Integrated We E D M a N a G E M E

Home , Bison Range, Bitterroot River, Clark Fork River, KTMF, Lake Missoula

United States Department of FINAL Environmental Impact Statement Agriulture Forest Service Integrated We e d Lolo National Forest M a n a g e m e n t

December, 2007 ON THE LOLO NATIONAL FOREST

Lolo National Forest

Final Environmental Impact Statement Integrated Weed Management

Missoula, Mineral, Sanders, Granite, Powell, Lewis and Clark, Flathead, Ravalli and Lake Counties, Montana

Responsible Agency: Responsible Official: For Further Information Contact: USDA Forest Service Deborah L. R. Austin Andy Kulla, Project Team Leader Forest Supervisor Lolo National Forest Lolo National Forest Building 24A, Fort Missoula Missoula, Montana 59804 (406) 329-3962 Abstract: This Environmental Impact Statement (EIS) documents the consideration of seven alternatives for improving weed management on the Lolo National Forest (Lolo NF). Two alternatives were given detailed study and five alternatives were considered but not given detailed study. The Forest is proposing to treat up to 15,000 acres of weeds per year using an integrated weed management approach as described in this EIS as Alternative 2. Alternative 2 would add sheep and goat grazing to the Lolo NF weed management “tool box”, allow weed managers to more promptly control new weed species and new infestations of weeds already found on the Forest, and would allow us to better prioritize our weed control efforts. The integrated weed management tools included in Alternative 2 would include education and awareness, prevention, ground and aerially applied herbicides, goat and sheep grazing, biological control agents, mowing, pulling, seeding and fertilizing. The purpose of this project is to establish beneficial vegetation and weed resistant plant communities by: • Preventing potential invaders from establishing • Eradicating small new infestations and reducing larger infestations of new invaders • Containing or reducing widespread weeds within infested areas • Allowing rapid, timely response to new, small or recently discovered infestations before they spread and become well established • Preventing or limiting the spread of established weeds to protect weed-free areas; and • Implementing the 2004 Forest Service National Strategy and Implementation Plan for Invasive Species Management Alternative 1 would take no action to change the current weed management on the Lolo NF. Only sites with existing weed management decisions could be treated and about 5,000 to 6,000 acres would be treated each year. Under Alternative two 78,443 weed infested acres were analyzed. Alternative 2 would allow treatment of a maximum of 15,000 of these 78,443 acres in any one year. This broad “menu” of infested sites would allow managers to better prioritize our weed control efforts and maximize our effectiveness. Alternative 2 would also allow treatment of more new and small infestations before they grew to unmanageable proportions. Five other alternatives; Prevention Only, No Herbicides, Low Herbicides, Concentrated Eradication Cells and No Aerial Herbicide Application, were considered but not given detailed study for the reasons described in Chapter 2. Alternative 2 is the Agency’s preferred alternative.

The U.S. Department of Agriculture (USDA) prohibits discrimination in all its programs and activities on the basis of race, color, national origin, age, disability, and where applicable, sex, marital status, familial status, parental status, religion, sexual orientation, genetic information, political beliefs, reprisal, or because all or part of an individual’s income is derived from any public assistance program. (Not all prohibited bases apply to all programs.) Persons with disabilities who require alternative means for communication of program information (Braille, large print, audiotape, etc.) should contact USDA’s TARGET Center at (202) 720-2600 (voice and TDD).

To file a complaint of discrimination, write to USDA, Director, Office of Civil Rights, 1400 Independence Avenue, S.W., Washington, D.C. 20250-9410, or call (800) 795-3272 (voice) or (202)720-6382 (TDD). USDA is an equal opportunity provider and employer.

Table of Contents

Summary...... 1 Chapter 1 Purpose of and Need for Action...... 1 Introduction...... 1 Purpose and Need for Action...... 3 Weed Law and Policy ...... 8 Relationship to the Forest Plan ...... 9 Supporting Documents and Analysis...... 9 Types of Actions Analyzed...... 12 Type of Impacts Analyzed...... 13 Decision to be Made ...... 14 Chapter 2 Public Participation, Issues and Alternatives...... 15 Introduction...... 15 Developing Alternatives ...... 16 Issues...... 16 Alternatives Considered But Not Studied in Detail...... 18 Alternatives Considered In Detail...... 21 Mitigation Measures ...... 28 Environmental Justice...... 31 Native American Treaty Rights ...... 31 Reasonable Foreseeable Actions...... 32 Monitoring ...... 32 Chapter 3 Affected Environment and Environmental Consequences ...... 35 Introduction...... 35 Resources ...... 41 Weeds...... 41 Range ...... 62 Forest Vegetation...... 64 Threatened, Endangered, and Sensitive Plants and Native Plant Communities ...... 70 Fish...... 103 Soils and Water...... 113 Recreation ...... 125 Human Health and Herbicide Drift...... 127 Wilderness and Inventoried Roadless Areas...... 148 Wild and Scenic Rivers...... 152 Heritage Resources ...... 154 Roads ...... 158 Wildlife ...... 160 Economics...... 196 Chapter 4 Consultation and Coordination...... 201 Preparers ...... 201 Agencies, Organizations, Businesses and People Notified...... 202 Literature Cited ...... 203 Glossary ...... 232

iii

Table of Contents

Chapter 5 DEIS Comments and FS Responses ...... 235 Comment 1 - Dick Artley...... 235 Forest Service Response...... 236 Comment 2 - Jim Brown ...... 238 Forest Service Response...... 238 Comment 3 – Phil Edholm Lookout Pass Ski and Recreation Area...... 238 Forest Service Response...... 238 Comment 4 – Ed Childers ...... 238 Forest Service Response...... 238 Comment 5 – Charles Miller and Family ...... 238 Forest Service Response...... 238 Comment 6 – Gloria Wegener and Michael Wegener ...... 238 Forest Service Response...... 239 Comment 7 - Robert F. Stewart / Office of Environmental Policy and Compliance US Department of Interior...... 239 Forest Service Response...... 240 Comment 8 - Bert Lindler ...... 240 Forest Service Response...... 240 Comment 9 - Janie Canton-Thompson...... 240 Forest Service Response...... 241 Comment 10 – Jeff Juel, WildWest Institute and Alliance for the Wild Rockies ...... 241 Forest Service Response...... 265 Comment 11 – Dwayne Garner...... 285 Forest Service Response...... 285 Comment #12 – Mack Long, Montana Fish Wildlife and Parks...... 285 Forest Service Response...... 287 Comment 13 - John F. Wardell, US Environmental Protection Agency...... 288 Forest Service Response...... 299 Comment 14 – Meggan Ryan...... 303 Forest Service Response...... 303 Appendices

A Project Area Maps By District ...... 1-5 B Screening/Adaptive Strategy Procedures...... 6 C Emergency Spill Plan ...... 7-14 D Weed Characteristics ...... 15-24 E Riparian Mitigation Measure Certification...... 25 F Fisheries Level 1 Team Recommendations ...... 26-31 G Best Management Practices for Weed Control...... 32-47 H Cost Estimate by Alternative ...... 48 I Herbicide and Target Weed Species ...... 49

List of Tables and Figures

Figures Figure 1-1 Vicinity Map ...... 2 Figure 2-1 Weed treatment decision tree ...... 24 Figure 3-1 Relationships between cover of spotted knapweed and richness and cover of native species, overall and by functional group, at knapweed invasion sites in western Montana, as determined by mixed models analysis. Functional group information is given...... 77 Figure 3-2 Average percent similarity between the experimental plots and reference stands defining potential natural communities (Rice & Toney 1998)...... 85 Figure 3-3 Average percent biomass (of total biomass) for grasses, spotted knapweed, and other forbs for eight Sawmill RNA double sampled test plots sprayed (1 pt/ac picloram) in fall of 1996 and fall 2004 (2/3 pt/ac clopyralid); as well as absolute canopy cover. * Values for cheatgrass are calculated from percent absolute canopy cover estimates, not biomass estimates...... 87 Figure 3-4 Average percent of lifeform total biomass after herbicide application Lolo National Forest winter range sites one time with a pint or less of herbicide (n = 3 to 4 sites depending on sample year) (Rice and Gauer 2007)...... 88 Figure 3-5 Average percent of life-form total biomass after herbicide application Lolo National Forest winter range one time with a quart or more of herbicide (n = 2 to 4 sites depending on sample year) (Rice and Gauer 2006)...... 89 Figure 3-6 Gibbons Pass plots in 2004 (G4), 2005 (G5), and 2006 (G6) compared to Mueggler & Stewart Idaho fescue/bluebunch wheatgrass habitat type definition plots (not labeled) (relative % canopy cover; NMS two dimensional ordination stress=15) (Rice and Gauer 2006)...... 91

Tables Table 1-1 Weeds on the Lolo NF...... 4 Table 1-2 Acres of Herbicide Treatment Allowed Under Existing NEPA ...... 7 Table 1-3 Acres of Weed Treatment Since 1997...... 13 Table 2-1 Acres proposed for treatment by Ranger District...... 23 Table 2-2 Herbicides included in Alternatives 1 and 2...... 25 Table 2-3 Comparing Alternatives ...... 27 Table 2-4 Comparing How the Alternatives Address the Purpose and Issues in This Analysis...... 27 Table 3-1 Disturbance Levels by Site Type*...... 36 Table 3-2 Acres by County ...... 37 Table 3-3 Public Knowledge about Weeds...... 40 Table 3-4 Seriousness of Weed Problem...... 40 Table 3-5 Number of Weeds Known ...... 40 Table 3-6 Ability to Name a Weed...... 41 Table 3-7 2005 Estimated Infested Acres on the Lolo NF ...... 43 Table 3-8 1998 NRIS Weed Inventory for the Lolo NF (Acres/ percent of Lolo NF) ...... 44 Table 3-9 Estimated Acres at Risk on the Lolo NF...... 45 Table 3-10 Estimated Infested Acres Compared to Acres at Risk on the Lolo NF ...... 46 Table 3-11 Broad Scale Cover Types in the Project Area and their Susceptibility to Invasion for 24 Weed Species...... 47 Table 3-12 Mormon Ridge Four Year Post Treatment Weed Canopy Cover ...... 49 Table 3-13 Status and Effects of Spotted Knapweed Biological Control Agents ...... 55 Table 3-14 Potential Vegetation Type Groups on the Lolo NF ...... 66 Table 3-15 Forest Dominance Types ...... 67 Table 3-16 G1-G3 Sensitive Species Known or Suspected to Occur on the Lolo NF ...... 70 Table 3-17 G4-G5 Sensitive Species Known or Suspected to Occur on the Lolo NF ...... 71 Table 3-18 Species of Concern Known or Suspected to Occur on Lolo National Forest...... 74 Table 3-19 Species of Interest Known or Suspected to Occur on Lolo National Forest...... 74 Table 3-20 TES Plant Species Biological Evaluation Summary ...... 96 Table 3-21 Species of Concern Biological Evaluation Summary...... 99 Table 3-22 Species of Interest Biological Evaluation Summary ...... 100 Table 3-23 Ecotoxicological Risks for Aquatic Species...... 109 Table 3-24 Characteristics of the Herbicides Proposed for Use Relative to Salmonid Fishes...... 109 Table 3-25 Determination of Effects for Bull Trout and Westslope Cutthroat Trout ...... 110

Tables Table 3-26 Biological Evaluation of Alternatives...... 111 Table 3-27 Water Characteristics ...... 115 Table 3-28 Representative Stream Flows ...... 115 Table 3-29 Margin of Safety on maximum Acute and short-term Chronic Exposure to Picloram ...... 122 Table 3-30 Montana Water Quality Human Health Standards for Herbicides...... 124 Table 3-31 Herbicides Included in Alternative 1 and 2 ...... 132 Table 3-32 Recommended Minimal Buffer Zones Associated with Different Herbicide Spray Applications in the Missoula Valley to Achieve a Reasonable Certainty of No Harm to Children (Felsot 2001)...... 144 Table 3-33 Summary of land in Wilderness, National Recreation Area and Inventoried Roadless Designation...... 149 Table 3-34 Annual Miles of projected/actual Construction and Reconstruction...... 158 Table 3-35 Management Indicator Species for the Lolo NF ...... 162 Table 3-36 Federally Listed Species Present on the Lolo National Forest ...... 167 Table 3-37 Treatment of Risk Factors as Outlined in LCAS ...... 181 Table 3-38 Sensitive Wildlife Species that are either suspected or known to occur on the Lolo NF* ...... 184 Table 3-39 Estimated Economic Impacts ...... 198

Summary

Integrated Weed Management on the Lolo National Forest

Proposed Action The Lolo National Forest (NF) proposes to implement an adaptive integrated weed management strategy to address the weed problem on the NF by: • Encouraging beneficial vegetation and weed resistant plant communities • Improving public awareness of the issues surrounding weeds • Use of weed prevention practices This integrated weed management strategy would use knowledge gained from each treatment and education program to improve on subsequent projects. It would allow managers to control new and small weed infestations before they expand, cause resource damage and become less manageable. This strategy includes: • Weed prevention and education • Combination of ground and aerially applied herbicides • Biological agents • Revegetation • Mechanical and cultural treatments • Sheep and goat grazing The weeds that would be addressed are displayed in Chapter 1 in Table 1-1. A maximum of 15,000 acres could be treated annually. This is less than 1 percent of the approximately 2.1 million acres administered by the Lolo NF. A total of 78,443 acres needing weed treatment have been identified in this analysis. Annual treatment may be less if weed budgets do not increase. Since weeds are often not continuous within a project area, actual treated acres would typically be less than the total area within any given project area. All of these weed management practices would be considered for each weed infestation. Acres proposed for treatment by Ranger District are shown in Table S-1.

Managers would select the maximum of 15,000 acres/year for treatment based on: • Funding • Invasiveness of the weed species • Size of the infestation • Potential for the seeds to move off-site • Ecological importance or rarity of a certain site The treatment priorities for this EIS include the priorities from the 1991 Lolo NF Plan Amendment 11 for noxious weed management and add others as follows: • Bunchgrass big-game winter ranges • Weed-free areas, and the roads, trails and trailheads that lead into them • Areas of concentrated public use (since people spread weeds) • Administrative sites

Final Integrated Weed Management EIS Summary-1 Summary

• National Forest lands bordering private lands with active weed-controlling programs • Disturbed areas (as described in Chapter 2) The purpose of treatments would be not only to kill the standing crop of weeds, but also to deplete the number of viable weed seeds in the soil, and encourage beneficial vegetation and weed resistant plant communities. Sites may need follow-up treatments until the viable weed seeds are depleted, regardless of what control method used. Eight to ten years is a typical viability period for the seeds of many weeds found on the Lolo NF. Treatments may not occur every year, depending on the control method used.

If herbicides were used, treatment return intervals would depend on the weed species, the size of the infestation, the soils and the persistence of the herbicide. If mechanical control or grazing was used the return period would be annually or even twice a year.

Treatment sites range in size from a few plants to infestations with a gross exterior boundary of 16,380 acres. In most areas, weeds do not cover all or even most of the project area. For example, on an infested area of two acres, only a half acre of ground might actually be occupied by scattered clumps of weeds.

Implementation could begin in calendar year 2008 and would continue until new information indicates the need for additional analysis.

Efforts would be made to coordinate treatments with private landowners managing weeds adjacent to national forest lands.

Purpose and Need for Action The overall purpose of this proposal is to raise public and agency awareness of weeds, increase both public and agency use of weed prevention practices, reduce and control existing weeds, prevent new weeds from establishing on the Lolo NF, and encourage beneficial vegetation and weed resistant plant communities. The purpose of direct weed control on the Lolo NF is to: • Prevent potential invaders from establishing • Eradicate small new infestations and reduce larger infestations of new invaders • Contain and reduce widespread weeds within infested areas • Allow rapid, timely response to new, small or recently discovered infestations before they become well established • Prevent or limit the spread of established weeds to protect weed-free areas • Implement the 2004 Forest Service National Strategy and Implementation Plan for Invasive Species Management The need for this project is: 1) Noxious weeds are present and spreading.

Populations of widespread invaders, such as spotted knapweed, sulfur cinquefoil, houndstongue and Canada thistle, are widespread and continue to expand. Since they became established, in some cases decades ago, they have increased on the Lolo NF. These weeds have also invaded both disturbed and undisturbed areas.

Summary-2 Final Integrated Weed Management EIS Summary

New invaders such as the hawkweeds and common toadflax have become established relatively recently. Some of these infestations were first seen on the Lolo NF in the last few years, while others have been here longer but have only recently been discovered. Most new invaders have limited distribution.

Potential invaders such as yellow starthistle, purple loosestrife, salt cedar and dyer’s woad have not been found on the Lolo NF yet, but do occur nearby. Based on the extent of the infestations in adjacent areas and states, these weeds have the potential to become established and impact other resources.

2) While prevention is an important weed management tool, we need to address existing weed populations on the Lolo NF.

Weed prevention and mitigation practices are used on the Lolo NF in association with direct control measures. Prevention efforts include practices such as washing off-road machinery to prevent the further spread of weeds, the 1995 Special Order requiring the use of certified weed-seed-free feed on the Lolo NF lands and the Northern Region weed prevention “Best Management Practices” (Appendix G).

Education is another prevention practice used on the Lolo NF and includes such activities as distributing posters and brochures, and the “Leave No Weeds” education program the Lolo NF developed for use in schools. These efforts need to continue to increase awareness of the weed problem and to help build a sense of personal responsibility to minimize the spread of weeds.

3) We are already doing a lot of weed management but weeds are still damaging forest resources, so we need to do more.

Forest resources are negatively impacted by existing and expanding populations of weeds. Because their naturally adapted competitors and native biological control agents are absent, weeds can out-compete and crowd out native plants.

Noxious weeds diminish the productivity, bio-diversity and appearance of Lolo NF lands. Weeds can cause habitat loss and adversely affect diversity and habitat function in plant communities.

4) We need to be able to respond quickly to new infestations and disturbances such as wildfire.

At present, we cannot respond rapidly to new infestations and disturbances. The Lolo NF is receiving more and more requests from the public to control weeds and is not able to respond to those requests.

We want to focus our weed control efforts on the infestations with the greatest potential impacts if left unchecked, those with the highest risk of further spread and those where we have the best likelihood of success. These areas are described in more detail in the Purpose and Need for Action section in Chapter 1.

5) We want to have an active weed program so we can implement and follow Federal, State and Forest Service law, regulation, policy and initiatives relating to weed management.

Final Integrated Weed Management EIS Summary-3 Summary

Forest plan and agency objectives for biodiversity, our responsibility to health and safety, cooperation with neighboring land owners and consistency with Federal and State laws require an aggressive weed program. The weed laws and policy described in Chapter 1 also direct us to have an aggressive program.

The Northern Region Overview, the Chief of the Forest Service and Montana Governor Brian Schweitzer have all identified weeds as high priority risks to National Forest lands. While we recognize that we will never eradicate all the weeds on the Lolo NF, we want to do a better job than we are doing now. We feel that this desire to do better is an integral part of responsible land management.

Issues Issues were categorized into three groups: 1) Issues outside the scope of the Proposed Action 2) Issues considered but not given further analysis 3) Issues that need further analysis – these are the issues that drive alternatives Issue outside the scope of the Proposed Action One commenter questioned the adequacy of the process used by the Environmental Protection Agency (EPA) to review and register herbicides for weed control. The EPA has regulatory responsibility regarding the evaluation, disclosure and registration of herbicides in the United States. This issue is beyond the legal authority of this analysis and the Forest Service and, as such, is outside the scope of the Proposed Action. For these reasons, this issue was not considered in this analysis.

Issue considered, but not given further analysis Why are we proposing a 15,000-acre cap on the area we could treat each year? This seems too small to address the weed infestations found on the 2.1 million acre Lolo NF.

While we recognize a biological need to treat more acres each year, this figure was proposed to provide an upper limit for effects analysis and it reflects our estimated maximum capacity to implement weed management activities on the Lolo NF.

Issues needing further analysis

Why don’t we analyze a Prevention Alternative? This issue was used to formulate an alternative, but since it wouldn’t address existing weed infestations, this alternative was not given detailed analysis. Refer to Chapter 2, Alternatives Considered, but Not Studied in Detail for more discussion. Prevention however, is included as one weed management tool in Alternative 2.

Summary-4 Final Integrated Weed Management EIS Summary

Why don’t we consider a Low Herbicide Alternative? We developed a Low Herbicide alternative to respond to this issue and it was considered, but not studied in detail. Refer to Chapter 2, Alternatives Considered, but Not Studied in Detail for more discussion.

What are the effects of weeds on wildlife and other natural resources? Many people feel that weeds are a very big problem because they’ve seen the impacts of these invasive non-native plants on vast areas. The effects of weeds on resources are discussed in Chapter 3.

How would actions taken to control weeds, especially herbicides, affect wildlife and other resources? Some people expressed concern about the effects of herbicides on wildlife and other natural resources. To respond to this issue, in Chapter 2, a No Herbicide Alternative is discussed, a Low Herbicide Alternative is considered, but not studied in detail and the No Action Alternative, which holds herbicide use at current levels, is analyzed. The effects of herbicides on natural resources, as well as models and past experience with herbicide drift from aerial application, are disclosed in Chapter 3.

What are the potential effects of herbicides on human health? Some people expressed concern about the effects of herbicides on human health. To respond to this issue, a No Herbicide Alternative is discussed in Chapter 2, Alternatives Considered, but Not in Detail. In addition, the potential human health effects of herbicides named in this proposal are disclosed in Chapter 3.

Alternatives Considered in Detail

Alternative 1, No Action Under the No Action Alternative, no new actions would be taken to respond to the Purpose and Need. Some weed control would continue to take place at current levels, namely, actions authorized in past site-specific decisions. This amounts to about 5,000 to 6,000 acres annually.

There would be no forest-wide plan to take action on new weed species or new infestations. Actions to control or eradicate new infestations not covered in existing decisions would require further NEPA analysis and documentation. These analyses would take from one to three years to complete and only if funding becomes available.

Alternative 2, Proposed Action Under this alternative, the Lolo NF would implement an adaptive integrated weed management strategy to control weeds at current levels. Adaptive Management, as used in this analysis, is defined as a dynamic approach to resource management in which the effects of treatments and decisions are continually monitored and used, along with research

Final Integrated Weed Management EIS Summary-5 Summary results, to modify management on a continuing basis to ensure that objectives are being met. It would also allow managers to control new and small weed infestations before they can expand, have a greater impact on other resources and become less manageable.

This strategy includes a prevention and education component, together with a combination of ground and aerially applied herbicides, biological agents, revegetation, mechanical and cultural treatments, and sheep and goat grazing on a maximum of 15,000 acres/year. Managers will be able to prioritize annual weed treatments to where we can have the greatest benefit.

Annual treatment priorities are described above under the Proposed Action heading in this summary. Most weed treatments would require re-treatments until the weed seed is depleted in the soil seed bank (for direct controls) and until public prevention measures are commonly used (for education measures). Re-treatment would be least frequent for herbicides (1 to 3 growing seasons). Re-treatment would be most frequent for grazing, mowing or pulling (1 to 2 times per growing season).

Table S-1 Acres proposed for treatment by Ranger District Ranger District Acres Missoula 7,082 Ninemile 31,351 Plains/Thompson Falls 35,949 Seeley Lake 2,292 Superior 1,769 TOTAL 78,443 When new weed infestations are found, they would be screened to determine priorities and what kind of treatment to use. A Decision Tree (Figure 2-1 page 24 in Chapter 2) would be used to determine the most appropriate method(s) to use to maximize weed control and minimize the risk of adverse environmental effects. Mitigation measures would be used to minimize any adverse effects on human health and other resources. • Herbicides would be used at or below the rates specified on the label or in the Lolo NF Plan, whichever is more restrictive. • Aerial application would be considered and used according to the mitigation measures and subject to the criteria described in Chapter 2. • This alternative includes an adaptive approach that would allow managers to learn from the past and improve treatments in the future.

Alternatives Considered but Not Given Detailed Study Additional alternatives were considered but not given detailed study. These alternatives were: • Prevention only • No herbicides • Low herbicide use • Concentrated eradication cells • No aerial herbicide application The reasons these alternatives were not given detailed study are discussed in Chapter 2.

Summary-6 Final Integrated Weed Management EIS Summary

Affected Environment The analysis area occurs across the Lolo NF. The Lolo NF stretches for 120 miles in a northwest to southeast direction and lies in west-central Montana. The forest extends from the Idaho border east. The forest is 40 to 80 miles wide in a southwest and northeast direction. The Clark Fork of the Columbia River is the major drainage. Interstate 90 bisects the forest. Chapter 3 describes which and how resources may be affected.

Environmental Consequences Chapter 3 discloses the potential environmental consequences of the alternatives. Direct effects are caused by the action and occur at the same time or place. Indirect effects are caused by the action and are later in time or farther removed in distance, but still reasonable foreseeable. Cumulative effects are the impacts which result from the action when added to other past, present, and reasonable foreseeable future actions regardless of which agency or person undertakes such actions.

Comparison of the Alternatives The main differences between Alternative 1 and Alternative 2 are that in Alternative 2: • We could add sheep and goat grazing in our weed management “tool box” • We would be able to better prioritize which weed infestations we treated each year, our intention being to get the greatest benefit for our weed management dollar • We would be able to treat new and recently discovered weed infestations • We could treat up to 15,000 acres/year if funding allowed. Otherwise we would treat our historical level of about 5,000 to 6,000 acres/year • Alternative 2 would put an increased emphasis on encouraging beneficial vegetation and weed resistant plant communities, and establishing native and desirable vegetation rather than just killing weeds The following table briefly summarizes how each alternative relates to the Purpose for this analysis and the effects of the alternatives by issue. More information on the issues is available in Chapter 2. A detailed disclosure of environmental consequences is presented in Chapter 3.

Final Integrated Weed Management EIS Summary-7 Summary

Table S-2 Comparison of how the alternatives address the Purpose and Issues in this analysis.

Alternative 1, No Action Alternative 2, Proposed Action Purpose: Prevent potential invaders from establishing. Includes full range of prevention measures and BMPs, but Same as Alternative 1 but allows for limits ability to treat new invaders treatment of new invaders as funded. Purpose: Eradicate small new infestations and reduce larger infestations of new invaders. Minimally meets purpose, because it does not allow rapid Best meets purpose, because it allows rapid response to small new infestations without additional treatment of both small and large new NEPA analysis. infestations. Purpose: Contain and reduce widespread weeds within infested areas. Moderately meets purpose on sites with existing NEPA Best meets purpose, because it allows rapid decisions, but does not allow containment or reduction of containment and reduction of more widespread widespread weeds unless they are covered by an existing weeds that are not covered by an existing NEPA NEPA decision. decision. Purpose: Allow rapid, timely response to new, small or recently discovered infestations before they become well established. Minimally meets purpose, because it does not allow rapid Best meets purpose, by allowing rapid, timely response without additional NEPA analysis, which usually response to widest range of situations. takes from 1 to 3 years. Purpose: Prevent or limit the spread of established weeds in weed free areas. Meets purpose only in places where there is an existing Best meets purpose, by allowing rapid response NEPA decision. to the widest range of situations. Purpose: Implement the 2004 Forest Service National Strategy and Implementation Plan for Invasive Species Management. Implements the strategy for sites with existing NEPA Implements the strategy by providing flexible decisions. Does not implement the strategy on new and rapid response while ensuring other infestations or after disturbances such as wildfire. resource protection. Issue: What are the effects of weeds on wildlife and natural resources? Impacts of weeds include: • Changes in plant community composition and structure. • Loss of sensitive plant populations. • Effects on water and soil quality. • Effects to people allergic to weeds Provides lower level of protection for native plants from Provides higher level of protection for native weeds. No reduction of weed-caused impacts to soil and plants from weeds. Reduces weed-caused water on sites without existing NEPA decisions. impacts to soil and water and people. Issue: How would actions taken to control weeds, especially herbicides, affect wildlife and natural resources? Impacts of herbicides include effects to: Fish and terrestrial animals, Non-target plants, and soil and water quality. Annual maximum treatment would occur on On average, treatment would occur on less than ¼ of 1% less than ¾ of 1% of the acres on the Lolo NF. of the acres on the Lolo NF. Effects would be extremely Low effect due to limited acres treated. low due to limited acres treated and mitigation measures Mitigation measures would further minimize applied. effects. Issue: What are the potential effects of herbicides on human health (chemically sensitive individuals, people allergic to herbicides and general public)? Annual average treatment would occur on less On average, treatment would occur annually on less than than ¾ of 1% of the acres on the Lolo NF. Effects ¼ of 1% of the acres on the Lolo NF. Effects would be would be below established toxicological below established toxicological thresholds due to limited thresholds due to limited acres treated and acres treated and mitigation measures applied. Mitigation mitigation measures. Mitigation would help would help sensitive people avoid treatment sites. sensitive people avoid treatment sites.

Summary-8 Final Integrated Weed Management EIS

Chapter 1 Purpose of and Need for Action

Introduction This Environmental Impact Statement (EIS) has been prepared to disclose the environmental effects of a Forest Service (FS) proposal to implement an adaptive integrated pest management strategy to control or reduce the presence of noxious and invasive weeds on the Lolo National Forest (Lolo NF). The effects of the No Action Alternative are also disclosed. The strategy proposed would be applied on lands administered by the Lolo NF in Flathead, Granite, Lake, Mineral, Missoula, Powell and Sanders Counties. The Lolo NF encompasses approximately 2.1 million acres in western Montana. The Lolo NF shares boundaries with the Beaverhead-Deerlodge, Bitterroot, Clearwater, Flathead, Helena, Kootenai, Idaho-Panhandle National Forests (NF), Bureau of Land Management (BLM), State of Montana, and private lands. Refer to the vicinity map (Figure 1-1) on the next page.

Weeds pose an increasing threat to native ecosystems, croplands, other plant communities, and human health. While weeds have long been recognized as a problem for agriculture, the potential impact to other plant communities, including wildlands, is receiving greater attention. An estimated 2,000 invasive and noxious weed species are already established in the United States. Twenty one invasive and noxious weed species have been found on the Lolo NF. Another 10 noxious and invasive species are found nearby but are not yet established on the forest (see Table 1-1). All types of ecosystems are vulnerable to invasive plants.

Experience and research show that these plants are no longer just a problem on disturbed sites. Noxious and invasive plant species have become established in undisturbed ecosystems, including specially designated areas such as wilderness and research natural areas.

In response to growing concerns about these plants, the U.S. Congress has passed federal laws addressing actions against noxious weed infestations. The State of Montana has the Montana County Noxious Weed Management Act. Forest Service policy has been developed to implement these laws and give direction for FS activities. The Lolo Forest Plan follows these laws and policies for weed management on the Lolo NF.

The Purpose and Need for this proposal and the components of the proposal described in this document were based on the situation described above, as well as knowledge of noxious and invasive weeds on the Lolo NF.

The analysis documentation is located in the Project Record and is available for public review at The Lolo National Forest Supervisor’s Office in Missoula, Montana.

Chapter 1 Purpose and Need

Figure 1-1 Vicinity Map

Page 2 Final Integrated Weed Management EIS Chapter 1 Purpose and Need

Purpose and Need for Action

Background Non-native plants have been a part of North America ecosystems for many decades. Some, such as mint or barley, have been accepted and used as part of this country’s agricultural heritage. Others have been designated as undesirable.

In order to maintain consistency during this analysis, use of the term “weed” refers to the following definitions: • The Federal Noxious Weed Act says a noxious weed is “a plant which is of foreign origin, is new to, or is not widely prevalent in the United States, and can directly or indirectly injure crops or other useful plants, livestock or the fish and wildlife resources of the United States, or the public health” (P.L. 93-629). • The Montana Noxious Weed Control Act says, “… any exotic plant species established or potentially established in the State and which may render land unfit for agriculture, forestry, livestock, wildlife, or other beneficial uses, and is further designated as either a state-wide or county-wide noxious weed” (MCA 7-22-2101). • Forest Service Manual 2080 defines noxious weeds as “those plant species designated as noxious weeds by the Secretary of Agriculture or by the responsible State official. Noxious weeds generally possess one or more of the following characteristics: aggressive and difficult to manage, poisonous, toxic, parasitic, a carrier or host to serious insects or disease and being non-native or new to or not common to the United States or parts thereof.” Exotic plants (those not native to North America) continue to become established, both on and off NF lands. Many occur only in disturbed areas, but some weeds are found almost everywhere.

A small percentage of exotic species have been designated noxious weeds, highly invasive, with the potential to colonize and dominate native plant communities, and are severely altering the natural ecological functioning of these environments. Many invasive plants have become established and are widespread invaders throughout the Lolo NF, whereas the new invaders are just getting a foothold.

Other weeds have yet to be seen on the Lolo NF, but are known to occur in adjacent counties or states. These potential invaders could greatly impact native plant diversity and wildlife habitat.

Table 1-1 shows a list of noxious and invasive weeds of specific concern on the Lolo NF, grouped as widespread, new and potential invaders. These categories would be used to help determine treatment priorities during project implementation. That process is described in Chapter 2.

Final Integrated Weed Management EIS Page 3 Chapter 1 Purpose and Need

Throughout the remainder of this document, the term weed will refer to one of the noxious or invasive plants included in Table 1-1.

Table 1-1 Weeds on the Lolo NF Scientific Name Common Name Potential invaders (currently absent) Goal is to prevent and eradicate promptly if found Bryonia alba White bryony Centaurea repens Russian knapweed Centaurea solstitalis Yellow starthistle Chondrilla juncea Rush skeletonweed Crupina vulgaris Common crupina Iris pseudacorus Yellowflag iris Isatis tinctoria Dyers woad Lythrum spp. Purple loostrife Polgonum cuspidatum Japanese Knotweed Tamarix spp. Salt cedar New invaders Goal is to eradicate small new infestations and reduce larger infestations Cardaria draba White top Centaurea diffusa Diffuse knapweed Echium vulgare Blue weed Hieracium aurantiacum Orange hawkweed Hieracium piloselloides / pratense/ et al Meadow/Yellow hawkweed complex Lepidium latifolium Perennial pepperweed Linaria vulgaris Common toadflax Ranunculus acris Tall buttercup Senecio jacobaea Tansy ragwort Widespread Weeds Goal is to contain inside infested area and reduce plant populations Cardus nutans Musk thistle Centaurea maculosa Spotted knapweed Chrysanthemum leucanthemem Oxeye daisy Cirsium arvense Canada thistle Cynoglossum officinale Houndstongue Euphorbia esula Leafy spurge Hypericum perforatum St. Johnswort Linaria dalmatica Dalmatian toadflax Potentilla recta Sulfur cinquefoil Sisymbrium altissimum Tumble mustard Tanacetum vulgare Common tansy Verbascum spp. Mullein Purpose The overall purpose of this proposal is to establish beneficial and weed resistant plant communities by raising public and agency awareness of weeds, increasing both public and agency use of weed prevention practices, reducing and control existing weeds, and preventing new weeds from establishing on the Lolo NF. The purpose of direct weed control on the Lolo NF is to:

Page 4 Final Integrated Weed Management EIS Chapter 1 Purpose and Need

• Prevent potential invaders from establishing • Eradicate small new infestations and reduce larger infestations of new invaders • Contain and reduce widespread weeds within infested areas • Allow rapid, timely response to new, small or recently discovered infestations before they become well established • Prevent or limit the spread of established weeds to protect weed-free areas • Implement the 2004 Forest Service National Strategy and Implementation Plan for Invasive Species Management Need 1) Weeds are present and spreading.

2) While prevention is an important weed management tool, we need to address existing weed populations on the Lolo NF.

Weed prevention and mitigation practices are used on the Lolo NF in association with direct control measures. Prevention efforts include practices such as washing off-road machinery to prevent the further spread of weeds, the 1995 Special Order requiring the use of certified weed-seed-free feed on the Lolo NF lands, and the Northern Region weed prevention “Best Management Practices” (Appendix G).

Education is another prevention practice used on the Lolo NF and includes such activities as distributing posters and brochures, and the “Leave No Weeds” program the Lolo NF developed for use in schools. These efforts need to continue to increase awareness of the weed problem and to help build a sense of personal responsibility to minimize the spread of weeds.

3) We are already doing a lot of weed management but weeds are still damaging forest resources, so we need to do more.

Final Integrated Weed Management EIS Page 5 Chapter 1 Purpose and Need

Weeds diminish the productivity, bio-diversity and appearance of NF lands. Weeds can cause habitat loss and adversely affect diversity and habitat function in plant communities.

Mechanical and cultural methods of weed control are both being used but have limited effectiveness because of the large scale of many weed infestations on the Lolo NF and the characteristics of the weeds found here. For example, roadside mowing does not prevent spotted knapweed from flowering and going to seed unless it is conducted during a very specific and short stage of growth. Hand pulling, the principal method of mechanical control, is effective on individual plants or small, isolated populations, but is unrealistic and unaffordable on large or widespread weed infestations. When soil conditions are dry, it can be hard to pull enough of the root to kill a plant. Hand-pulling is also not effective on deep- rooted rhizomatous weeds like leafy spurge, St. Johnswort, common toadflax and Canada thistle.

Biological control agents, organisms that feed on weeds, have been released on the Lolo NF since 1978. Most of the biological control agents released on the Lolo NF have been insects. They can consume or weaken weeds, or at least reduce seed production. Since 1978, we have conducted more than 400 releases of 35 different biological control agents, targeting Canada thistle, spotted knapweed, diffuse knapweed, leafy spurge, St. Johnswort, Dalmatian toadflax and musk thistle. Biological agents have also spread onto the Lolo NF from releases on adjacent lands.

Biological management in the best case can control but not eradicate or prevent the spread of weeds. Biological agents can reduce weed density and rate of spread, but not eradicate weeds. Optimal biological management is a combination of several agents – some to attack the flowers and seeds, others stressing the stems and leaves, and others boring into the roots.

The results of releases on the Lolo NF have been sporadic. Some biological agents have noticeably reduced some weed infestations. Others have had little apparent effect so far. While some biological agents can fly and readily spread to other weed infestations, others have to crawl from weed to weed, which slows their rate of spread. Some biological agents cannot survive the annual variations in our weather. Other insects are difficult to obtain. Testing for others is in experimental stages, so they aren’t widely available yet. Since biological control is considered a long-term effort, it may be too early to see the effects of some biological agents. Some researchers are even beginning to question the wisdom of introducing exotic insects into our ecosystems.

Herbicides have been used in limited areas, as shown in Table 1-2. Selective herbicides, those that do not kill all the plants they contact, are what we commonly use. To date, herbicides have been a useful weed control tool. As with all weed control measures, herbicides have their detractors – some people are concerned about the ecosystem and human health impacts of herbicides.

Page 6 Final Integrated Weed Management EIS Chapter 1 Purpose and Need

Table 1-2 Acres of Herbicide Treatment Allowed Under Existing NEPA Ranger District Area Missoula 2,700 acres Ninemile 760 acres Plains/Thompson Falls 2,610 acres Seeley Lake 1,730 acres Superior 2,530 acres Total 10,330 acres 4) We need to be able to respond quickly to new infestations and disturbances such as wildfire. At present, we cannot respond rapidly to new infestations and disturbances. The Lolo NF is receiving more and more requests from the public to control weeds and is not able to respond to those requests. Many recently located infestations or new weed species are not covered under existing decisions.

We are committed to doing sound, comprehensive analysis and planning, but funding and staff is limited. Environmental analysis leading to a decision to treat new infestations takes a long time to complete, from one to three years, and can cost hundreds of thousands of dollars. During this time, new infestations can become well established and spread to surrounding areas, becoming difficult or impossible to control. We feel it is important to commit a significant proportion of our weed funding to on the ground management.

We want to focus our weed control efforts on infestations where we can realize the greatest resource benefits; those with the highest risk of further spread and those where we have the best likelihood of success.

Under Lolo NF Plan Amendment 11, we prioritized our weed control efforts on: • Bunchgrass big-game winter ranges • Weed-free areas, and the roads, trails and trailheads that lead into them • Areas of concentrated public use (since people spread weeds) • Administrative sites • NF lands bordering private lands with active weed-controlling programs We need to respond to weeds quickly and expand our priorities to include: • Other disturbed sites • Undisturbed sites where weeds have invaded Developing prompt treatment methods that protect other resources, like water quality and rare plants and animals, would improve resource benefits, cost effectiveness, and respond better to public requests for weed treatments.

5) We want to have an active weed program so we can implement and follow Federal, State and Forest Service law, regulation, policy and initiatives relating to weed management.

Forest plan and agency objectives for biodiversity, our responsibility to health and safety, cooperation with neighboring land owners and consistency with Federal and State laws require an aggressive weed program. The weed laws and policy described in the next section of this chapter also direct us to have an aggressive program.

Final Integrated Weed Management EIS Page 7 Chapter 1 Purpose and Need

Our present inability to keep up with weed infestations has lead to a continuing, compounding annual loss of desirable habitat to weeds and the potential that NF lands could infest adjacent lands or reduce the effectiveness of control efforts by neighboring landowners.

This is inconsistent with Lolo NF Plan Amendment 11 and Northern Region goals for sustaining native plant diversity and desired community function.

The Northern Region Overview, the Chief of the Forest Service and Montana Governor Brian Schweitzer have all identified weeds as high priority risks to National Forest lands. While we recognize that we would never eradicate all the weeds on the Lolo NF, we want to do a better job than we are doing now.

Weed Law and Policy Direction and authority for weed management on National Forest lands comes from a wide variety of laws, policy and direction. These include the National Forest Management Act (PL 94-588, NFMA), the National Environmental Policy Act (PL 91-190, NEPA), the Federal Land Policy and Management Act (PL 94-579, FLPMA), the Carlson-Foley Act (PL 90-583), the Federal Noxious Weed Control Act (PL 93-629) and Presidential Executive Order (EO 13112). • NFMA, NEPA and FLPMA provide general land management and environmental analysis direction. • The Carlson-Foley Act authorizes and directs federal agencies to permit control of noxious plants by state and local governments on a reimbursement basis in connection with similar weed control programs carried out on adjacent non- federal land. • The Federal Noxious Weed Act defines weeds, and authorizes cooperative weed control agreements between federal agencies and other agencies, organizations or individuals. • In Montana, the Montana County Noxious Weed Management Act (MCA 7-22- 2101) says that it’s unlawful for any person to allow weeds to propagate or go to seed on their land unless they have an approved weed management plan. This act directs the counties to develop weed control districts to plan and implement weed control efforts. • 36 CFR Subpart A, Section 222.8 directs the Forest Service to cooperate with local weed control districts to develop control programs where there are National Forests and Grasslands. • Executive Order 13112 of 1999, directs federal agencies to reduce weeds. In addition to the above, weed management direction and support is provided in the following: • On January 25, 2005, Montana Governor Brian Schweitzer told the Montana Weed Control Association that weeds were Montana’s greatest environmental threat. • The 1998 Forest Service Natural Resource Agenda placed a strong emphasis on conserving and restoring degraded ecosystems as a management priority for the 21st Century, including actions to “attain desirable plant communities and prevent exotic organisms from entering or spreading in the United States.”

Page 8 Final Integrated Weed Management EIS Chapter 1 Purpose and Need

• The 1998 Northern Region Overview addressed priority needs for ecosystem health and recreation, saying, “Noxious weeds are one of the most serious threats to ecological integrity.” It identified integrated weed control as a priority action. • The 1998 Forest Service Strategy for Noxious and Nonnative Invasive Plant Management provided a “roadmap into the future for preventing and controlling the spread of noxious weeds and non-native invasive plants.” • Forest Service Manual 2259.03 states “Forest officers shall cooperate fully with State, County and Federal officials in implementing 36 CFR 222.8 and sections one and two of Public Law 90-583. Within budgetary constraints, the Forest Service shall control to the extent practical, noxious farm weeds on all National Forest System lands.” • The 2004 National Strategy and Implementation Plan for Invasive Species Management identifies the Forest Service as one of the lead agencies in the fight against invasive plants. It provides long-term direction to reduce, minimize or eliminate invasive species across all landscapes and ownerships by improving the management of invasive species using science-based technology, by emphasizing partnerships, and by increasing performance and accountability, as well as communication and education. Relationship to the Forest Plan This is a site-specific analysis. The Proposed Action complies with the 1986 Lolo National Forest Plan and Lolo NF Plan Amendment 11, which addresses weed management. This analysis focuses on specific issues pertaining to the project areas; it is not a general management plan for the project areas. It provides site-specific direction for implementing the Lolo NF Forest Plan in compliance with NEPA requirements.

Lolo NF Plan Amendment 11 directs weed control projects to focus where they would have the greatest effect on preventing the spread of weeds or damage to natural resources and where they would have the greatest benefit to people who are actively trying to control weeds on land adjacent to the Lolo NF. Plan Amendment 11 further says that priority would be given to weed control projects in areas relatively free of weeds, the roads that lead into those areas, and to new infestations and small patches that threaten areas at high or moderate risk of weed invasion. The actions included in this analysis are designed to implement that direction.

Supporting Documents and Analysis The potential impacts of proposed weed control activities are analyzed in this EIS. This EIS would incorporate by reference the guidelines, findings, and analysis described in the following documents: • Lolo Forest Plan (USDA-FS 1986) • Lolo NF Plan Amendment 11, the FEIS and ROD for Noxious Weed Management on the Lolo NF (USDA-FS 1991) • FS Strategy for Noxious and Nonnative Invasive Plant Management (USDA-FS 1998) • Risk Assessment for Herbicide Use in FS Regions 1, 2, 3, 4 and 10 and on Bonneville Power Administration Sites (USDA-FS 1992) • FEIS and ROD for Big Game Winter Range and Burned Area Weed Management on the Lolo NF (USDA-FS 2001)

Final Integrated Weed Management EIS Page 9 Chapter 1 Purpose and Need

• Herbicide Ecological and Human Health Risk Assessments prepared by/for the USDA-FS by Syracuse Environmental research associates (SERA) (http://www.fs.fed.us/foresthealth/pesticide/) These documents are available for review at the Lolo NF Supervisor’s Office or on the internet. Findings covered by these documents are included or considered in this EIS.

Proposed Action A more detailed discussion of this proposal is found in Chapter 2, where the alternatives are described in detail. The Proposed Action is described in Chapter 2 as Alternative 2.

The Lolo NF proposes to implement an adaptive integrated weed management strategy to address the weed problem by establishing beneficial vegetation and weed resistant plant communities, improving public awareness of the issues surrounding weeds and the use of weed prevention practices. The strategy would use knowledge gained from each program and treatment to improve on subsequent projects. It would allow managers to control new and small weed infestations before they can expand. This strategy includes a prevention and education component, together with a combination of ground and aerially applied herbicides, biological agents, revegetation, mechanical and cultural treatments, and sheep and goat grazing on a maximum of 15,000 acres/year. All of these weed control practices would be considered for each weed infestation.

The weeds listed in Table 1-1 would be considered for treatment subject to a number of factors, such as funding, invasiveness of the species, size of the infestation, potential for the seeds to move off-site, and the ecological importance or rarity of a certain site. Acres proposed for treatment by Ranger District are shown in Table 2-1.

The treatment priorities for this EIS include the priorities from the Lolo NF Plan Amendment 11 and others as follows: • Bunchgrass big-game winter ranges • Weed-free areas, and the roads, trails and trailheads that lead into them • Areas of concentrated public use (since people spread weeds) • Administrative sites • National Forest lands bordering private lands with active weed-controlling programs • Disturbed areas, which include but are not limited to administrative pastures, ranger stations, work centers and lookouts, burned areas, airstrips, helispots, fire camps, landings, staging areas, project camps, campgrounds, recreation areas, trailheads, recreation and electronic sites, gravel pits, mines, rock operations, power and utility rights-of-way, firelines, roads, trails, ski areas and range sites • Less or undisturbed areas where weeds are found. Weeds found in roadless and wilderness areas would be considered for treatment in compliance with wilderness legislation. Weeds found in National Research Areas would be considered for treatment in compliance with Forest Service Direction Newly discovered infestations could be added and screened to determine the type of treatment and treatment priority. The decision trees in Chapter 2 would be used to make sure water and TES plants and animals were protected. Mitigation measures described in Chapter 2 would be utilized to protect human health and other resources.

Page 10 Final Integrated Weed Management EIS Chapter 1 Purpose and Need

If herbicides were determined an appropriate control method for a particular site and weed species, herbicides considered for use would be limited to those that: • Are registered for weed control by the Environmental Protection Agency (EPA) • Have a current Forest Service Ecological and Human Health Risk Assessment (http://www.fs.fed.us/foresthealth/pesticide/) or a Line Officer review and approval • The herbicides proposed for use are listed in Table 2-2 in Chapter 2. This list could be expanded if and when other products are registered by the EPA. Dyes would be used in some case to indicate where herbicides were applied All herbicides would be applied by or supervised by state-licensed applicators or operators following current laws.

Herbicides would be handled following EPA label requirements and other state and federal laws for storage, mixing and application. EPA label requirements have the force of law and include measures to reduce risks to the environment and human health.

Under the Proposed Action: • A maximum of 15,000 acres would be treated annually. This is less than on1 percent of the approximately 2.1 million acres administered by the Lolo NF. A total of 78,443 acres needing weed treatment have been identified in this analysis. No more than 15,000 acres would be treated in one year. Since weeds are often not continuous within a project area, actual treated acres would typically be less than the total area within any given project area. • The purpose of treatments would be not only to kill the standing crop of weeds, but to deplete the number of viable weed seeds in the soil. Sites may need follow-up treatments until the viable seeds are depleted, regardless of what control method is used. Eight to ten years is a typical viability period for the seeds of many weeds found on the Lolo NF. Treatments may not occur every year, depending on the control method used. • If herbicides were used, treatment return intervals would depend on the weed species, the size of the infestation, the soils and the persistence of the herbicide. Treatment sites range in size from less than an acre to infestations with a gross exterior boundary of 16,380 acres. In most areas, weeds do not cover all of the project area. For example, on an infested area of two acres, only a half acre of ground might actually be occupied by scattered clumps of weeds.

Implementation could begin in calendar year 2008 and would continue until new information indicates the need for additional analysis.

Scope of Proposed Action The scope of this analysis is framed by the Purpose and Need earlier in this chapter. The scope of this analysis does not include eliminating people or human activities, even though people are often the means by which weed seeds are spread. Human activities, such as recreation, wildland fire suppression, road use, and forest access, would not be changed by this action, nor would it change land use direction in the Lolo Forest Plan.

Final Integrated Weed Management EIS Page 11 Chapter 1 Purpose and Need

The Proposed Action is limited to the management activities described in Chapter 2. The Proposed Action includes weed treatments only on lands administered by the Lolo NF. Activities could occur in any of the management areas described in the Lolo Forest Plan. The goals and objectives of these management areas are described in the Lolo Forest Plan.

This EIS documents the site-specific effects of the Proposed Action and the No Action Alternative. It is not a general management plan, nor is it programmatic. No further decisions would need to be made under NEPA before the selected alternative could be implemented.

The treatments described in this analysis would not all occur during a single year. Actual treatment scheduling would depend on budgets, weather, the changes in each infestation, and natural and man-made disturbances.

Any resource management decisions other than weed management as described in the Proposed Action are beyond the scope of this analysis.

Scope of the Analysis The analyses of effects disclosed in this document include any occurring from the entire project. Scope is defined in 40 CFR 1508.25 as “the range of actions, alternatives and impacts, such as connected and cumulative actions and impacts” considered in this EIS.”

Types of Actions Analyzed

Connected Actions Connected actions are closely related to the Proposed Action. Actions are connected if they: • Automatically trigger other actions that may require environmental analysis • Cannot or would not proceed unless other actions are taken previously or simultaneously • Are independent parts of a larger action and depend on the larger actions for their justification The Proposed Action includes activities necessary to fulfill the Purpose and Need. No other connected actions have been identified.

Cumulative Actions Cumulative actions are actions which, when viewed together with past, other present and reasonably foreseeable actions, have cumulatively significant impacts.

Past Actions: Past actions include considering weed treatments conducted inside and adjacent to the Lolo NF. Cumulative actions would be included at least back to the 1991 Lolo Weed EIS. The acres of direct weed control on the Lolo NF since 1997 are shown in Table 1-3.

Page 12 Final Integrated Weed Management EIS Chapter 1 Purpose and Need

Table 1-3 Acres of Weed Treatment Since 1997 Direct Control 1997 1998 1999 2000 2001 Herbicide 662 804 871 237 1,757 Mechanical 10 5 5 5 3 Biological 1,650 1,300 1,800 590 1,425 Conducted Conducted Conducted Conducted Conducted but but not but not but not but not Cultural not quantified quantified quantified quantified quantified Acres of control methods funded by Weed Funds Weed Fund Totals 2,322 2,109 2,676 832 3,194 Acres herbicide control funded by other funds Other Fund Totals 830 792 469 1,231 470 Total All Funds 3,152 2,901 3,145 2,063 3,664

Table 1-3 Acres of weed Treatment Since 1997 continued

Direct Control 2002 2003 2004 2005 Herbicide 1,728 2,293 482 1,641 Mechanical 11 13 17 36 Biological 1,930 360 230 320 Cultural 490 441 2,082 422 Acres of control methods funded by Weed Funds Weed Fund Totals 4,159 3,107 2,811 2,419 Acres herbicide control funded by other funds Other Fund Totals 4,399 1,845 4,028 2,120 Total All Funds 8,558 4,952 6,839 4,539 Present Actions: Present actions include weed treatment conducted within and adjacent to the Lolo NF by the FS, by private landowners, and by state, county or city governments.

Reasonably Foreseeable Actions: Reasonably foreseeable actions include weed treatments anticipated to occur inside and adjacent to the Lolo NF during the next five to ten years. Weed treatment on lands acquired through land exchange might be an example.

Similar Actions Similar actions are actions which, when viewed with other reasonably foreseeable or proposed actions, have similarities that provide a basis for evaluating their environmental consequences together, such as common timing or geography.

There are no known actions similar to the Proposed Action.

Type of Impacts Analyzed The scope of the analysis includes consideration of three types of impacts, direct, indirect and cumulative. The definitions of these impacts or effects are contained in 40 CFR 1508.7 and 1508.8.

Final Integrated Weed Management EIS Page 13 Chapter 1 Purpose and Need

Direct Effects Direct effects are those caused by the Proposed Action and occurring at the same time and place. They are disclosed in Environmental Consequences.

Indirect Effects Indirect effects are caused by the Proposed Action, but are later in time or farther removed in distance. Still, they are reasonably foreseeable, and are disclosed in Environmental Consequences alongside direct effects in most cases.

Cumulative Effects Cumulative effects result from incremental impacts of the Proposed Action when added to other past, present and reasonably foreseeable future actions, regardless of what agency or person undertakes the other actions. Cumulative effects are considered for each resource and disclosed in Environmental Consequences.

Decision to be Made The Lolo NF Forest Supervisor is the responsible official for making the decision on this proposal. The decision would address the following points: • Whether to authorize the Lolo NF to expand efforts to control weeds using an adaptive strategy • What specific monitoring requirements are necessary to ensure that treatments and mitigation measures are implemented Decisions that would not be made based on this analysis are: • Changes in land use and forest plan direction • Changes in the level of wildland suppression, strategies, tactics and whether or not to control wildfire • Changes in travel management, road use and forest access • Prevention measures that minimize the establishment and spread of weeds that are already part of FS policy and recent decisions • Environmental Protection Agency established Reference Doses, related EPA toxicological thresholds, and herbicide registration processes • Ecological and toxicological conclusions and data included in the Forest Service/Syracuse Environmental Research Associates Human Health and Ecological Risk Assessments (RAs) and related documents

Page 14 Final Integrated Weed Management EIS

Chapter 2 Public Participation, Issues and Alternatives

Introduction This chapter describes the public involvement and issue development process used to design and develop weed management alternatives for the Lolo NF. Issues that were identified by the public and the agency are described. It also summarizes how the public participated in the analysis process and how the interdisciplinary team (ID team) used this input.

The Proposed Action and other alternatives are described and compared. Features of the alternatives are discussed. Mitigation measures are described.

Chapter 2 changes made in response to comments on the DEIS Fourteen comments were received on the DEIS. These comments can be found verbatim along with responses and specific changes that were made in the FEIS Chapter 5, pages 236-303.

DEIS Page 18 - Deleted two bullets under “This alternative was not given detailed analysis because:” as follows: • We need to be able to quantify the amount of weed management to be able to analyze effects – we can not analyze the effects of an unknown quantity of weed control. • While the FS does not dispute the contention that human activity is a primary means by which weeds are spread, excluding humans from the Lolo NF would be inconsistent with the Lolo Forest Plan. DEIS Page 24 – Aerial applications were clarified DEIS Page 27 – Mitigation number 35 added to read: If commercial applicators are used for the application of restricted use pesticides, Forest Service contract administrators will check to make sure their Montana commercial restricted use pesticide license is current. (With the addition of this mitigation measure, the subsequent mitigation numbers also changed) DEIS Page 27 – Clarified Sheep and Goat Grazing mitigation measure number 15

DEIS Page 28 - Herbicide mitigation number 28 was clarified.

DEIS Page 29 - Under, “Aerial Application Mitigation Measures,” clarified number 40 and added 45-48. DEIS Page 31 – Clarified the fourth bullet.

Chapter 2 Public Participation, Issues and Alternatives

Developing Alternatives

Public Participation The FS identifies public concerns and possible effects of a proposal on NF land through public participation. It is also a means to disclose the nature and probable consequences of the Proposed Action to the public. The public involvement strategy for this project provided government agencies and interested people with timely information so they could fully participate in the planning process.

A scoping letter was sent out on July 16, 2004 to 268 people, government agencies, tribal interests, news media, businesses and organizations that have shown interest in similar projects on the Lolo NF. This letter described the Purpose and Need for the project, the Proposed Action, and asked for comments. People were asked to comment within a 30 day period ending August 20, 2004.

During the same time frame, the scoping letter was sent as a news release to six local news organizations. In addition, the proposal was the subject of a news article in the Missoulian on August 1, 2004 and an interview on KTMF-TV on August 11, 2004.

A Notice of Intent was published in the Federal Register on August 4, 2004. People were asked to comment within a 45 day period, ending September 20, 2004.

The Flathead Culture Committee of the Consolidated Salish and Kootenai tribes was asked for comment.

This media and scoping publicized the proposal. Comments received from the public identified specific areas of concern and issues during the early phase of this environmental analysis. Twenty-nine letters, personal comments and phone calls were received and considered by the ID team and the responsible official. These documents are included in the Project Record at the Lolo NF Supervisor’s Office.

Issues Issues are identified through scoping, as they are raised by the public, other government agencies and the FS. Important environmental issues are identified to focus the environmental analysis. Potential environmental issues are discussed in proportion to their importance, and issues that don’t pertain to the action are eliminated.

The ID team’s review of the comments showed that two-thirds agreed weeds are an urgent concern on the Lolo NF and that steps should be taken quickly to reduce or eliminate them. All of these comments supported the use of herbicides as part of the proposal. Another concern was that the maximum annual acres treated would be too small. Some people had questions or concerns about herbicides. They questioned the need for herbicides or expressed concern about the effects of herbicides.

The ID team found the primary concerns with the Proposed Action centered on whether there was a need and urgency to control weeds, and whether herbicides should be used as part of the Integrated Pest Management (IPM) strategy. The ID team evaluated these

Page 16 Final Integrated Weed Management EIS Chapter 2 Public Participation, Issues and Alternatives concerns and incorporated measures to make sure the effects of the Proposed Action would not be significant.

Issues are categorized into three groups, namely: • Issues outside the scope of the Proposed Action • Issues considered but not given further analysis • Issues that need further analysis – these are the issues that used to develop alternatives

Alternatives are of two sorts, • Alternatives considered, but not studied in detail • Alternatives studied in detail, such as the Proposed Action Issues outside the Scope of the Proposed Action One commenter questioned the adequacy of the process used by the EPA to review and register herbicides for weed control. The ID team included a requirement in the Proposed Action that all herbicides would be handled following EPA label guidelines and other state and federal laws governing storage, mixing, application and disposal.

The EPA has regulatory responsibility regarding the evaluation, disclosure and registration of herbicides in the United States. This issue is beyond the legal authority of the Forest Service.

This concern is outside the scope of the Proposed Action, and was not considered further

Issues Considered, But Not Given Further Analysis Why are we proposing a 15,000-acre cap on the area we could treat each year? This seems too small to address the weed infestations found on the 2.1 million acre Lolo NF.

While we recognize a need to treat more acres each year, this figure was proposed based on anticipated weed management budgets and to provide an upper limit for effects analysis. Our 10-year budget average has been about $100,000/year for weed management and we do not anticipate receiving substantially more funding in the foreseeable future.

Issues Needing Further Analysis Public and agency comments raised issues that guided the development of alternatives. Each issue is stated as a question and is followed by a brief discussion of how it is addressed in the FEIS.

Why don’t we analyze a prevention alternative? This issue was used to formulate an alternative, but for the reasons discussed in the next section, this alternative was not given detailed analysis. See Alternatives Considered, but Not Studied in Detail. Prevention is however, included as a weed management tool in Alternative 2.

Final Integrated Weed Management EIS Page 17 Chapter 2 Public Participation, Issues and Alternatives

Why don’t we consider a low herbicide alternative? We developed a low herbicide alternative to respond to this issue, it was considered, but not studied in detail. See Alternatives Considered, but Not Studied in Detail.

What are the effects of weeds on wildlife and other natural resources? Many people described weeds as a very big problem because they have seen the impacts of non-native plants on vast areas. The effects of weeds on resources are discussed in Chapter 3.

How would actions taken to control weeds, especially herbicides, affect wildlife and other resources? Some people expressed concern about the effects of herbicides on wildlife and other natural resources. To respond to this issue, a No Herbicide Alternative is discussed, a Low Herbicide Alternative was considered, but not studied in detail, and the No Action Alternative, which holds herbicide use at current levels, is analyzed.

The effects of herbicides on natural resources, as well as models and past experience with herbicide drift from aerial application, are disclosed in Chapter 3. Effects on terrestrial animals, plants, soil and aquatic organisms are also addressed in the FS Ecological and Human Health Risk Assessments (http://www.fs.fed.us/foresthealth/pesticide/) which are incorporated into this analysis by reference.

What are the potential effects of herbicides on human health? • Some people expressed concern about the effects of herbicides on human health. To respond to this issue, a No Herbicide Alternative is discussed in Alternatives Considered, but Not Studied in Detail. • The potential human health effects of herbicides named in this proposal are disclosed in Chapter 3. Effects on human health are also addressed in the FS Ecological and Human Health Risk Assessments (http://www.fs.fed.us/foresthealth/pesticide/) which are incorporated into this analysis by reference. Alternatives Considered But Not Studied in Detail This section describes alternatives considered by the ID team, but not analyzed in detail for the reasons given. These alternatives receive no further consideration in this document.

Prevention Alternative One commenter suggested we consider a Prevention-Only Alternative. This Prevention Alternative did not pre-suppose any particular amount of a given method of weed management, featuring instead, intensive monitoring. This alternative would act on the logic that preventing weeds and eradicating new infestations would be effective.

In this alternative, the commenter said, “that human activity is the major vector of weeds into the Forest” and that by significantly controlling human activity, the FS would be able to stop weeds from spreading.

Page 18 Final Integrated Weed Management EIS Chapter 2 Public Participation, Issues and Alternatives

This alternative includes a monitoring program that would measure: • The interaction of invading and native species in various ecological niches • Soil parameters, such as hydrology and nutrients • The impact of human activity on weed introduction and spread • The effects of controlling human activity • The effects of weeds • The effects of herbicide drift • The “true impact of poisons to biological systems” This alternative was not given detailed analysis because: • Although preventing weeds and eradicating new infestations would control some weeds, this approach would not reduce or control existing infestations. As such, this alternative would not meet the Purpose and Need. • Changing land use/Forest Plan direction or travel management, road use and forest access are outside the scope of this analysis. For these reasons, this alternative was not given further consideration.

No Herbicides Alternative Some people expressed concerns about using herbicides, as a result the ID team considered an alternative that would respond to the Purpose and Need without using herbicides. A No Herbicides Alternative is different from the No Action Alternative because the No Action Alternative allows the use of herbicides authorized under previous NEPA decisions – see Table 1-2.

A No Herbicides Alternative was analyzed in detail in the Lolo NF 2001 Big Game Winter Range and Burned Area Weed Management EIS. That analysis, incorporated into this one by reference, found the No Herbicide Alternative to be cost prohibitive. That analysis indicated it would cost almost $73,000,000 to treat a maximum of 21,750 acres. It found effective control would require at least one treatment a year and, in most cases, two (e.g. pulling or mowing).

In this analysis we are considering a maximum of 15,000 treatment acres/year, smaller maximum annual treatment acreage than considered in the 2001 Lolo NF analysis. Using the 2001 aggregate cost, it would cost more than $50,000,000 to treat a maximum of 15,000 acres/year.

A No Herbicide Alternative simply would not allow enough annual treatment to minimally meet the Purpose and Need for this analysis. Furthermore, hand pulling and mowing have not proven effective at containing or reducing widespread infestations.

For these reasons, this alternative was dropped from further consideration.

Low Herbicide Alternative Under a Low Herbicide Alternative, herbicides would be used only to eradicate new species or new infestations. There would be no aerial application of herbicides. To minimize herbicide use, prior decisions authorizing herbicides would be withdrawn.

Final Integrated Weed Management EIS Page 19 Chapter 2 Public Participation, Issues and Alternatives

New infestations would be defined as infestations smaller than one acre in perimeter. Mechanical weed control (pulling) would be used on about half the projects and herbicides on the other half.

This alternative assumes a maximum of 250 acres of small, new infestations would be treated each year (50 acres/Ranger District). Weed infestations larger than one acre would not be treated with herbicides or by pulling or mowing. Biological controls would be released at current levels. Treated sites would be seeded only if there was not enough native or desirable vegetation. Weed education and prevention would continue at current levels.

This alternative would include an adaptive approach that would allow managers to learn from the past and improve effectiveness. It would provide quick, effective treatment of small, new infestations before they grow while still addressing other resource concerns. Since not every acre has been inventoried, many small existing sites have yet to be identified.

The herbicide treatments assume an average cost halfway between backpack pedestrian ($125/acre) and truck application ($30/acre), for a cost of $77.50/acre. For 125 acres, this would amount to $9,687.50.

Pulling would cost an average of $8,370/acre for two pulls – to be effective pulled sites have to be pulled at least twice a year. Pulling costs are estimated at the rates for paid employees.1 For 125 acres, this would amount to $1,046,250.

This alternative would not meet the Purpose and Need, would not address larger infestations and would exceed present and anticipated weed funding. For these reasons, it was dropped from further consideration.

Concentrated Eradication Cells Alternative One Interdisciplinary Team Member suggested we focus on eradicating weeds in limited locations, or cells. Under a Concentrated Eradication Cells Alternative, weed management would be concentrated on a limited number of small infestations. Control efforts would not begin in new places until the weeds were eradicated in the first cells.

While one of the specific purposes of this proposal is to eradicate small new infestations of new invaders, it is not the purpose of this proposal to eradicate widespread weeds. The Purpose and Need says widespread weeds are to be contained and reduced within infested areas.

Concentrating all our efforts on limited areas would reduce our ability to rapidly respond to new infestations and potential invaders. New infestations and potential invaders could become established before we eradicated weeds in previously identified cells.

1 Experience has shown that volunteers often lose interest and won’t pull weeds for more than a few hours. Costs are from Appendix J of the 2001 FEIS for Big Game Winter Range and Burned Area Weed Management.

Page 20 Final Integrated Weed Management EIS Chapter 2 Public Participation, Issues and Alternatives

This alternative does not meet the Purpose and Need, and would conflict with the 1991 Amendment 11 to the Lolo NF Plan. For these reasons, it was dropped from further consideration.

No Aerial Application Alternative A No Aerial Application Alternative was analyzed in detail in the 2001 Big Game Winter Range and Burned Area Weed Management EIS.

Under this alternative we would not be able to economically or safely treat large, remote winter ranges or other infestations where there are few or no roads, and where pedestrian treatment is inefficient, dangerous or impractical.

Because of these limitations and the costs documented in the 2001 EIS, this alternative would not meet the Purpose and Need, nor would it effectively implement the 2004 National Strategy. For these reasons, it was dropped from further consideration.

Alternatives Considered In Detail The alternatives below include weed education, prevention, mechanical, cultural, biological and herbicide weed management practices. These weed management practices are described in Chapter 2 of the 2001 Big Game Winter Range and Burned Area Weed Management EIS, which is incorporated into this analysis by reference.

Alternative 1, No Action The No Action Alternative is required by NEPA. It provides a point of reference for evaluating the effects of the other alternatives.

Under the No Action Alternative, no new actions would be taken to respond to the Purpose and Need described in Chapter 1. Weed control would continue to take place at current levels, namely, actions authorized in site-specific decisions since the 1991 Lolo NF Plan Weed Amendment 11. This amounts to about 5,000 to 6,000 acres annually, depending on funding. See Table 1-2, Acres of Herbicide Treatment Allowed Under Existing NEPA in Chapter 1.

Alternative 2, Proposed Action This alternative is the Proposed Action referred to in Chapter 1. The Lolo NF proposes to implement an adaptive integrated weed management strategy to control weeds as budgets allow. Adaptive Management, as used in this analysis, is defined as a dynamic approach to resource management in which the effects of treatments and decisions are monitored and used with research results, to modify management on a continuing basis to ensure objectives are met. It would also allow managers to control new and small weed infestations before they can expand. This strategy combines a prevention/education component; a

Final Integrated Weed Management EIS Page 21 Chapter 2 Public Participation, Issues and Alternatives combination of ground and aerially applied herbicides; biological agents, revegetation, mechanical and cultural treatments; and sheep and goat grazing on a maximum of 15,000 acres/year.

The desired outcomes of this strategy are to: • Establish beneficial vegetation and weed resistant plant communities • To improve public awareness of weeds • Expand the use of both public and agency weed prevention practices • Protect weed free areas • Reduce or deplete the number of viable weed seeds in the soil Sites treated with herbicides, biological agents, seedings, mechanical controls or sheep or goat grazing would receive follow-up treatments until the viable weed seeds in the soil are depleted or reduced to an acceptable level. When treatments or programs are repeated, they could occur either annually or every two or three years until the desired outcomes are achieved. Herbicide treatments would typically be repeated once every one to three years, depending on the control method, size of infestation, persistence (if a herbicide is used) and weed species, until the desired outcome is achieved.

Education and awareness programs would be repeated, once or more a year, in hopes of incorporating them into the way people act and think without being prompted.

The weeds listed in Table 1-1 in Chapter 1 would be treated considering a number of factors, such as funding, invasiveness of the species, potential for the seeds to move off-site, and the ecological importance or rarity of a certain site. The 78,443 acres of known weed- infested sites are shown on the maps in Appendix A.

The treatment priorities for the Proposed Action include the priorities from the 1991 Weed Amendment to the 1986 Lolo Forest Plan, and other sites at risk for weed invasion or spread, as described in Chapter 1 under the Need Section. The priority site types are: • Big-game winter ranges which include generally south-facing, dry, relatively open lands under 5,000 feet elevation that support ponderosa pine or Douglas fir habitat types. • Weed-free areas, and the roads, trails and trailheads that lead into them. • Areas of concentrated public use (since people spread weeds) which includes both developed and dispersed undeveloped campsites; trailheads where system trails join forest roads and serve as a transition point from automobile travel to pedestrian, bicycle, horseback, ski or motorized recreational vehicle travel. • Administrative sites, which include Forest Service pastures, Ranger Stations, work centers, lookouts, backcountry cabins and owned or leased airstrips and helibases and helispots. • National Forest lands bordering private lands with active weed-control programs. These areas include FS lands where the neighboring landowner has an active weed control program and lack of weed management on Lolo NF lands is affecting their program; relatively weed free lands where adjacent weed infested Lolo NF lands threaten the adjacent land and/or where adjacent landowners might request treatment on NF lands to protect weed management investment on private land. • Disturbed areas, which include but are not limited to log landings, fire crew or other crew staging areas and field project camps.

Page 22 Final Integrated Weed Management EIS Chapter 2 Public Participation, Issues and Alternatives

• Other disturbed areas include: Fire camps which include permanent or occasionally used locations we have used or expect to use for fire camps or helispots in the event of a wildfire; Burned areas which include areas of the forest that have burned in either a wild or prescribed fire and firelines and contingency firelines; Electronic sites which include designated and undesignated high elevation commercial or non commercial sites used for microwave, radio, television, cellular or navigational facilities; Gravel pits, mining and rock operations which include active and inactive mines; sapphire or crystal digging pits; borrow pits; prospecting holes and common rock gathering sites; Linear features which include power and utility corridors and rights of ways; ski areas which include Marshall Mountain and Montana Snowbowl ski areas; range sites which include permitted grazing allotments and other areas impacted by livestock; and roads and trails which include open and closed system and non system roads and trails, cross country ski trails, snowmobile trails, historic roads and trails, National Historic Trails and un-maintained roads. • Less or undisturbed areas such as riparian areas, roadless or wilderness areas where weeds are found. Riparian areas are lands adjacent to streams, lakes and wetlands identified as generally flat and where the vegetation present requires large amounts of water normally available from a high water table. Riparian areas are transition zones between aquatic and upland terrestrial environments. Weeds are found in roadless and wilderness areas, and could be considered for treatment as long as such treatments are in compliance with the acts designating each wilderness area. There would be no aerial herbicide treatments or sheep or goat grazing for weed control in designated wilderness areas. Alternative 2 would use the same variety of treatment methods as at present except that sheep and goat grazing would be added. Weeds within a gross total area of 78,443 acres have been identified as needing treatment. This is the total of all the exterior boundaries of the project areas. Actual, on the ground treatment would be less because only weed infested areas within the gross project areas would be treated and there is not 100 percent weed coverage in all the project areas. Roads, trails and linear features, for example would usually only be spot treated where weeds occur. Acres proposed for treatment on each District are shown below in Table 2-1.

Table 2-1 Acres proposed for treatment by Ranger District Ranger District Acres Missoula 7,082 Ninemile 31,351 Plains/Thompson Falls 35,949 Seeley Lake 2,292 Superior 1,769 TOTAL 78,443 It is inevitable that new infestations would be found. They would be screened to determine priorities and what kind of treatment to use. Screening would be documented as shown in Appendix B.

The Decision Tree, Figure 2-1, would be used to determine the most appropriate method(s) to use to minimize the risk of adverse environmental effects. Mitigation measures would be used to minimize any adverse effects on human health.

Final Integrated Weed Management EIS Page 23 Chapter 2 Public Participation, Issues and Alternatives

Figure 2-1 Weed treatment decision tree

Riparian No Yes zone?

Yes High Use appropriate Leaching No broadcast Leaching treatments – potential? herbicide Manual, biological, or potential? treatment cultural methods or herbicides labeled for use near water No Low Use appropriate treatments – Manual, biological, Unique Unique or cultural methods elements Use appropriate No elements No or herbicides treatments – present? present? labeled for use Manual, biological, near water cultural or herbicides

Yes Yes

Mitigation Mitigation Use appropriate possible? possible? treatments – Manual, biological, or Yes Yes cultural methods or Use appropriate herbicides labeled for treatments – use near water Manual, biological, cultural or herbicides No No

Avoid herbicides – Avoid herbicides – Use manual, biological or Use manual, biological or cultural methods cultural methods

The following definitions apply:

Riparian areas are lands adjacent to streams, lakes and wetlands identified as generally flat and where the vegetation present requires large amounts of water normally available from a high water table. Riparian areas are transition zones between aquatic and upland terrestrial environments.

Low potential would be as described in the ‘environmental hazards’ section of the herbicide product label. Depending on the herbicide leaching potential, areas that are not riparian may need herbicide mitigation. To identify leaching potential, consider soil type, the presence of wet-site plants, slope position and the potential for herbicides to run off down slope.

Unique elements are sensitive resources other than water.

If herbicides were determined to be an appropriate control method for a particular infestation, they would be limited to those listed in Table 2-2. Other herbicides could only

Page 24 Final Integrated Weed Management EIS Chapter 2 Public Participation, Issues and Alternatives be used if registered for weed control by the Environmental Protection Agency (EPA) and if they had a current FS Ecological and Human Health Risk Assessment (http://www.fs.fed.us/foresthealth/pesticide/) or a Line Officer review and approval. Dye markers would be used in some cases to see where herbicide was applied.

Herbicides would be used at or below the rates specified on the label or in the Lolo Forest Plan, whichever is more restrictive. Selective broadleaved herbicides would be used in most cases. Only one non-selective herbicide (glyphosate) would be used. The non-selective herbicide would be used only in specific situations, usually as wick or individual plant treatments. The weeds controlled by each herbicide are listed on the product labels and in Appendix I Herbicide and Target Weed Species.

Table 2-2 Herbicides included in Alternatives 1 and 2

Chemical Selective? Common Product Names Aminopyralid Yes Milestone and ForeFront R & P 2,4-D Yes Formula 40, Hi-Dep, Weedar 64 Chlorsulfuron Yes Telar Clopyralid Yes Transline Dicamba Yes Banvel, Veteran 10-G Glyphosate No Roundup, Rodeo, Accord Imazapic Yes Plateau Imazapyr* Yes Arsenal, Chopper, Contain Metsulfuron methyl Yes Escort, Ally Picloram Yes Tordon Triclopyr* Yes Garlon In Place, LI-700, MSO, Preference, Adjuvants N/A R-11, Stay Put, Syltac Dyes N/A SPI Max, Hi-Light, Dye marker *Herbicides included in Alternative 2 but not in Alternative 1. Project labels included in the project file

Aerial application would be considered and used according to the mitigation measures described later in this chapter if: • Sensitive resources (such as open water, or neighboring landowners adverse to herbicides) will be effectively protected and buffered • The overstory is open enough that the herbicides would reach the weeds (generally less than 25 percent canopy cover) • The infestation covers a large enough area to make aerial treatment efficient This alternative includes an adaptive approach that would allow managers to learn from the past and improve effectiveness. The steps and necessary documentation required for a new infestation before treatment are identified in Appendix B. This alternative would provide quick, effective treatment of small, new infestations before they grow to

Final Integrated Weed Management EIS Page 25 Chapter 2 Public Participation, Issues and Alternatives unmanageable populations. Since every acre on the Lolo NF has not been inventoried, many small existing sites have yet to be identified.

A maximum of 15,000 acres, less than on1 percent of the 2.1million acre Lolo NF could be treated annually.

Comparison of Alternatives

Cost Alternative 1, No Action, is the current Lolo weed program and the cost is what is presently spent on weed control.

Alternative 2, the Proposed Action, would allow treating weeds on a maximum of almost three times more acres than the No Action Alternative.

The cost comparison uses the present combination of treatment methods and site prioritization. Costs are estimated at the maximum number of acres proposed. The actual acres treated will be dependent on funding.

Present funding for all aspects of weed management on the Lolo NF, including planning, education, coordination, treatment and reporting, is about $150,000/year. The average over the last 10 years is about $100,000/year.

Other differences between the Alternatives The primary differences between the No Action Alternative and the Proposed Action are as follows: • The Proposed Action uses standardized, comprehensive methodology to consider critical resource information; the No Action Alternative does not. (refer to Figure 2- 1) • The Proposed Action would allow quick response to new weed species and infestations and control while infestations are small. The No Action Alternative necessitates a delay of one to three years for new environmental analysis and a decision before new species or infestations could be treated. • The Proposed Action could treat up to 15,000 acres/year; the No Action Alternative would treat only about 5,000 to 6,000 acres/year. These figures include all treatment methods, not just herbicides. • The Proposed Action would allow carefully managed sheep or goat grazing to control weeds. • Under the Proposed Action, if funding did not increase, acres treated/year could remain the same as under Alternative 1, but weed managers would have a larger “menu” of approved projects to chose from and could thereby prioritize better to get the maximum benefit for the weed management dollar spent. At present weed managers are limited to projects with existing decisions. There are 78,443 acres of high priority known weed infestations identified in this analysis that can not be treated because they do not have past decisions on them. • The Proposed Action would add two herbicides (imazapyr and triclopyr) to those available under the No Action Alternative.

Page 26 Final Integrated Weed Management EIS Chapter 2 Public Participation, Issues and Alternatives

Table 2-3 Comparing Alternatives Alt 1 Alt 2 Provision No Action Proposed Action Average Annual treatments 5,700 acres 15,000 acres Maximum Annual cost $150,000 $450,000 Standardized methodology to consider critical resource No Yes information Rapid response to new No Yes infestations Aerial applications Yes Yes Herbicides available for use 9 11

Table 2-4 Comparing How Alternatives Address Purpose and Issues in This Analysis

Alternative 1, No Action Alternative 2, Proposed Action Purpose: Prevent potential invaders from establishing. Includes full range of prevention measures and BMPs, but Same as Alternative 1 but allows for limits ability to treat new invaders. treatment of new invaders as funded. Purpose: Eradicate small new infestations and reduce larger infestations of new invaders. Minimally meets purpose, because it does not allow rapid Best meets purpose, because it allows rapid response to small new infestations without additional treatment of both small and large new NEPA analysis. infestations. Purpose: Contain and reduce widespread weeds within infested areas. Moderately meets purpose on sites with existing NEPA Best meets purpose, because it allows rapid decisions, but does not allow containment or reduction of containment and reduction of more widespread widespread weeds unless they are covered by an existing weeds that are not covered by an existing NEPA NEPA decision. decision. Purpose: Allow rapid, timely response to new, small or recently discovered infestations before they become well established. Minimally meets purpose, because it does not allow rapid Best meets purpose, by allowing rapid, timely response without additional NEPA analysis, which usually response to widest range of situations. takes from 1 to 3 years. Purpose: Prevent or limit the spread of established weeds in weed free areas. Meets purpose only in places where there is an existing Best meets purpose, by allowing rapid response NEPA decision. to the widest range of situations. Purpose: Implement the 2004 Forest Service National Strategy and Implementation Plan for Invasive Species Management. Implements the strategy for sites with existing NEPA Implements the strategy by providing flexible decisions. Does not implement the strategy on new and rapid response while ensuring other infestations or after disturbances such as wildfire. resource protection. Issue: What are the effects of weeds on wildlife and natural resources? Impacts of weeds include: • Changes in plant community composition and structure. • Loss of sensitive plant populations. • Effects on water and soil quality. • Effects to people allergic to weeds Provides lower level of protection for native plants from Provides higher level of protection for native weeds. No reduction of weed-caused impacts to soil and plants from weeds. Reduces weed-caused water on sites without existing NEPA decisions. impacts to soil and water and people. Issue: How would actions taken to control weeds, especially herbicides, affect wildlife and natural resources? Impacts of herbicides include effects to: Fish and terrestrial animals, Non-target plants, and Soil and water quality. Annual maximum treatment would occur on On average, treatment would occur on less than ¼ of 1% less than ¾ of 1% of the acres on the Lolo NF. of the acres on the Lolo NF. Effects would be extremely Low effect due to limited acres treated. low due to limited acres treated and mitigation measures Mitigation measures would further minimize applied. effects.

Final Integrated Weed Management EIS Page 27 Chapter 2 Public Participation, Issues and Alternatives

Alternative 1, No Action Alternative 2, Proposed Action Issue: What are the potential effects of herbicides on human health (chemically sensitive individuals, people allergic to herbicides and general public)? Annual average treatment would occur on less On average, treatment would occur annually on less than than ¾ of 1% of the acres on the Lolo NF. Effects ¼ of 1% of the acres on the Lolo NF. Effects would be would be below established toxicological below established toxicological thresholds due to limited thresholds due to limited acres treated and acres treated and mitigation measures applied. Mitigation mitigation measures. Mitigation would help would help sensitive people avoid treatment sites. sensitive people avoid treatment sites. Mitigation Measures

Mitigation for the No Action Alternative Mitigation measures for the No Action Alternative are described in the 1991 Amendment 11 to the Lolo NF Plan.

Mitigation for the Proposed Action Alternative

General Mitigation Measures 1) We will coordinate with and/or notify neighboring landowners who may want to, or are already treating adjacent lands. 2) Information learned from each project will be used to improve safety, resource protection, reduce human health risks and improve efficacy on subsequent projects. 3) The effectiveness of weed treatment and the impact on non-target plants will be monitored at representative project areas. Information and methods to mitigate impacts to non-target plants that are learned from previous treatments will be applied to future treatments. 4) New invaders, as identified by local and state agencies, will be given high priority for treatment. 5) All weed treatment information will be provided and coordinated with a Forest Service botanist prior to field season or as soon as possible in the event of discovery of a new invader species.. All treatment sites will be evaluated for sensitive plant habitat suitability and suitable habitats will be surveyed as necessary. Site-specific treatment guidelines will be developed and followed for infestations within or adjacent to known TES plant populations to protect those populations. Site specific herbicide mitigation to protect TES plant populations may include herbicide selection, application timing, application method, herbicide rate, buffers or avoidance. 6) Prehistoric trails, remnants of historic structures and other heritage resources will be protected if deemed necessary by a Forest Service Archaeologist (with Tribal coordination when needed). 7) All weed treatment activities will comply with state and local laws and agency guidelines. 8) Region 1 weed prevention practices will be applied to forest management activities. New weed prevention Best Management Practices will be implemented as they are developed.

Page 28 Final Integrated Weed Management EIS Chapter 2 Public Participation, Issues and Alternatives

9) Revegetation on disturbed or treated sites should include native plant species as recommended by the USFS-R1 native species policy (USDA-FS 1994). This policy emphasizes the use of locally adapted native plant seed, whenever possible. Seeding should be used as a reclamation tool only where resource damage will occur without it. Otherwise, sites should be allowed to re-vegetate naturally from the localized adjacent seed source. 10) Treatments in all Research Natural Areas (RNAs) will be designed to conserve native species diversity while meeting the goals and objectives of the RNA and will be coordinated with the Rocky Mountain Research Station. 11) Biological agents will not be released until they are screened for host specificity and approved by the USDA Animal Plant Health Inspection Service.

Sheep and goat grazing mitigation measures - include the above measures, plus 12) Sheep or goat grazing will not be used for weed control in recovery zones or in occupied grizzly bear habitat. 13) When using livestock for weed control, in areas outside of the recovery zones and occupied grizzly bear habitat, a herder and guard animal(s) will be required. Livestock must be contained at night. 14) Sheep and goat herders will notify the local District Ranger of any livestock losses within 24 hours of discovery. 15) Wildlife Services (USFWS) will be the first contact in the event of a sheep or goat death caused by a known or suspected wild predator. Sheep or goat carcasses of animals that die on the Lolo NF must be removed from the Forest within 24 hours of the site visit by Wildlife Services to avoid habituation by predators. Upon a determination by Wildlife Services of a confirmed or probable wild predator predation, all sheep or goats in that herd will be removed from the Lolo NF lands within 24 hours. 16) To prevent disease transmission between domestic sheep or goats and wild bighorn sheep, weed grazing projects will be coordinated with the Montana Fish, Wildlife and Parks biologists. 17) In areas of known wolf den and rendezvous sites on the Lolo NF, sheep and goat grazing will be coordinated with wildlife biologists.

Herbicide mitigation measures - include the above measures, plus 18) Signs will be posted at access points to heavily used recreation areas and heavily used areas of the forest where herbicides are applied to let the general public, chemically sensitive people and others who may wish to avoid herbicides know the estimated date(s) of treatment, the herbicide to be used and a name and phone number of who to call for more information. 19) Herbicide applications will be coordinated with permittees on permitted project areas. 20) All application of herbicides will be performed by, or supervised by, a state licensed applicator following all current legal application procedures administered by the Montana Department of Agriculture.

Final Integrated Weed Management EIS Page 29 Chapter 2 Public Participation, Issues and Alternatives

21) All herbicides will be handled following the EPA label guidelines and other state and federal laws for storage, application, and disposal methods. 22) Mixing will take place at least 150’ from open water unless spill containment devices are readily available and an anti back siphoning device is used when drafting water. 23) Applicators will review stream and wetland areas to assure that herbicides will not be applied to open water. 24) Herbicides will be used to waters edge only when absolutely needed and provided the product label allows such use. 25) Herbicide applications near live water or in areas with shallow water tables will follow label directions. 26) Herbicide applicators will not initiate spraying when heavy rains are forecast that could cause offsite herbicide transport into sensitive resources such as streams. 27) Herbicides applicators will be familiar with and carry a Herbicide Emergency Spill Plan (Appendix C) to reduce the risk and potential severity of an accidental spill. Herbicides applicators will also carry spill containment equipment. 28) All herbicide applications in riparian areas will be documented on the Mitigation Measure Certification form included as Appendix E to document protection of surface waters.. 29) Each Ranger District will submit a proposed list of herbicide project information each year to the Forest Service project wildlife and fisheries biologists and botanist at least 45 days prior to herbicide application or as soon as possible in the event of discovery of a new invader species. These specialists will review this information in relation to the current or needed Biological Assessments and may modify or prepare additional Biological Assessments as needed. Required project information will include a map, acres proposed by 6th code HUC, proposed herbicide, application rate, application method and any special resource or other circumstances. 30) Herbicides will not be applied if snow or ice covers the target vegetation. 31) Spray equipment will be calibrated each season and recalibrated as needed during the spray season. 32) Low boom pressure (less than 40 pounds per square inch) will be used to reduce drift. 33) Drift reduction products will be used as needed near sensitive resources. 34) Ground-based herbicide application will occur only when wind speed is 10 mph or less. 35) If commercial applicators are used for the application of restricted use pesticides, Forest Service contract administrators will check to make sure their Montana commercial restricted use pesticide license is current.

Aerial application mitigation measures - include the above measures, plus 36) Temporary road and area closures will be used as needed to ensure public safety during aerial spray operations.

Page 30 Final Integrated Weed Management EIS Chapter 2 Public Participation, Issues and Alternatives

37) A “no-fly” zone will be designated to avoid disturbance to active bald eagle and peregrine falcon eyries. 38) Aerial spray project areas will be marked or delineated so treatment boundaries are clear to the application pilot. 39) Buffer zones and treatment areas will be reviewed with the application pilot prior to application. 40) To mitigate drift, there will be a 300-foot buffer next to sensitive resources unless drift card monitoring or other information indicates a smaller buffer is effective. Even if drift card monitoring indicates smaller buffer will be effective, the buffer will never be less than 150 feet wide. 41) Application will occur only when wind speed is 6 mph or less. 42) Aircraft smokers, smoke bombs or on-site wind monitoring will be used as needed to determine wind direction near sensitive resources. 43) Spray detection cards will be used as needed inside representative or critical buffer zones near sensitive resources. 44) Low level aerial herbicide applications will be restricted to two days per Grizzly Bear Management Unit/year (a likely home range for a female grizzly bear). 45) Do not conduct aerial herbicide applications during temperature inversions. 46) Maintain close communications between the helicopter pilot and the ground field observers to monitor drift, deposition and wind speeds during aerial applications of herbicides. 47) GPS systems should be used in spray helicopters in association with flagging or field marking of treatment areas to ensure accuracy of aerial treatments (i.e. to better assure that only areas marked for treatment are treated). 48) Drift reduction agents and nozzles that create large droplets to reduce drift to non target areas should be used during aerial herbicide applications Environmental Justice Executive Order 12898, issued in 1994 ordered Federal Agencies to identify and address any adverse human health and environmental effects of agency programs that disproportionately impact minority and low-income populations. At this time, no minority or low-income communities have been identified in southwest Montana.

Native American Treaty Rights While the alternatives may have differing impacts on wildlife and fish, as described in Environmental Consequences, none of the alternatives will alter opportunities for subsistence hunting and fishing by Native American tribes. Tribes holding treaty rights for hunting and fishing on the Lolo NF were contacted during scoping and have an opportunity to comment on this EIS.

Final Integrated Weed Management EIS Page 31 Chapter 2 Public Participation, Issues and Alternatives

Reasonable Foreseeable Actions The BLM is authorized to apply herbicides on adjacent Federal lands. Aerial and ground applications are currently being used by the BLM. This practice is expected to continue into the foreseeable future. The Missoula Field Office of the BLM annually treats an average of about 800 acres per year with herbicides.

Counties adjacent to and within the boundary of the Lolo NF have weed control programs that include aerial and ground application of herbicides. The State of Montana also applies herbicides on State lands near or adjacent to the Lolo NF. These programs treat adjacent areas and roads, State roads and highways within and around the Lolo NF. County and State weed control is likely to continue.

Adjacent private landowners actively control weeds and some use herbicides. Methods often include both aerial and ground application of herbicides, especially on large ownerships. Weed control on these lands is likely to continue.

Forest projects, such as timber sales, post and pole sales, road maintenance, mining permits, etc., will also affect weed populations. Through recently adopted Best Management Practices (appendix G) weed prevention is incorporated into mitigation measures in those projects.

Prevention and education programs will continue. The Weed-Seed-Free feed requirement on National Forest lands will remain in place and participation in State, private and county weed programs will continue.

Monitoring A monitoring program will be incorporated as part of the adaptive management approach. Monitoring will be done after education programs, every five years for weed prevention BMPs and before, during and after direct weed control treatments.

Monitoring Education Programs After formal weed education programs, participants will be asked for feedback, whether the program felt effective, what they learned and how the program could be improved. Program locations and participants will be reported annually.

Monitoring Weed Prevention Best Management Practices Every five years, District weed coordinators and Supervisors Office specialists will participate in a weed prevention Best Management Practices review. Each practice will be reviewed and reported on as to whether and how it is being used, if it is effective and how we can improve weed prevention practices. Weed prevention practices will be added or improved based on this review. The review will be documented in a Best Management Practices report. The last such review on the Lolo NF was conducted in 2005.

Page 32 Final Integrated Weed Management EIS Chapter 2 Public Participation, Issues and Alternatives

Monitoring before Treatment Surveys for threatened, endangered or sensitive (TES) plants or animals will be conducted prior to grazing and herbicide weed control projects. These surveys will not be conducted for biological agent, pulling, mowing and seeding weed control projects. Site specific recommendations will be developed, and monitored as needed to protect TES populations.

Relative abundance of native, beneficial, and desirable vegetation present on representative treatment sites will be evaluated to determine if revegetation will be needed after weed treatment.

Distance to ground and surface water will be evaluated prior to treatment. Mitigation measures for herbicide applications in riparian zones will be monitored (Appendix E).

Water quality will be monitored before treatments as deemed necessary by the unit hydrologist based on site conditions and experience with past projects.

Herbicide treatments in riparian areas will be documented on a Riparian Mitigation Measure Verification Report (Appendix E) and signed off on by the applicator.

Monitoring during Treatment During herbicide applications, critical parameters such as wind speed and direction, temperature, and humidity will be monitored to assure that conditions are within project standards (see mitigation measures for the herbicide use). Spray drift cards will be placed as needed between sensitive resources (such as but not limited to water) and treatment areas to monitor drift during aerial spray applications.

Monitoring after Treatment Non-target vegetation and effectiveness of treatments on weeds will be monitored on a representative sample of the treatment areas to determine changes in plant species and life form as a result of weed treatments. This will include monitoring herbicide effects on TES plant populations if they are treated. TES plant monitoring should include herbicide used, rate, application type, date of application and effects on TES plants.

Typical monitoring parameters include absolute canopy cover, relative canopy cover, biomass production (lb/ac dry wt) by life form, frequency of occurrence (%) of weeds, and photo points.

Biological agents will be monitored with sweep transects, stem count transects and photo points to evaluate both insect and weed population changes.

Water quality will be monitored after treatments as deemed necessary by the unit hydrologist based on site conditions and experience with past projects.

All water samples will have laboratory analysis conducted by an independent lab to determine herbicide content in relation to the limits defined by the Montana Department of Environmental Quality. Water samples will be collected and analyzed before and after aerial applications if a project is planned in any of the following circumstances:

Final Integrated Weed Management EIS Page 33 Chapter 2 Public Participation, Issues and Alternatives

• unsettled weather patterns • moderate to high likelihood of precipitation is forecast in next 24 hours • ambient air conditions that would alter modeled drift patterns • any aerial application where the buffer has been reduced to 150’ (see mitigation measure #40 • aerial applications in State designated municipal watersheds.

Page 34 Final Integrated Weed Management EIS

Chapter 3 Affected Environment and Environmental Consequences

Introduction This chapter describes the current condition of the physical, biological and human resources that could be affected by the Proposed Action and then discloses the direct, indirect, and cumulative effects of the alternatives described in Chapter 2. Direct effects are caused by the action and occur at the same time or place. Indirect effects are caused by the action and occur later in time or farther removed in distance, but still reasonably foreseeable. Cumulative effects are the impacts that result from the action when added to other past, present, and reasonably foreseeable future actions regardless of which agency or person undertakes such actions. The level of detail in each description is commensurate with the importance of potential impacts and the amount of information necessary to understand the effects of the actions.

For ease of reference, the alternatives given detailed analysis from Chapter 2 are summarized below:

Chapter 3 changes made in response to comments on the DEIS DEIS Page 112 – added:

(Montana Water Quality Standards prohibit toxic or harmful concentrations to aquatic life in surface waters,

DEIS Page 115 – added:

Since 1992, the Lolo NF has conducted over 30 aerial herbicide application projects with an excellent safety record, no reported public or work injuries and no accidental resource damage.

DEIS Page 64 - Sensitive Plant Species List and Table 3-16 were updated to include: Arabis fecunda (a forb), Botrychium paradoxum (a fern ally), and Phlox kelseyi var. missoulensis (a forb).

DEIS Page 169 – Change:

The northern goshawk was taken off the Regional Sensitive Species List July, 17, 2007.

Alternatives Alternative 1: No action-Under this alternative, no new actions would be taken to respond to the Purpose and Need described in Chapter 1. Weed control would continue to take place

Chapter 3 Affected Environment and Environmental Consequences at current levels, namely, actions authorized in site-specific decisions since the 1991 Lolo NF Plan Weed Amendment 11. This amounts to about 5,000 to 6,000 acres annually, depending on funding. This alternative would use a combination of prevention, education, herbicides (ground and aerially applied), biological agents, mechanical and cultural treatment methods. There would be no sheep or goat grazing to control weeds. There would be no forest-wide plan to take action on new weed species or new infestations. Actions to control or eradicate new infestations not covered in existing decisions would require further NEPA analysis and documentation before they could be treated. These analyses would take from one to three years to complete.

Alternative 2: Proposed Action – Under this alternative, we would implement an adaptive weed control program and strategy that would include the same weed control methods as in Alternative 1 with the addition of sheep and goat grazing. Alternative 2 is essentially an expansion of the same weed management practices used in Alternative 1. There would be no aerial treatments or sheep or goat grazing for weed control in wilderness areas.

Greater emphasis would be placed on treating new infestations and new weed species when they are found. More known weed infestations could be treated than under Alternative 1. A maximum of 15,000 acres could be treated per year, depending on funding. If funding remained constant however, acres treated would remain the same as Alternative 1, but we would be able to prioritize better to determine which weed infestations to treat.

The main differences between Alternative 1 and Alternative 2 are: • Sheep and goat grazing could be added to our weed management “tool box” • We would be able to better prioritize which weed infestations to treat each year, our intention being to get the greatest benefit for our weed management dollar • New and recently discovered weed infestations could be treated • Maximum acres would increase to 15,000 acres. Otherwise we would still only treat approximately 5,000 to 6,000 acres/year • Alternative 2 would put an increased emphasis on establishing native and desirable vegetation rather than just killing weeds Acres to be Treated, Site Types and Disturbance Levels Our analysis assumed that the maximum acres allowed under each alternative would be treated each year. However, this would actually be influenced by budget. While there are 78,443 acres of projects identified in Chapter 2 and shown on the maps in Appendix A, only infested areas within each project area would be treated. This means that there would typically be untreated (weed free) zones within each project area and that less than the gross acres in each project area would typically be treated. The environmental consequences for each resource considered both the site types described in Chapter 2 and assumed the typical disturbance levels displayed below in Table 3-1.

Table 3-1 Disturbance Levels by Site Type* Permanently Temporarily Disturbed Sites, Sites With No Surface Disturbed Sites Where Objective Is To Disturbance Restore Administrative Sites Burned Areas Adjacent to Private Land Campgrounds, Other Disturbed Areas (such as Big Game Winter Ranges

Page 36 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Permanently Temporarily Disturbed Sites, Sites With No Surface Disturbed Sites Where Objective Is To Disturbance Restore Recreation Areas, landings, staging areas, project Trailheads camps) Electronic Sites Riparian Areas Gravel Pits, Mines, Rock Wilderness, backcountry,

Operations roadless Linear Features Range Sites Roads and Trails Ski Areas *Refer to Chapter 2 for a description of each site type. Weeds within a gross total area of 78,443 acres have been identified as needing treatment. Gross project areas range from less than 1 acre to 16,380 acres. This is the total of all the exterior boundaries of the project areas. Actual on the ground treatment would be less because only weed infested areas within the gross project areas would be treated and there is not 100 percent weed coverage in all the project areas. Roads, trails and linear features, for example, would usually only be spot treated where weeds occur. A mosaic of treatments within larger project areas would be typical.

Most of the actual contiguous treatment areas (as opposed to gross project areas) would be less than 100 acres. Spots and areas within watersheds would be treated where weeds are found, but entire watersheds would not be treated.

Cumulative Effects The cumulative effects analysis for each resource considers the on-going or proposed projects/activities listed in the Lolo National Forest Schedule of Proposed Actions in addition to ongoing and past projects on the Lolo NF.

Physical Setting, Topography, Precipitation and Climate The analysis area occurs across the Lolo NF. The Lolo NF stretches for 120 miles in a northwest to southeast direction and lies in west-central Montana. The forest extends from the Idaho border east. The forest is 40 to 80 miles wide in a southwest and northeast direction. The Clark Fork of the Columbia River is the major drainage. Interstate 90 bisects the forest.

The exterior boundary of the Lolo NF lies in nine different Montana counties (Flathead, Granite, Lake, Lewis and Clark, Mineral, Missoula, Powell, Ravalli, and Sanders). The Lolo NF also shares a common boundary with three Idaho counties (Clearwater, Idaho, and Shoshone). The number of Lolo NF acres in each county is shown below in Table 3-2.

Table 3-2 Acres by County County Acres Flathead 18,907 Granite 218,081

Final Integrated Weed Management EIS Page 37 Chapter 3 Affected Environment and Environmental Consequences

County Acres Lewis and Clark 74,332 Mineral 646,857 Missoula 519,620 Powell 144,788 Ravalli 8,131 Sanders 484,469 The Lolo NF lies in the Northern Rocky Mountains, which are characterized by north-to- south orientated mountain ranges separated by flat valley bottoms, foothill benches and terraces. Principal rocks in the general area are metamorphosed sediments of the Precambrian Belt Super group. Clayey, silty, and sandy soils make up the predominant lithology of the Belt, with some impure carbonate rocks as well. Quaternary-age unconsolidated alluvium is found in most valley bottoms. Pleistocene glaciation is evident throughout the area. Glacial drift deposits and lakebed sediments from Glacial Lake Missoula are common in valley bottoms and terraces.

Terrain is highly dissected, steep, and rugged. Most of the land is heavily forested with conifers, but many south-facing slopes are grassy, open, and park-like. Mountains are often higher than 7,000 feet and show the effects of alpine glaciation: bowl like cirque basins, U- shaped valleys, and narrow, serrated ridgelines or arêtes. While mid-slopes are generally steep (slope greater than 50 percent), gentle slopes are often found near larger streams and some ridge tops.

Atmospheric conditions are modified by aspect and slope, and become progressively cooler and moister as elevation increases. Climatic zones range from the semiarid and relatively warm valley bottoms through a broad range of cool, moist coniferous forests to the cold, moist sub-alpine and alpine mountaintops. The average annual precipitation on the forest is 42 inches, two thirds of which falls as snow. In valley bottoms along the Clark Fork, average annual precipitation is closer to 14 inches. Both Pacific Maritime and Continental weather systems affect the area.

Visual Character The landscapes of the Lolo NF are largely within the Columbia Rockies Character Type described in Visual Character Types of the Northern Region (U.S. Department of Agriculture-FS 1980). Roughly 30 percent of these lands outside of wilderness have distinctive scenic quality. Approximately 80 percent of the Lolo NF has a relatively natural appearance.

The Lolo National Forest Plan (U.S. Department of Agriculture-FS 1986) provides overall direction for visual quality on the forest. The forest plan specifies Visual Quality Objectives (VQOs) in the standards section of each forest plan Management Areas (MA) description. The project areas have been allocated Retention, Partial Retention, and Modification VQOs. Activities within these VQOs are described as follows: • Retention: Activities may only repeat form, line, color, and texture that are frequently found in the characteristic landscape, and should not be evident to casual forest visitors.

Page 38 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

• Partial Retention: Activities may repeat form, line, color, or texture, which are found infrequently or not at all in the characteristic landscape, but remain visually subordinate to the visual strength of the characteristic landscape. • Modification: Activities of vegetative and landform alteration must borrow from naturally established line, form, color, and texture so that their visual characteristics are those of natural occurrences within the surrounding area when viewed as middle ground or background. Activities may visually dominate the original characteristic landscape. All lands on the Lolo NF are visible to varying extents by the viewing public. Except for weed species that put on massive floral displays, most weeds are visually prevalent only when seen within foreground viewing distances of less than one-quarter mile. Primary travel corridors, including state and federal highways, major forest roads, and major forest trails have been assigned a VQO of foreground retention. Communities, campgrounds, lakes, and rivers often have a VQO of Retention. These foreground landscapes are often at higher risk to weeds because of ground disturbance and introduction of weeds by vehicles, people, and livestock along these corridors and activity centers.

Social Environment The social environment affected by the Lolo NF’s weed management program is described in Chapter III-7 through 10 of the LNF Plan FEIS.

A changing social and economic environment is partially responsible for the increase of weeds on private land and on the adjacent National Forests. The changes are assumed to be the result of a decrease in agricultural land and increased subdivision and residential and business development.

The following counties are included in the Lolo National Forest economic impact area: Granite, Mineral, Missoula, Powell, Ravalli and Sanders Counties, Montana.

Population The population grew from 160,284 during 1993 to 188,850 by 2003 in the seven county impact area. This is an increase of 28,566 people during that time period. The growth represents roughly eighteen percent growth in a decade, for an average annual increase of 1.8 percent. Every county experienced an increase in population. Missoula and Ravalli counties led growth with about 12,000 and 10,000 people added to each county.

Lifestyles Natural resource amenity values attract people to this region, even when employment opportunities are limited. Earnings are lower here, compared with other parts of the country, reflecting a “quality of life” premium that people are willing to pay to live in this region. Hunting, fishing, mushroom and berry gathering and other outdoor recreation are all amenities that are part of this quality of life premium.

Some people who use these outdoor amenities place high value on “naturalness.” Depending on a person’s beliefs, this leads some people to oppose weed invasion as a

Final Integrated Weed Management EIS Page 39 Chapter 3 Affected Environment and Environmental Consequences degradation of “natural” ecosystems. Others may view weed invasion as a “natural” process and weed management activities as “unnatural.”

This natural life-style takes on a spiritual aspect for some, including Native Americans. Many Native Americans practice religious and spiritual ceremonies on lands administered by the Lolo NF. These include such activities as vision quests, gathering traditional foods, and hunting.

Public Awareness of Weeds In 1994, 2002 and 2005, telephone surveys were conducted for the Statewide Noxious Weed Awareness and Education Campaign (Hoopes, 2005) to better understand Montanan’s awareness of the weed issue and whether educational intervention between the surveys had any effect. The survey found that between 1994 and 2005, self proclaimed knowledge of weeds increased.

Table 3-3 Public Knowledge about Weeds Knowledge Level 1994 2005 Not Knowledgeable 27.6% 20.0% Know a Little 40.1% 31.6% Know Something 24.4% 35.8% Know a Lot 8.8% 12.6% There was an increase in the percentage of respondents who rated themselves as knowing something or a lot about weeds and a decrease in those who rated themselves as knowing little or not being knowledgeable about weeds.

Assessment of the seriousness of the noxious weed problem was high and generally increased.

Table 3-4 Seriousness of Weed Problem Seriousness Level 1994 2005 Not Serious 4.8% 2.0% Somewhat Serious 19.4% 14.1% Serious 29.5% 37.1% Very Serious 46.3% 46.9% The statistically significant increase in respondent assessment of the seriousness of the weed problem indicated that the general public at large considers the weed problem to be serious.

The number of dangerous weeds known increased.

Table 3-5 Number of Weeds Known Number of Dangerous Weeds 1994 2005 Known 0 32.8% 21.6% 1 34.3% 39.9% 2 32.8% 38.5%

Page 40 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

The ability to name a dangerous weed increased.

Table 3-6 Ability to Name a Weed 1994 2005 73.4% 89.8% The 16.4 percent increase in respondents who could name a dangerous weed of any kind and the increase in respondents who could name a second dangerous weed showed a marked increase in weed awareness.

In addition, the surveys indicated high or increasing public awareness of how weeds spread, of the problems caused by weeds and how people can prevent weed spread.

At the local level, in June 2000, Missoula County voters passed a mil levy increase for weed management. This increase roughly tripled the previous mil levy element of the Missoula County weed budget. Powell County also voted an increase in their County weed management budget in 2000. The Powell County Commissioners sent in a letter of support (12/9/00) for the Proposed Action. County weed management mill levy increases in Missoula and Powell counties indicate additional public support for weed management on the Lolo NF.

Resources Weeds

Existing Conditions Dynamics of Weed Invasion. According to Cousens and Mortimer in Dynamics of Weed Populations (1995), weeds generally invade a region through a three-phase process: • Introduction As a result of dispersal, seeds or plant fragments arrive at a site beyond their previous geographic range and establish populations of adult plants. Potential new invaders such as yellow starthistle, tansy ragwort or rush skeleton weed could become a serious problem on the Lolo NF if allowed to establish and advance beyond the introduction phase. • Colonization The plants in the founding population reproduce and increase in number to form a self-perpetuating colony. Houndstongue and Dalmatian toadflax are examples of weeds in the colonization phase on the Lolo NF. • Naturalization The species establishes new self-perpetuating populations, undergoes widespread dispersal and displace native flora. Spotted knapweed is the most prevalent species in this stage in the Lolo NF ecosystem. Invasion and range expansion by a weed involves all three phases. Typically, weed invasions do not occur along a single front. Instead, new outbreaks initiated by long distance dispersal become the centers for shorter distance dispersal that may eventually fill the gaps between them.

The rate at which weed populations expand can be difficult to determine, and may be exponential, (i.e. a constant proportional rate of increase) or two-phased (with sudden range expansion following a period of little or slow increase in abundance).

Final Integrated Weed Management EIS Page 41 Chapter 3 Affected Environment and Environmental Consequences

Typically, only when the naturalization phase is reached, is when a species is likely to be recognized as a nuisance and considered a weed. Weed control efforts are then focused on limiting further spread of naturalized weeds into areas previously not infested. Eradication is a common goal to prevent naturalization. Thus, while a weed species may be considered naturalized within some parts of the Lolo NF, the same species may still be in the colonization or introduction phase on other areas of the forest.

How Weeds Spread Weeds spread from travel ways such as roads, trails, creeks, and rivers. Airborne dispersal occurs from wind and birds. Weed seeds and plant parts are moved along road and trail systems by vehicles, people and livestock allowing the establishment of weeds into areas previously not infested. As corridors, road systems allow weeds to invade both disturbed and undisturbed areas.

Creeks and rivers function similar to roads and trails in regards to weed spread, except that they are natural features and they can carry and spread weeds in the absence of human travel or other activity. Creek and riverside areas are critical routes of weed spread because high water tables and dense riparian vegetation reduce weed management options once a weed becomes established. Once established from waterways, weeds can spread upland onto disturbed and undisturbed areas.

Wildlife and domestic livestock also transport weeds. The spread of houndstongue, for example, is an indirect effect of livestock grazing and domestic dogs walking because the velcro-like seed sticks tenaciously to the hair of these animals. Weed seeds consumed by animals or attached to their fur are carried off road and trail corridors into less or undisturbed areas. Through the influences of wind, waterways and wildlife, weeds have been able to occupy undisturbed habitats far removed from road or trail systems.

Viable spotted knapweed seeds can also pass through the digestive systems of sheep and deer. Wallander, et. al. (1995) found that sheep manure contained viable spotted knapweed seeds from 2 to up to 7 days after dosing. Viable seeds were recovered from mule deer manure 10 days after dosing. Viability of seeds was reduced, but not eliminated by passing through sheep and mule deer. Sheep and especially mule deer, are likely to transport viable seeds of spotted knapweed and thus disseminate weed seeds. Wallander (1995) recommended that sheep be contained for at least 7 days after grazing a spotted knapweed infested area to allow viable seed to pass. Mule deer movement and subsequent weed dispersal cannot be controlled.

Present Infestations Infested acres are described by a combination of estimates, section surveys and site specific inventories. These values are then compared to current and past evaluations to determine acres at risk. These two value sets can then be compared to estimate current infested acres in relation to potential infested acres. Growth and biological characteristics for the weeds on the forest can be found in Appendix D. The status of each weed on the Lolo NF is displayed in Chapter 1 in Table 1-1.

Infested acres are periodically estimated for weed infestations on the Lolo. These estimates are done by District weed program coordinators because they have the most on the ground

Page 42 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences knowledge of the weed situation on each Ranger District. There is not a comprehensive weed inventory for the Lolo NF. While a comprehensive regularly updated inventory would be most desirable, at a conservative estimate of $2/acre (based on recent contract inventories and administration) this would cost approximately $4,310,370 for an initial inventory (not including periodic updates and data management) and far exceeds our capabilities. An estimate was conducted in 1999 for 23 species of weeds and indicated that 258,300 acres were infested. Another estimate was made in 2005 for the same 23 species and totaled approximately 391,700 acres (See Table 3-7 below). The duplicity (more than one weed species occupying the same acre) in these estimates has not been aggregated or estimated.

Table 3-7 2005 Estimated Infested Acres on the Lolo NF Weed species Missoula Ninemile Plains Seeley Superior Forest white top Cardaria draba 0 0 10 0 20 30 musk thistle Cardus nutans 75 15 25 50 150 315 diffuse knapweed Centaurea diffusa 0 0 0 0 1,000 1,000 spotted knapweed Centaurea maculosa 75,000 22,000 100,000 39,000 100,000 336,000 Russian knapweed Centaurea repens 0 0 0 0 0 0 yellow starthisle Centaurea solstitalis 0 0 0 0 0 0 rush skeletonweed Chondrilla juncea 0 0 0 0 0 0 oxeye daisy Chrysanthemum leucanthemem 250 200 200 150 2,500 3,300 Canada thistle Circium arvense 260 200 200 1,500 3,000 5,160 common crupina Crupina vulgaris 0 0 0 0 0 0 Houndstongue Cynoglossum officinale 250 40 50 75 800 1,215 blue weed Echium vulgare 0 0 10 0 0 10 leafy spurge Euphorbia esula 1,000 700 50 22 800 2,572 orange hawkweed Hieracium aurantiacum 5 5 3 0 5 18 yellow hawkweed Hieracium pratense 1 2 3 0 300 306 St. Johnswort Hypericum perforatum 1,500 600 5,000 1,000 5,500 13,600 dyers woad Isatis tinctoria 0 0 0 0 0 0 Dalmatian toadflax Linaria dalmatica 90 0 100 1 50 241

Final Integrated Weed Management EIS Page 43 Chapter 3 Affected Environment and Environmental Consequences

Weed species Missoula Ninemile Plains Seeley Superior Forest common toadflax Linaria vulgaris 2 50 7 100 100 259 purple loostrife Lythrum spp. 0 0 0 0 0 0 sulfur cinquefoil Potentilla recta 2,500 1,500 15,000 100 5,000 24,100 tansy ragwort Senecio jacobaea 0 0 0 0 0 0 common tansy Tanacetum vulgare 140 250 75 575 2,500 3,540 TOTAL 81,073 25,562 120,733 42,573 121,7255 391,666 Section inventories are another acreage estimate source. This data is collected and resides in the Natural Resource Information System (http://nris.state.mt.us/interactive.html) presence/absence section data base. The Lolo NF did a section by section presence/absence report in 1998 for five weeds (leafy spurge, Dalmatian toadflax, sulfur cinquefoil, Russian knapweed, and spotted knapweed). In this survey (Table 3-8), each weed was identified as present or absent in each section. If the weed were known to be present, the entire section acreage was counted as infested, so this inventory may exaggerate the number of infested acres on the forest.

Table 3-8 1998 NRIS Weed Inventory for the Lolo NF (Acres/ percent of Lolo NF) Status Leafy Spotted Russian Dalmatian Sulfur Spurge Knapweed Knapweed Toadflax Cinquefoil 73,276/ 1,616,389/ 43,104/ 206,898/ Present 0/0% 3.4% 75% 2% 9.6% 1,323,283/ 239,226/ 1,446,129/ 1,355,611/ 769,401/ Absent 61.4% 11.1% 67.1% 62.9% 35.7% 762,935/ 306,036/ 713,366/ 760,780/ 1,183,196/ Unknown 35.4% 14.2% 33.1% 5.3% 54.9% 2005 Estimate 2,572 336,000 0 241 24,100

Source: http://maps2.nris.mt.gov/mapper/ThemeList.asp?Profile=2088894&qLayer1=USFSDISTRI CT&qField1=Forest&qValue1=Lolo&Oper1=&Buffer1=0&TabName=Land%20Information

County infestation information for the Lolo NF is also displayed in map form on the Invader Database (http://invader.dbs.umt.edu. The Invader Database does not however estimate infested on the Lolo NF. The Invader database indicates that weeds are widespread across Montana. The Montana Weed Plan (http://agr.state.mt.us/weedpest/pdf/2005weedPlan.pdf) reports that about 8.2 million acres (or 9 percent) of the 93,153,280 acres in Montana are infested with weeds.

The mapping strategy on the forest is to site specifically map project areas and areas of recent disturbance where we anticipate the capability of conducting weed control activities rather than mapping areas where we do not anticipate control activities and would therefore not be able to use the mapping data for active weed management. The Lolo NF has conducted site specific detailed weed inventories on about 111,900 acres. These inventories have focused on burned area, big game winter ranges, areas of concentrated public use and related high risk weed areas.

Page 44 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

In the absence of a forest wide site specific weed inventory, the Lolo NF identifies areas to conduct weed control projects based on the priority site types identified in Amendment 11 to the Lolo NF Plan. These site types are: • Weed-free areas, and the roads, trails and trailheads that lead into them; • Areas of concentrated public use (since people spread weeds); • Administrative sites; • NF lands bordering private lands with active weed-controlling programs; and • Bunchgrass big-game winter ranges. Under this analysis, the priority site types would be expanded to include those described in the Proposed Action. The site types identified in the Proposed Action (Chapter 2) and sites shown on the project area maps in the Appendix A are the areas where weeds are presently found and those considered at risk of initial invasion and additional weed spread.

Acres at Risk Different vegetative cover types have a large influence in weed invasion. Vegetative cover types vary in susceptibility to different weeds based on precipitation, aspect, soil and competition. The availability of weed spread agents also plays a large role in weed risk. Susceptibility of various vegetative cover types on the Lolo NF has been evaluated using two methods.

In 2005, the Lolo NF conducted a risk evaluation for 15 species of weeds using the Maria Mantas/Cohesive Strategy Team/Nature Conservancy of Montana risk evaluation method entitled Evaluating Risk to Native Plant Communities from Selected Exotic Plant Species (http://www.fs.fed.us/r1/cohesive_strategy/datafr.htm). A three-tiered approach was developed in this evaluation to access risk to native plant communities in Montana and Northern Idaho from exotic plant species. This approach involved determining the susceptibility of areas to species’ establishment, assigning level of threat to susceptible areas, and factoring in the probability of exposure of each site to plant propagules by considering the various aspects that affect dispersal. Native plant communities were represented by various biophysical settings that were modeled as 34 potential natural vegetation (PNV) groups. Susceptibility, threat, and probability of exposure were then combined to spatially model degree of risk across the study area from some of the most threatening extant exotic species in Montana and Northern Idaho.

These 15 species covered by this risk evaluation and risk rating is displayed in Table 3-9 below.

Table 3-9 Estimated Acres at Risk on the Lolo NF Weed High Moderate Low Unknown No Risk Water Risk Risk Risk Risk White top - Cardaria draba 875 532 29,373 1,905,521 287,994 4,161 Spotted Knapweed - Centaurea maculosa 63,857 7,708 694,530 0 1,458,200 4,161 diffuse knapweed - Centaurea diffusa 224,103 12,856 2,558 4,932 1,831,993 4,161 rush skeltonweed - Chondrilla juncea 11,869 2,142 616,256 20,296 1,425,879 3,725

Final Integrated Weed Management EIS Page 45 Chapter 3 Affected Environment and Environmental Consequences

Weed High Moderate Low Unknown No Risk Water Risk Risk Risk Risk Canada thistle - Circium arvense 630 284 678,078 6 1,397,443 0 houndstongue - Cynoglossum officinale 0 0 364,029 349,892 1,362,520 0 leafy spurge - Euphorbia esula 61,564 8,180 306,108 20,666 1,679,923 0 orange hawkweed - Hieracium aurantiacum 327,221 93,554 202,372 56,726 1,396,568 0 yellow hawkweed - Hieracium canadense 361,593 102,867 5,688 238,759 1,367,534 0 St. Johnswort - Hypericum perforatum 54,100 7,220 654,990 2,206 1,357,927 0 dyers woad - Isatis tinctoria 52,545 12,864 0 27,984 1,983,049 0 dalmatian toadflax - Linaria dalmatica 24,663 6,773 326,470 11,006 1,697,529 0 common toadflax - Linaria vulgaris 12,770 4,237 662,045 25,370 1,372,020 0 sulfur cinquefoil - Potentilla recta 55,761 5,901 587,972 1,933 1,424,874 0 common tansy - Tanacetum vulgare 25,223 0 588,238 4,219 1,458,760 0 In April 1987, the Lolo NF Ecologist conducted An Evaluation of Weeds on the Lolo, Bitterroot and Flathead Forests with recommendations for implementing a Weed Control Program (Losensky 1987). That evaluation included information on the expected acres that could become infested for nine common weeds based on the acres of the various habitat types found on each forest. This provides an estimate of the severity of the present weed infestations in relation to the potential. Table 3-10 below compares present estimated infested acres to acres at risk for two different risk evaluations.

Table 3-10 Estimated Infested Acres Compared to Acres at Risk on the Lolo NF Species 2005 Acres At Moderate Acres At Moderate or Estimated or High Risk High Risk Infested Acres (Losensky Method) (Mantas Method) Spotted knapweed 336,000 431,547 71,565 Diffuse knapweed 1,000 267,580 236,959 Canada thistle 5,160 62,261 914 Musk thistle 315 4,475 Not Rated St. Johnswort 13,600 428,948 61,320 Houndstongue 1,215 307,700 0 Common Tansy 3,540 225,762 25,223 Dalmatian toadflax 241 230,940 31,436 Leafy spurge 2,572 436,013 69,744 While the acres at risk vary widely between the two risk estimate methods, it is apparent that the acres presently infested are still far below the acres estimated to be at risk for most weed species. This indicates we are still early in the invasion cycle and there is a lot of area that can be protected from weed invasion under the Proposed Action.

The Upper Columbia Basin Ecosystem Management Project (DEIS 5/1997 and FEIS 12/2000) also included a susceptibility evaluation for broad scale cover types in the

Page 46 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences analysis area. Although this evaluation was not specific only to the Lolo NF, it provides a broader look at weed risk over the Columbia River Basin area.

Table 3-11 Broad Scale Cover Types in the Project Area and their Susceptibility to Invasion for 24 Weed Species

Veg Sulfur Rush Cheat Spotted Diffuse Musk White Russian Yellow Oxeye Canada Common cinqu skeleton Cover grass knapweed knapweed thistle top knapweed starthistle daisy thistle crupina Type efoil weed Grassland H H H H, M M M H H M M H M Dry Forest/ H H M H M M M H M M H M Shrub Moist Forest M M M H, M M M M M M H, M H L, M

Wetland M H M H, M M M M H L M H L

Burned Area H H H H H H H H H H H H

Table 3-11 continued Veg Leafy Orange Yellow St. Purple Common Tansy Dyers Dalmatian Common Hounds Blue- Cover spurg hawk hawk Johns loose tansy ragwort woad toadflax toadflax tongue weed Type e weed weed wort strife Grassland H, M U U L L U H H U L U U Dry Forest/ M U U M L, M U L M U L U U Shrub Moist Forest M U U M M U M M U M U U Wetland M U U H, M M U M M U H U U Burned Area H H H H H H H H H H H H (Adapted from An Assessment of Ecosystem Components in the Interior Columbia Basin and Portions of the Klamath and Great Basins, Volume II, Table 3.174, June 1997). H =High susceptibility to invasions - Weed species invades the cover type successfully and becomes dominant or co-dominant even in the absence of intense or frequent disturbance.

M =Moderate susceptibility to invasion - Weed species is a “colonizer,” and invades the cover type successfully following high intensity or frequent disturbance which impacts the soil surface or removes the normal canopy.

L =Low susceptibility to invasion - Weed species does not establish because the cover type does not provide suitable habitat.

U =Unknown susceptibility to invasion - Ecological requirements of the weed species are not known, or there was a lack of distribution records for the weed species, or the extent of the cover type in the project area might be so minor as to prevent or restrict the probability of obtaining distribution records for the weed species within that cover type.

Final Integrated Weed Management EIS Page 47 Chapter 3 Affected Environment and Environmental Consequences

Biological Agents At the end of calendar year 2005 there had been 413 releases of 34 biological agents on six weed species on the Lolo NF since 1978. A complete listing of the weed species and biological agents is on file at the Lolo NF Supervisors Office.

Method of Analysis Since weed treatment results may vary based on soils, vegetation type and local climate, the effects on weeds for both alternatives are based on local (Lolo and Bitterroot NF specific) monitoring. The effects and monitoring are based on test plots and projects using the same practices and on the same sites and site types as proposed for treatment in this analysis. This local site specific information is used to describe environmental consequences on weeds.

The primary sources in addition to those previously cited in this analysis that were used to assess the effects of the alternatives on weeds are: • FEIS and ROD for Noxious Weed Management on the Lolo National Forest (U.S. Department of Agriculture-FS 1991c). This analysis includes estimated weed spread rates. • Cost and efficiency of Spotted Knapweed Management with Integrated Methods (Duncan, et al. 1999). This study provides Lolo NF site specific measurements of control that can be expected from various weed control methods*. • Weeds on the Lolo, Bitterroot and Flathead Forests – with Recommendations for Implementing a Weed Control - An Evaluation of Noxious Weed Program, (Losensky 1987). • The Montana Weed Management Plan, (Lacey-Duncan 2001) which includes estimated spread rates. • Results and observations from the Mormon Ridge Winter Range Restoration Project (USDA 1996)*. • Results from restoration attempts on the Sawmill Research Natural Area on the Bitterroot National Forest (Rice and Harrington 2003)*. • Observations and monitoring by Lolo NF weed and other resource specialists*. • Monitoring results from projects included in the FEIS and ROD for Big Game Winter Range and Burned Area Weed Management on the Lolo (USDA-FS 2001)*. *denotes local site specific projects with the same objectives and methodologies as proposed in this analysis

Environmental Consequences - Weeds

Herbicide Control

Direct and Indirect Environmental Consequences Common to Both Alternatives The weed effects described below are what we have found from on the ground monitoring on the Lolo and Bitterroot NF. Under the Proposed Action, a slightly broader selection of herbicides would be available for use. Herbicides are selected based on which one(s) provide the best potential weed control, the least non-target effects, and considering site

Page 48 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences conditions or restrictions. The anticipated effects on weeds would be as good as or better than, described below. The recent registration of aminopyralid is expected to reduce or replace the use of picloram and 2,4-D in many situations.

Herbicide resistance is a greater risk when weeds are treated several times a year with the same herbicide. While there is a possibility that some weeds may have or develop resistant characteristics, this is considered low risk for the weeds found on the Lolo NF because they are not treated more than once a year and herbicides can be alternated to broaden efficacy on a subpopulation that may be suspected of resistance.

Mormon Ridge: Using aerially applied herbicides (picloram at 1.5 pints/acre) and biological management agents (Urophora affinis, Urophora quadrifasciata, Agepeda zoegana and Cyphocleonus achetes), 95 percent spotted knapweed control was achieved on the Mormon Ridge pilot project (Henry 1998). Pioneering infestations of leafy spurge were suppressed. Four years after the initial treatment spotted knapweed control was as shown in Table 3-12:

Table 3-12 Mormon Ridge Four Year Post Treatment Weed Canopy Cover Four year post Pretreatment Pretreatment spotted treatment spotted Plot spotted knapweed knapweed Absolute % knapweed Absolute % density Canopy Cover Canopy Cover 1 High 80% 0% 2 Medium 60% 0.5% 3 Absent 0% 0% 4 Low 10% 0% 5 High 80% 0.5% * U.S. Department of Agriculture/Rice, Peter; Lavelle, Darlene; Greenlee, Jack May 2003 Ordinations were performed using Detrended Correspondence Analysis (DCA) to determine if knapweed treatments caused plant community change trajectories toward or away from potential natural plant communities as defined by Mueggler and Stewarts (1980) potential natural habitat type definitions. Plots 2, 3 and 4 were within or near that definition pretreatment. In the years post spray, these plots all became more typical of the Mueggler and Stewart standard. In the high density spotted knapweed plots (Plots 1 and 5), the treatments did not drive these plots toward the Mueggler and Stewart standard. Treatment primarily exchanged knapweed cover for cheatgrass standard due to the absence of bunchgrass propagules in severely degraded sites where knapweed competition has eliminated native plants (U.S. Department of Agriculture/Rice, Peter; Lavelle, Darlene; Greenlee, Jack May 2003).

These findings indicate the need to: 1) treat spotted knapweed earlier in the invasion cycle before native and desirable plants have been competed out of the plant community and 2) include cheatgrass control and/or revegetation treatments on weed communities with a cheatgrass understory and/or an absence of desirable vegetation to replace spotted knapweed after treatment.

Big Game Winter Ranges and Areas Burned in the 2000 Wildfires: Herbicide efficacy on the Lolo NF was monitored in relation to species composition and production on nine burned areas and big game winter range monitoring sites since the summer and spring of

Final Integrated Weed Management EIS Page 49 Chapter 3 Affected Environment and Environmental Consequences

2002. Individual weed species monitored on these sites were spotted knapweed (Centaurea maculosa), sulfur cinquefoil (Potentilla recta), St. Johnswort (Hypericum perforatum), Dalmatian toadflax (Linaria dalmatica), and cheatgrass (annual Bromus spp.). After herbicide treatment, these sites were re-sampled in the first, second and third years post- spray to assess the response of herbaceous life forms to herbicide treatment.

On average, forest-wide target weed suppression was 94 percent in the first and second year post-spray; ~100 percent in the third year post-spray; and 63 percent in the fourth year post-spray compared to the pre-spray weed abundance for the sampled plots in spray project areas. Production of desirable graminoid species increased 31 percent of the total biomass before spraying to 67 percent in the first year post-treatment, 50 percent in the second year post-treatment, 67 percent in the third year post-treatment, and 60 percent in the fourth year post-treatment. There was post-spray recovery of non-target forbs, with the highest rates of recovery coinciding with lower rates of herbicide application where non- target forb production in the second year post-spray was equivalent to the abundance before spraying. Cheatgrass increased in abundance in five of the eleven sprayed plots. The greatest cheatgrass increases were at those sites that were sprayed with higher rates of herbicides because they were more degraded, were dominated by rhizomatous weeds, lacked perennial grasses and had a high frequency of occurrence of cheatgrass before treatment. In the O’Brien Creek project the two monitoring plots originally established in 2002 and one plot established in 2004 were subjected to a secondary treatment aimed at decreasing cheatgrass abundance and obtaining further control of knapweed. These O’Brien Creek monitoring plots had little to no decrease in abundance of cheatgrass following the second spray treatment, but the knapweed control was over 99 percent (Rice and Gauer, 2007).

Four other big game winter ranges on the Lolo NF were monitored and aerially treated with picloram at a rate of 1 pint/acre in September 2002. Each project area contains two study sites, one of which has been comprehensively monitored since 1999, prior to treatments, and a second of which has been sampled primarily after treatments were applied. Sites in the Eddy and Madison winter ranges contained heavier infestations of weeds prior to treatment while those located in the Blacktail and Greenough winter ranges had only light levels of weeds. In addition, conditions were monitored in eight other control (i.e., no-treat) sites to document annual changes in vegetation unrelated to treatments (Ortega 2005).

In 2003, the first year following treatment, weed canopy cover was low at knapweed- invaded and especially native-dominated sites that had been treated in 2002. However, by 2004, cover of the cheatgrass (Bromus tectorum) at treated sites had increased markedly relative to prior years, replacing knapweed as the most abundant weed at seven of eight sites. Other exotic forbs, including yellow salsify (Tragopogon dubius) and Jim hill mustard (Sisymbrium altissimum), remained at low levels in the two years following treatment. Less common weeds (on these project areas) such as houndstongue (Cynoglossum officinale) and sulphur cinquefoil (Potentilla recta), occurring at trace levels prior to treatment, remained at comparable levels after treatment. Comparison of untreated to treated sites also indicates drought conditions may have also affected weed control and cover. However, treated versus control sites saw a sharper decline in weed cover over the study period, indicating that herbicide succeeded in reducing abundance of target weeds beyond natural declines. This pattern was not perceptible at native-dominated sites where weeds occurred at low abundance prior to treatment (Ortega 2005).

Page 50 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

On heavily infested knapweed winter ranges, spotted knapweed was reduced from 17 percent canopy cover in 2000 to 1 percent in 2004 on Eddy South and from 27 percent canopy cover in 2000 to 3 percent in 2004 on Madison West (Ortega 2005).

Gibbons Pass Weed Control Monitoring: The Gibbons Pass elk winter range on the Bitterrroot NF was helicopter sprayed with picloram at 1 pt/ac plus R-11 non-ionic surfactant on September 30, 2004. First and second growing seasons post spray responses were measured on August 3, 2005 and August 7, 2006, respectively. Both pre and post treatment monitoring occurred. This elk winter range is an Idaho fescue/bluebunch wheatgrass habitat type at 5,980 ft. elevation with an east/southeast aspect and an average 25 percent slope. The dominant plant on the site at the time of baseline sampling in 2004 prior to treatment was spotted knapweed.

Two growing season later, the results were monitored after the site was aerially treated with picloram at 1 pt/ac with a non-ionic surfactant. The fall treatment was 100 percent effective in suppressing the spotted knapweed in the monitoring plot for two growing seasons post-spray. Total absolute canopy cover at Gibbons Pass of 35 percent to 44 percent during this current cumulative drought period is less than half of the 108 percent average canopy cover measured by Mueggler and Stewart in the early 1970’s for this habitat type. The minimum absolute canopy cover observed by Mueggler and Stewart was 52 percent, and most of their plots had more than 100 percent absolute canopy cover. Species richness in the Gibbons Pass plot sample was 22 in 2004 then down to 16 in 2005 and increased to 19 in 2006. Average species richness is 25 for the Mueggler and Stewart Idaho fescue/bluebunch wheatgrass habitat type definition plots. The species observed at Gibbons Pass in 2004 but not present in the 2005 and 2006 samples were mostly forbs in the Asteraceae family. The composite family has numerous species, including spotted knapweed, that are susceptible to picloram. However these species typically start to reestablish from seed in the soil in subsequent growing seasons (Rice 2006).

Elimination of the growth of spotted knapweed by spraying was followed by a three fold plus increase in absolute production of native grasses in 2005 and no knapweed. Absolute graminoid production dropped greatly from 2005 to 2006 (57.5 percent decrease). This 2005 to 2006 decline in graminoid production was measured at numerous other bunchgrass sites being studied by the report author (Rice). As an example, at the intensively studied Sawmill Research Natural Area on the Bitterroot National Forest, the graminoid production in 2005 was 1,222 lbs/ac but only 387 lbs/ac in 2006 (68 percent decrease). Dry and warm weather in May 2006 appears to have arrested growth of perennial grasses, and although June 2006 rainfall was above average, the grasses did not recover from the May suppression.

The large increases in biomass of non-target forbs was due in part to inclusion of the large forbs Balsamorhiza sagittata and Lupinus sericeus in some of the randomly place clipping sample frames.

The pre-spray community compositions at Gibbons Pass in 2004 was atypical of the Idaho fescue/western wheatgrass habitat type as defined by Mueggler and Stewart’s Grassland and Shrubland Habitat Types of Western Montana. The floristic trajectory after spraying showed progression towards the potential natural community in both 2005 and 2006. Most of the dissimilarity of the 2004 plot at Gibbons Pass from the habitat type definition

Final Integrated Weed Management EIS Page 51 Chapter 3 Affected Environment and Environmental Consequences plots was due to the low abundance of Idaho fescue, western wheatgrass, and other native grasses, plus the pre-treatment dominance of spotted knapweed.

Sawmill Creek Research Natural Area Weed Control Monitoring: The Bitterroot National Forest established the 270 acre Sawmill Research Natural Area (RNA) in the Sapphire Mountain Range because it was one of the best remaining examples of native bunchgrass and open canopy forest habitat types. The RNA is also an important elk winter range. However, invasion of the RNA by weeds is degrading the grasslands, and Douglas-fir encroachment of the forested ravines (because of fire suppression) is altering the species composition and structure of the forest types. Weed distribution was mapped in 1994 and a restoration plan developed in 1995. The overall strategy was to first suppress the weeds in the grasslands, reduce stand density in the forested ravines by thinning, and then return fire through prescribed burning to the grasslands and the forested areas. This is an integrated weed control and restoration demonstration project at the management unit scale, in addition to a focused restoration effort for the altered ecosystems of this RNA. Extensive networks of vegetation monitoring plots were established to provide scientific documentation of progress in returning the potential natural vegetation. The same integrated weed control is proposed in Alternative 2. (Rice and Harrington 2003)

Implementation began in 1995 by setting up herbicide test plots and introducing knapweed biocontrol insects. Knapweed biocontrol insects, Agapeta zoegana (knapweed root moth) and Cyphocleonus achates (knapweed root weevil) were released on the site starting in 1996 (Rice 2000). Urophora (knapweed seed head gall flies) species were already well established on the spotted knapweed as of 1994. Chrysolina (Klamath weed beetles) were already present on goatweed at the start of the project. (Rice and Harrington 2003)

In the fall of 1996, broadcast treatments for spotted knapweed control were applied on 14 experimental monitoring plots. The timing corresponded to fall regrowth of knapweed rosettes. Each plot was treated with picloram at the recommended rate for control of spotted knapweed (0.25 lb active ingredient/acre or 1 pint/acre picloram). The application dates were from September 26 through 30. An 80 by 80 foot area centered on each 1/10 acre plot was treated so there was a sprayed buffer between the surrounding weeds and the sampling plot. (Rice and Harrington 2003)

During the period 9/26-10/31/98 the extensive grassland portions of the RNA were broadcast sprayed. The extensive grassland treatments had to be applied in a manner that avoided spraying a certain number of the permanent monitoring plots and biocontrol release plots. Herbicides were sprayed when weather conditions were favorable. ATV boom, truck-mounted handguns, and Solo backpacks were used. Most of the acreage was treated with the ATV-mounted boom. Picloram was applied at 1 pint/acre to 133.5 acres and clopyralid at 2/3 pint/acre to 18.9 acres. Total calculated coverage was 152.4 acres of the nominal 156 acres targeted for treatment. (Rice and Harrington 2003)

Hand digging, pulling, and limited spot spraying have been used to suppress small colonies of other weeds found on the RNA. The road prisms have also been periodically sprayed with clopyralid or picloram. (Rice and Harrington 2003)

The 1996 trial spray treatments were successful in driving the plant community composition to a state more similar to the potential natural vegetation as defined by the Mueggler and Stewart habitat types found on the RNA. In 1998 nearly the entire grassland

Page 52 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

(~160 acres) was sprayed. Bunchgrass cover was restored to the site and native forb diversity was maintained. The biomass that had been represented as coarse-stemmed knapweed was converted to fine grass. (Rice and Harrington 2003)

Production of habitat type indicator bunchgrasses and other bunchgrasses were greatly increased by the fall 1996 herbicide spraying. Bunchgrasses were the three habitat type indicator species, Agropyron spicatum, Festuca idahoensis, and Festuca scabrella. Other Grasses were primarily Koeleria cristata, Poa secunda and Stipa spp. Increases in perennial bunchgrass production were as much as two to three fold depending on the grass species and number of years post spray. Approximately a pound of bunchgrass was realized for every pound of knapweed that was suppressed by spraying. Knapweed dominated no-spray Idaho fescue/bluebunch wheatgrass check plots on the lower slopes had almost no production of habitat type indicator grasses in 1999 and less than 100 lbs/ac in 2002. (Rice and Harrington 2003)

Cheatgrass (Bromus tectorum and Bromus japonicus) had been a minor component of the sprayed test plots in every year sampled. For the ten full community plots that were sprayed in fall 1996 and sampled on a near yearly schedule the cheatgrass canopy cover averaged 1.8 percent. (Rice and Harrington 2003)

The suppression of knapweed production was huge, dropping from 497 lbs/acre in 1996 to 11 lbs/acre at one-year post spray (1997). Resurgence of knapweed has been slow and increased perennial grass production maintained through 2002. Knapweed began to reenter the stands in significant amounts in the third year post spray (1999) but had not increased proportionately as of 2002. (Rice and Harrington 2003)

Complete coverage of the infested areas would be enhanced with the use of GPS guided aerial herbicide applications. There would be the least amount of skips and misses on treatment acres where aerial application is used on large infestations under both alternatives. Good coverage could reduce skips, increase treatment accuracy, and reduce the number of re-treatments. This could in turn reduce both the volume of herbicide applied and the number of herbicide applications needed over the long term. More effective and accurate treatments would reduce the chance of new invaders establishing and growing to large infestations before they are detected. Therefore, the weed soil seed bank would be reduced more effectively with aerial application. This is not a new program on the Lolo NF. The Lolo NF has a proven history and excellent safety record in the use of aerial herbicide application since 1992.

Herbicides and Hand Pulling

Direct and Indirect Environmental Consequences Common to Both Alternatives Ground Applied Herbicides on Missoula Valley Open Space Lands: St. Johnswort and associated weeds were treated on Mount Jumbo, a City of Missoula Open Space property with 1) hand pulling in 1999, and 2) 3 pints/acre of picloram in spring 2000.

St. Johnswort plots had significantly less cover one year after hand pulling. Public participation in the hand pulling was poor. Nine people showed up to pull in May 1999

Final Integrated Weed Management EIS Page 53 Chapter 3 Affected Environment and Environmental Consequences and the progress was slow. Weed cover on the herbicide plots decreased from 33 percent to 3 percent cover in 2000. (Marler 2004)

Spotted knapweed and associated weeds were also treated on Mount Jumbo, a City of Missoula Open Space property with 1) hand pulling in 1999, and 2) 1 pint/acre of picloram in fall 2000, with spot treatments in 2004.

Spotted knapweed plots had showed no change in cover one year after hand pulling. Participation in the hand pulling was unorganized and poorly documented. Weed cover on the herbicide plots decreased from 32 percent to 1 percent cover by 2004. (Marler 2004)

Herbicide Alone and in Combination with Hand Pulling, Mowing and Biological Control

Direct and Indirect Environmental Conditions Common to Both Alternatives Blue Mountain Recreation Area Spotted Knapweed Control Test Plots: Spotted knapweed control test plots were set up on the Lolo NF at the Blue Mountain Recreation Area in 1997. Weed practices examined included herbicides alone and in combination with hand pulling, mowing and the biological control agent Cyphocleonus achetes. Post-treatment vegetation cover data for all treatments were collected in August 1997, June and August 1998, and June 1999. Visual percent control evaluations of each treatment were made in August 1997 and 1998, and June 1999.

Insects alone or in combination with picloram at ¼ pint/acre did not provide good spotted knapweed control. Mowing twice for two consecutive years and Cyphocleonus achates alone did not control spotted knapweed. Mowing and insects alone had no significant effect on grass cover, litter, or forbs compared to the control. Insects alone and combined with herbicides may prove more cost effective if insects establish and maintain long-term control. (Duncan, et. al. 1999)

Picloram at rates of ½ pint/acre and above, and Curtail (a mixture clopyralid and 2,4-D) at 2 quarts/acre provided greater than 90 percent control one year following application. Mowing combined with Curtail at 1 quart/acre provided significantly better knapweed control than this herbicide rate alone. Insects alone or in combination with picloram at ¼ pint/acre did not provide good spotted knapweed control. Herbicide treatments alone and in combination with hand pulling and mowing increased grass cover by greater than 44 percent.

Herbicides alone provide the most effective spotted knapweed control for the lowest cost. The most cost-effective treatments were picloram at ½ to 1 pint/acre, and Curtail at 2 quarts/acre. Herbicides at half the standard rate can be combined with mowing or hand pulling to improve initial control or maintain control over a longer period of time. (Duncan, et. al. 1999)

Insects alone and combined with herbicides may prove more cost effective if insects establish and maintain long-term control.

Page 54 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Differences between Alternative 1 and Alternative 2 for Herbicides, Hand Pulling and Mowing: Alternative 2 would treat up to three times the weed infested acreage and would therefore better meet the Purpose and Need of this project.

Biological Control Only

Direct and Indirect Environmental Consequences Common to Both Alternatives Spotted knapweed is the most widespread weed on the Lolo NF. A research program has been underway since 1974 at the Montana State University Western Agricultural Research Center (MSU/WARC) at Corvallis, MT, to biologically control spotted knapweed by introducing Eurasian-collected insects proven to be host-specific to the plant. To date, 12 insect species (eight flower head insects and four root insects) have been introduced against spotted knapweed in Montana. Many of these insect species have been released within just the last few years.

Table 3-13 Status and Effects of Spotted Knapweed Biological Control Agents Part of weed Insect / Biological the insect Status Effect Control Agent affects reducing knapweed seed Urophora affinis* seed head fly well established production by 50-90 percent Urophora reducing knapweed seed seed head fly well established quadrifasciata* production by 50-90% Metzneria seed head established in low reducing knapweed seed paucipunctella* moth numbers production established in good reducing knapweed Agapeta zoegana* root moth numbers at some vigor and density in locations some areas reducing knapweed Cyphocleonus established in root weevil vigor and density in achetes* moderate numbers some areas Pelochrista established in low root moth NA medullana numbers Pterolonche root moth not established NA inspersa* seed head established in low reducing knapweed seed Larinus obtusus* weevil numbers production causing decline of seed head established in Larinus minutus* diffuse knapweed in weevil moderate numbers some areas established in low Terellia viren*s seed head fly NA numbers Chaetorellia established in low seed head fly NA acrolophi numbers reducing knapweed Sphenoptera established in root beetle vigor and density in jugoslavic* moderate numbers some areas seed head Bangasternus fausti* status uncertain NA weevil

Final Integrated Weed Management EIS Page 55 Chapter 3 Affected Environment and Environmental Consequences

Source: MSU/ARC at http://ag.montana.edu/warc/biocontrol_agents_of_knapweed.htm The Lolo NF has released 11 biological control species on spotted knapweed. MSU/WARC studies have shown that the two seed head flies, Urophora spp., are reducing knapweed seed production by a minimum of 50 percent where the two flies coexist. Likewise, recent studies indicate that the root moth, Agapeta zoegana, and the root weevil, Cyphocleonus achates, are causing significant reductions of spotted knapweed vigor and density at several release sites. (http://ag.montana.edu/warc/biological_control_of_knapweed.htm)

Two other spotted knapweed biological control agents, Agapeta zoegana and Cyphocleonus achates have slow reproductive rates so MSU/WARC has initiated a mass-rearing program to hasten their distribution and population increase around the state. Both insects have now been released in every knapweed-infested county in the state. (http://ag.montana.edu/warc/biological_control_of_knapweed.htm)

Biological releases alone can in some cases control leafy spurge, for example on some southerly, dry, lower elevation sites such as one spurge infestation near Hayes Creek at the Blue Mountain Recreation area where the Apthona flea beetle has provided over 95 percent spurge control (U.S. Department of Agriculture-FS 1996). This result is not consistent however on all Apthona releases on leafy spurge on the Lolo NF.

Another researcher suggests that biological control insects may increase competitive ability of spotted knapweed. (Callaway et. al. 1999) found that when one biocontrol insect (Agapeta zoegana) fed on the roots of spotted knapweed neighboring Idaho fescue (Festuca idahoensis) plants actually did more poorly than when grown with un-attacked spotted knapweed. Instead of releasing the neighboring grass from competition and allowing it to thrive as intended, the insect had no detectable impact on knapweed and may have indirectly suppressed the fescue. Some folks suggest that some other biocontrol agents could be having negative indirect effects on some native species that are not yet recognized.

Biological controls have been released extensively on the Lolo National Forest. Since 1978, the Lolo has conducted over 400 releases of 35 different biological control agents on seven weed species (Canada thistle, spotted knapweed, diffuse knapweed, leafy spurge, St. Johnswort, Dalmatian toadflax and musk thistle). On some sites, biological agents have spread onto the Lolo from releases on adjacent lands.

The results of these releases have been variable. Some biological agents have noticeably reduced some weed infestations. Other releases have had little apparent effect so far. While some biological agents can fly and readily spread to other weed infestations, others have to hike from weed to weed, which slows their rate of spread. Some biological agents cannot survive annual variation in our weather. Biological control also appears cyclic, meaning that weed populations vary as the populations of biological control agents wax and wane in response to the weed population and other factors. Biological control monitoring files at the Lolo National Forest indicate anywhere to over 90 percent to little or no weed control from biological control agents depending on factors such as weed species, biological control species, aspect and elevation (USDA, Lolo NF 1996, 1998).

Page 56 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Differences between Alternative 1 and Alternative 2 for Biological Control The Lolo NF already has an aggressive biological management program. While this program would expand further under Alternative 2, due to the level of the current program, the effects of biological control agents and biological control work on adjacent non Lolo NF lands, the difference between the two alternatives would not be significant in relation to biological agent weed control.

Hand Pulling and Mowing Alone and With Herbicides

Direct and Indirect Environmental Consequences Common to Both Alternatives Spotted knapweed control test plots were set up on the Lolo NF at the Blue Mountain Recreation Area in 1997. Weed practices included hand pulling and mowing. Post- treatment vegetation cover data for all treatments were collected in August 1997, June and August 1998, and June 1999. Visual percent control evaluations of each treatment were made in August 1997 and 1998, and June 1999.

Mowing alone did not provide control of knapweed plants, but reduced flowering. Mowing combined with Curtail at one quart/acre provided significantly better knapweed control than this herbicide rate alone. Hand pulling alone eliminated flowering and provided 56 percent control of spotted knapweed plants. Mowing twice for two consecutive years and Cyphocleonus. achates alone did not control spotted knapweed.

Hand pulling for two years significantly increased bare ground, but did not affect grass, litter and other forb cover compared to the control. Mowing and insects alone had no significant effect on grass cover, litter, or forbs compared to the control.

Hand pulling twice for two consecutive years was the most expensive treatment and provided less than 60 percent spotted knapweed control after two seasons. Hand pulling is not an economically viable option on dense and/or large acreage knapweed infestations (Duncan, et. al. 1999).

Based on cost estimates from the 2001 Lolo NF Big Game Winter Range and Burned Area Restoration EIS, it could cost more than $50,000,000 to treat a maximum of 15,000 acres/year by hand pulling and mowing alone without the use of herbicides (Refer to Chapter 2, Alternatives Considered But Not Given Detailed Study). This would equate to approximately $3300/acre/year, but hand pulling under both alternatives would only be conducted on a spot basis and not on an acre and larger infestations.

Final Integrated Weed Management EIS Page 57 Chapter 3 Affected Environment and Environmental Consequences

Differences between Alternative 1 and Alternative 2 for Hand Pulling and Mowing Alone and With Herbicides: None.

Sheep or Goat Grazing

Direct and Indirect Environmental Consequences Specific to Alternative 1 There is no sheep or goat grazing under Alternative 1.

Direct and Indirect Environmental Consequences Specific to Alternative 2 Montana State University and the Montana Sheep Institute (MSU/MSI) are involved in 22 projects on 100,000 spotted knapweed and leafy spurge infested acres that include 31 monitoring sites, 31 sheep producers, 30,000 ewes and lambs, and 1000 landowners (Kott 2005). Results from 2003 and 2004 indicate that sheep will graze about twice as much leafy spurge as grass on high (78 percent spurge) and medium (50 percent spurge) infested sites as grass. Spurge utilization on low (26 percent spurge) infested sites was over 60 percent on spurge and just over 40 percent on grass. On spotted knapweed infested sites, sheep will graze about twice as much spotted knapweed as grass on high (89 percent knapweed), medium (52 percent knapweed) and low (10 percent knapweed) infested sites as grass. Spurge utilization on low (26 percent spurge) infested sites was over 60 percent on spurge and just over 40 percent on grass (Kott 2005). Kott reports that over time this type of grazing should favor re-establishment of grass and forbs.

MSU/MSI (Kott 2005) report that sheep can selectively graze leafy spurge plants consuming 50 to 95 percent of the leafy spurge biomass. The goal for sheep grazing to control leafy spurge is to remove the yellow. Sheep should begin grazing leafy spurge as soon as it reaches about three to four inches in height. This prevents the majority of leafy spurge plants from going to seed and prevents dense canopies of cover to develop.

Three years of repeated sheep grazing on leafy spurge significantly reduced spurge seeds in the seed bank and seedling densities, potentially minimizing the rate of spread by seeds. Although stem densities were not reduced, stem heights were reduced indicating that the competitive ability of leafy spurge may have been reduced in the short term (Olson and Wallander 1998). Olson and Wallander (2003) found that leafy spurge was less likely to be grazed than Idaho fescue and bluebunch wheatgrass in the early summer, but were more likely to be grazed in mid to late summer. In general, sheep in their study removed more of the canopy of leafy spurge than these grasses.

Sheep grazing leafy spurge infested rangeland pick up seed in their fleece and it is unlikely that this seed will fall on site. On the other hand, seeds that are consumed and passed through the digestive tract may enhance the dispersal of leafy spurge. Land managers using sheep to graze leafy spurge are advised to graze leafy spurge so that it does not flower, or if it does set seed, to not move sheep from infested to non-infested areas unless the animals are confined at least 5 days (Olson, et. al. 1997). Viable spotted knapweed seeds also passed through the digestive systems of sheep. Sheep manure contained viable spotted knapweed seeds from 2 to up to 7 days after dosing. Viability of seeds was reduced, but not eliminated by passing through sheep. Sheep are likely to transport viable seeds of spotted knapweed

Page 58 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences and thus disseminate weed seeds. Wallander (1995) recommended that sheep be contained for at least 7 days after grazing a spotted knapweed infested area to allow viable seed to pass. (Wallander, et. al. 1995)

Sheep will graze spotted knapweed. Repeated sheep grazing initially reduces current flower stem production; although flower stem production may increase in subsequent years if grazing is not continued. By controlling the time of grazing and by repeatedly grazing spotted knapweed while associated grasses were going dormant, the age class distribution of spotted knapweed was altered (i.e. there were fewer, albeit older and larger plants), and it’s reproduction was reduced with minimal impact on associated grasses. Sheep grazing may slow the rate of increase of spotted knapweed in native plant communities (Olson, et. al. 1997). Three years of repeated sheep grazing on a spotted knapweed and Idaho fescue community reduced shoot and root biomass of Idaho fescue, where as shoot and root biomass of spotted knapweed were not reduced (Olson and Wallander 1997).

Sheep and goats have different dietary preferences and grazing behavior, but both will readily graze leafy spurge (USDA ARS APHIS 2002). Grazing by sheep or goats stresses leafy spurge, taxes its root system, reduces seed production and increases it’s vulnerability to other control tools (such as biological control or herbicides.) It can be used to prevent the spread and reduce the density of leafy spurge infestations. When incorporated as a long term management tool, multi species grazing can reduce leafy spurge densities by 80-90 percent after three to five years of grazing. It is also useful in riparian areas where herbicide use is often restricted. Sheep and goats will also consume other weeds such as spotted knapweed, ragwort, and some thistles and mustards (USDA ARS APHIS 2002).

Differences between Alternative 1 and Alternative 2: Alterative 1 does not include sheep and goat grazing for weed control, but Alternative 2 does. Under carefully managed sheep and goat grazing situations, Alternative 2 could provide additional weed control, another weed management tool, and better meet the Purpose and Need for this project.

Summary of Direct and Indirect Weed Environmental Consequences Specific to Alternative 1 This alternative would reduce weeds on sites covered under previous analyses but would not meet the Purpose and Need as stated in Chapter 1. Weeds would spread unchecked on the project acres listed in Chapter 2. Spread rates would vary between less than 1 to 10 percent per year depending on the weed species, size of current infestation, disturbance history, present disturbances and site productivity at each site.

Site productivity is a measure of how much total plant material a site will produce and is a function of the moisture and nutrients available. For example, spotted knapweed production on the Mormon pilot project was 1,075 lbs/acre prior to treatment. After biological agents and initial herbicide treatments spotted knapweed production decreased by 95 percent to approximately 50 lbs/acre. Most or all of this weed reduction is attributed to the herbicide treatment.

Final Integrated Weed Management EIS Page 59 Chapter 3 Affected Environment and Environmental Consequences

On project areas where cheatgrass is well represented and only broadleaf selective herbicides are applied, there will be an increase in cheatgrass. This effect would be reduced if practices are used that control cheatgrass.

Weeds presently occurring at smaller infestation levels but without previous analyses would infest additional acres. For example, a four acre sulfur cinquefoil infestation, at a spread rate of 10 percent per year could occupy approximately 4.4 acres in one year, and approximately 6.4 acres in 5 years. In another example, a 388 acre spotted knapweed infestation could spread to 427 acres in one year and 624 acres in 5 years. While both these weeds spread at 10 percent per year, the sulfur cinquefoil would spread by 2 acres whereas spotted knapweed would spread by 236 acres. This demonstrates the exponential nature that weeds would expand within the project areas.

Rhizomatous weeds would spread faster than tap rooted weeds because they have both above ground (seed) and below ground (rhizomes) reproductive abilities. This dual spread mechanism would allow rhizomatous weeds to spread in a wider range of geographic and competition regimes.

Weeds could spread more rapidly in burned areas because characteristics are more favorable. These characteristics include recent soil disturbance, opening of the overstory canopy, soil damage from heat, removal of competing vegetation and fire suppression activities by persons, equipment and vehicles that could spread existing and introduce new weed seed and rhizomes.

Under this alternative, weeds would spread and expand more rapidly on the project areas than under Alternative 2. The density of existing infestations would also increase. New weed species would establish and co-dominate or replace existing weed species. Weeds would spread into portions of the project areas that are presently weed free. The project areas would serve as weed seed sources for other areas of the forest and nearby non- National Forest System lands. Weeds could eventually occupy all suitable habitats within and outside the project areas.

The direct and indirect effects of this alternative would be less control and spread of weeds into suitable habitats within and adjacent to the project areas.

Cumulative Effects The cumulative effects would be less control of weed spread into adjacent public and private un-infested areas. Future control costs would increase if managers elected to initiate a control program at a later date. Chances of effective control would diminish as the infestation grew larger and more acres became infested with (more difficult to control) rhizomatous weeds. New invader species could become common and widespread. Weed colonization and infestation of all suitable habitats and acres at risk could occur within 50 years or less.

Page 60 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Summary of Direct and Indirect Weed Environmental Consequences Specific to Alternative 2 This alternative would best meet the Purpose and Need in Chapter 1 and would be the most effective in controlling weeds in the project areas. Anticipated levels of weed control are as described in the “Past Monitoring Results” section above.

Access to weed infestations is a major challenge in the project areas. Using a combination of aerial and ground-based treatments on additional project areas would allow weed control on the more infested acres in the project areas.

Additional large and small infestations could be contained or suppressed and smaller infestations could be eradicated. Topography, slope, and remoteness of each infestation would be less of a treatment limitation. All control measures would require follow up treatments after the initial treatment.

For example, where herbicides were used, large infestations of common weeds such as spotted knapweed would be 95 percent controlled as was achieved on the Mormon pilot project area. Over large areas, 94 percent of the weeds would be controlled in the first year post-spray and 99 percent in the second year post-spray (compared to the pre-spray weed abundance) as has been achieved in areas treated under the 2001 Lolo NF Big Game Winter Range and Burned Area Weed Management EIS. Rhizomatous weeds such as St. Johnswort which was treated on the Prospect Creek Winter Ranges would be contained or suppressed with over 90 percent control. This would protect adjacent un-infested public and private lands.

Where grazing was used, sheep and goats could reduce standing weed biomass and seed production over time.

A positive effect has been observed on some sites on the Lolo NF when herbicide control is combined with biological control. On Strawberry Ridge, for example, in the Rattlesnake National Recreation Area on the Lolo NF, leafy spurge flea beetles were released in combination with herbicide application. Leafy spurge control was better one year after treatment than with either control method alone (U.S. Department of Agriculture-FS 1998). Under this alternative biological management would be complemented more by herbicide applications than under Alternative 1. Aerial herbicide application would allow managers to utilize this synergistic efficiency over a larger number of infested acres. Discovery, screening and release of new biological management agents would increase this effect and could lead fewer weed treatments in the future.

Direct and Indirect Effects Direct effects would be a reduction in the infested acreage in the project areas. Small infestations of new invaders could be eradicated as they are detected. Indirect effects would be a reduction in spread onto adjacent public and private lands. Weed control programs on adjacent non National Forest System lands with active control programs would be more effective since we would be “treating both sides of the fence”. The

Final Integrated Weed Management EIS Page 61 Chapter 3 Affected Environment and Environmental Consequences maximum number of “at risk” yet uninfested acres as described in Chapter 2 would be protected.

Cumulative Effects The cumulative effect of the proposed action on weeds would be fewer infested acres, reduced rate of spread and protection of more adjacent un-infested areas. Total weed seed production on the forest would also be reduced which in turn would reduce the rate of weed spread and seed availability.

Forest Plan Consistency Both alternatives are consistent with the Weed Management Direction in Amendment 11 of the Lolo NF Plan. Alternative 2 better meets the intent because it could potentially treat more acres.

Range

Existing Condition Livestock grazing allotments on the Lolo NF occur primarily on transitory sites where the timber overstory has been reduced by logging or fire and in constricted riparian areas across the Lolo NF. The Lolo NF permits only limited livestock grazing. There are 21 active allotments on the forest which provide approximately 2,100 AUM’s annually. All of the permitted livestock grazing on the Lolo NF is by cattle and horses. There are no permitted sheep or goat grazing allotments. Range sites which include permitted grazing allotments and other areas impacted by livestock use are one of the weed infested and at risk site types identified in Chapter 2.

Method of Analysis Range effects were analyzed by comparing the impacts of cattle or horse grazing to the disturbance types and seed vectors that encourage weeds. This analysis does not include effects of sheep or goat grazing designed to control weeds.

Environmental Consequences - Range

Direct and Indirect Environmental Consequences Common to Both Alternatives Livestock activities such as hoof action, removal of desirable and competitive vegetation and travel through weed infested areas where weed seed can attach to cattle or horses, put riparian and transitory ranges at high risk of weed invasion and/or spread. This risk may increase when grazing occurs in areas with favorable site characteristics such as open canopy (increased sunlight) and limited vegetation cover for weed invasion. Excessive livestock grazing (in excess of proper use standards) under both alternatives increase the risk of encouraging weeds.

Page 62 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

The rates of herbicide use under both alternatives are below toxic levels for livestock. Herbicide labels include livestock management provisions for pastures treated with herbicides, so close coordination and information sharing with permittees may be necessary.

Direct and Indirect Environmental Consequences Specific to Alternative 1 Under Alternative 1, weeds would spread at a more rapid rate on grazing allotments because weed treatments would be limited to sites with existing decisions. Newly discovered or known infestations without a decision would be allowed to spread and expand unchecked until additional NEPA analysis(s)were completed. This would reduce livestock and wildlife forage. As weeds increased unchecked, there would be a higher likelihood of permitted livestock transporting weed seed off site into un-infested areas.

Direct and Indirect Environmental Consequences Specific to Alternative 2 Under Alternative 2 more range sites could be treated. Livestock and wildlife forage would increase and be protected more than under Alternative 1. This would reduce the risk of livestock spreading both existing and new weed species into adjacent or nearby un-infested areas. This would have a localized direct beneficial impact on livestock forage and grazing, but only in very limited areas (grazing allotments).

Cumulative Effects The adverse cumulative effects would be greater under Alternative 1 and would include a continued reduction in available livestock and wildlife forage and increased potential for livestock to carry weed seed to other areas. Adverse cumulative effects would be less under Alternative 2 because weed infestations without existing weed management decisions that are included in this analysis could be treated and small new infestations could be controlled. Livestock (sheep or goats) could be used as a weed control tool under Alternative 2, however cattle and/or horse grazing would have to be adjusted downward so as to not exceed proper use limits.

The cumulative effects of Alternative 2 would be a long-term increase in livestock forage and reduced potential for livestock to spread weed seed to other areas.

Forest Plan Consistency Both alternatives are consistent with the Lolo NF Plan. Alternative 2 however, complies better with Amendment 11 of the Lolo NF Plan because it treats more acres and therefore better mitigates the weed spreading potential of livestock grazing.

Final Integrated Weed Management EIS Page 63 Chapter 3 Affected Environment and Environmental Consequences

Forest Vegetation

Existing Condition - Vegetation The Lolo National Forest includes over two million acres of diverse plant communities. These plant communities are the result of a wide range of environmental conditions and are influenced by topography with elevations ranging from 2,400 feet to over 9,000 feet above sea level and slopes on all aspects ranging from flat to steep slopes exceeding 45 degrees. This topography and the influence of the continental and maritime climates prevalent over the Lolo NF produce very diverse growing conditions for plants. Over 1,500 plant species including 250 exotic non-native plant species grow within or near the Lolo National Forest (Lackschewitz, 1991). Eighteen native conifers (including two hybrids) and seven native hardwood (including one hybrid) tree species occur within the Lolo NF boundary.

Forest plant communities are commonly grouped and classified by their dominance type, a grouping by dominant plant. Dominance types identify general species groupings that are currently occupying various sites of the Lolo NF. They are described by mapped polygons of similar species and density. Dominant tree cover is derived from satellite imagery and rendered into a mapped classification by a project referred to as Northern Region Vegetation Mapping Project, R1-VMP. (Brewer, 2004)

Additionally, a plant communities’ growing environment is commonly classified by Pfister and others (1977) into habitat types. Potential vegetation types are groupings of those habitat types. Potential vegetation types give a good indication of what species and where these various species will grow. Their grouping classification is listed by Jones (2004). Potential vegetation groups are very useful in predicting the types of vegetation that will occupy a particular site especially after disturbance. These groupings relate only to potential vegetation and do not directly determine the specific species that are growing on a particular site at the current time. Potential vegetation type groups as listed by Jones (2004) are mapped for the Lolo NF through computer modeling of the environmental conditions of habitat types.

Both these classifications, dominance type and potential natural vegetation type, have been mapped and summarized for the Lolo National Forest. From these mapped classifications over 90 percent of the Forest is tree dominated (see Table 3-15, Forest Dominance Types) and 98 percent of the Lolo NF has the potential natural vegetation to be dominated by conifers (see Table 3-14, Potential Vegetation Type Groups on Lolo NF). Currently, only about 2 percent of the Lolo NF is non-forest comprised of sparse tree cover, rock, alpine tundra, water or warm, dry grass/shrub communities.

Risk of invasion by exotic weeds increases in plant communities that are warm, dry and in other communities affected by disturbances such as insects, disease, fire or management activities that create open forest canopy. Other exotic species can invade wet areas such as deciduous and coniferous riparian areas when disturbed.

The following Potential Vegetation Type Groups are derived from a much broader habitat type grouping listed by Jones (2004).

Warm-Dry Potential Vegetation Type Group - This group is characterized by multi-aged stands of ponderosa pine, Douglas fir, and occasional western larch with grass and shrub

Page 64 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences understory. Historically these stands were mostly single-storied due to a frequent fire regime (5-25 years). Today lacking fire or forest management, many of these types are characterized by multi-storied and often quite dense stands. Most of these sites occur at mid to lower elevations on southerly and westerly aspects. These sites are high risk for exotic plant invasion especially when current conditions are in an open forest canopy condition or when disturbances such as fire, disease, insects or management activities create open forest conditions. Once established, many exotics can occupy these sites as long as open forest conditions are maintained. This group comprises over 680,000 acres or 33 percent of the Lolo landscapes.

Warm-Moist Potential Vegetation Type Group - This group is characterized by multi-aged and even-aged stands of primarily shade-intolerant western larch, Douglas fir, western white pine, lodgepole pine, and Engelmann spruce. Stands also include, and occasionally are dominated by, shade-tolerant western red cedar, western hemlock, and grand fir. Most of these sites occur on lower slopes and valley bottoms, or mid-slopes on northerly and easterly aspects. These sites contain some high risk sites for weed invasion when fire, disease, insects or management activities create open forest conditions. Some weeds will not continue to grow within these as disturbance related forest openings close and shade understory weeds. Approximately 380,000 acres, 18 percent of the Lolo NF is in this group.

Cool-Moist Potential Vegetation Type Group - This group is intermediate between the warmer and more moist types generally below it in elevation and the colder and drier types above it. Conifer species diversity can be high, with shade intolerant species including western larch, Douglas fir, western white pine, lodgepole pine, and occasionally ponderosa pine. Shade tolerant conifer species found in this potential vegetation group include subalpine fir, grand fir, and Engelmann spruce. Most of these sites are low risk to weed invasion except for a short time when disturbance places them in early seral stages with open forest canopy. Approximately 385,000 acres, 18 percent of the Lolo NF is in this vegetation group.

Cool & Moderately Dry Potential Vegetation Type Group - This group is typified by even- aged and multi-aged stands of primarily shade intolerant lodgepole pine, Douglas fir, and western larch. Stands can be occasionally dominated by shade tolerant subalpine fir and Engelmann spruce. Most of these sites are low risk to weed invasion except for a short time when disturbance places them in early seral stages with open forest canopy. Approximately 510,000 acres, 25 percent of the Lolo NF is in this group.

Cold Potential Vegetation Type Group - This group is typified by even-aged and multi-aged stands of primarily shade intolerant whitebark pine and alpine larch. Stands can be dominated by shade tolerant subalpine fir and Engelmann spruce. This potential vegetation group occurs at high elevations on severe sites, extending above the tree line. There is currently low risk to these sites from weed invasion unless the canopy is removed, the soil disturbed and weed seed introduced. Over 80,000 acres, four percent of the Lolo NF is in this group.

Sparse Vegetation Potential Vegetation Type Group – This consists of areas of rock to poor soil dry sites. The lower elevation sites are somewhat susceptible to weed invasion where enough soil for plant growth is present, but at upper elevations this environment has low risk of weed invasion. Overall this type is at low weed invasion risk. This group is approximately 18,000 acres, less than one percent of the Lolo NF.

Final Integrated Weed Management EIS Page 65 Chapter 3 Affected Environment and Environmental Consequences

Riparian Potential Vegetation Type Group - This group is found at lower elevations, valley bottoms, and lower slope positions near streams, lakes, ponds, etc. This group occupies less than one percent of the Lolo. The deciduous riparian sites are high risk for weed invasion. Other riparian sites are high risk when in early seral stages of open forest canopy.

Dry Grasslands/Shrubs Potential Vegetation Type Group - This group includes communities of dry site forbs, grasses and shrubs most commonly on southerly aspects of the Lolo NF. Many are high risk to weed invasion. These sites are at the highest risk for weed invasion of any vegetation group on the Lolo NF. These dry sites occupy nearly 20,000 acres or one percent of the Lolo NF.

Table 3-14 Potential Vegetation Type Groups on the Lolo NF Percent Lolo N.F. PVT Acres of Lolo Weed Risk Rating Groups N.F. Warm-dry 680,000 33% H M except after disturbance during early seral Warm-moist 380,000 18% stages L except after disturbance during early seral Cool-moist 385,000 18% stages L except after disturbance during early seral Cool Mod-Dry 510,000 25% stages L except after disturbance during early seral Cold 80,000 4% stages L except after disturbance where soil exists, Sparse Vegetation 18,000 0.9% then high Riparian Deciduous 2,000 0.9% H Dry Grass/Shrubs 20,000 1.0% H L Mesic Shrubs 5,000 0.2% except during early seral stages U Water 3,000 0.1% High if aquatic weeds are introduced H =High susceptibility to invasions - Weed species invades the cover type successfully and becomes dominant or co-dominant even in the absence of intense or frequent disturbance.

L =Low susceptibility to invasion - Weed species does not establish because the cover type does not provide suitable habitat.

Page 66 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Table 3-15 Forest Dominance Types Forest Dominance Acres Percent Types Douglas-fir - Dry 260,000 12 Douglas-fir - Wet 160,000 8 Mixed Species - Dry 490,000 24 Mixed Species - Wet 350,000 17 Lodgepole Pine 260,000 12 Subalpine fir, Spruce, Mtn 310,000 15 Hemlock Grand fir, Red cedar, 90,000 4 Western Hemlock Other - Rock, Non-Forest, 160,000 8 burn, etc. TOTAL 2,080,000 100 Method of Analysis The Forest Vegetation affected environment describes vegetation b PVG and displays the susceptibility of each PVG to weed invasion. Forest and grasslands vegetative effects were analyzed by grouping potential vegetation groups within the same susceptibility rating and comparing how the alternatives would affect species compositions, vegetation density, structure and function in relation to the potential natural community.

Environmental Consequences - Vegetation

Direct and Indirect Environmental Consequences Specific to Alternative 1 Under Alternative 1 weeds will continue to expand into native plant communities not currently covered with existing weed control decisions (see Chapter 1, Table 1-2 Acres of herbicide treatment allowed under existing NEPA). While weeds have a limited effect on tree establishment and growth they can displace the native understory vegetation. Of most concern are the approximate 700,000 acres of the Lolo NF in the warm-dry and dry grass- shrub potential vegetation groups that have a high susceptibility to weed invasion (See Table 3-9 Estimated Acres at Risk on the Lolo NF). Within these groups weeds will continue to increase, even without disturbances such as fire, due to the open environments of these communities and the ability of weeds to produce large seed crops, disperse their seed varying distances, germinate under a wide variety of conditions, establish quickly, exhibit fast seedling growth, and out-compete native and desirable species for water and nutrients. Over time weeds would be expected to replace much of the native understory vegetation (grasses and forbs) leading to departures from the current structure and function in these types.

The vegetation groups with moderate risk to weeds include approximately 380,000 acres. Under Alternative 1 the presences of weeds in the native plant communities without existing weed management decisions is expected to increase more slowly than the high risk sites. As the tree canopies on these types are disturbed by insect, disease, weather or humans, weeds are expected to occupy the site until shaded out by the developing tree canopy decades later. Though weeds will be sporadically present, they are not expected to

Final Integrated Weed Management EIS Page 67 Chapter 3 Affected Environment and Environmental Consequences displace native vegetation to the degree to have an effect on the current structure and function.

The low risk to invasion group includes approximately 1,000,000 acres. Under Alternative 1 weeds are not expected to have an effect on the native plant communities in the low risk to invasion groups.

Weeds are known to invade and dominate native plant communities following fire (Asher, 1999). Dalmatian toadflax (Linaria dalmatica) in the Henry Creek drainage (Plains / Thompson Falls R.D.) has steadily increased within the burned areas since the 1994 Henry Peak fire. Alternative 1 would increase noxious weed establishment and displacement of native vegetation even in the low risk weed invasion sites following stand replacing or partial stand replacing fires.

Cumulative effects Loss of native plant communities will occur as weeds occupy and out compete native species. As weeds displace native forb and grass communities, effects on related ecosystem processes such as wildlife and fire ecology will occur. This effect would cumulatively increase over time due to lack of weed control on sites without existing weed management decisions.

Direct and Indirect Environmental Consequences Specific to Alternative 2 The direct effect of Alternative 2 would be a reduction in the density of existing weed populations within currently infested native plant communities and a reduced spread of weed seed from these populations to un-infested forest communities on the Lolo NF. These effects would be greatest in open canopy forest and following canopy opening and/or ground disturbing events.

Of the acres proposed for treatment approximately 54 percent are in the high susceptibility to weed invasion groups. The direct effects of herbicides on native plants are expected to be limited to species such as common yarrow and arrowleaf balsamroot, which commonly occur in the dry grass-shrub and warm-dry, high susceptibility to weed invasion groups. Many of the herbicides proposed for use in Alternative 2 (glyphosate, hexazinone, imazapyr, triclopyr) degrade quickly once they enter the environment and are not persistent and do not bioaccumulate (Tatum, 2004). Though population densities have been reduced in the short-term, native forbs have been shown to recover to pre-spray levels within 2 to 3 years (Rice and Toney, 1997). Woody species, such as ninebark, snowberry and ceanothus spp., which may also be found on these sites, are not likely to be affected.

Burned area treatments account for approximately 28 percent of the proposed acres in Alternative 2 and occur on all the weed susceptibility groups (high to low). Though actual treatment acres will be less than the maximum proposed by project area, the direct effect will be a reduction in the weed density within the native plant communities. Indirectly, treatment within burned areas will reduce the scale weeds spread to and the period of time weeds displace native vegetation prior to being shaded out by redeveloping overhead shrub and coniferous canopies.

Page 68 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Road and trail treatments account for approximately four percent of the proposed acres in Alternative 2 and also occur on all the weed susceptibility groups (high to low). Though the vegetation adjacent to roads and trails does not fully represent the native vegetative communities they pass through, weed treatments along roadways and trails directly reduces the dispersal of weed seed and subsequent opportunity for weed establishment in the adjacent native vegetative communities.

The remaining, approximately 16 percent, proposed treatment sites in Alternative 2 occur on all weed susceptibility groups. These areas are represented by sites such as power line corridors, campgrounds, electronic sites, etc. which have modified native and/or desirable nonnative species present. Similar to the treatment of roads and trails, treatment of these sites directly effect the surrounding native vegetation by reducing weed seed and therefore, weed establishment in the native vegetation communities.

The treatment of a maximum of 15,000 acres per year results in the return to species compositions, vegetation density, structure and function more reflective of potential natural communities than what currently exist. Even if there is treatment of fewer than 15,000 acres per year, managers will be able to prioritize treatments where gains in native vegetation density and composition may be maximized or where treatment of isolated and/or new weed infestation may protect larger areas of native vegetation.

Alternative 2 will allow response to small new weed infestation, including weeds which do not yet occur on the Lolo NF, in all vegetation groups. Early treatment will result in less native vegetation displacement and a reduce use of control measures, be it herbicide, biological or grazing methods.

Alternative 2 includes the use of sheep or goat grazing to control weeds. Proper grazing will reduce the nonnative vegetation density and thereby allow native plant species to expand their density. Ingestion by sheep and goats reduces weed seed viability and germination which in turn will minimize further spread of weeds into the adjacent native plant communities.

Cumulative Effects – Alternative 2 (Proposed Action) In addition to the native plant species that may be adversely affected by herbicides, other on-going activities such as timber harvest, recreational use, grazing, trail and road construction can also adversely affect native plants. In the absence of weed control, weeds will displace and adversely impact native plant communities. Though non-target native plants could be affected by actions in Alternative 2, there is greater potential for adverse effects to structure and function of native plant communities if weeds are not controlled. The preservation of native plant community diversity, structure and function in turn benefits all living organisms dependent on those ecosystems.

Forest Plan Consistency Alternative 1: Control of noxious weeds and other invasive exotics is consistent with the Lolo N.F. forest plan within those areas currently authorized for treatment. Lack of activities to control noxious weeds and other invasive exotics outside of authorized treatment areas in Alternative 1 would not be consistent with the Lolo N.F. forest plan.

Final Integrated Weed Management EIS Page 69 Chapter 3 Affected Environment and Environmental Consequences

Alternative 2: Control of noxious weeds and other invasive exotics is consistent with the Lolo N.F. forest plan.

Threatened, Endangered, and Sensitive Plants and Native Plant Communities

TES Plants and Native Plant Communities

Existing Conditions

Native Plant Communities The species composition of native plant communities of grassland and Shrubland habitat types found on the Lolo National Forest are described in qualitative and quantitative detail by Mueggler and Stewart (1980). (Pfister and others 1977) provide the herbaceous layer composition of forest habitat types. (Hansen and others 1995) characterized the herbaceous species composition of riparian plant communities found in western Montana.

Threatened, Endangered, and Sensitive Plants Threatened, endangered, and sensitive (TES) plant species known or suspected to occur on the Lolo NF: There are no federally-listed endangered or threatened plant species known to occur on the Lolo NF. However two threatened plant species are suspected to occur have habitat on the forest. These species are Howellia aquatilis and Silene spaldingii.

G1-G3 plant species: The Regional Forester’s Sensitive Species List (2004) contains 21 Forest Service sensitive plant species found on the Lolo NF. Of these species, three species are ranked as G1 (“globally at risk”) to G3 (“globally potentially at risk”). Grimmia brittoniae, a moss, is globally ranked as a G1 species, meaning it is considered at high risk to global extinction due to extremely limited populations or potentially declining population numbers and/or habitat (Montana Natural Heritage Program 2006). Waldsteinia idahoensis and Grindelia howellii, both forbs, are ranked as G3 species meaning they are potentially at risk because of limited range, populations and/or habitat.

A fourth globally rare species, Epipactis gigantea, an orchid, was found on the Lolo NF in 2005 (D. Lavelle, personal communication 2006). It is ranked as a G3G4. In addition, three sensitive species that are ranked as G1 to G3 are suspected to occur on the Lolo NF: Arabis fecunda (a forb), Botrychium paradoxum (a fern ally), and Phlox kelseyi var. missoulensis (a forb).

These nine species and their global/state ranks are shown in Table 3-16.

Table 3-16 G1-G3 Threatened or Sensitive Species Known or Suspected to Occur on the Lolo NF Scientific Name Common Name Known or Global State suspected Rank Rank Arabis fecunda Sapphire rockcress S G2 S2 Botrychium paradoxum Peculiar moonwort S G2 S2 Epipactis gigantea Giant helleborine K G3/G4 S2

Page 70 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Scientific Name Common Name Known or Global State suspected Rank Rank Grimmia brittoniae Cushion moss K G1 S1 Grindelia howellii Howell’s gumweed K G3 S2S3 Howellia aquatilis Water howellia S G2 S2 Phlox kelseyi var. Missoula phlox S G2 S2 missoulensis Silene spaldingii Spalding’s campion S G2 S1 Waldsteinia idahoensis Idaho barren strawberry K G3 S1 (“K” signifies species with known occurrences; “S” signifies species suspected to occur due to suitable habitat). In addition, all nine species are ranked as either “S1” or “S2” or S2S3, meaning they are vulnerable to extirpation within the State of Montana because of limited numbers of known populations and potentially declining population numbers and/or habitat (Montana Natural Heritage Program 2006).

G4-G5 Plant Species The Regional Forester’s List (2004) contains 17 G4 to G5 sensitive plant species with known occurrences on the Lolo NF. In addition, another 10 G4 to G5 plant species are suspected to occur on the forest. (Table 3-17)

G4 and G5 ranks mean the species are globally “apparently secure” or “secure” (Montana Natural Heritage Program 2006). However, these species have state rankings of either “S1” or “S2,” meaning they are considered to be at risk from extirpation in the state of Montana because of limited numbers of known populations and potentially declining population numbers and/or habitat.

Table 3-17 G4-G5 Sensitive Species Known or Suspected to Occur on the Lolo NF Scientific Name Common Name Known or Global State Suspected Rank Rank Adoxa moschatellina Musk-root K G5 S2 Allium acuminatum Tapertip onion K G5 S1 Amerorchis rotundifolia Round-leaved orchis S G5 S2S3 Arabis fecunda Sapphire rockcress S G2 S2 Athysanus pusillus Sandweed S G4 S1 Bidens beckii Beck water marigold K G4 S2 Botrychium paradoxum Peculiar moonwort S G2 S2 Brasenia schreberi Watershield K G5 S2 Carex chordorrhiza Creeping sedge S G5 S2 Carex rostrata Beaked sedge K G5 S1 Clarkia rhomboidea Common clarkia K G5 S2 Claytonia arenicola Sand springbeauty K G4 S1 Cypripedium fasciculatum Clustered lady’s slipper K G4 S2 Cypripedium passerinum Sparrow’s-egg lady’s slipper S G4G5 S2 Cypripedium parviflorum Small yellow lady’s slipper S G5 S2S3 Drosera anglica Great sundew K G5 S2 Dryopteris cristata Wood fern K G5 S2 Eupatorium occidentale Western boneset S G4 S2 Gentianopsis simplex Hiker’s gentian S G4 S1 Heterocodon rariflorum Western pearlflower S G5 S1

Final Integrated Weed Management EIS Page 71 Chapter 3 Affected Environment and Environmental Consequences

Scientific Name Common Name Known or Global State Suspected Rank Rank Kalmia polifolia Pale laurel K G5 S1 Meesia triquetra A moss S G5 S2 Mertensia bella Oregon bluebells K G4 S1 Orogenia fusiformis Taper-root orogenia K G5 S2 Potamogeton obtusifolius Blunt-leaved pondweed S G5 S2 Scheuchzeria palustris Pod grass K G5 S2 Scirpus subterminalis Water bulrush K G4G5 S2 Trifolium eriocephalum Woolly-head clover S G5 S2 Trifolium gymnocarpon Hollyleaf clover K G4 S2

Habitat for TES plant species on the Lolo NF Riparian areas contain some of the largest concentrations of TES plants on the Lolo NF. Riparian areas include green zones along intermittent and perennial creeks as well as those zones that may be seasonally moist such as seeps, sloughs, lowland depressions, and vernally moist potholes. Riparian areas may be affected by historical patterns of use as well as current use. They are susceptible to invasion by species such as common tansy, leafy spurge, spotted knapweed, and other weeds.

Roadsides and other habitats that contain known populations of Grindelia howellia (Grinhow), or provide suitable habitat for the species, are of special concern during weed treatments. Grinhow is a globally rare species of limited distribution in Montana. The species occurs frequently in disturbed habitats, and on the Lolo NF many populations have been found along roadside right-of-ways.

There are 11 presently or past known Grinhow sites (separated by at least one mile) on the Lolo NF. Grinhow was found at five of these sites that were monitored in 2006. Monitoring will be continued in future years on the known sites to assess the status of the species.

Other habitat of special concern on the Lolo NF includes native perennial grassland habitat such as occurs in woodland parks in valleys and open montane meadows. Many of these sites have been highly impacted by historical use and are also highly susceptible to invasive species such as spotted knapweed.

The 21 known and 13 suspected sensitive plant species designated on the Lolo NF, and the two federally listed threatened plant species that are suspected to occur are known from a variety of different habitats. The species have been categorized into broad habitat groups based on their affinity for both macro- and microhabitats (Lavelle 2006). Microhabitats are ecologically specialized sites that occupy limited areas on the landscape. These habitat groupings will be used in the Environmental Consequences section of this EIS for the effects analysis discussion.

Page 72 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

The habitat groupings for sensitive and threatened plants are shown below:

Aquatic and Vernal Pools Subalpine Forests or lower subalpine zones Bidens beckii Eupatorium occidentale Brasenia schreberi Mertensia bella Howellia aquatilis* Orogenia fusiformis Potamogeton obtusifolius* Waldsteinia idahoensis Scirpus subterminalis Dry Grasslands and Openings in Ponderosa Peatlands (fens) and Fen Margins Pine and Dry Douglas Fir Forests Amerochorchis rotundofolia* Allium acuminatum Carex chordorrhiza* Arabis fecunda* Carex rostrata Botrychium paradoxum* Cypripedium parviflorum Clarkia rhomboidea Cypripedium passerinum* Grimmia brittoniae Drosera angelica Grindelia howellii Epipactis gigantean Orogenia fusiformis Gentianopsis simplex* Phlox kelseyi v. missoulensis* Kalmia polifolia Silene spaldingii* Meesia triquetra* Trifolium eriocephalum* Scheuchzeria palustris Trifolium gymnocarpon Seasonally Wet Grass/Forb Community Other Wetlands (marshes/wet Botrychium paradoxum* meadows/seasonally wet meadows/lake Grindelia howellii margins) Orogenia fusiformis Amerochorchis rotundofolia* Trifolium eriocephalum* Carex chordorrhiza* Disturbed Areas (Including human-caused Cypripedium parviflorum disturbance) Cypripedium passerinum* Dryopteris cristata Clarkia rhomboidea Epipactis gigantean Grindelia howellii Gentianopsis simplex* Heterocodon rariflorum* Heterocodon rariflorum* Mertensia bella Kalmia polifolia Orogenia fusiformis Orogenia fusiformis Calcareous Grass and Forb Scheuchzeria palustris Trifolium eriocephalum* Arabis fecunda* Phlox kelseyi var. Missoulensis* Moist/Wet Coniferous Forests Fire Related (plant species that may be fire Adoxa moschatellina tolerant or fire dependent) Athysanus pusillus Amerochorchis rotundofolia* Botrychium paradoxum* Botrychium paradoxum* Cypripedium fasciculatum Claytonia arenicola Allium acuminatum Cypripedium parviflorum Trifolium eriocephalum* Cypripedium passerinum* Trifolium gymnocarpo Dryopteris cristata Epipactis gigantean Waldsteinia idahoensis

Final Integrated Weed Management EIS Page 73 Chapter 3 Affected Environment and Environmental Consequences

Riparian Associates (adjacent to stagnant Vernally Moist Cliffs or Mossy Talus Slopes surface water Adoxa moschatellina Amerochorchis rotundofolia* Athysanus pusillus* Cypripedium parviflorum Claytonia arenicola Cypripedium passerinum* Grimmia brittoniae Epipactis gigantean Heterocodon rariflorum* Heterocodon rariflorum* Dry Rocky Places Kalmia polifolia Waldsteinia idahoensis Allium acuminatum Eupatorium occidentale

(Species with an “*” are not currently known to occur on the Lolo NF, but are suspected to occur based on suitable habitat according to the Regional Forester’s Sensitive Species List, 2004)

Plant Species of Concern and Species of Interest In the ongoing proposed forest plan, the Lolo National Forest also recognizes eight plant Species of Concern (Table 3-18) and 36 Species of Interest (Table 3-19). "Species of Concern" includes plants that are at-risk or potentially at-risk due to rarity, restricted distribution, habitat loss, and/or other factors. These Species of Concern have high Global risk rankings (G1 through G3). Management actions may be necessary to prevent future listing under the Endangered Species Act. Species of Interest have high State risk ranking (usually S1 or S2) or are plants of cultural significance to Native Americans. There is some species overlap between the Sensitive Plants list and the lists of Species of Concern and Species of Interest.

Table 3-18 Species of Concern Known or Suspected to Occur on Lolo National Forest. Scientific Name Common Name Known or Global State Suspected Rank Rank Botrychium paradoxum Peculiar moonwort S G2 S2 Grimmia brittoniae A moss K G2 S2 Grindelia howellii Howell's gumweed K G3 S2S3 Howellia aquatilis Water howellia S G3 S2 Phlox kelseyi var. Missoula phlox S G2 S2 missoulensis) Silene spaldingii Spalding's campion S G2 S1 Synthyris canbyi Mission Mountain kittentails K G3 S3 Waldsteinia idahoensis Idaho barren strawberry K G3 S1

Table 3-19 Species of Interest Known or Suspected to Occur on Lolo National Forest. Known or Global State Scientific name Common name suspected rank Rank Adoxa moschatellina Musk-root K G5 S2 Allium acuminatum Tapertip onion K G5 S1 Allotropa virgata Candystick K G4 S3 Bidens beckii Beck water-marigold K G4 S2

Page 74 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Known or Global State Scientific name Common name suspected rank Rank Brasenia schreberi Watershield K G5 S2 Camassia quamash Small camas K G5 SNR Carex rostrata Beaked sedge K G5 S1 Carex scoparia Pointed broom sedge K G5 S2 Cirsium brevistylum Short-styled thistle K G4 S1S2 Clarkia rhomboidea Common clarkia K G5 S2 Claytonia arenicola Sand springbeauty K G4 S1 Cypripedium fasciculatum Clustered lady's-slipper K G4 S2 Cypripedium parviflorum Small yellow slipper S G5 S3 Drosera anglica English sundew K G5 S2 Dryopteris cristata Buckler fern K G5 S2 Epipactis gigantea Stream orchid K G3G4 S2 Western joepye-weed Eupatorium occidentale S G4 S2 (snakeroot) Gentianopsis simplex Hiker's gentian S G4 S1 Hierochloe odorata Sweetgrass K G3G5 SNR Kalmia polifolia Pale laurel K G5 S1 Lagophylla ramosissima Slender hareleaf K G5 S1 Lewisia rediviva Bitterroot K G5 SNR Mertensia bella Oregon bluebells K G4 S1 Nymphaea tetragona ssp. Leibergii (N. Leiberg's waterlily K G5 S1 leibergii) Orobanche fasciculata Clustered broomrape K G4 S4 Orogenia fusiformis Tapered-root orogenia K G5 S2 Potamogeton obtusifolius Blunt-leaved pondweed K G5 S2 Ribes cognatum Shinyleaf gooseberry K G4Q S1 Ribes triste Swamp red currant K G5 S1 Satureja douglasii Yerba buena K G4 S2 Scheuchzeria palustris Pod grass K G5 S2 Scirpus subterminalis Water bulrush K G4G5 S2 Sphagnum mendocinum A moss K G4 S1 Sphagnum riparium A moss K G5 S1 Trifolium gymnocarpon Hollyleaf clover K G4 S2 Method of Analysis The scientific literature was searched for research and review papers addressing the impact of weeds on native plant communities and TES plant species, and the effect of herbicide spraying on native plant communities and individual TES plant species. The relevant scientific publications were then scrutinized in detail for data and implications for native plant community diversity and maintenance or restoration of the potential natural community with and without use of herbicides for suppression of the targeted invasive plants.

Final Integrated Weed Management EIS Page 75 Chapter 3 Affected Environment and Environmental Consequences

Environmental Consequences - Weeds

Direct and Indirect Environmental Consequences Common to Both Alternatives The commonality between Alternatives 1 and 2 is the degree to which weeds continue to degrade native plant communities; the differences relate to the potential for increased active management to maintain and restore native vegetation. Alternative 1 would allow less treatment, and the ongoing spread of weeds would have greater potential to adversely affect native and rare plants, while Alternative 2 would allow more treatment and more prompt treatment of new invaders and recently discovered infestations. The negative effects of unchecked weed infestations and spread and the negative and beneficial effects of active weed management are discussed in the following sections that review the available scientific studies.

Weed Impacts on Native Plant Communities and Rare Plants Invasive plants are known to reduce native plant abundance, diversity, and richness in the grasslands of the western United States (Tyser and Key 1988; Belcher and Wilson 1989; Randall 1996; Kedzie-Webb and others 2001; Ortega and Pearson 2005; Sheley and Denny 2006). A specific example of competitive exclusion of native plants by an invasive exotic weed in western Montana is provided in Figure 3-1: as the percent cover of spotted knapweed increases the species richness and percent cover of native plants declines.. Spotted knapweed (Centaurea maculosa) is the most common invasive weed on the Lolo National Forest. Competition from invasive species is the second greatest danger, after habitat loss, to threatened and endangered species in the United States (Wilcove and others 1998; Stein and others 2000). Invasive species impact 57 percent of the imperiled native plant species in the United States. Several studies have looked specifically at the impact of invasive plants on rare native plant species and they concluded that the invasives were reducing the rare plant populations (Thompson and others 1987; Rosentreter 1994; Lesica and Shelly 1996) .

Page 76 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Figure 3-1 Relationships between cover of spotted knapweed and richness and cover of native species, overall and by functional group, at knapweed invasion sites in western Montana, as determined by mixed models analysis. Functional group information is given

AF=annual forbs, PF=perennial forbs, and PG=perennial grasses (Ortega and Pearson 2005).

Herbicide Application and Community/Non-target Plant Response

Reduction in Competitive Exclusion The herbicides being used for noxious weed control are rate-selective. Forb species exhibit varying degrees of susceptibility and tolerance at the lower rates of application, while high rates kill many broadleaf species and also can injure grasses. In addition to rate, phenological stage (timing/season of application), environmental factors controlling physiological activity, adjuvants, and application method all influence the spectrum of direct species kill and tolerance. More importantly, in natural plant communities the herbicide treatment alters the complex competitive interactions between plant species. The non-target species that are susceptible to herbicide application may recover from initial herbicide injury and even increase in abundance in the absence of intense pressure from the herbicide-targeted aggressive and dominating invasive weeds. The herbicide pulse is transitory, decaying over time (Wauchope and Leonard 1980; Watson and others 1989; Rice and others 1997; Vencill 2002), but competition from dominating invaders is perpetual and increasing over time. The theoretical basis for maintenance or increase in diversity after herbicide application is that by causing a short term reduction in overall plant biomass, the resultant prevention or slowing of competitive exclusion by dominant species favors niche occupation by other plant species (Grime 1979). In addition, persistent seed banks are typical of annual and perennial forbs in grasslands subjected to summer drought (Daubenmire 1968; Grime 1979; Rice 1989); and underground perennial regenerative meristematic tissues, below 12 to 22 inches in depth, receive little or no herbicide dose (Watson and others 1989; Rice and others 1997).

Final Integrated Weed Management EIS Page 77 Chapter 3 Affected Environment and Environmental Consequences

In an example of an extremely strong ecological benefit obtained by using a herbicide to reverse competitive exclusion by a dominant invader Getsinger and others (1997) treated a Eurasian watermilfoil infestation in the Pend Oreille River, Washington, with triclopyr. Non-target native plant biomass increased 500-1,000 percent by one year post-treatment, and remained significantly higher in the treated river cove two years after treatment. Native species diversity doubled following herbicide treatment, and the restoration of this robust native plant community delayed the re-establishment and dominance of Eurasian water milfoil for three growing seasons.

Herbicide application may decrease short-term species richness while simultaneously increasing or not effecting the calculation of Shannon-Weaver diversity, or other abundance weighted diversity indices, because the reduced abundance of the dominating target weed increases the evenness component of these mathematical indices.

Studies of Forest Understory in Reforestation Projects The largest block of longer-term studies on the response of natural non-target plant communities to herbicide application is from the use of herbicides for reforestation. High rates of relatively broad spectrum herbicides, usually glyphosate or hexazinone, are used for preparing a site for tree planting or for release of conifers from competition with other species. The target species are most often shrubs or grasses. There is usually a marked decline in overall herbaceous layer abundance and diversity shortly after herbicide application, and managers have been concerned about the longer term effect of these reforestation practices on grass, forb, and shrub diversity.

Glyphosate was aerially applied at 2 lb a.i./ac to early seral sub-boreal spruce sites in British Columbia to suppress shrub growth and release conifers (Sullivan and others 1998). Glyphosate is the least selective of the herbicides used on the Lolo National Forest. Shrub and herbaceous layer diversity was quantified within these two life-forms by species richness, Simpson’s index, and the Shannon-Weaver index before herbicide application and for five years after herbicide application. Species richness of shrubs was reduced in the first year after treatment and remained lower on treated sites throughout the five year period. The Simpson’s and Shannon-Weaver indices of shrub diversity, although depressed in the first year after herbicide application, were not significantly different over the second through fifth years post-spray. Herbicide treatment initially reduced crown volume index of herbaceous vegetation, but values quickly recovered to untreated levels by the second year after treatment. Herbaceous species diversity was not affected by herbicide treatment. In general, diversity of plant communities seemed to be maintained, and the investigators concluded that these treatment sites may not lower overall diversity of a forested landscape (Sullivan and others 1998).

Sullivan et al (1998) reviewed other plant diversity response studies of glyphosate use for conifer release. Overall there was no longer term reduction in herbaceous and shrub diversity. Temporary reductions in relative abundance were reported, and sometimes short- term diversity declines. Some other investigators suggested that the glyphosate treatments may be increasing herbaceous diversity in forest habitats, and although species turnover tended to be higher after herbicide application, no species were being extirpated from the treated sites.

Page 78 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Post-emergent treatment of competing herbaceous vegetation with glyphosate was used as part of site preparation for reforestation with green ash in Illinois old fields (Baer and Groninger 2004). The glyphosate treatment increased the diversity and richness of herbaceous vegetation reoccupying the site after the tree planting.

Ristau and Horsley (2002) tested the effects of glyphosate plus sulfometuron methyl on understory diversity at partially cut forest sites on the Allegheny National Forest. Half of each plot received herbicide and half remained untreated. At each site, richness and abundance measurements were conducted biannually for six years. Although abundance of all species was initially reduced, no species was consistently eliminated by herbicide across sites. Targeted species did remain less abundant on the sprayed vs. no-spray control plots after six years. Grasses and sedges had higher abundance on treated plots after four years. One forb species had significantly lower abundance after six years. Sprayed plot dominance, evenness and the Shannon diversity index did not differ from no-spray plots before or at any time after herbicide application. Pre-treatment species richness was not different; however immediately after herbicide there were five fewer species on average in sprayed plots. The species richness difference was gone after four years, and after six years, the sprayed plots had one more species than the no-spray controls. Results indicate that herbicide control of native invasives does not permanently reduce understory plant diversity Ristau and Horsley (2002).

Miller and others (1999) sampled all plants in planted pine test plots in the southeastern U.S. for the effects of reforestation site preparation using six herbicides on plant species richness and diversity. The herbicides evaluated in this study included picloram which is the herbicide most used on the Lolo National Forest. The herbicides had been applied at high labeled rates eleven years earlier. Species richness and diversity were the same in sprayed stands as in stands that had not been sprayed. This was true for all life-form subgroups including understory forbs and legumes. The Miller et al (1999) brief literature review of herbicide use for reforestation in the southeastern U.S. indicates that post-spray declines in herbaceous diversity are short term.

DiTomaso and others (1997) determined species richness and diversity in California wildfire areas that had been sprayed with hexazinone and planted with ponderosa pine. Hexazinone is a relatively broad spectrum herbicide that has soil residual activity for several years when applied at the high rates used for reforestation. The initial effect, measured at two years post-spray, was to reduce native forb and grass richness. The second measurement interval at eight years post-spray indicated full recovery of the native community, and the third measurement period, at twelve years post-spray, indicated that richness and diversity were highest in the areas that had been sprayed (DiTomaso and others 1997).

Changes in Richness and/or Diversity in Varied Habitats Sweet fennel (Foeniculum vulgare), a perennial Mediterranean forb, was present on Santa Cruz Island, California, for nearly 150 years before becoming invasive and replacing populations of native species (Erskine and Rejmanek 2001). An experiment evaluating the effects of fennel control on species diversity began with a fall prescribed burn and two subsequent spring herbicide treatments. The herbicide was triclopyr, a relatively wide spectrum broadleaf herbicide. The burn and spray regime significantly decreased sweet fennel cover and density. Canonical Correspondence Analysis has shown herbicide-induced

Final Integrated Weed Management EIS Page 79 Chapter 3 Affected Environment and Environmental Consequences succession in the treated plots becoming less similar to untreated fennel plots, and more similar to untreated grassland plots in which fennel has not invaded. Species diversity increased with the removal of fennel. The data indicate that the removal of a monospecific stand of an invasive species such as fennel can increase overall species diversity, but may also increase species diversity of other invasive species more significantly than native species (Erskine and Rejmanek 2001).

A grass specific herbicide, fluazifop-p-butyl (Fusilade), was used to reduce the strong dominance of exotic annual grasses on abandoned farmland in California (Allen and others 2005). The canopy cover of both native and exotic annual forbs increased because of herbicide application although the benefit was greatest for exotic stork’s-bill species (Erodium spp.). Native forb richness also increased relative to no-spray controls (Allen and others 2005).

Russian knapweed is known to displace other plant species. This invasive perennial has deep reproductive rhizomes and is difficult to suppress with herbicides. Laufenberg and others (2005) provided a peer reviewed paper on a community level study of herbicide applications to Russian knapweed in two severely degraded river bottom sites in central Montana. They used three herbicides (glyphosate, fosamine, or a mix of clopyralid plus 2,4-D) at three rates and three seasonal timings. Community level data was collected for two years after herbicide application. Two other reports discussed this study (Laufenberg and others 2002; Sheley and Laufenberg 2006). The herbicides effects on nonnative plant biomass, other than Russian knapweed, depended on site and application timing, but none of the spray treatments had less nonnative forb biomass than the no-spray control (Laufenberg and others 2005). Some herbicide treatments increased the non-target nonnative forb biomass component. Too few native forbs were present on these degraded sites to statistically analyze that community group. The researchers also suggested that herbicide treatments could not have increased the native forb abundance at these sites because of the pre-treatment paucity of native forbs (Laufenberg and others 2005). Density of each species was also recorded twice per year for two years post-spray (Sheley and Laufenberg 2006). Species richness and Simpson's Reciprocal Index, a diversity measurement, were calculated. At the last sample period, only the glyphosate (4.6 species m2) yielded greater total species richness over that of the no-spray control (3.5 species m2). At that last sample period, diversity after applying clopyralid plus 2, 4-D remained similar to that of the no-spray control, but glyphosate and fosamine increased total diversity. Non- native grasses and forbs accounted for the increases in richness and diversity (Sheley and Laufenberg 2006).

Chinese bush-clover (Lespedeza cuneata) is a well-established invasive plant of old fields and tall-grass prairie. In a southern Illinois study glyphosate application by brushing on Chinese bush-clover increased non-target abundance (native and non-native species) in the year after treatment although a measured four-fold increase in species richness was not statistically significant (p.>0.05) (Brandon and others 2004).

Spot spraying herbaceous vegetation in small plots in a French Mediterranean region old field with glyphosate increased species richness in the year after herbicide application, but in the second year the species richness of sprayed plots reverted to the same level as the no- spray controls (Lavorel and others 1994).

Page 80 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Carlson and Gorchov (2004) spot sprayed glyphosate with the results of decreasing invasive garlic mustard (Alliaria petiolata) density in both of two different forest stands in Ohio. The density of other species (mostly exotic winter annuals) in leaf while spraying in an old-growth stand was also reduced. Native plant responses were greater cover of spring ephemerals in one stand and increased reproduction of the native late-summer perennial, American lopseed (Phryma leptostachya), in another. These researchers concluded that glyphosate reduces garlic mustard without negatively impacting native species, and that some native species respond positively to a single-year reduction in this invasive biennial.

Nebraska wetlands with 95 percent or more purple loosestrife canopy cover and 5 percent or less of grass and other forb canopy cover were sprayed with sixteen different herbicide treatments (Knezevic and others 2004). Grass and other non-target forb canopy cover increased relative to the no-spray controls during the two years of post-spray measurement. Knezevic and others (2004) did not separate native from exotic non-target species.

Reever Morghan and others (2003) measured all plant species responses to clopyralid- treated yellow starthistle-dominated Central Valley California grasslands for two years post-spray. The entire data set was analyzed at family level, and at the species level for the most abundant forbs. Four family groups declined in canopy cover or frequency of occurrence, eleven were not significantly changed, and none increased significantly including the grass family. Ten species, including yellow starthistle, had a lower frequency of occurrence in the first year post-spray and six in the second year post-spray when recovery was evident for all species except one nonnative thistle. The authors concluded that “native species appeared to recover more quickly than the exotic species, so it is possible that applications of clopyralid every two or three years could control yellow starthistle without reducing diversity of natives” (Reever Morghan and others 2003).

Twenty-five of 30 species in a prespray sample were present two years after treatment of knapweed-infested Idaho fescue grassland with 1 pt/ac of picloram in the fall near Bozeman, Montana (Kedzie-Webb and others 2002). Four of the species that were not captured in the last sample period were native forbs. A regression model developed from the field data on plant density indicated that Shannon Weaver diversity (i.e. abundance weighted diversity) increased in the second year post-spray. The Kedzie-Webb et al 2002 study did not have spatial no-spray controls, just pre-spray data.

Quinclorac (Facet), imazapic (Plateau), and glyphosate plus 2,4-D were applied in the fall for two consecutive years to leafy spurge-infested tallgrass prairie in North Dakota (Kirby and others 2003). Changes in native forb cover were assessed as well as graminoid cover, and leafy spurge cover, density, and biomass. At 22 months after the first spray treatment, quinclorac and imazapic, but not glyphosate plus 2,4-D, reduced leafy spurge cover, density, and yield without causing deleterious effects to associated native forb cover and biomass (Kirby and others 2003).

Interactions with Introduced Perennial Grasses A large presence of perennial exotic rhizomatous grasses or certain robust exotic tufted grasses can prevent enhancement of native bunchgrasses and non-target forbs. A complex factorial experiment used combinations of two herbicides, two grass reseeding methods, three seeding rates of native grasses, and a cover crop in an attempt to promote native succession in a grassland site infested with spotted knapweed and sulfur cinquefoil (Sheley

Final Integrated Weed Management EIS Page 81 Chapter 3 Affected Environment and Environmental Consequences

2006). Exotic rhizomatous grasses were also a large component of this degraded site with low species richness near Kicking Horse Reservoir in Lake County, Montana. Responses of all species were measured. The reestablishment of seeded native bunchgrasses varied by treatment combination, but in the majority of combination treatments including herbicides, the initially limited native forb abundance declined and the exotic grass abundance increased. Species richness was similar for all treatment combinations in the third year post-spray (Sheley 2006).

Stringer (2003) reported that his thesis experiments in Glacier National Park showed that fall spot spray herbicide application of clopyralid at 0.5 lb/ac or 1.33 pt/ac reduced spotted knapweed canopy cover without significantly reducing existing native forbs. However, a repeat spot herbicide application in the second year increased exotic graminoid canopy cover (timothy [Phleum pretense] and rhizomatous Kentucky bluegrass [Poa pratensis]).

Roadside topsoil was salvaged and re-spread in Glacier National Park, then various native plant reseeding treatments were evaluated for sprayed and no-spray plots (Tyser and others 1998). Clopyralid (2/3 pt/ac) was sprayed in June for three consecutive years and plant establishment measured in the summer immediately after the third successive annual herbicide application. Herbicide treatments decreased mean canopy coverage of alien forbs (sprayed = 4.2 percent, no-spray = 23.4 percent) and the herbicide treatments reduced mean coverage of native forbs (sprayed = 3.9 percent, no-spray = 8.9 percent). Some native forbs were less abundant on sprayed plots while others established at similar levels with and without herbicide application. Herbicide application increased mean canopy coverage of native graminoids slightly (sprayed = 6.3 percent, no-spray = 4.0 percent), however coverage of rhizomatous introduced grasses consistently dominated the sprayed plots at 29.4 percent average cover, and were 20.2 percent on no-spray plots. Timothy (Phleum pratense) was at a relative canopy cover of 50 percent on the sprayed plots, and also dominated the no-spray plots although at a lower level of abundance (27 percent relative canopy cover). Kentucky bluegrass (Poa pratensis) was the second most dominant species in the sprayed plots.

Sheley and Denny (2006) sprayed five of the most commonly used broadleaf herbicides or combinations of herbicides (2,4-D amine, metsulfuron, picloram, clopyralid, 2,4-D plus clopyralid) at three rates (high, medium, low) on southeastern Montana, rangeland plant communities. However these sites were not infested with any invasive broadleaf weed so this experiment is not an explicit test of herbicide release of a native community from competition with a dominating invasive plant. However one of the two spray test sites was dominated by introduced rhizomatous Kentucky bluegrass (Poa pratensis). Canopy cover data was collected for all species the first year after herbicide application, then canopy cover, density, and biomass data were collected two years after the herbicide treatments. Picloram reduced forb cover relative to the no-spray controls at both sites, but forb cover for the other four herbicide treatments did not differ significantly from the no-spray controls. Total forb biomass was similar for all spray treatments and the no-spray controls at both sites. Total forb density was not affected by herbicide application at one site, but at the other site total forb density was reduced by picloram, metsulfuron, and clopyralid. Shannon-Weaver diversity was not affected by herbicide application at either site. Measured species richness was nine for the no-spray controls, and ranged from approximately eight for picloram to 11 for metsulfuron with no significant differences. The investigators also concluded that herbicide-induced increases in Kentucky bluegrass

Page 82 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences abundance at the site which was dominated by that introduced grass shifted the community composition of that site even further away from the potential natural community. The study was discontinued after two post-spray years (Sheley and Denny 2006).

Release of Natives in Weed-Infested Western Montana Grasslands Bedunah and Carpenter (1989) reported that there were few significant differences in native perennial forb densities between no-spray plots and plots treated with picloram or clopyralid two years after herbicide application for spotted knapweed control at two bunchgrass sites and one old field site in western Montana.

An extensive and detailed field experiment compared the effects of herbicide treatments (picloram, clopyralid and clopyralid + 2,4-D), each at two timings, at the recommended rates for control of spotted knapweed (Centaurea maculosa) on the structure and species diversity of native plant communities during an eight-year period in western Montana (Rice and others 1997). Floristic composition of replicated treatment plots was sampled before herbicide application and for three years after the initial herbicide applications at two grassland and two early seral forest sites. Following the third year post-spray measurements, half the treatment plots were randomly selected to be re-sprayed and community sampling was continued for two more years. Diversity was quantified as species richness and Shannon Diversity Index. Standing crops by growth form were estimated by double sampling at the conclusion of the experiment. Total standing crops of non-target forbs were largely unchanged by the herbicide treatments. Herbicide treatments had high efficacy on the target weed, shifting the plant communities back to a grass-dominated structure. Depressions in plant community diversity were small and transitory. In the third year after the initial applications, there were no significant diversity differences among treatments and some herbicide-treated plots had begun to surpass the untreated plots in community diversity measures. With most treatments, re-spraying three to four years after the initial applications did not reduce plant diversity compared to untreated levels. Late season applications, made after most herbicide-susceptible forbs had entered summer drought-induced dormancy, minimized the initial impacts on plant community diversity (Rice and others 1997).

Rice and Toney (1996) examined the frequency of occurrence of the 48 most common individual non-target forb species at three years after the initial herbicide applications used in the above Rice and others (1997) study. Statistical analyses of these types of data are constrained because most forbs have a low abundance and patchy distribution so the data sets have high variance. A nonparametric test using a high probability level (p≤0.20) was used. Nine non-target forbs had declined significantly in their measurements of frequency of occurrence and three had increased significantly relative to the no-spray control. Thirty- six non-target forb species did not have statistically significant differences in measurement (p>0.20); among these forbs the actual sample counts for 20 forb species were lower and 16 were higher relative to the no-spray controls. No non-target forb species were eliminated by herbicide applications to the test sites, although some of the most sensitive forbs were no longer found in the samples taken on all of the individual replicated test plots where they had been sampled prior to herbicide application. Total non-target forb canopy cover, which had initially declined due to herbicide application, had recovered to no-spray control plot levels by the third year after spraying. The investigators concluded that the

Final Integrated Weed Management EIS Page 83 Chapter 3 Affected Environment and Environmental Consequences herbicide treatments were very selective and had low overall impact on non-target forb species (Rice and Toney 1996).

Rice and Toney (1998) used the canopy cover data from the same experiment described above in Rice and others (1997) to compare the overall plant community similarity of herbicide sprayed grassland plots to the potential natural vegetation in absence of weed infestation. Historic data sets (Mitchell 1958; Mueggler and Stewart 1980) taken before extensive knapweed invasion were used as quantitative reference points to determine if herbicide application was making the plant communities of the current weed-infested sites more or less natural in regard to their potential native species composition. Herbicide treatment generally increased similarity to the potential natural community, relative to the no-spray controls (p= 0.024, Figure 3-2). In the no-spray controls similarity to the potential natural community continued to decline over the five-year period of measurement at one of the two sites analyzed. The site with lower pretreatment cover of exotic species and higher pretreatment composition of native species had the largest positive response to herbicide application in terms of increasing similarity to the potential natural vegetation. Although some treatments caused relative initial depressions in species richness one year after herbicide application, recovery of forbs was evident in the second year, and in the third year after herbicide application all six spray treatments had the same or a higher percent of pretreatment species present compared to the untreated control plots.

Page 84 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Figure 3-2 Average percent similarity between the experimental plots and reference stands defining potential natural communities (Rice & Toney 1998)

Final Integrated Weed Management EIS Page 85 Chapter 3 Affected Environment and Environmental Consequences

Rice and Harrington (2005) studied the effects of integrated picloram and prescribed fire treatments on four native grassland communities in western Montana. Prescribed fire alone, herbicide alone, and various timing combinations of prescribed burning and herbicide were applied to weed-infested bunchgrass communities. All plant species were measured both pre-treatment and post-treatment. The primary objective of this study was to measure target weed responses to herbicide or burning alone or in combinations, as well as the entire plant community-level effects of the various treatments. The paper did not report on individual non-target forb species, but did calculate the overall community response and floristic trajectories relative to the presumed potential natural community for each site. Prior to treatments the species composition at the four sites differed considerably from their respective Mueggler and Stewart habitat type definition plots. The herbicide treatments made the species composition more similar to the potential natural communities for the plots that still had remnant habitat type indicator bunchgrasses (Rice and Harrington 2005).

The Rice and Harrington (2005) sprayed plots that responded with progression towards the habitat type definitions were in several blocks at the Blue Mountain site and the North Hills site, and at the National Bison Range. The greatest improvements in similarity to the potential natural community were the result of the 1 pt/ac picloram treatments at Blue Mountain. The 1 qt/ac picloram applications at the National Bison Range and even the 2 qt/ac picloram applications in the North Hills had small increases in similarity to the potential natural vegetation in some plots. Blue Mountain plots that had co-dominant introduced rhizomatous grass species, a full block of plots in the North Hills which was dominated by introduced rhizomatous grasses, and the Henry Creek site which was lacking in indicator bunchgrass species went on floristic trajectories that will not lead to the native habitat type definition even with spraying. Adding a burn to the spray treatments did not increase the progression towards the potential natural communities, nor did burning alone foster that restoration goal. Burning at the Bison Range site was retrogressive relative to the habitat type definition in the first growing season after the treatments, but the first year retrogression attributed to the prescribed burning did not persist in subsequent years. Unfavorable precipitation and temperatures suppressed herbaceous growth during the period of this study. Total absolute canopy cover was much less than that measured in the 1971-1973 period when the Mueggler and Stewart reference plots were sampled. This drought-induced suppression of total growth may be restricting the magnitude of the progression to the habitat type definitions and confounding interpretation of the degree of floristic shifts among treatments (Rice and Harrington 2005).

Plant community responses to herbicide treatments for suppression of spotted knapweed were followed from 1996 through 2006 at the Sawmill Creek Research Natural Area in the Sapphire Range (Rice and Harrington 2006). Picloram (1 pt/ac) was ground broadcast to a series of permanent monitoring plots in the fall of 1996 and then re-sprayed by helicopter with clopyralid (2/3 pt/ac) in the fall of 2004. During the ten years following the initial herbicide treatment knapweed reestablishment never exceeded 50 percent of its pre-spray abundance, native bunchgrasses were greatly enhanced, and cheatgrass never exceeded ~1 percent canopy cover. Relative biomass production of non-target forbs was depressed slightly in the first year after the initial picloram treatment, began to recover to pre-spray proportions in the second year post-spray, and had greatly exceeded the pre-spray relative abundance by the seventh year post-spray (Figure 3-3). Then the site was aerially re- sprayed with clopyralid in the fall of 2004. Non-target forb relative biomass then declined

Page 86 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences in 2005 and recovered strongly in 2006, again to a non-target forb level greater than before the initial 1996 herbicide treatment (Rice and Harrington 2006).

Figure 3-3 Average percent biomass (of total biomass) for grasses, spotted knapweed, and other forbs for eight Sawmill RNA double sampled test plots sprayed (1 pt/ac picloram) in fall of 1996 and fall 2004 (2/3 pt/ac clopyralid); as well as absolute canopy cover.

100 All Grasses Knapweed Other Forbs 80 Cheatgrass ACC

60 spray

40 spray % of Biomass 20

1994 1996 1998 2000 2002 2004 2006 2008 Year

* Values for cheatgrass are calculated from percent absolute canopy cover estimates, not biomass estimates.

Final Integrated Weed Management EIS Page 87 Chapter 3 Affected Environment and Environmental Consequences

The same pattern of post-spray non-target forb recovery from low rate herbicide treatments (aerial application of 1 pt/ac or less of picloram or clopyralid) for control of spotted knapweed and sulfur cinquefoil was measured in a forest-wide multiple-site winter range spray monitoring study on the Lolo National Forest (Rice and Gauer 2007).

Figure 3-4 Average percent of lifeform total biomass after herbicide application Lolo National Forest winter range sites one time with a pint or less of herbicide (n = 3 to 4 sites depending on sample year) (Rice and Gauer 2007). Average Percent of Total Biomass for Sites Treated with a Pint or Less

100

40 Average % Average 20

Pre 1st yr post 2nd yr post 3rd yr post 4th yr post Annual Treatment schedule

Grasses/sedges Weeds Other Forbes Cheatgrass

Page 88 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Lolo National Forest winter range sites infested with rhizomatous noxious weeds (dalmatian toadflax, St. Johnswort) were helicopter sprayed with 2 pt/ac of picloram or 2 pt/ac of picloram plus 2 pt/ac of 2,4-D. The native plant composition of these sites, including the non-target forb component, was strongly degraded prior to herbicide application; and they had high pre-spray proportions of cheatgrass and/or introduced rhizomatous grasses (Rice and Gauer 2007). Relative abundance of non-target forbs was strongly suppressed by the high rate herbicide treatments, and non-target forb recovery at four years post-spray was limited to 45 percent of pre-spray levels.

Figure 3-5 Average percent of life-form total biomass after herbicide application Lolo National Forest winter range one time with a quart or more of herbicide (n = 2 to 4 sites depending on sample year) (Rice and Gauer 2006). Average Percent of Total Biomass for Multi-pint Treated Sites

Average % Average 20

pre 1st yr post 2nd yr post 3rd yr post 4th yr post Annual Treatment schedule

Grasses/sedges Weeds Other Forbes Cheatgrass

Final Integrated Weed Management EIS Page 89 Chapter 3 Affected Environment and Environmental Consequences

The Bitterroot National Forest helicopter sprayed a spotted knapweed dominated winter range site in the fall of 2004 with picloram at 1 pt/ac (Rice 2006). This Gibbons Pass site is an Idaho fescue/bluebunch wheatgrass habitat type (Mueggler and Stewart 1980). Pre- spray relative canopy cover of non-target forbs was 30 percent in 2004, declining to 20 percent in 2005, and then recovering to 30 percent in 2006. Species richness in the Gibbons Pass plot sample was 22 in 2004, then down to 16 in 2005, and increased to 19 in 2006. Relative biomass of non-target forbs increased considerably in both post-spray years compared to their pre-spray biomass. The investigator used graphic ordination analysis to compare how similar the pre-spray and post-spray communities were to the potential natural vegetation (Figure 3-6). In this graph each point represents a single plot or specific plot-year. The exact position of the plot relative to the two ordination axes is determined by the overall species composition and relative abundance of every species in that plot. Plots (points) that are close to each other have similar overall species composition; plots that are more widely separated in the graph have increasingly dissimilar species composition. The pre-spray community composition at Gibbons Pass in 2004 (G4) was strongly atypical of the Idaho fescue/western wheatgrass habitat type as defined by Mueggler and Stewart 1980. The Mueggler and Stewart plots are not labeled in Figure 3-6. The relative canopy cover floristic trajectory after herbicide application showed marked progression towards the potential natural community in the first year post-spray (G5), and definitive similarity to the potential natural community in the second year post-spray (G6). Herbicide spraying was a restorative treatment creating a more “natural” native plant community (Rice 2006).

Rice and Harrington (2003) examined the similarity of sprayed plots to the potential natural community by ordination graphics for ten separate plots scattered across the Sawmill Creek RNA. These ten plots represented three different habitat types. The ordination analysis covered the pre-spray community composition absolute canopy cover data (1995 or 1996) through six years (2002) after the initial fall 1996 treatment with 1 pt/ac of picloram. By the second year post-spray, 1998, the herbicide treatment had made nine of the ten plots more similar to the potential natural communities as defined by the composition of the Mueggler and Stewart (1980) habitat type definition plots. One sprayed plot actually matched its reference plot set after spraying. The initial progression toward the potential natural communities after one herbicide treatment was partially reversed from 1999 through 2002 in seven of the ten plots. Plots that were enclosed by fencing in the fall of 1996 to prevent elk overgrazing following the test herbicide application had less retrogression (Rice and Harrington 2003).

All three studies of low rate herbicide treatment of spotted knapweed infested sites in western Montana that quantitatively examined species composition similarity to the potential natural community indicated that herbicide application was restorative of the natural vegetation (Rice and Toney 1998; Rice and Harrington 2003; Rice and Gauer 2007).

Page 90 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Figure 3-6 Gibbons Pass plots in 2004 (G4), 2005 (G5), and 2006 (G6) compared to Mueggler & Stewart Idaho fescue/bluebunch wheatgrass habitat type definition plots (not labeled) (relative % canopy cover; NMS two dimensional ordination stress=15) (Rice and Gauer 2006).

Restoration of TES Plant Habitats with Herbicides Carlsen and others (2000) demonstrated that restoration of bunchgrass structure to intermediate abundance in California grassland plots previously dominated by exotic annual grasses creates a high habitat value for the potential establishment of the endangered native annual large-flowered fiddleneck (Amsinckia grandiflora). Nutlets of large-flower fiddleneck were sown in test plots to establish a new California grassland population (Pavlik and others 1993). Introduced annual grasses had significant negative effects on survivorship to reproduction, plant size, and nutlet production in the new large- flower fiddleneck population. Annual grass cover was effectively reduced by a grass-

Final Integrated Weed Management EIS Page 91 Chapter 3 Affected Environment and Environmental Consequences specific herbicide fluazifop-p-butyl (Fusilade). The herbicide treatment had no effect on large-flower fiddleneck mortality rates or survivorship to reproduction, but it significantly increased plant size and, therefore, nutlet output per plant and per plot.

Leafy spurge, which forms monotypic stands, is invading the habitat of the threatened western prairie fringed orchid (Platanthera praeclara). Fall-applied quinclorac (Facet) herbicide controls leafy spurge, but does not cause any injury to the Federally listed orchid (Erickson and others 2006). Fall timing is optimum for leafy spurge control with quinclorac and imazapic as the orchid generally has senesced by mid-September, making injury less likely. In this North Dakota grasslands study the density and frequency of occurrence of western prairie fringe orchid was not affected by imazapic (Plateau), also used for leafy spurge suppression, but the imazapic sprayed orchids were shorter and had fewer flowers than no-spray control orchids. The researchers concluded that quinclorac application could be a method maintaining the habitat of this threatened orchid. They advised that the weed control obtained by careful use of selected herbicides might ultimately save a threatened or endangered plant species (Erickson and others 2006).

Cumulative Effects of Both Alternatives on Native Plant Communities, TES Plant Species, Species of Interest, and Species of Concern Forest-wide the native plant communities will continue to degrade and individual native plant species (TES, Species of Concern and Interest) will continue to be at risk from weed invasions over the next several decades under both alternatives, but the rate of weed spread and the threat to native plants will be reduced under Alternative 2, which provides for increased integrated weed management of invasive plants and more prompt treatment of new weed species on the Lolo NF.

Direct and Indirect Consequences for Alternative 1 Under Alternative 1, potential impacts to native plant communities and individual native plant species would include a greater risk than under Alternative 2. Under Alternative 1, fewer acres of weeds in fewer locations could be treated throughout the forest. New infestations of weeds would have greater potential to become established and spread due to the longer time required to complete NEPA analysis and issue a Decision for newly discovered infestations. The Forest would not have the ability to treat as many new weed species or quickly eradicate small new infestations of weeds already found on the forest. A potentially greater impact to native communities and individual native plant species from weed introduction and spread at a higher rate would occur in all site types under Alternative 1.

Cumulative Effects Under Alternative 1, there would be continued longer term weed impacts to the integrity of native plant communities and individual native plant populations and/or their habitat from untreated infestations. Native communities and individual native plant species in lower- elevation dry grassland and open canopy ponderosa pine and Douglas-fir habitats would continue to be particularly vulnerable to weeds such as spotted knapweed, sulfur cinquefoil, dalmatian toadflax, and leafy spurge if these weeds were not controlled. Weed

Page 92 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences spread in isolated and remote locations without weed management decisions would also continue unchecked. Important big game winter range habitat that was not identified for treatment under previous decisions could not be treated and the weed-induced loss of native bunchgrass for winter elk forage would continue. New or isolated grassland sites would require more time and expense to conduct a NEPA analysis prior to a Decision and treatment. Weeds would continue to spread and increase in density in vulnerable habitats. The overall extent of native plants and intact native plant communities will decline more under Alternative 1 than under Alternative 2.

Direct and Indirect Consequences for Alternative 2 Alternative 2 would provide a wider menu of treatment sites and would allow managers to better prioritize where weed control is conducted each year. Aerial application costs ($15/ac) are much lower than ATV ($ 100/ac) or backpack ($245/ac) applications so more acres could be protected. Protection of general native plant communities, TES plant occurrence and habitats and plant communities containing Species of Concern or Interest would be considered when prioritizing each year’s treatment sites.

Under Alternative 2, more acres of native plant communities could be maintained or restored to similarity with the potential natural vegetation. Areas supporting TES plants and Species of Concern or Interest will have a higher level of protection from weed infestation. Alternative 2 would allow treating up to three times the weed infested acreage and would therefore better meet the Purpose and Need of this project.

The TES plants listed in the Existing Condition are grouped by various sets of higher-order habitat type affinities that were derived from the Lolo NF TES Plant Strategy (2006). The following TES plant habitat groups may be affected by the weed treatments proposed under Alternative 2: • Dry grasslands and openings in ponderosa pine and dry Douglas fir forests. • Ponderosa pine and Douglas fir forest/dry coniferous forest. • Dry rocky sites. • Disturbed sites (human-caused). • Disturbed sites (fire-related). • Seasonally wet grass/forb communities. • Calcareous grass/forb. • Riparian areas (adjacent to flowing or stagnant surface water). The largest number of TES plants on the Lolo NF occurs in riparian habitats. Riparian TES species include those found in the “Riparian Associates” habitat group (seven known or suspected TES plants), the “Other Wetlands” habitat group (12 known or suspected TES plants), and the “Peatlands (fens) and Fen Margins” habitat group (11 known or suspected TES plants).

Occurrence records for known populations of TES plant species, Species of Concern, and Species of Interest were reviewed with respect to potential weed treatment polygons. There are presently seven species that have known occurrences within or immediately adjacent to one or more of these potential treatment polygons.

Final Integrated Weed Management EIS Page 93 Chapter 3 Affected Environment and Environmental Consequences

These are musk-root (Adoxa moschatellina), sand springbeauty (Claytonia arenicola), clustered lady’s slipper (Cypripedium fasciculatum), Howell’s gumweed (Grindelia howellii), slender hareleaf (Lagophylla ramosissima), yerba buena (Satureja douglasii), and hollyleaf clover (Trifolium gymnocarpon).

Under Alternative 2, weed treatments in remote areas would be more effective than Alternative 1 since more sites are included in Alternative 2 and it includes treatment of new infestations. Alternative 2 would help reduce the threat of weeds to individual native plant species on all site types.

Alternative 2 also includes aerial spraying (as does Alternative 1) that which, while allowing more weed control on large and/or remote sites, also increases the amount of area that could be treated. More acreage could involve treatment by broadcast vs. spot application. Alternative 2 therefore has the potential to affect more populations and suitable habitat of TES species and Species of Concern or Interest unless mitigation measures are implemented to protect these native plant species.

Mitigation measures for Alternative 2 are included in Chapter 2 and reduce the potential impacts to native plant populations that could result from the increased acreage and types of weed treatments included in Alternative 2. If TES, Concern, or Interest plant populations are present in a proposed treatment area, a Forest Service Botanist would evaluate the population and mitigations would be used or designed to protect these native plant species.

Treatment timing may also be used to avoid injury to the TES, Concern or Interest plant species. Many TES, Concern or Interest plants on the Lolo NF are early season species and may be affected by early season herbicide applications. Late summer or fall herbicide application would usually result in avoidance of direct initial herbicide exposure for many Lolo NF TES, Concern or Interest plants except for warm season blooming members of the Asteraceae family. Herbicide application while native plant species are in later season dormancy should reduce or prevent injury from herbicides, although residual herbicide activity could still affect TES, Concern or Interest and other native plant species. Herbicide selection to reduce potential residual effects is a mitigation measure that would be used. Sheep grazing may have a greater long-term adverse impact on native forbs than herbicides. Seasonal vulnerability of TES, Concern or Interest and other native plant species would also be considered when grazing was considered for weed control.

Mechanical, biological, or cultural control methods may also be used to minimize impacts to the TES, Concern or Interest plant species. Digging or hand pulling may be part of the integrative approach proposed under Alternative 2: however these methods may also adversely impact TES, Concern or Interest plants. Grazing by goats or other herbivores may provide control to certain invasive species such as knapweed and leafy spurge while providing an alternative to herbicide application in sensitive areas. A Forest Service Botanist would be involved in evaluation of treatments where TES, Concern or Interest plant populations or habitat is present.

Page 94 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Statement of TES Plant Findings Under Alternative 2

Threatened and endangered plant species Spalding’s catchfly (Silene spaldingii) has potential habitat on the Plains Ranger District; however there are no known populations at this time on the Lolo NF. In addition, water howellia (Howellia aquatilis) has potential habitat on the Seeley Lake Ranger District, but there are no currently known populations. Previous surveys for these species on the Lolo NF have not detected any occupied habitat.

Suitable habitat for both species may occur within some of the proposed project areas. Areas with suitable habitat for these species will be field surveyed prior to the treatment. Treatments in or near any populations will be designed to minimize or eliminate impacts to these species.

In the long term, the treatments proposed under Alternative 2 should provide benefits for Spalding’s catchfly by reducing the threats from weeds invasion in its grassland habitat.

Based upon available information, the mitigation measures, and this evaluation, Alternative 2 will have no effect on the Threatened plant species discussed above.

Sensitive plant species Alternative 2 may impact individuals of some sensitive plants known to be present on the Lolo NF, but with the applied mitigations the proposed treatments will not likely contribute to a trend towards federal listing or cause a loss of viability to the populations or species. These are musk-root (Adoxa moschatellina), tapertip onion (Allium acuminatum), common clarkia (Clarkia rhomboidea), sand springbeauty (Claytonia arenicola), clustered lady’s slipper (Cypripedium fasciculatum) wood fern (Dryopteris cristata), giant helleborine (Epipactis gigantean), Howell’s gumweed (Grindelia howellii), Oregon bluebells (Mertensia bella) taper-root orogenia (Orogenia fusiformis), hollyleaf clover (Trifolium gymnocarpon), , and Idaho barren strawberry (Waldsteinia idahoensis). The Chapter 2 mitigations would be applied to protect sensitive plant species not known to occur on the Lolo NF at this time if their suitable habitat was subjected to weed invasion and was included in a proposed treatment area. In the long term, the treatments proposed under Alternative 2 should provide benefits to all sensitive plant species by reducing the threats of invasive plant species to their habitats.

Summary: The actions proposed under this EIS should not adversely affect TES plant species since an evaluation of known populations and suitable habitat would be conducted by a Forest Service botanist prior to treatments. Any populations found as a result of TES plant surveys would be evaluated by the Forest Botanist and site-specific provisions would be used to mitigate impacts and protect the species. Protection measures may vary by site and species. Alternative 2 will utilize weed control methods that are compatible with conservation of populations of TES plant species. Treating more infested acres would increase the potential risk that individuals of some TES plant species may not be detected during plant surveys and may be impacted by treatments. However, new infestations of invasive plant species on inaccessible areas such as winter range, burned areas, and remote roads would be reduced and this should protect or enhance habitat for TES plant species

Final Integrated Weed Management EIS Page 95 Chapter 3 Affected Environment and Environmental Consequences

and suitable habitat in these areas. Overall degradation of native plant communities would be slowed by utilization of integrated weed management methods.

Table 3-20 TES Plant Species Biological Evaluation Summary Sensitive Habitat Conclusion Justification Species Most known occurrences on Vernally moist places in the Adoxa Lolo NF near roads; one such mountains at the bottom of moschatellina May impact** occurrence adjacent to undisturbed, open rock slides Canada thistle; mitigation that have cold air drainage measures should protect. Habitat highly susceptible to Allium weeds. Only one known Dry, open forests and acuminatum May impact** population on Lolo NF, grasslands in the montane zone mitigation measures should protect. Habitat generally not Amerorchis Spruce forest around seeps or impacted by weeds; no known rotundifolia* along streams, often in soil No impact populations on Lolo NF, derived from limestone mitigation measures should protect. Open rocky, often eroding Suitable habitat for this Arabis fecunda* slopes developed from May impact** species is highly vulnerable to

calcareous parent material weeds. Suitable habitat for this Athysanus Vernally moist, shallow soil on species is highly vulnerable to pusillus* open, rocky shelves and along May impact** weeds, mitigation measures small drainages should protect. Still or slow-moving water of Habitat currently not Bidens beckii lakes, rivers, and sloughs in the No impact impacted by weeds; mitigation

valleys measures should protect. Grassland and meadow Botrychium Mature western red cedar habitat susceptible to weeds, paradoxum* stands or grasslands and May impact** mitigation measures should meadows protect. Brasenia Still or slow-moving water of Habitat currently not schreberi lakes, rivers, and sloughs in the No impact impacted by weeds; mitigation valleys measures should protect. Carex Habitat currently not chordorrhiza* Sphagnum fens No impact impacted by weeds; mitigation measures should protect. Wet, organic soils of fens in the Habitat currently not Carex rostrata montane zone, and floating peat No impact impacted by weeds; mitigation

mats. measures should protect. Dry, open ponderosa pine, Habitat highly susceptible to Clarkia Douglas-fir forests with May impact** weeds; mitigation measures gravelly soils in the montane rhomboidea should protect. zone Claytonia Mossy, forested, north-facing Some populations on Lolo NF arenicola talus slopes in the lower May impact** have weeds; mitigation montane zone measures should protect. Habitat for this species is Warm, dry mid-seral montane Cypripedium vulnerable to weeds, forest in the Douglas fasciculatum May impact** particularly if disturbed, fir/ninebark and grand mitigation measures should fir/ninebark habitat types protect.

Page 96 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Sensitive Habitat Conclusion Justification Species Fens, damp mossy woods, Cypripedium Habitat currently not seepage areas, and moist forest- parviflorum* No impact impacted by weeds; mitigation meadow ecotones in the valley measures should protect. to lower montane zones. Habitat generally not Cypripedium impacted by weeds; no known Moist coniferous forests, often passerinum* No impact** populations on Lolo NF, on calcareous substrates mitigation measures should protect. With Sphagnum moss in wet, Habitat currently not Drosera anglica organic soils of fens in the No impact impacted by weeds; mitigation

montane zone measures should protect. Moist to wet, often organic, Two known populations on Dryopteris soils at the forest margins of Lolo NF, one impacted by cristata May impact**t fens and swamps in the weeds, mitigation measures

montane zone should protect. Known population occurs on Epipactis Stream banks, lake margins, disturbed site. Habitat gigantea fens with springs and seeps, May impact** vulnerable to weeds, often near thermal waters. mitigation measures should protect. Habitat on Lolo NF currently not impacted by weeds; found Eupatorium Rocky outcrops and slopes in on old road on private land occidentale* the montane and lower No impact with weeds within Lolo NF; subalpine zones. mitigation measures should protect. Fens, meadows, and seeps; Gentianopsis Habitat currently not usually in areas of crystalline simplex* No impact impacted by weeds; mitigation parent material, in the montane measures should protect. and subalpine zones Calcareous cliffs and large Grimmia boulders in low elevation Habitat not impacted by brittoniae montane zone. Habitat usually No impact weeds; mitigation measures red cedar and Douglas- should protect. fir/ninebark. Habitat highly impacted by weeds. Herbicide use may restore habitat but may impact growing plants. 11 Vernally-moist, lightly presently or past known sites Grindelia disturbed soil adjacent to ponds on Lolo NF. Grinhow was howellii May impact** and marshes as well as found at five of these sites

roadsides and grazed pastures monitored in 2006. Monitoring will continue to assess the status of the species. Mitigation measures will be used to protect. Vernally-moist grassland slopes, road cutbanks, and other Heterocodon Habitat often highly invaded disturbed sites in forested rariflorum* May impact** by weeds, mitigation measures habitats; also loose soil areas should protect. under rock ledges and talus slopes

Final Integrated Weed Management EIS Page 97 Chapter 3 Affected Environment and Environmental Consequences

Sensitive Habitat Conclusion Justification Species Habitat currently not Howellia Oxbow sloughs and vernal impacted by weeds; adjacent aquatilis glacial pothole ponds which No impact vegetation can be impacted by may dry by late summer. weeds; mitigation measures should protect. Sphagnum peatlands, including Habitat currently not Kalmia polifolia spruce forest margin and outer No impact impacted by weeds; mitigation lake margin, in the montane measures should protect. zone Habitat generally not Fens with high pH and calcium impacted by weeds. Meesia triquetra* No impact concentration Mitigation measures should protect. Wet, seepy, open or partially Habitat currently not shaded slopes in the upper impacted by weeds, but some Mertensia bella montane or lower subalpine May impact** populations occur in

zones and disturbed wet areas disturbed sites, mitigation in and adjacent to roads. measures should protect. Known populations on Lolo Orogenia Open slopes, ridges, and NF occur in disturbed habitat, fusiformis meadows from lower foothills May impact** mitigation measures should to mid-elevations. protect. Phlox kelseyi var Open, wind-swept slopes in the Habitat highly susceptible to missoulensis* lower foothills to mid- May impact** weeds, mitigation measures elevations. should protect. Potamogeton Shallow water of lakes, ponds, Habitat currently not obtusifolius and sloughs in the valley, No impact impacted by weeds, mitigation foothill, and montane zones. measures should protect. Scheuchzeria Wet, organic soil of fens in the Habitat currently not palustris valley and montane zones, No impact impacted by weeds; mitigation usually with sphagnum moss measures should protect. Open water and boggy margins Scirpus Habitat currently not of ponds, lakes, and sloughs in subterminalis No impact impacted by weeds; mitigation the valley, foothills, and measures should protect. montane zones Bluebunch wheatgrass-fescue Habitat highly susceptible to Silene spaldingii* grasslands in the valleys and May impact** weeds, mitigation measures

foothills. should protect. Trifolium Dry meadows, woods and Habitat highly susceptible to eriocephalum* margins in the foothill and May impact** weeds. lower montane zones Habitat highly susceptible to Trifolium Open woods and slopes, usually weeds. Known population on gymnocarpon in dry soil of sagebrush to May impact** Lolo NF has weeds present, ponderosa pine forests. mitigation measures should protect. Habitat susceptible to weeds. Waldsteinia Meadows and moist grasslands Known population on Lolo NF idahoensis along streams, wet seeps, and May impact** occurs on disturbed site, adjacent forests. mitigation measures should protect. * - species with suitable habitat but no known populations on the Lolo NF. May impact** - may impact individuals or habitat but will not likely contribute to a trend towards federal listing or cause a loss of viability to the population or species; if treatments are proposed, Chapter 2 mitigations will be applied

Page 98 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Species of Concern and Species of Interest Of the eight Species of Concern (Table 3-21) seven are analyzed as sensitive species. The eighth species, Mission Mountain kittentails (Synthyris canbyi), occurs in subalpine an alpine habitat that is at low risk for significant weed invasion and thus should not be negatively impacted by any proposed treatments.

Table 3-21 Species of Concern Biological Evaluation Summary Species of Habitat Conclusion Justification Concern Grassland and meadow Mature western red cedar Botrychium habitat susceptible to weeds, stands or grasslands and May impact** paradoxum* mitigation measures should meadows protect. Calcareous cliffs and large boulders in low elevation Habitat not impacted by Grimmia montane zone. Habitat usually No impact weeds; mitigation measures brittoniae red cedar and Douglas- should protect. fir/ninebark. Habitat highly impacted by weeds. Herbicide use may restore habitat but may impact growing plants. 11 Vernally-moist, lightly presently or past known sites disturbed soil adjacent to on Lolo NF. Grinhow was Grindelia howellii May impact** ponds and marshes as well as found at five of these sites roadsides and grazed pastures monitored in 2006. Monitoring will continue to assess the status of the species. Mitigation measures will be used to protect. Habitat currently not Oxbow sloughs and vernal impacted by weeds; adjacent Howellia glacial pothole ponds which No impact vegetation can be impacted aquatilis* may dry by late summer. by weeds; mitigation measures should protect. Open, wind-swept slopes in Habitat highly susceptible to Phlox kelseyi var. the lower foothills to mid- May impact** weeds, mitigation measures missoulensis* elevations. should protect. Habitat highly susceptible to Wheatgrass-fescue grasslands Silene spaldingii* May impact** weeds, mitigation measures in the valleys and foothills. should protect Open, rocky, usually Habitat currently not calcareous soil of talus slopes impacted by weeds; Synthyris canbyi No impact and windswept ridge tops in mitigation measures should the subalpine and alpine zones protect. Habitat susceptible to weeds. Meadows and moist grasslands Known population on Lolo NF Waldsteinia along streams, wet seeps, and May impact** occurs on disturbed site, idahoensis adjacent forests. mitigation measures should protect. * - species with suitable habitat but no known populations on the Lolo NF. May impact** - may impact individuals or habitat but will not likely contribute to a trend towards federal listing or cause a loss of viability to the population or species; if treatments are proposed, Chapter 2 mitigations will be applied.

Final Integrated Weed Management EIS Page 99 Chapter 3 Affected Environment and Environmental Consequences

Species of Interest There are 35 plant species of Interest on the Lolo NF (Table 3-19), some of which were analyzed as sensitive species.

Of the 35 Species of Interest, 17 are known to be present on the Lolo NF and may be impacted by the proposed treatments as their suitable habitat is at risk to weed invasion (Table 3-22). These are musk root (Adoxa moschatellina), tapertip onion (Allium acuminatum), small camas (Camassia quamash), short-styled thistle (Cirsium brevistylum), common clarkia (Clarkia rhomboidea),sand springbeauty (Claytonia arenicola), clustered lady's-slipper (Cypripedium fasciculatum), wood fern (Dryopteris cristata), giant hellebornine (Epipactis gigantea), slender hareleaf (Lagophylla ramosissima), bitterroot (Lewisia rediviva), Oregon bluebells (Mertensia bella) tapered-root orogenia (Orogenia fusiformis), shinyleaf gooseberry (Ribes cognatum), swamp red currant (Ribes triste), and hollyleaf clover (Trifolium gymnocarpon). Alternative 2 may impact individuals or habitat, but with the applied mitigations the proposed treatments will not likely contribute to population declines. In the long term, the treatments proposed under Alternative 2 should provide benefits to all plant Species of Interest by reducing the threats of the targeted weeds to their habitats.

Table 3-22 Species of Interest Biological Evaluation Summary Species of Habitat Conclusion Justification Interest Vernally moist places in the Most known occurrences mountains at the bottom of on Lolo NF near roads; one Adoxa moschatellina undisturbed, open rock May impact such occurrence adjacent to slides that have cold air Canada thistle; mitigation drainage measures should protect. Habitat highly susceptible to Dry, open forests and weeds. Only one known Allium acuminatum grasslands in the montane May impact** population on Lolo NF, zone mitigation measures should protect. Habitat currently not Alpine to subalpine forests, impacted by weeds; Allotropa virgata No impact primarily in lodgepole pine. mitigation measures should protect. Habitat currently not Still or slow-moving water impacted by weeds; Bidens beckii of lakes, rivers, and sloughs No impact mitigation measures should in the valleys protect. Habitat currently not Still or slow-moving water impacted by weeds; Brasenia schreberi of lakes, rivers, and sloughs No impact mitigation measures should in the valleys protect. Moist meadows with Habitat susceptible to seasonally high water Camassia quamash May impact** weeds, mitigation measures tables from valleys to lower should protect subalpine zone. Four known occurrences on Lolo NF; habitat currently Wet soil along rivers and Carex scoparia No impact not impacted by weeds; sloughs in valleys. mitigation measures should protect.

Page 100 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Species of Habitat Conclusion Justification Interest Habitat currently not Wet, organic soils of fens in impacted by weeds; Carex rostrata the montane zone, and No impact mitigation measures should floating peat mats protect. Meadows and disturbed Habitat susceptible to Cirsium brevistylum forests in the valley and May impact** weeds. montane zones Dry, open forest ponderosa Habitat highly susceptible to pine, Douglas-fir forests Clarkia rhomboidea May impact** weeds; mitigation measures with gravelly soils in the should protect. montane zone. Mossy, forested, north- Some populations on Lolo Claytonia arenicola facing talus slopes in the May impact** NF have weeds; mitigation lower montane zone measures should protect. Warm, dry mid-seral Habitat for this species is montane forest in the vulnerable to weeds, Cypripedium Douglas fir/ninebark and May impact** particularly if disturbed, fasciculatum grand fir/ninebark habitat mitigation measures should types protect. Fens, damp mossy woods, Habitat currently not seepage areas, and moist Cypripedium impacted by weeds; forest-meadow ecotones in No impact parviflorum* mitigation measures should the valley to lower montane protect. zones Habitat currently not With Sphagnum moss in impacted by weeds; Drosera anglica wet, organic soils of fens in No impact mitigation measures should the montane zone protect. Moist to wet, often organic Two known populations on soils at the forest margins Lolo NF; one impacted by Dryopteris cristata May impact** of fens and swamps in the weeds; mitigation measures montane zone should protect. Known population occurs Stream banks, lake on disturbed site. Habitat margins, fens with springs Epipactis gigantea May impact** vulnerable to weeds, and seeps, often near mitigation measures should thermal waters. protect. Habitat on Lolo NF currently not impacted by Rocky outcrops and slopes Eupatorium weeds; found on old road in the montane and lower No impact occidentale* on private land with weeds subalpine zones within Lolo NF; mitigation measures should protect. Fens, meadows, and seeps, Habitat currently not usually in areas of impacted by weeds; Gentianopsis simplex crystalline parent material, No impact mitigation measures should in the montane and protect. subalpine zones Habitat currently not Moist ground on shores, impacted by weeds; one meadows, low prairies, and known population on Lolo Hierochloe odorata No impact at the edges of woods, bogs NF with few weeds and marshes observed; mitigation measures should protect.

Final Integrated Weed Management EIS Page 101 Chapter 3 Affected Environment and Environmental Consequences

Species of Habitat Conclusion Justification Interest Sphagnum peatlands, Habitat currently not including spruce forest impacted by weeds; Kalmia polifolia margins and outer lake No impact mitigation measures should margins, in the montane protect. zone Open or disturbed places in Habitat highly susceptible to Lagophylla dry grasslands in the May impact** weeds, mitigation measures ramosissima valleys should protect. Habitat highly susceptible to Harsh exposed grassland Lewisia rediviva May impact** weeds, mitigation measures sites in foothills to 6,000’. should protect. Wet, seepy, open or Habitat currently not partially shaded slopes in impacted by weeds, but the upper montane or Mertensia bella May impact** some populations occur in lower subalpine zones and disturbed sites, mitigation disturbed wet areas in and measures should protect. adjacent to roads. Habitat currently not Nymphaea tetragona Quiet, fresh water of lakes impacted by weeds; ssp. leibergii and backwater sloughs in No impact mitigation measures should (N. leibergii) the valleys protect. Habitat susceptible to Orobanche Open slopes in foothills May impact** weeds; mitigation measures fasciculata and valleys should protect. Open slopes, ridges, and Habitat highly susceptible to meadows, from lower Orogenia fusiformis May impact** weeds, mitigation measures foothills to the mid- should protect. elevations. Shallow water of lakes, Habitat currently not Potamogeton ponds, and sloughs in the impacted by weeds; No impact obtusifolius valley, foothill, and mitigation measures should montane zones protect. Rocky and sandy shores, stony banks, talus slopes Some habitats when and outcrops and in disturbed are susceptible to Ribes cognatum May impact** clearings, moist woods, and weeds, mitigation measures thickets, generally at low should protect elevations Some habitats when Moist soil of forest disturbed are susceptible to Ribes triste openings in the montane to May impact** weeds, mitigation measures lower subalpine zones should protect. Habitat not generally Partial or deep shade of impacted by weeds; some moist forests in the populations on Lolo NF Satureja douglasii No impact montane zone, sometimes found on disturbed sites; in second growth mitigation measures should protect. Wet, organic soil of fens in Habitat currently not Scheuchzeria the valley and montane impacted by weeds; No impact palustris zones, usually with mitigation measures should Sphagnum moss protect. Open water and boggy Habitat currently not margins of ponds, lakes, impacted by weeds; Scirpus subterminalis No impact and sloughs in the valley, mitigation measures should foothill, and montane zones protect.

Page 102 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Species of Habitat Conclusion Justification Interest Habitat currently not Sphagnum impacted by weeds; Wetland meadows, fens No impact mendocinum mitigation measures should protect. Habitat currently not impacted by weeds; Sphagnum riparium Wetland meadows, fens No impact mitigation measures should protect. Habitat highly susceptible to Open woods and slopes, weeds. Known population Trifolium usually in dry soil of May impact** on Lolo NF has weeds gymnocarpon sagebrush to ponderosa present, mitigation pine forests. measures should protect. * - species with suitable habitat but no known populations on the Lolo NF. May impact** - may impact individuals or habitat but will not likely contribute to a trend towards federal listing or cause a loss of viability to the population or species; if treatments are proposed, Chapter 2 mitigations will be applied.

Cumulative Effects Alternative 2 would allow additional management of weed populations on otherwise untreated areas of the Lolo NF. Overall degradation of native plant communities would be slowed by more utilization of integrated weed management methods.

Treating more infested acres would increase the potential risk that individuals of some TES plant species, Species of Concern, or Species of Interest may not be detected during plant surveys and may be impacted by treatments. However, new infestations of invasive plant species on inaccessible areas such as winter range, burned areas, and remote roads would be reduced and this should protect or enhance habitat for TES plant species, Species of Concern and Interest and suitable habitat in these areas.

Regulatory Framework and Forest Plan Consistency Both alternatives are consistent with the Lolo NF Plan. Alternative 2 however complies with the Lolo NF Plan to a greater degree than Alternative 1 because it may treat more acres and allow better prioritization of weed treatments.

Fish

Existing Conditions - Fish Fisheries descriptions are based largely on the Section 7 Middle Clark Fork River Bull Trout Baseline report produced by the Lolo National Forest in 2000, site-specific field reviews, surveys, monitoring of past projects, and fish population surveys conducted by the Lolo NF and Montana Fish, Wildlife and Parks.

Final Integrated Weed Management EIS Page 103 Chapter 3 Affected Environment and Environmental Consequences

The fisheries affected environment addresses: • Current fish population status and threats and resiliency as related to potential chemical contamination • Riparian and associated instream woody debris and nutrient, and shade conditions • Stream sediment relative to weed impacts across the forest No other habitat or population variables have the potential to be significantly impacted by this proposal, and they would not be addressed.

Fish Populations: The Lolo NF contains about 3500 miles of fishable perennial streams, several of which are of national, regional, and local significance. Portions of the Blackfoot, Bitterroot, Upper Clark Fork, and Lower Flathead Rivers flow through the forest. The entire Middle Clark Fork River flows through the Lolo NF, with many important spawning and recruitment streams such as Fish Creek, Rattlesnake Creek, Trout and Cedar Creeks, and the St. Regis River draining mostly National Forest lands. The lower 35 miles of Rock Creek, a recognized blue ribbon destination trout stream and critical native fish stronghold, are almost entirely Lolo NF lands. The forest also contains numerous small, high-mountain lakes. These are mostly along the Idaho state line, in the Thompson River/Cabinet Mountains area, and in the Rattlesnake and Scapegoat Wildernesses. These lakes provide local fishing opportunities for small westslope cutthroat, rainbow, and brook trout. Five larger, low-elevation lakes are found on the Seeley Lake Ranger District. These lakes contain adfluvial populations of westslope cutthroat and bull trout. They support local fisheries, primarily for non-native kokanee, pike, and bass.

The dominant factors affecting fish populations across the forest are habitat quality and non-native fish species. Non-native species are discussed here, and habitat quality as it relates to factors potentially affected by herbicides is discussed below in the riparian and stream habitat and sediment sections. The only native trout on the LNF are westslope cutthroat and bull trout. Other native species include mountain whitefish, longnose and large scale sucker, slimy sculpins, northern pikeminnow, redside shiner, and dace.

Westslope cutthroat trout are common in nearly every stream on the forest. They also occupy most of the lakes. Populations of westslope cutthroat trout in tributary streams on the Lolo NF are generally stable and relatively robust. However, fluvial populations in the larger streams and rivers are not as healthy due to hybridization with rainbows and habitat degradation.

Bull trout are present in most of the larger stream systems across the forest. They do not occur in high mountain lakes, but are present in the low elevation Chain Lakes on the Clearwater River system. Populations of bull trout are variable, ranging from relatively strong in streams like Rock Creek, Rattlesnake Creek, and the North Fork Blackfoot River, to very weak in streams like Petty Creek and Ninemile Creek. Habitat quality – primarily deep pools, large woody debris, and water temperature – are the key factors affecting bull trout populations. Hybridization with brook trout is a minor issue.

Mountain whitefish, sucker, pikeminnow, sculpin, and dace populations on the forest appear to be relatively stable. However, little trend data is available for these species. For more information on fish species found on the Lolo NF, visit the Fisheries website at http://www.fs.fed.us/r1/lolo/resources-natural/fish.

Page 104 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

The most common non-native fish species on the Lolo NF are rainbow, brown, and brook trout. Rainbow and brown trout populations are widespread in all the major river systems listed above, including Rock Creek. They also exist in most of the larger tributary streams, and use these extensively for spawning and rearing. Rainbow and brown trout are not as prevalent in small, headwater streams. Brook trout populations are just the opposite – they generally are minor or non-existent in the larger rivers, but are very common in headwater streams.

Rainbow trout pose a significant risk to native westslope cutthroat populations through competition and hybridization where populations overlap in the larger spawning tributaries and river systems. Most of the westslope cutthroat populations in the larger rivers are hybridized with rainbow to some extent.

Brown trout do not hybridize with cutthroat or bull trout, but they are very competitive and can displace these species when habitat conditions are degraded. An example of this displacement is currently occurring in Rock Creek, where brown trout populations have displaced cutthroat in the middle reaches over the last 4-6 years. This situation is most likely related to warm water temperatures associated with the ongoing drought and whirling disease in the system.

Brook trout also pose a significant threat to westslope cutthroat and bull trout, primarily in the headwater streams where local populations exist. Brook trout are very adaptable, and do well in small, cold streams. But they can also do well where habitat conditions are slightly degraded, giving them a competitive advantage over more sensitive species like cutthroat and bull trout. This can result in a species compositional shift over time in slightly degraded streams. Brook trout can also interbreed with bull trout, causing hybridization and an eventual loss of the genetic purity of bull trout. This doesn’t appear to be happening to a significant extent at the current time.

In general, chemical contamination is not a significant factor influencing current fish populations on the forest. However, potential contamination from inadvertent entry of a herbicide to a stream or lake could impact populations. Westslope cutthroat populations are widespread and robust enough, in general, to recover quickly from this type of accident. Bull trout populations, on the other hand, could take much longer to recover due to the limited distribution and present population sizes.

Riparian and Stream Habitat: Riparian zones and wetlands are lands between terrestrial and aquatic systems where the water table is at or near the surface of the land. They are extremely important in supporting stream functions such as sediment filtration, flow moderation, and nutrient cycling, and are also sources of shade and large woody debris to the aquatic system.

Most riparian zones on the Lolo NF are relatively narrow and well shaded, dominated by conifer forests with shrub understories. In general, riparian vegetation is healthy, but valley bottoms often have roads that compromise some of the function of the riparian zone. This, in turn, affects instream woody debris, nutrient, and shade conditions. Past harvest in riparian zones, along with stream channel clearing of large woody debris in the 1970’s, also impacted the ability of these areas to provide shade, woody debris, and nutrient cycling to their potential. At the landscape scale, these past and ongoing perturbations result in chronic stresses to fish habitat across the forest.

Final Integrated Weed Management EIS Page 105 Chapter 3 Affected Environment and Environmental Consequences

One of the main sources of nutrients to small, forested headwater streams is from riparian shrub leaf fall. Because these streams are well shaded, with little sunlight penetration, there is typically very little internal food production. The input of nutrients from riparian vegetation is therefore critical to the overall food chain in these systems.

Sediment: Instream sediment is affected by the amount of runoff in a watershed and the amount of sediment delivery to the stream network. There is some potential for heavily infested weed sites to result in increased runoff through site conversion from a native vegetative community with higher amounts of ground cover and therefore higher sediment filtering capabilities. The main areas where this could occur are on dry, relatively open hillsides where knapweed takes over the site. In general, these areas are upslope of riparian zones that typically still contain healthy vegetation with the ability to trap and filter runoff. This issue is therefore assumed to be relatively insignificant from a fish population and fish habitat perspective.

Method of Analysis Effects on aquatic habitat, fish and threatened, endangered or sensitive fish were analyzed by comparing the alternatives to: • Past ground based and aerial herbicide applications on the Lolo NF on the same site types as proposed in this analysis. • Effects from the Mormon Ridge Big Game Winter Range Restoration Project on the Lolo NF. • Herbicide soil movement studies from western Montana and other parts of the northwestern United States. • Analysis and monitoring results from other National Forests in Montana. • The fishery impact mitigation buffers applied by the forest and environmental hazards section of the herbicide label. • The fishery mitigation measures in Lolo NF Plan Amendment 11 (USDA Forest Service 1991) and the Lolo NF 2001 Big Game WR and BA FEIS and ROD. • The risk assessment section for aquatic organisms in the Human Health and Ecological Risk Assessments for Herbicide Use in Regions 1,2,3,4 & 10. • Herbicide Ecological and Human Health Risk Assessment prepared by SERA for the USDA-FS (SERA 1997-2004) under section 4.4.1, Risk Characterization http://www.fs.fed.us/foresthealth/pesticide/ ). An interagency team of fish biologists, weed management specialists, hydrologists, an environmental toxicologist, and an environmental fate research specialist met in 2003- 2004 to finalize a standard modeling procedure to utilize throughout Region 1 for herbicide application. The group’s final recommendations (Level I Team 2004) are incorporated into this document (Appendix F) and will be utilized in determining maximum treatment acres and locations under the selected alternative on an annual basis. In calculating maximum treatment acres, the model uses 1/20th of the LC50 as a limit for maximum concentration of herbicide within a watershed. This measure provides a large margin of safety to prevent negative effects to fish and accounts for other unforeseen activities. Potential instream concentrations are also modeled at low flows, further limiting treatment acres and increasing the margin of safety. Finally, the LC50 values used are for 96-hour exposure periods – this is typically much longer than exposure in a stream

Page 106 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences environment would be expected, and therefore provides another precautionary buffer in the determination.

Consultation with the U.S. Fish and Wildlife Service will occur annually, based on the treatment acres proposed, however, remaining below the maximum acreages is anticipated to result in a Not Likely to Adversely Affect determination for bull trout.

Environmental Consequences - Fish

Direct and Indirect Environmental Consequences Specific to Alternative 1 The limited control of weeds under Alternative 1 could lead to increased soil erosion and runoff. This would likely occur in small areas, primarily in dry upslope habitat types that are not directly connected to stream channels. Therefore, the potential effect is relatively small. There would be no additional potential effects to fish populations from herbicides, since no additional herbicides other than those already reviewed through previous projects would be applied. Some weeds (such as common tansy, houndstongue, and Canada thistle) would continue to invade riparian areas, especially on terraces and slightly drier and disturbed sites within riparian zones. This would lead to an overall reduction in the health of riparian communities, and consequently a negative impact on nutrient production and shade to stream channels. The magnitude of this effect would be small, since most valley bottoms are relatively narrow, well shaded, and dominated by moist microhabitats.

Cumulative Effects – Alternative 1 There may be slight negative cumulative effects associated with no action if weeds continue to spread and erosion on dry sites increases. Fish populations and habitat quality may decline from increases in sediment that could be transported to stream channels, however, the potential magnitude of this impact is small. Similar small negative impacts could be seen from cumulative changes in riparian vegetation over time, but the extent would be minimal. No negative effects or risks from herbicide application would occur.

Direct and Indirect Environmental Consequences Specific to Alternative 2 Under this alternative, weeds could be treated with combinations of ground and aerial applied herbicides, biological agents (including sheep or goat grazing), mechanical and cultural treatments. The amount of acres treated would range from 5,000 to 6,000 (roughly the amount treated annually at present) to an annual maximum of 15,000 acres across the forest. New infestations not covered by existing decisions could be controlled while they are small.

The potential effects of this alternative relative to increases in sediment are minimal. There may be potential for slight reductions in sediment if weeds are reduced, but, as described in the effects of Alternative 1, the magnitude of this impact is very small because most sites are dry and not connected to stream channels. There is also a slight increase in the potential for sediment erosion associated with sheep or goat grazing. This potential negative impact is

Final Integrated Weed Management EIS Page 107 Chapter 3 Affected Environment and Environmental Consequences expected to be small on dry sites, but could be greater if these methods were used in riparian sites.

The effects on riparian vegetation (and the associated habitat factors of nutrient production, shade, and large woody debris recruitment) associated with this alternative are also relatively small. The potential negative effects are associated with mowing and grazing in riparian areas (as described above). There is also some potential for non-target riparian vegetation to be killed through herbicide application as is the case with Alternative 1. This impact is expected to be minimal in both extent and duration, based on past projects on the forest. The positive aspects of this alternative relative to riparian vegetation include healthier riparian communities, which translate into healthier stream channels, fish habitat, and fish populations. This mainly applies to the drier and more disturbed microsites within riparian zones, which comprise a relatively small percentage of riparian habitats across the forest.

The greatest potential effect of this alternative on fish populations and habitat is associated with the potential for herbicide to either directly or indirectly enter fish bearing waters. Potential effects to aquatic habitat and fisheries could result from toxic concentrations of herbicides reaching and sustaining a 96 hour concentration that can cause fish mortality. However, by staying below the treatment acreage limits determined through the Level 1 modeling process and following all mitigation measures outlined in Chapter 2, herbicide use should have no measurable direct effect on fish and aquatic resources.

The mitigation measures in Chapter 2 were designed to minimize the possibility of herbicides entering aquatic systems. Application of these measures on the Mormon Ridge pilot project, and subsequent Lolo NF drift monitoring indicates that these measures are effective (refer to Chapter 3 Mormon Ridge Field Drift Monitoring results).

In a report prepared for the U.S. Department of Agriculture-FS (SERA 2003c) under section 4.4.1, Risk Characterization, it states "[that] Longer term water concentrations associated with the normal application of picloram (anticipated to be the most commonly used herbicide) at an application rate of 1 lb (a.i.)/acre are likely to be in the range of 0.01 to 0.06 mg/L in areas with substantial rain fall or as the result of applications in which some initial incidental contaminations of water occurs. All of these concentrations are substantially below concentrations that have been shown to impact aquatic plants or animals. Even at the highest plausible application rate, however, no effects would be anticipated in aquatic animals…" (U.S. Department of Agriculture-FS 1992). It is important to note that the most common application rate on the Lolo NF is ¼ to ½ pound per acre, not 1 pound so effects would be commensurately less.

Most of the project areas are infiltration dominated sites, where little or no run off is expected. On infiltration-dominated sites, herbicides are less likely to move in rain generated run-off events. Mitigation measures are included to avoid the possibility of applications immediately prior to extreme rain events since such an occurrence has a higher probability to cause a run-off event.

In a literature review conducted by the Flathead National Forest (USDSA FS 2000), the only studies that showed short-term levels of herbicide higher than the calculated no observed effect levels were those where herbicides were applied 8 to 18 times higher than the

Page 108 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences prescribed rate. In one case, the herbicide was applied directly to an intermittent stream channel.

Direct effects could include mortality, or reduced growth or survival of organisms resulting from exposure to herbicides. Based on the analysis methods described above and the mitigation measures in Chapter 2, no direct or indirect effects to fish or other aquatic organisms are anticipated through implementation of this alternative.

There is however, always a potential risk associated with a spill of herbicide directly into a live water body. This risk is expected to be slightly higher with ground than aerial application due to the time of exposure being higher with ground equipment. The risk of any spill, however, is extremely low, based on experience and past projects on the forest, which have demonstrated that both aerial and ground based spraying can be accomplished effectively without accidental spills or inadvertent application to live water. The Lolo NF has a excellent safety record in that we have never had an accidental herbicide spill during either ground or aerial herbicide applications.

The following tables were adopted from the Helena National Forest Herbicide Use Biological Assessment. They describe some of the most commonly used herbicides proposed under this alternative and their toxicity to fish. The toxicity classification provides a relative ranking of the toxicity of each herbicide to aquatic species. Of the herbicides proposed, picloram and glyphosate (in the Roundup formulation) have the lowest LC50s (which equates to the highest toxicity) for salmonids (glyphosate in the Rodeo formulation is labeled for aquatics, and has a very low toxicity to salmonids). Mitigation measures in Chapter 2 would minimize the potential for adverse effects on aquatic resources.

Table 3-23 Ecotoxicological Risks for Aquatic Species. Toxicity Classification LC50(mg/L) Very Highly Toxic <0.1 Highly Toxic 0.1 – 1 Moderately Toxic >1 – 10 Slightly Toxic >10 – 100 Practically Non-toxic >100

Table 3-24 Characteristics of Herbicides Proposed for Use Relative to Salmonid Fishes Typical Active Ingredient Application Toxicity 96- Product Toxicity Species Lit. Cited for and Soil Half-life Rates lb ai/ac hour LC50 Name Classification Tested LC50 (Range in Days) (Lolo typ Rate (mg/L)4 if different) 2, 4-D (amine) Amine 4 64 Cutthroat Johnson & 1.0 lb Slightly Toxic (2-16) (Weedar) (Weedar) trout Finley 1980 Chlorsulfuron Practically Rainbow Telar 0.5 oz (1 oz) 250 USDOE 2000 (28-42) Non-toxic trout Clopyralid Practically Rainbow Transline 0.38 >100 USDOE 2000 (12-70) Non-toxic trout Johnson & Finley Rainbow Banvel 0.75 lb (1 lb) 28 Slightly Toxic 1980 & USGS Dicamba trout 2003 (3-90) Practically Rainbow Vanquish 0.75 (1 lb) >100 USDOE 2000 Non-toxic trout

Final Integrated Weed Management EIS Page 109 Chapter 3 Affected Environment and Environmental Consequences

Typical Active Ingredient Application Toxicity 96- Product Toxicity Species Lit. Cited for and Soil Half-life Rates lb ai/ac hour LC50 Name Classification Tested LC50 (Range in Days) (Lolo typ Rate (mg/L)4 if different) Moderately Rainbow Roundup 2 lb 1.3 USGS 2003 Glyphosate Toxic trout (47) Practically Rainbow Rodeo 2 lb >1000 Monsanto 2001 Non-toxic trout 0.25 oz Practically Rainbow Johnson & Finley Imazapic Plateau >100 (0.1 lb) Non-toxic trout 1980 Metsulfuron- 0.5 oz Practically Rainbow Escort >150 USDOE 2000 methyl (120) (0.75 oz) Non-toxic trout Picloram Tordon 0.25 lb Moderately Cutthroat Johnson & Finley 1.5-100 (20-277) 22K (0.5 lb) Toxic trout 1980 Triclopyr 10 oz Practically Rainbow Johnson & Finley (triethylamine salt) Garlon 3A 117 (1 lb) Non-toxic trout 1980 (46)

TES Fish

A biological evaluation/biological assessment of the selected alternative will be completed for the federally listed threatened bull trout, and the FS listed sensitive westslope cutthroat trout.

The following table format was developed by the Montana Level 1 Bull Trout Team to document project effects to key populations and habitat parameters for bull and westslope cutthroat trout (USDA 2000). It provides a comparison, by alternative, for this project. The effects are considered the same for both species since they are both coldwater salmonids that have similar habitat requirements. The major difference between the two species is the westslope cutthroat trout are spring spawners, and bull trout are fall spawners. This factor should not result in a differential effect to either species.

Table 3-25 Determination of Effects for Bull Trout and Westslope Cutthroat Trout Diagnostic/Pathways: Effects of the Action(s) (R,M,D*) Indicators Alternative 1 Alternative 2 Subpopulation Characteristics Subpopulation Size M M D (extremely low risk – only if Growth & Survival M a spill) Life History M M Diversity & Isolation Persistence & M M Genetic Integrity Water Quality Temperature M M Sediment D (extremely low risk) D (extremely low risk) D (extremely low risk – only if Chem. Contam/Nutrients M a spill) Habitat Access Physical Barriers M M

Page 110 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Diagnostic/Pathways: Effects of the Action(s) (R,M,D*) Indicators Alternative 1 Alternative 2 Habitat Elements Substrate Embeddedness D (extremely low risk) D (extremely low risk) Large Woody Debris M M Pool Frequency & Quality M M Large Pools M M Off-Channel Habitat M M Refugia M M Channel Conditions and Dynamics Wetted Width/Max M M Depth Ratio Streambank Condition M M Floodplain Connectivity M M Flow Hydrology Change In Peak/Base M M Flows Drainage Network M M Increase Watershed Conditions Road Density & Location M M Disturbance History M M Riparian Conservation M M Area Disturbance Regime M M Integration Of Species & M M Habitat Cond. *Restore (R), Maintain (M), Degrade (D)

Table 3-26 Biological Evaluation of Alternatives Alternative Bull trout Westslope cutthroat trout Alt. 1 NE NI Alt. 2 MA-NLAA* MIIH *Call based on the extremely low potential risk of an accident or spill. No Effect (NE); May Affect – Not Likely to Adversely Affect (MA-NLAA); May Affect – Likely to Adversely Affect (MA-LAA); No Impact (NI); May Impact Individuals, but will not lead toward listing or loss of viability to the population or species (MIIH); Will Impact Individuals, and may lead toward listing or loss of viability to the population or species (WIFV). Call base on the potential (extremely low) risk of a spill or accident.

Cumulative Effects – Alternative 2 Cumulative effects include past activities in riparian zones that have affected riparian vegetation, leading to weed establishment. Activities such as road construction, grazing, mining, and timber harvest have all led to the establishment and spread of weeds in riparian areas, but some relatively undisturbed riparian areas have also been infested. Other activities such as spraying herbicides on private lands within the same watersheds that would be sprayed under this alternative could theoretically lead to herbicide levels in streams that approach, or possibly exceed, lethal limits to fish and other aquatic organisms.

Final Integrated Weed Management EIS Page 111 Chapter 3 Affected Environment and Environmental Consequences

Since picloram (the most commonly used herbicide on the Lolo NF) degrades quickly when exposed to sunlight and because the chances of multiple different activities taking place the same day, in the same drainage, and before a peak rainfall event are unlikely, the potential for cumulative effects is minimal. Increasing downstream flows also exponentially increase the dilution factor, which would further detoxify any herbicide that might potentially reach a stream.

Many minor effects documented at lower levels of exposure are from continuous exposure such as those that occur in flow through studies in a laboratory. In natural systems, it is more realistic to expect short duration exposures due to a storm event. Therefore, the practical amount of herbicide expected to reach a stream, even in worst case scenarios, is likely to be well below the level expected to affect fish.

An accidental spill is one scenario in which larger volumes of herbicide could enter an aquatic system. In this scenario, the duration of the event would not be expected to threaten the long-term persistence of fish populations, but could affect individual fish. Unknown factors include potential effects on bull trout egg and fry development (no testing has been done specifically on bull trout). The effects of an accidental spill would also be influenced by whether such a scenario was in large volume flowing water (which would flush) or in smaller volume standing water such as a pond. Lolo NF spray trucks however use injectors, which further minimize the chance and potential volume of herbicide that could be released in the event of a spill.

Forest Plan Consistency and Applicable Laws and Regulations Forest Plan: The Lolo National Forest Plan (1986) established broad objectives for the conservation and protection of aquatic habitat and fish populations. In all projects, “land management practices shall be designed to have a minimum impact on the aquatic ecosystem, free from permanent or long-term unnatural imposed stress.” In addition, “all threatened and endangered species occurring on the Lolo NF would be managed for recovery to non-threatened status.

Inland Native Fish Strategy: The Inland Native Fish Strategy was signed in 1995 and was intended as interim direction to maintain options for inland native fish by reducing the risk of loss of populations and reducing potential negative impacts to aquatic habitat. This document is incorporated as part of the Lolo National Forest Plan. It sets specific standards and guidelines for management within Riparian Habitat Conservation Areas (RHCA’s).

Land-use Strategy for Implementation of the 1999 Memorandum of Understanding and Conservation Agreement for Westslope Cutthroat Trout in Montana: The Memorandum of Understanding and Conservation Agreement (MOUCA) for Westslope Cutthroat Trout in Montana includes as objectives: • Protecting all pure and slightly introgressed (90 percent or greater purity) westslope cutthroat trout populations; and • Ensuring the long-term persistence of westslope cutthroat within their native range. The Land-use Strategy for Implementation of the 1999 Memorandum of Understanding and Conservation Agreement for Westslope Cutthroat Trout in Montana (Strategy) for the MOUCA, adopted by the FS and BLM in 2002, further defines how the MOUCA would be

Page 112 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences implemented by federal land management agencies. For new activities, the Strategy stipulates that the FS would: • Provide watersheds supporting conservation populations of westslope cutthroat trout with the level of protection necessary to ensure their long-term persistence; • Defer any new federal land management action if it cannot be modified to prevent unacceptable aquatic/riparian habitat degradation; and • Maintain westslope cutthroat trout habitat at 90 percent of optimum habitat conditions. When this 90 percent of optimum condition criteria is not met, only activities resulting in habitat improvement are to be considered. The Strategy also states that FS Biological Evaluations (FSM 2670) prepared for new activities should, in most cases, conclude that there would be a beneficial effect or no effect to the westslope cutthroat trout population or its habitat.

Sensitive Species: Sensitive species are those animal species identified by a Regional Forester for which population viability is a concern as evidenced by a significant current or predicted downward trend in population numbers, density, or in habitat capability that would reduce a species’ existing distribution (FSM 2670.5.19). There are two fish species listed as sensitive for Region 1 – westslope cutthroat and bull trout.

Protection of sensitive species and their habitats is a response to the mandate of NFMA to maintain viable populations of all native and desired non-native vertebrate species (36 CFR 219.19). The sensitive species program is intended to be pro-active by identifying potentially vulnerable species and taking positive action to prevent declines that would result in listing under the Endangered Species Act.

As part of NEPA decision-making process, proposed Forest Service programs or activities are to be reviewed to determine how an action would affect any sensitive species (FSM 2670.32). The goal of the analysis should be to avoid or minimize impacts to sensitive species. If impacts cannot be avoided, the degree of potential adverse effects on the population or its habitat within the project area, and on the species as a whole, needs to be assessed.

Soils and Water

Existing Conditions

Soils Soils across the Lolo National forest are primarily developed from residuals of the Belt Series rocks and glacial materials. Belt Series residuals and glacial soils are both moderately fertile. Their fertility is enhanced in areas where a surface “cap” of weathered volcanic ash is present. Once the ash cap is denuded from either soil, the fertility is reduced. Surface layers of these soils often contain numerous rock fragments and are relatively stable (Sasich and Lamotte-Hagen 1989).

Soils considered less stable than those developed from Belt Series and glacial materials are those derived from intrusive granites, Glacial Lake Missoula sediments and shaley sediments. These more sensitive soil types make up approximately 10 percent of the soil on the Lolo NF.

Final Integrated Weed Management EIS Page 113 Chapter 3 Affected Environment and Environmental Consequences

Granitic soils are susceptible to accelerated erosion and require special management attention. Natural erosion rates are high; when combined with human activities soil stability decreases. There is evidence of dry creep and raveling of road cuts and disturbed slopes, and cuts exceeding 35 percent would not stabilize. Granitic soils on the Lolo NF are localized in the upper Lolo Creek area and the Sapphire Mountains.

Glacial Lake Missoula sediments and materials derived from shaley sediments are relatively unstable because of their typically fine texture. Roads developed in these materials require surface stabilization and good drainage to prevent mass failure and landslides.

The localized surface layer of volcanic ash is a critical consideration in the Lolo NF’s soil management program. The chemical properties and high water holding capacity greatly enhance the moderately fertile granitic and Belt Series soils. The erosion potential of the ash cap is high when disturbed.

The Lolo Land Systems Inventory (LSI) is an integrated soil survey of all national forest land within the forest boundary. Soil “mapping units” identify landform, parent material and climate/vegetation characteristics.

The detailed information included in the Land Systems Inventory allows interpretations to be drawn as to which potential treatment areas may have soil characteristics suggesting specific sensitivities. For example, Landforms “10”, “61”, “42”, “46” and “72” have characteristics drawing attention to a particular need to protect surface and ground water. Landform “10” denotes stream bottoms where there is low relief, porous subsoil, near surface water tables and close proximity to stream channels. Landforms “42” and “46” are associated with glaciated headwater basins or valley trains where the potential for seeps, bogs and near-surface water is high. Landform “61” is the descriptor of steep dissected stream breaklands. There are frequently numerous stream channels in this landform (up to 25 percent of the surface area). In Landform “72”, glacial moraines, pothole lakes and saturated soils are common features.

Water The streamside or “riparian areas” adjacent to the nearly 700 named creeks, streams, rivers and 96 lakes are unique watershed resources on the Lolo NF and are designated Management Areas 13 and 14 in the Lolo National Forest Plan. These Management Areas have specific riparian-related goals and objectives. Unless grazing, mining, or other intrusive activities have disturbed riparian systems, the spread of weeds has been slow and would continue to be so because of the vigorous competition provided by native species.

The Lolo NF receives an average of 42 inches of annual precipitation, half of which is released as stream flow yielding 3.5 million acre-feet of water per year. The Forest consists of approximately 10,000 miles of stream channels flowing to the Clark Fork River. Most drainages within the Forest are classified as B-1 by the Montana Department of Environmental Quality. This classification indicates that the primary beneficial use of these waters is for support of cold-water fish. The municipal supply watersheds for Missoula (Rattlesnake Creek) and Thompson Falls (Ashley Creek) are classified as A- closed (drinking water supply) above the cities’ water system intakes.

Page 114 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

The chemical water quality of the streams on the Lolo NF is generally excellent. Table 3-27 displays a water quality survey of an “average” stream on the Lolo NF (University of Oklahoma, 1972). These are physical attributes of natural streamflow that are primarily a function of a watershed’s geology. These attributes tend to remain constant over time.

Table 3-27 Water Characteristics Quality Parameter Mean Value PH 7.8-8.3 (range) Temperature 11.9° C (53.4° F) Dissolved Oxygen 10.2 mg/1 Total Dissolved Solids 79.8 mg/1 Alkalinity 67.3 mg/1 Nitrate 0.4 mg/1 Phosphate (Total) 0.126 mg/1 Phosphate (Ortho) 0.048 mg/1 Tannin and Lignin 0.239 mg/1 Since 1972, periodic monitoring of streams on the Lolo NF has shown consistent comparable values with some extreme values. Summer temperatures over 20° C on some streams, total dissolved solids over 400 mg/1 on streams within drainages subjected to underground mining and over 200 mg/1 occurring naturally in watersheds with soluble mineralized bedrock such as limestone.

Sediment is the water quality parameter most affected by forest land management. Any activity that disturbs the surface of the soil has the potential to produce sediment. Unlike the chemical constituents of water, sediment is naturally a highly variable parameter, both among watersheds and within the same watershed at different times. The unstable or sensitive soils previously noted have a higher risk of producing sediment when disturbed.

High and low flow rates for representative streams on the Lolo NF are shown in Table 3-28. The table shows average daily flow for the high and low flow months in cubic feet per second (cfs).

Table 3-28 Representative Stream Flows Low (mean daily) High (mean daily) Stream Month cfs Month cfs Clark Fork River (3 miles below Blackfoot January 1,500 June 9,876 mouth) Clark Fork River (6 miles below Flathead October 7,885 June 70,885 mouth) Blackfoot River January 548 June 5,280 (6 miles above mouth) Rattlesnake Creek November 17 June 374 (Missoula, Vine St. Bridge)

Final Integrated Weed Management EIS Page 115 Chapter 3 Affected Environment and Environmental Consequences

Low (mean daily) High (mean daily) Stream Month cfs Month cfs St. Regis River September 135 May 2,150 (3 miles west of St. Regis) Thompson River January 199 May 1,534 (1 mile above mouth) Rock creek January 191 June 1,954 (at mouth)

Method of Analysis The analysis methodology used to assess potential effects of proposed actions on soil and water resources is based on the Lolo NF Noxious Weed EIS (1991), Appendix C.

The soil and water effects analysis is based on actual on-the-ground weed project monitoring on the Lolo NF and established and successful procedures followed for assessments of noxious weed control programs on neighboring National Forests. Weed control activities, land types and mitigating measures documented in previous neighboring National Forest assessments are similar to those on the Lolo NF. Effects would be analyzed by considering: • Research results and other literature reports on individual herbicide characteristics and impacts for a range of soil and water considerations. • Studies evaluating potential for herbicide entry into surface and/or groundwater, via different routes (leaching, overland flow, direct application and drift). • Results of recent analyses conducted by other National Forests in Region 1. • Treatment methods proposed within alternatives. • Proximity of proposed treatments to streams, wetlands and other sensitive water bodies. • Mitigation measures for specific treatment methods. • Range and scope of proposed treatments. • SERA Risk Assessments. Environmental Consequences – Soil and Water

Soil Quality The effect of weed infestations on soil quality varies with the community type, weed species, soil, and ecological setting.

A review of information displayed by the Beaverhead-Deerlodge National Forest in their 2002 Final EIS for their noxious weed control program concludes that the introduction of weeds has the following effects on soil processes and soil quality: • On treated sites, where treatment has been effective, soil quality indicators such as structure, organic carbon levels, and baseline erosion rates appear to be normal and stable. • On weed-dominated sites, where treatment has not occurred or has not been effective, organic matter levels are low and structure in the surface may have been altered. Erosion rates appear to have increased in some cases. (FEIS p 3-42).

Page 116 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Weeds affect the structure of ecosystems by altering soil properties. Soil in areas dominated by weeds may have lower amounts of organic matter and available nitrogen than areas supporting native grasslands (Olson 1999). Organic matter can be affected in various ways. For example, spotted knapweed has a deep taproot, which tends to decompose more slowly than the fine roots of native grasses, reducing the annual input of organic matter into the soil (Olson 1999). Biologically active organic matter occurs within the top 1 to 4 inches of soil and may be more prone to loss even during minor runoff events.

Soil nutrient levels may be affected by the presence of weeds. For example, potassium, nitrogen, and phosphorous levels were 44 percent, 62 percent, and 88 percent lower on soil from a spotted knapweed-infested site than from adjacent soil with a perennial grass cover in a study conducted by Harvey and Nowierski (Olson 1999). Plants that reduce soil nutrient availability to very low levels have a competitive advantage over neighboring plants (Olson 1999).

Soil microorganisms can either benefit or be impacted by the presence of secondary compounds produced by some weedy species. Most microbial populations adapt to secondary compounds by increasing their populations, thereby increasing the rate of breakdown of secondary compounds (Olson 1999). Conversely, these secondary compounds may limit activity and growth of aerobic soil microbial populations, resulting in thick litter layers and slowed nutrient cycling (Olson 1999).

Cryptogrammic soil crust is a layer of lichens and mosses important for nitrogen fixation, soil stabilization and soil-moisture retention. Such favorable conditions for native grass seed establishment are necessary for optimal total biomass for bunchgrass sites. This important layer has been found to be almost completely eliminated, compared to native grassland sites, shortly after spotted knapweed is introduced (Anderson et al. 1982).

Soil moisture can also be altered by the presence of tap-rooted, weedy species. Tap-rooted forbs may reduce infiltration because they do not have the dense, fine root systems of grasses, which contribute organic matter and enhance soil structure (Olson 1999). Infested sites may also have more extreme temperature changes because of lower soil water content; poorer soil aggregation; and greater exposure of soil to direct sunlight (Olson 1999). Water has a high capacity to store heat, which helps to maintain a more uniform soil temperature. By reducing soil water content in surface soil, greater evaporation enhances rapid heating and cooling of near surface layers. This will increase runoff but lower infiltration, again reducing thermal conductivity and capacity of the soil to store heat, resulting in greater temperature extremes at the soil surface (Olson 1999). The weeds with deep taproots are active later into the summer than native cool-season grasses. This can affect soil water and nutrient availability the following growing season.

Difference between Alternative 1 and 2 Weed effects on soil quality would be less under Alternative 2 because of the potential to treat up to 15,000 acres/year and prioritize treatment sites using soil quality as a prioritization criteria.

Final Integrated Weed Management EIS Page 117 Chapter 3 Affected Environment and Environmental Consequences

Soil Erosion On slopes, surface runoff and sediment production may be increased when bunchgrasses are replaced by weeds (Lacey et al. 1989). Even modest erosion of the soil surface layer will cause disproportionate losses of biologically active, nutrient holding organic matter. Most of the soil organic content is concentrated in the top one to four inches (Woods 1989).

Hickenbottom (2000) studied the effects of road reclamation on runoff and sediment in O’Brien Creek (west of Missoula) on the Lolo National Forest. A site 80 percent infested with knapweed yielded nearly five times the amount of sediment as bunchgrass covered sites. Hickenbottom’s rain simulator was calibrated to reproduce the effect of a 20 minute thunderstorm of 100 year return period. Hickenbottom’s results extended to the 1,648 acres of knapweed infested winter range treated in O’Brien Creek in 2003 would indicate such a storm could have produced an additional 160 tons of sediment if the site had been left untreated.

Research reported by Lacey et. al. (1989) was conducted in site conditions comparable to those of the Lolo NF proposed weed treatment areas. Simulated rainfall amounts representative of about half the expected storm intensities for this climatic region were applied to 12 paired plots having similar soil surface characteristics. In each pair, vegetation on one plot was 90 percent bunchgrass while the other plot was 90 percent knapweed. The runoff and sediment produced by the simulated rainfall from the plots were captured. The differences between runoff volume and sediment yield from the bunchgrass and knapweed plots was statistically significant (P<0.10). The runoff volume on the knapweed plots averaged 1 1/2 times the volume on the bunchgrass plots. The sediment yield was nearly 3 times greater on the knapweed plots. Incremental measurements of the runoff during the simulated rainfall events revealed how the bunch grass sites protected the soil surface allowing for higher infiltration rates than on the knapweed dominated sites. When the vegetation was clipped from the sites and the test re-run, the paired plots did not differ in their infiltration rates, confirming that the difference was attributable to the vegetation.

Because the No Action Alternative (Alt No. 1) maintains the current level of weed treatment on the forest (5,000 to 6,000 acres per year) about one-third the maximum level of the Proposed Action, Alternative 2, the lower treatment level in turn, results in potential rates of soil erosion up to 3 times higher and correspondingly reduced rates of infiltration. The quantitative increase in sediment from knapweed sites over bunchgrass sites reported by Lacey et al. (1989) averaged 75 pounds per acre for a simulated rainfall event of 30 minutes. The total number of weed infested acres proposed for treatment across the Lolo NF is 78,443. The total sediment increase found by Lacey applied to 78,443 acres would total nearly 3,000 additional tons of sediment contributed to Forest streams without treatment. On a per year basis there would be up to 10,000 fewer acres treated under Alternative 1 as compared to Alternative 2 for a calculated 375 additional tons of sediment.

Differences between Alternative 1 and 2 Weed effects on soil erosion would be less under Alternative 2 because of the potential to treat up to 15,000 acres/year and prioritizes treatment sites using soil quality as a prioritization criteria.

Page 118 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Groundwater Most herbicide groundwater contamination results from “point sources”, such as spills, leaks, and at storage and handling facilities. Point sources are discrete, identifiable locations that discharge relatively high localized concentrations of herbicides. These contaminations can be avoided through proper handling of herbicides and regular equipment maintenance.

Non-point source groundwater contamination is relatively uncommon. It can occur when a mobile herbicide is applied in areas with a shallow water table. In such areas, the choice of an appropriately labeled herbicide or an alternative control strategy can prevent contamination of the groundwater. Although herbicides can move chemically unaltered, via soil macropores or preferential flow paths to groundwater, following herbicide label requirements can prevent negative soil and water effects from occurring. In addition to label requirements, mitigation measures in Chapter 2 include provisions to keep all herbicides from being applied on sites where herbicides could enter the water table or surface water. No reported over-treatment effects on soil or groundwater contamination have occurred since the implementation of the current noxious weed treatment program on the Lolo National Forest in 1991.

Picloram is the most commonly used herbicide on the Lolo NF and also the most persistent and one of the most mobile herbicides proposed for use in the Lolo NF program and, thus, has the greatest potential to move through the soil and into the ground water. By analyzing soil and water effects for this herbicide, if potential effects are acceptable, then by conservative extension, less persistent, less mobile herbicides would also be acceptable. It is anticipated however that aminopyralid, which is labeled for use to waters edge, will reduce or replace much of the picloram use on the Lolo NF. At two sites in western Montana, Watson et al. (1989) examined picloram soil mobility after weed treatment following the label restrictions. The first site was a flat, valley bottom at Fort Missoula in Missoula, Montana. The Fort Missoula site was located 1/3 mile from the Bitterroot River, had a loamy topsoil, and had groundwater 30 feet from the surface. The site was treated with picloram at 1 pint/acre. At Fort Missoula, the authors reported a 44 percent photo decay of the picloram after the first seven days due to high sunlight and low precipitation. Follow up soil samples indicated herbicide persistence of 36 percent three months later, 13 percent twelve months later, and 10 percent fifteen months later. No herbicide was detected in soil deeper than 20 inches.

The second study site was on the North Fork of Elk Creek, 35 miles east of Missoula, Montana. This site had a sandy loam soil (61 percent sand compared to 33 percent sand at the Fort Missoula site). Herbicide was applied along a roadside adjacent to 2.5 mi of Elk Creek. Application averaged 110 feet from the creek but in some spots the sprayed site was as close as three feet from the stream. The groundwater table varied from three to nine feet deep throughout the application area. The area was treated with picloram at 2 quarts/acre (four times higher than the rate applied at the Fort Missoula site and two-three times higher than proposed in this analysis).

At the Elk Creek site, the authors reported that photo decay during the first seven days was half that observed at Ft. Missoula. The difference was attributed to tree canopy cover and less available sunshine. After three months, 78 of the applied picloram was still persistent in

Final Integrated Weed Management EIS Page 119 Chapter 3 Affected Environment and Environmental Consequences the soil. Although soil sampling detected picloram at depths of 40 inches, there was no detection of the herbicide in the groundwater or the stream.

Differences between Alternative 1 and 2 While Alternative 2 would treat more acres and potentially increase the amount of picloram used, the above studies indicate little likelihood of picloram reaching ground water as used on the Lolo NF. The potential effects on ground water are further decreased by use of aminopyralid instead of picloram.

Surface Water On a third site located on the Lolo National Forest, herbicides were aerially applied to the Mormon Ridge winter range in 1997 and 1999. Mormon Ridge presented a difficult treatment scenario in that it is extremely steep, has rolling topography, considerable microclimate variability, and aerial application occurred upslope of Mormon Creek, a bull trout –spawning stream. Mormon Creek flows along the bottom of the roughly three miles by ½ to ¾ - mile wide treatment area.

Picloram was aerially applied on Mormon Ridge in 1997. Buffer zones and water quality were monitored and continuous automated water samples collected. Analysis of the water samples (conducted by the Montana Department of Public Health and Human Services Chemistry Lab) indicated no herbicide entered the stream to a detection level of 0.1 parts per billion (USFS 1997). The Maximum Contamination Level as set by the EPA for drinking water is 500 parts per billion (Dow AgroSciences 1999). No picloram was detected in Mormon Creek when tested at the level 5,000 times lower than the EPA Maximum Contamination Level. Drift cards were also placed along Mormon Creek to monitor drift. The cards indicated no detectable drift reached the Creek.

The Mormon Ridge pilot project area was also aerially treated with picloram three growing seasons after the initial application to control weeds that germinated from the soil seed bank after the herbicide decomposed. Drift cards used during this subsequent treatment did not detect picloram in the riparian area or creek.

The research by Lacey (1989) and Hickenbottom (2000) previously cited supports the assertion that weed infested slopes are more of a sediment source than natural grass and forb dominated slopes. Weed treatments that effectively reduce the density of weeds and promote natural vegetative recovery also restore surface protection and erosion control benefits provided by the natural vegetation.

Surface water contamination with herbicides may occur when herbicides are applied intentionally or accidentally to ditches, irrigation channels, or are carried away in runoff to surface waters (DiTomaso 1999).

Norris (1971) describes five ways by which herbicides may enter water: • direct application • drift • mobilization in ephemeral stream, channels • overland flow • leaching

Page 120 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Each route of entry has different characteristics that influence both the potential magnitude and duration of contamination. Aerial application of herbicide in or near aquatic zones without adequate buffer zones offers the greatest opportunity for introducing herbicides into the aquatic environment by either drift or direct application. Herbicides that are applied from ground-based equipment can enter streams by direct application or drift when applied in or near aquatic zones. According to Norris (1981), “Aerial applications do not offer any greater opportunity for chemical entry to water than any other type of application when made in areas away from aquatic zones.” Based on Norris (1981), ground and aerial application of herbicide may thus be considered concurrently with respect to aquatic resources.

Direct application to surface waters is the route most likely to introduce significant quantities of herbicide into surface waters. This route has the potential to produce the highest concentrations, and therefore the most pronounced effects. Drift from nearby spray areas is similar to direct application except that concentrations of chemicals in the water are lower and the probability of impacts is reduced (Norris 1981).

Mobilization in ephemeral stream channels is another mechanism by which herbicides may enter streams. Ephemeral stream channels are difficult to see from the air and may be sprayed inadvertently. Thus, this mechanism may be more of a problem during aerial applications because ground application provides greater opportunity for identifying and avoiding these areas. With subsequent sufficient precipitation, herbicide residue remaining in ephemeral stream channels is available for movement in runoff to connected perennial streams.

Overland flow occurs infrequently on most well vegetated forest and rangelands because soil infiltration capacity is usually greater than precipitation (Rothacher and Lopushinsky 1974 as cited by Norris 1981). Norris acknowledges that denuded and compacted soil may yield “surface runoff”.

Leaching of herbicides through the soil profile is the least likely route for herbicides to enter the aquatic environment (Norris 1981). The author states, “Herbicides disappear quickly from both the forest floor and soil.” Microbial decomposition and volatilization are the predominant breakdown processes. Picloram is also adsorbed by organic matter in the surface layers of soil (Norris 1970).

Norris et al. (1976) determined the persistence and leaching characteristics of picloram at several sites on power transmission line rights-of-way in Oregon and Washington. Both temperature and precipitation ranged from high to low among the study sites. An extensive monitoring effort for picloram in forest streams flowing across the power line right-of-way failed to show measurable levels of the herbicide. In several cases, intensive sampling was done with automatic equipment for periods in excess of six months after application (Norris et al. 1976).

Haas et al. (1971) treated study plots adjacent to a small stream with 1.1 kg/ha (1 lb/ac) of picloram and collected water samples at several locations downstream after each rainfall for 5 months after application (rates proposed in this analysis are ½ to ¼ of that amount). Picloram was detected at a concentration of 0.029 mg/l in stream samples only during the first significant runoff. Kenaga (1969) exposed several species of trout to 1,000 times this concentration of picloram for 96 hours with no mortality.

Final Integrated Weed Management EIS Page 121 Chapter 3 Affected Environment and Environmental Consequences

Norris and others (1983) suggest that a margin of safety calculated as the quotient of no- effect level/exposure level is a good index to the probability that the use of a specific herbicide will result in direct affects on fish. The larger the margin of safety, the less likely direct effects will occur. Norris calculated margins of safety (MOS) based on the maximum acute and short-term chronic exposures likely to occur in operational use of picloram.

Table 3-29 Margin of Safety on maximum Acute and short-term Chronic Exposure to Picloram Herbicide Acute MOS Short-term Chronic MOS Picloram 43 50 Norris further notes, “These margins of safety will be 5 to 10 times greater when streams are not included in areas to be treated, buffer strips are used, and full attention is given to the details of application to prevent drift and direct application to surface water.” (Norris et al. 1983).

Most herbicide groundwater contamination results from "point sources", such as spills, leaks, storage and handling facilities, improperly discarded containers, or rinses of equipment in loading and handling areas. Point sources are discrete, identifiable locations that discharge relatively high local concentrations of herbicides. These contaminations can be avoided through proper calibration, rinsing and cleaning of equipment.

At two sites in western Montana, Watson, Rice and Monnig (1989) examined picloram soil mobility after weed treatment following EPA label restrictions. The first site was a flat, valley bottom at Fort Missoula, Missoula, Montana. The Fort Missoula site was located 1/3 mile from the Bitterroot River, had a loamy topsoil and groundwater 30 feet from the surface. The site was treated with one pt/acre of Tordon 22K (picloram).

At Fort Missoula, the authors reported a 44 percent photodecay of the Tordon 22K after the first seven days due to high sunlight and low precipitation. Follow up soil samples indicated herbicide persistence of 36 percent three months later, 13 percent twelve months later, and 10 percent fifteen months later. No herbicide was detected in soil deeper than 20 inches.

The second study site was on the North Fork of Elk Creek, 35 miles east of Missoula, Montana. This site had a sandy loam soil (61 percent sand compared to 33 percent sand at the Fort Missoula site). Herbicide was applied along a roadside adjacent to 2.5 mi of Elk Creek. Application averaged 110 feet from the creek but in some spots the sprayed site was as close as three feet to the stream. The groundwater table varied from three to nine feet deep throughout the application area. The area was treated with 4pt/acre of Tordon 22K (four times higher than the rate applied at the Fort Missoula area and two-three times higher than what is proposed in this analysis).

At the Elk Creek site, the authors reported photo decay during the first seven days to be half of what was observed at Ft. Missoula. The difference was attributed to tree canopy cover and less available sunshine. After three months, 78 percent of the applied picloram was

Page 122 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences still persistent in the soil. Although soil sampling detected picloram at depths of 40 inches, there was no detection of the herbicide in the groundwater or the stream.

Differences between Alternative 1 and 2 While Alternative 2 increases the number of acres that could be treated, a review of past Lolo NF herbicide applications and the mitigation measures in Chapter 2 indicate little difference between alternatives in relation to surface water quality.

Direct and Indirect Environmental Consequences Specific to Alternative 1 In upland areas under Alternative 1, weedy forbs with single or spreading stems and low basal cover would have more opportunity to replace perennial grasses which have multiple stems and relatively high basal cover. Soil cover, biomass production, and accumulations of organic matter drop dramatically when weeds dominate a site. Erosion rates would increase, and soil biochemical processes and soil organism habitats would be altered (Willard et al. 1988). A structure of perennial grasses more readily dissipates energy associated with falling raindrops, provides cover to the soil surface to minimize erosion, and traps blowing soil particles and snow (Olson 1999). Under Alternative 1 native bunchgrasses, which have dense fibrous root systems that hold the soil, would be more rapidly displaced than in Alternative 2 by spotted knapweed (which is the most widespread weed on the Lolo NF and has a taproot) and other weeds.

Direct and Indirect Environmental Consequences Specific to Alternative 2 Alternative 2, which allows a higher maximum number of treated acres per year, would thus also allow for inclusion of more selection criteria when prioritizing the annual program of work. For example, minimizing erosion on slopes where perennial grasses have been replaced by weeds, would have more of an opportunity for treatment when more acres are available for targeting in a given year.

Cumulative Effects Intensive scientific research, including studies in western Montana within the broad periphery of the Lolo NF, demonstrates that herbicides can be used to control weeds with no affect on soil or water resources. The mitigation measures for the Proposed Action are even more restrictive than EPA label restrictions and assure higher Margins of Safety than calculated by Norris et al. (1983) and displayed in the above table. There are no direct, indirect nor cumulative effects on soil or water resources from this alternative. Further, the Proposed Action meets the intent of Lolo Forest Plan.

Forest Plan Consistency and Regulatory Framework Most of the Lolo NF is classified B-1 by the Montana Department of Environmental Quality (DEQ) (ARM 16.20.604). The associated beneficial uses of B-1 water are drinking, culinary and food processing purposes and conventional treatment: bathing, swimming, and recreation; growth and propagation of salmonid fishes and associated aquatic life,

Final Integrated Weed Management EIS Page 123 Chapter 3 Affected Environment and Environmental Consequences waterfowl, furbearers, and other wildlife; and agricultural and industrial water supply (ARM 17.30.607 & 623). Municipal watersheds are classified A-1.

Applicable standards for Montana’s B-1 streams and rivers include maximum allowable increase in naturally occurring turbidity is 5 nephelometric turbidity units (NTU); and no increases are allowed above naturally occurring concentrations of sediment, settleable solids, oil, or floating solids, which would or are likely to create a nuisance or render the water harmful, detrimental, or injurious to public health, recreation, safety, welfare, livestock, wild animals, birds, fish, or other wildlife (ARM 17.30.623).

In Montana, numeric water quality standards are specific in Circular DEQ-7, Montana Numeric Water Quality Standards (MDEQ February 2006) as human health water quality standards for herbicides that could be used on the Lolo NF (Table 3-30). (Montana Water Quality Standards prohibit toxic or harmful concentrations to aquatic life in surface waters, including concentrations of weed control herbicides.)

Table 3-30 Montana Water Quality Human Health Standards for Herbicides Human Health Standard Herbicide Category (micrograms/liter) Groundwater Surface water 2,4-D toxin 70 70 Chlorsulfunon toxin 1,750 1,750 Clopyralid toxin 3,500 3,500 Dicamba toxin 210 210 Imazapyr toxin 21,000 21,000 Methsulfuron methyl toxin 1,750 1,750 Picloram toxin 500 500 The Clean Water Act requires that Total Maximum Daily Loads (TMDLs) be established by States/tribes/U.S. territories, and EPA for water bodies where water quality standards are not being attained(Federal Pollution Control Act as amended by the Clean Water Act, 1987). The EPA, States, tribes and territories refer to these as “impaired water bodies” or Water Quality Limited Segments (WQLS). To achieve the goals of the Clean Water Act for watersheds having forest and range lands, land management objectives must include protecting and maintaining watersheds that are healthy and functioning, plus improving those watersheds that are unhealthy and not functioning. The Forest Service policy is to participate with states, tribes, private land owners, and the United States Environmental Protection Agency (EPA) in the preparation and implementation of TMDL’s which are the responsibility of States/tribes/EPA to establish; and to encourage and assist States, tribes, and EPA to develop and implement effective programs for controlling non-point sources of pollution. The Montana Department of Environmental Quality designates stream segments they determine to be water quality limited within the Lolo NF. Primary causes for stream impairment include flow alteration, siltation, land development, roads, and other habitat alterations. None of the Lolo NF 303(d) listed streams had herbicides listed as a cause for impairment.

Page 124 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Recreation

Existing Conditions Weeds can affect the recreation experience. Invading weeds such as spotted knapweed, thistles, toadflax, leafy spurge, houndstongue, and oxeye daisy detract from the desirability of recreation sites and public enjoyment of the forest environment. These weeds diminish the usefulness of sites because the stiff plant stalks, thorns, or toxic sap can discourage or prevent walking, sitting, or setting up a camp. Weeds also detract from the recreation experiences by reducing the variety and abundance of native flora to observe or study and reducing forage availability for wildlife and recreational livestock.

Weeds are frequently spread through recreational activities, particularly along roads and trails, and at campgrounds, and dispersed recreation sites. The Lolo NF provides a variety of recreational experiences and opportunities including camping, hiking, hunting, fishing, mountain biking, snowmobiling, horseback riding, skiing, and driving for pleasure. Passenger vehicle roads provide primary transportation routes into and throughout the Forest. These roads provide access for a variety of purposes (commercial, residential, administrative), but their primary public benefit may be for recreational purposes.

Method of Analysis The area for recreation analysis is confined to developed and dispersed recreation sites on the Lolo National Forest. Environmental consequences are based on the potential for spread and expansion of weeds if left untreated and proposed weed treatment activities to impact recreational opportunities on the Lolo NF.

Environmental Consequences - Recreation

Direct and Indirect Environmental Consequences Common to Both Alternatives Some degree of visitor displacement is the most likely direct effect of weed treatment. Some people chose to relocate their recreation activities during weed treatments while others either don’t mind or appreciate knowing and seeing active weed control actions. Effects on recreation from actual weed control would be minor and short-term (one to seven days) under both alternatives. Nearly all people encountered during weed control activities on the Lolo NF do not leave when they encounter an operation.

Direct and Indirect Environmental Consequences Specific to Alternative 1 Direct and indirect effects under Alternative 1, weed infestations not covered by existing decisions would not be treated. Untreated weed infestations would be allowed to spread into weed free and relatively weed free areas. The spread of existing biological weed control agents may provide some degree of control on otherwise untreated sites, but this effect would be negligible in most situations. Consequently, the long-term impact of limited weed control will substantially increase weed density throughout recreation sites that are not

Final Integrated Weed Management EIS Page 125 Chapter 3 Affected Environment and Environmental Consequences covered under existing weed control decisions, which in turn, will cause weeds to spread into adjacent non-infested areas.

Direct and Indirect Environmental Consequences Specific to Alternative 2 Direct and indirect effects on recreation from Alternative 2 would include short-term (one to seven days) encounters with herbicide treatment crews, short-term odors from herbicides, and visual impacts from wilting plants. Additional effects resulting from this alternative would enhance the recreation experience by protecting adjacent non-infested areas and maintaining native and desirable plant communities. Concern over herbicides may cause some forest users to choose to recreate in areas that have not been recently treated with herbicides while other forest users would be attracted to areas where weeds are eliminated or reduced. All weed treatment activities would be conducted in compliance with Lolo Forest Travel Plan regulations, which allow for administrative use.

Weed infestations at dispersed sites, trailheads and along trails, and campgrounds within the project areas could be treated. The mitigation measures in Chapter 2 would reduce or eliminate recreation/weed control conflicts. Controlling weeds along roads and at recreational sites will reduce the likelihood of recreational activities spreading weeds into adjacent areas.

Under Alternative 2 weed treatments would reduce establishment and expansion of weed species into un-infested areas and reduce weed-related impacts on recreation. The visual impact of spraying would be temporary and on most sites only last a few hours. Dying and wilting weeds and some non target vegetation after herbicide treatments would be apparent.

Long-term improvements include an overall reduction of stiff plant stalks and sharp stickers and an increase in the variety and amount of native and desirable flora. Treating weeds would be an improvement in the overall quality and aesthetics of the recreational sites.

Cumulative Effects Cumulative effects from weed management actions and the alternatives described in Chapter 2 would continue to impact recreation by affecting the location where and times when people can recreate at various locations across the Lolo NF.

Some people chose to relocate their recreation activities during weed treatments while others either don’t mind or appreciate knowing and seeing active weed control actions. Effects on recreation from actual weed control would be minor and short-term (one to seven days) under both alternatives. Adverse long term recreation setting effects (fewer weeds and more desirable vegetation) are greater under Alternative 1. Some degree of visitor displacement is the most likely direct effect of weed treatment. An expanded weed control program (as in Alternative 2) will maintain the native plants and current visual quality of native plant communities. The present weed control program (Alternative 1) would allow a greater increase in weed species and a decrease in native plants, resulting in a diminished visual quality for the landscape.

Page 126 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Forest Plan Consistency and Regulatory Framework Lolo National Forest Plan Standard No. 28 states in part, “Land management practices shall be designed to have a minimum impact on the aquatic ecosystem, free from permanent or long-term un-natural imposed stress.” While Standard No. 28 was written to address management actions, the intent was to achieve an aquatic ecosystem free from unnatural imposed stress. Weeds that alter natural watershed function to the extent that sediment yields in the order of magnitude of hundreds of tons are imposed on Forest streams (as in this alternative) are not compatible with the intent of this Forest Plan Standard.

The goal of the Lolo National Forest Plan (1986) relative to recreation is to maintain existing dispersed and developed recreational opportunities. The Forest Service Manual, FSM 2300, describes the Forest Service Authority, Objectives, Policy, and Responsibility for recreation management. Pertinent Federal Laws are the Forest Rangeland Renewable Resources Planning Act of 1974, as amended by the National Forest Management Act, and the Wilderness Act of 1964. Alternatives 1 and 2 are consistent with the Forest Plan, but Alternative 2 better meets the weed management objectives in the Lolo NF Plan.

Human Health and Herbicide Drift

Existing conditions The Lolo NF lies mostly in Missoula, Mineral and Sanders counties. Missoula is a trade center for western Montana, and with a population of 57,035, it is the largest city in Western Montana (2000 Census). In the 2000 census the Missoula County population was 95,802; Sanders County 10,227; and Mineral County 3,884. Missoula has the second largest population in Montana (behind Billings).

The human health environment is the Lolo NF lands in Flathead, Granite, Lake, Lewis and Clark, Mineral, Missoula, Powell, Ravalli and Sanders counties. Effects are related both to the impacts of weeds on humans and the impacts of weed control. For weeds, concerns are related to the impacts from exposure to pollens and plant chemicals. For weed control, concerns are related to the exposure to toxicants found in the herbicides used in ground and aerial applications.

The human populations affected by the weed control actions in the alternatives include the workers applying the herbicide, recreationists walking through treated areas (e.g. roadsides, trailheads, recreation areas, campgrounds), berry-pickers or plant gatherers using areas that have been sprayed, and private landowners that may use areas adjacent to Lolo NF land that have been treated. In relation to herbicides, there has never been a reported instance of acute toxic reactions by herbicide applicators on the Lolo NF.

Since 1992, the Lolo NF has conducted over 30 aerial herbicide application projects with a excellent safety record, no reported public or work injuries and no accidental resource damage.

Method of Analysis The Forest Service contracted with Syracuse Environmental Research Associates (SERA) to evaluate ecological and toxicological data and human health effects based on

Final Integrated Weed Management EIS Page 127 Chapter 3 Affected Environment and Environmental Consequences

Environmental Protection Agency (EPA) studies and other current peer-reviewed scientific literature. Human health and safety risks for the use of herbicides in this analysis are based on SERA Human Health and Ecological Risk Assessments (RAs) and other documents. The SERA RAs are incorporated into this analysis, are summarized in this chapter and can be found at http://www.fs.fed.us/foresthealth/pesticide/.

RAs have been prepared for all these herbicides except for aminopyralid. Aminopyralid is a herbicide recently reviewed and registered (August 10, 2005) for use under the Reduced Risk Pesticide Initiative of the U.S. Environmental Protection Agency. This designation is reserved for compounds that demonstrate lower risk to the environment and humans than marketplace standards. Decisions on which herbicides to use for weed control are based not only on their effectiveness on weeds, but also on the risks associated with their use. Toxicity information was reviewed to determine the levels of these chemicals that would be harmful to human health. The herbicides proposed for use in this analysis and common product names are listed below in Table 3-31.

As part of each risk assessment, a set of general exposure scenarios was developed based on the normal use of the herbicides. For workers, projected exposures are calculated from the amount of the herbicide handled per day and an estimate of possible exposure via various routes. Estimates of probable worker doses are based on empirical relationships of absorbed dose to pesticide exposures (Durkin 2005; training manual, p.3-32). For the general public, three general exposures scenarios are considered: walking through a sprayed area shortly after treatment, the consumption of ambient water from a sprayed watershed, and the consumption of sprayed fruit or fish.

Some, if not all, of these general exposure scenarios for the general public are very conservative. For example, in many cases, compounds are applied in relatively remote areas. Thus, under most situations, it is not likely that members of the general public would be exposed to substantial levels of any of these herbicides or treated plants right after application. (As an additional precaution, signs informing of herbicide use are and would continue to be placed at access points to treatment areas prior to initiating treatment.) The scenarios however, include exposure right after application to address exceptions to applications in remote areas.

Similarly, the estimates of longer-term consumption of contaminated water are based on estimated application rates and monitoring studies that can be used to relate levels in ambient water to treatment rates in a watershed. In most herbicide applications, however, substantial proportions of a watershed are not likely to be treated.

Finally, the exposure scenarios based on longer-term consumption of contaminated vegetation assume that an area of edible plants is inadvertently sprayed and that these plants are consumed by an individual over a 90-day period.

The most common wild food collected in large quantities on the Forest would be huckleberries. The weeds listed in Chapter 1 do not commonly occur among huckleberry shrubs. In addition, if huckleberry plants were sprayed, it would occur along roadsides, a site type pickers may tend to avoid because of the dust and dirt associated with roads. Treated huckleberries would have to be collected and stored to be available for the 90 day time frame. Treated plants fruit, in the field after spraying, would look wilted and would rapidly become unappealing to foragers. Also, it would not remain fresh and available for

Page 128 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences the 90 days. Finally, treated areas would be posted so that pickers entering the area would know that the site or roadside had been sprayed.

In addition, a set of accidental exposure scenarios are also assessed in case of gross over- exposure associated with a spill or mishandling of a chemical (Durkin 2005; training manual, p.3-33). The worker exposure scenario involves immersion of the hands for a 1- minute period and wearing contaminated gloves for an hour. Accidental exposures of the general public are evaluated in three scenarios: • A naked child is sprayed directly with a pesticide as it is being applied and no steps are taken to remove the pesticide from the child for 1 hour • A woman is accidentally sprayed on her feet and legs • One in which there is an accidental spill into a small pond and a young child consumes contaminated water shortly after the spill Although these scenarios may seem unlikely, they are useful as a uniform comparison among herbicides and as an indicator for concern; i.e. if the ‘naked child’ scenario indicates no basis for concern, other dermal spray scenarios with lesser exposure will also pose little basis for concern.

Environmental Consequences – Human Health

Comparing Direct and Indirect Effects – Alternatives 1 and 2 Risks to workers and the public from manual and mechanical treatment methods. Both alternatives include incidental amounts of manual and mechanical weed treatments such as pulling and mowing. Human health and safety risks for manual and mechanical weed control is based on review of job hazard analyses, consideration of the specific actions involved in these activities and the tools and machines involved.

Manual (hand) and mechanical (power mowers) treatments pose hazards to weed control workers. Adverse weather and terrain commonly create unfavorable working conditions and increased hazards. Hazards associated with adverse weather conditions include extreme heat and cold, which can be exacerbated by very dry and very wet conditions. Other hazards include: falling objects; tripping or slipping on hazards on the ground; protruding objects such as branches and twigs; poisonous plants and insects, dust inhalation and dangerous wildlife such as yellow jackets.

Tools and equipment present inherent hazards such as sharp edges on the tools, exhaust fumes, dust generation and the hazardous nature of fuels and lubricants used in mechanized equipment. Manual and mechanical methods present potential ergonomic hazards related to lifting and carrying equipment, and pulling vegetation.

Injuries can vary from minor cuts, sprains, bruises, and abrasions to respiratory impairment (from dust or fumes) to major arterial bleeding, compound bone fractures, serious brain concussions, and death. Workers are subject to heat-related illness or hypothermia when working in extreme weather conditions, and may incur musculo- skeletal injuries related to improper body mechanics. Equipment operators could be injured from improperly operating the equipment or losing control of equipment on steep or slippery terrain. Operators, nearby workers and the public may suffer hearing damage or be struck by flying debris around some machinery.

Final Integrated Weed Management EIS Page 129 Chapter 3 Affected Environment and Environmental Consequences

The potential for hazard exposure, i.e. risk of injuries, may be exacerbated when workers are fatigued, poorly trained, or poorly supervised, and do not follow established safety practices. Appropriate training, together with monitoring and intervention to correct unsafe practices, would minimize risk of worker injury and illness. Compliance with safety standards and regular review and discussion of the Job Hazard Analysis along with agency, industry and manufacturers’ recommendations reduces the potential exposure and risk of injury to workers. Members of the public are usually not at risk from manual and mechanical methods unless they are too close to machinery that is producing flying debris during treatment.

Comparison of effects between Alternative 1 and 2: Since mechanical and manual weed control is a minor component of both alternatives, effects of the alternatives would be essentially the same.

Risks to Workers and the Public from Livestock Grazing to Control Weeds. Human health and safety risks for livestock grazing are anticipated to be minimal or non existent. Sheep or goats could technically attack and injure or kill a person but while such risk is recognized, that risk is considered low.

Comparison of effects between Alternative 1 and 2: Alternative 1 does not include livestock grazing for weed control. Risk of any adverse impacts from livestock grazing for weed control would be greater under Alternative 2.

Risks to Workers and the Public from Biological Controls: Release of biological control agents may have potential effects on human health. Pearson et. al. (2004) reported that release of the knapweed seed head gall fly (Urophora affinis and U. quadrifaciata) may increase deer mouse (Peromyscus maniculatus) abundance and the prevalence of Hantavirus. This may increase the risk of humans contracting this disease.

Biological control agents may change ecosystems on a large scale in ways that have unknown potential effects on human health. These changes may occur on a similar scale to how weeds change ecosystems.

Comparison of effects between Alternative 1 and 2: Since both alternatives include the release of biological control agents, and biological control agents can spread to Lolo NF lands from adjacent lands even in the absence of Forest Service releases, there is no difference between the alternatives.

Risks to the General Public and Workers from Weeds. Some weeds can cause allergies and minor skin irritations in humans. Some weeds, such as thistles, cause minor scrapes and irritations. Other complications may result from hand pulling. For example, leafy spurge contains a latex-bearing sap that irritates human skin and rarely causes blindness in humans upon contact with the eye (Callihan et al. 1991).

Allergic individuals can have serious complications when exposed to allergens (weeds or pollen), including constriction of the airway and anaphylactic shock, the significance of which should not be underestimated since forest workers are often working some distance from medical assistance.

Page 130 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Approximately 10 to 15 percent of the U.S. population suffers from allergy symptoms associated with weed species such as knapweed. Knapweed contains a common and powerful allergen that peaks in August (Gillespie and Hedstrom 1979). Allergies to weeds such as knapweed may complicate or trigger asthma. It may take up to two years after getting a person’s allergies under control to see a benefit in reduced asthma symptoms (Nielson 1999).

In 1992, a man in northern Idaho reported cuts on his hand after pulling spotted knapweed. The cuts eventually developed into what doctors called a very aggressive benign tumor. The man eventually had to have his little and ring fingers removed (Neihoff 1997).

After this incident, the Bureau of Land Managements National Applied Resource Sciences Center (BLM 1997) followed up and researched the topic. They reported that many plants are poisonous in one way or another. Some plants are poisonous on ingestion to livestock or humans. Other plants cause inflammation (irritation or dermatitis) on contact. The physical type from spines is called either skin irritation or is sometimes called contact dermatitis; whereas the chemical type via allergy or sensitization (e.g., poison ivy, poison oak) is called allergic contact dermatitis.

Plants have evolved these defenses as a means of protection. It is not unusual to observe injury from contact with certain plants. However, it is unusual to associate tumors with contact from plants. Russian knapweed, Centaurea repens, contains a pharmacologically active chemical complex called repin, a member of the family referred to generally as sesquiterpene lactones (SQLs). These SQLs have been found in many plants (e.g., the poisonous range plants bitterweed, small head sneezeweed, and yellow starthistle, the latter of which is on the Lolo NF weed list) and have been responsible for livestock grazing losses. After poison ivy and poison oak, SQLs account for most cases of allergic human contact dermatitis. Unlike poison ivy, which sensitizes a large part of the population, however, those who have casual contact with plants containing SQLs require long and intimate contact before sensitization occurs. Persons pulling weeds are advised to wear leather work gloves to protect the hands from abrasions and blisters when pulling weeds.

When ingested by horses, repin from Russian knapweed has been associated with a movement disorder and encephalomalacia simulating Parkinson’s disease. Repin has been found to be toxic to animal brain cells. The BLM report concluded that while knapweed does appear to be poisonous to livestock which ingest sufficient amount but, there is no evidence that dermal exposure to knapweed causes cancer.

Several knapweeds occur throughout the west. Spotted knapweed and diffuse knapweed are found on the Lolo NF while Russian knapweed and yellow starthistle occur on lands near to and hence are considered threats to the Lolo NF. Persons pulling weeds are advised to wear leather work gloves to protect the hands from abrasions and blisters when pulling weeds.

Comparison of effects between Alternative 1 and 2: Alternative 2 could potentially treat approximately three times the annual acreage and reduce the human health impact of weeds better than Alternative 1. Neither alternative, however, would eliminate weeds. Alternative 2 would slow the spread of weeds better than Alternative 1 and would provide better weed control in disturbed areas (areas of concentrated use) and places with the

Final Integrated Weed Management EIS Page 131 Chapter 3 Affected Environment and Environmental Consequences greatest human activity, so would reduce the human health impact of weeds better than Alternative 1.

Risks to the General Public and Workers from Herbicides. Both alternatives include herbicide use but Alternative 2 would allow use of two additional herbicides (see Table 3- 31 below) and up to 15,000 acres of treatment per year. For the six year period from 2000 through 2005, annual herbicide treatment on the Lolo NF varied from a low 1,428 acres in 2000 to a high of 4,510 acres in 2004. Only herbicides reviewed and registered by the EPA would be used in both alternatives and label directions (which include such things as hazard identification, potential health effects, first aid, handling and storage precautions, exposure guidelines, personal protective equipment and toxicological information) would be strictly followed. The herbicides included in Alternatives 1 and 2 are displayed below.

Table 3-31 Herbicides Included in Alternative 1 and 2 Common name Selective? Common product names Aminopyralid Yes Milestone, ForeFront R&P 2,4-D Yes Formula 40, Hi-Dep, Weedar 64 Chlorsulfuron 4-11 Yes Telar Clopyralid Yes Transline Dicamba Yes Banvel, Veteran 10-G Glyphosate No Roundup, Rodeo, Accord Imazapic Yes Plateau Imazapyr* Yes Arsenal, Chopper, Contain Metsulfuron methyl Yes Escort, Ally Picloram Yes Tordon Triclopyr* Yes Garlon In Place, LI-700, MSO, Preference, R-11, Stay Put, Adjuvants N/A Syltac Dyes N/A SPI Max, Hi-Light, Dye marker * indicates herbicides included in Alternative 2 but not Alternative 1. The EPA develops Reference Doses (RfDs) for chronic exposures (which are an estimate of a daily dose over a 70 year lifespan) that a human can receive without an appreciable risk of deleterious effects. RfDs include a “safety factor” where the No Observed Effect Level is divided by a factor, usually 100 fold, to account for uncertainty and hypersensitive individuals. The 100 value is derived by including a safety margin of 10 for using the results from mammals other than humans, and another safety factor of 10 for variation in the population to a particular compound. The chronic RfD is a very conservative toxicological threshold in relation to this analysis because it assumes daily exposure over a 70 year lifespan. Actual potential environmental exposures for the projects in this analysis for both workers and the general public would not be every day for 70 years. The chronic RfD is used to error on the side of caution and display potential effects if a person was exposed daily for a 70 year lifespan. Actual chronic effects are less if exposure is less.

RAs were used to evaluate whether worker or public exposure scenarios could potentially exceed the RfD. A background discussion of RA hazard identification, exposure assessment, dose-response assessment risk characterization considered in Forest Service RAs is included in the project file (Durkin / SERA, 2005).

Herbicide application workers are exposed to many of the same hazards as manual/mechanical treatment workers when working in an outdoor wildland

Page 132 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences environment. They may also be exposed to machinery noise from pumps, vehicles, and aircraft. The RA scenarios consider only risks associated with the herbicides themselves. Risk assessment scenarios are discussed above under “Methods of Analysis”.

The human health hazards associated with each herbicide active ingredient were evaluated with a thorough review of relevant toxicological studies. In all SERA RAs developed since 2000, the EPA’s Confidential Business Information files were accessed and this literature was cited where legal to do so in the assessments. The EPA develops Reference Doses (RfDs) to serve as a benchmark for estimating risk of health effects from a lifetime exposure to herbicides. These RfDs generally reflect the most conservative (health protecting) No Observed Adverse Effect Level (NOAEL) or No Observed Effect Level (NOEL) (used in lieu of a NOAEL) and are made even more conservative by the application of precautionary factors which further increase conservatism in these estimates. The precautionary factors also consider and account for data uncertainty and other factors representing corrections for both intra- and inter-species variability. The FS/SERA RAs normally adopt EPA’s RfD values unless compelling evidence exists indicating that the EPA erred in its selection of NOAEL or NOEL. For the purposes of the chemicals evaluated in this EA, EPA’s evaluations for RfD are used in all cases.

Judgments about the potential hazards of herbicides to humans are based, in large part, on the results of toxicity tests on laboratory animals. Information on actual human poisoning incidents and effects on human populations supplements the laboratory animal test results, where such information is available. For a background discussion of all toxicological tests and endpoints considered in FS/SERA RAs, refer to SERA, 2001.

Detailed discussion of the toxicity endpoints evaluated and a review of relevant toxicological literature is presented in the cited SERA documents which are available to the public at www.fs.fed.us/foresthealth/pesticide. No information of concern is presented in these documents relating to carcinogenicity, mutagenicity or teratogenicity. All other toxicity endpoint discussion for the variety of possible scenarios described above is subsumed in the RA which combines the herbicides inherent hazard, an estimate of exposure, and a dose-response assessment modeled together to generate an estimate of risk for each scenario for each chemical – referred to as the Hazard Quotient (HQ). The herbicides that would be available for invasive plant treatment under each alternative are compared based on the HQ, which is the ratio between the estimated dose (the amount of herbicide received from a particular exposure scenario) and the RfD. When a predicted dose is less than the RfD, then the HQ (estimated dose/RfD) is less than 1.0, and significant toxic effects are unlikely for that specific herbicide application.

Public Potential Health Risk: For the typical application rates, only two of the herbicides (2,4-D and dicamba) and NPE in R11 exceeds a HQ of 1. All eight of the problem scenarios are acute exposures and result from direct spray of a naked child who does not wash off the herbicide for 1 hour (2,4-D), an adult female eating 1 lb. of sprayed fruit (2,4-D), a child drinking l liter of water from a pond where 200 gallons of mixed herbicide was spilled (2,4-D, dicamba and Nonylphenol Polyethoxylate -NPE), a child drinking 1 liter of water

Final Integrated Weed Management EIS Page 133 Chapter 3 Affected Environment and Environmental Consequences from a stream contaminated by runoff or percolation (2,4-D) and an adult male recreational angler and subsistence population member eating fish from a pond after a herbicide spill (2,4-D). The plausibility of these scenarios is low because applicators are unlikely to both encounter and spray a naked child and in the event of a pond spill, precautions would be taken to prevent public access following a spill (reducing the chance of drinking the water and fishing). Mitigation measures are described in Chapter 2 make runoff or percolation contamination unlikely.

In the NPE scenario, the HQ>1 was for a child drinking contaminated water after a spill of herbicide and surfactant containing NPE at 25 ounces/acre (Baake 2003, p.35). This scenario is highly unlikely because typical rates of the surfactant used on the Lolo NF are 1 ounce or less/acre.

Consumption of sprayed fruit may be slightly more plausible. Huckleberries are the most commonly collected wild fruit on the Lolo NF. Weed infestations do not typically occur on aspects, elevations and habitat types where huckleberries are found. Sprayed areas would be posted and fruit contaminated with 2,4-D would look wilted and have an unusual odor that is affiliated with 2,4-D. In the event a wild fruit gatherer did not notice the posting, wilting of the plant and odor of 2,4-D, the fruit would have an uncharacteristic taste that would make it unlikely the woman in the scenario would consume 1 lb. of fruit.

The worst case scenarios include the acute exposure scenarios described above plus an acute exposure in which a woman is directly sprayed on her lower legs and feet and does not wash for 1 hour. Again, as stated above, applicators are instructed in situational awareness and never spray a person. The one chronic worst case scenarios not described above results from a woman eating 1 lb. of contaminated wild fruit per day for 90 days. Posting the site, and the appearance and taste of sprayed wild fruit would make this scenario unlikely.

No other scenario presenting a greater potential exposure to the general public is expected. There were no general public scenarios with HQs greater than 1 for imazapyr. Triclopyr had four general public worst case scenarios with HQs greater than 1. These scenarios were direct spray of a naked child who does not wash off the herbicide for 1 hour, a woman being directly sprayed on her lower legs and feet and not washing it off for 1 hour, a child drinking l liter of water from a pond where 200 gallons of mixed herbicide was spilled and a woman walking through a recently sprayed area wearing shorts for 1 hour and not washing for 24 hours. The plausibility of the first three scenarios is discussed above. The fourth scenario may be more plausible. Treated areas would be posted (that herbicides will be applied) as a mitigation measure to reduce the likelihood of this scenario occurring. The HQ of this worst case scenario was relatively low, 1.3 however, indicating that while it may

Page 134 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences be plausible, mitigation would reduce the likelihood and in the event of it occurring, the HQ is low, indicating the exposure is very close to the conservative EPA RfD.

Five of the herbicides: chlorsulfuron, imazapic, imazapyr, sulfometuron methyl and picloram (the most commonly used herbicide on the Lolo NF) have a HQ of less than one for all scenarios. A HQ less than one indicates significant toxic effects are unlikely for these herbicide scenarios.

Worker Potential Health Risk: For the typical application rates, only one of the herbicides (2,4-D) exceeds a HQ of 1. All three of these HQs are also relatively low and close to one. All three of the problem scenarios are chronic exposures. These scenarios are for general exposure for backpack (HQ = 1.3), ground (HQ = 2) and aerial (HQ = 1.5) application. Chronic exposures in these scenarios assume an applicator applies this herbicide every day for a 70 year lifespan. These scenarios are mitigated and unlikely to occur because applicators will apply herbicides, at most, 30 to 40 days a year (usually less) for a maximum 20 to 30 year career. In addition workers will wear personal protective equipment and observe safe herbicide handling practices.

The worst case scenarios include the chronic exposure scenarios described above plus several acute exposure scenarios. These acute exposure scenarios are a worker wearing contaminated gloves for 1 hour (HQ = 1.7) and a worker spilling herbicide on their lower legs and not washing it off for 1 hour I HQ = 2). Both of these scenarios have relatively low HQs that are close to 1. The likelihood of these scenarios occurring is low because applicators carry clean wash water on the job, and are instructed to wash immediately upon contamination or spilling herbicide on themselves.

Seven of the herbicides: chlorsulfuron, clopyralid, glyphosate, imazapic, imazapyr and sulfometuron methyl and picloram (the most commonly used herbicide on the Lolo NF) have HQs of less than one for all scenarios. A HQ less than one indicates significant toxic effects are unlikely for these herbicide scenarios.

Comparison of Potential Worker Effects between Alternative 1 and Alternative 2: The possible use of imazapyr and triclopyr differentiate Alternative 1 from Alternative 2. Alternative 2 includes these two herbicides. Aminopyralid is a recently registered herbicide (August 10, 2005) and, as such has the studies required by EPA for registration but does not have a Forest Service RA yet. According to the laboratory studies, it has the lowest toxicity of any of the proposed herbicides and requires a much lower application rate than most of the other proposed herbicides (http://www.epa.gov/opprd001/factsheets/aminopyralid.pdf). Because of its low toxicity no acute Reference Dose (RfD) has been established; the chronic RfD is 0.5 mg/kg/day. Based on its low toxicity and the low application rate needed, worker exposure to aminopyralid would not exceed the chronic RfD. There were no scenarios with worker HQs greater than 1 for imazapyr. Triclopyr had three worst case scenarios with worker HQs greater than 1. These scenarios were all chronic exposures for general exposure for backpack (HQ = 1.6), ground (HQ = 3) and aerial (HQ = 1.6) application. These are all relatively low HQs close to 1. Chronic exposures in these scenarios assume an applicator applies this herbicide every day for a 70 year lifespan. These scenarios are mitigated and unlikely to occur because applicators will apply herbicides, at most, 30 to 40 days a year (usually less) for a maximum of a 20 to 30 year career. In addition, workers will wear personal protective equipment and observe safe herbicide handling practices.

Final Integrated Weed Management EIS Page 135 Chapter 3 Affected Environment and Environmental Consequences

The most commonly used herbicide under both alternatives, picloram, has a HQ less than one, which indicates significant toxic effects to humans are unlikely under any of the general public or worker scenarios.

Impurities and Metabolites. Formulated herbicides as applied in Forest Service weed treatments may contain additional compounds besides the active herbicide ingredient that are called impurities or inert ingredients. Other additives, called surfactants or adjuvants, may be mixed with the diluted formulation before spraying to either enhance the herbicide activity, or to modify undesirable properties of the spray mixture. Additionally, when organisms in the environment internalize herbicide formulation chemicals in their physiologic systems, they may be transformed into other compounds called metabolites.

In addition to the analysis of potential hazards to human health from each herbicide active ingredient, Forest Service/SERA RAs evaluate any available scientific studies of potential hazards of these other substances associated with herbicide applications: impurities, metabolites, inert ingredients, and adjuvants. While there is often less information available on these substances because they are not subject to the extensive testing that is required for the herbicide active ingredients under Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA), it must be noted that testing of formulated products is relatively common, and where it is done, the test results include the effects of the additives which are included in the formulated product (inert ingredients, surfactants, penetrants or other chemicals added to enhance product efficacy). In some cases, additional information is available concerning inerts and adjuvants if they have been tested to comply with other federal laws, such as the Federal Food, Drug, and Cosmetic Act.

Toxicity studies generally account for impurities in the active ingredient, except in the case of carcinogens associated with the following impurities: • Hexachlorobenzene (HCB) in both clopyralid and picloram • Ethylene oxide in NPE (nonylphenol polyethoxylate)-based surfactants • 1,4-dioxane in some formulations of glyphosate containing NPE-based surfactants Analyses of the carcinogenic risk of these three impurities are presented in the corresponding FS/SERA or Bakke RA. In addition to carcinogenic risks, acute and other chronic risks from exposure to HCB are also specifically analyzed. In Bakke (2003) the risk of cancer from exposure to ethylene oxide in NPE-based surfactants was considered negligible for workers, based on the EPA standard of acceptable risk of less than 1 in 1 million. Ethylene oxide is not analyzed further in this human health RA.

Borrecco and Neisess (1991) found the risk of cancer from exposure to 1,4-dioxane in glyphosate was considered negligible for workers, based on the EPA standard for acceptable risk of less than 1 in 1 million. 1,4-dioxane is not analyzed further in this human health RA.

Similar to impurities, the potential health effects of herbicide metabolites are often accounted for in the available toxicity studies on the herbicide active ingredient, assuming that the toxicological effects of metabolism in the test animal species would be similar to the toxicological effects in humans. Uncertainties in this assumption are encompassed in the precautionary uncertainty factor used in calculating the RfD and may sometimes influence the selection of the study used to derive the RfD.

Page 136 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

The herbicide triclopyr presents a special case: its principal metabolite, called TCP, is also a metabolite of chlorpyrifos, an insecticide. The risks for direct, indirect, and cumulative effects of human exposure to TCP from both sources are discussed in the SERA triclopyr herbicide RA (SERA 2003e).

Comparison of Alternatives for impurities and metabolites: The effects are the same between Alternative 1 and 2 with the exception that Alternative 2 includes triclopyr (refer to preceding paragraph).

Inert Ingredients. FS/SERA RAs analyze the human health risks of inert ingredients and full herbicide formulations by the process described below: • Comparison of acute toxicity data between the formulations (that include inert ingredients) and the active ingredients alone to address whether the inert is likely to constitute a substantial fraction of the formulation toxicity • Disclosing whether or not the formulated products have undergone chronic toxicity testing • Identifying, with the help of EPA and the herbicide registrants, ingredients of known toxicological concern in the formulated products and assess the risks of those ingredients Researchers who have studied the relationships between acute and chronic toxicity have found that relationships do exist and acute toxicity data can be used to give an indication of overall toxicity (Zeise, et al., 1984). The court in NCAP v. Lyng, 844 F.2d 598 (9th Cir 1988) decided that this method of analysis provided sufficient information for a decision maker to make a reasoned decision. In SRCC v. Robertson, Civ.No. S-91-217 (E.D. Cal., June 12, 1992) and again in CATs v. Dombeck, Civ. S-00-2016 (E.D. Cal., Aug 31, 2001) the district court upheld the adequacy of the methodology described above for disclosure of inert ingredients and additives.

The EPA has categorized approximately 1,200 inert ingredients into four lists. Lists 1 and 2 contain inert ingredients of known or suspected toxicological concern. List 4 contains nontoxic substances such as corn oil, honey and water. List 3 includes substances for which EPA has insufficient information to classify as either hazardous (List 1 or 2) or non-toxic (List 4). Use of formulations containing inert ingredients on List 3 and 4 is preferred for weed treatment under current Forest Service policy and in the alternatives considered in the EIS.

Much of the information about inert ingredients that is submitted to EPA for pesticide registration is classified as “Confidential Business Information” (CBI). CBI is not generally released or available for public review. SERA risk assessors, however, have reviewed the identity and data on inerts in the CBI files when preparing all herbicide RAs except 2,4-D (which RA predates their clearance to review the data). This RA is being revised to include CBI materials available to our assessor at the present time. The revision is expected in mid- summer 2006 and if significant changes are made in the assessment of risk based on this review, the new information will be used to reevaluate conclusions drawn in this document.

All publicly released information from herbicide registrants and Freedom of Information Act requests to EPA has also been reviewed for its relevance in Forest Service/SERA risk analyses. Comparison of acute toxicity (LD50 values) data between the formulated products

Final Integrated Weed Management EIS Page 137 Chapter 3 Affected Environment and Environmental Consequences

(including inert ingredients) and their active ingredients shows (due to dilution when incorporated into less toxic carrier liquids or granulating materials) the formulated products are generally less acutely toxic than their active ingredients (even as concentrated formulated products are poured from the container). Further dilution generally is done prior to loading the herbicide into the tank of the application tool which further reduces the inherent hazard per unit of volume of the chemicals being applied. Forest Service/SERA RAs review the acute toxicity comparisons, the EPA’s review, of the toxicity information on inert ingredients in each formulated product.

Comparing Alternatives for inert ingredients: The effects are the same for Alternatives 1 and 2. For all formulations containing inert ingredients that have been reviewed in the FS /SERA RAs, the reviews have concluded that the formulations do not significantly increase the risk to human health and safety over the risks identified for the active ingredients. Further, in all cases (where both separate information on the active ingredient and formulation inert ingredients toxicology are known, and there is information available concerning the formulated products toxicological profile) in the formulated product data accurately portrays the toxicology spectrum predicted from assessing a hypothetical equivalent mixture (of appropriate proportion) of the two separate components. That is, a model which considers the active ingredient and the inerts separately and adds projected HQs for the two in proportions equivalent to the formulated product for which data is available, yields HQs equivalent to those projected using the data developed concerning the formulated product.

Comparing Alternatives for surfactants: The effects are the same for Alternatives 1 and 2. Surfactants, or surface-acting agents, are a broad category of activator adjuvants that facilitate and enhance the absorbing, emulsifying, dispersing, spreading, sticking, wetting, and/or penetrating properties of herbicides. Surfactants are most often used with herbicides to aid spread and penetration over the waxy cuticle (outer layer) of a leaf or to penetrate through the small hairs present on the leaf surface. Because of the high surface tension of water, spray mixture droplets can maintain their roundness and sit on the leaf hairs or waxy surface without much of the herbicide actually contacting the leaf. The primary purpose of a surfactant is to reduce the surface tension of the spray solution and allow more intimate contact between the spray droplet and the plant surface. They may also change the permeability of the leaf surface.

Most surfactants used with herbicides are considered non-ionic surfactants. This means that these compounds have no electrical charge. There are cationic (positive charge) and anionic (negative charge) surfactants, but they are not as commonly used, with the exception of the cationic surfactant in the Roundup® formulation of glyphosate. Surfactants have the physical characteristics of both oil and water. Most surfactant molecules contain a water-loving (hydrophilic) head and an oil-loving (lipophilic), long-chain hydrocarbon tail.

The surfactants used on the Lolo NF are In Place, LI 700, Mentholated Seed Oil, Preference, R-11 and Syl-tac. None of these contain ingredients found on U.S. EPA’s inerts list 1 or 2 (Baake 2002 and http://www.epa.gov/opprd001/inerts/lists.html). This is either based on the identified ingredients, or if these ingredients are not sufficiently identified, by information given by the manufacturers. The assessment of hazards for these adjuvants is limited by the proprietary nature of the formulations, and also by the simple fact that study of their toxicity is not required under the current version of FIFRA. Unless the U.S. EPA

Page 138 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences classifies a compound in the formulation as ether an active ingredient or as hazardous, the manufacturer is not required to disclose its identity. At the current time, the disclosure of whether a material is hazardous is based primarily on acute toxicity. All of the adjuvants discussed here are no more than slightly toxic when ingested, inhaled, or absorbed through the skin (ibid).

The primary summary statement that can be made is that the more common risk factors for the use of these adjuvants are through skin or eye exposure. These adjuvants all have various levels of irritancy associated with skin or eye exposure. Thus emphasizes the need for good industrial hygiene practices while utilizing these products, especially when handling the concentrate during mixing. Chemical resistant gloves and goggles are used at all times during application, but especially while mixing products and loading equipment.

Of the surfactants proposed for use, only surfactants containing NPE (nonylphenol polyetholxylate) have been tested and data produced that identifies specific and quantifiable hazards to human health. NPEs are also used in products designed for human exposure (such as cosmetics and spermacides). NPE is often referred to as nonoxynol (Bakke 2003). The FS RA (USDA FS 2003) identified that NPE may contain nonylphenol, an EPA List 11 inert, which has potential for toxic effects. The potential for toxic effects from use on the Lolo NF were analyzed for public and workers scenarios.

Comparing Alternatives for Additive or Synergistic effects from Herbicide Mixtures: The effects are the same for Alternatives 1 and 2 except that Alternative 2 includes triclopyr. (Refer to the discussion in this section on triclopyr). Synergistic effects are those effects resulting from exposure to a combination of two or more chemicals that are greater than the sum of the effects of each chemical alone (additive). No indication of potential synergisms among the chemicals in the alternatives, or between these chemicals and other environmental pollutants has been seen in the literature relating to these chemicals, with a single exception. Picloram in combination with 2,4-D is known to produce a mild skin rash (unpredicted from the behavior of either chemical alone) in a limited percentage of individuals. There have also been no reports of xenoestrogenic response in these chemicals.

Effects on Sensitive Individuals. The uncertainty factors used in the development of the RfD take into account much of the variation in human response. The uncertainty factor of 10 for sensitive subgroups is sufficient to ensure that most people would experience no toxic effects. “Sensitive” subgroups are those that might respond to a lower dose than average, such as women, children or hypersensitive or allergic persons. The National Academy of Sciences (NAS), 1993, the quantitative differences in toxicity between children and adults are usually less than a factor of approximately one order of magnitude (10-fold). Human susceptibility to toxic substances can vary by two to three orders of magnitude, so an uncertainty factor of 10 for sensitive subgroups may not cover every individual who may be unusually sensitive to herbicides. Unusually sensitive individuals may experience effects even when the HQ is equal to or less than 1. Individual susceptibility to the herbicides in the alternatives cannot be specifically predicted. Factors affecting individual susceptibility include diet, age, heredity, pre-existing diseases, life style, and other environmental factors.

There is anecdotal information (case histories) suggesting that some individuals may be sensitive to 2,4-D. These individuals report neuropathy as impaired nerve function after exposure to 2,4-D at levels that are not expected to cause adverse health effects in the general population. The effects reported in the case studies are debilitating, and recovery

Final Integrated Weed Management EIS Page 139 Chapter 3 Affected Environment and Environmental Consequences may be prolonged and incomplete. On the other hand, no case study to the present has been scientifically definitive; the case studies do not rule out the possibility that the neuropathy was caused by other unidentified agents.

There is no information to suggest that specific groups or individuals may be especially sensitive to the systemic effects of chlorsulfuron, clopyralid, imazapic, imazapyr, metsulfuron methyl, picloram, or sulfometuron methyl.

The only identified sensitive subgroup for dicamba appears to be children. Since the RfD for dicamba explicitly considers the increased sensitivity of children with an additional safety factor and since exposure assessments for children are conducted in the risk assessment, this sensitive subgroup is adequately addressed in the SERA risk assessment of dicamba. (SERA 2004e, 2004f)

Because triclopyr may impair blood filtration, individuals with pre-existing kidney diseases are likely to be at increased risk. Women of childbearing age have been recognized by EPA as a particularly sensitive subgroup for acute exposures to triclopyr by reducing the RfD from 1.0 mg/kg/day to 0.05 mg/kg/day.

There is limited information to suggest that specific groups or individuals may be especially sensitive to the systemic effects of NPE-based surfactants. NPE can cause increases in kidney and liver weight, and effects to kidney function and structure. Thus, individuals with preexisting conditions that involve impairments of the kidney or liver may be more sensitive to this compound. There is some indication that sensitive individuals may develop contact allergies. People with a history of skin allergic reactions to soaps and detergents may be especially sensitive to dermal exposures of NPE-based surfactants.

The potential of NPE to induce reproductive effects should be considered low. Based on the available dose/duration/severity data, it appears that exposure levels below those associated with the most sensitive effect (i.e., kidney effects) are not likely to be associated with reproductive toxicity. However, as shown in the exposure scenarios, there is the potential for acute exposures to be in the range (considering a 100X safety factor) where effects to the developing fetus may occur; therefore, pregnant women could be considered a sensitive population.

Persons who are chemically sensitive usually know that they are. The mitigation measure in Chapter 2 to post sites that will and have been treated with herbicides provides sensitive individuals the opportunity to avoid sprayed areas and avoid adverse effects. Use of herbicide dyes can further identify treated areas and aid those who wish to avoid such areas.

Herbicide Drift Methods of Analyzing Herbicide Drift. Herbicide drift is an issue with both the public and agency. To assure adequate disclosure of the potential impacts from aerial herbicide drift, four separate methods of analysis were conducted. The methods and sources used were: • The Risk Assessment for Herbicide Use in Forest Service Regions 1,2,3,4 and 10 and on Bonneville Power Administration Sites (May 1992) (RAHUFS). This assessment

Page 140 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

includes a Human Health Exposure Analysis, Spray Drift Modeling, Potential Exposures and a health risk assessment. • Forest Service Cramer-Barry-Grim aerial spray modeling (FSCBG) for three specific helicopters that do aerial herbicide spraying in western Montana. These models were run in 1999 by the Forest Service Pacific Southwest Region Remote Sensing Lab and Forest Service National Aerial Herbicide Application Specialist Jack Barry. These models assure that the specific configurations and aircraft that could operate on the Lolo NF are considered. • A review of spray drift potential for the Missoula Valley Region conducted by Environmental Toxicologist Alan Felsot at Washington State University in 2001. Felsot conducted his review using a model called AGDRIFT, which was developed with the cooperation of EPA and the USDA for use by industry to estimate distance that sprays may drift beyond the target field. Parts of AGDRIFT were based on a model developed and field tested by the Forest Service. • Actual on-the-ground drift monitoring in 1997, 1999 and 2003 on the Mormon Ridge aerial spray project on the Lolo NF and drift monitoring results from 11 aerial spray projects conducted on the Lolo NF between 2002 and 2005. These 11 projects included 36 drift card lines. This method of analysis is very valuable because it ensures that the models and reviews listed above are effective for actual on the ground local field conditions on projects identical in type and location to those in the proposed action. Aerial herbicide application has a greater potential for drift than ground based applications because release heights are higher. In this analysis it is assumed that drift potential from ground applied applications are less than aerial applications. Therefore, if human health exposures for aerial applications are below toxicological thresholds, ground applications would be even less.

Drift models assume worst-case scenarios. In the case of aerial or ground herbicide applications this worst-case scenario is an adverse wind (meaning wind blowing from your point of release toward the resource or area you are trying to avoid). Aerial applications would not be conducted near buffer zones when the wind was blowing toward the buffer zone and a sensitive resource you want to avoid, so potential drift from applications proposed in this analysis are exaggerated. Treatment buffers are required in the mitigation measures to further protect sensitive resources from spray drift.

Human health impacts from drift are considered in relation to 1) whether and how far they are below the toxicological thresholds set by the EPA and 2) the Montana Water Quality Human Health Standards for Herbicides proposed for use in this analysis (refer to Table 3- 30).

Herbicide Drift Dynamics. Spray drift is largely a function of droplet particle size. The largest particles, being the heaviest, will fall to the ground quickly upon exiting the sprayer. Medium size particles can be carried beyond the sprayer swath (the fan shape spray under a nozzle), but virtually all of the particles fall within a short distance of the release point. The physics of sprayers dictates that there will always be a small percentage of the spray droplets that are small enough to be carried in wind currents to varying distances beyond the point of release. Since these smallest droplets are a minor proportion of the total spray volume, their toxicological significance beyond the project area boundary rapidly declines as they are diluted in increasing volumes of air (Felsot 2001).

Final Integrated Weed Management EIS Page 141 Chapter 3 Affected Environment and Environmental Consequences

Drift characteristics differ widely between different kinds of pesticides and application objectives. With insecticides for example, the objective is to cover the leaf as thoroughly as possible with the pesticide being applied. In the case of mosquito fogging, particle size is so small that a fog is intentionally formed to contact the airborne pest. This is best achieved with very small droplets that in turn have farther offsite drift potential. With herbicides proposed in this analysis, it is not critical to coat the entire leaf since some of the products can be absorbed by the plant roots. Good efficacy can also be achieved with larger droplets on leaves of the target weed. Therefore, herbicide drift can be intentionally reduced by generating larger droplets without reducing efficiency.

Spray nozzle diameter, pressure, the amount of water applied with the herbicide, and herbicide release height are important controllable determinants of drift potential by virtue of their effect on the spectrum of droplet sizes emitted from the nozzles (Felsot 2001). Meteorological conditions such as wind speed and direction, air mass stability, temperature and humidity and herbicide volatility also affect drift.

Commercial drift reduction agents are available that are designed to reduce drift beyond the capabilities of the determinants previously described. These products create larger and more cohesive droplets that are less apt to break into smaller particles as they fall through the air. They reduce the percentage of smaller, lighter particles that are the size most apt to drift off the treatment.

Higher wind speed increases the concentration of droplets leaving the treated area, especially if the wind is adverse (blowing toward an area you want to avoid). If the wind is favorable (blowing back into the treatment area) drift can be reduced. However, even if winds are adverse, a high wind speed can cause a greater dilution or dispersion of the herbicide in the air further away from the field, and thus reduce its concentration when it impacts the ground or foliage.

The mitigation measures described in Chapter 2 take all these factors into consideration and are intended to reduce or offsite drift.

Since spray drift is largely a matter of physics, graphs of the relationship between fractions (or the percentage) of herbicide applied in the project area and deposition distance downwind all look very similar. Well over 90 percent of the spray droplets land on a field being treated, and about 10 percent or less move off-target. However, of the droplets that move off-target, typically deposit within 100 feet of the field border (Felsot 2001).

In summary, while spray drift is impossible to completely eliminate, it is possible to minimize drift to levels that do not exceed toxicological thresholds. Establishing no spray buffer zones between the sprayed area and sensitive resources further reduce unintentional exposure.

The most obvious and effective measure to avoid drift toward sensitive resources is to use the wind to an advantage. Aerial applications will be made when there is wind less than 6 mph blowing away from the sensitive area. This practice combined with a 150’ to 300’ “no spray” buffer areas adjacent to sensitive areas and a drift reduction agent will reduce the likelihood of offsite drift that exceeds EPA toxicological thresholds and Montana Water Quality Human Health Standards for Herbicides.

Page 142 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

RAHUFS Drift Estimations. The 1992 Risk Assessment for Herbicide Use in Forest Service Regions 1,2,3,4 and 10 and on Bonneville Power Administration Sites, or RAHUFS, determined spray drift distances downwind of an application site for aerial, backpack and ground mechanical application equipment. Spray drift from hand application equipment was considered negligible. A detailed methodology is included in RAHUFS (pg III-D-2-E-6). RAHUFS is incorporated by reference into this analysis (see Chapter 1). The results of the RAHUFS spray drift analysis indicates “low” health risk to the public from both ground based and aerially applied for the herbicides proposed in this analysis (RAHUFS pg III-E-7). “Low risk” means drift from the herbicides present a less than 1 in 1 million systemic or reproductive risk or cancer risk (RAHUFS pg III-E-5).

FSCBG Drift Estimations. Forest Service Cramer-Barry-Grim aerial spray modeling (FSCBG) was conducted using specific details for three helicopters that do aerial herbicide spraying in western Montana. These were a Bell 47 Saloy, a Bell 206BIII and a Hiller 12E. The Forest Service Pacific Southwest Region Remote Sensing Lab and Forest Service National Aerial Herbicide Application Specialist Jack Barry ran these models. The FSCBG model was used to generate X, Y plots and isopleths of downwind drift. The input variables included helicopter type and also included flight speed, nozzle type and atomization, wind direction, swath width and release height. The outputs were expressed in ounces of herbicide per acre deposition on a horizontal plane. These models indicated that material could drift about 200 feet on flat terrain in a 6 mph adverse wind. The aerial application projects in this proposal would be on predominantly sloped terrain and only in favorable winds, which would blow herbicide back into the slope and the treatment area. Aerial application proposed in this analysis also includes a 300-foot no-treatment buffer next to sensitive resources unless drift card monitoring indicates a smaller buffer is effective. Even if drift card monitoring indicates smaller buffer will be effective, the buffer will never be less than 150 feet wide.

AGDRIFT/FELSOT Drift Estimations. Felsot (2001) used the EPA / USDAFS AGDRIFT model that uses a combination of physical equations that describe spray drift and field measurements. AGDRIFT simulations of herbicide sprays were made for several application scenarios, including a truck mounted spray boom set at two heights and a helicopter at two heights. These simulations included crosswinds blowing at ten and six mph. The model output was a list of the fraction of the applied application rate that deposited at a defined distance from the edge of a spray swath. A spray deposition curve was then developed. To calculate dose that a bystander could potentially receive if standing within the drift zone of an application, the whole body surface area was assumed exposed to a drifting spray. In reality, only a small fraction of the body would be exposed given a person would usually have at least a shirt and pants on. A simple algebraic transformation allowed the depositing micrograms of pesticide per unit area to be expressed as milligrams of pesticide per unit of body dose. The bystanders were assumed to be an adult weighing 70 kg and a child weighing 10 kg.

Studies with human volunteers show that only a minor portion of the herbicides picloram and 2,4-D can be absorbed through the skin. For example, only 0.2 percent and 6 percent of an applied dermal dose of picloram and 2,4-D, respectively is absorbed. To build an extra safety factor into the exposure estimate, absorption of the depositing dose was assumed to be 10 percent and the person exposed was assumed to be a small child weighing 10 kg. Calculations were made to determine the percentage of the depositing spray that a child could be exposed to on a daily basis over a 70-year life span and still be

Final Integrated Weed Management EIS Page 143 Chapter 3 Affected Environment and Environmental Consequences within the EPA safety guidelines as defined by the RfD (i.e., the “equivalent safe deposit”). In reality, however, potential exposure in the project areas would only occur at most on very few days per year, not as a daily lifetime exposure upon which the RfDs are based. (Public access to aerial spray project areas is usually restricted too, so the chance of a person being in an aerial project area and coming into contact with herbicide at the time of application is very low.)

The equivalent safe deposits were 1831 percent, 251 percent, and 9.7 percent of the recommended application rate for clopyralid, picloram, and 2,4-D, respectively. For an aerial application, these equivalent safe deposits corresponded to distances from the edge of the spray field of 0, 0, and about 60 feet, respectively, for clopyralid, picloram, and 2,4-D. For a ground application, the “equivalent safe deposit” of 2,4-D would occur within 27 feet of the sprayed field edge.

Although direct exposure to clopyralid or picloram spray is unlikely to cause any adverse effects on bystanders, such exposure is certainly not desirable and would be inconsistent with the product label directions. Furthermore, prudence dictates that the buffer zone should always be designed based on the compound with the lowest RfD, which is 2,4-D in this example, even though picloram would be the most commonly aerially applied herbicide as proposed in this analysis. Recommended minimal buffer zones to achieve a reasonable certainty of no harm to children were developed and are displayed below in Table 3-32. The buffer zone would be the length in feet extending from the edge of the effective spray swath in a downwind distance to a bystander.

Table 3-32 Recommended Minimal Buffer Zones Associated with Different Herbicide Spray Applications in the Missoula Valley to Achieve a Reasonable Certainty of No Harm to Children (Felsot 2001) Spray One Spray Swath Twenty Spray Swaths Parameters Clopyralid 2,4-D Picloram Clopyralid 2,4-D Picloram Ground spray 0 5 0 0 5 0 Low Boom (2 ft) Ground spray 0 15 0 0 30 0 High Boom (4 ft) Aerial Spray (10-ft release - - - 0 65 0 height) Aerial Spray (25-ft release - - - 0 65 0 height) Source: Felsot 2001 Bystanders outside of the calculated buffer zones in Table 3-32 are predicted to have no exposure in excess of EPA guidelines. Actual buffer zones of 150 to 300 feet would be required for aerial applications in this analysis adjacent to sensitive resources. A 300 foot buffer is 4.6 times greater than the largest buffer recommended by Felsot to protect a child during worst case applications of 2,4-D, a herbicide that would be used very little in the project areas. The guidelines are conservatively set to protect people because they assume daily exposures over a lifetime of 70 years and an adverse wind. But the actual potential exposure in the analysis areas would only occur very few days each year, no applications

Page 144 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences would be made in adverse wind, and most of the treatment areas are away from populated areas. The exposures that may inadvertently occur with spray drift are very short term, and in the modeling scenario take place within a 24-hour time interval.

Mormon Ridge and other Actual Lolo NF Aerial Application Drift Monitoring. Herbicides were aerially applied with helicopters to the Mormon Ridge winter range in 1997, 1999 and 2003. A Bell 47 Saloy was used in 1997 and a Bell 206 BIII was used in 1999 and 2003. This project was a pilot project in 1996 which was more complex than most of the projects in this analysis. Mormon Ridge was more complex than most of the projects in this analysis because it is extremely steep, has rolling topography, considerable microclimate variability and aerial application occurred just upslope of Mormon Creek, a bull trout spawning stream. Mormon Creek runs along the bottom of the roughly three miles by ½ to ¾ mile wide treatment area. Many projects proposed in this analysis are less complex and have much lower change of herbicide drift into no spray areas or to sensitive resources.

Picloram was first aerially applied on Mormon Ridge in 1997. Buffer zones and water quality were monitored in Mormon Creek. The mitigation measures described in Chapter 2 were used. Continuous automated water samples were taken.

Analysis of the water samples (conducted by the Montana Department of Public Health and Human Services Chemistry Lab) indicated no herbicide entered the stream to a detection level of 1 part per 10 billion (U.S. Department of Agriculture-FS 1996). The Montana Water Quality Human Health Standards for Herbicides in surface water is 500 parts per billion. No picloram was detected in Mormon Creek when tested at a level 5,000 times lower than the Montana Water Quality standard. Drift cards were also placed along Mormon Creek to monitor drift toward the creek. The cards indicated that no detectable drift reached the creek.

The Mormon Ridge pilot project area was aerially treated with picloram again in 1999 and 2003 to control weeds that germinated from the soil seed bank after the herbicide decomposed. The same mitigation measures were used to protect Mormon Creek. Drift cards did not detect herbicide within 150’ of Mormon Creek in 1997 and 300’ in either of the successive applications.

In addition to Mormon Ridge, 24 other aerial herbicide projects have been conducted on the Lolo NF between 1992 and spring 2005. On 12 of those projects, herbicides were aerially applied within 300’ of live water or other sensitive resources. Thirty-six drift card lines were placed and monitored to ensure herbicides did not reach the identified sensitive resource (usually water). Thirty-five of these lines indicate that herbicide did not reach the sensitive resource.

One drift card line on the Prospect Creek project in June 2002 showed 1 percent detection at 0’ from a small low flow unnamed creek. The situation on Prospect Creek was evaluated and additional wind monitoring and buffering was done for the next aerial application in June 2005. In the 2005 aerial application, the situation was remedied and herbicide was detected no closer than 270’ from the sensitive resource.

Bitterroot NF Aerial Application Monitoring: On 9/30/04 and 10/1/05, the Bitterroot National Forest (USDA 2005) implemented an aerial spray project in two areas to control weed species. Water quality sampling was carried out before, during, and after the aerial

Final Integrated Weed Management EIS Page 145 Chapter 3 Affected Environment and Environmental Consequences spraying on both of the sites treated that fall 2005. Three hundred foot wide riparian buffer strips were established and marked with white feed sacks for pilot identification. Drift cards were placed at set intervals across the buffer, between the stream and the marked line. Results from the lab indicated herbicide detection as “below detection limit” for all tested chemicals in all samples. No herbicides used in the spraying operation were detected in the water samples. All drift cards were visually scanned for droplet presence. No cards were found with visible droplets after the spray treatments.

Spray Drift Summary. Based on the above information, herbicide spray drift under Alternative 1 or 2 would not adversely effect general public outside the treatment areas or adversely effect water quality. Herbicide drift outside the project area would be substantially below EPA RfD’s or Montana Water Quality Human Health Standards for Herbicides. Chemically sensitive persons who may react to herbicide concentrations far below EPA RfD’s or Montana Water Quality Human Health Standards for Herbicides may be affected but the likelihood is very low because we would prevent public access during aerial applications and posting treated sites to minimize this potential impact. Herbicide applications may have a low short-term, very localized impact on extremely chemically sensitive individuals as a result of drift. Most of the drift would settle to within 100-200 feet of the point of release in adverse conditions but applications would not occur in adverse conditions.

Cumulative Effects Alternatives 1 and 2. The proposed (Alternative 2) and existing (Alternative 1) use of herbicides could result in cumulative doses of herbicides to workers or the general public. Cumulative doses from the same herbicide may result from (1) additive doses via various routes of exposure resulting from a single invasive plant treatment project and (2) additive doses if an individual is exposed to other herbicide treatments. Possible exposure scenarios are described and analyzed in this chapter. Additional sources of exposure include: use of herbicides on adjacent other non Lolo NF lands or home use by a worker or member of the general public.

Forest Service risk assessments consider the effects of both single (acute) and repeated (chronic) exposure(s) by assessing the exposure levels for each herbicide. Chronic long term exposure scenarios are based on a lifetime of repeated doses from consuming contaminated water, fruit, or fish and repeated exposure to herbicides and direct spray.

This analysis specifically considers the effect of repeated exposure in the chronic exposure scenarios and through the use of the chronic RfD as an index of acceptable exposure. Where individuals could be exposed by more than one route, the cumulative risk of such cases can be quantitatively characterized by adding the HQs for each exposure scenario. Using glyphosate as an example, the typical levels of exposure for a woman being directly sprayed on the lower legs, staying in contact with contaminated vegetation, eating contaminated fruit, and consuming contaminated fish leads to a combined (acute) HQ of 0.012. Similarly, for all of the chronic glyphosate exposure scenarios, the addition of all possible pathways lead to HQs that are two orders of magnitude less than 1, indicating an acceptable level of cumulative risk. Consequently, repeated exposure to levels below the toxic threshold should not be associated with cumulative toxic effects. Mitigation measures are included to address cumulative HQs that exceed 1.

Page 146 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Similar scenarios can be developed with the other herbicides. Since these herbicides persist in the environment for a relatively short time (generally less than 1 year), do not bioaccumulate in humans and are rapidly eliminated from the human body, no significant cumulative health effects from repeated treatments in the past, present or foreseeable future are predicted. The herbicides clopyralid, picloram, and sulfometuron methyl persist in the environment for more than one year; however, re-treatment in the following year is not expected; thus, no additive doses from re-treatment are predicted.

Hexachlorobenzene, an impurity found in herbicide formulations containing clopyralid and picloram (trade names include Transline and Tordon), was evaluated for cancer risk based on long term exposures. The scenarios which include consumption of contaminated fruit, fish, and water resulted in exposures at least 100 times below the level of concern (a cancer risk of 1 in 1 million).

Estrogenic effects (a common toxic action, and a form of endocrine disruption) can be caused by additive amounts of nonylphenol (NP), NPE (an ingredient in R-11, one of the surfactants used on the Forest), and their breakdown products. In other words, an effect could arise from the additive dose of a number of different xenoestrogens (estrogens from outside the body), none of which individually have high enough concentrations to cause effects, except perhaps in especially sensitive or allergic individuals. This can also extend out to other xenoestrogens that biologically react the same. Additive effects, rather than synergistic effects, are expected from combinations of these various estrogenic substances.

When assessing cumulative effects of exposure to NP and NPE, there must be some consideration of the contribution from other sources, such as personal care products (skin moisturizers, makeup, deodorants, perfumes, spermicides), detergents and soaps, foods, and from the environment away from the forest herbicide application site. In addition to xenoestrogens, humans are exposed to various phytoestrogens, which are hormone- mimicking substances naturally present in plants. In all, more than 300 species of plants in more than 16 families are known to contain estrogenic substances, including beets, soybeans, rye grass, wheat, alfalfa, clover, apples, and cherries.

The Forest Service, Pacific Southwest Region, recently analyzed the risks of cumulative estrogenic effects from proposed Forest Service use of NPE, plus worst-case environmental background and consumer product exposures (USDA FS 2002). Adding together the cumulative contributions from the worst-case background environment and consumer products, the risk assessment estimated that backpack applicator exposure would add from 0.1 (typical rate) to 10 (maximum rate) to the cumulative HQ, which ranged from 3 (low dermal exposure assumptions) to 270 (high dermal exposure assumptions). For the public chronic exposures at the maximum application rate, the doses of NPE would add 0.00002 to 0.2 to any HQ. These may be negligible depending upon the background exposures, lifestyles, absorption rates, and other potential chemical exposures that are used to determine overall risk to environmental xenoestrogens.

The potential for cumulative human health effects from any herbicide use proposed in this EIS, combined with other potential herbicide applications in the analysis area, is encompassed in the chronic exposure scenarios, which consider the effect of repeated exposures. The risk of toxic health effects from repeated exposure to any of these herbicides at doses that are less than the chronic toxicity benchmark (chronic RfD) is low.

Final Integrated Weed Management EIS Page 147 Chapter 3 Affected Environment and Environmental Consequences

Difference in Cumulative Effects between Alternative 1 and 2: Alternative 2 includes all the herbicides in Alternative 1 plus imazapyr and triclopyr. Triclopry’s principal metabolite, called TCP, is also a metabolite of chlorpyrifos, an insecticide. The risks for direct, indirect, and cumulative effects of human exposure to TCP from both sources are discussed in the SERA triclopyr herbicide RA (SERA 2003e)).

Forest Plan Consistency and Regulatory Framework Regulatory Framework. Safety standards for herbicide use are set by the Environmental Protection Agency, Occupational Health and Safety Administration, Code of Federal Regulations (40 CFR part 170), and individual states. In addition, several sections of the Forest Service Manual (FSM 2150, 2153, 2156, 6709, FSH 2109, 6709) provide guidance to the safe handling and application of herbicides.

Forest Service Manual direction (FSM 2109.16.3) recommends and defines, Pesticide Risk Assessment as “Another method of helping to ensure safety in pesticide use is to conduct risk assessments. Analyses estimate the possible pesticide dose to workers and the public who may be affected by a pesticide application; and the potential effects on fish, wildlife, and other non-target organisms. These estimated doses are then compared with levels of no observed effects based on tests of laboratory animals.”

These analyses are usually incorporated into the decision-making documents prepared in compliance with the National Environmental Policy Act (FSM 1950). A pesticide risk assessment does not, in itself, ensure safety in pesticide use. The analysis must be tied to an action plan that may provide additional mitigation measures to minimize or avoid potential risks identified by the risk assessment.

Wilderness and Inventoried Roadless Areas

Existing conditions Wilderness Areas are areas of federally owned land that have been designated by Congress as Wilderness, in accordance with the Wilderness Act of 1964. These areas are protected and managed so as to preserve their natural conditions which: • Generally appear to have been affected primarily by forces of nature with the imprint of man’s activity substantially unnoticeable; • Have outstanding opportunities for solitude or a primitive and confined type of recreation; • Have at least 5,000 acres or is of sufficient size to make practical their preservation, enjoyment, and use in an unimpaired condition; and • May contain features of scientific, educational, scenic, or historical value as well as ecologic and geologic interest. A Wilderness Study analysis is conducted on candidate areas to determine an area’s appropriateness, cost, and benefits for addition to the National Wilderness Preservation System. Inventoried Roadless Areas (IRAs) are areas identified in a set of inventoried roadless area maps, contained in Forest Service Roadless Area Conservation, Final Environmental Impact Statement, Volume 2 dated November 2000 (USDA), which are held at the national

Page 148 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences headquarters office of the Forest Service or any subsequent update or revision of those maps.

The analysis area for wilderness and inventoried roadless areas is the extent of the individual wilderness area and/or roadless area. The Lolo National Forest includes about 940,133 total acres of Wilderness and Inventoried Roadless Areas.

Table 3-33 Summary of land in Wilderness, National Recreation Area and Inventoried Roadless Designation. Total Forest Rattlesnake Welcome Scapegoat Inventoried Acres Wilderness and Creek Wilderness Roadless National Wilderness (Lolo NF Recreation Area Portion) 2,077,823 60,976 acres 28,135 acres 74,832 acres 776,190 acres acres

Rattlesnake Wilderness and National Recreation Area: Congress designated the Rattlesnake Wilderness in 1980 and it now has a total of 32,976 acres. It lies adjacent to the Rattlesnake National Recreation Area and is often referred to cumulatively as the Rattlesnake National Recreation Area and Wilderness (RNRAW). The southern boundary of the RNRAW is just four miles north of Missoula, Montana. The approximately 28,000 acre National Recreation Area (NRA) receives heavy human use, but primarily in the South Zone, within about three miles of the main trailhead. Far fewer people venture into The Rattlesnake Wilderness which is in the more remote northern portion of the RNRAW. A primitive road (which is used by the public as a trail) leaves the main trailhead into the RNRAW along Rattlesnake Creek and heads up a NRA corridor to within about three miles of a scenic Wilderness cluster of high alpine lakes. Near the NRA entrance at about 3,600 feet, the elevation rises to 8,620 feet on McLeod Peak and a picturesque mountain setting.

The northern boundary of the wilderness abuts the Confederated Salish and Kootenai Indian Reservation, upon whose sacred ground only tribal members are allowed.

Eight trailheads provide access to the RNRAW. Several of the small lakes and lake access trails are closed to livestock. Camping and campfires are restricted in the south zone of the NRA but are allowed in the Wilderness. Rattlesnake Creek is a municipal watershed for the City of Missoula. Several of the lakes in the upper wilderness are dammed and maintained by the local water company.

There are many weed infestations peripheral to the RNRAW. Trails used throughout the RNRAW provide conduits for weed spread by bikers, hikers, joggers and horseback riders. Some of these trails continue on into the Wilderness. There are also several roads ending at or near the RNRAW boundary which have existing weed infestations. These weeds have the potential to spread into the Rattlesnake Wilderness and destroy naturally functioning ecosystems.

Welcome Creek Wilderness: Congress designated the Welcome Creek Wilderness in 1978 and it has a total of 28,135 acres. The Welcome Creek Wilderness lies in the Rock Creek drainage about 25 miles east of Missoula, Montana. This 28,135 acre wilderness measures about nine miles by seven miles. The ridges are very steep and the valleys are narrow and

Final Integrated Weed Management EIS Page 149 Chapter 3 Affected Environment and Environmental Consequences uneven, Most of the land is heavily timbered with pine, fir and larch (although the south- facing slopes have a few open but very steep grassy slopes) and is extremely rocky and rough. Welcome Creek flows south and east, providing a home to native trout. Elk hunters, bear hunters, and a few adventurous anglers from Rock Creek are the most common visitors.

About 25 miles of steep trails provide foot and horse access, but overnight backpackers are few and far between. The main travel route is the Welcome Creek Trail, which crosses the area southeast to northwest for approximately seven miles, passing no lakes (this Wilderness has none).

Scapegoat Wilderness: Congress designated the Scapegoat Wilderness in 1972 and it has a total of 239,936 acres. This wilderness lies on the Lolo, Helena and Flathead NFs. Approximately 74,832of the total 239,936 acres lie on the Lolo NF. The long northwest border of Scapegoat Wilderness is shared with Bob Marshall Wilderness. Together, the Great Bear Wilderness, the Bob Marshall Wilderness and the Scapegoat Wilderness form the Bob Marshall Wilderness Complex, an area of more than 1.5 million acres.

Scapegoat’s rugged ridge tops slope down onto alpine meadows, heavily forested hillsides, and timbered river bottoms. Fish are plentiful in the 14 lakes and 89 miles of streams. Elevations range from about 5,000 feet on the Blackfoot River to about 9,400 feet on Red Mountain. Wildlife includes wolverines, moose, deer, elk, mountain goats, mountain sheep, mountain lions, black bears, and numerous grizzly bears.

Hundreds of miles of trails are suitable for backpacking and horse packing, and most of them follow drainages.

Inventoried Roadless Lands: Approximately 776,190 acres of inventoried roadless in 36 separate areas are located on the Lolo National Forest. The inventory was displayed in the Lolo Forest Plan EIS, Appendix C (USDA 1986). In 2000 the Clinton Administration completed a nationwide study of “roadless” lands on public land, and maps of record included in the final rule (USDA 2001). The final rule acknowledges that this inventory may not be perfectly accurate, and likely included lands that no longer retained their roadless characteristics.

Inventoried roadless lands are scattered throughout (found in all mountain ranges and most often along ridge tops) the Lolo National Forest. Inventoried Roadless Area maps are available in the Forest Service Roadless Area Conservation, FEIS, Volume 2, USDA (2000). The Final Roadless Area Conservation Rule (Jan. 15, 2001) does not prohibit any treatment of invasive species.

Management activities (proposed, past, present, and reasonably foreseeable) were evaluated for their potential effects on the Wilderness attributes listed in the Forest Service Northern Region “Our Approach to Effects Analysis” for assessing the impacts on Wilderness and roadless characteristics. This method will be used for designated Wilderness and Inventoried Roadless Areas. The attributes include: natural integrity, apparent naturalness, remoteness and solitude, management, and boundaries. Natural integrity is the extent to which long-term ecological processes are intact and operating. Apparent naturalness is a measure of how natural the environment appears. Impacts to natural integrity and apparent naturalness are measured by the presence and magnitude of human-induced

Page 150 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences change to an area. Solitude is a personal subjective value defined as isolation from the sights, sounds, presence of others, and the developments of man. Management and boundaries will not be affected by proposed activities and will not be discussed further.

Environmental consequences – Wilderness and Inventoried Roadless

Direct and Indirect Environmental Consequences Specific to Alternative 1 The Wilderness and Inventoried Roadless areas on the Lolo NF currently have fewer weeds than non wilderness general forest areas. Difficulty of access for weed management however can allow introduced weed populations to spread undetected and unchecked. But difficulty of access also reduces the potential for weeds to be introduced. Ground-based application of herbicides, where needed, would result in protection of native vegetation and wilderness characteristics in the wilderness areas on the Lolo NF. Depending on weed density, associated vegetation and special location, herbicide treatment could retard or kill non-target vegetation and result in a short-term reduction in plant community diversity. However, native and desirable vegetation has been found to reestablish after treatment. (Rice et al. 1992)

Cumulative Effects Since additional weed treatments would not be considered in Alternative 1, weeds in burned areas and along trails in wilderness that do not have existing weed control would not be controlled and non-target vegetation in these areas will not be affected by herbicides. The spread of existing biological weed control agents from outside wilderness and roadless areas may provide some degree of control on otherwise untreated sites, but this effect would be negligible in most situations. Native species diversity would not be affected by herbicides or physical disturbance associated with mechanical weed control in these areas, but would be affected by weed competition. As time progresses, the native plant communities in Wilderness and roadless areas would become affected (reduced) by weeds.

Direct and indirect Environmental Consequences Specific to Alternative 2 Alternative 2 includes all current weed treatment methods and would allow treatment of new weed infestations and infestations not covered by past weed management Decisions. Effects of this alternative are the same as the effects of Alternative 1 except that new infestations could be treated more promptly before they spread and become larger. Early treatment of weeds in the wilderness and roadless areas while the infestations are small would preclude the need for larger scale and more impactive treatments in the future. There would be less chance of weed spread in these areas into un-infested and difficult to detect areas. Wilderness and roadless area characteristics would be better protected in the long term than under Alternative 1.

Final Integrated Weed Management EIS Page 151 Chapter 3 Affected Environment and Environmental Consequences

Cumulative Effects Several reasonably foreseeable, past, present, and future activities could contribute to cumulative effects to Wilderness and Inventoried Roadless Areas. The past, present, and future effects of recreation use in wildland areas have and will continue to lead to introductions and spread of weeds in wild land areas. The ability to efficiently respond to known and future infestations both in and adjacent to Wilderness and roadless areas is critical to maintaining the character of these areas and native plant communities in wildlands. Alternative 2 would have fewer cumulative effects from weeds in these areas than Alternative 1 by allowing managers to treat new weed infestations and those not covered by past decisions.

Forest Plan Consistency and Regulatory Framework Designated Wilderness is mandated to be administered so that its community of life is untrammeled by man, its primeval character retained, and naturally functioning ecosystems preserved (Public Law 88-577).

Wilderness areas are managed as directed by the Wilderness Act of 1964. Management actions within Wilderness focus on maintaining naturally functioning ecosystems, providing access through appropriate means (typically trails), and managing some pre- existing uses like outfitter operations. Examples of management activities include trail construction and maintenance, fire suppression or management of naturally ignited fires, removal of existing structures, and noxious weed treatment.

Forest Service Manual (FSM) 2323.26b allows plant control for “noxious farm weeds by grubbing or with chemicals when they threaten lands outside Wilderness or when they are spreading within the Wilderness, provided that it is possible to effect control without causing serious adverse impacts on Wilderness values.” FSM 2109.14 (13.4) requires Regional Forester approval of pesticide use in designated Wilderness Areas (USDA). Generic direction for Wilderness and proposed wilderness Management is found in the Lolo Forest Plan, page III-37 (USDA 1986).

Inventoried Roadless Lands: There is currently no specific congressional oversight of inventoried roadless lands. Weed treatments on inventoried roadless lands would not need special approval simply because of the area’s roadless status.

Wild and Scenic Rivers

Existing Conditions The analysis area for Wild and Scenic Rivers (WSRs) are those streams and adjacent lands within the Lolo National Forest that are currently suitable for protection under the WSRs Act.

Portions of nine streams and rivers were identified as “eligible” for WSR designation during the WSRs Eligibility Study (USDA 1991). Those included were segments of Cache Creek, the North Fork Blackfoot River, South Fork Lolo Creek, West Fork Fish Creek, Rock Creek, Morrell Creek, Rattlesnake Creek, Clark Fork River, and the Clearwater River. Congress has not designated these river segments, as “Wild and Scenic,” to date.

Page 152 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

The WSR Act was enacted to preserve in a free-flowing condition rivers that possessed outstanding scenic, recreational, geologic, fish and wildlife, historic, cultural, or other similar values. Congress declared that it was important to manage certain rivers in their free-flowing condition, and to manage them and their immediate environment to protect those qualities for the benefit and enjoyment of present and future generations. The presences of weeds along the river corridor can detract from the aesthetic and recreational opportunities. Weeds adjacent to creeks and rivers allow for as much or greater spread than weeds along roads because weeds along creeks and rivers are more difficult to detect (isolated and remote) and more difficult to control (due to the presence of water and shallow water tables) than weeds along roads. The eligible river segments are assigned a potential classification of wild, scenic, or recreational. Characteristics of these classifications are: • Wild River areas -free of impoundments, generally accessible only by trail, shorelines primitive and the water unpolluted. • Scenic River areas - free of impoundments, shorelines largely undeveloped but accessible in places by road. • Recreational River areas –readily accessible by roads, some development and may have impoundment or diversion. Method of Analysis The source of information for the Affected Environment was the 1986 Forest Plan (USDA), Amendment 12 (1991) and the 1996 EIS (USDA) for a proposed change to the Lolo Forest Plan regarding WSR designation. The analysis is based on the potential for the proposed weed treatment activities to impact the values inherent to rivers or streams on the Lolo National Forest that are suitable for protection under the WSR Act.

Environmental Consequences – Wild and Scenic Rivers

Direct and Indirect Environmental Consequences Specific to Alternative 1 Direct and indirect effects. Weeds will not be controlled within any suitable WSR Corridor that does not have a past weed management decision. Treatments would be limited by the mitigation measures listed in Amendment 11 to the 1986 Lolo NF Plan. The spread of existing biological weed control agents may provide some degree of control on otherwise untreated sites, but this effect would be negligible in most situations. There would be no substantial direct effects under Alternative 1 to the outstandingly remarkable attributes that make these rivers eligible for inclusion in the system. Weeds are present along all or portions of these streams. Water is a very effective weed spread vector. These established weed populations are difficult to treat effectively within close proximity to water. Alternative 1 would allow for more weed spread along these waterways than Alternative 2. This could reduce Outstandingly Remarkable Values (ORVs) such as recreation, fisheries and scenery.

Indirectly, the lack of timely weed control may affect the natural appearance (scenery) of these corridors, as weeds occupy suitable habitats. The presence of weeds could have a negative effect on the experience of recreationists who expect a natural environment without the presence of exotic plant species. Weeds can also increase sediment level, thus

Final Integrated Weed Management EIS Page 153 Chapter 3 Affected Environment and Environmental Consequences affecting fish habitat. Weeds can also decrease forage quality and composition and thus affect wildlife in the river corridor.

Direct and Indirect Environmental Consequences Specific to Alternative 2 Direct and indirect effects. There would be additional weed control along rivers/creeks within eligible/suitable WSR Corridors using the same mix of management tools as included in Alternative 1. New infestations could be treated before they grew to unmanageable proportions. There would be no direct effects of herbicide to the outstandingly remarkable features of these rivers due to the mitigation measures listed in Alternative 2 and those in the Lolo NF Plan. There would be less direct effects than under Alternative 1 because un-infested downstream areas would have a lower likelihood of being infested from upstream water transported weed seed.

Indirectly, the effective treatment of weeds along these corridors would improve ORVs such as recreation, fisheries and by restoring and protecting the native and desirable vegetation component of these corridors.

Cumulative Effects Cumulative effects would relate to the effects of weeds on ORVs such as recreation, fisheries and scenery. For Alternative 1 there would be the cumulative effect of a greater likelihood of weed spread along waterways off Lolo NF lands from weeds whose seeds were carried in the water from weeds on Lolo NF lands in WSR corridors. If current recreation trends persist and recreation use continues to increase, the cumulative effect would increase in the future. This effect would be less under Alternative 2 because weed control could occur on additional infestations.

Forest Plan Consistency and Regulatory Framework The Wild and Scenic Rivers Act (16 US1271) and Interagency Guidelines provided in the Wild and Scenic Rivers (WSRs) Reference Guide (USDA and others 1995) provide the general direction for management of these rivers. Additional discussion and interim management considerations are found in the Lolo Forest Plan, Amendment 12, WSR Eligibility Study, 1991 and the subsequent Final WSR Suitability Study and EIS, August 1996.

Heritage Resources

Existing Conditions The Lolo National Forest contains a rich and diversified cultural history represented in prehistoric and historic properties. Prehistoric resources date to at least the Late Paleo- Indian Period (8,000 - 10,000 years ago) and include early Native American campsites, rock art, quarries, cambium peeled trees, and other activity areas. Historically, this area was used by the Salish-speaking Pend d'Oreille and Salish people who now live on the Confederated Salish and Kootenai Reservation. The historic period begins in 1805 with the entry of the Lewis and Clark’s expedition into western Montana and the lands that today

Page 154 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences comprise the Lolo National Forest. The historic period ends 50 years ago in 1956. Fifty years is the accepted time for events or historic properties to achieve historical significance. In the event of exceptional significance an event or historic property may gain significance in less than 50 years, however this is rare. Historic properties include; prehistoric archeological sites, historic roads and trails (the Mullan Military Road and the Lolo trail National Historic Landmark are two examples), historic mines, homesteads, trapper’s cabins and US Forest Service administrative sites.

Consultation with the CSK Tribal Preservation Office (CSKTHPO) normally occurs on an annual basis for all proposed undertakings for that year on the Lolo NF. Consultation with the CSKTHPO for the proposed action occurred in December 2005. The Forest Archeologist contacted Mr. Francis Auld from the CSKTHPO and discussed the project. The CSKTHPO was provided with maps and project descriptions of the proposed treatment areas for their review and comment. Weed treatment concerns by the CSKTHPO are generally associated with areas where traditional plant gathering occurs. Consultation specific for this project has been initiated and will continue throughout the Section 106 process when the preferred alternative is selected. Protecting cultural resources is of paramount concern to the both the CSK Tribes and the Lolo NF.

Forty-eight Heritage Sites have been recorded within the analysis area. Twelve historic properties have been listed or determined eligible for listing on the National Register for Historic Places (NRHP) in consultation with the Montana State Historic Preservation Office and the Keeper of the National Register. Eight others have been determined to be “Not Eligible” and the remaining twenty-five have not been evaluated. These sites are included in the project file.

Method of Analysis Heritage resource inventories are conducted for all projects on the Lolo National Forest that may adversely affect historic properties. Usually these are limited to ground disturbing activities such as timber sales, road construction, prescribed fire and land exchanges where Federal lands will transfer into non Federal ownership. Heritage resource inventories are conducted using an inventory strategy outlined within the Site Identification Strategy (SIS) for the Bitterroot, Flathead and Lolo National Forests: (McLeod, Light and Williams: 2003). This document outlines the inventory methods used on lands throughout the Bitterroot, Flathead and Lolo National Forests based upon a variety of variables including topography, vegetative communities and prehistoric and historic site probability. The SIS has been accepted as the inventory standard for the Bitterroot, Flathead and Lolo National Forests by the Montana State Historic Preservation Office as well as the CSK Tribal Preservation Office.

All previous inventories essentially used similar field methods by focusing the search for heritage properties in areas of high to medium probability for site occurrence. The Prehistoric Overview for the Lolo National Forest (McLeod and Melton: 1986) and the Site Identification Strategy for the Bitterroot, Flathead and Lolo National Forests (McLeod, Light and Williams: 2003) defines such areas as major ridges, river terraces, prominent mountain peaks, lakeshores, natural meadows, and areas adjacent to springs. Other less common landforms with moderate potential for site occurrence include large rock faces, talus slopes, and possible rock shelters. Moderate probability areas can hold historic mining and logging activity oftentimes located in secondary draws and even mid-slopes.

Final Integrated Weed Management EIS Page 155 Chapter 3 Affected Environment and Environmental Consequences

Given the size of the analysis area and the diffuse nature of heritage resources, it is neither practical nor necessary to undertake a 100 percent inventory of the analysis area for this level of decision making. The Lolo NF Heritage Program uses a phased compliance process as provided for in 36 CFR 800.3 and 36 CFR 800.4 to identify and evaluate cultural resources and comply with NFMA, NEPA, and Section 106 of NHPA. At this stage of analysis, known historic properties and the location of possible historic properties are identified and the history of the area is researched. With this information, the Lolo NF can identify areas of concern for historic properties and Native American concerns to develop alternatives. Additional field inventory, analysis, and Section 106 compliance will occur as treatment areas are selected and closer to project implementation. The Montana SHPO, the ACHP, and the Confederated Salish and Kootenai Tribe are and will continue to be consulted and provided an opportunity to comment on the proposed undertaking and its effects on significant historic properties as provided in NHPA and 36 CFR 900 (R-1 Analysis and Guidance for Cultural Resources and NEPA March 1992).

Environmental Consequences - Heritage

Direct and Indirect Environmental Consequences Common to Both Alternatives

Physical non living heritage resources The likelihood of affecting significant prehistoric archeological sites is remote. The exception is the Hogback archeological site (24GN0013) which surrounds the Hogback Homestead. Weed treatment has occurred under previous decisions to restore the historic landscape immediately adjacent to this site. Spotted knapweed infestations adjoin the previously treated area. Additional weed treatment near site 24GN0013 would not affect the subsurface components of the site that make it eligible for National Register listing and would further restore the historic landscape, which was weed free.

Weed treatment at historic sites such as the Hogback Homestead (24GN0167), the Morgan Case Homestead (24GN0195), Ninemile Ranger Station (24M0300), the Ninemile CCC Camp (24MO0142), Savenac Nursery (24MN0145), Quartz Guard Station (24MN0203) and Monture Guard Station (24PW0047) would bring benefits to the historic resource by restoring the historic vegetative appearance on the landscape.

Subsistence Gathering and Culturally Important Plants Human health effects for subsistence gathering for American Indians for both acute and chronic exposure scenarios is addressed in the human health section of Chapter 3.

Picloram and clopyralid are two of the most commonly used weed control herbicides used on the Lolo NF. Measured residues of picloram and clopyralid in bitterroot roots and camas bulbs (two culturally important plants) are below EPA tolerance limits for crops when the plant foods are collected two and three years post spray. Clopyralid concentrations are lower than picloram levels at the same elapsed time after treatment. Under the climatic conditions of western Montana the half life of picloram is about 100 days while the half life of clopyralid is about 30 days (Rice 2002). Treatment sites are also posted to ensure that subsistence gatherers know if a site has been treated.

Page 156 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

The difference between Alt 1 and 2 for subsistence gathering are minimal. Whether 5,000 to 6,000 or 15,000 acres a year are treated the treatment areas are often roadsides, gravel pits or other disturbed areas and not places where the gathering of subsistence and culturally important plants normally occurs.

Direct and Indirect Environmental Consequences Specific to Alternative 1 Alternative 1 would treat roughly 5,000 to 6,000 acres a year. This would have a minimal effect to heritage resources. However if weeds are considered an “adverse effect” to historic heritage resources and their cultural landscapes this is a very limited amount of historic landscape restoration.

Direct and Indirect Environmental consequences Specific to Alternative 2 Alternative 2, on the other hand, could treat up to 15,000 acres a year from approximately 78,443 acres identified for treatment. This would provide a much larger menu from which to select treatment areas and more readily restore the historic landscape at many of our significant historic heritage resource sites. If additional funding does not materialize only 5,000 to 6,000 acres may be treated annually. However under Alternative 2, the Lolo NF could better respond to treat newly discovered infestations and protect heritage resources.

Cumulative Effects A total of 48 historic properties lie within or near the project areas. Twelve of these historic properties have been listed or determined eligible for listing on the National Register of Historic Places. Twenty five historic properties remain unevaluated for National Register listing and eight have been determined ineligible for listing on the National Register.

With the above exceptions and considering the remaining known heritage resources within the Forestwide Integrated Weed Management Area, there are no significant differences in effects between alternatives. There are no cumulative effects.

Forest Plan Consistency The goals of the Lolo National Forest Heritage Program are to manage heritage resources using the various Federal cultural resource laws and regulations to integrate these resources on an equal footing within the Forest’s resource management program. The Lolo National Forest Plan directs that heritage resources be considered during the planning process on all proposed undertakings. Inventories are conducted prior to ground disturbing activities, heritage resource sites are identified and evaluated for eligibility to the National Register of Historic Places, and eligible properties are managed in a manner consistent with the standards specified by the State Historic Preservation Office (SHPO), the Advisory Council on Historic Preservation (ACHP), and USDA Forest Service regulations and guidelines.

The Lolo NF Forest Plan further directs that the Lolo NF will consult on a yearly basis with representatives from the Confederated Salish and Kootenai (CKS) Tribes to discuss the type

Final Integrated Weed Management EIS Page 157 Chapter 3 Affected Environment and Environmental Consequences and location of proposed Forest undertakings to identify areas of traditional religious or cultural concern. Additional consultation occurs with the CSK Tribal Preservation Office whenever there is an indication a project may affect a historic property of interest to the CSK Tribe within their aboriginal homeland. Coordination with other American Indian groups such as the Nez Perce and the Shoshone Bannack will occur as needed (Lolo National Forest 1986: II-20).

The Lolo National Forest Heritage Program complies with all applicable state and federal laws and regulations including the Antiquities Act of 1906, the National Historic Preservation Act of 1966 (NHPA) and its amendments, the Archaeological Resources Protection Act (ARPA), the American Indian Religious Freedom Act (AIRFA), the Native American Graves and Repatriation Act (NAGPRA), 36 CFR 800, and FSM 2360.

Roads

Existing Conditions Road construction on the Lolo NF is generally associated with timber sales. Road construction however has decreased with reduced timber harvests. There are currently 6,400 miles of Forest Development Road (FDR) under Lolo NF jurisdiction. Thirty-six percent of these roads are closed for wildlife and other resource protection, or to provide walk-in hunting opportunities. In addition to inventoried roads, there are about 1,800 miles of old logging spurs that were built for temporary project needs. These kinds of roads are now usually closed, obliterated or recontoured and revegetated as soon as the project is completed.

The arterial road system for the forest is nearly complete, more than three quarters of the collector system is in place. Less than one quarter of the local project road system is completed. Table 3-34 below shows annual miles of construction and reconstruction projected in the forest plan compared to miles actually contracted in fiscal year 1998.

Table 3-34 Annual Miles of projected/actual Construction and Reconstruction Type of Road Average Annual Projected Actual FY 2004 Arterial & Collector / Local 89/174 103 Method of Analysis Effects of the alternatives on roads was conducted by a literature search of the role of road in weed spread, review of weed mapping projects and observations of roadside weed infestations both on and off the Lolo NF.

Environmental Consequences - Roads

Direct and Indirect Environmental Consequences Common to Both Alternatives The introduction, establishment, and spread of weeds throughout Montana are closely tied to transportation corridors (Chicoine 1984; Forcella and Harvey 1983; Losensky 1989). A study conducted on spotted knapweed by Montana State University indicated that a vehicle

Page 158 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences driven several feet through a knapweed site acquires up to 2,000 seeds (Trunkle and Fay 1991). These seeds are dispersed along highways, with about 200 seeds remaining on a vehicle after driving 10 miles.

Forest road corridors are high-risk sites for introduction, establishment, and spread of weeds on and off Lolo NF lands. Weeds can be carried on vehicles, in the loads they carry, and on construction and maintenance equipment. They can be inadvertently introduced into rights-of-way during maintenance, reconstruction and construction projects by use of contaminated construction equipment, mulch, soil or gravel or revegetation seed. Once established on roadsides, weeds spread quickly into un-infested areas. The 6,400 miles of FDRs under Lolo NF jurisdiction traverse many lands that are susceptible to weed invasion. In addition, roads on adjacent lands may be infested with weeds that can serve as sources for spread onto the Lolo NF, surrounding National Forest, counties and states.

Whether weed infestations are limited to road shoulders or expand out beyond the road into adjacent wildlands, roadside weed control is a critical element in reducing the spread of weeds. Regardless of whether weeds are limited to roadsides or expand into adjacent wildlands, weed control along roads 1) reduces the spread of already present weeds onto un-infested roads and wildlands and 2) can prevent new weed species from establishing on either weed free or already weed infested roads and adjacent wildlands.

Direct and Indirect Environmental Consequences Specific to Alternative 1 Under this alternative weed control along roadsides would be limited to roads in existing weed management decisions. As existing weeds spread into un-infested areas and new weed species are introduced, control would be delayed for as long as it takes to conduct environmental analysis, make a decision, and deal with potential appeals and/or lawsuits. This delay would allow weeds to spread further along roads and into adjacent wildlands. This in turn would increase the cost and complexity in comparison to controlling weeds when the infestations are small and/or localized. On and off road resource impacts from weeds would increase.

Direct and Indirect Environmental Consequences Specific to Alternative 2 Under this alternative, weed managers would be able to reduce existing weed species along more miles of roadside and consequently better protect uninfested roads and adjacent wildlands. New weed species could also be controlled in a timelier manner, which in turn would protect more acres from weeds and from new invaders than under Alternative 1. The delay between discovery and treatment would be less than under Alternative 1.

Cumulative Effects Alternative 1 would have continually increasing cumulative adverse effects (in regards to the resource effects weed spread) as time progressed. Small and new weed infestations would grow with less control than under Alternative 2 and cumulatively distribute the adverse effects of weeds onto more acres. Alternative 2 would potentially protect more acres; weed managers would have a greater “menu” of infested sites to develop a program

Final Integrated Weed Management EIS Page 159 Chapter 3 Affected Environment and Environmental Consequences from and would be able to concentrate weed treatments on roads where the benefit was greatest rather than where there were existing decisions. Alternative 2 would reduce the chance of new invaders from cumulatively establishing both on and off Lolo NF lands. Alternative 2 would also better comply with existing weed policy and law.

Improved weed management along Lolo NF roads under Alternative 2 would improve the cumulative benefits resulting from new roadside vegetation management plans developed or developed by County and State agencies. The Lolo NF could be more responsive in terms of coordinated agency roadside weed and vegetation management. The Montana Department of Transportation, for example, completed a new roadside weed and vegetation management plan in 2002 and the Missoula County Weed District and Missoula County Department of Public Works are also cooperatively in the final stage of finishing a Missoula County roadside vegetation management plan. Improved roadside weed management by the Lolo NF would enhance these other agency weed management efforts.

Forest Plan Consistency Both alternatives would be consistent with the Lolo NF Plan weed management direction.

Wildlife

Existing Condition The Lolo National Forest provides habitat for over 370 different species of wildlife (P. Sweeney, WMPZ, Lolo NF species List 2005, project files). The presence or absence of these wildlife species depends on the amount, distribution and quality of each animal’s preferred habitat. In addition to habitat changes, many of these animals are impacted by hunting or trapping. Montana’s Department of Fish, Wildlife and Parks (MTFWP) regulates game- animal populations. The Forest Service and MTFWP work together to ensure that an appropriate balance is maintained between habitat capability and population numbers. The Forest Service also works closely with the U.S. Fish and Wildlife Service (USFWS) to assist in the recovery of animals listed under the Endangered Species Act (ESA). Proposed federal projects which have the potential to impact species protected by ESA require consultation with the USFWS.

Method of Analysis A number of wildlife species were selected for detailed analysis. The species chosen represent a combination of fine-filter (species specific) and course-filter (management indicator species) analyses. The USFWS requires that endangered, threatened and proposed species be included in effects analysis. The Regional Forester designates sensitive species. Any effects to sensitive species present or potentially present in a project area must be disclosed. Management Indicator Species (MIS) are identified by the Lolo National Forest Plan (USDA Forest Service 1986, pg. II-14) and represent a particular habitat or habitat complex. Each MIS represents a group of species that share common habitat components required for substantial growth and successful reproduction.

Page 160 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Other species that would not be affected by any of the alternatives are reviewed, but not discussed in detail. The wildlife portion of this document is discussed in four sections: • Threatened and Endangered, • Sensitive, • Management Indicator species, and • Neo-tropical migratory birds. Effects of herbicides on wildlife are determined as the relationship between the inherent hazard of the herbicide(s) proposed for use, projections of the animal’s actual exposure to it, and estimates of the dose/response of the animal. These three considerations are modeled and compared to published levels of concern (toxicity endpoints [with additional safety margins included] such as the No Observed Effect Level (NOEL), or the Lowest Observed Effect Level (LOEL). The results of these mathematical comparisons are presented in the form of Hazard Quotients (HQ).

A detailed explanation of the methodology used and the interpretation of the numbers are presented by SERA in web accessible documents and worksheets. For our analyses of risk to wildlife the SERA Risk Assessments (RAs) and worksheets, available at www.sera-inc.com, and at www.fs.fed.us/foresthealth/pesticide/ were used.

The SERA RAs are based on wildlife exposure scenarios. Each RA combines the herbicides inherent hazard, an estimate of exposure, and a dose-response assessment modeled together to generate an estimate of risk for each scenario for each chemical – referred to as the HQ. The herbicides that would be available for weed treatment under each alternative are compared based on the HQ, which is the ratio between the estimated dose (the amount of herbicide received from a particular exposure scenario) and the NOAEL for terrestrial animals and the LD50 for bees. When a predicted dose is less than the NOAEL for terrestrial animals and the LD50 for bees, then the HQ (estimated dose/NOAEL for terrestrial animals and birds and the LD50 for bees) is less than 1.0, and significant toxic effects are unlikely for that specific herbicide application.

Since any RA is based on a number of assumptions, readers and decision-makers should not make the conclusion that the risk values are absolute. If the assumptions are changed, the risk values change. However, the relative risk among herbicides or application methods should remain the same unless new toxicity data becomes available. The number of wildlife exposure scenarios with HQs that range between 1 and 10, and the number of HQ’s that exceed 10 can be found in the project files.

Environmental Consequences - Wildlife

Cumulative Effects The cumulative effects analysis considered the on-going or proposed projects/activities outlined in the Proposed Action (see Chapter 3 Introduction). All projects shown on the maps in Appendix A were considered for each species evaluated; however, it is possible for a particular species that there were no cumulative effects from one or more of these projects.

Final Integrated Weed Management EIS Page 161 Chapter 3 Affected Environment and Environmental Consequences

Management Indicator Species

Applicable Laws, Regulations and Forest Plan Guidelines NFMA provides for balanced consideration of all resources. It requires the Forest Service to plan for diversity of plant and animal communities based on the suitability and capability of the specified land area.

The Lolo National Forest Plan (pages II-13 thru II-14) provides forest-wide standards for management of MIS and habitats on the Forest.

Method of Analysis MIS are species whose habitat is most likely to be affected by forest management activities and serve as indicators of change. There are three terrestrial MIS for the Lolo National Forest (Lolo NF Plan page II-14). Pileated woodpeckers were selected as an indicator for old-growth and mature forests. The northern goshawk is a Region 1 sensitive species as well as an indicator for old growth associated species on dry Douglas fir sites, and is discussed under the Sensitive Species section of this document. Elk are indicators for big game species.

Table 3-35 Management Indicator Species for the Lolo NF Common Name Scientific Name Other Classifications Northern Goshawk Accipiter gentilis R1, USFS, Sensitive Pileated Woodpecker Dryocopus pileatus Elk Cervus elaphus

Northern Goshawk The northern goshawk is a large forest-dwelling hawk. Their prey may include grouse, smaller birds such as jays and woodpeckers, snowshoe hares, and squirrels (Reynolds et.al. 1992, page 4). Reynolds et al. (1992, page 3) identified the 3 components of a goshawk nesting home range as being the nest area, post-fledging area (PFA), and foraging area. Generally, nest areas are composed of older-aged forests with a closed canopy and larger diameter trees located on northern aspects at lower elevations with gentle to moderately steep slopes. PFA’s contain a large percentage of mature forest habitat. Closed crowns forming a matrix enable young fledged birds to branch from one tree to the next and move throughout the forest canopy. Foraging areas are increasingly larger and more diverse than either nesting or post-fledgling areas (Hargis et al. 1994). A diverse complex of vegetation within the foraging area supports a varied and abundant prey base.

The 2005 regional goshawk surveys showed that 40 percent (confidence interval 30 percent-50 percent) of forested lands in roaded areas of the Northern Region are occupied by goshawks. Goshawks are present across the Lolo National Forest and nests have been located in a variety of forested habitat.

The goshawk’s preferred habitat of mature, closed canopy, mixed-conifer forest would not be a target for weed treatment. Goshawks are predators of forest birds and mammals (Reynolds et al. 1992). The herbicides proposed for use in both alternatives have virtually

Page 162 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences no potential for bioaccumulation and are rapidly excreted by animals. No impacts are expected. No further analysis is required.

Pileated Woodpecker Bull and Jackson (1995) and Bergeron (1992) list pileated woodpeckers as both breeding and wintering in the west-central portion of Montana (which would include the Lolo NF). Winter roosting habitat and summer breeding habitat appear to be similar (Bull and Jackson 1995). As noted by Samson (2005), although these woodpeckers are associated with mature forests (Conner 1980, Shackelford and Conner 1997), the presence of large trees for nesting is reported to be more important than forest age (Kirk and Naylor 1996, Giese and Cuthbert 2003). In northwestern Montana, pileated woodpeckers have been found to prefer western larch, but also use Ponderosa pine, black cottonwood, aspen western white pine, grand fir and Douglas-fir as nest trees (in that order; McClelland and McClelland, 1999). Pileated woodpeckers may also forage in young and fragmented forests with abundant remnant (older) structure (Mellon et al. 1992). In summer, ants, particularly carpenter ants (Hoyt 1950), and other insects are obtained on or near the surface of live and dead trees or by extensive excavation in partially decayed wood (Bull et al. 1986, 1992). Data from the northern region land bird program (http://avianscience.dbs.umt.edu/research_landbird.htm) shows the average number of pileated woodpeckers detected at survey points on the Lolo National Forest from 1994- 2000 to be relatively stable. The mature forests that pileated woodpeckers prefer for nesting would not be targeted for noxious weed treatment. No impacts are expected. No further analysis is required.

Elk

Existing Condition The analysis boundary for project impacts to individuals and their habitat is the Lolo National Forest. The boundary for cumulative effects and determining trend is the MFWP elk management unit and the Lolo National Forest, respectively. The Geographical Information System (GIS) map for elk winter range (1999) developed by MTFWP and critical elk winter range (1999) developed by Region 2, MTFWP provide baseline information. Analysis focuses on factors important in big game habitat quality including impacts of herbicide use on big game winter range and elk security. Potential impacts of noxious weed invasion on grazing animals will be measured by projected loss of foraging habitat due to conversion to weeds. Impacts of weed control will be estimated based on national risk analyses and potential effects of herbicides on large mammals. Assumptions for these analyses and potential effects were also reviewed to ensure that they applied to proposed local application of herbicides by the Lolo NF. Disturbance to elk due to implementation of noxious weed treatments is also reviewed.

Elk population ecology, biology, habitat description and relationships identified by research are described in Toweill and Thomas (2002).That information is incorporated by reference. Elk data comes primarily from MFWP data. Additional information used is from recent District wildlife observation records and Forest historical data (NRIS FAUNA).

Final Integrated Weed Management EIS Page 163 Chapter 3 Affected Environment and Environmental Consequences

Elk are a highly adaptable species that annually use a wide variety of habitats including open sagebrush and grasslands as well as all forest types. Nearly the entire Forest provides habitat for elk during some time of the year. Elk generally summer on National Forest lands in the higher elevation mountain ranges and migrate to winter ranges on National Forest, state, BLM, and private lands in lower elevation valleys. They are capable of grazing or browsing a wide range of plants during different seasons, but in Montana, grasses or forbs are a critical dietary component for much of the year (Nelson and Leege 1982). Weeds are typically not eaten by elk at all, or are of generally low palatability. Important winter ranges for elk on the Lolo National Forest normally occur in bunchgrass habitats that are at high risk for weed infestation. Infestations of weeds such as spotted knapweed can lead to 60-90 percent decreases in forage production on winter ranges (Rice et al. 1997, page 628). The decrease in forage production may in-turn result in a decrease in the number of ungulates on weed-infested winter ranges (Trammel and Butler 1995, page 814). Elk populations on the Forest are currently at or above objectives set by the MTFWP.

Maintaining quality of big game habitat on winter range is of primary concern in providing for the elk herds that inhabit the Lolo NF. Habitat use areas for elk that would be targeted for noxious weed control would be lower-elevation, open forest/grasslands and shrub lands considered to be winter range. Approximately 414,000 acres of elk winter range exist on the Lolo National Forest which supports approximately 12,000-15,000 elk (USDA Forest Service 2001, B. Henderson, MTFWP, personal communication). On 70,000 acres (of the 414,000) grass is the predominate forage type (USDA Forest Service 2001). Weed treatments would be proposed on a maximum of 5,000 to 6,000 acres on big game winter range/year under Alternative 2. Other areas on the Forest, which could be treated by ground-based of herbicides would consist mainly of road and power line rights-of-way, administrative sites, etc. where vehicle traffic disturbance occurs regularly. Roads, utility rights-of-way and other sites could be sprayed aerially. A maximum of 15,000 acres/year could be treated for weeds under Alternative 2; the amount of winter range treated annually would vary based on project priorities and funding.

Elk security is another factor important in maintaining big game habitat quality. Elk security consists of areas of hiding cover greater than 250 acres and more than ½ mile from any road open during hunting season (Hillis et al. 1991). These areas give bull elk a higher probability of surviving the hunting season. Implementation of proposed weed treatments would not affect elk security and will not be discussed further in this analysis.

Effects of noise on wild animals can be classified as those affecting auditory physiology and sensory perception, those affecting behavior, and those affecting populations (Bowles 1995). Wild animals are free to move away from a disturbance, while negative impacts on physiology or sensory perception are generally from chronic exposure. Noise generated from weed control operations is likely to be of short duration and sporadic. Ungulates typically avoid areas where noises are present and return when they are not. If noises are of brief duration and the animal has good cover, change in home range size is not detectable. No detectable changes in population size or growth rate have been documented due to noise, as most effects of noise disturbances are relatively mild (Bowles 1995).

Weeds have been shown to completely replace native grass communities (USDA FS 2001, Duncan 1997, Losensky 1987) resulting in a substantial change in ecosystem dynamics. The change in forage production is significant ranging from over 1,000 pounds produced annually by native grass communities to around 100 pounds of forage produced annually

Page 164 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences in sites dominated by spotted knapweed, leafy spurge, and sulfur cinquefoil (USDA Forest Service 2001). As a result of weed invasion, several researchers documented low usage of spotted knapweed-dominated sites by elk and deer (Lavelle 1986, Guenther 1989, Thompson 1996).

Environmental Consequences - Elk

Direct and Indirect Environmental Consequences Common to Both Alternatives Alternatives 1 and 2: Alternative 1 would allow use of fewer herbicides than the proposed action (i.e. impazapyr and triclopyr are not currently being used on the Lolo NF). Effects due to these two herbicides would be considered only under Alternative 2. All other herbicides considered for use are considered in both alternatives.

2,4-D and dicamba have scenarios with a HQ>1 for acute exposure for large mammals eating contaminated vegetation. Herbicides would be used on winter range in the spring or fall, when elk are often no longer present in the area so the plausibility of a 100 percent contaminated diet for a highly-mobile, large mammal is low. Currently 2,4-D is used for control on approximately ¼ of the total acres of weeds treated/year on the Lolo NF. The only application of 2,4-D on winter range would be for treatment of leafy spurge and leafy spurge is only present on about 2,500 acres (1/10 of 1 percent) of the entire Forest. The Lolo NF does not plan any application of dicamba over 1 lb/acre and dicamba is only typically used for spot bead applications (i.e. would be limited in occurrence and area) and not broadcast applications.

If elk were present on winter range where herbicides were sprayed, they would not likely consume 100 percent contaminated vegetation. However, if this scenario occurred, a 100 percent contaminated diet could result in mild signs of systemic toxicity and sub-clinical signs of neurologic toxicity, increased thyroid weight, decreased testicular weight, decreased body weight gain, and damage to several organs.

2,4-D and triclopyr have scenarios with a HQ >1 in the chronic exposure scenarios for large mammals eating contaminated vegetation. The chronic exposure scenarios involving applications of 2,4-D and triclopyr could result in sub-clinical effects.

Aminopyralid is an herbicide recently reviewed and registered (August 10, 2005) for use under the Reduced Risk Pesticide Initiative of the U.S. Environmental Protection Agency. This designation is reserved for compounds that demonstrate lower risk to the environment and animals than marketplace standards. According to the laboratory studies, it has the lowest toxicity of any of the proposed herbicides (acute NOAEL 1,000 mg chemical /kg body weight [based on a rat study], and a chronic NOAEL of 232 mg chemical / kg body weight [based on a dog study] and will be applied at a much lower rate per acre than most of the other proposed herbicides. Based on low toxicity to animals and low application rate, aminopyralid is not expected to result in exposure scenarios with adverse effects on elk. The use of aminopyralid in Alternative 2 will also reduce the need to use 2,4-D.

Other weed treatment techniques included in the alternatives (bio control, reseeding, sheep and goat grazing and ground-based mechanical treatment) are not expected to have an effect on elk.

Final Integrated Weed Management EIS Page 165 Chapter 3 Affected Environment and Environmental Consequences

Cumulative Effects Common to Both Alternatives In addition to herbicide application by Forest Service applicators, ongoing herbicide application will be conducted by other agencies and on private lands within the Lolo NF boundary. These inputs will be dispersed over the forest and are most commonly associated with roads, railroads, or power line corridors. The only herbicide with the possibility of mild toxicity effects on elk that has a high likelihood of use by other herbicide applicators is 2,4-D. Because of the high mobility of large herbivores like elk and the disturbance associated with human travel corridors, the likelihood of a 100 percent herbicide- contaminated diet would be very low. All the herbicides proposed and/or likely to be used are rapidly excreted from the body. Thus, cumulative buildup of herbicides in animal tissue would not be plausible. In addition no synergisms are known for any of the chemicals proposed for use.

Forest management activities and public use of Lolo NF lands will continue to contribute to the spread of weeds. Best Management Practices for weed control (Appendix G) will reduce the impacts of forest management activities; however, well-established weed populations like knapweed will continue to spread and cumulatively reduce winter range forage for elk and other big game species, but less so in Alternative 2. Under the No Action and Proposed Action Alternatives, herbicide treatments on 5,000 to 6,000 and 15,000 acres, respectively, will help reduce weed spread on big game winter ranges and in other areas of the Forest (see list of site types; Chapter 2). Thus, the cumulative increase in weed populations will be lowest for Alternative 2.

Forest Plan Consistency and Regulatory Consistency Alternative 1 and 2 comply with direction in the Lolo NF Plan to provide habitat for increasing populations of big-game animals (FP II-1, Goal #4).

Involve the Forest’s wildlife biologist in reviewing and recommending vegetative objectives for managing and protecting all winter range whenever activity is proposed within it (FP II- 13, Forest-wide Standard #22).

Optimize forage production for elk (FP III-83-112, Goal #1 for Management Areas 18, 19; Goal #2 for MAs 22 and 23).

Neither alternative would change overall cover/forage levels or open road densities.

Statement of Findings Weed treatments for both alternatives will limit expansion of weeds and maintain native grasses that provide forage for big game species such as elk. Since sufficient general forest habitat is available on the Forest, populations of elk and other big game species are expected to remain viable.

Page 166 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Threatened and Endangered Species

Applicable Laws, Regulations and Forest Plan Guidelines Section 7 of the ESA, 1973, states that All Federal agencies are required to undertake programs for the conservation of threatened and endangered species, and are prohibited by the ESA amendment of 1978 from authorizing, funding or carrying out any action that may jeopardize a listed species or destroy its critical habitat.

The National Forest Management Act (36 CFR 219.19) directs the Forest Service to manage habitat to maintain viable populations of existing native and desired non-native vertebrate species. A viable population is defined as one that has the estimated numbers and distribution of reproductive individuals to insure its continued existence is well distributed in the planning area, which for this analysis, is the Lolo National Forest.

Species List – Existing Conditions The U.S. Fish and Wildlife Service determined that the following listed wildlife species are present on the Lolo National Forest (USDI 2005a).

Table 3-36 Federally Listed Species Present on the Lolo National Forest Common Name Scientific Name Status in Analysis “Area Bald eagle Haliaeetus leucocephalus Threatened Endangered or Gray wolf (Northern Rocky Canis lupus Experimental/Non-essential Mountain) (depends on location) Grizzly bear Ursus arctos Threatened Lynx Lynx canadensis Threatened Bald Eagle

Existing Condition The statewide bald eagle population is increasing. Between 1978 (when it was put on the endangered species list) and 1995 the number of known breeding pairs increased from 12 to 166, well above the downlisting goal of 99 breeding pairs cited in the 1986 Bald Eagle Recovery Plan (MTFWP; http://fwp.state.mt.us/wildthings/tande/baldeagle.html) From 1995 through 2004, the number of known successful nests has risen from 129 to 223 and nest success during those years have ranged from 78-85 percent (MTFWP, annual bald eagle summaries, project files).

Most breeding bald eagles are thought to remain within the analysis area year-round although they may move out of their territories in response to food availability; they may have to move to areas along streams and rivers where ice does not limit their ability to fish and/or to areas where carrion is available (K. Dubois, MTFWP, personal communication). There are presently 19 bald eagle nests (that have been active one or more years within the last 10 years) within the National Forest boundary. Most of these nests are located along the Clark Fork River, Clearwater River and larger lakes like Salmon Lake. Often these nest sites are on private lands that are located within the National Forest boundary.

Final Integrated Weed Management EIS Page 167 Chapter 3 Affected Environment and Environmental Consequences

Wintering bald eagles on the Forest are generally found along portions of the Clark Fork and Clearwater Rivers that have not frozen. Wintering eagle surveys are also done each year in conjunction with Montana Department of Fish Wildlife and Parks. Wintering eagle numbers within the boundaries of the Lolo National Forest are generally higher than the number of breeding eagles. Bald eagles migrate into the area from Canada and will winter in the area as long as the rivers and lakes stay relatively free of ice. Eagles can occur at higher densities during the winter because breeding pairs are not defending territories. If/when the rivers and lakes freeze over in the vicinity of the Lolo National Forest, most bald eagles will move to other locations in (or out of) the state where open water is still available.

In Montana, bald eagle nest sites are generally distributed around the periphery of lakes and reservoirs greater than 80 acres (32.4 ha) as well as in forested corridors within one mile (1.6 km) of major rivers [Montana Bald Eagle Working Group (MTBEWG), 1994, pg 2]. The annual breeding cycle from initiation of courtship and nest building through fledging of young usually occurs in Montana from February 1 – August 15 and once fledged, young are dependent on adults for six to ten weeks (MTBEWG, 1994, pg 22 and 3). Bald eagles are opportunists and will prey on fish, waterfowl, small mammals and carrion. Wintering bald eagles in the state occupy areas near unfrozen portions of lakes and free flowing rivers, or upland areas where ungulate carrion and rabbits are available (MTBEWG, 1994, pg 4).

Bald eagles may be affected by a variety of human activities (MTBEWG, 1994, pg 4). Responses of eagles may range from abandonment of nest sites to temporary temporal and spatial avoidance of human activities. Responses may also vary depending on type, intensity, duration, timing, predictability and location of human activities. Individual pairs may respond differently to human disturbances because some birds are more tolerant than others (MTBEWG, 1994, pg 4). Generally, eagles are most sensitive to human activities during nest building, egg-laying and incubation from February 1- May 30 (MTBEWG, 1994, pg 22). Human activities during this time may cause abandonment and reproductive failure. Once young have hatched, a breeding pair is less likely to abandon the nest. However, eagles may leave the nest due to prolonged disturbances, exposing young to predation and adverse weather conditions (MTBEWG, 1994, pg 22). Weed treatment activities have the potential to cause disturbance to nesting bald eagles if they occur within nesting territories.

The MBEWG guidelines recognize four general categories related to the management of habitat for bald eagles. These categories are nesting habitat, perch and roost sites, feeding habitat, and winter habitat. Nesting habitat is typically associated with mature forest stands in close proximity (less than 1 mile) to large bodies of water, including lakes and fourth order streams, which provide a prey base.

Zone 1: Nest Site Area – Includes the area within a ¼ mile (400m) radius of existing and all alternate nests. Eagles are most sensitive to human activity within this area, and will react to intrusion. Minimize nesting disturbance from February 1 to August 31. Aerial and/or ground herbicide application could occur in areas suitable for eagle nests and could cause displacement of nesting eagles if treatment occurs in the nest site area from February 1 to August 31.

Page 168 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Zone 2: Primary Use Area – Includes the area within a ½ mile radius (800 m) from the occupied and alternate nests in the territory. Area is heavily used by the nesting pair. Aerial herbicide application has the potential to be disruptive to nesting/foraging eagles if treatment occurs from February 1 to August 31.

Zone 3: Home Range – represents all areas used by the eagles during the nesting season. In the absence of behavioral data, the area with a 2.5 mile-radius (4km) of the nest site should be considered as a minimum home range. It should include a 200 foot or sight-distance buffer along foraging habitat. Aerial application occurring within line of sight of eagle nests has the potential to cause displacement of nesting eagles from February 1 to August 31. Studies by Grubb and King (1991) and Grubb et al. (1992) indicate that helicopter use over ¾ mile from nest sites did not displace bald eagles.

Method of Analysis The analysis area lies within the boundaries of the Lolo National Forest. The Forest falls within the Upper Columbia Basin Management Zone (Zone 7) of the Pacific Bald Eagle Recovery Area (USDI, Fish and Wildlife Service 1986). Based on the Montana Bald Eagle Management Plan (MBEMP) (MBEWG 1994), four general habitat management guidelines are considered for bald eagles including nesting habitat, perch and roost sites, feeding habitat and winter habitat.

Environmental Consequences – Bald Eagle

Direct and Indirect Environmental Consequences Common to Both Alternatives Potential impacts on bald eagles from Alternatives 1 and 2 include exposure to herbicides through diet and disturbance caused by herbicide treatment efforts.

All of the herbicides proposed for use have virtually no potential for bioconcentration. (SERA 1997-2004) Exposure scenarios were evaluated for all the proposed herbicides using the SERA worksheets for each of the herbicides. These worksheets, and their output tables, are available in the project files. At the rates proposed for use, no significant effects are projected for any of the herbicide application scenarios.

Direct and Indirect Effects - Effects common to Alternatives 1 and 2: Activities in addition to herbicide treatment, (including bio control, reseeding, sheep and goat grazing and ground- based mechanical treatment) are not expected to have an effect on bald eagles based on mitigation measures (Chapter 2).

Direct and Indirect Environmental Consequences Specific to Alternative 1 Herbicides already being used to treat weeds, on a limited basis on the Lolo, include 2,4-D, chlorsulfuron, clopyralid, dicamba, glyphosate, imazapic, metsulfuron methyl and picloram. At the rates of application proposed for use, all HQs projected for predatory birds with a diet of 100 percent contaminated foodstuff are below the level of concern (a HQ <1). In addition, in the chronic scenario, only 2,4-D was projected to exceed a HQ of 1.

Final Integrated Weed Management EIS Page 169 Chapter 3 Affected Environment and Environmental Consequences

The plausibility of exposure to contaminated prey over the lifetime of a predatory bird in the application scenarios proposed for implementation using this analysis is very low. The short half-lives of the herbicides being considered coupled with the limited acreages proposed for treatment limit the availability of contaminated foods, and mitigate the conclusion that there is risk in this chronic exposure scenario.

The potential for disturbance from the weed spraying activities on nesting bald eagles would be low to non-existent because spraying activities would be restricted in the nesting and primary use areas of bald eagle pairs. Any helicopter spray operations would be limited to areas outside the home range area of nesting pairs.

Use of these herbicides on approximately 5,000 to 6,000 acres/year would result in no change in habitat effectiveness because of the low-level of disturbance and dispersed nature of the weed control measures proposed.

Direct and Indirect Environmental Consequences Specific to Alternative 2 In addition to the herbicides currently being used on the Lolo, this proposal would include the use of imazapyr and triclopyr. At the rates of application proposed for use, all HQs projected for predatory birds with a diet of 100 percent contaminated foodstuff are below the level of concern (a HQ <1). In addition, in the chronic scenario, only 2,4-D and tryclopyr were projected to exceed a HQ of 1. The plausibility of exposure to contaminated prey over the lifetime of a predatory bird in the application scenarios proposed for implementation using this analysis is very low. The short half-lives of the herbicides being considered coupled with the limited acreages proposed for treatment limit the availability of contaminated foods, and mitigate the conclusion that there is risk in this chronic exposure scenario.

The use of herbicides on a maximum of 15,000 acres/year, with limited use of some herbicides in riparian areas, would result in no change in habitat effectiveness because of the low-level of disturbance and dispersed nature of the weed control measures proposed.

Cumulative Effects Common to Both Alternatives Alternatives 1 and 2 will not contribute to cumulative effects from herbicide use by other agencies and private landowners because the herbicides included in this analysis have virtually no potential for bioaccumulation in the animal tissues of prey species.

Forest Plan Consistency and Regulatory Consistency The project complies with Forest Plan direction on T&E species that applies to the bald eagle (Lolo FP page II-1 Forest-wide goal #7; pages II-13 and II-14, Forest-wide standards #24 and #27).

Page 170 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

The project is consistent with the Endangered Species Act as evidenced through consultation with the FWS and receipt of concurrence.

The project is consistent with the Bald Eagle Protection Act (16 U.S.C. 668-668C 1978).

Statement of Finding Alternatives 1 and 2 may effect, but are not likely to adversely affect, the bald eagle. This determination is based on: 1) a very low potential for adverse effects from 2,4-D and tryclopyr and 2) spray activities would be limited around active bald eagle nesting territories.

Gray Wolf

Existing Conditions Strategies to protect and recover wolf populations in Montana, as well as the ecology, biology and habitat descriptions are outlined in the Northern Rocky Mountain Wolf Recovery Plan (USDI FWS 1987). The Lolo National Forest supports both endangered and experimental/non-essential wolf packs. A total of sixty-six wolves were reintroduced in Yellowstone National Park and Central Idaho during 1995-1996 (Joe Fontaine pers. comm.). Wolves inhabiting the Missoula District south of I-90 and south of HWY 12 are considered to be part of the reintroduced populations in Central Idaho (status: Experimental/non-essential). Wolves on the remainder of the Missoula District and all other districts on the Forest (north of I-90 and HWY 12) are considered naturally re- colonizing (status: Endangered; see Montana map of wolf distribution; project files). The Gray Wolf Recovery Weekly Progress Report (USDI, Fish and Wildlife Service 2005b) identifies 5 wolf packs that currently have territories within or overlapping the Lolo National Forest. These include the Big Hole (ID/MT), Fish Creek/Burdette Creek (NWMT), Fish Trap (NWMT), Superior (NWMT) and Ninemile (NWMT) wolf packs. Wolf occurrence data comes from recent District wildlife observation records, Forest historical data (NRIS FAUNA), and other agencies (USDI Fish and Wildlife Service, and MFWP).

Wolf territories are variable in size and may range from 60 to 900 square miles and wolf packs may occupy a variety of habitats including grasslands, sagebrush steppes, coniferous and mixed forests and alpine areas. Wolf distribution and habitat use is more closely tied to denning areas and availability of food (especially ungulate prey) than to vegetation cover type (USDI Fish and Wildlife Service 1987, pages 7-8, 72-74). Pups are whelped in a den during the spring (Mech, 1970, pg 123) and moved to a rendezvous site several months later when they are able to leave the den. Pups leave rendezvous sites when they are mobile enough to travel with the pack (Mech, 1970, page 146 – 148).

Current information about wolves on the Lolo NF is included in the Rocky Mountain Wolf Recovery 2004 Annual Report and is incorporated here by reference (USDI Fish and Wildlife Service 2005c). Some areas on the Lolo National Forest that have been invaded by weeds are within wolf pack territories.

Wolves prey on elk, deer, moose, bison, and other ungulates, but elk are their primary prey (USDI et al. 2003, pg 1). Wolves follow big game movements and may concentrate on elk

Final Integrated Weed Management EIS Page 171 Chapter 3 Affected Environment and Environmental Consequences winter ranges or elk calving areas (USDI Fish and Wildlife Service 1993). The Lolo National Forest supports yearlong habitat for elk. The lower elevations are used year-round and provide wintering habitat. The rest of the Forest is used during the spring, summer, and fall until snow conditions force animals to lower elevations. Mule deer, white-tailed deer and elk are the most abundant big game species found on the Lolo National Forest. Moose and bighorn sheep are also found on the Forest; however, in fewer numbers than deer and elk. Together, this mix of big game species provides a good year-round prey base for wolves.

Human disturbance and accessibility of wolf habitats (i.e. road densities) are the principle factors limiting wolf recovery in most areas (Leirfallom 1970, USDI FWS 1978 and 1987 all in Frederick 1991, Thiel 1978). These components are important by maintaining open road density standards required by the Forest Plan as well as maintaining any security habitat recommended in the big game habitat recommendations. This project does not propose any changes in road management.

Herbicide treatment proposed under Alternatives 1 and 2 would include treatment of a variety of site types (see Chapter 2). Some of the propose treatment areas would be within wolf pack home ranges. Den and rendezvous sites or areas of pack concentration known to occur on the Forest would be avoided by ground and aerial spray operation.

Method of Analysis The analysis boundary for project impacts and cumulative effects on individuals and their habitat is the Lolo National Forest.

Measurement indicators of wolf habitat that provide survival and recovery value and which were used for this analysis are 1) adequate prey base and 2) habitat security/minimizing mortality risk. Additional indicators essential for recovery and which are also analyzed include special habitats such as dens and rendezvous sites. These special habitats are important to the wolf and should be managed for low disturbance levels. These indicators are in accordance with the Northern Rocky Mountain Wolf Recovery Plan (USDI 1987, pg 7).

Environmental Consequences – Gray Wolf

Direct and Indirect Environmental Consequences Common to Both Alternatives Three potential factors resulting from Alternatives 1 and 2 which could affect wolves include exposure to herbicides, benefits to prey species from a reduction in weed infestations, and disturbance from weed treatment efforts.

The methodology for herbicide analysis is discussed under the MIS section (elk) and exposure scenarios based on proposed application rates of the various herbicides can be found in the project files.

Treatment of weeds on winter ranges and in fire areas could result in higher quality and more evenly distributed forage for big game prey species. This could result in increased productivity of prey species.

Page 172 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

The other potential effect from herbicide application is disturbance caused by ground- based or aerial application. Both aerial and ground application methods have the potential to displace individual wolves from an area; however, the disturbance would be short- duration (2 hours to 3 days) and no more than annually, in any given drainage. Known rendezvous areas would be protected by a no-fly, one mile buffer in aerial spray operations.

Direct and Indirect Effects - Alternatives 1 and 2: The acute herbicide exposure scenario is based on a medium-sized carnivorous mammal that consumes small mammal prey that has been contaminated by direct spray, assuming 100 percent absorption for the prey. Doses to a larger carnivorous mammal like a wolf would be even lower on a per kilogram body weight basis. Of the eleven herbicides proposed for application on the Lolo NF, only two – 2,4-D and dicamba have the potential for adverse effects based on an acute exposure scenario. The scenario includes the assumption that the wolf will consume a diet consisting solely (100 percent of its diet) of small animals that received a direct spray with 100 percent absorption rate. While this is the simplest worst-case scenario the plausibility of its occurrence is very low. Because wolves have a relatively large home range in comparison to the proposed weed treatment areas, it is more likely that only a small proportion of a wolf’s diet might consist of prey that have been sprayed directly with herbicide.

The chronic herbicide exposure scenario is based on a medium-sized carnivorous mammal that consumes small mammal prey that has been contaminated by direct spray throughout its lifetime. Nine of the eleven herbicides proposed for use when evaluated using this scenario had a HQ < 1. However, chronic dietary exposures from the evaluated application rate of 2,4-D and triclopyr could result in a HQ > 1. As for the chronic scenario, the plausibility of a wolf having available a source of, and consuming, contaminated prey with the regularity required for chronic toxic effects is very low.

Other weed treatment methods (bio control, reseeding, sheep and goat grazing and ground-based mechanical treatment) are not expected to have an effect on wolves assuming implementation of mitigation measures (Chapter 2).

The effects of controlling the spread of invasive plant species would have beneficial effects on ungulate prey species. Alternative 2 would have the greatest potential to benefit ungulate species because it could result in treating invasive plants on up to 2-3 times the acres as treated under Alternative 1. If restored to native plant communities, these areas would provide greater amounts of palatable forage species (ungulate prey) for the wolves.

Cumulative Effects Specific to Alternative 1 Well-established weeds, like spotted knapweed will continue to spread with human-caused or natural disturbance and cumulatively reduce winter range forage for elk and deer. This will occur at a higher rate under Alternative 1 because fewer acres could be treated.

Cumulative Effects Specific to Alternative 2 Under Alternative 2, weed treatments could occur on up to 15,000 acres/year and will help reduce weed spread along human travel corridors and more remote wildlands. The Proposed Action Alternative will also improve the Lolo NF’s ability to reduce the spread of established weed populations on winter ranges, thus the cumulative reduction in weed spread and an improvement in forage quantity.

Final Integrated Weed Management EIS Page 173 Chapter 3 Affected Environment and Environmental Consequences

Forest Plan and Regulatory Consistency The project complies with Forest Plan direction on T&E species that applies to the bald eagle (Lolo FP page II-1 Forest-wide goal #7; pages II-13 and II-14, Forest-wide standards #24 and #27).

Endangered Species Act: The project is consistent with the Endangered Species Act as evidenced through consultation with the FWS and receipt of concurrence. Northern Rocky Mountain Wolf Recovery Plan (USDI FWS 1987): The project is consistent with the Wolf Recovery Plan as it will help maintain productivity of winter ranges and ungulate numbers.

Statement of Findings Alternatives 1 and 2 may affect, but are not likely to adversely affect the gray wolf or its habitat based on: • A very low risk of acute adverse effects from 2,4-D and dicamba and chronic dietary effects from 2,4-D and triclopyr. • Long-term cumulative adverse effects on native plant communities (if limited treatments in Alternative 1 occur) which provide high quality and quantity of forage for big game (prey for the wolf). • Mortality risk to the wolf is not expected to change during proposed activities. • Alternatives will not affect known denning/rendezvous sites. • Short-term avoidance of areas of activity (although transient use could still continue). • Alternatives meet forest plan big game management recommendations. Grizzly Bear

Analysis Area and Methods The analysis boundary for project impacts to individuals and their habitat are the bear management units in the recovery zones and grizzly bear occupied habitat (USDA et al. 2002). The boundary for cumulative effects and making the effects determination are the recovery zones and/or the Lolo National Forest.

Factors considered in the effects analysis for grizzly bear include: • Access management. • Food/garbage storage. • Livestock grazing. • Linkage Zones. • Vegetation management.

Page 174 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Environmental Consequences – Grizzly Bear

Existing Condition The grizzly bear was once found throughout much of the lower 48 states west of the Mississippi river. Currently their distribution is restricted to five discreet populations in established recovery zones: The Greater Yellowstone Ecosystem (GYE) in portions of Montana, Wyoming and Idaho; the Northern Continental Divide Ecosystem (NCDE) in Montana, the Cabinet-Yaak Ecosystem (CYE) in Montana and Idaho; the Selkirk Mountains in Idaho and Washington; and the North Cascades in Washington (Servheen 1993, pgs 11- 13).

The Lolo National Forest provides important habitat for grizzly bears in parts of both the NCDE and CYE. Over 580 grizzly bears are estimated to be in the NCDE (USDI Fish and Wildlife Service 2005d). The first study to determine the actual number of grizzly bears in the NCDE, through DNA analysis of hair samples, was initiated by the U.S. Geological Survey in 2003 and is scheduled to be completed in 2006. More grizzly bears have been consistently observed adjacent to the NCDE in recent years [in what has been identified as Occupied Habitat (USDA et al. 2002)]. The minimum number of grizzly bears estimated to be in the CYE is 35 (Kasworm et al. 2005, pg 11). The probability that this population is declining is 89.4 percent (Kasworm et al. 2005, pg. 57).

Grizzly bears are large omnivores that typically utilize a wide variety of foods. Vegetation such as roots, tubers, bulbs, berries, nuts and green herbaceous plants are seasonally important to grizzly bears. Additionally, high calorie animal food sources such as ungulates, ground squirrels carrion, fish and insects are highly valuable to them when they can be obtained (Servheen 1993, pg. 7). To utilize such a wide variety of foods, bears use a wide variety of vegetation types spread out over large distances. These vegetation types include lower elevation sagebrush/grass or Douglas-fir stands as well as higher elevation lodgepole pine, Engelmann spruce/subalpine fir and whitebark pine forests.

Grizzly bears occasionally occur outside the recovery areas and occupied habitat on the Lolo National Forest. During early summer 2001, a young male grizzly bear was present in the Ninemile Valley and in during the fall of 2002 a grizzly bear was observed near the head of the Rock Creek drainage on the Missoula Ranger District. The bear in the Ninemile Valley was eventually destroyed and the location of the bear that was in Rock Creek is unknown (Jonkel, pers. com.). During the fall bear season, a grizzly bear was accidentally shot in North Kelly Creek, in Idaho, approximately 3 miles from the Forest boundary in the bitterroot Ecosystem.

Access management Seven subunits of the NCDE recovery zone lie on the Lolo National Forest. Five subunits (North Scapegoat, South Scapegoat, Monture, Morrell-Dunham and Swan) are within the South Fork Flathead Bear Management Unit (BMU) and two subunits (Mission and Rattlesnake) lie within the Rattlesnake BMU. Open motorized route density (OMRD) and total motorized route density (TMRD) standards (USDI Fish and Wildlife Service 1996) for six of the seven subunits within the recovery zone have been met. Reduction in road densities in the Swan subunit is still in progress and the Forest is close to meeting standards for OMRD and TMRD (USDA Forest Service 2004). Within the CYE (Mount Headley Bear

Final Integrated Weed Management EIS Page 175 Chapter 3 Affected Environment and Environmental Consequences

Unit), OMRD and TMRD are below established standards but improvement is expected with road closures in future projects. The Lolo NF will be required to meet OMRD and TMRD standards in this Bear Unit by 2013.

The area where grizzly bears are now thought to reside outside the NCDE recovery zone on the Lolo National Forest consists of approximately 71,060 acres of National Forest System lands (USDA Forest Service 2004). Road densities in this area, identified as “occupied habitat” are higher than in the recovery zones. Standards for road density levels are not required outside the recovery zones, but road-related management activities within occupied habitat are being monitored according to USFWS reporting requirements (USDI Fish and Wildlife Service 2004).

Food Storage: A food storage order within the NCDE, on the Lolo National Forest, has been enforced since 1998. A Food Storage Order is also enforced in the Rattlesnake Wilderness and a new Food Storage order has been signed (2005) for the entire Seeley Lake Ranger District outside of the NCDE. All campgrounds and trailheads within the Mount Headley Bear Unit of the CYE are pack-it-in/pack-it-out and signs are posted to encourage proper food/garbage storage. The Forest has developed a Forest-wide Voluntary Food Storage Strategy. The strategy encourages proper food and garbage storage not already addressed in Food Storage Orders currently in effect on the Forest.

Livestock Management: During the 1990’s many sheep and livestock allotments on the Lolo National Forest were closed. In 1995, an amendment was made to the Forest Plan that closed a number of allotments. Review of Allotment Management Plans (AMPs) continue on a scheduled basis and it is possible (10 allotments were closed in 2007) that additional allotments may be closed in the future. At this time there are no sheep allotments on the Lolo National Forest and many of the cattle allotments that are active have relatively low stocking rates. The Two Creek allotment is the only cattle grazing allotment for the NCDE portion of the Lolo. One special use permit (for a few horses) and one livestock allotment (for 15 cow/calf pairs) occur within the Mount Headley Bear Unit of the CYE.

Linkage Zones: Linkage zones are the main habitat factor concerning grizzly bear habitat outside of recovery areas and occupied habitat. Grizzly bear populations primarily exit in large isolated islands of habitat such as the NCDE, CYE and Yellowstone Ecosystem. Human developments such as roads, highways, railways, clearing for agriculture, and commercial and residential construction have divided what was once contiguous habitat between these ecosystems. Providing linkage zones for grizzly bear habitat fragmented by major highways is important to grizzly bear management on the Lolo National Forest.

Vegetation Management: Weed treatments are proposed within grizzly bear recovery zones and within designated habitat. There are less than 4,100 acres within the NCDE recovery area on the Seeley Lake Ranger District and 1,300 acres within the CYE recovery area on the Plains-Thompson Falls District that are identified for weed treatment under the Proposed Action Alternative. Approximately 20,000 acres within occupied habitat may be treated in addition to those treatments proposed within the recovery zones.

Page 176 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Direct and Direct Environmental Consequences Common to Both Alternatives Aerial or ground application of herbicides would not change existing access management, food storage orders, livestock management or linkage zones for grizzly bears under Alternatives 1 and 2.

Three potential factors resulting from the proposed vegetative management action, however, could affect grizzly bears, including exposure to herbicides, benefits to preferred forage species from a reduction in weed infestations, and disturbance from weed treatment efforts.

Grizzly bears would have to be exposed to high levels of most of the herbicides to result in an adverse effect. The grizzly habitat most likely to be treated includes road beds in core areas. SERA risk assessment worksheets were used to determine HQs for a large mammal; it assumes contaminated vegetation is 100 percent of the animals’ diet. Spring foraging habitat is the main area where bears would be grazing on 100 percent grasses and forbs and would consequently have the potential to be exposed to the highest levels of herbicide residues. Of the herbicides with HQ >1 (under the acute scenarios),- only two dicamba and 2,4-D – have the potential to be used in spring habitat for grizzly bears. Consumption of 100 percent contaminated vegetation is unlikely because bears are highly mobile and heavy weed infestations would not be attractive as a food source. Dicamba is used primarily for spot treatments in backcountry areas and 2,4-D will be used along roadside and for limited broadcast treatments.

Inside Recovery Zone Proposed treatments under Alternatives 1 and 2 may involve ground vehicle traffic on restricted roads but no change in OMRDs or TMRDs is expected for any of the 8 BMUs potentially affected because treatments would occur within the permitted administrative use window. Aerial application using a helicopter could be used on restricted roads that can not be accessed with ground-based equipment in and outside of core within those BMUs. Aerial application would only be used in the event that a large, widespread weed infestation is discovered and cannot be effectively or economically treated with ground-based equipment or biological control agents. Treated areas would have reduced foraging capacity for grizzly bears if non-target plants would be killed by broad-spectrum herbicides. Use of narrow spectrum herbicides and/or application of herbicides during less impactive seasons could mitigate this potential effect. A return of and increase in foraging capacity would occur within 2-3 years of herbicide treatment (Rice et al. 1997, p. 631).

A wide range of impacts to wildlife due to aircraft over-flights have been reported in the literature. Grizzly bears have been noted to panic and flee areas from over-flights in nearly all cases where they have been observed (USDI 1994). In Glacier National Park observations have indicated a similar flight response (ibid). Little research has been conducted regarding long-term impacts of frequent over-flights; however, indications are that frequent and repeated over-flights may impose a burden on the energy and nutrient supply for animals (ibid). For this reason, a conservative design criterion is proposed: restrict low-level aerial herbicide spraying operations to 2 days per Bear Management Unit (a likely home range for a female grizzly bear) per year. A similar restriction was imposed

Final Integrated Weed Management EIS Page 177 Chapter 3 Affected Environment and Environmental Consequences under the Final Conservation Strategy for Grizzly Bears in the Yellowstone Area (IGBC 2003, p. 43) and is proposed for weed treatment purposes on the Kootenai National Forest.

To minimize potential for grizzly bear mortalities due to implementation of weed treatments, no sheep or goat grazing will be authorized within recovery zones.

Occupied Grizzly Bear Habitat The main effect would be possible disturbance and/or displacement any ground or aerial weed treatments. Proposed herbicide treatments under Alternatives 1 and 2 would involve ground application along open and restricted roads. During treatments, restricted roads would remain closed to the public. Thus, no change in road densities would be expected. Aerial application would include treatment of winter ranges and areas affected by past wildfire. Adjacent displacement areas will be available in adjacent undisturbed areas. To minimize potential for grizzly bear mortalities due to implementation of weed treatments, no sheep or goat grazing will be authorized within occupied grizzly bear habitat.

Direct and Indirect Effect - Effects common to Alternatives 1 and 2: Activities in addition to herbicide treatment, (including bio control, reseeding, sheep and goat grazing and ground- based mechanical treatment) are not expected to have an effect on grizzly bears assuming implementation of mitigation measures (Chapter 2).

Direct and Indirect Environmental Consequences Specific to Alternative 1 Alternative 1: Under this alternative most herbicide application would be associated with open roads. The existing core areas provide displacement habitat for on-going projects. This alternative will not cause additional incidental take because OMRD, TMRD, and core standards will not be changed in any of the affected BMUs. No additional incidental take is expected because baseline linear open and total road densities are maintained in all of the affected BMUs and occupied habitat.

Direct and Indirect Environmental Consequences Specific to Alternative 2 Alternative 2: Under this alternative most herbicide application would also be associated with open roads. These treatments would result in no decrease in habitat effectiveness for the grizzly bear. Aerial spray related disturbance would be short-term (2 days or less) and only one annual disturbance per BMU. Displacement of bears is likely within one mile of the flight paths used in aerial operations. Design criteria will provide adjacent undisturbed core areas for displaced bears. No long-term adverse effects are expected. Additional incidental take is not expected because OMRD, TMRD, and core standards will not be changed in any of the affected BMUs and terms and conditions for road densities in occupied grizzly will be met.

Cumulative Effects In addition to herbicide application by Forest Service applicators, ongoing herbicide application will be conducted by other government agencies and on private lands within

Page 178 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences the Lolo National Forest boundary. These inputs will be dispersed over the forest and are most commonly associated with roads, railroads or power line corridors. The only herbicides with HQs >1 that have a high likelihood of use by other herbicide applicators are 2,4-D and dicamba. Because of the high mobility of grizzly bears and the disturbance associated with human travel corridors, the likelihood of a 100 percent herbicide- contaminated diet would be very low. All the herbicides proposed and/or likely to be used are rapidly excreted from the body, thus cumulative buildup of herbicides in animal tissue would be unlikely.

Forest management activities and public use of forest lands will continue to contribute to the spread of weeds. Best Management Practices for weed control (Appendix G) and mitigation measures for weed treatments (Chapter 2) will help minimize the impacts of forest management activities, however, widespread weeds like spotted knapweed will continue to spread with human-caused and natural disturbance and cumulatively reduce spring forage for grizzly bears. Although both alternatives would reduce impacts of weeds, Alternative 1 would treat approximately 1/3 of the maximum acres possible in Alternative 2 and would not reduce weed spread as effectively over time. Alternative 2 will be the most efficient in helping reduce the spread of established weed populations on winter ranges (spring range for bears). The cumulative increase in weed populations will be lowest for Alternative 2.

A large variety of human activities – timber sales, livestock grazing, recreation, etc. – occur on the Lolo National Forest, and will continue into the future. Grizzly bear access management in the recovery zone is designed to balance these effects by providing core habitat characterized by a low level of human activity. Each planned activity involves analysis to determine and account for impacts to core areas. Aerial spraying in core habitat proposed under Alternative 2 could temporarily displace grizzly bears from localized areas. However, cumulative effects resulting from such actions would be discountable, due to their short duration and localized nature. Adjacent areas of core habitat would continue to be managed to provide secure grizzly habitat.

Forest Plan and Regulatory Consistency The project is in compliance with ESA. This statement is based on: 1) this project meeting all terms and conditions established in consultation on the Lolo Forest Plan (UDSI 1996, USDI 2001, USDI 2004) and 2) consultation with FWS subsequent concurrence.

Statement of Findings Alternatives 1 and 2 may affect, but are not likely to adversely affect, the grizzly bear. This determination is based on: • Short-term non-recurring disturbances in core habitat. • No changes in road densities. • Short-term reductions in spring forage occurring on and along roads and on big game winter range. • Long-term benefits of reducing established weed populations. • No use of sheep or goats for weed treatment in grizzly bear recovery areas.

Final Integrated Weed Management EIS Page 179 Chapter 3 Affected Environment and Environmental Consequences

Lynx

Analysis Area and Methods The Lynx Conservation Assessment and Strategy (LCAS; Ruediger et al. 2000) Project planning objectives and planning guidelines (pp. 7-1 thru 7-17) were considered and used, when appropriate to the proposed project activities, during alternative development.

The effects analysis follows the standards and guidelines established in the LCAS (Ruediger et al. 2000). Only the standards and guidelines applicable to the proposed project are analyzed, and they are only applied to lynx habitat on Federal lands (in compliance with the LCAS). Those considered but found “not applicable” are found in the project file.

The scale for direct effects analysis is the impacted Lynx Analysis Unit(s) (LAU). The scale for indirect effects is the impacted LAUs and adjacent LAUs for connectivity effects.

Existing Condition The final lynx listing rule (Clark 2000) gives population and habitat status on a national scale. Lynx occurrence data comes from Forest historical records (NRIS Fauna) and other agencies (MNHP, MFWP, USFWS).

Optimal lynx habitat can generally be described as a mosaic of early-successional forest stands for foraging and late-successional forests with deadfall for security cover and denning habitat (Koehler and Aubry 1994, p. 86). Lynx inhabit the mid to high elevations where snow excludes most other predators during winter. Denning habitat occurs most often in subalpine fir forests where there is a high amount of down material (Koehler and Aubry 1994, p. 89). Snowshoe hares are the primary prey for lynx. Primary forest types that support snowshoe hare are subalpine fir, Engelmann spruce, Douglas-fir and lodgepole pine. The key component of snowshoe hare habitat is dense understory vegetation. In winter, lynx forage for hares in vegetation that provides a high density of young conifer stems or branches that protrude above the snow (Ruediger et al. 2000, p. 1-4 and 1-7). Snowshoe hares appear to avoid clear-cuts and very young stands (Ruediger et al. 2000, p. 1-7).

Lynx habitat and weed infestations generally do not overlap, because lynx are typically found in dense forested stands in which weeds are not able to compete with native vegetation. Approximately 37,600 acres of lynx habitat (subalpine fir forests above 4,500 feet elevation) are proposed for weed treatment. Generally, only forest openings, previously harvested (generally clear-cut that have not yet regenerated enough for weeds to be shaded out) or burned subalpine fir would be proposed for treatment.

The following table identifies where, or if, risk factors identified in the LCAS are addressed by this analysis.

Page 180 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Table 3-37 Treatment of Risk Factors as Outlined in LCAS Risk Factors How Treated in the Herbicide Treatment Project? I. Risk Factors Affecting Lynx Productivity: A. Timber Management Cumulative Effects Section B. Wildland Fire Management Cumulative Effects Section C. Recreation Cumulative Effects Section D. Forest/Backcountry Roads and Trails Biological Assessment E. Cattle Grazing Cumulative Effects Section F. Other Human Developments Not Applicable II. Risk Factors Affecting Lynx Mortality: A. Trapping Biological Assessment B. Predator Control Not Applicable C. Incidental or Illegal Shooting Biological Assessment D. Competition & Predation…Human Activities Biological Assessment E. Highways Not Applicable III. Risk Factors Affecting Lynx Movement: A. Highways, Railroads and Utility Corridors Not Applicable B. Land Ownership Pattern Not Applicable C. Ski Area and Large Resorts Not Applicable IV. Other Large-Scale Risk Factors: A. Fragmentation/Degradation of Refugia Biological Assessment B. Movement…Across Shrub-Steppe… Not Applicable C. Habitat Degradation…Invasive Plant Species Biological Assessment Environmental Consequences - Lynx

Direct and Indirect Environmental Consequences Common to Both Alternatives Three potential factors resulting from the two alternatives being considered which could affect lynx include exposure to herbicides, benefits to prey species from a reduction in weed infestations, and disturbance from weed treatment efforts.

Exposure scenarios were evaluated for proposed rates of herbicide application and the results of these analyses are briefly summarized below in the Effects of Action Alternatives. Full disclosure of the analyses is found in the project file.

Another potential effect from herbicide application is disturbance caused by ground-based or aerial weed control activities. Both aerial and ground treatment methods have the potential to displace individual lynx from an area; however the disturbance would be of short-duration and no more than annually in any given drainage.

Final Integrated Weed Management EIS Page 181 Chapter 3 Affected Environment and Environmental Consequences received a direct spray with 100 percent absorption rate. While this is the simplest worst- case scenario the plausibility of its occurrence is very low.

The chronic herbicide exposure scenario is based on a medium-sized carnivorous mammal that consumes small mammal prey that has been contaminated by direct spray throughout its lifetime. Nine of the eleven herbicides proposed for use when evaluated using this scenario have HQs < 1). However, chronic exposures from this scenario for 2,4-D and triclopyr have a HQ > 1. In the worst case acute scenario, the plausibility of a lynx actually having available a source of, and consuming, contaminated prey over its lifetime is very low. Triclopyr and dicamba would not be widely used either within or outside of lynx habitat.

Activities in addition to herbicide treatment, (including bio control, reseeding, sheep and goat grazing and ground-based mechanical treatment) are not expected to have an effect on lynx assuming implementation of mitigation measures (Chapter 2).

The effects of controlling the spread of invasive plant species may have beneficial effects on lynx prey species. Alternative 2 would have the greatest potential to benefit lynx prey species because it would result in treating invasive plants on 2-3 times the potential acres treated under Alternative 1. Treatment of weed infested sites would reduce weeds and increase native and desirable palatable vegetation which would provide forage for lynx prey (small mammals such as snowshoe hares).

Cumulative Effects Alternatives 1 and 2: These alternative’s non-herbicide control methods would have limited results on many weed species which would compete for preferred spring/summer forage for snowshoe hares. Long-term effects on hare populations would be minimal; however, individual hares in heavily infested areas could be adversely affected. If weed populations are not effectively controlled in the long-term, this would be one more factor to consider in future cumulative effects analysis.

Well-established weed populations will continue to spread as a result of both human- caused and natural and cumulatively reduce spring/summer forage for snowshoe hare. This will occur at a higher rate under the no Action Alternative because fewer acres will be treated. Under Alternative 2, weed treatments on up to 15,000 acres/year will be more effecting at reducing weed spread along human travel corridors. The cumulative increase in weed populations and reduction in spring forage quantity and quality will be lowest for Alternative 2.

Alternatives 1 and 2 will not contribute to cumulative effects from herbicide use by other agencies and private landowners because the herbicides included in this analysis have virtually no potential for bioaccumulation in the animal tissues of prey species.

Forest Plan and Regulatory Consistency The project complies with Forest Plan direction on T&E species that applies to the lynx (Lolo FP page II-1 Forest-wide goal #7; pages II-13 and II-14, Forest-wide standards #24 and #27).

Page 182 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

The project is consistent with the Endangered Species Act as evidenced through consultation with the FWS and receipt of concurrence.

Statement of Findings Alternatives 1 and 2 may affect, but are not likely to adversely affect the lynx. This determination is based on: 1) the HQs > 1 for 2,4-D and triclopyr, and 2) possible short- term, non-repetitive displacement from weed control activities.

Sensitive Species

Applicable Laws, Regulations and Forest Plan Guidelines The sensitive species analysis in this document meets the requirements for a biological evaluation as outlined in FSM 2672.42.

Sensitive species are administratively designated by the Regional Forester (FSM 2670.5) and managed under the authority of the National Forest Management Act. FSM 2670.22 requires the maintenance of viable populations of native and desired non-native species and to avoid actions that may cause a species to become threatened or endangered.

Sensitive species are those animal species identified by the Regional Forester for which population viability is a concern as evidenced by a significant downward trend in population numbers, density, or in habitat capability that will reduce a species existing distribution (FSM 2670.5.19). There are 13 terrestrial wildlife species (including 3 amphibians) listed as sensitive for the Region 1 National Forests including the Lolo, and which are discussed in this section.

Protection of sensitive species and their habitats is a requirement of Forest Service Manual 2670. The sensitive species program is intended to be proactive by identifying potentially vulnerable species and taking positive action to prevent declines that will result in listing under the Endangered Species Act.

As part of the National Environmental Policy Act (NEPA) decision-making process, proposed Forest Service programs or activities are to be reviewed to determine how an action will affect sensitive species (FSM 2670.32). The goal of the analysis should be to avoid or minimize impacts to sensitive species. If impacts cannot be avoided, the degree of potential adverse effects on the population or habitat within the project area (and on the species as a whole) needs to be assessed.

The Lolo National Forest Plan provides forest-wide standards for management of wildlife species, including threatened, endangered and sensitive (TES) and MIS and habitats on the Forest (pages II-13 thru II-14). This project complies with the standards and guidelines in Northern Rockies lynx Amendment.

Final Integrated Weed Management EIS Page 183 Chapter 3 Affected Environment and Environmental Consequences

Table 3-38 Sensitive Wildlife Species that are either suspected or known to occur on the Lolo NF* Potential Status in Common Name Scientific Name for Analysis Area Impacts? Peregrine Falcon Falco peregrinus Known No Common Loon Gavia immer Known No Harlequin Duck Histrionicus histrionicus Known No Flammulated Owl Otus flammeolus Known No Black-backed Picoides arcticus Known No woodpecker Northern Goshawk Accipiter gentilis Known No Fisher Martes pennanti Known No Wolverine Gulo gulo Known No Northern Bog Lemming Synaptomys borealis Known No Northern Leopard Frog Rana pipiens Suspected Yes Boreal (Western) Toad Bufo boreas Known Yes Coeur d’ Alene Plethodon idahoensis Known Yes Salamander Western (Townsend’s) Corynorhinus townsendii Known Yes big-eared bat *USFS, Region 1 Sensitive Species List, 2004 The following discussion focuses on the use of herbicides on sensitive species. Activities in addition to herbicide treatment, (including bio control, reseeding, sheep and goat grazing and ground-based mechanical treatment) are not expected to have an effect on any of the sensitive species assuming implementation of mitigation measures (Chapter 2).

Peregrine Falcon Peregrine falcon populations in western Montana appear to be increasing based on the increased number of known nest sites (eyries) (Montana Peregrine Institute 2005). Surveys of potential peregrine falcon nesting habitat are completed on the Forest each year to monitor known nest sites and document new breeding pairs. As of the 2005 nesting season, 8 known active or historic eyries on the Forest with 1 located on the Missoula District, 3 on the Ninemile District and 4 on the Plains District.

Peregrine falcons occupy a variety of habitat but are typically found near water because of the abundance of prey associated with such sites. Nests are generally located <8500’ in elevation, <3000’ from water or a wetland, on a >150 percent slope, and on a cliff ledge that is 3000’ in length and >4000’ in height. Prey consists almost entirely of birds, which are usually taken on the wing.

Environmental consequences – Peregrine Falcon It appears that peregrine falcons are sensitive to human activities, especially those occurring above the nest site. They are more tolerant of activities that occur below the nest site if there is pronounced relief from the valley floor to the nest site (Ellis 1982). Human disturbance at the nest may lead to abandonment and interference with care of the chicks. Guidelines for minimizing disturbance to nesting peregrine falcons are to restrict human

Page 184 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences activities and disturbances in excess of what historically occurred during the nesting season within 1 mile of nest cliffs (USDI Fish and Wildlife Service 1984, page 34). The use of pesticides that persist in the environment and magnify through the food chain also presents a risk to peregrines (USDI Fish and Wildlife Service 1984, pages 9-10). Because peregrines may forage in a variety of habitats, some areas used by these birds for foraging may be at risk of weed infestation while others would not be. Peregrine eyries may be located near weed infestations. There is virtually no potential for bioaccumulation of the herbicides proposed for use over multiple years (SERA 1997-2004) and a “no fly” zone will be in place around an active eyrie between April 1 – August 15, no impacts are expected. No further analysis is required.

Common Loon The breeding range for common loons, in Montana, is primarily restricted to low elevation glacial lakes in the northwest corner of the state. This area supports the highest density of nesting Common Loons in the western U.S. (NHS-NRIS, animal guide). On the Lolo National Forest, common loons are currently found only on selected lakes in the Clearwater River drainage generally between April and October. Seventeen loons (adults and chicks) were documented in the Clearwater River drainage (all on National Forest System lands), and 6 loons were present in the Blackfoot drainage (some of which were on lakes on private land) during the 2005 breeding season (L. Kelly, personal communication). Loons that breed in Montana are thought to winter along the coasts of Oregon and California (NHS-NRIS). Some loons may occasionally over-winter in Montana; this has been documented in Flathead, Lake and Lincoln counties (NHS-NRIS).

Environmental Consequences – Common Loon Herbicide treatments near any bodies of water would involve only herbicides that break down quickly and have a low toxicity to loon prey species such as fish and invertebrates. The herbicides proposed for use in both alternatives have virtually no potential for bioaccumulation and are rapidly excreted by animals. (USDA Forest Service 1992b) No impacts are expected. No further analysis is required.

Harlequin Duck The harlequin duck population winters along the north Pacific Coast, and migrates inland to breed east to the Rocky Mountains. They are found in mountain streams that have boulder or cobble substrate, maintain swift currents and provide adequate “loafing” sites during the breeding season (USDA Forest Service 1992a, Kuchel 1977). Trees and shrubs (the “overstory”) along stream banks do not appear to affect habitat use and, although overhanging vegetation along stream banks may be an important component of Harlequin Duck habitat, the amount of overhanging vegetation along stream banks is variable (MT NHP – NRIS 2005, Diamond and Finnegan 1993, Ashley 1994). Harlequin ducks generally arrive at their breeding sites from mid-April to early-May; females and young may remain on the inland streams until August or early September (Reichel, et al 1997). Females usually return to the same breeding sites each year (Reel et al. 1989) and often nest in log jams or along stream banks behind overhanging vegetation. Harlequins feed on benthic aquatic insects such as stoneflies, mayflies, and caddis flies, and are often seen diving in turbulent stream reaches.

Final Integrated Weed Management EIS Page 185 Chapter 3 Affected Environment and Environmental Consequences

Historically, the harlequin duck was found along numerous streams on the Lolo NF. However, recent survey efforts on the Forest indicate the harlequin duck continues to nest only on one stream on the Seeley Lake Ranger District.

Environmental Consequences – Harlequin Duck Herbicide treatments near streams would involve herbicides that degrade quickly and application rates would be well below toxic concentrations to benthic invertebrates, the main food source for harlequins during the breeding season (USDA Forest Service 1992b). No impacts are expected. No further analysis is required.

Flammulated owl The flammulated owl is a small, secretive owl that is known to occur over a wide geographic area in interior mid-elevation montane forests from British Columbia (Christie and Woudenberg 1997) south, into Mexico (American Ornithologists’ Union 1998). It is an obligate secondary cavity nester, often using cavities originally excavated by pileated woodpeckers or northern flickers (McCallum 1994) in open ponderosa pine, Douglas fir, or mixed species forests. They are nocturnal hunters that feed on moths, beetles, caterpillars and crickets (Hayward and Verner 1994). Flammulated owls show a preference for foraging in relatively open stands of older forests but appear to require some patchy understory growth for roosting near the nest tree (Goggins 1986). Fire suppression in ponderosa pine stands has led to dense Douglas fir understories that make foraging more difficult for these owls.

Approximately 79,826 hectars of flammulated owl habitat is estimated in the Northern Region; approx. 7,003 hectares (11.4 percent) is estimated to be on the Lolo National Forest (Samson 2005). Flammulated owl surveys on the Lolo National Forest (1990, 1993-1995, 2003-2005) indicate these owls are present on all Districts primarily in ponderosa pine/Douglas fir forests. During the 2005 flammulated owl surveys, a total of 45 different flammulated owls were found along 18 transects.

Environmental Consequences – Flammulated owl Flammulated owl foraging habitat in stands with a more open canopy could be in areas at risk of weed infestation and might warrant weed treatment. Most of the herbicides proposed for use are relatively nontoxic to insects, based on toxicity studies of bees (USDA Forest Service 1992b). One herbicide, Dicamba, is moderately toxic to insects; however, field application rates would result in contact doses well below toxic levels (USDA Forest Service 1992b). No impacts are expected. No further analysis is required.

Black-backed woodpecker In Montana, the black-backed woodpecker is found primarily west of the continental divide and east of the divide along the south-west border of the state.

Populations of black-backed woodpeckers appear to be irruptive in nature and correspond with the sporadic abundance of wood-boring beetle larvae and bark beetle larvae (Bull, 1986). These beetles feed on the woody parts of trees. Several studies have shown that

Page 186 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences black-backed woodpeckers prefer ponderosa pine, western larch and Douglas fir that have burned at high to moderate/high severities (Caton 1996, Hejl et al. 2000, Hutto 1995, Powell 2000). Even fire burned areas of less than 100 acres are valuable for these woodpeckers and one black-backed woodpecker nest was found on the Lolo National Forest in a 20 acre burn (O’Connor and Hillis 2000). Black-backed woodpeckers move into a stand shortly after trees are burned; woodpecker numbers usually peak in two or three years (Caton 1996) then begin to decline within five to six years (Hutto 1995). Although black-backed woodpeckers are also found in timber stands affected by insect and disease outbreaks, it is generally thought that black-backed woodpeckers inhabit these areas at a lower density than the intensely burned areas. As noted by Samson (2005) however, few studies have considered all habitats used by black-backed woodpeckers in proportion to availability nor considered the comparative difficulty in observing these woodpeckers in open post-fire habitats versus the more closed and structurally complex live forest environments (Hyot and Hannon 2002).

Post-fire burned habitat increased on the Lolo National Forest since 2000. An estimated 74,000 acres burned on the Lolo in 2000. Of this total, 2,300 acres were analyzed for salvage in the Post Burn Environmental Impact Statement. Due to litigation, only 325 acres (approx.) were salvaged. Approximately 2,500 acres burned on the LNF in 2001 and 2002. No salvage harvest was conducted on the forest. In 2003, approximately 128,000 acres burned on and adjacent to the LNF. Roughly 60 percent of this acreage was on LNF lands. At most, 3 percent of this burned timber has been or will be salvaged.

On the Lolo National Forest, biologists have found black-backed woodpeckers in stands following prescribed burns, site preparation burns, and wildfires. In the spring of 1999, Hugh Powell a graduate student at the University of Montana conducted surveys for black- backed woodpeckers on the Lolo National Forest. He found black-backed woodpeckers in burned stands but did not find them in beetle killed stands. Surveys for black-backed woodpeckers were also conducted on the Lolo National Forest after the wildfires in 2000. Transect surveys conducted in portions of four wildfires on the Ninemile and Superior Ranger Districts from 2001 through 2003 resulted in the discovery of several black-backed woodpecker nests and numerous observations of foraging black-backed woodpeckers (Monson 2003).

Environmental Consequences – Black-backed woodpecker Recent burned areas may be treated for weed infestation and subsequent weed treatments may be proposed in these areas. However, predominant prey species of black-backed woodpeckers would not be affected. No impacts are expected. No further analysis is required.

Northern goshawk The northern goshawk was taken off the Regional Sensitive Species List in July, 17, 2007. It is a management indicator species (MIS) on the Lolo National Forest. See previous section on MIS species for discussion of this species.

Final Integrated Weed Management EIS Page 187 Chapter 3 Affected Environment and Environmental Consequences

Fisher Fishers in northwestern Montana are most often found in moist grand fir and cedar habitat types (Heinemeyer 1993). Mature forests are probably preferred, but younger stands are often used in the winter (Jones and Garton 1994). Banci (1989) believes the best fisher habitats are multi-aged stands interspersed with small openings containing riparian habitats. Complex understory structure with abundant woody debris also appears to be an important habitat factor in habitat selection. Fishers feed on snowshoe hares, porcupines, carrion, squirrels, small mammals and birds (Banci 1989, Powell and Zielinski 1994).

Sighting and trapping records show that fisher are present on the Lolo National Forest. The Montana Department of Fish, Wildlife and Parks (MTFWP) administers state trapping regulations. Most of the Lolo National Forest is within MTFWP Region 2 with all of the Plains-Thompson Falls District within MTFWP Region 1. According to the 2005 Montana Trapping Regulations Region 1 has a quota of 2 fishers and Region 2 has a quota of 5 fishers. Since 1991, a total of 32 fishers have been trapped in Region 1 and 66 fishers have been trapped in Region 2. The trapping quota for fisher in Region 1 has not been met or exceeded every year between 1997-2004 (MTFWP 2006) and the trapping quota for fisher in Region 2 has been met or exceeded every year between 1997-2004 (MTFWP 2004).

Environmental Consequences – Fisher The fisher does not generally inhabit openings such as meadows and grasslands or clear- cuts where herbicide treatments would occur. The herbicides proposed for use in both alternatives have virtually no potential for bioaccumulation and are rapidly excreted by animals (USDA Forest Service 1992b). No impacts are expected. No further Analysis is required.

Wolverine Wolverines are the largest member of the weasel family. Although few studies have been conducted on them, they appear to utilize a wide variety of food sources including carrion, rodents, berries, insects, and birds (Reel et al. 1989, page 32; Ruggiero et al. 1994, page 111-113)). In the western United States they occupy a variety of mostly remote montane habitats throughout the year including alpine areas, boulder and talus fields, mature and intermediate forests adjacent to natural openings, big game winter ranges, and riparian areas (Reel et al. 1989, page 32; Ruggiero et al. 1994, pages 100-115). Extensive travel by wolverines is not unusual and home ranges are typically very large (Ruggiero et al. 1994, page 117). Although wolverine populations have increased in western Montana since the 1920’s, they occur at low densities even where habitat is optimal (Ruggiero et al. 1994, page 103). Even with the little research, wolverine appear to be most dependent on undisturbed denning habitat, big game as a food source and trapping pressure.

The limited research on wolverines indicates that they have a huge home range, possibly between 140 and 770 square miles (Copeland and Harris 1994, Halter 1989). During the study in northwestern Montana, Hornocker and Hash (1981) found most wolverines in medium to scattered timber, and wolverines crossed large clear-cuts but did not appear to hunt in them. Copeland and Harris (1994) found two natal dens in cirque basins, and large boulders and down trees may be important features. Denning wolverines with young appear to be very sensitive to human disturbance.

Page 188 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

In both Montana and Idaho, big game carrion appears to be the major food source with snowshoe hares, squirrels and small mammals making up most of the rest of the diet (Copeland and Harris 1994, Hornocker and Hash 1981). Winter is probably the most difficult time for wolverines to find food. Marked wolverines did not move to big game winter range as expected, but feed on carcasses of animals that probably died during the late summer and fall (Copeland and Harris 1994). The presence of large predators such as mountain lions and wolves, and wounding losses from hunting may be important sources of carrion.

Sightings and trapping records show that wolverines are present on the Lolo National Forest. Since 1991, a total of 33 wolverines have been trapped in MTFWP Region 1 (MTFWP 2006) and a total of 23 wolverines have been trapped in MTFWP Region 2 (MTFWP 2004).

Environmental Consequences – Wolverine The main potential for exposure to herbicide would be on winter range where wolverines forage on big game carrion. Deer and elk that may have consumed herbicide treated vegetation would have very low concentrations and it is highly likely that excretion or biodegradation of the herbicide would occur long before wolverines would be foraging on winter range habitat. No impacts are expected. No further analysis is required.

Northern Bog Lemming The Rocky Mountains of Montana are the southern margin of the global distribution of northern bog lemming. Potential habitat in Montana includes sphagnum bogs, wet meadows, moist mixed and coniferous forests, montaine sedge meadows, krummholz spruce-fir forest with dense herbaceous and mossy understory, alpine tundra and mossy streamsides (MT NHP-NRIS, 9/2005). The northern bog lemming feeds on herbaceous vegetation (grasses, sedges, etc.) and invertebrates (snails, slugs, etc.; West 1999, Foresman 2001) and is thought to have a home range of less than 1 acre (NHP-NRIS, animal guide). The only documented observation of northern bog lemming on the Lolo National Forest was in 1978 in the Rattlesnake drainage (MT Natural Heritage Program; M. Miller, pers. comm.).

Environmental Consequences – Northern Bog Lemming Herbicide treatment of preferred bog/fen habitat would involve application of herbicides that degrade quickly and concentrations typical of field application levels would be well below even low toxicity levels (USDA, Forest Service 1992b). No impacts are expected. No further analysis is required.

Townsend’s Big-eared Bat, Northern Leopard Frog, Boreal (Western) Toad and Coeur d’Alene Salamander

Analysis Area and Methods The boundary of the analysis area for these four species is the Lolo National Forest. Methods for analysis include:

Final Integrated Weed Management EIS Page 189 Chapter 3 Affected Environment and Environmental Consequences

• Presence of these species (and/or their prey) in areas where herbicides will be applied. • Review of impacts of herbicides and surfactants on these species (and their prey) from SERA. Affected Environment

Bats

Western (Townsend’s) big-eared bat The distribution of the western big-eared bat is strongly correlated with the availability of caves or abandoned mine shafts where they winter in communal roosts and where females roost with their young. If suitable roosting areas are available, they occur within a wide variety of habitats from arid pine forests to high-elevation mixed coniferous forests (Reel et al. 1989, page 38). Foraging habitat includes riparian areas, forest edge, and diverse forest stands during early mornings, late evenings and at night. They are insectivorous and feed primarily on moths. The disruption of roosting habitat can cause permanent abandonment of roost sites (Reel et al. 1989, page 39). There is one known roost site on the Plains- Thompson Falls District. Because surveys for these bats have not been completed across the entire Forest, they may be present in other areas. Western big-eared bat foraging habitat could include areas that may receive herbicide treatment for weed infestation.

Amphibians A number of factors that may affect amphibian populations include: timber harvest, grazing, fires and fire management activities, non-indigenous species, road and trail development, recreational facilities and water impoundments, harvest and commerce, habitat fragmentation, fungus outbreaks and other diseases, roadkill, predation, heavy metal pollution and pesticides (Maxell 2000).

Northern leopard frog Northern leopard frogs were once found throughout Montana. However, recent surveys indicate major extirpation of populations west of the continental divide (Werner et al. 2004). Habitat for this species includes wet meadows, cattail marshes and grassy shorelines. Adults remain within a wetland during the summer but may move up to two- tenths of a mile away along wet drainages. During spring and fall migrations and juvenile dispersals, individuals have moved up to 5 miles (Werner et al. 2004).

Historical observations of Northern Leopard Frogs on the Lolo NF were primarily on the western side of the Forest. There have been no documented observations of Northern Leopard frogs on the Lolo NF in the past 5 years (project files).

Boreal (Western) Toad The subspecies of Western toad in Montana is the boreal toad. Boreal toads were once common and widespread in western Montana, but their status in this state is uncertain and recent surveys on several National Forests and Glacier National Park found few breeding

Page 190 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences sites (Werner et al. 2004). The Western toad has declined over the last 25 years in other western states including California, Colorado, Utah, southeast Wyoming New Mexico and Oregon (Maxell 2000, pages 86-87). Boreal toads could be present across the Lolo NF, however, recent surveys on the Forest only documented 15 boreal toad breeding (Bryce Maxell, pers. comm.).

Adult boreal toads are largely terrestrial and can move relatively long distances using coniferous forest, wetland, subalpine meadow, lake, pond and marsh habitats. (Werner et al. 2004). They can breed in almost any clean standing water from roadside ditches to large lakes (Werner et al. 2004). Egg laying usually takes place one to three months after the snow melts (Reichel and Flath 1995).

The most vulnerable site for herbicide exposure for the boreal toad would be in pond and wetland breeding sites. Because boreal toads may inhabit upland areas outside of the breeding season, it is possible that they may occur in areas treated with herbicides.

Couer d’ Alene Salamander The Couer d’ Alene salamander is found in Montana, adjacent to the Idaho/Montana border, from the Bitterroot Valley to the Canadian Border. These salamanders have been documented in about 50 locations in Montana (Werner et al. 2004). This salamander is found on the Plains-Thompson Falls and Superior District of the Lolo National Forest (Maxell et al 2003, district wildlife records).

Coeur d’Alene salamanders depend on a specialized microhabitat, and are limited to springs and seeps, waterfall spray zone and damp stream banks in talus or fractured rocks sites (Werner et al. 2004). Although little is known about this species, it is thought to breed in late summer or fall with egg-laying occurring in April or May (Werner et al. 2004).

The only potential exposure of this salamander to herbicides would be infiltration or runoff of herbicides into occupied stream habitat. It is highly unlikely that exposure to direct herbicide spray would occur since adult salamanders come above ground at night or during moist weather conditions (Werner et. al 2004) when herbicide treatments would not be scheduled. This salamander’s habitat of fractured rock in waterfall spray zones or wet, seeping areas would not be targeted by herbicide treatments.

Environmental Consequences – Bats and Amphibians

Direct and Indirect Environmental Consequences Common to Both Alternatives All of the herbicides proposed for use have virtually no potential for bioconcentration. (SERA 1998-2004) Exposure scenarios for terrestrial animals and birds were evaluated for all the proposed herbicides using the SERA worksheets for each of the herbicides. These worksheets, and their output tables, are available in the project files. At the rates proposed for use, no significant effects are projected for any of the herbicide application scenarios.

Final Integrated Weed Management EIS Page 191 Chapter 3 Affected Environment and Environmental Consequences

Bats Townsend’s big-eared bat would not be exposed directly to herbicide application since they roost during the day when herbicide applications would occur. However, the caterpillar stage of the small moths this bat preys upon may be exposed to herbicides during the treatment of big game winter range. Most studies involving the effects of herbicides on insects involve honey bees, thus the honey bee will be the laboratory surrogate for wild insects such as caterpillars.

Amphibians Amphibians (frogs, salamanders, and toads) spend time both on land and in the water, but breeding typically occurs in water; thus the area of focus for this effects analysis is the potential contamination of breeding sites. Data on toxicity of herbicides to amphibians are limited. Several studies have found that amphibians are less sensitive, or about as sensitive, as fish to other less-toxic herbicides (Berrill et al. 1994, Berrill et al. 1997, Perkins et al. 2000). Consequently, for most of the herbicides proposed for use, the toxicity data available for fish and other aquatic species are used as surrogates.

Direct and Indirect Effects

Bats Insects as a Prey Species for Townsend’s Big-eared bat: The honeybee is an insect of concern to many people so there is research available regarding its response to herbicides. SERA’s worksheets include an exposure scenario for this insect. For all of the herbicides, except 2,4-D, proposed for use on the LNF, hazard quotients related to honey bee exposure are below 1, the level of concern that would warrant further analysis. The exposure of honey bees to 2,4-D compared to the range of LD50s reported indicates that some populations of bees subject to a direct spray could evidence substantial mortality while other populations would not. An additional factor not considered involves interception of 2,4-D by vegetation. This would tend to reduce the level of exposure of insects depending on the proportion of 2,4-D intercepted and thus becoming essentially unavailable to the bee, although caterpillars would still potentially be exposed to this fraction of the herbicide via contact and ingestion. 2,4-D maybe used on winter ranges where leafy spurge is present (on a relatively small portion of the Forest; see Environmental consequences for elk) however, 2,4-D would primarily be used along roadsides. Approximately 62,000 acres of winter range are proposed for treatment under Alternative 2. This is a small portion (15 percent) of the total winter range and only a maximum of 15,000 total acres could be treated annually under Alternative 2. Adequate populations of prey for Townsend’s big-eared bats would still be available even if some unintended losses of caterpillars were to occur. The high LD50 for aminopyralid indicates a low likelihood of potential exposure exceeding the adverse effect level.

Amphibians SERA risk assessments are based on acute exposure scenarios which result when animals are exposed directly and immediately following herbicide application. Chronic exposure scenarios result from exposure to herbicides over an animal’s lifetime. Acute exposure

Page 192 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences scenarios are based on an herbicide spill into a small pond or stream. Chronic studies are based on the ambient water contamination after an entire watershed is sprayed. Most of the acute and chronic exposure scenarios for aquatic species do not have a HQ > 1, which reduces concerns over potential adverse effects to amphibians.

Data provided in the SERA risk assessment indicate that chlorsulfuron, clopyralid, imazapic, imazapyr, metsulfuron methyl, and picloram have acceptably low potential to cause adverse effects in aquatic animals. Data provided in the EPA registration process for recently approved aminopyralid indicates that it is practically non-toxic to fish, frogs, and aquatic insects. Data available for studies of the effects of glyphosate and dicamba indicate that at the highest application rates, the resulting herbicide concentration is below the LC50s by an order of magnitude of 2.5 or 3 (USDA FS 2005). Formulations of glyphosate are used in very limited circumstances. Over the last three years, an average of 0.4 percent of the total acres treated with herbicides on the Lolo NF were treated with glyphosate. Glyphosate was applied over the last three years on an average of 0.00095 percent of the Lolo NF. Dicamba is used on an even smaller acreage and percentage of the forest.

At the highest application rate for triclopyr acute exposure from runoff could adversely affect responsiveness of some tadpoles, increasing the risk of predation (Bautista 2005). However, triclopyr at above about 1.5 lbs ae/ac will have so many unacceptable HQs that it is highly unlikely to be used at even higher rates. Triclopyr is only proposed at substantially lower rates in this analysis.

Unlike other herbicides in this analysis, 2,4-D may be more toxic to some species of amphibians than to fish. Lethal doses of 2,4-D on frogs, toads, and newts are displayed in SERA (1998). The estimated concentration of 2,4-D in runoff and after an accidental spill is well below any dose reported to cause mortality in amphibians (Bautista 2005). However, mortality to amphibians could result from an accidental spill of a large volume of 2,4-D (SERA 1998).

Cumulative Effects Herbicide use by other agencies and private land owners within and adjacent to the Lolo NF will be concentrated mainly along dry pastures, road sides, railroad rights-of-way, and power line corridors. The likelihood of Northern Leopard Frogs, Coeur d’Alene salamanders, western toads, or the caterpillar prey species of Townsend’s Big-eared bat being exposed to herbicides along human-travel corridors is dependent on the availability of suitable habitat. This exposure scenario is non-existent to low. Herbicide labels (which applicators are required to follow) for herbicides that are particularly toxic to aquatic species would direct applicators to observe a buffer around open water sources. Since multiple applications from more than one herbicide applicator in the same location are not planned and habitat of these sensitive species is limited along the major human travel corridors, no cumulative effects from the application of herbicides are expected.

Direct and Indirect Environmental Consequences Specific to Alternative 1 Under Alternative 1, there would be greater weed spread due to limitations on acres and areas that could be treated, this would result in a cumulative reduction in native plant

Final Integrated Weed Management EIS Page 193 Chapter 3 Affected Environment and Environmental Consequences diversity. Salamanders, frogs toads, and bats all rely on insect populations for their food requirements; insects rely on the native plant community for at least part of their life cycle. Eventually replacement of native plant communities with weeds would have an effect on insect populations potentially resulting in adverse effects on the food resource for western toads, Coeur d’Alene salamanders, northern leopard frogs and Townsend’s big-eared bat.

Direct and Indirect Environmental Consequences Specific to Alternative 2 Alternative 2 applies integrated weed management to a greater area and allows better prioritization of sites to be treated (to improve the effectiveness of treatments) and would have a substantially higher likelihood to maintain native plant communities on up to 15,000 acres per year as compared to Alternative 1 (5,000 to 6,000 acres/year).

Forest Plan and Regulatory Consistency Both alternatives meet Forest Plan direction to provide habitat for all indigenous wildlife species (FP page II-1 Goal #2).

Both alternatives are consistent with Forest Plan direction for aquatic ecosystems (FP II-14, Forest-wide standard #28).

Both alternatives meet Forest Plan standards for snags and dead material (FP II-14, Forest- wide standard #28 N1-3, Appendix N).

Statement of Finding Implementation of the No Action and Action Alternatives may impact individuals or habitat, but would not likely contribute to a trend towards federal listing for Townsend’s big-eared bat, northern leopard frog, western toad, and Coeur d’Alene salamander. This finding is based on: • a low potential for increasing the risk of predation on western toad tadpoles subjected to runoff from high applications of 2,4-D, • a low risk of herbicide spills into pond or stream habitat for western toads or Coeur d’Alene salamanders, • a low potential for mortality of the caterpillar prey species of Townsend’s big-eared bat, and • no change in suitable habitat occurring on the Lolo National Forest. Migratory Birds and Biodiversity

Existing Condition Neotropical migratory birds are those bird species that migrate to more northerly latitudes to breed on the Lolo National Forest each summer. Come fall, these species migrate south to spend the winter months. Of the approximately 275 bird species known to occur on the Forest as breeders, migrants, winter visitors, or transients, over 100 species could be classified as neotropical migratory land birds (Pat Sweeney, WMPZ Wildlife Biologist, personal communication)

Page 194 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

An executive order signed by President Clinton on 01/10/01 requires the Forest Service and other federal agencies to evaluate the effects of agency actions on migratory birds. Additionally, migratory birds have a variety of life history strategies that tie in a wide variety of other plant and animal species. Therefore, they were used to assess biodiversity for this analysis. Numerous species of migratory birds seasonally inhabit a variety of habitats across the Forest.

The North American Landbird Conservation Plan (Rich et al. 2004) identifies bird species within grasslands and western shrublands that are of continental importance in the Intermountain West Avifaunal Biome. Seven bird species (Swainson’s Hawk, Brewer’s Sparrow, Rufous Hummingbird, Calliope Hummingbird, Dusky Flycatcher, Sage Sparrow, and Mountain Bluebird) are species of continental importance that inhabit western shrubland habitats that can be found on the Lolo NF (Rich et al. 2004, pg 52).

Analysis Methods and Area The analysis area lies within the boundaries of the Lolo National Forest. Potential impacts on migratory birds from the proposed and alternative actions include exposure to herbicides through diet; disturbance caused by herbicide treatment efforts; and benefits from a reduction in weed infestations.

Environmental Consequences – Migratory Birds and Biodiversity

Direct and Indirect Environmental Consequences Common to Both Alternatives The potential hazard to migratory birds was evaluated using SERA worksheets designed to perform risk assessment for herbicides at the rates of application proposed for a project. The representative bird used to determine the effects of herbicides in the Forest Service Risk Assessments (SERA 1997-2004) is a small insectivorous bird. Direct spray of insects could occur, as they are present on vegetation and would not necessarily flee during treatment operations. The exposure scenario explored in the SERA assessments involves a bird whose diet consists of 100 percent contaminated insects.

Direct and Indirect Effects Alternatives 1 and 2: Of 11 herbicides proposed for use on the Lolo National Forest, three (2,4-D, dicamba and tryclopyr) could result in a HQ > 1. The application of any of these three herbicides has the potential for adverse effects on small insectivorous birds if they consume contaminated insects as 100 percent of their diet. However, foliar interception would reduce the actual amount sprayed on almost all insects present. 2,4-D is also not widely used on the Lolo NF and is used in only certain situations. The use of aminopyralid, which is a very low toxicity herbicide, will further decrease the use of 2,4-D on the Lolo NF. If the treatment area involved most of one or several territories established by insectivorous birds during the breeding season, it is feasible for an insectivorous bird to consume all or most of its daily diet within the treatment area. The young of these bird species are highly dependent on insects for their growth and development. Therefore, while the actual doses received by insectivorous birds will most likely be lower than the exposure scenarios predict (due to foliar interception, application method, and other variables) the

Final Integrated Weed Management EIS Page 195 Chapter 3 Affected Environment and Environmental Consequences consumption of contaminated insects by young birds feeding in a relatively small area may offset this advantage. Broadcast application of 2,4-D on limited areas on the Lolo NF could occur. Use of 2,4-D would be infrequent and patchy reducing the potential for adverse effects on individual insectivorous birds.

Activities in addition to herbicide treatment, (including bio control, reseeding, sheep and goat grazing and ground-based mechanical treatment) are not expected to have an effect on migratory birds assuming implementation of mitigation measures (Chapter 2).

One of the benefits of reducing weeds and increasing native and desirable palatable vegetation includes providing a diversity of insects that provide forage for birds.

Cumulative Effects The cumulative impacts of application of 2,4-D by the Forest Service, other agencies and private landowners would mostly occur along human travel corridors and pastures. Human travel corridors are less likely to contain entire bird territories thus the likelihood of multiple adverse effects on insectivorous birds from broadcast spray programs along road sides is very low.

Forest Plan and Regulatory Consistency There are no specific goals or standards for migratory land birds in the Lolo Forest Plan. However, it does contain the forest-wide goal to: “Provide habitat for viable populations of all indigenous wildlife species ….” (FP, II-1, Goal #3). Both alternatives are consistent with the Lolo Forest Plan, as a wide range of successional habitats would be available. The alternatives are also in compliance with the Executive Order “Responsibilities of Federal Agencies to Protect Migratory Birds”.

Economics

Existing Conditions The total employment for the seven county impact area increased from 87,751 full and part-time jobs during 1993 to 116,930 by 2003. This was an absolute increase of 29,179, translating into a 33 percent increase in total employment. Every county also experienced an increase in total employment during the last 10 years. Missoula and Ravalli counties led growth with about 17,000 and 6,000 jobs added to each county.

Per capita personal income (PCPI) is a measure of economic well-being and is calculated by dividing total personal income by the population. This figure is typically adjusted for inflation using the gross domestic product deflator to show changes in real dollars. The average real PCPI across the zone increased from $19,873 to $24,441 during the 10 year period. This represents an increase of just under $5,000, or 23 percent. The ranking of the counties based on the PCPI changed only slightly with four counties retaining the same rank at the end of the period. Missoula retained the highest real PCPI. Mineral County increased its income faster than its neighbors, whereas Lake and Powell counties did not grow as fast as other counties in the impact area.

Page 196 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

Method of Analysis IMPLAN Modeling -- The analytical technique used by the Forest Service to estimate employment and income impacts is "input-output" analysis using the IMPLAN Pro software system (IMPLAN 2004). Input-output analysis is a means of examining relationships within an economy both between businesses and between businesses and final consumers (Maki and Lichty 2000). It captures all monetary market transactions for consumption in a given time period. The resulting mathematical representation allows one to examine the effect of a change in one or several economic activities on an entire economy, all else constant. This examination is called economic impact analysis.

The IMPLAN model estimates economic effects, such as employment and labor income within a defined geographical area (called a functional economic area). The IMPLAN modeling system allows one to build functional economic area models of one or more counties for a particular year. Economic impacts associated with the action alternatives for this analysis area are based on an IMPLAN model for the seven county impact area of the Lolo NF. The functional economic area model for this analysis used 2002 IMPLAN data. Impacts are measured in terms of employment (full- and part-time jobs) and labor income.

Environmental Consequences - Economics

Direct and Indirect Environmental Consequences Common to Both Alternatives The presence of wildlife is an important part of many outdoor recreation activities. A study done by Hirsch and Leitch (1996) evaluated the impacts of spotted knapweed on the Montana economy. Economic impacts affected the economic sectors of grazing, wildlife, and soil and water conservation. The effect of weeds on wildlands in reducing the availability of wildlife for many outdoor recreation activities has a direct negative relationship to the amount of expenditures spent on those activities.

Total annual expenditures for consumptive and non-consumptive wildlife associated recreation in Montana, adjusted to 1994 dollars were $217,000,000 (Hirsch and Leitch 1996). Direct economic impact is the result of changes in expenditures that impact suppliers of recreation related goods and services (e.g., fees, lodging, etc.). The direct negative economic impacts from spotted knapweed on wildlife-associated values in Montana are about $1,177,000 a year, consisting mostly of impacts on retail trade ($883,000) and business and personal services ($294,000) (Hirsch and Leitch 1996). Direct impacts from spotted knapweed infestations also have secondary negative economic impacts on Montana's economy. Reduced economic activity in one sector can have substantial effects on employment, income, and expenditures in other sectors. When annual direct economic impacts are added to the secondary economic impacts of spotted knapweed on wildlife associated benefits in Montana, there is a total loss of about $2,641,000 of labor income and a loss of 34 full time equivalent jobs (Hirsch and Leitch 1996). Most of this loss comes from reduced retail trade ($1,257,000), reduced household spending ($567,000), and business and personal services ($326,000) (Hirsch and Leitch 1996). These are just the economic impacts of spotted knapweed, the most common noxious weed invading Montana (Hirsch and Leitch 1996).

Final Integrated Weed Management EIS Page 197 Chapter 3 Affected Environment and Environmental Consequences

Direct and Indirect Environmental Consequences Specific to Alternative 1 Under this alternative, weeds would continue to impact economic values related to wildlife. Alternative 1 would cost $150,000 per year and treat approximately 5,000 to 6,000 acres per year (refer to Appendix H – Cost Estimate by Alternative) for additional cost information. This combination of treatments would not accomplish the project objective of improving long-term forage and protecting native plants.

Table 3-39 below presents the estimated total employment and labor income impacts (direct, indirect and induced effects) of implementation if all the project areas were treated in one year. Table 3-39 indicates that Alternative 1 would support approximately 9 full and part-time jobs and $144,000 of labor income in the functional economic area, with most of economic effects occurring in the agricultural and support services sector of the economy.

Direct and Indirect Environmental Consequences Specific to Alternative 2 Alternative 2 would cost $450,000 per year and treat a maximum of 15,000 acres per year (refer to Appendix H for additional cost information). Although grants are used to supplement some aspects of the present weed program on the forest, strictly ground based controls may not compete well in the grant process due to the high unit costs and lack of a successful model for such vast acreages needing treatment. Expenditures for erosion related impacts would decrease as the spread of weeds was reduced.

Table 3-39 below presents the estimated total employment and labor income impacts (direct, indirect and induced effects) of implementation if the maximum acres were treated in one year. Table 3-39 indicates that Alternative 2 would support approximately 28 full and part-time jobs and $432,000 of labor income in the functional economic area, with most of economic effects occurring in the weed control and support services sector of the economy.

Table 3-39 Estimated Economic Impacts Estimated Estimated Total Alt Cost Total Employment Labor Income 1 $150,000 9 $144,000 2 $450,000 28 $432,000 Cumulative Effects

Benefits Common to Weed Treatment The costs of the weed treatments by alternative are presented in Chapter 2. There would be additional intangible and difficult to quantify economic benefits other than those described above. Since many other anticipated economic benefits are not easily converted to dollar amounts they are summarized here.

Page 198 Final Integrated Weed Management EIS Chapter 3 Affected Environment and Environmental Consequences

The proposed weed treatments are expected to produce the following non-market benefits: • Improved wildlife habitat by increasing the amount of forage produced per acre, • Improved biodiversity and protection of several threatened or endangered species, • Reduced potential for increased erosion and soil loss associated with weed dominated plant communities, • Protection of aquatic species resulting from reduced runoff and sedimentation, • Improved esthetic value of the landscape, • Potential increases in the amount of recreation use that would occur in relatively weed free areas, • Protection of adjacent lands that are either weed free or have active weed control programs. Forest Plan Consistency Both alternatives are consistent with the Lolo NF Plan.

Final Integrated Weed Management EIS Page 199

Chapter 4 Consultation and Coordination

Preparers The Forest Service consulted the following individuals, Federal, State, and local agencies, tribes and non-Forest Service persons during the development of this environmental assessment:

ID Team Members: Name Title Role Anderson, Carl Forestry Technician, Missoula RD Missoula RD Weed Coordinator Forest Vegetation / Ecology Applegate, Vic Lolo NF Ecologist Analysis Bacon, Sydney Archaeologist Heritage Resources Analysis Beckes, Barbara Program Officer for Planning Project Staff Officer Seeley Lake RD Weed Blackburn, Jim Forestry Technician, Seeley Lake RD Coordinator Blackburn, Lisa GPS/GIS Specialist, East Zone Lolo Maps Burwasser, Jim Forester, Superior RD Superior RD Weed Coordinator Lolo NF Landscape Recreation / Roadless / Casselli, Elizabeth Architect/Recreation Specialist Wilderness Analysis Domsalla, Ryan Resource Forester, Ninemile RD Ninemile RD Weed Coordinator Acting Lolo NF Ecologist, Forest Vegetation / Ecology and Errecart, John Plains/Thompson Falls RD Plains RD Weed Coordinator Kratville, Sandy Lolo NF Wildlife Biologist Wildlife Analysis ID Team Leader, Weeds, Social Environment, Range, Roads Kulla, Andy Lolo NF Weed Program Leader Analysis Mcleod, C. Milo Lolo NF Archaeologist Heritage Resources analysis Millar, Sarah Forestry Technician, Missoula RD Tables, Figures and Editing Niccolucci, Mike Region One Economist Economics Analysis Pike, Lynda Writer / Editor, East Zone Lolo Editor Rice, Peter Botanist/Plant Ecologist, U of M TES Plants/Botany Analysis Riggers, Brian Lolo NF Fisheries Biologist Fisheries Analysis Rosquist, Skip Lolo NF Hydrologist Soil and Water Analysis Steve Shelley Region One Botanist Botany Analysis Stadler, Don Lolo East Zone NEPA Specialist NEPA Coordinator Tomson, Scott Acting Lolo NF Wildlife Biologist Wildlife Analysis

Agencies, Organizations, Businesses and People Notified Citizens for a Weed Free Future Agencies Four Rivers Shuttle Advisory Council on Historic Preservation Friends of Pattee Canyon Colorado State University Library Missoula County Redevelopment Environmental Protection Agency, Region 8 Missoula Outdoor Learning Federal Aviation Administration, Mormon Creek Weed Group NW Mountain Region Open Space Flathead County Commissioners Outback Rides Flathead County Weed District Parson’s Pony Rides Granite County Commissioners Westech Granite County Weed District WildWest Institute Granite County Extension Woodland Restoration Idaho Dept. of Agriculture Weed Management Services Lake County Commissioners Westech Lake County Weed District Lewis and Clark County Commissioners People: Lewis and Clark County Weed District Artley, Dick Mineral County Commissioners Birch, Joan Mineral County Weed District Brown, Jim and Sue Missoula County Commissioners Canton-Thompson, Janie Missoula County Weed District Childers, Ed Missoula Parks and Recreation Corti, Dan Montana State Highway Administration Doyle, Kevin Mt. Dept. of Fish Wildlife and Parks Edholm, Ed Natural Resources Conservation Service, Garner, Dwayne National Environmental Coordinator Goldberg, Stuart Northwest Power Planning Council Henry, Teresa Powell County Commissioners Hillis, Mike Powell County Weed District Knudsen, Kathy Ravalli County Commissioners Linder, Bert Ravalli County Weed District Livingston, Don Sanders County Commissioners McInnis, Logan Sanders County Weed District Miller, Charles Sula Ranger District Moore, Vicki and Jane University of Montana, Olson, Duane Division of Biological Sciences Peterson, Tom US Army Engr., Northwestern Division Reiter, Sharon US Coast Guard, Environmental Mgmt. Ryan, Meggan US Department of Energy, Swingendorf, Joyce Office of NEPA Policy and Compliance. Tawney, David US Department of Interior, Thibodeau, Gwen Office of Environmental Policy and Compliance Thomas, William US Environmental Protection Agency Tweedale, Tony USDA APHIS PPD/EAD Voermans, Margo USDA, National Agriculture Library Wegener, Gloria and Michael Organizations/Businesses: Whitesitt, David Wilson, Sara Chemical Injury Comm. Net

Page 202 Final Integrated Weed Management EIS

Literature Cited

Introduction The references listed here are inclusive of all references used during the analysis. Not all of these are necessarily referenced in the Integrated Weed Management EIS, but are included in the individual specialist’s reports contained in the Project File.

Weeds Borrecco, J. and J. Neisess, 1991. Risk Assessment for the impurities 2-butoxyethanol and 1,4-dioxane found in Garlon 4 and Roundup herbicide formulations. USDA FS PSW Region, Forest Pest Management. Report # R91-2. 33 pp.

Callaway, R.M., DeLuca, T. and W.M. Belliveau, 1999. Ecology, 80(4):1196-1201.

Cousens, R. and M. Mortimer, 1995. Dynamics of Weed Populations. Cambridge University Press. Cambridge, Great Britain.

Duncan, C.A, Brown, M. and M.B. Halstvedt, 1999. Cost and Efficacy of Spotted knapweed Management with Integrated Methods, Report from Weed Management Services, Helena, MT.

Henry, C., 1998. Mormon Ridge Elk Winter Range Restoration Project. Techline. August. 11 pp.

Hoopes, C., 2005. Noxious Weed Pre and Post Test Telephone Surveys 1994, 2002, 2005, Comparative Charts. Final report submitted to Carla Hoopes, Montana State Weed Education Project Coordinator, March 7, 2005. 24 pp.

Kott, R., 2005. Montana State University Sheep Institute Progress Report, 2004 Summary.

Kott, R., 2005. Sheep Grazing and Leafy Spurge Control, Montana Sheep Institute Bulletin. August. 20 pp.

Losensky, J.B., 1987. An evaluation of noxious weeds on the Lolo, Bitterroot and Flathead Forests. Unpublished Report. On file at Lolo National Forest, Missoula, MT 59804. pg. 64

Marler, M., 2004. Response of an Idaho fescue grassland to hand pulling and Tordon for control of spotted knapweed. Report to Missoula Parks and Recreation.

Marler, M., 2004. Response of Mt. Jumbo grasslands to integrated control of St. Johnswort (Hypericum perforatum) from 1999-2004. Report to Missoula Parks and Recreation.

Mueggler, W.F. and W.L. Stewart, 1980. Grassland and shrub land habitat types of Western Montana, USDA Forest Service General Technical Report INT-66, Intermountain Forest and Range Experiment Station. 154 pp.

Literature Cited

Ortega, Y.K., 2005. Winter Range Restoration on the Lolo National Forest: Monitoring Vegetation Response to Herbicide Treatments, Rocky Mountain Research Station, FSL, Missoula, MT.

Rice, P., 2005. Gibbons Pass Spray Monitoring Plot Report For 2005.

Rice, P. and R.S. Cochran, 2005. Winter Range Weed Treatment and Monitoring, Lolo National Forest: Second Year Post-Treatment.

Rice, P. and M. Harrington, 2003. The Impact of Prescribed Fire Following Either Herbicide or Thinning Treatments on Bunchgrass and Forested Ecosystems. In: Rice, P., Lavelle, D. and Greenlee, J., 2003. USDA, Monitoring Analysis of Mormon Ridge Plant Community Plots (1996-2000).

Sawmill Creek Research Natural Area, FY 2003 Final Report, Agreement Number 02-JV- 11222048-141. 43 pp.

USDA, Lolo National Forest, 1996. Hayes Creek Biological Control Monitoring Photo points.

USDA, Lolo National Forest, 1998. Strawberry Ridge Biological Control Monitoring Photo points.

USDA Forest Service, 1996. Mormon Ridge Big Game Winter Range Restoration. Environmental Impact Statement. Lolo National Forest, Missoula, MT.

USDA Forest Service, 1980. Visual Character Types of the Northern Region.

USDA Forest Service, 1997. Upper Columbia River Basin Draft Environmental Impact Statement. Forest Service Intermountain and Northern Regions and USDI Bureau of Land Management, Idaho, Montana, Nevada, Utah and Wyoming.

USDA, Agricultural Research Service Animal Plant Health Inspection Service, 2002. Multi- Species Grazing and Leafy Spurge. A product of the USDA ARS TEAM Leafy Spurge Area- Wide IPM Program. 28 pp.

Weed Web Sites Cited: Invader Database, weed presence in Montana, http://invader.dbs.umt.edu

USDA Forest Service, Evaluating Risk to Native Plant Communities from Selected Exotic Plant Species (http://www.fs.fed.us/r1/cohesive_strategy/datafr.htm

USDA Forest Service, Lolo National Forest web site, weed characteristics, http://fsweb.lolo.r1.fs.fed.us/silv/noxious-weeds/index.shtml http://ag.montana.edu/warc/biocontrol_agents_of_knapweed.htm http://nris.state.mt.us/interactive.html

Page 204 Final Integrated Weed Management EIS Literature Cited http://maps2.nris.mt.gov/mapper/ThemeList.asp?Profile=2088894&qLayer1=USFSDISTRI CT&qField1=Forest&qValue1=Lolo&Oper1=&Buffer1=0&TabName=Land%20Information http://agr.state.mt.us/weedpest/pdf/2005weedPlan.pdf

Sheep Information Olson, B.E., Wallander, R.T. and Lacey, J.R., 1997. Effects of sheep grazing on a spotted knapweed infested Idaho fescue community. Journal of Range Management, 50:386-390.

Olson, B.E. and Wallander, R.T., 1997. Biomass and carbohydrates of spotted knapweed and Idaho fescue after repeated grazing. Journal of Range Management, 50:409-412.

Olson, B.E., Wallander, R.T. and Kott, R.W., 1997. Recovery of leafy spurge seed from sheep. Journal of Range Management, 50:10-15.

Olson, B.E., and Wallander, R.T., 1998. Effects of sheep grazing on a leafy spurge-infested Idaho fescue community. Journal of Range Management, 51:247-252.

Olson, B.E., and Wallander, R.T., 2003. Comparing Indicators of Sheep Grazing Leafy Spurge and Perennial Grasses. Sheep and Goat Research Journal, 18:101-108.

Wallander, R.T., Olson, B.E. and Lacy, J.R., 1995. Spotted knapweed viability after passing through sheep and mule deer. Journal of Range Management, 48:145-149.

Forest Vegetation Asher, J., Dewey, S., Johnson, C. and J. Olivarez, 1999. Protecting Relatively Uninfested Lands: Reducing Weed Spread Following Fire. Tech Line (2004):2-3.

Brewer, K.C., Berlund, D., Barber, J.A. and R. Bush, 2004. Northern Region Vegetation Mapping Project Summary Report and Spatial Datasets. USDA Forest Service, Northern Region, Lolo National Forest, Missoula, MT. 61 pp.

Jones, J., 2004. Potential Natural Vegetation Types. Northern Region Cohesive Strategy Team.

Krogstad, D., 2005. Mapping of Lolo National Forest Potential Natural Vegetation Type Groups and Forest Dominance Types. GIS for Western Montana Planning Zone. Data available at the Lolo National Forest Supervisor’s office, Fort Missoula, Missoula, MT.

Lackschewitz, K., 1991. Vascular Plants of West-Central Montana-Identification Handbook. USDA Forest Service. General Technical Report INT-277. Intermountain Research Station. Ogden, Utah. 648 pp.

Pfister, R.D., Kovalchik, B.L., Arno, S.F. and R. Presby, 1977. Forest Habitat Types of Montana. USDA Forest Service. General Technical Report INT-34. Intermountain Forest and Range Experiment Station. Ogden, Utah. 174 pp.

Final Integrated Weed Management EIS Page 205 Literature Cited

Rice, P.M. and J.C. Toney, 1997. Exotic Weed Control Treatments for Conservation of Fescue Grasslands in Montana. Biological Conservation, 85 (1998):83-95.

Sheley, R., Petroff, J. and M., Borman, 1999. Biology and Management of Noxious Rangeland Weeds.

Sheley, R., Manoukian, M. and G. Mark, 2004. Preventing Noxious Weed Invasion. Montana State University Extension Service, A-5. 3 pp.

Tatum, V.L., 2004. Toxicity, transport and fate of forest herbicides. Wildlife Society Bulletin, 32(4):1042-1048.

Fish Johnson, W. W. and M. T. Finley. 1980. Handbook of acute toxicity of chemicals to fish and aquatic invertebrates. US Dept. of Interior FWS, Resource Publication 137.

Lacey, J.R., Marlow, C.B. and J.R. Lane, 1989. Influence of spotted knapweed (Centaurea maculosa) on surface runoff and sediment yield. Weed Technology, 3:627-631.

US Department of Energy (USDOE), 2000. Chlorsulfuron: herbicide fact sheet. Prepared by US Department of Energy – Bonneville Power Administration.

USDOE-Bonneville Power Administration. 2000. 2, 4-D: Herbicide Fact Sheet. Bonneville Power Administration:

USGS Open-File Report. 2003. Reconnaissance data for glyphosate, other selected herbicides, their degradation products, and antibiotics in 51 streams in nine Midwestern states, 2002.

SERA 1997. Selected commercial formulations of hexazinone - human health and ecological risk assessment. Submitted to the USDA FS by Syracuse Environmental Research Associates, Inc. p. 3-28-3-33

SERA 1998. Human health and ecological risk assessment for 2,4-D. Submitted to the USDA FS by Syracuse Environmental Research Associates, Inc.

SERA 2001a. 2,4-D worksheets for human health and ecological risk assessments. Submitted to the USDA FS by Syracuse Environmental Research Associates, Inc. p. 38-41

SERA 2001b. Hexazinone – worksheets for human health and ecological risk assessments. Submitted to the USDA FS by Syracuse Environmental Research Associates, Inc. p. 42-45

SERA 2001c. Preparation of environmental documentation and risk assessments. Prepared by Syracuse Environmental Research Associates, Inc.

SERA 2003a. Glyphosate - human health and ecological risk assessment. Submitted to the USDA Forest Service by Syracuse Environmental Research Associates, Inc. p. 3-47-3-61

Page 206 Final Integrated Weed Management EIS Literature Cited

SERA 2003b. Chlorsulfuron - human health and ecological risk assessment. Submitted to the USDA FS by Syracuse Environmental Research Associates, Inc. p. 3-22-3-30.

SERA 2003c. Picloram - human health and ecological risk assessment. Submitted to the USDA FS by Syracuse Environmental Research Associates, Inc. p. 3-33-3-37.

SERA 2003d. Picloram (Tordon 22 and 22K) – worksheets for human health and ecological risk assessments. Submitted to the USDA FS by Syracuse Environmental Research Associates, Inc.

SERA 2003e. Triclopyr - human health and ecological risk assessment. Submitted to the USDA FS by Syracuse Environmental Research Associates, Inc. p. 3-29-3-34, 3-50-3-51

SERA 2004a. Chlorsulfuron – worksheets for human health and ecological risk assessments. Submitted to the USDA Forest Service by Syracuse Environmental Research Associates, Inc. p. 38-41

SERA 2004b. Clopyralid - human health and ecological risk assessment. Submitted to the USDA Forest Service by Syracuse Environmental Research Associates, Inc. p. 3-33-3-38

SERA 2004c. Clopyralid – worksheets for human health and ecological risk assessments. Submitted to the USDA Forest Service by Syracuse Environmental Research Associates, Inc. p. 43-46

SERA 2004d. Hexachlorobenzene as an impurity in clopyralid – worksheets for human health and ecological risk assessments. Submitted to the USDA FS by Syracuse Environmental Research Associates, Inc. p. 37-40

SERA 2004e. Dicamba - human health and ecological risk assessment. Submitted to the USDA Forest Service by Syracuse Environmental Research Associates, Inc. p. 3-29-3-32

SERA 2004f. Dicamba - worksheets for human health and ecological risk assessments. Submitted to the USDA Forest Service by Syracuse Environmental Research Associates, Inc. p. 43-46

SERA 2004g. Imazapic - human health and ecological risk assessment. Submitted to the USDA FS by Syracuse Environmental Research Associates, Inc. p. 3-22-3-25

SERA 2004h. Imazapic – worksheets for human health and ecological risk assessments. Submitted to the USDA FS by Syracuse Environmental Research Associates, Inc. p. 43-46

SERA 2004i. Imazapyr - human health and ecological risk assessment. Submitted to the USD A FS by Syracuse Environmental Research Associates, Inc. p. 3-25-3-28

SERA 2004j. Imazapyr – worksheets for human health and ecological risk assessments. Submitted to the USDA Forest Service by Syracuse Environmental Research Associates, Inc. p. 43-46

Final Integrated Weed Management EIS Page 207 Literature Cited

SERA 2004k. Metsulfuron methyl - human health and ecological risk assessment. Submitted to the USDA Forest Service by Syracuse Environmental Research Associates, Inc. p. 3-23-3- 25

SERA 2004l. Metsulfuron methyl - worksheets for human health and ecological risk assessments. Submitted to the USDA Forest Service by Syracuse Environmental Research Associates, Inc. p. 38-41

SERA 2004m. Sulfometuron methyl - human health and ecological risk assessment. Submitted to the USDA FS by Syracuse Environmental Research Associates, Inc. p. 3-25-3- 28.

SERA 2004n. Sulfometuron methyl – worksheets for human health and ecological risk assessments. Submitted to the USDA FS by Syracuse Environmental Research Associates, Inc. p. 38-41

USDA Forest Service, 1986. Lolo National Forest Plan. Lolo National Forest, Missoula, MT.

USDA Forest Service, 1991. Final Environmental Impact Statement, Noxious Weed Management, Amendment to Lolo National Forest Plan. Lolo National Forest, Missoula, MT.

USDA Forest Service, 2000. Section 7 Consultation Watershed Baseline, Middle Clark Fork River. Prepared by the Lolo National Forest. Lolo National Forest, Missoula, MT.

USDI Forest Service, 1991. Lolo National Forest Plan, Amendment 11.

USDA Forest Service. 1992. Risk Assessment for Herbicide Use in Forest Service Regions 1, 2, 3, 4, and 10 and on Bonneville Power Administration Sites.

Fish Websites Cited: Monsanto. 2001. Material Safety Data Sheet. Monsanto Company, St. Louis Mo. http://www.fs.fed.us/foresthealth/pesticide/data/Rodeo_msds.pdf http://www.efw.bpa.gov/portal/Organizations/Government/Federal/Dept_of_Energy/BP A/Environment/NEPA/Vegetation/appenH.pdf. Also Fact sheets for Chlorsulfuron, Clopyralid, Dicamba, Glyphosate, Hexazinone, Imazapyr, Metsulfuron-methyl, Picloram, Sulfometuron, Triclopyr.

USDA Forest Service. Lolo National Forest Fisheries Webpage. http://www.fs.fed.us/r1/lolo/resources-natural/fish

Threatened and Endangered Species-Plants Allen, Edith B.; Cox, R. D.; Tennant, T.; Kee S.N.; Deutschman, D. H. 2005. Landscape restoration in southern California forblands: Response of abandoned farmland to invasive annual grass control. Israel Journal of Plant Science. 53(3-4): 237-245.

Page 208 Final Integrated Weed Management EIS Literature Cited

Baer, Sara G.; Groninger, John W. 2004. Herbicide and tillage effects on volunteer vegetation composition and diversity during reforestation . Restoration Ecology. 12(2): 258-267.

Bedunah, Donald J.; Carpenter, Jeffrey L. 1989. Plant community response following spotted knapweed control (Centaurea maculosa) on three elk winter ranges in western Montana. In: Fay, P. K., and Lacey, J. R., eds. Proceedings: Knapweed Symposium Proceedings EB45 Bozeman, MT. Bozeman MT: Plant and Soil Science Department and Extension Service, Montana State University: p 205-212.

Belcher, Joyce W.; Wilson, Scott D. 1989. Leafy spurge and the species composition of a mixed-grass prairie. Journal of Range Management. 42(2): 172-175.

Brandon, Alice L.; Gibson, David J.; Middleton, Beth A. 2004. Mechanisms for dominance in an early successional old field by the invasive non-native Lespedeza Cuneata (Dum. Cours.) G. Don. Biological Invasions. 6(4): 483-493.

Carlsen, Tina M.; Menke, John W.; Pavlik, Bruce M. 2000. Reducing competitive suppression of a rare annual forb by restoring native California perennial grasslands8: 18- 29.

Carlson, Adriane M.; Gorchov, David L. 2004. Effects of herbicide on the invasive biennial Alliaria petiolata (garlic mustard) and initial responses of native plants in a southwestern Ohio forest. Restoration Ecology. 12(4): 559-567.

Daubenmire, Rexford. 1968. Plant communities: a textbook of plant synecology. New York, NY: Harper & Row Publishers, Inc. 300 p.

DiTomaso, Joseph M.; Marcum, Daniel B.; Rasmussen, Michelle S.; Healy, Evelyn A.; Kyser, Guy B. 1997. Post-fire herbicide sprays enhance native plant diversity . California Agriculture. 51: 6-11.

Erickson, A. M.; Lym, Rodney G.; Kirby, D. R. 2006. Effect of herbicides for leafy spurge control on the western prairie fringed orchid. Rangeland Ecology and Management. 59(5): 462-467.

Erskine, Jennifer; Rejmanek, Marcel. 2001. Changes in plant community structure following the removal of a dominant perennial invasive species. Proceedings: ESA 2001 Annual Meeting

Getsinger, K. D.; Turner, E. G.; Madsen, J. D.; Netherland, M. D. 1997. Restoring native vegetation in a Eurasian water milfoil-dominated plant community using the herbicide triclopyr . Regulated Rivers Research and Management. 13(4): 357-375.

Grime JP. 1979. Plant strategies and vegetation processes. New York: Wiley.

Hansen, Paul L.; Pfister, Robert D.; Boggs, Keith; Cook, Bradley J.; Joy, John; Hinckley, Dan K. 1995. Classification and management of Montana's riparian and wetland sites Volume Misc. Pub. 54. School of Forestry, University of Montana, Missoula, MT 59812: Montana

Final Integrated Weed Management EIS Page 209 Literature Cited

Forest and Conservation Experiment Station: Miscellaneous Publication No. 54. xvi + 646 pp.

Kedzie-Webb, S. A.; Sheley, Roger L.; Borkowski, John J.; Jacobs, James S. 2001. Relationships between Centaurea maculosa and indigenous plant assemblages. Western North American Naturalist. 61(1): 43-49.

Kedzie-Webb, Susan A.; Sheley, Roger L.; Borkowski, John J. 2002. Predicting plant community response to picloram. Journal of Range Management. 55(6): 573-583.

Kirby, D. R.; Lym, R. G.; Sterling, J. J.; Sieg, C. H. 2003. Observation: Leafy spurge control in western prairie fringed orchid habitat. Journal of Range Management. 56(5): 466-473.

Knezevic, Stevan Z.; Smith, Doug; Kulm, Ralph; Doty, Don; Kinkaid, Dick; Goodrich, Mick; Stolcpart, Rod. 2004. Purple loosestrife (Lythrum salicaria) control with herbicides: single- year application. Weed Technology. 18: 1255-1260.

Kulla, A., 2006. Lolo National Forest Weed Coordinator. Personal Communication.

Laufenberg, Stephen; Sheley, Roger L.; Jacobs, James S. 2002. Restoring species diversity and richness along river bottom areas using herbicides. Proceedings: ESA 2002 Annual Meeting

Laufenberg, Stephen; Sheley, Roger L.; Jacobs, James S.; Borkowski, John. 2005. Herbicide effects on density and biomass of russian knapweed (Acroptilon repens) and associated plant species. Weed Technology. 19(1): 62-72.

Lavelle, D., 2006. Lolo National Forest Botanist. Personal Communication.

Lavorel, Sandra; Lepart, Jacques; Debussche, Max; Lebreton, Jean-Dominique; Beffy, Jean- Luc. 1994. Small scale disturbances and the maintenance of species diversity in Mediterranean old fields. Oikos. 70(3): 455-473.

Lesica, Peter; Shelly, Stephen J. 1996. Competitive effects of Centaurea maculosa on the population dyamics of Arabis fecunda. Bulletin of the Torrey Botanical Club. 123(2): 111- 121.

Miller, James H.; Boyd, Robert S.; Edwards, M. B. 1999. Floristic diversity, stand structure, and composition 11 years after herbicide site preparation. Canadian Journal of Forest Research. 29: 1073-1083.

Mitchell, William W. 1958. An ecological study of the grasslands in the region of Missoula, Montana Missoula: University of Montana (MSU). 111p

Mueggler, W. F.; Stewart, W. L. 1980. Grassland and shrubland habitat types of western Montana. Ogden, UT: U.S.D.A. Forest Service, Intermountain Forest and Range Experiment Station. General Technical Report INT-66. Peter has a copy on the shelf. 154 p.

Page 210 Final Integrated Weed Management EIS Literature Cited

Ortega, Y.K., 2005. Winter range restoration on the Lolo National Forest: Vegetation response to herbicide treatments. Unpublished report for the Rocky Mountain Research Station, Missoula, MT. 11pp.

Ortega, Yvette K.; Pearson, Dean E. 2005. Weak versus strong invaders of natural plant communities: assessing invasibility and impact. Ecological Applications. 15(2): 651-661.

Pavlik, Bruce M.; Nickrent, Daniel L.; Howald, Ann M. 1993. The recovery of an endangered plant. I. Creating a new population of Amsinckia grandiflora . Conservation Biology. 7(3): 510-526.

Pfister, Robert D.; Kovalchik, Bernard L.; Arno, Stephen F.; Presby, Richard C. 1977. Forest habitat types of Montana. Ogden, UT: U.S.D.A. Forest Service, Intermountain Forest and Range Experiment Station. General Technical Report INT-34. Peter has a copy on the shelf. 174 p.

Randall, John M. 1996. Weed control for the preservation of biological diversity. Weed Technology. 10(2): 370-383.

Reever Morghan, Kimberly J.; Leger, Elizabeth A.; Rice, Kevin J. 2003. Clopyralid effects on yellow starthistle (Centaurea solstitialis) and nontarget species. Weed Science. 51(4): 596- 600.

Rice, K. J. 1989. Impact of seed banks on grassland community structure and population dynamics. In: Leck, Mary A.; Parker. V. Thomas; Simpson, Robert L., eds. Ecology of soil seed banks. Volume Chapter 10. San Diego: Academic Press: 211-230.

Rice, P.M and J.C. Toney, 1996. Plant population responses to broadcast herbicide applications for Spotted Knapweed control. Down to Earth, 51. 6 pp.

Rice, P.M. and J.C. Toney, 1998. Exotic weed control treatments for conservation of fescue grassland in Montana. Biological Conservation, 85:83-95.

Rice, P.M. and M.G. Harrington, 2003. The impact of prescribed fire following either herbicide or thinning treatments on bunchgrass and forest ecosystems in the Sawmill Creek Research Natural Area. FY 2003 Final report to the Forest Service. Agreement #02-JV-11222048-141. 43 pp.

Rice, P.M. and M.G. Harrington, 2005. Stabilization of plant communities after integrated picloram and fire treatments. FY 2005 Final report to the Forest Service. Agreement #03- CA-11011600-029. 71 pp.

Rice, P.M. and R.S. Cochran, 2005. Winter range weed treatment and monitoring, Lolo National Forest: Second year post-treatment. Monitoring report for the Lolo National Forest, Missoula, MT. 19 pp.

Rice, P.M., 2005. Gibbons Pass Spray Monitoring Report for 2005. Monitoring report for the Lolo National Forest, Missoula, MT. 6 pp.

Final Integrated Weed Management EIS Page 211 Literature Cited

Rice, P.M., Bedunah, D.J. and C.E. Carlson, 1992. Plant community diversity after herbicide control of Spotted Knapweed. Research Paper INT-460. USDA Forest Service, Intermountain Research Station, Ogden, UT. 6 pp.

Rice, P.M., Toney, J.C., Bedunah, D.J. and C.E. Carlson, 1997. Elk winter forage enhancement by herbicide control of Spotted Knapweed. Wildlife Society Bulletin, 25(3):627-633. 19 pp.

Rice, P.M., Toney, J.C., Bedunah, D.J. and C.E. Carlson, 1997a. Plant community diversity and growth form responses to herbicide applications for control of Centaurea maculosa. Journal of Applied Ecology, 34:1397-1412.

Rice, Peter M. 2006. Gibbons Pass spray monitoring plot report for 2006. Hamilton, MT: Bitterroot National Forest. 9p

Rice, Peter M.; Gauer, Steven. 2007. Winter range weed treatment and monitoring, Lolo National Forest: pre-spray through fourth year post-treatment results (2002-2006). Missoula, MT: Lolo National Forest & University of Montana Division of Biological Sciences. Agreement #06-CS-11011600-026. 34p

Rice, Peter M.; Harrington, Michael G. 2003. The impact of prescribed fire following either herbicide or thinning treatments on bunchgrass and forested ecosystems in the Sawmill Creek Research Natural Area. Missoula, MT: USFS Rocky Mountain Research Station Fire Sciences Laboratory & University of Montana Division of Biological Sciences. FY 2003 Final Report Agreement # 02-JV-11222048-141. 43p

Rice, Peter M.; Harrington, Michael G. 2005. Integration of herbicides and prescribed burning for plant community restoration. Missoula, MT: USFS Rocky Mountain Research Station Fire Sciences Laboratory & University of Montana Division of Biological Sciences. Final Report P.A. Nu. 01-PA-11222048-172 Pesticide Impact Assessment Program (PIAP) Project Number RM-6. 71 p

Rice, Peter M.; Harrington, Michael G. 2006. The impact of prescribed fire following either herbicide or thinning treatments on bunchgrass and forested ecosystems in the Sawmill Creek Research Natural Area. Missoula MT: USFS Rocky Mountain Research Station Fire Sciences Laboratory & University of Montana Division of Biological Sciences. Progress Report for Sampling Through 2006, Agreement # 02-JV-11222048-141. 19p

Rice, Peter M.; Toney, J. C. 1996. Plant population responses to broadcast herbicide applications for spotted knapweed control. Down to Earth. 51(2): 14-19.

Rice, Peter M.; Toney, J. C. 1998. Exotic weed control treatments for conservation of fescue grassland in Montana. Biological Conservation. 85: 83-95.

Rice, Peter M.; Toney, J. C.; Bedunah, Donald J.; Carlson, Clinton E. 1997. Plant community diversity and growth form responses to herbicide applications for control of Centaurea maculosa. Journal of Applied Ecology. 34: 1397-1412.

Page 212 Final Integrated Weed Management EIS Literature Cited

Ristau, Todd; Horsley, Stephen. 2002. Impact of chemical control of native invasive plants on understory diversity in northern hardwood forests. Proceedings: ESA 2002 Annual Meeting Tucson, AZ.

Rosentreter, Roger. 1994. Displacement of rare plants by exotic grasses. In: Monsen, Stephen B., and Kitchen, Stanley G., compilers. Proceedings: Ecology and management of annual rangelands. Gen. Tech. Rep. GTR-INT-313 Boise, ID. Intermountain Research Station, Ogden, UT: U.S. Forest Service: p 170-175.

Sheley, Roger L. M. J. m. A. J. L. 2006. Potential for successional theory to guide restoration of invasive-plant-dominated rangeland. Ecological Monographs. 76(3): 365-379.

Sheley, Roger L.; Denny, Kirk M. 2006. Community response of nontarget species to herbicide application and removal of the nonindigenous invader Potentilla recta L. Western North American Naturalist. 66(1): 55-63.

Sheley, Roger L.; Laufenberg, Stephen. 2006. Restoring species richness and diversity in a russian knapweed-infested riparian plant community using herbicides: Eastern Oregon Agricultural Research Center . SR1057. 36-37

Stein, B. A.; Kutner, L. S.; Adams, J. S., eds. 2000. Precious heritage: the status of biodiversity in the United States. New York, NY: Oxford University Press. 399p

Stringer, Lewis T. 2003. Restoration of spotted knapweed infested grasslands in Glacier National Park Bozeman, MT: Thesis, Montana State University. 73p

Sullivan, Thomas P.; Wagner Robert G.; Pitt, Douglas G.; Lautenschlager, R. A.; Chen, Din G. 1998. Changes in diversity of plant and small mammal communities after herbicide application in sub-boreal spruce forest. Canadian Journal of Forest Research. 28(2): 168- 177.

Thompson, D.; Stuckey, R.; Thompson E. 1987. Spread, impact and control of purple loosestrife (Lythrum salicaria) in North America. Washington D.C.: U.S. Fish and Wildlife Service. Res. Rep. 2.

Tyser, R.; Key, C. H. 1988. Spotted knapweed in natural area fescue grasslands: An ecological assessment. Northwest Science. 62(4): 151-160.

Tyser, Robin W.; Asebrook, Jennifer M.; Potter, Rachel W.; Kurth, Laurie L. 1998. Roadside revegetation in Glacier National Park, U.S.A.: effects of herbicide and seeding treatments. Restoration Ecology. 6(2): 197-206.

USDA Forest Service, 2004. Regional Forester’s Sensitive Plant Species List for the Northern Region. Unpublished.

Vencill, William K., Editor. 2002. Herbicide Handbook: Eighth Edition. Lawrence, KS: Weed Science Society of America. 493p.

Watson, V. J.; Rice, P. M.; Monnig, E. C. 1989. Environmental fate of picloram used for roadside weed control. Journal of Environmental Quality. 18(2): 198-205.

Final Integrated Weed Management EIS Page 213 Literature Cited

Wauchope, R. D.; Leonard, R. A. 1980. Maximum pesticide concentrations in agricultural runoff: a semiempirical prediction formula. Journal of Environmental Quality. 9(4): 665- 672.

Wilcove, D. S.; Rothstein, D.; Dubow, J.; Phillips, A.; Losos, E. 1998. Quantifying threats to imperiled species in the United States. BioScience. 48: 607-615.

Soil and Water Anderson, D.C., Harper, K.T. and R.C. Holmgren, 1982. Factors influencing development of cryptogrammic soil crusts in Utah deserts. Journal of Range Management, 35:180-185.

DiTomaso, J.M., 1999. Risk analysis of various weed control methods. Weed Science, as repeated in Techline. University of California, Davis, CA.

Dow AgroSciences, 1999. Picloram and the Environment. Technical Profile, November.

Haas, R.C, Scifres, C.J., Merkle, M.G., Han, R.R. and G.W. Hoffman, 1971. Occurrence and persistence of picloram in grassland water sources. Weed Research, 11:54-62.

Hickenbottom, J.A.S., 2000. A Comparative Analysis of Surface Erosion and Water Runoff from Existing and Recountoured Forest Service Roads: O’Brien Creek Watershed, Lolo National Forest, MT. M.S. Thesis, University of Montana, Missoula, MT. 178 pp.

Kenaga, E.E., 1969. Tordon Herbicides-Evaluation of Safety to Fish and Birds. Down to Earth, 25(1):5-9.

Lacey, J.R., C.B. Marlow and J.R. Lane, 1989. Influence of Spotted Knapweed (Centaurea maculosa) on surface runoff and sediment yield. Weed Technology, 3:627-631.

Norris, L.A., 1970. Degradation of Herbicides in the Forest Floor. In: Tree Growth and Forest Soils. Eds. Youngberg, C. T. and C.B., Davey. Proceedings, Third North American Forest Soils Conference. Oregon State University, Corvallis, OR. pg. 397-411.

Norris, L.A., 1971. The Behavior of Chemicals in the Forest. In: Pesticides, pest control and safety on forest and range lands. Witt, J. and J. Capizzi, eds. Oregon State University, Corvallis, OR. pg. 90-106.

Norris, L.A., 1981. The Behavior of Herbicides in the Forest Environment and Risk Assessment. Proceedings of the John S. Wright Forestry Conference. pg. 192-215.

Norris, L.A., H.W., Lorz and S.V. Gregory, 1983. Influence of Forest and Rangeland Management on Anadromous Fish Habitat in Western North America-Forest Chemicals. USDA Forest Service General Technical Report PNW-149. PNW Forest and Range Experiment Station, Portland, OR. 95 pp.

Olson, B.E., 1999. Impact of Noxious Weeds on Ecological and Economic Systems. In: Shelly, R.L. and J.K. Petroff (eds.), 1999. Biology and Management of Noxious Rangeland Weeds. pg. Oregon State University Press. Corvallis, OR. 438 pp.

Page 214 Final Integrated Weed Management EIS Literature Cited

Rothacher, J. and W. Lopushinsky, 1974. Soil stability and water yield and quality. In: Environmental effects of forest residues management in the Pacific Northwest: a state-of- knowledge compendium. USDA Forest Service General Technical Report PNW-24, p. D-1 to D-19. PNW Forest and Range Experiment Station, Portland, OR.

Sasich, J. and K. LaMotte-Hagen, 1989. Lolo National Forest Land Systems Inventory. USDA Forest Service, Lolo National Forest, Missoula, MT.