mechanisms between the reservoir and the treatment facility and the addition of organic matter or other substances that might affect treatment needs; neither appears to represent a major threat in this case. Recreational impacts would be limited to fishing and boating in the northern portion of the reservoir, and then mainly to shoreline fishing and actual launching of boats. The potential for rapid response success for this reservoir is low; the reservoir already contains a substantial amount of EWM, more than any normal rapid response technique could address. Chemical controls could be employed, but are not viewed as acceptable in this drinking water supply situation. 151

172 Table 8-2. Threat analysis matrix for EWM in Reservoir A. FACTOR YES NO THREAT EVALUATION HIGH MEDIUM LOW A large area could Extent and speed of possible X be affected infestation Plant density could be high X X Spread could be rapid X Water supply may be impacted X Swimming may be impacted Boating may be impacted Fishing may be impacted Aesthetics may be impacted Sensitive species may be impacted Protected species may be impacted Spread by water flow likely Spread by birds likely Spread by boating likely Spread by other human activities likely Eradication is possible Confinement is possible X X X X X X X? X X X X Nature of possible impacts Potential to clog intake low Limited volume of reservoir directly affected Treament of water, so water impacts not felt Minor boating and fishing impacts Minor impacts to other human activities Some concern over possible link of plants and algal mats that cause taste and odor Ability to spread Only boating allowed and restricted areas are in place Potential success of rapid response X X X Annual macrophyte assessments X Additional Data Needs There are no immediate additional data needs for this reservoir relating to EWM infestation and control. It would be appropriate to elucidate the potential link between EWM (or other rooted plant growths) and benthic algal mats that cause taste and odor, but past management and current treatment capacity limit the extent of any threats from such algal mats. 152

173 Rapid Response Assessment An active Rapid Response Plan (RRP) would have been beneficial prior to the initial colonization of EWM. There are no known upstream colonies of EWM, but EWM does exist in other waterbodies within 5-10 miles, so introduction to the reservoir by some means was to be expected. A RRP with a pre-determined set of control techniques may have been able to eradicate the pioneering plants. As the initial colonization occurred at least 25 years ago, a RRP for Reservoir A would only be useful now to help identify potential control techniques that could be implemented if an increase in EWM was observed in key areas, like the southern end of the lake, where the intake is located. The process of creating a RRP to address these issues would ensure that utility managers were already aware of potential drawbacks (e.g., toxicity) and holdups (e.g., any permitting by the state) associated with any new control techniques that may be needed. This does not appear to be a high priority action in this case, however. Longer Term Response Assessment Aquatic macrophyte surveys on an every other year basis are recommended to track coverage by EWM and other invasive species in the reservoir. In a system where no routine macrophyte surveys are conducted, colonies of pioneering plants or increased distribution of already established plants may go unnoticed. Routine surveys may identify the potential for control techniques to address a new or existing issue before it begins causing problems for drinking water managers. As with the RRP, however, there does not appear to be a strong impetus for such an effort in this system. SEBAGO LAKE Sebago Lake is natural lake reservoir located in Cumberland County, Maine, in the Presumpscot River Basin. The lake is used year round as a drinking water supply. Direct flow from tributaries accounts for 70% of the total raw water source, with 30% direct groundwater inflow. Sebago Lake consists of 30,000 surface acres with maximum and mean depths of 312 and 100 feet respectively. Maximum reservoir volume is 978 billion gallons. The watershed area for Sebago Lake is approximately 300,000 acres, with an unquantified but substantial number of residences along the 90 mile shoreline. The majority of the watershed area was forested (85%) as of 1990, but there are residential and commercial zones and transportation corridors as well. In addition to the many private boat launches, three major and six smaller launches are available to the public. Roughly 10% of the lake is contained in a restricted area where bodily contact is prohibited, and there is a smaller zone from which all recreation is excluded. The Portland Water District (PWD) patrols this area and manages the water supply, but does not have control over the entire lake or watershed. Human activities allowed on the unrestricted portion of the lake include swimming, fishing, passive recreation, and boating. There are no restrictions on the size of boats or the horsepower of motors. Due to the size of Sebago Lake, there is a fairly large amount of suitable habitat for milfoil colonization, but it represents a small portion of lake area overall. The 0-15 foot depth contour is less than 5% of the total lake area, but this is still 1500 acres. With exceptional water clarity in Sebago Lake, plant growth at up to 30 ft of water depth is quite possible, but growths are generally sparse. Bottom type consists of 50% rock, 30% silt/muck, and 20% sandy sediment. Variable water milfoil (VWM) favors silt/muck sediments while Eurasian water 153

174 milfoil (EWM) prefers sandy sediments, although there can be overlap in habitat occupied. Given sediment preferences of these two species, about 450 acres are favorable for VWM colonization and about 300 acres are favorable for EWM colonization, although some additional, deeper area of colonization would be possible in this clear lake. EWM is known from only one lake in Maine, and the third invasive milfoil species, parrotfeather, has not been found in Maine, so the primary species of immediate interest is VWM. Sebago Lake water levels fluctuate during the year due to hydropower generation. Lake level drawdowns discourage milfoil growth in areas <5 ft deep. Historically, with drawdowns of as much as 20 ft, the restriction would have been greater, but water levels are now more tightly controlled under the power company s license. Cover and biovolume of aquatic plants in the 0-15 foot contour are not precisely quantified for Sebago Lake, but are known to be low relative to most other lakes in the region. The drinking water intake at Sebago Lake is in very deep water, well below the depth at which plants grow, so there is no direct threat to the intake from aquatic macrophytes. Recreational uses of Sebago Lake increase the number of potential transport vectors for milfoil. The major transport vector in Sebago Lake is recreational boating. Fragments of milfoil attached to boats, motors and trailers, or in bait wells of boats, may result in multiple introductions to the lake and may aid movement around the lake. Water flow and waterfowl may also contribute to the spread of milfoil. In particular, there is a lock on the Songo River that allows passage between Sebago Lake and several upstream waterbodies, and the movement of boats and water through that lock is a known route of invasion. Variable water milfoil is the only milfoil species currently found in Sebago Lake. VWM was first positively identified in the early 1970s. At that time the awareness of milfoil species and their invasive nature was low, and the plant was only identified because someone reported it as being odd for Sebago Lake, where relatively few plants were found in general at that time. There is a high likelihood that VWM existed in Sebago Lake prior to that report, with accidental introduction by boat as the likely source, but no records are known. Milfoil Assessment In 2000, the Portland Water District (PWD) began to assess the VWM infestation through the use of GPS with the assistance of lake residents and users. When reports of VWM were reported to the PWD, PWD personnel would investigate, confirm the report, and use a GPS to mark the location. This GPS mapping of the VWM colonies continues to date (Figure 8-3), but a true macrophyte survey has never been completed. Although the VWM colonies are concentrated in the northern portion of the lake and the water intakes are in the southern portion of the lake, PWD personnel began examining management actions to control the VWM at specific locations and prevent its spread in During the fall and winter of , management techniques were considered and selected for implementation in the spring and summer of The impetus for this effort does not come from any perceived threat to the water supply, but rather from concern by the PWD and lake users over the impacts of VWM on lake ecology and recreational uses. Five management techniques were considered for use at the public boat launch in Sebago Lake State Park. Herbicides were rejected as a consequence of public opinion and the Maine Department of Environmental Protection (MDEP) stance on herbicide use in a public water supply such as Sebago Lake. Harvesting by commercial sized machines and cutting with even handheld devices were rejected as options as a consequence of the potential for fragments to be 154

175 generated and cause further spread of VWM. Benthic barriers were deemed an appropriate option for control but were not implemented at the boat launch area, mainly out of concern for 155

176 Figure 8-3. Variable water milfoil distribution in Sebago Lake as of

177 possible fouling of propellers, although expense may have also been a factor. The final decision was to use hand harvesting by divers to remove the VWM in With the passage of legislation focused on preventing invasions in Maine, the state became more actively involved in invasive plant management in A major goal of the PWD and state agencies working in Sebago Lake is to prevent the spread of VWM to other waterbodies. One surface supply of water to Sebago Lake is the Songo River. The Songo River enters Sebago Lake at the north end and flow can be interrupted by operation of the oldest lock in the country. The first positive identification of VWM in the Songo River occurred in 2004 by PWD personnel. Movement into this area was apparently upstream through the lock. A coalition of agencies and more local interest groups has worked to slow its advance in the upstream direction where there are several susceptible lakes Brandy Pond and Long Lake. The gradual movement of VWM upstream from the lock has been documented over three growing seasons, despite some success with control efforts. To date, plants have been located in one marina in Brandy Pond, and control measures have been used to eradicate this colony. Management of VWM in the Songo River is a test of the ability to build appropriate coalitions and effectively address invasive plant issues. Management Efforts Sebago Lake VWM Management In 2001, all boat launches on Sebago Lake were posted with warning signs to help boat owners identify VWM and other invasive species, and recognize the importance of inspecting boats and trailers for plant material. Also in 2001, MDEP funded boat inspections at select boat launches around the lake, with PWD conducting the inspections. Standish boat launch is approximately 0.75 miles from the water intake on Sebago Lake, and is a heavy traffic boat launch area. The Standish boat launch is the only boat launch in close proximity to the drinking water intake, and the PWD has had an inspector in this area since the 1970s to discourage bodily contact with the water. In 2001, the PWD added boat inspection to the duties of the inspector stationed at the Standish boat. The inspector works 8-10 hours per day, seven days per week, inspecting boats for milfoil and other invasive plant species between Memorial Day and Labor Day. The annual cost of the inspector is $12,500. In 2002, the PWD in conjunction with the MDEP and MDOC completed hand harvesting at the Sebago Lake State Park boat launch, a heavily infested area 12 miles away from the water intake. The MDOC marked a boat channel to ensure boats weren t traveling through the heavily infested areas, and divers were used to hand pull plants around the launch area and inside the marked channel. The hand pulling effort lasted two days and cost approximately $3000. Two additional hand pulling efforts have occurred since the initial effort in One half-day event occurred in 2003 and another half-day event in An additional source of milfoil control and prevention is the large number of local associations on Sebago Lake. As opposed to one large lake association, most interest lies within multiple smaller associations linked to a part of the lake by home ownership. Approximately 3-5 associations are currently doing work to prevent or control VWM each year on Sebago Lake. These associations are implementing control measures at beaches, canals, boat launches and channels. While the majority of these efforts are privately funded, the MDEP does award grants 157

178 to assist associations with their control activities. Benthic barriers and hand pulling are the two techniques utilized by these private associations. The total area treated, the number of plants removed, and the associated costs of these efforts are not known. Songo River VWM Management After VWM was identified in the Songo River in 2004, control measures were implemented with the hope of eradicating the infestation. A quantitative assessment of the 0.5 miles upstream of the Songo lock was funded by the PWD. A GPS map was created of the surveyed area, and the infestation was estimated to be three acres in size, in scattered plots with varying density. River level drawdown was attempted to control the VWM infestation, as Sebago Lake is drawn down by about 5 ft by late autumn and it was thought possible to lower the river by keeping the lock open. However, results were not satisfactory due to the limited magnitude and length of the drawdown. The lock on the Songo River was not opened far enough to allow adequate water level drawdown to expose the VWM. It is not clear that a better arrangement could not be reached, but alternative methods were sought. The use of herbicides, specifically 2,4-D, was discussed at length among all interested parties, but was not pursued as a function of the public perception that such use could create in a drinking water supply. As 2,4-D is the only currently registered herbicide with a proven record of success for VWM, this was the only herbicide given serious consideration for use in the Songo River. While dilution alone should be sufficient to prevent any ill effects in Sebago Lake and no detectable concentration would be expected at the PWD intake, 2,4-D is rarely used in drinking water supplies or lakes with hydrologic connections to supply wells as a function of human health risk. While some local residents favor a chemical approach, the MDEP, PWD and certain other local groups oppose this approach without first exhausting alternative options. In 2005 a combination of benthic barriers and hand pulling was used to control the infestation. Polyethylene tarps (40 ft X 60 ft) with holes to allow venting and anchors at the corners and weights in more central locations were placed over the more dense infested areas. In addition, hand harvesting was applied to remove EWM plants in less dense areas. These techniques were repeated in 2006, and currently 16 tarps are being used to control the VWM in the Songo River. Also in 2006, a suction harvester was built for use in the Songo River. It was tried out in 2006, but is expected to represent a bigger part of the control effort in The cost of the control effort in 2005 and 2006 was approximately $60,000, with $15,000 used to build the suction harvester. Some control was achieved, but not eradication. Continued effort is needed. Overall Result of Management Sebago Lake may seem like an unusual case because of the strong perceived need for VWM control by most stakeholders involved, despite minimal threat perceived by the PWD to drinking water quality or supply. The MDOC and MDEP are concerned that Sebago Lake will act as a source population of VWM that will spread to other waterbodies throughout the state of Maine. Local stakeholder groups are either concerned with current growths near their properties or key use points, or are worried that VWM will reach their location and impair recreational use and ecological function. This is, however, not unusual at all, and in the course of evaluating possible case studies, this pattern was one of the most common. Lake George in New York has a 158

179 nearly identical situation with EWM. This creates a management problem, as most control techniques carry some risk of impact to non-target species or resources, and utilities have a very low tolerance for risk when there is no perceived threat to water supply. Water level control and herbicide application are controversial issues directly related to macrophyte control and strongly debated among the stakeholders at Sebago Lake. In the past, Sebago Lake was lowered by as much as 20 feet as a combination of drought and generation of hydro-power, although the normal water level fluctuation is about five feet. Drawdown restricts VWM colonization and establishment, while more stable water levels favored by some stakeholders will favor greater coverage by VWM. In Maine, only the MDEP or its agent may apply herbicides, and any treatment of a Public Water Supply (PWS) must be accompanied by consent of the operating utility. Herbicides have been proposed to control VWM in the Songo River, and could be applicable in other parts of Sebago Lake, but the preferred herbicide for VWM control is 2,4-D, a chemical that does represent a health risk in a water supply at the recommended dose. Dilution and other mechanisms could greatly reduce risk, but there will always be some risk associated with use of 2,4-D. The water utility did not offer its consent and the MDEP did not authorize the use of herbicides, although both have been supportive of alterative control efforts. The perception of risk to a major supply is a powerful motivator in decision-making. The main efforts to manage VWM in Sebago Lake have focused on education of the public and minimization of the spread of VWM. Posting of signs is intended to make lake users aware of the threat and hopefully elicit behaviors that reduce the chance of spread of VWM. Hand pulling near the state park boat launch is intended to limit removal of VWM on boats or trailers leaving the lake. The effort in the Songo River represents the most intensive action to control an existing population of VWM, and yet only limited techniques and resources have been brought to bear on the problem in that location. Blue poly tarps have been applied instead of benthic barrier made for this purpose to lower costs. Drawdown was attempted only once and without complete planning, apparently a consequence of difficult coordination among parties which need to be involved. Herbicides have been rejected out of perceived risk without a complete evaluation of possible approaches. It will be difficult to control VWM or any invasive plant without a more comprehensive program. Lessons Learned and Recommendations Threat Analysis A threat analysis matrix for Sebago Lake is supplied in Table 8-3. Overall, the potential extent and spread of the infestation is high because the size of the area that could be infested is large, even though this area is only a small percentage of the lake area. Although the water supply is not threatened by the VWM, other human uses such as swimming and boating are threatened, especially in confined cove areas (e.g., Kettle Cove to the northeast). Consequently, the nature of the possible impacts is considered medium. High boat traffic on the lake and prevailing inflows and currents results in a high potential for the plant to spread to other parts of the lake. The potential success for rapid response is low because of the high number of transport vectors, an apparent inability to confine the new growths, and limitations of the current survey program. VWM has already spread to enough places in Sebago Lake to put this lake beyond the rapid response phase overall, although rapid response to new growths is warranted and should be pursued with vigor. 159

180 Additional Data Needs Although there is little concern over the impact of VWM on the water supply, a clear understanding of the current location and magnitude of the VWM infestation may be useful to all stakeholders. The PWD, state employees and any local groups that can be trained through the Volunteer Monitoring Program should continue to monitor the growth and spread of VWM in Table 8-3. Threat analysis matrix for VWM in Sebago Lake. FACTOR YES NO THREAT EVALUATION HIGH MEDIUM LOW A large area could Extent and speed of possible X be affected infestation Plant density could be high X X Spread could be rapid X Water supply may be impacted X Swimming may be impacted Boating may be impacted Fishing may be impacted Aesthetics may be impacted Sensitive species may be impacted Protected species may be impacted Spread by water flow likely Spread by birds likely Spread by boating likely Spread by other human activities likely Eradication is possible Confinement is possible X X X X?? X X X X X X Nature of possible impacts No threat to water supply, intakes too deep, dilution too great Sensitive species may be impacted Substantial impacts to human activities are possible Note that some forms of fishing might actually be improved by greater plant density in this lake Ability to spread All possible transport vectors exist Recreational activities with high potential for spread Potential success of rapid response X X X Annual macrophyte assessments X 160

181 Sebago Lake and its tributaries. An attempt should be made to conduct macrophyte surveys at least every two years. If whole-lake macrophyte surveys are not possible due to the effort required on such a large lake, macrophyte surveys should be concentrated in the areas of the lake where depth and sediment features are optimal for VWM. Continued monitoring of the Songo River and upstream waterbodies for the spread of VWM from Sebago Lake is especially warranted. Additional stewards at boat ramps are also needed to evaluate actual movement of VWM by boats. This has worked well in other situations to document the perception that boat transport is a prime vector, but some site specific quantification is desirable. These stewards also raise awareness among boaters and minimize the potential for spread of VWM. Rapid Response Assessment The state of Maine currently has a statewide Rapid Response Plan (RRP) for non-native invasive plant species. A RRP specific to Sebago Lake or the Songo River does not exist. A RRP for the Songo River may be a bit late, but one that includes upstream Brandy Pond and Long Lake would be beneficial to guide actions quickly and avoid delays of even a season in controlling newly discovered growths of VWM. A RRP for Sebago Lake would similarly allow for pre-selection of control techniques based on costs, funding, permits, size of the new infestations, and physical characteristics of the area of new infestation. Speed of response is important to success, and delays in action to the year following discovery of a new infestation should be avoided wherever possible. Longer Term Response Assessment Current management techniques should be supported, and efforts to eradicate new infestations should be maintained or increased. It is a good sign that no new areas of growth were detected in 2006 (Figure 8-3), but it is not clear whether this is a result of proper management or incomplete survey. Additional or more extreme controls should be examined and discussed before a situation arises in which they are proposed (e.g., as has already occurred in the Songo River). This will be controversial, and there is a need for all interested parties to set personal or organizational priorities aside and work as a group to determine what techniques can or cannot be applied. This does not mean that there will be some democratic process with a majority vote, and it is not meant as an opportunity for one group to strong-arm another into an unfavorable decision. Rather, it should be viewed as an opportunity to look at all factors (scientific, economic, social and regulatory) that have bearing on each technique and determine the controlling influence on a decision to use or avoid each method. The use of Sebago Lake as a drinking water supply and the apparent absence of a major threat to that supply from VWM is perhaps the most critical consideration in any assessment of longer term response to the VWM invasion. The PWD has been actively supporting assessment and control of VWM, but cannot be expected to accept risk (actual or perceived) of impact to the water supply to control an infestation that has a negligible probability of influencing water supply. It is therefore incumbent on all parties to look for ways to minimize that risk (or perception of it) if more severe (i.e., herbicides) controls are to be considered. On the other hand, the use of physical techniques has merit in this system, but requires a greater level of funding, commitment and speed of response than has been witnessed to date. The prevention of the spread of VWM is a goal that all parties appear to agree on. In addition to the data collection and educational efforts underway but in need of augmentation, an 161

182 additional preventive technique that could be employed at all boat launch areas is mandatory use of a washdown station. Boat owners would be required to wash down their boats and trailers and empty all bait and live wells. This could be for incoming or outgoing boats, but may not be justified until data are collected to indicate the clear risk associated with specific boat ramps on Sebago Lake. Data collection at each ramp is therefore emphasized as a need. LAKE MASSABESIC Lake Massabesic is a natural lake located in Auburn, New Hampshire in the Merrimack River Basin, and is used year round for water supply, operated by Manchester Water Works (MWW). The raw water supply for Lake Massabesic is derived entirely from tributaries. Lake Massabesic consists of 2560 surface acres with maximum and mean depths of 51 and 17 feet, respectively, and a highly irregular shape (Figures 8-4 and 8-5). The maximum reservoir volume is 15 billion gallons, while the average withdrawal for water supply is 17 million gallons per day. The drinking water intake is located in feet of water, within the range of depths that milfoil will colonize. The watershed of Lake Massabesic covers 46 square miles. Current land use consists of 60% forest, 23% wetlands and lakes, and <10% residential/urbanized. Some shoreline residences are present; the percentage of developed shoreline was not available, but development is not intense. The MWW controls the reservoir; public access to the lake is permitted, but some uses are restricted. Three boat launches provide access for general boating and fishing. Personal watercraft, sailboards, and inflatable craft are not allowed on the lake. Direct contact activities (i.e., swimming, waterskiing) are not permitted on Lake Massabesic. The physical characteristics of Lake Massabesic are suitable for growth of aquatic macrophytes, including milfoil species. Within the 0-15 foot contour, 25-50% of the bottom substrate is currently covered with aquatic macrophytes, of which there are 17 species. Aquatic macrophytes can be found growing out to depths of 15 feet, but high color and occasional algal blooms in this lake restricts light penetration, so dense growths are not common at water depths greater than 10 ft. At least 60% of the total surface area, or 1,536 acres, is located within the 0-15 foot contour, the zone considered in most analyses to be susceptible to dense milfoil growths. If a 10 ft limit on dense growths is accepted in this case, the total colonizable area based on depth for this lake might be reduced to about 1100 acres, or 43% of the reservoir area. Bottom type within the area of potential plant growth consists of 80% silt/muck, 10% sand, and 10% rocky substrate. Based on depth and substrate properties, at least 880 acres are highly suitable for variable water milfoil (VWM), and possibly as much as 1230 acres. About 89 acres are especially suitable for Eurasian water milfoil (EWM), but while EWM might survive in Lake Massabesic, water quality is not ideal for this species. The third invasive milfoil species, parrotfeather, is not known from New Hampshire. Consequently, VWM is the primary milfoil threat, and could cover more than a third of the lake area. Milfoil Assessment Currently, VWM is the only milfoil species inhabiting Lake Massabesic. The first VWM plants were positively identified in 1994, although earlier infestation was suspected and prompted planning by the MWW. Since 1994, VWM has spread to various parts of the lake and is approaching the intake area. Issues of water quality and possible intake clogging are of concern to MWW personnel. Additionally, Lake Massabesic already experiences swampy, 162

183 earthy/musty, and fishy odors in association with occasional algal blooms, and there is concern that VWM might exacerbate this problem. Figure 8-4. Lake Massabesic and its immediate watershed in New Hampshire. 163

184 Figure 8-5. Bathymetry of Lake Massabesic. Contour interval = 10 ft. 164

185 A whole-lake assessment of the Lake Massabesic was performed in 2000 by the New Hampshire Department of Environmental Services. Such assessments are performed about once a decade, not necessarily in response to invasive species introductions, but they do include evaluation of the plant community. Organized annual assessments to track the growth and spread of VWM in Lake Massabesic have not been performed, but the MWW does investigate reports of VWM growths, and the state Department of Environmental Services has taken an interest in the spread of VWM in the lake. The distribution of VWM as of 2006 is shown in Figure 8-6. There are four major areas of growth, although it is possible that undetected patches or scattered plants exist elsewhere in this large lake. The patch in the far south of the western basin has largely been controlled, although not eliminated. The patches in the connector channel between the eastern and western basins, near the Route 128 overpass, have also been greatly reduced in area, but some plants remain. The long finger of infestation on the eastern shore of the eastern basin is the most extensive infestation and includes the first area where milfoil was found in the lake, Clair s Landing. Management Effort The decision was made in 1992/1993 to begin management planning to control invasive plants, which were suspected of already being in the lake. The planning process took approximately one year to complete, and VWM was documented a year later. Four management techniques were considered to control the VWM in Lake Massabesic. Herbicides were rejected for use because the lake is a source for drinking water. Mechanical harvesting was rejected because of the high likelihood of regrowth and possible fragment dispersal. Hand pulling and benthic barriers were both deemed appropriate to control invasive growths. The chosen management techniques were first employed in 1996 at the Clair s Landing boat launch in Auburn, where the first infestation was found. The delay in response was partly a matter of budgeting and partly a lack of a sense of urgency by the MWW. Hand pulling and benthic barriers were used at Clair s Landing to remove VWM in that initial effort. Signs were erected at boat ramps to educate the public about the VWM threat in The MWW also roped off areas to restrict boats from the most infested areas in Benthic barriers have been used every year since 1996 to address VWM patches as they are found. Hand pulling was initially used to remove scattered plants and small patches of VWM, but has been abandoned due to the size and density of the current patches in Lake Massabesic. Approximately $25,000 of benthic barrier material has been purchased since 1996, and each year about $10,000 is spent replacing or augmenting the mats. In addition, utility personnel spend manhours each year on VWM control, mainly providing labor for the benthic barrier program. While actual pre- and post-control data are not available, the impression is that the control measures reduce VWM at a targeted site, but are generally ineffective for eradicating it. Efforts devoted to stopping the spread of VWM to new areas of the reservoir seem to have helped, but the lack of a comprehensive mapping and plant monitoring program limits conclusions about how well the infestation has been contained. Recently, utility personnel have been in contact with the New Hampshire Department of Environmental Services (NHDES) regarding possible alternative, non-herbicidal controls for VWM in Lake Massabesic. The soil chemistry in Lake Massabesic indicates high levels of phosphorus, which may facilitate rapid expansion of aquatic macrophytes and algal blooms through resuspension of the bottom sediments. One treatment being considered to control the 165

186 phosphorus is bauxite, which it is hoped might also limit plant growth by binding phosphorus (bauxite has a high aluminum content). Experimentation with alum, sand and clay treatments Figure 8-6. Distribution of variable water milfoil in Lake Massabesic. 166

187 elsewhere have shown some physical interference with rooted plant growth, but no limitation on nutrient uptake by rooted plants. Currently, MWW personnel are conducting experiments to determine the impact of bauxite on sediment chemistry. Overall Result of Management The control program for Lake Massabesic appears to have limited the spread of VWM to some degree, but has not prevented some spread and has not eliminated VWM at any site. The MWW staff commits a fairly large amount of time to the control of VWM, but began a full two years after VWM was documented in the lake and three years after having gone through a response planning exercise. The perception of the threat of VWM was apparently not strong enough to override the tendency to study the problem for some time or to circumvent the budgeting process. Inability to eradicate VWM at any of the four major sites in the lake, coupled with the absence of a focused monitoring program to detect VWM at its earliest stage of infestation, creates a perception that the current management program is inadequate. The MWW is therefore looking for alternatives. Phosphorus inactivation and physical interference with growths through application of bauxite is being researched, but it appears that additional, more severe control measures will be necessary to get VWM under control anytime soon. Lessons Learned and Recommendations Threat Analysis A threat analysis matrix for Lake Massabesic is supplied in Table 8-4. The overall threat of VWM infestation is high because of the large amount of suitable habitat for VWM in this lake. The nature of the impacts is high; all human uses could be impacted, most notably water supply, and sensitive species that are present in Lake Massabesic are likely to be at risk as well. Lake Massabesic already experiences taste and odor problems, and an increase in the VWM infestation could increase taste and odor issues. Dominance by VWM could certainly impact later water quality, including greater fluctuations in oxygen and ph and higher organic content. The potential for intake clogging also exists in this case. Public boating and fishing serve as a major transport vector for VWM, resulting in a high potential for the milfoil to spread. Due to the current level of infestation, continued spread of VWM, and the low frequency of whole-lake plant assessments, the probability of rapid response success is low. However, it should be noted that the spread of VWM is slower in this lake than many others where measures have not been taken; the MWW may very well have slowed the spread by its actions. Additional Data Needs An annually updated plant map is needed for Lake Massabesic. The NH DES was able to provide one, but the MWW was apparently not aware of its existence and should be cooperating in the effort to track VWM in this lake. In addition, a more detailed map of substrate type overlaid on a bathymetry map would be very helpful in assessing area-specific risk from VWM. This map would allow MWW personnel to identify areas likely to be colonized by VWM based on ideal substrate and depth, streamlining monitoring needs. 167

188 Table 8-4. Threat analysis matrix for VWM in Lake Massabesic. FACTOR YES NO THREAT EVALUATION HIGH MEDIUM LOW A large area could Extent and speed of possible X be affected infestation Plant density could be high X X Spread could be rapid X Water supply may be impacted X Swimming may be impacted Boating may be impacted Fishing may be impacted Aesthetics may be impacted Sensitive species may be impacted Protected species may be impacted Spread by water flow likely Spread by birds likely Spread by boating likely Spread by other human activities likely Eradication is possible Confinement is possible Annual macrophyte assessments X X X?? X X X X X X? X Nature of possible impacts Threat to water supply via water quality and possible clogging Boating and fishing could be substantially impacted Sensitive species may be impacted Ability to spread Several major transport vectors exist Recreational activities with high potential for spread Potential success of rapid response No longer an early infestation, but some limited potential for confining growths may exist X X X Rapid Response Assessment It appears that a Rapid Response Plan (RRP) was generated in some form by the end of 1993, but failure to implement any controls gave VWM time to become well established at several sites in the lake. An actively implemented RRP would have fostered response upon initial discovery of any invasive plant, including actual management of identified infestation 168

189 areas and an education program aimed at minimizing the spread while eradication was being achieved. As it was, no direct remedial action was taken until two years after discovery, and the education component was not implemented until four years after VWM was found. Due to the extent of the current infestation, a true RRP for VWM cannot be implemented for Lake Massabesic. However, a RRP could be developed for the un-infested areas of greatest concern to MWW personnel. The immediate area of the intake is not currently inhabited by VWM, but the plants are creeping in that direction. Macrophyte surveys should be conducted on at least an annual basis to ensure immediate recognition of VWM plants in the intake area, and for as large an associated area as possible. Plant density and the size of the infested area may trigger different approaches to control. Chapter 6 discusses how to create a RRP and includes a flow diagram for control techniques based on plant density and the size of the infested area. Proximity to the intake will limit some options, but careful planning is still appropriate in this case. Longer Term Response Assessment An important part of the longer term response assessment is macrophyte surveys. MWW personnel or someone selected by MWW should be conducting annual macrophyte surveys. If the plant communities are not being monitored throughout Lake Massabesic, newly colonized areas may go unnoticed until the infestation is too large or dense to control by typical RRP protocols. New colonies, identified early, allow for more choices for control. There are very few non-herbicide options for control when the infested area is large or the plants are very dense. Given what appears to be limited coverage in Lake Massabesic, some consideration should be given to hydraulic dredging in selected areas. While expensive and logistically complicated, dredging could remove the plants and sediment in VWM-infested areas with a high degree of efficiency. A lot of additional benefits can accrue with dredging as well. Expense will undoubtedly be an issue, but if the true cost of ongoing management is considered, and the areas to be dredged are not more extensive than appears to be the case (Figure 8-6), selective areal dredging warrants consideration. There are two preventive techniques that could be implemented at Lake Massabesic. Boat inspectors could be stationed at some or all of the boat launches. The job of the inspector is to check incoming and outgoing boats and trailers for invasive plants. This helps prevent the spread of VWM to other water bodies and also helps prevent the introduction of new and perhaps more invasive aquatic macrohpytes. The second preventive measure is a washdown station. Stations could be maintained at the boat ramps for boaters to wash down boats and trailers before entering or leaving Lake Massabesic to limit the spread of VWM and other invasive plants. As VWM is already in Lake Massabesic, these methods are not as applicable as in-lake actions at this time, but protect the investment made in other controls. LAKE AUBURN Lake Auburn is a natural lake located in Auburn, Maine, and is used year round for water supply. Lake Auburn is an unfiltered source of supply. The breakdown of raw water sources for Lake Auburn includes 42% flow from tributaries, 36% direct precipitation, and 22% direct groundwater inflow. Of the 42% flow from tributaries, 27% comes from the Basin located north of Lake Auburn, which is dammed and connected to Lake Auburn via a stream. Lake Auburn consists of 2290 surface acres with maximum and mean depths of 120 and 35 feet respectively. 169

190 The maximum reservoir volume is 26 billion gallons, and the average daily withdrawal is 8 million gallons per day, well below the safe yield limit of 17 million gallons per day. The watershed of Lake Auburn is 15.2 square miles, with only 15 residences along the shoreline. Land use has not been quantified but cultivated land, livestock use, forest, wetlands and lakes are present in the watershed. The majority of the watershed is located north of Lake Auburn. Currently, public access to the Lake is restricted to one boat launch. Lake Auburn is susceptible to colonization by milfoil species. Twenty-nine percent of the surface area of Lake Auburn is located within the 0-20 foot contour. Bottom type within this contour consists of 90% sandy sediment and 10% silt/muck. Based on depth and substrate features, 66 surface acres in Lake Auburn are optimal for variable water milfoil (VWM) colonization, and 598 surface acres are ideal for Eurasian water milfoil (EWM) colonization (Figure 8-7). Additional coverage by each species is possible, with some overlap in possible habitats, and the susceptibility assessment does not include connected waterbodies. Currently, within the 0-20 foot contour, 1-25% of the bottom substrate is covered by aquatic macrophytes, while <5% of the water column is occupied. The drinking water intake is located in >30 feet of water and not within the likely depth range for milfoil colonization. Additional water uses and potential transport vectors within Lake Auburn are limited to boating, fishing, and passive recreation. Approximately one-third of the Lake is closed off to all activity. There are no horsepower or size restrictions on boat motors, but personal watercraft are not permitted. No direct contact (i.e., swimming, water skiing) is permitted anywhere in Lake Auburn. Shoreline access is limited, with no direct contact permitted. The Lake Auburn Watershed Protection Commission (LAWPC) owns approximately 80% of the lake shoreline. In addition to human transport vectors, water flow and animals such as birds and beavers have potential to spread milfoil fragments. The Basin, located north of Lake Auburn, is a dammed shallow water basin that flows into Lake Auburn. The dam is strictly an overflow weir, although water flows from the Basin to Lake Auburn most of the year. The Basin is 106 surface acres, has a mean depth of 5 feet, and the bottom substrate consists entirely of organic silts. These features make the entire Basin susceptible to VWM colonization. VWM is the only milfoil species found in Lake Auburn today. VWM was first discovered in the 1980s in the stream between the Basin to the north and Lake Auburn. It is possible that VWM was also present in the Basin and the northern sections of Lake Auburn at that time as well, but no detailed survey was conducted. It is believed that VWM may have been present in these areas since the 1970s. Although steps were taken prior to 2001 to assess on the aquatic macrophyte community present in Lake Auburn, the decision was made in 2001 that measures needed to be implemented to control the VWM. Milfoil Assessment In 2000, the LAWPC worked with the Maine Department of Conservation to conduct a macrophyte survey of Lake Auburn using proper techniques, to identify and catalogue all aquatic macrophytes present in the lake and its major tributaries. During this survey, a dense stand of VWM approximately 15 acres in area was mapped in the northern end of Lake Auburn. VWM was also present in the Basin to the north and the stream connecting the two waterbodies. To assess the extent of possible spread of milfoil, LAWPC staff examined substrate and depth to identify areas in the lake conducive to VWM and EWM growth. As parrotfeather has not been 170

191 found in Maine or in adjacent New Hampshire, this milfoil species has not been considered a threat. In 2001, a whole-lake macrophtye survey was conducted, and included Lake Auburn and the Basin. The conclusion of the survey was that the VWM had not spread outside of North Auburn Cove, the initial area of infestation. In 2001, LAWPC closed one private boat launch area in North Auburn Cove (later purchased) due to the VWM infestation and used buoys to 171

192 Figure 8-7. Milfoil susceptibility in Lake Auburn and the Basin, with known VWM distribution as of late

193 restrict access by motor boats to this area. The southern portion of the Basin was also restricted using buoys. In 2001, management planning occurred to select additional control techniques to be employed in the lake. Three management techniques were considered to control the VWM in Lake Auburn. Benthic barriers, hand pulling and prevention in the form of boat inspections and signs were all selected for implementation. During the planning phase, LAWPC staff contacted the Maine Department of Conservation, Maine Department of Environmental Protection and the Volunteer Lake Monitoring Program for input. The plan was to implement the preventive techniques during the first year, then follow up with more active control methods. Management Efforts 2002 VWM Management In 2002 another whole-lake aquatic macrophyte survey was conducted to determine the distribution of all plants, including the number and size of VWM patches. The survey included the Basin north of Lake Auburn. VWM was documented in the north end of Lake Auburn, as well as in the Basin and the connector stream. Approximately 140 patches of VWM were located and mapped within the Basin in 2002 using GPS and GIS technology. The southernmost seven acres were considered completely infested and were not surveyed in further detail. Due to the mean depth (5 feet) and the substrate types, 100% of the Basin is considered prime habitat for VWM. The extent of the VWM coverage in Lake Auburn in 2002 was similar to The prevention and education portion of the management plan was first implemented in LAWPC staff posted signs around the lake and at the public and private boat launch, warning the public about invasive plants in general and VWM specifically. They also conducted inspections of boats and trailers at the public boat launch area to identify and remove fragments of VWM and any other invasive plants. Again in 2002, buoys were deployed to identify areas of restricted boat traffic VWM Management In 2003, a whole-lake aquatic macrophyte survey was again conducted to determine the distribution of all plants, including the number and size of VWM patches. The survey included the Basin north of Lake Auburn. VWM was documented in the north end of Lake Auburn, as well as the Basin and the connector stream. The extent of the VWM coverage in Lake Auburn in 2003 was similar to 2001 and In addition to boat launch inspections and buoy deployment, the LAWPC staff implemented a VWM removal plan for the Basin in Divers were used to install marker buoys and hand pull VWM plants in the Basin. Divers were trained by the Maine Department of Environmental Protection on the proper techniques to hand pull the milfoil to ensure entire root crown removal. The total cost of the two separate hand pulling efforts was $700. Benthic barriers were not installed in

194 2004 VWM Management In 2004, a whole-lake aquatic macrophyte survey was again conducted to determine the distribution of all plants, including the number and size of VWM patches. Yearly assessments of the plant community using GPS allow the LAWPC staff to determine if and to where the VWM is spreading. The survey included the Basin and Lake Auburn. The extent of VWM coverage in the Basin was greater in 2004, compared to previous years. VWM levels in Lake Auburn also increased in North Auburn Cove. In 2004, hand pulling, benthic barriers, and buoys were employed in the Basin, while boat launch inspections and buoys were used in Lake Auburn. Six 10 ft x 10 ft benthic barriers constructed of Geotech Cell were deployed by divers in the Basin. Hand pulling resulted in the removal of enough VWM to completely fill two 16 ft boats. Staff from LAWPC conducted 377 boat inspections in 2004, but did not report finding any invasive plant material. The total cost of all management techniques to control milfoil in 2004 was $ VWM Management Another whole-lake macrophyte survey was conducted in 2005 and included the Basin and Lake Auburn and its associated wetlands. For the first time, the LAWPC staff located patches of VWM in two wetland areas in Lake Auburn. These areas were recorded using a GPS for inclusion in the mapping efforts and for later location in removal efforts. VWM levels in North Auburn Cove remained similar to 2004 levels, but the Basin once again contained greater amounts of milfoil. In 2005, the decision was made that hand pulling in the Basin would no longer occur, but buoys were deployed to mark restricted areas. The decision was made based on the physical demand of the removal, the lack of funding and staff, and ultimately the overall ineffectiveness of the removal based on the level of infestation. Eleven 10 ft x 10 ft benthic barriers were installed in the Basin to control the VWM in specific locations. Also, the only boat access was gated to control motorized boats from entering the Basin. Hand pulling was used to control the VWM colonies identified in two of four major wetland areas surrounding the lake. Hand pulling was employed in both wetland areas to remove the VWM. In the southwest wetland, furthest from the intakes, the need for additional measures was recognized. Monitoring later in the season revealed that some plants were re-growing in the pulled area and a larger patch of VWM plants was observed near the pulled area. This area was exposed due to lower lake level from evaporative loss. The second wetland, closest to the intake area, had a very small patch of VWM and was hand pulled on two separate dates. Once again, LAWPC staff members were stationed at the boat launch area to conduct inspections of boats and trailers. A total of 499 boat inspections were conducted in 2005, with one of those inspections resulting in VWM plant material being detected. The total cost for milfoil control efforts in 2005 was $ VWM Management A whole-lake macrophyte survey was conducted again in 2006 and included the Basin and Lake Auburn and its associated wetlands. LAWPC staff observed VWM in the southwest wetland area. This wetland area containing VWM in 2006 was the wetland furthest from the intake that had milfoil in Hired divers were able to thoroughly survey the 174

195 wetland using snorkel gear. VWM levels in the North Auburn Cove of Lake Auburn were similar to those observed in In 2006, physical milfoil control efforts were restricted to the southwest wetland area in Lake Auburn; no work was completed in the Basin other than buoy deployment. Buoys were again deployed in North Auburn Cove to identify restricted areas, and boat inspections were also completed at the boat launch area. Ten 10 ft x 10 ft benthic barriers were positioned in the wetland area to control the VWM growth and were left in place over the winter. The total cost for milfoil control efforts in 2006 was $3035. Overall Result of Management The success of the current control program at Lake Auburn varies between locations. The Basin is already infested, so while benthic barriers are effective in small areas where they are deployed, the VWM is not controlled and there is no major impact on the infestation. Hand pulling is no longer used in the Basin because of its ineffectiveness, given the extent of infestation, but will be considered in the future for very small patches. No physical removal techniques have been employed in Lake Auburn proper or North Auburn Cove, although effort has been made to control VWM in the adjacent wetlands. Currently, a PhD student with the support of the LAWPC is conducting research on the VWM in North Auburn Cove to determine the best control technique for the lake. Benthic barriers constructed of various materials are being examined, as well as other physical removal techniques. The control of VWM in the wetland areas of Lake Auburn has been effective, because annual surveys have resulted in rapid identification of new VWM growths and control measures have been rapidly implemented. Lessons Learned and Recommendations Threat Analysis A threat analysis matrix for Lake Auburn is supplied in Table 8-5. The extent of infestation by VWM is low due to the limited area in Lake Auburn that possesses suitable substrate for VWM growth. Some growth beyond that small optimal growth area within the lake is possible, but apparently not enough to have a major impact; this is an assumption from the current data that is worth verifying. The nature of the possible impacts is also rated low because the area suitable for VWM is small compared to the size of the lake. However, if all of the Basin and the wetlands associated with Lake Auburn are counted in the total potential acreage, the impact to water supply and other uses may not be minimal. Boating and fishing are already impacted in certain areas, and habitat impacts in the adjacent wetlands may be significant. The potential for VWM to spread to all hospitable areas is high because there are active transport vectors and an established population just upstream of the lake and in the northern cove of the lake. Finally, the potential for rapid response success is medium in the lake. North Auburn Cove contains a large area of VWM, so rapid response is not really feasible in this area. However, annual surveys and rapid implementation of control measures in other parts of Lake Auburn and associated wetlands have minimized the spread of VWM outside of established areas. The use of annual surveys is especially commendable, as this facilitates informed management response. 175

196 Additional Data Needs The LAWPC conducts annual macrophyte surveys, and is continually monitoring when personnel are on the lake for other activities (i.e., bird harassment to minimize bacterial inputs). This level of vigilance is needed, but the current approach is sufficient. One interesting observation of LAWPC staff and others around the lake is the apparent association of VWM growths and beaver lodges. When VWM is found in the wetland areas of Lake Auburn and the Table 8-5. Threat analysis matrix for VWM in Lake Auburn. FACTOR YES NO THREAT EVALUATION HIGH MEDIUM LOW A large area could be affected X Extent and speed of possible infestation Plant density could X X be high Spread could be rapid X Water supply may Nature of possible impacts X be impacted Sensitive species may be impacted Swimming may be Minor impacts to other human X impacted activities Boating may be impacted X Fishing may be impacted X X Aesthetics may be impacted X Sensitive species may be impacted? Protected species may be impacted? Spread by water Ability to spread X flow likely No direct human contact Spread by birds Large numbers of seagulls X likely Spread by boating likely X X Spread by other human activities X likely Eradication is Potential success of rapid X possible response Confinement is possible X X Annual macrophyte assessments X 176

197 Basin, there is almost always a beaver lodge in the vicinity. Observations of the public living around and using the lake indicate that the beavers may be intentionally gathering the VWM for use in their lodges. Also, trees brought through infested areas catch VWM and transport it to new areas; beavers transport trees from as far away as one mile. Further investigation of this association may be warranted as an additional indicator of areas for probable milfoil introduction. It would also be appropriate to carefully document substrate composition at water depths <20 ft, to be certain that VWM cannot occupy a larger area than previously estimated. Rapid Response Assessment The importance of an active RRP is illustrated in Lake Auburn. Granted, at the time of the initial discovery in the 1980s, the knowledge and understanding of the importance of prevention and immediate control of invasive species was muted compared to today. An active RRP may have allowed LAWPC staff and other interested stakeholders to eradicate the VWM growths in their early stages when the number of effective control techniques is at its highest and the potential for success is greatest. Suitable control measures would be pre-selected based on the size, density, and location of the VWM growth. In the case of Lake Auburn, the initial growths of VWM may have been controlled by hand harvesting and benthic barriers. If initial eradication was successful in the Basin area, this area would have acted as an additional buffer zone for Lake Auburn. Instead, with the current conditions, the Basin acts as a constant source of VWM flowing directly into Lake Auburn. The current goal of the LAWPC is to control VWM spread in Lake Auburn; for at least new areas of Lake Auburn, the current practices constitute a RRP. When new areas of growth are identified, LAWPC staff act quickly to hand pull the plants or install benthic barriers. The annual macrophyte surveys in conjunction with quick response after identification have prevented VWM from becoming established elsewhere in Lake Auburn and its associated wetlands. Lake Auburn Rapid Response Plan for EWM Threat Analysis A threat analysis for EWM in Lake Auburn is supplied in Table 8-6. The large area of suitable habitat for EWM compared to the overall size of Lake Auburn (Figure 8-7) is of interest to the LAWPC, as EWM has been found in Maine. At least 26% of the lake area, or 598 surface acres, are highly suitable for colonization by EWM. The calculation of suitable habitat includes depths from 0-20 feet, although EWM has been found growing at depths of 30 feet in clear waterbodies. In terms of suitable habitat, this is nearly 10 times the amount of habitat in the lake that is highly suitable for VWM. Therefore, the impacts of EWM infestation are likely to be much greater than those of VWM in this reservoir and could impact water quality as relates to water supply. Boating and fishing could be impacted as well, and the threat to sensitive species and lake ecology overall would be substantial, although no information on sensitive or protected species was available. The spread of EWM could be rapid, and there are at least three common vectors in Lake Auburn: water flow, birds and boats. Yet as EWM has not yet appeared, the potential for successful rapid response is high. The current practice of annual macrophyte assessment is especially encouraging, as the probability of early detection is high. 177

198 Monitoring to Detect EWM Invasion During plant surveys, it is important to ensure that areas of highest probability for initial introduction are assessed. These include mouths of tributaries, boat ramps and areas of high bird concentration. The initial introductions of EWM are typically in shallow water, and may not be mature enough to have reached the surface of the water. Therefore, viewing tubes, underwater cameras and snorkeling are useful tools to help identify early EWM infestations and should be included in the annual surveys. Table 8-6. Threat analysis matrix for EWM in Lake Auburn. FACTOR YES NO THREAT EVALUATION HIGH MEDIUM LOW A large area could Extent and speed of possible X be affected infestation Plant density could be high X X Spread could be rapid X Water supply may Nature of possible impacts X be impacted Sensitive species may be impacted Swimming may be Major impacts to other human X impacted activities Boating may be Large area optimal for infestation X impacted along entire shoreline Fishing may be impacted X X Aesthetics may be impacted X Sensitive species may be impacted? Protected species may be impacted? Spread by water Ability to spread X flow likely No direct human contact Spread by birds likely X Spread by boating X X likely Spread by other human activities X likely Eradication is Potential success of rapid X possible response Confinement is X possible X Annual macrophyte assessments X 178

199 If EWM is identified in Lake Auburn during a macrophyte survey it is important to examine the pattern of occurrence, as it may provide useful clues as to the source of the plants. In the case of new growths in the vicinity of boat launch areas, boats are the likely vector for introduction. If the EWM is first identified near an inflow area, a survey of the upstream waterbodies should be performed to identify additional source growths of EWM. Large numbers of gulls use Lake Auburn, so EWM growths found in remote areas suggest the birds could be acting as the transport vector for the plant material. After identification of a new growth of EWM, a thorough inspection of the surrounding area needs to be conducted to ensure any associated plants are located. Tightly spaced transects that move outward from the new growth or in the direction of likely current or wind transport should be conducted. This more concentrated survey will help ensure the entire population is identified for complete removal. The survey of the new growth should include some estimate of cover and biomass. Stem counts are useful, but are very time consuming, but some estimate of density should be made to support appropriate control method selection. Also, surveyors should record the other plants in the area because the presence of protected or rare species may affect selection of appropriate control measures. Communication and Quarantine If EWM is identified in Lake Auburn or upstream in the Basin, the LAWPC should notify all stakeholders of the introduction of EWM. These stakeholders include but are not limited to the 15 home owners around the lake, the Town of Auburn where the lake is located, and boaters and anglers who use the lake. A posting of the discovery in a local newspaper may facilitate widespread notification of the problem. Additional signs should be posted at the boat launch area that indicate the exact location(s) of the growth and encourages boaters to stay out of infested areas. The signs should include detailed descriptions and pictures of EWM and provide a contact number in case of additional discovery of plants by lake users. The signs should tell boaters to mark the area where they identified the EWM but should not encourage them to remove the plant, as they are unlikely to get the entire root crown. To minimize the spread of EWM in Lake Auburn, quarantine of the new growth areas is recommended. The options for quarantine of the new growth vary with the size of the affected area and available funding. The area could be marked as restricted using buoys; this technique is currently employed in the Basin and North Auburn Cove to keep boaters away from VWM growths. Sequestration curtains or screens could be deployed, and although this technique has been very effective for small infestations, it is also very expensive. Additional options to limit the spread of the EWM are available, depending on the location of the infestation. If the new growth is located in the boat launch area, the boat launch could be closed. The boat launch property at Lake Auburn is owned and maintained by the LAWPC. If deemed necessary due to the size and location of the infestation, the LAWPC could close the boat launch area without approval of state or local authorities. If the boat launch is kept open to the public, increased staffing may be necessary to ensure all boats are inspected for plant material. Due to the high numbers of birds present on Lake Auburn, bird controls should be implemented to limit bird contact with infested areas; current bird control is limited to gulls only, but efforts may have to be increased to minimize the threat of EWM spread. While eradicating any EWM that appears in the Lake Auburn system will undoubtedly have a high priority with the LAWPC, it is also important to prevent the export of EWM out of Lake Auburn and into other waterbodies. Possible expansion routes should addressed 179

200 immediately. Aside from the boat ramp and bird migration, the outflow area of Lake Auburn could be screened to prevent EWM fragments from moving downstream. Frequent cleaning of the screens may be necessary to prevent blockage. Early Eradication The potential to eradicate EWM quickly relies heavily on the ability to control the spread of the plant. During the growing season plant expansion is through the root crowns, while during the late season plant fragmentation occurs. Removal of the EWM prior to fragmentation during the late season is essential to control the dispersal of propagules. To eradicate an early infestation of EWM, the LAWPC staff will have five primary management options to consider; herbicides, water level drawdown, hand pulling, benthic barriers and suction harvesting. The LAWPC is not expected to support the use of herbicides in Lake Auburn, but it should be noted that two potentially effective herbicides are approved by the USEPA for use in drinking water supplies at levels known to be effective against EWM. While such use would represent a departure from the norm for Maine, and indeed most potable water providers, the potential for these herbicides to provide control should not be ignored. Approaches are available that would minimize any risk of any of the herbicide reaching the intakes, but some risk will always exist. Water level control through drawdown may not be technically possible, and keeping the reservoir in a lowered conditions for much of the winter, the normal refill period, would not be consistent with normal operating procedures and could endanger adequate supply in a dry year. Therefore, three primary options would be considered first to control any arriving EWM in Lake Auburn. If the density of EWM plants is low (typically <500 plants per acre) and the size of the area infested is small (<10 acres, preferably <2 acres), hand pulling can be a very effective technique. At higher densities or area, the risk of spread by fragmentation due to hand pulling increases, and the probability of thorough removal declines drastically. This is why hand pulling of VWM was not successful previously, but does not negate the value of hand harvesting under the right conditions. To reduce the chance of spread through fragmentation, even at low densities appropriate for hand pulling, a fragment barrier should be installed around the infested area. One major advantage of hand pulling is the limited impact to non-target species. For Lake Auburn, divers are already trained on how to hand pull milfoil species, so this technique should be a preferred technique. Benthic barriers are effective at controlling EWM infestations, and are especially appropriate for small (<1 acre) patches that are too dense to hand pull. However, benthic barriers are not species specific and will also eliminate non-target species in Lake Auburn. Benthic barriers are also quite expensive on an areal basis ($30,000 to $50,000 per acre), although the material is useful for many years, justifying the capital cost on at least a small scale installation. The use of benthic barriers is labor intensive, as crews at Lake Auburn already know, and maintenance (at least annual cleaning) is recommended to ensure that rooted plants do not become established on the top of the barrier. Consequently, widespread use is not preferred. Benthic barrier application should be considered at Lake Auburn to control areas of growth that are too dense to control with hand pulling. Another control technique for consideration in Lake Auburn to control a new growth of EWM is suction harvesting. Suction harvesting is best performed in conjunction with hand pulling where divers can remove milfoil plants and feed them directly into the suction harvester tube. The divers at Lake Auburn have been trained in proper identification and removal of 180

201 milfoil species, so the removal of non-target species or incomplete removal of the milfoil root crown should not be serious problems. Suction harvesting is most effective at moderate densities; it can be applied at high densities, but the likelihood of eradication declines and repeated efforts are necessary. Typically, the cost of suction harvesting ranges between $5,000 and $15,000 per acre depending on the specific equipment, plant density and total area to be harvested. Turbidity is the major concern with suction harvesting, and may limit its applicability in Lake Auburn because of the impact to the drinking water, but this is a viable option at least at a substantial distance from the intake. The economic, regulatory and social limits on larger scale control techniques (e.g., herbicides or dredging) at Lake Auburn reinforces the importance of addressing new growths of EWM early in the infestation process. In just a few growing seasons, an infestation of EWM could be too dense and too large to control with any of the techniques described above. In general terms, once a growth of EWM is greater than 3-5 acres in size and denser than stems per square foot, herbicides are the most practical means to control the infestation. Small scale hydraulic dredging might be considered, but the cost is very high. Any EWM infestation in Lake Auburn that is not identified quickly enough to employ one of the above described techniques could quickly spread throughout the entire hospitable area within the lake, which is on the order of 600 acres or 26% of the lake area, enough to potentially impair both water supply and recreational uses. GENERAL OBSERVATIONS FROM CASE STUDIES While each case is to some extent unique, there are a few features that many case studies share. These include: Lack of a complete Rapid Response Plan. A tendency to wait at least a year and often 2-3 years between the discovery of an invasive milfoil and the first control action. An avoidance of more aggressive controls in drinking water supplies, in comparison to recreational lakes. Rapid response planning is not a new thing, but it is not a widely publicized process. Some utilities have made an effort to construct a RRP, but none of the ones evaluated in this project were complete. Early detection methods are most often lacking; infestations are found by chance or as part of infrequent monitoring; not enough value is place on early warning systems with regard to invasive species. Quarantine options seem to be well understood, although not always implemented as quickly or completely as warranted. Actual control techniques are often not evaluated prior to the discovery of infestation, and where they have been, implementation is still routinely slow. Delays in implementation of controls are sometimes a result of a wait and see attitude, but are also often a function of approval procedures for implementation that take too much time. Budgeting for labor and materials is a particular problem that often results in at least a one-year delay in control actions. There is a surprising lack of economic analysis with regard to the cost of immediate action versus later costs to address a more widespread infestation. Many states are willing to address new infestations in recreational lakes, having developed statewide Rapid Response Plans, restricting human access that might lead to spread of the invasive species and applying the full range of techniques where deemed appropriate. The addition of a drinking water supply to the mix of uses, where the invasion does not represent an 181

202 immediate threat, appears to shift the balance away from aggressive effort to gain control over the invasive plant. Similar shifts have been observed when a protected species is present and believed to be sensitive to a possible control method for an invasive species. This hesitancy to act swiftly and decisively can allow the infestation to progress beyond the means of most rapid response methods to control. The use of herbicides is particularly controversial with regard to drinking water supplies, but more on a social acceptance basis than scientific grounds. At least two herbicides (Fluridone and Triclopyr) are approved under their federal labels for use in drinking water supplies and can be used in approaches that minimize the potential for the chemical to enter any supply intake. There will always be some risk, but each can be effective on at least one species of nuisance milfoil at levels allowable in drinking water. Even other herbicides that cannot be allowed to enter drinking water at concentrations necessary to effectively kill milfoil may have a place in remote treatments, as dilution can be sufficient to minimize risk. Use of herbicides on a regular basis in drinking water supplies is not a valid approach, but use as part of a RRP aimed at preventing a widespread infestation in a reservoir that could be impaired by such growths deserves consideration. The blanket policy of many utilities and state agencies against herbicide applications is not completely justified. On the other hand, an appropriate monitoring program should allow detection at a level that is amenable to eradication without the use of chemicals, so herbicides are not a necessary component of the RRP if the early warning system is adequate. It does appear that drinking water supplies may be less threatened overall than recreational lakes, for reasons that include greater average depth (to support adequate supply) and fluctuating water levels (from withdrawals) that generate drawdowns that can kill milfoil and also tend to maintain less hospitable substrates for milfoil growth. It is also true that recreational uses are more restricted in many drinking water reservoirs, limiting milfoil vectors. However, these factors will not provide adequate protection in all cases, and the use of the threat analysis described in this report is strongly encouraged to determine the potential for supply impairment. Development of a RRP is encouraged even if the threat is minor, but the elements and implementation schedule will vary with the severity of the threat. 182

203 CHAPTER 9 CONCLUSIONS AND RECOMMENDATIONS This milfoil scoping study was commissioned to determine the extent of perceived problems with species of milfoil in drinking water supplies and to evaluate methods of preventing or mitigating such problems. This document is intended to raise awareness by drinking water supply managers of issues relating to invasive milfoil species and actions they might take to assess the susceptibility of their reservoirs to invasion and impacts, prevent such invasions, or mitigate an infestation when it occurs. This is not meant to be an alarmist document. The information collected on current milfoil problems does not indicate that invasive milfoils are a widespread, major threat to drinking water supplies, but it does indicate that there is distinct potential for invasion in susceptible systems and that the impacts of milfoil have bearing on water supply quality and cost of treatment. The major conclusions of this investigation are summarized as follows: 1. There are three species of milfoil that are invasive and represent the vast majority of problem growths within this species: Myriophyllum spicatum (Eurasian water milfoil), M. heterophyllum (variable water milfoil), and M. aquaticum (parrotfeather). 2. Problems include elevated disinfection byproduct precursors (mainly organic carbon compounds), taste and odor, altered and/or fluctuating oxygen and ph, and clogging of intake screens. 3. From the available data, about 10 to 15% of water supplies, representing <10% of the corresponding water volume, may be subject to milfoil-related problems. While not a large portion of water supplies, problems may result in significant costs. 4. Features of reservoirs that tend to make them susceptible to milfoil invasions include sediments of intermediate organic content, low slope, water depth <15 ft, unrestricted boat access, large migratory bird populations, a sparse native plant assemblage, and upstream waterbodies with milfoil present. 5. Fluctuating water levels associated with most water supply reservoirs tend to limit milfoil colonization and foster less hospitable sediments. 6. Measures to prevent milfoil infestation focus on preventing or inspecting incoming boats, bird controls and management of upstream milfoil populations. 7. Rapid response plans work best in the earliest stages of an invasion; monitoring for detection and avoidance of implementation delays are essential to successful rapid response. 8. Physical and chemical techniques have both proven effective against milfoil, but most utilities do not apply herbicides to drinking water, even when allowable by the label and applicable laws. 9. Once an invasive species of milfoil becomes established, it is extremely difficult to eradicate it. 10. The level of awareness of milfoil issues and control options among utilities is low, and effective rapid responses are rare. 183

204 Recommendations for drinking water suppliers can be summarized as follows: 1. Conduct a threat analysis for the reservoir to determine the likelihood of milfoil invasion and the potential for resultant water supply problems. 2. Map the aquatic plant assemblage in the reservoir; this knowledge is not expensive to acquire and can be very helpful in management planning. 3. If invasive milfoil is not present, take steps to prevent its introduction and establishment. Inspections at any boat launch and development of a practical rapid response plan are two important actions in this regard. 4. Monitor logical locations of invasion (inlets, boat launches, areas of bird congregation) and be prepared to implement the rapid response plan without delay. 5. If invasive milfoil is established, evaluate the potential for water quality impacts and effective control. Implement a control plan if warranted. The value in milfoil awareness is largely related to the opportunity to prevent possible problems for relatively little cost or effort. This awareness appears to be low among utilities, however. While the potential for widespread infestations and major costs across the nation may not be high, problems in susceptible reservoirs could carry significant costs. The situation could be characterized as analogous to the zebra and quagga mussel problem. As source water management gains higher priority in utility operations, the need to evaluate system biology and manage to avoid problem species such as milfoil will become apparent. Anything that can be done now to raise awareness and prompt preventive action at limited cost is worthwhile. 184

205 APPENDIX A: SPECIES KEY AND PICTURES Identification Key for Myriophyllum This identification key for the North American species of Myriophyllum is adapted from Aiken Myriophyllum ussuriense is not included in this key. 1 Leaves scale-like or absent; terminal flowers alternate. 1. M. tenellum 1 Leaves pinnately divided, the segments filiform; terminal flowers alternate or whorled. 2 Flowers and fruits in axils of limp submersed leaves. 3 Fruits mm long with two prominent tuberculately roughened dorsal ridges; turions (specialized winter buds) present September to April. 2. M. farwellii 3 Fruits mm long, rounded on back, never dorsally ridged, smooth or only minutely roughened; turions never present.3. M. humile 2 Flowers in erect spikes above water, the subtending bracts firm. 4 Emergent leaves usually cm long, with uniform linear divisions, each 4-8 mm long; petiole 5-7 mm; plants dioecious, only the pistillate, white flowers known in North America.. 4. M. aquaticum 4 Emergent leaves always less than 2.0 cm long, entire, serrate or pectinate, the teeth or divisions less than 4 mm long; petiole absent, or less thatn 2 mm; plants monoecious, the lower flowers pistillate, usually pink, the upper staminate. 5 Uppermost flowers and all submersed leaves whorled. 6 Upper bracts entire; lower bracts entire, pectinate, or serrate, not more than twice the length of the adjacent pistillate flowers. 7 Inflorescence commonly forking; bracteoles 1-2 mm long, lanceolate 5. M. quitense 7 Inflorescence simple; bracteoles less than 1 mm long, ovate. 8 Stem thickened below the inflorescence to almost double the width of the lower stem, characteristically curved to lie parallel with the water surface; scales at inflorescence nodes 2-3, black, distinct at least in fresh material; plants never forming turions, dying back to root crowns over winter 6. M. spicatum 8 Stem not thickened below the inflorescence, straight; scales at inflorescence nodes 0-2, black or brown, indistinct even in fresh material; plants forming oblanceolate turions from October to June, the black green turion leaves thicker than summer leaves. 7. M. sibiricum 6 Upper bracts pectinate or dentate; lower bracts pectinate, usually more than twice as long as the adjacent pistillate flowers. 9 Bracts without a distinct lamina, the lower pinnatifid, the upper pectinate; bracteoles less than 0.1 mm long or absent; turions clavate, dark yellow-green, present on vertical stems October to April 8. M. verticillatum 9 Bracts laminate, the lower more deeply toothed than the upper; bracteoles mm long; overwintering buds if present not clavate, green, arising from the rhizomes, or at the base of the stem. 10 Bracteoles mm long; bracts mm wide, lanceolate to elliptic; stigma recurved; mature mericarps terete, rounded on the sides, prominently breaked by the persistent recurved stigma.9. M. heterophyllum 10 Bracteoles mm long; bracts 1-2 mm wide, linear to lanceolate; stigma erect; mature mericarps deeply sulcate, flattened on the sides, lacking a beak, the stigma withering as the fruit matures..10. M. hippuroides 5 Uppermost flowers alternate; submersed leaves whorled or alternate. 11 Submersed leaves whorled; terminal flowers alternate; bracts entire or serrate, all about as long as the flowers..11. M. alterniflorum 11 Submersed leaves alternate or pseudowhorled (a series of short internodes alternating with a long one); bracts with filiform divisions, the lower longer than the adjacent flowers. 12 Fruit smooth or only minutely roughened (Canada and northern United States) M. humile 13 Upper bracts shorter than the adjacent flowers, subentire, spathulate.12. M. laxum 13 Upper bracts much longer than the adjacent flowers, pinnately divided, filiform M. pinnatum 185

206 Representative Photographs and Diagrams of the Genus Myriophyllum Myriophyllum spicatum From the ENSR photo library Myriophyllum heterophyllum Photographer: Donald Cameron 186

207 Myriophyllum aquaticum Photographer: Donald Cameron Myriophyllum humile Photographer: Donald Cameron 187

208 Myriophyllum sibiricum Photographer: Donald Cameron Myriophyllum hippuroides Photographer: Dennis Woodland and located on the Robert W. Freckmann Herbarium website. 188

209 Myriophyllum verticillatum Photographer: Donald Cameron Photographer: Donald Cameron Myriophyllum laxum Photographer: John Nelson, University of South Carolina Herbarium 189

210 Myriophyllum farwellii Photographer: Donald Cameron Myriophyllum quitense Photographer: Jenifer Parsons, Washington Department of Ecology 190

211 Myriophyllum ussuriense Photographer: John A. Christy, Oregon Natural Heritage Information Center, Oregon State University. Myriophyllum tenellum Photographs taken by Robert W. Freckmann and Robert H. Read respectively. The photographs are located on the Robert W. Freckmann Herbarium website. 191

212 Myriophyllum alterniflorum Photographer: Donald Cameron 192