The preceding section provided the rationale for consistent and accountable policies, plans and programs for invasive species management, along with a means for evaluating their success or failure. This section discusses the disclosure requirements for planning documents in the context of invasive species management. Disclosure requirements provide valuable predictions and descriptions of the expected and likely effects of actions, which are critical to insure that policies and regulations are being followed. Case examples illustrate where disclosure requirements have not been adequately performed, with the result that projects may not have been carried out as intended and policy direction may not be met.

Section A. Planning documents must disclose all potential significant impacts, and provide detailed discussions and mitigation measures for all reasonably foreseeable impacts.

Invasive species management is a serious undertaking, with high potential for economic and resource losses and impacts. Such projects require prior disclosure of likely effects in order to insure that programs are effective, and remain within the limitations of policies and regulations. The National Environmental Policy Act (NEPA) is the guiding policy that provides for the descriptive documentation of a project through the development of an Environmental Impact Statement (EIS) or an Environmental Assessment (EA). These documents also provide the basis from which project inputs and outputs will be evaluated.

Pursuant to NEPA (§ 102 (2)(C)), every federal agency must prepare a complete detailed analysis of the environmental consequences and impacts of their proposed actions (see also 40 CFR § 1508.25, 1508.8, & 1508.9). The Council on Environmental Quality (CEQ) regulations clarifies what is meant by impacts (40 CFR § 1508.8):

an “impact” or “effect” includes ecological (such as the effects on natural resources and on the components, structures, and functioning of affected ecosystems), aesthetic, historic, cultural, economic, social, or health, whether direct, indirect, or cumulative.

[Alternatives shall] rigorously explore and objectively evaluate all reasonable alternatives, and for alternatives which were eliminated from detailed study, briefly discuss the reasons for their having been eliminated.

Environmental impact statements shall serve as the means of assessing the environmental impact of proposed agency actions, rather than justifying decisions already made.

Despite NEPA regulations, existing planning documents regarding invasive species management have not portrayed an accurate model of actual, expected impacts. Instead, planning documents are being used to rationalize predetermined decisions, while avoiding full disclosure. Weeds are vilified to the extent that chemical impacts appear acceptable in comparison, while strategies for preventing invasions are largely ignored and replaced by mitigation measures. The effect of such documents is to bias the choice of alternatives, because decision-makers must choose the alternative that appears to have the least impact. Without full disclosure of chemical impacts or prevention options, the choice of alternatives is prejudiced, policies are compromised, and projects are likely to fail.

Case example: Boulder Creek on the Okanogan NF in Washington

In the Okanogan NF Environmental Assessment for noxious weeds (1997), many required NEPA disclosures were never completed. Sensitive plant surveys were never performed in conjunction with the project, as required.

Herbicide label directions were not followed, resulting in a Notice of Correction from the Washington Department of Agriculture (2000). The Biological Evaluation accompanying the project (Molesworth, 1997) provided incorrect documentation to the National Marine Fisheries Service (NMFS) relating to likely effects of the project, biasing the determination of “no effect” on listed species.

Erosion increased into riparian areas following loss of native species from herbicide treatments ( Photo 4, p. 3), and adding to the burden of sediment in the stream. A swath of killed and damaged native vegetation averaging 10 feet wide followed the road for five miles ( Photo 2, p. 3), yet significant large infestations were left alongside the treated areas as a source of reinfestation.

The treatment was ineffective at killing weeds on many sites, and many of the weeds were able to go to seed. Livestock in the area browsed the seedlings of grasses planted to revegetate the roadside, leaving sites as poor as they were before. When forage became scarce in the fall, the livestock resorted to eating seeded plants of diffuse knapweed (Centaurea diffusa) and spotted knapweed (Centaurea maculosa) ( Photo 1, p. 3).

Monitoring was practically nil. The response to a year 2000 FOIA by Kettle Range Conservation Group, was that no monitoring at all was done on the Methow District, which had about half of the treated acreage on the Forest. Following the treatment, several new exotic species were found along the road (common tansy, Tanacetum vulgare, and spotted knapweed), leading one to question whether the original baseline monitoring was incorrect, or whether the treatment opened up new ground for these new invaders.

Warning signs were never posted and some sensitive individuals were never contacted. When these problems were pointed out to the Forest Supervisor (Wooten, 2000), the response was (O’Neal, 2000):

There is nothing in my staff’s monitoring report that suggests to me that the project was not implemented as planned, and that the results of the project were what we had expected.

• Planning documents must disclose all potential significant impacts.
• Planning documents must provide detailed discussions and mitigation measures for all reasonably foreseeable impacts.
• Planning documents should include a means for evaluating and reporting project expenditures and outcomes, to be used in determining project effectiveness and in adjusting plans and policies.

Many laws already have been enacted with site-specific requirements for management, e.g., the Clean Water Act, INFISH (Forest Service, 1995), PACFISH (Forest Service and BLM, 1995), and Best Management Practices (BMPs). To insure effectiveness, these measures have site-specific requirements (Mosley et al., 1998). With limited resources at their disposal, failure to prioritize treatments on a site-specific basis can result in wasting funds on lost causes while other controllable plant infestations get out of control.

Decisions based on Integrated Pest Management also require site-specific measures. True Integrated Weed Management begins with an honest, unbiased appraisal of the problem, including an examination of the reasons why invasive species are out of control, and it develops a solution based on the use of all available tools, which includes prevention, site-specificity, and adaptive management designed to respond to quantifiable, repeatable monitoring.

The importance for invasive species managers to take site-specific factors into account has been confirmed by biologists (Lonsdale, 1999; Woods, 1997; Hengeveld, 1989; Chicoine et al., 1988; Tyser and Worley, 1992; Weaver et al., 1989). Furthermore, species invasions are dependent on characteristics of both the invading species as well as the invaded ecosystem (Hobbs and Humphries, 1995; Randall, 1997). Thus, site- and species-specific measures should be incorporated into invasive species management.

• Planning documents must include site-specific analyses of the effects of all proposed chemical treatments.
• Planning documents must include site-specific analyses of the response of invasive species to alternatives.
• Planning documents must include species-specific analyses of the response of invasive species to alternatives.

Each management or multiple-use prescription must: 1) conserve soil and water resources and not allow significant or permanent impairment of the productivity of the land; 2) minimize serious or long-lasting hazards from flood, wind, wildfire, erosion, or other natural physical forces; 3) maintain diversity of plant and animal communities; 4) provide for adequate fish and wildlife habitat to maintain viable populations of native vertebrate species; and 5) maintain air quality at a level that is adequate for the protection and use of National Forest System resources and that meets or exceeds applicable Federal, State, and/or local air quality standards or regulations (ibid. at § 219.27(a)).

The National Environmental Policy Act (NEPA) requires that impacts be attended to, whether through elimination, avoidance, or reduction. The Council on Environmental Quality (CEQ) has promulgated regulations implementing NEPA which all federal agencies are required to follow. Again, these regulations specify that “environmental information” relevant to federal actions must be “available to public officials and citizens before decisions are made and before actions are taken.” 40 CFR § 1500.1(b).

In the case of invasive species, impacts caused by the invasion of non-native plant species, while bad, often result in a perceived need to use aggressive chemical controls with even greater negative effects, for which the need for analysis is correspondingly greater.

Forest Service planning documents are deficient in that they conceal, rather than inform, the public and decision-makers of known hazards of chemicals. According to Kovach et al., (1992):

Because of the EPA pesticide registration process, there is a wealth of toxicological and environmental impact data for most pesticides that are commonly used in agricultural systems. However, these data are not readily available or organized in a manner that is usable to the IPM practitioner.

Even when sufficient information is available to base an effects analysis upon, the Forest Service is unlikely to include any information that would contradict predetermined outcomes, such as a predetermined need to treat weeds with herbicides. In this way, even the most fundamental regulations regulating the use of poisons are ignored, such as the label directions required under 40 CFR § 152.

For instance, the labels for Tordon^© , a restricted-use pesticide, contain information that it is hazardous to nontarget plants, both crop and noncrop, and that it is restricted from use on subirrigated soils. Existing planning documents rarely provide an analysis of the occurrence of such plants and soils, nor possible impacts from Tordon^© applications on them. Instead, existing Forest Service documents have proposed broad-scale applications, combined with vague effects analyses so that herbicide treatments will be made to appear tightly controlled by numerous safeguards and the presence of qualified personnel. Furthermore, the broad hand waving and document padding that accompanies planning documents such as those of the Okanogan NF (1997, 1999), and Colville NF (1998) is a screen to hide the lack of required analyses of significant effects.

Nor can the Forest Service claim that methods for conducting impacts analyses are unavailable. The legal requirements to provide impacts analyses in planning documents can be accomplished with standardized methods of rating chemicals (Kovach et al., 1992):

A rating system was developed for the environmental impact quotient of pesticides called the Environmental Impact Quotient (EIQ) model, where l = least toxic or least harmful, 5 = most toxic or harmful. Data included Mode of Action, Acute Dermal LD50 for Rabbits/Rats (m&/kg), Long-Term Health Effects, Plant Surface Residue Half-life, Soil Residue Half-life, Toxicity to Fish-96 hr LC50, Toxicity to Birds-8 day LC50, Toxicity to Bees, Toxicity to Beneficials, Groundwater and Runoff Potential . . . The impact of pesticides on terrestrial systems is determined by summing the toxicities of the chemicals to birds, bees, and beneficial arthropods. . . .. . . After the data on individual factors were collected, pesticides were grouped by classes (fungicides, insecticides/miticides, and herbicides), and calculations were conducted for each pesticide. When toxicological data were missing, the average for each environmental factor within a class was determined, and this average value was substituted for the missing values. Thus, missing data did not affect the relative ranking of a pesticide within a class. . . .

Traditional Pest Management Strategy - Total Environmental Impact = 938 Integrated Pest Management (IPM) Strategy - Total Environmental Impact = 167

Surfactants labeled as “inert” constituents of herbicide formulations are now believed to behave as endocrine-disrupting compounds in wildlife and humans. Cumulative effects of multiple chemicals used over time have rarely been analyzed. In addition, the identity of formulations containing “inert” ingredients has been kept as a trade secret from the public, in violation of NEPA disclosure requirements.

A 1996 ruling by District of Columbia Federal District Court found that pesticide manufacturers’ concealment of the identity of “inerts” as “trade secrets” was unsubstantiated (NCAP et al. v. Browner, 1996). In addressing the case brought by Northwest Coalition for Alternatives to Pesticides (NCAP) and National Coalition Against the Misuse of Pesticides (NCAMP), the court opinion clarified that these chemicals are not exempt from the Freedom of Information Act.

In NCAP et al. v. Browner, the Court ruled that EPA must disclose secret ingredients in pesticides. The ruling protects the right of public access to secret chemicals in pesticides by requiring the EPA to provide information about the identity of so-called “inert” ingredients in pesticide products.

Unfortunately the ruling came too late to classify the risks of harm from hundreds of project plans based on incomplete testing of the environmental and health effects of only “active” ingredients. Since pesticide manufacturers were not required to reveal the identity of “inert” ingredients before 1996, the toxicity of formulations applied on the ground was generally not analyzed in documents prepared before that time. Pesticides also are generally not tested for synergistic or cumulative effects, are inadequately tested for neurotoxicity and immunotoxicity, and only recently have begun to be tested for their ability to disrupt the endocrine system.

The so-called “inert ingredients” revealed following this decision (NCAP et al. v. Browner), have been shown in some cases to be more toxic than the active ingredients. According to the EPA web site (www.epa.gov/opprd001/inerts), a chemical that is an active ingredient in one pesticide product may be considered an “inert” in another. The toxic nature of full herbicide formulations has not been adequately disclosed as required by NEPA. Instead, the disclosure of harmful effects from herbicides had to wait for citizens, such as this statement presented to the Forest Service by O’Brien (1997):

The Region 6 herbicide information profile for picloram indicates that . . . ‘No ingredient in any picloram formulations was categorized by EPA to have evidence or suggestion of toxic effects.’ In fact, Region 6 doesn’t know what is in Tordon K or Tordon 22K, but published its assurance of low concern on the fact that Tordon’s inert ingredients are either on List 4 of inerts, which are generally recognized as safe, or on List 3, which [Forest Service] Region 6 characterized as being, ‘low priority for health effects testing based on absence of data or chemical structures that would indicate toxic effects.’

This is not true. List 3 is the list of "Inerts of Unknown Toxicity", and as Holly Knight, an intern at NCAP discovered, even this is not true: 1,981 pesticide inerts hide on List 3, including 264 pesticide active ingredients, some of which are known to be highly toxic, including naphthalene (which can cause brain damage, convulsions, and death in children), chlorothalonil (a probable carcinogen), and chloropicrin (a respiratory tract irritant that can cause asthma). Other inerts that are not active ingredients, are likewise well-known to be of toxicological concern.

In other words, what you don't know about the constituents of Tordon K or 22K, or any other pesticide formulation could adversely affect your workers, wildlife, and humans who are exposed in water, air, or food to these chemicals.

Even if you do know what you are spraying, you do not necessarily know how the diffet formulation components interact. For instance, you know that Roundup contains a surfactant in addition to the active ingredient, glyphosate. A surfactant enhances contact of the active ingredient with the plant's surface cells. Does it enhance uptake of glyphosate by cells in wildlife or the workers spraying it? Since chronic effects testing is not required for full formulations, you know little about the consequences of exposing wildlife or workers to the combination of glyphosate and a surfactant.

Over 650 chemicals that have been identified as hazardous by federal, state, or international agencies are hiding behind the misleading word “inert” in pesticide products, according to a report released today by the Northwest Coalition for Alternatives to Pesticides. “Worst Kept Secrets: Toxic Inert Ingredients in Pesticides” documents the hazards of so-called “inert” ingredients, over 2,500 substances that are added to pesticides but are not named on product labels. Regulatory agencies have few requirements for toxicological or ecological effects testing of inerts. Despite this lack, the new report shows that over 25% of the chemicals used as “inerts” actually have been identified as hazardous.

NCAP’s report recommends that all pesticide ingredients be fully disclosed on product labels. In addition, all health and safety testing required for pesticides should use the complete pesticide product, including all so-called “inert” ingredients.

The decision to continue the use of herbicides without analyses of “inert” ingredients appears to be an arbitrary and capricious decision by the Forest Service. Now that herbicide formulations have been shown to contain toxic ingredients which formerly masqueraded as “inert”, the use of planning documents which ignore those effects violates NEPA disclosure requirements. As described by NCAP’s Grier (1994):

Most safety tests for pesticides are made only on the active ingredient and not on the whole product. “inert” ingredients can be more toxic than the active ingredient and comprise up to 99% of a pesticide product. “inert” ingredients are any of over 2,300 substances that are added to pesticides but are not named on product labels. Despite their name, they are neither biologically, chemically, or toxicologically inert. According to the EPA web site (www.epa.gov/opprd001/inerts), a chemical that is an active ingredient in one pesticide product may be considered an “inert” in another. Furthermore, the toxicity of most of the “inert” ingredients allowed by the EPA in registered pesticides is unknown. So-called “inert” ingredients laws allow the application of toxic compounds such as kerosene, diesel fuel or fungicides to be used as 98% of a mixtures application rate.

. . . more than additive and this raises doubt that the LC50s [concentration required to kill half of the test group] reported for Roundup^© in reconstituted water are applicable to natural waters.

Case example: Appeal of the Okanogan NF Environmental Assessment (1997) for inadequate disclosure of effects

Risks from the herbicide picloram have resulted in recommendations from the EPA to withdraw its registration; however, neither the EPA recommendations nor the risks appear in Forest Service sanctioned specimen labels, (e.g., the specimen label for Tordon 22K^© , revised 05-24-00). Neither are the risks disclosed in the Okanogan National Forest’s Environmental Assesments (EAa) on noxious weeds (1997, 1999, 2000).

Yet the risks of picloram harm to non-target plants is extremely high, as shown by the following EPA Ecological Effects Branch (EEB) recommendations to withdraw its registration (Abramovitch, date unknown):

Due to the extreme phytotoxicity, its persistence under typical environmental conditions, and its extreme propensity to leach into groundwater in all soil types, the EEB is strongly recommending against the reregistration of all active ingredients of Picloram. This conclusion is based on the extreme exceedance of the acute levels of concern for non-endangered and endangered terrestrial plants. The risk quotients (RQ) are exceeded as follows for the various application methods. . . .. . . In 1989 EEB received incident data from a private citizen who cited 30 incident reports of plant damage resulting from surface runoff or leaching into groundwater. Based on these reports, EPA is requesting additional phytotoxicity data for potatoes and other sensitive crops including tobacco, soybeans, corn, pasture, watermelons, tomatoes, bell peppers, and hay. Further, there are sufficient data to state that picloram will likely cause serious adverse effects in nontarget terrestrial plants. These additional data are confirmatory and will support our risk assessment.

Based on all available data EFGWB believes that picloram should not be reregistered because its use would pose unreasonable adverse effects to the environment. Because of picloram’s mobility in all soil types and its persistence under normal ambient conditions, no practical use restriction can prevent it from contaminating the environment surrounding the target site.

The lack of disclosure of picloram’s negative effects to wildlife, plants and soils were brought to the attention of the Forest Service in a notice of appeal filed on October 10, 1997, by the Lands Council, Kettle Range Conservation Group, Methow Forest Watch and the Leavenworth Audubon Adopt-A-Forest:

The FEIS, the Mediated Agreement (1989) and the Guide give direction for site-specific risk assessment/risk analysis, but that direction was not followed in the ONF EA. Herbicide profile information is not included in the EA. The USDA Forest Service rating of marginal adequacy for the quality of health effects testing for the proposed herbicides, picloram and glyphosate, is not included in the EA or discussed. Inert ingredients are not disclosed, when even the EA acknowledges that, “Technical grade glyphosate was less toxic (LC 50,140 mg/L), than Glyphosate formulations or the surfactant (LC50, 2 mg/L) (sic)”. The significant 1996 amendment to the Federal Insecticide, Fungicide, and Rodenticide Act is not included or acknowledged as a reference (EA p.9).

In appealing the EA, the groups reminded the Forest Service that analyses must be available for public review. In defiance, the appeal was denied by the Regional Forester and the required analyses were never provided.

Case example: Boulder Creek on the Okanogan NF in Washington

The Environmental Assessment on noxious weeds (Okanogan NF, 1997) was accompanied by a Biological Evaluation (BE) (Molesworth, 1997) that determined, without substantial review, that there would be “no effect” from the chemical treatment of nearly 6,000 acres in the project.

No damage thresholds or risk analyses were presented for the herbicide applications. The EA presented biased effects analyses and failed in some cases to follow its own prescriptions, for instance in implementing treatments in violation of label directions. The EA failed to disclose important effects information provided by the Washington office (Syracuse Environmental Research Associates, 1996):

The primary hazard to non-target terrestrial plants is from unintended direct deposition or spray drift. Unintended direct spray will result in exposure equivalent to the application rate. As discussed in the dose-response assessment for terrestrial plants (Section 4.3.3), such exposures are likely to result in adverse effects to a number of plant species. . . . . . . Direct deposition, either through unintentional direct spraying or spray drift does present a plausible hazard. If plants are accidentally sprayed at the application rates used by the Forest Service, they are likely to be damaged, particularly in the upper ranges of anticipated application rates. This kind of exposure may be regarded as an accidental scenario, which is relatively easy to control with proper management and application. The extent and duration of damage will depend on the time of application and plant species. . . .

. . . (1992), this could damage some sensitive plant species. . . .

. . . Glyphosate can reduce the emergence and weights of progeny seedlings on crops such as corn, soybeans, and Johnson grass (Jeffery et al., 1981). It is not clear whether this effect is caused by direct toxic action on the seeds or simply reduced vigor in the parent plant as the seeds develop. . . .

. . . Gross signs of toxicity, which may not be apparent for 2–4 days in annuals or for more than 7 days in perennials, include wilting and yellowing of the vegetation, followed by browning, breakdown of plant tissue, and, ultimately, root decomposition. . . .

. . . In addition to these laboratory bioassays, there are several field studies that have assessed the effects of glyphosate on terrestrial organisms (Appendix 2-2). . . In most cases, the effects noted were changes in population density that reflected changes in food availability or suitable habitat. . . .

. . . Glyphosate residues or perhaps residues of adjuvants used with glyphosate have been shown to affect grazing preference in cattle (Jones and Forbes, 1984) but not sheep (Kisseberth et al., 1986).

Case example: Boulder Creek on the Okanogan NF

Hand application of glyphosate within the buffers, the spill plan and mixing requirements, the small amount of herbicide near water at any given time, the rapid dilution of chemicals if they do enter the water all keep the potential effect to listed fish at a non?measurable level. Any herbicide that does enter the water should be an insignificant amount.

The findings of Newton should have been enough to garner a determination that the project “may affect” threatened or endangered species. In fish exposed to 2.0 mg/L of Roundup© (another glyphosate formulation) the fillets contained 80 mg/kg of glyphosate and the eggs contained 60 g/kg. Roundup© is four times more toxic to rainbow trout fry and fingerlings than to larger fish. Significant increases in stream drift of midge larvae were observed after the 2.0 mg/L of Roundup© treatment. The toxicity of Roundup© to rainbow trout and bluegill increases with increasing temperature. Roundup© was about twice as toxic to rainbow trout at 17 degrees C than at 7 degrees C. It is also more toxic to bluegills at 27 degrees C than at 17 degrees C. Roundup© was more toxic to rainbow trout and bluegills at pH7.5 than at pH6.5. Technical glyphosate was less toxic to fish at a higher pH, but the surfactant appears to be more toxic at the higher pH. Solutions of Roundup© aged for up to 7 days in reconstituted water at 12 and 22 degrees C did not change in toxicity to midge larvae, rainbow trout, or bluegills. Applications of Roundup© to ditchbanks near aquatic ecosystems may be hazardous to resident fauna, particularly if the water temperatures are elevated or the pH exceeds 7.5. This is important because glyphosate causes water temperatures to increase for several years following treatment (Holtby and Baillie, 1987).

Since the publication of Newton in 1984, reports of the toxicity of glyphosate formulations on fish and wildlife have become more widespread in the scientific literature. The Forest Service was given ample information through publicly submitted scoping comments on the EA to avail themselves of this literature, yet they chose to ignore it and implement the project without full disclosure, in violation of National Environmental Policy Act (NEPA) procedures, and Endangered Species Act (ESA) consultation procedures.

• Planning documents must disclose and analyze the full range of potential impacts.
• Planning documents must present the full range of potential outcomes, pro and con.
• Planning documents must disclose and analyze the full range of potential environmental effects of proposed chemical applications.
• Planning documents must disclose and analyze the full range of adverse effects on humans resulting from proposed chemical applications.
• Planning documents must be comprehensive; issues analyzed must address the full scope of project impacts; all known cumulative and indirect effects must be disclosed and analyzed.
• Planning documents must include a comprehensive list of citations referenced in effects analyses.

Staff in charge of data collection and interpretation have become so fearful of criticism that all negative connotations in their data are anticipated and counteracted prior to public release. This is unfortunate, in that the attachment of value judgments should follow, not precede data interpretation. The fault lies partly in our system, which assigns funding priority based on target attainment, and reinforces a military-like code of conduct to punish any case where the public complains. Managers need to be free to state facts as they are, and realize that; ‘any disturbance and any management regime will be good for some species and bad for others.’

. . . (i)t is intriguing to contemplate how the Forest Service would approach its management of the public’s National Forests differently if the agency knew that it would no longer have herbicides available. The simple availability of herbicides is a siren song to continue noxious weed-favoring activities.

The buzzword in all these projects has been the claim that they are using “Integrated Weed Management”. However, true Integrated Weed Management (IWM) begins with an honest, unbiased appraisal of the problem, including an examination of the reasons why invasive species are out of control, and then develops a solution based on the use of all available tools, which includes prevention, site-specificity, and adaptive management, in response to quantifiable, repeatable monitoring.

In contrast to the use of true IWM, the documents above are deficient in their analyses of undesirable effects of chemical treatments but spend an inordinate amount of time demonizing weeds in a way that appears to justify the benefits of short-term chemical treatments (resulting in the unstated need for long-term program funding for herbicide projects). The Okanogan NF (1997, 1999, 2000) readily listed the harmful environmental effects resulting from invasive species, including loss of native communities, loss of endangered species and loss of wildlife habitat. The remedies for these problems are obvious—removal of noxious weeds. Yet this solution is often complicated. The majority of weed populations are not pure infestations, and any non-specific measures will result in killing as much or more native species during the weed removal process.

Following short-term removal of weeds and off-target species, the result may be an environment of unstable denuded, soil, which is highly susceptible to re-infestation from the existing seed bank. The site may become too harsh for revegetation using native species, requiring that restoration be accomplished by encouraging rhizomatous species to take over, perhaps followed by reseeding with “beneficial” species such as timothy, orchard grass, Kentucky bluegrass, and Dutch clover. Yet these “beneficial” European species can be more invasive than the weeds replaced and are relished by livestock for their high protein content.

The result of such short-term treatment is that either the site gets worse, or it becomes converted into pasture, which rapidly becomes denuded again as livestock and wild ungulates are preferentially drawn to the palatable forage. In the process, soil quality worsens, and it becomes more disturbed and erosive—perfect conditions for initiating a new round of weed invasion, brought by seeds carried on livestock and wildlife. This cycle of ever-worsening invasion effects and stock-mediated reinfection has brought many acres of rangeland in the western states to a state of near or total ecological collapse (Belsky and Gelbard, 2000).

Prevention measures are rarely discussed in planning documents, an effort that tends to isolate chemical treatment as the “best alternative”. Indeed, planning documents such as the weed treatment EA on the Okanogan NF (1997) summarized this biased viewpoint on p. 87 to be,

Chemical treatments would provide a much increased long term noxious weed control success [sic].

Nearly all systems are likely to be nonequilibrial in the future; we must be activists in determining which species to encourage and which to discourage. We cannot just manage passively, or for maximal diversity, but must be selective and tailor management to specific goals.

Case example: Untested assumptions and biased analyses

The EA then built upon this assumption by adding that the restoration seedings would effect restoration of the site, resulting in an “increase in native plant populations and grasses.” In reality, the revegetation attempts failed in many instances, as corroborated by an independent photographic survey of the area ( Photo 1, p. 3; Wooten, 1999d) as well as photographs taken by the Forest Service (Bennett, 1999). This could be due to several reasons: (1) because cattle grazing is nearly ubiquitous on the Forest and the young seedlings of European grasses were hungrily devoured ( Photo 1, p. 3); or (2) because the seeds fell on unproductive soil and never germinated; or (3) because the seeds were never planted as stated.

The implication that native bunchgrass communities and soil stability would benefit from herbicide treatments was a naive assumption, and in fact the herbicides destroyed native species at a much higher rate than weeds were killed as documented in the photographs (Wooten, 1999d).

Case example: Use of research with untested assumptions

However, in an independent, controlled study of the effectiveness of hand pulling Dalmatian toadflax, Jeffries (2000) found that hand pulling was very effective and resulted in a reduction of over 96% of toadflax plants with very little tendency to re-sprout from seed or root fragments. Furthermore, Jeffries found that many other methods were also effective against toadflax, including mowing, burning, and sheep grazing.

This demonstrates how the Forest Service uses research as a tool to skew projections, rather than as an unbiased estimate of likely results.

The Harrod study made a number of assumptions that should have been discussed in the EA. Harrod stated that toadflax was not palatable to livestock, however in contrast, Jeffries found that sheep seek it out. Harrod’s study did not describe how hand pulling was accomplished; yet concluded that hand pulling was ineffective. In contrast, Jeffries notes that if the crown is removed, the plants cannot resprout in the time period observed by Harrod.

Case example: Failure to disclose information results in biased decision

In an appeal of the Environmental Assessment, Decision Notice, and FONSI for the 1999 Okanogan NF Integrated Noxious Weed Management Program, (Lands Council et al., 1999) the appellants noted that:

The Purpose of the Integrated Weed Management (IWM) Project is to treat 15 sites that have noxious weeds, including riparian areas. The Okanogan NF attempted and failed to disclose why Alternative C was chosen over Alternative B. By selectively evaluating the two alternatives without a full analysis of treatment effectiveness, cost, and ability to maintain watershed integrity, the Forest has biased its Decision.

Appellants rightfully contended that the analyses in the EA were incomplete, since under Alternative C, herbicides would also kill substantial numbers of non-target species, whereas Alternative B would primarily only kill those species that were accidentally trampled or uprooted during treatment. The failure to compare the effects of these two alternatives fairly may have swayed the decision-maker to choose the alternative that was falsely portrayed to have less effects on native plants.

Case example: Boulder Creek on the Okanogan NF in Washington

The EA claimed that, “areas that would be successfully treated would provide more native vegetation and thus be beneficial to wildlife.” (p. 112). Yet the EA failed to mention that herbicides would denude the roadsides of native vegetation as well, during typical applications (Wooten, 1999d; Photos 2, 4, p. 3).

The EA is quick to point out the insidious nature of noxious weeds wherever possible, while ignoring the same effect when it is likely to happen from herbicide treatments, e.g., the following statement about aquatic habitat: “noxious weeds hinder the following objective: Maintain and restore the sediment regime under which aquatic ecosystems evolved.” (p. 104). Thus, Alternative A (no treatment) was rejected because (p. 106),

. . . this alternative would not meet the objectives of the Aquatic Conservation Strategy listed above because of the possibility of increased sedimentation and decreased water quality, and the threat to native riparian vegetation by noxious weeds.

. . . (t)his alternative would improve water quality by controlling noxious weed populations contributing to sedimentation. This alternative would improve fish habitat by improving the sediment indicator.

There is nothing in my staff’s monitoring report that suggests to me that the project was not implemented as planned, and that the results of the project were what we had expected [sic emphasis ours . . . a subconscious slip of the pen?].

In the case of the Boulder Creek road slopes, the problem is not noxious weeds; it is erosion into the stream caused by the road and livestock being in too close to riparian habitats. The EA was not true IWM; the plan was biased, it didn’t examine the causes of the problems, it avoided preventive measures, it did a poor job of incorporating monitoring into the management system, and it relied primarily on herbicide treatments rather than using all available tools. The EA overtly attempted to debunk any treatment other than herbicide and was driven by a biased motive to force the selection of a predetermined alternative.

The risks of ground-based herbicide applications exceeding damage thresholds to nontarget species was not presented. Effects analyses were biased, treatments analyzed were not the ones that occurred, and different herbicide formulations were used from ones specified in the EA. The EA provided unreliable information about the impacts of herbicide applications on the Okanogan NF.

In the zeal to obtain project funding on the Okanogan NF, the EA biased the discussion of impacts and failed to disclose the actual harm that would occur to aquatic habitat from increased erosion due to the off-target herbicide killing of valuable streambank vegetation. The Forest should consider performing a comprehensive restoration program designed to correct or prevent the damage caused by weeds, livestock, roads, and now herbicides on the National Forest. The EA wasted the allocated funds and should serve as an example of how not to manage invasive species.

• Decisions must not be biased toward the choice of a predetermined alternative.
• Planning documents should analyze studies that offer different viewpoints.
• Project objectives should follow the goals of an effective system of invasive species management.
• Projects should follow a process of true Integrated Weed Management, which begins with an honest, unbiased appraisal of the problem, including an examination of the reasons why invasive species are out of control. Actions should be developed based on the use of all available tools, including prevention, site-specificity, monitoring and adaptive management.
• Invasive species management projects should incorporate the results of monitoring. Monitoring should include operational indices such as the ratio of increasers and decreasers used in range management.
• Ineffective programs should be replaced with comprehensive restoration programs that address and correct the causes of plant invasions, while preventing further damage caused by invasive species and inappropriate treatments.

Forest Service personnel are discouraged from providing references that might contradict their agency’s agenda to use herbicides without adequate review. The Washington Office of the Forest Service funded an extensive review of research on the effects of glyphosate (Syracuse Environmental Research Associates, 1996); however, individual Forests appear oblivious to this resource available within their agency and fail to cite it in planning documents proposing the use of the herbicide (Okanogan NF 1997, 1999; Colville NF, 1998; Wenatchee NF, 1998). A letter from the Washington Office (McDougle, 1999) makes it apparent that the Washington Office does not actually want Forests to have new information that implicates adverse effects from glyphosate, and in fact, that they blocked a request to provide an additional 78 references to glyphosate toxicity to humans and the environment.

On the Santa Fe NF (2000), agency personnel used a risk assessment procedure based on outdated studies from the 1980s in their noxious weed EA. If Forests are going to rely on outdated data, then it should be only because current data is unavailable. In any case, there needs to be a far greater margin of safety incorporated into actions. According to Tickner (1997):

Given the limitations of science to address emerging environmental problems, such as endocrine disruption, there is a significant need for the development of new public policy approaches to anticipate and prevent harm to human health and the environment. The question of what society should do in the face of uncertainty regarding cause and effect relationships is necessarily a question of public policy, not science.

In recommending changes to the National Forest Management Act (NFMA), the Committee of Scientists (1999) recommended that the Forest Service should establish a science and technology advisory board with a primary goal of helping collaborative planning become a reality on the national forests and grasslands. This board would provide highly qualified and independent advice to the Forest Service to assure that the most current and complete scientific and technical knowledge is used as the basis of land and resource management.

Case example: Label violations

The label violation is by itself alarming. However, in trying to counter claims that the application’s effectiveness had been compromised by violating the label directions, the Forest Service then disclosed that a second label violation had occurred, namely the use of a different pesticide than the one called for. Photographs provided by the Forest Service of riparian areas (Bennett, 1999) clearly state that Tordon 22K was used, in violation of riparian buffers set up in the EA.

As to the receipt of a Notice of Correction for not following label directions, the Forest Supervisor indicated that he did not consider this significant (O’Neal, 2000). The ready acceptance of label violations as a cost of doing business may very well be because it was actually the lesser of the second, more sinister crime—using picloram on riparian soils where it is not allowed—which was only uncovered later and is still under investigation by the Washington Department of Agriculture.

• Planning documents should be prepared by qualified personnel.
• Planning documents that propose herbicide treatments should be reviewed by qualified medical and chemical experts.
• Planning documents must be based on a thorough, comprehensive review of up-to-date and peer-reviewed scientific studies.
• Planning documents should include references to peer-reviewed studies that provide a wide range of conclusions on potential effects.
• Risk assessments should be supplemented with precautionary principles that act in advance of scientific certainty to prevent harm to humans and the environment ( Appendix A).
• The Forest Service should coordinate projects through a scientific advisory board that can offer qualified and independent advice about projects.

Projects which manage invasive species, must have a stated purpose and need, from which goals and objectives follow. Projects should also include benchmarks to validate whether the goals and objectives are effective in accomplishing the purpose of the project.

Despite large expenditures, invasive species are increasing beyond our means of control. The prospect of reversing this trend is not likely, even with vastly increased expenditures. Existing decisions are based on assumptions that actions will bring invasive species under control, when in fact, this hasn’t happened. For the vast majority of cases, the chosen means have not justified the ends.

Because of the great deal of complexity and uncertainty that accompanies the fledgling “science” of invasive species management, it is imperative that managers err on the side of precaution, follow laws designed to protect human health and the environment and assure that decisions are made through an open public process.

As agencies gain experience in managing invasive species along with the conflict this generates in the affected public, a more solid foundation for choosing alternatives will develop. The Committee of Scientists (1999) defined the mechanism for assuring that the purpose and need for projects are being accomplished by the treatment methods used:

Validation monitoring asks, are the basic assumptions about cause-and-effect relationships used to predict the outcomes of strategies and pathways of treatments valid?

Case example: Okanogan NF Weed Programs, 1997-2000.

The Okanogan NF Integrated Weed Management Environmental Assessment (EA) (2000) stated as its purpose to, “control or eliminate the existing noxious weed populations on the included sites.” Essentially this same statement was included in the Okanogan NF’s EAs for the 1997 and 1999 programs, which cost over $300,000. In the 2000 EA, the Okanogan NF program proposed to continue the program already begun, at the cost of $763,300 to control only 75% of the identified infestations over a five-year period.

It can safely be stated that the 1997 EA did not result in control or elimination of existing weed populations, except in a limited area of the Forest for a limited amount of time. Most of the sites treated under the 1997 EA were roadsides, because the majority of Okanogan NF weed infestations away from roads have never even been inventoried.

The purpose and need was a necessary statement which can be agreed upon by all concerned, but nonetheless complete elimination of weeds is unattainable under the methods given, except for very limited infestations of weeds, and then only until re-infestation occurs. Such a regular program for weed control is termed a maintenance program, and such programs are attainable only in limited circumstances. In addition, maintenance treatments typically need to be repeated every year until the seed bank is exhausted. An example of a successful maintenance program would be along railroads, where the railbed is kept in a continually sterile state though regular applications of herbicides.

The goals for invasive species management should follow from the stated purpose and need, followed in turn by objectives that can accomplish the goals. The Okanogan NF 2000 EA is a clear indication that the goals of the 1997 EA were never attained and should be revisited before proceeding further with costly approaches destined to fail. However, instead of documenting this failure through validation monitoring and adjusting the program through adaptive management, the Forest hardens its resolve to ignore all warnings and repeat its past mistakes.
 
 

• Projects which manage invasive species must have a stated purpose and need from which goals and objectives of the project will follow.
• Projects should include benchmarks, which can validate whether the goals are accomplishing the purpose of the project.
• Projects should incorporate evaluation and reporting procedures which will allow program goals and implementation procedures to conform to regional and national policies.

Under FIFRA, the use of pesticides involves a weighing of both the harm as well as the benefits resulting from their use. In deciding to use herbicides, the Forest Service has a responsibility to provide risk analyses that weigh the costs and benefits of herbicides. Without a risk analysis or worst-case scenario, it is impossible to weigh the relative costs and benefits of using an herbicide.

The need for a risk-based approach was recently emphasized by the National Invasive Species Council Management and Restoration Working Group (2000):

The effective management of invasive species depends upon several underlying capabilities: 1) the establishment of priorities based upon a science-based assessment of risks, . . .

Most people seriously overestimate the amount of protection given them by governments regarding pesticide safety. Congress found that 90% of the pesticides on the market lack even minimal required safety screening (American Defender Network, 1989). Of the 34 most used lawn pesticides, 33 have not been fully tested for human health hazards (Davidson, 1994). If any tests are done, they are performed by the chemical manufacturers, not the EPA. ‘If a chemical company wanted to, they could start with a desired conclusion, and skew the data, and the EPA would never know’, notes David Welch, an entomologist with the EPA’s Office of Pesticide Programs. Welch did a random sampling of 15 pesticide files and found 13 without proper reviews (Sayan, 1991). One third of the most commonly used lawn pesticides were illegally registered for use. Despite the fact executives of Industrial Bio-Test labs were given jail terms for faking pesticides tests, the chemicals are still on the market (American Defender Network, 1989). Shortages in funding, personnel, and interference from business has slowed re-evaluation of these chemicals (General Accounting Office, 1993). Even when the EPA does refuse a pesticide registration, the manufacturer often files a lawsuit, which keeps the chemical on the market (Sayan, 1991).

An important scientific principle worth following is that when opinions differ, both sides get to see the other side’s methodology. The practice of ignoring or concealing data that does not agree with the status quo is not only unscientific, but can lead to disastrous consequences, as explained in this story in the Bend Oregonian (Sunday, December 5, 1999):

For decades, scientists had believed that people reacted the same way to toxic chemicals, regardless of their age. But as scientists began noticing learning disabilities in children, they suspected the old assumptions might be wrong. Then, in the 1970s, one major study after another proved children were in fact far more sensitive to lead than adults. Within a few years, scientists found they had dramatically underestimated the way lead hurts the developing brains of children.

Instead of asking how much damage or harm we will tolerate (which is the approach taken by risk assessment), the precautionary principle asks how to reduce or eliminate hazards, and it considers all possible means for achieving that goal, including scrapping the proposed activity. (Of course, alternatives to a hazardous activity must be scrutinized as carefully as the hazardous activity itself.)

Case example: Okanogan NF Noxious Weed EA, 1997

For instance, under the No-action Alternative, the EA warns against the likely hazards of weeds:

This alternative [A] would not meet the objectives of the Aquatic Conservation Strategy listed above because of the possibility of increased sedimentation and decreased water quality, and the threat to native riparian vegetation by noxious weeds. [p. 106] . . . Alternative A, “may effect but is not likely to adversely effect” peregrine falcon, gray wolf, grizzly bear, northern spotted owl, and bald eagle. [p. 112]

Weed species that are resistant to non-chemical control would continue to spread [p. 106] . . . Fish spawning and rearing habitat may be impacted slightly in the short-term due to ground disturbance caused by manual and mechanical treatments. . . . Water quality would continue to be slightly degraded by sediment from adjacent noxious weed populations. [p. 107]

This is not an honest appraisal of available research, it is blatant rationalization. The lone reference cited in support of their action is badly out of date. The analysis lacks quantification; instead it uses terms like “rapidly” and “tends”. Unfortunately, the Forest Service continues to deny the public an objective risk assessment of the use of herbicides.

Case example: Inadequate worst case analysis leads to injunction.

In 1984, the Ninth Circuit Court of Appeals heard a case in which plaintiffs challenged the spraying of herbicides on Forest Service and BLM lands (Save Our Ecosystems v. Clark, and Merrill v. Block, 1984). The court ruled against the Forest Service and concluded that it could not simply rely on the EPA registration process for herbicides under FIFRA because that process is inadequate to address environmental concerns under NEPA. The Court concluded that the Forest Service must do research if no adequate data exists. This requirement follows from the requirement to do a worst case analysis.

The District Court had previously enjoined portions of the Forest Service spray program in Merrill v. Block and portions of the BLM program in Save Our Ecosystems v. Clark, but the injunctions were modified by the Ninth Circuit to enjoin all spraying in their respective regions until the agencies complied with NEPA. In the BLM case, subsequent to a prior requirement by the Ninth Circuit Court (SOCATS v. Clark, 1983), the BLM prepared a worst case analysis for its spray program on the Eugene District. But the District Court concluded that the worst case analysis was inadequate because it was brief and cursory and proceeded from the wrong assumption. The Court concluded that the worst result that could occur as a result of preceding in the face of uncertainty as to whether the herbicide could cause cancer, is that it does, in fact, cause cancer. That the agency considers it speculative or remote is not sufficient to justify failure to analyze the worst case.

In abdicating their responsibilities, the agencies have lost a large degree of public confidence.

• Action alternatives which propose herbicide use must demonstrate an overwhelming public need.
• Planning documents must provide risk assessments for herbicides.
• Planning documents should provide risk assessments for invasive species.
• Planning documents should consider supplementing risk assessments with precautionary principles.

The avoidance of herbicides is more than a requirement—it is a basic tenet of integrated pest management, or IPM (Wooten, 1999b). According to Pimentel (1999),

The basic premise of IPM centers on employing first biological and other non-chemical pest controls, with the use of chemical pesticides only as a last resort.

True IPM is an interdisciplinary system of techniques for controlling invasive plants that is both practical and environmentally sensitive.

Other physical methods may involve modification of the natural disturbance regime. Methods include controlled burning, flame weeding or flameless radiant energy (Prull, 2000). Commercially available radiant heat weed control equipment is available through the Swiss company Messerili Sessa, with inexpensive units beginning around $300. The principle of heat is to burst plant cells, causing loss of fluids and thermal denaturation.

Site modification methods include the amount and timing of shading (Elmore, 1993b) and mulching. Straw mulch 3.5 inches thick gave 98 percent control of yellow star-thistle, Centaurea solstitialis,(Dremann, 1996).

Chemical methods of plant control include using allelopathic chemicals (compounds that retard other plants from growing nearby). Some allelopathic chemicals are found in the weeds themselves, e.g., cnicin, a knapweed extract being tested under an EPA grant to the University of Colorado as a chemical agent to increase the effectiveness of biological control insects on knapweeds (Centaurea spp.) Another natural chemical is corn gluten meal, a byproduct of the corn milling process, which has been found to be more effective in garden weed inhibition than chemical herbicides (Consumer Reports, March 1999).

Modification of climatic and diurnal factors also have a strong influence on weeds. Wind entrainment through the use of barrier or “snow” fences has been successfully used to capture “tumbleweed” forms of weeds, preventing their spread, and allowing easier control by burning or plant removal (Roché, 1994).

Restoration grass seeding has been used to decrease leafy spurge (Euphorbia esula) by 67 percent in two years in Minnesota (Biesboer et al., 1994). Competitor enhancement can also help improve the chances for outcompeting weeds, for example by pruning desirable species, whether using a mechanical tool, or through selective grazing (Elmore, 1993). Range experience is needed in this technique however, because livestock can also benefit weeds ( Photo 1, p. 3).

Prevention methods are perhaps the single most important tool for controlling the spread of invasive species. Prevention methods include prioritization of new invaders, use of signage to alert the public, and perhaps most importantly, constraints on soil disturbance and seed transport that avoid the most common entry routes for seeds: contaminated seeding mixtures, hay, mulch, topsoil, road gravel, nursery stock, and manure (Quarles, 1999; Olivarez, 1995).

Prevention opportunities are lost if attention is not given to adjacent infestations. Vehicle quarantines and vehicle washing should be part of regular operations. Prevention means taking a hard look at the cause, not the symptoms, of the spread of invaders. If the causes are eliminated, weed spread can often be eliminated, and eventually reversed. Causes that should be examined on public lands include soil disturbance processes such as logging, grazing, roading and mining.

• Herbicides should be considered only a last resort after all other viable alternatives have been considered.
• Planning documents will present a range of non-chemical alternatives.
• Planning documents will provide an analysis of the long-term effectiveness and environmental costs of all alternatives.

In order for decision-makers to allocate limited funds where they are most critically needed, it is imperative that land managers provide comprehensive effects documentation for all projects which affect invasive species, not just weed and vegetation management projects. Almost all projects can be expected to have an effect on the extent of invasive species including road construction and reconstruction, logging, livestock grazing, “forest health” restoration projects, and recreation (particularly motorized recreation).

Congress has provided the means to protect resources from declining through the National Forest Management Act (NFMA), which mandates the Forest Service to prepare and periodically review a comprehensive Renewable Resource Assessment, which,

. . . must consider, among other things, the important . . . laws, regulations, and other factors expected to influence and affect significantly the use . . . and management of forest, range, and other associated lands [16 USC § 1601(4)]. . . . As part of the Assessment, an inventory of all National Forests and renewable resources must be maintained and kept current [16 USC § 1603]. . . . Public participation in the development of the Assessment is also required 16 USC § 1600 (3)].

The Council on Environmental Quality (CEQ) has promulgated regulations in 40 CFR § 1500.1(b) specifying that, “environmental information” relevant to federal actions must be “available to public officials and citizens before decisions are made and before actions are taken.” These include “new and continuing activities, including projects and programs entirely or partly financed, assisted, conducted, regulated or approved by federal agencies” [40 CFR § 1508.18(a)].

But in all too many cases the Forest Service has attempted to satisfy these regulations through the use of documents such as the Integrated Weed Management Environmental Assessments of the Okanogan (1997, 1999) and Colville NFs (1998). Unfortunately, these documents have been self-serving, top-down treatment programs. Their methods have been primarily symptomatic in nature, they have failed to address individual projects that are causing weeds to spread and they have been based on inaccurate inventory information.

The Okanogan NF (1997, 1999, 2000) continues to use the same outdated inventory information collected by personnel driving along Forest roads and nowhere else. Rather than improving the quality of the inventory, the Forest Service continues to spend its funds on chemical treatment of roadsides using the “shotgun approach”. This approach treats small areas of weeds alongside roads throughout the Forest while ignoring infestations away from roads, which later serve to re-infest treated areas.

Instead of acknowledging such shortfalls, the agency continues to issue direction as if the problems are being attended to (Forest Service, 1999):

The environmental analysis for any project with the potential to introduce or spread noxious weeds must consider and analyze weed prevention strategies. . . . Issue clear direction with budget advice from Region to National Forests stating program areas responsible for ground disturbing activities are to build prevention costs into the project on the front end.

Beyond the mere fact that roads provide an extreme amount of soil disturbance in which weeds flourish, they also provide access to relatively undisturbed areas, acting as a conduit for weed spread into more remote areas. Studies show that exotic weeds don’t remain along roads. When road surveys for noxious weeds are extended to 50 meters, the invasive species are still present 70-100% of the time (Gelbard, unpublished). In Montana, it was found that exotic weeds spread outward from roadsides by invading adjacent relatively undisturbed communities (Forcella and Harvey, 1983). Within the interior Columbia Basin, cheatgrass (Bromus tectorum) originating on roadways has successfully invaded shrublands, ponderosa pine forests, and pinyon-juniper woodlands (Monsen, 1994).

The Forest Service has displayed a lack of regard for any acknowledgment that their primary activities of logging and associated roading could be responsible for weed invasions. It is a widely held view that logging provides the type of open, disturbed habitats favorable for weed establishment and research confirms this to be true. Both spotted and diffuse knapweed prefer open, disturbed habitats including roads and over-shaded areas (Watson and Renney, 1974). Yet Environmental Assessments written for timber sales rarely include more than cursory analyses of the effects of these projects on invasive species. Instead of addressing the issue squarely where it begins, logging projects often marginalize invasive species threats as “public concerns” that never see the light of day, and to date, timber budgets have not shouldered their fair share of the management of invasive species.

The Interior Columbia Basin Ecosystem Management Project objective to prevent invasions by reducing disturbance activities would be a good idea if only it applied to the majority of cases where logging occurs—in timber sales (ICBEMP, 2000, p. 30):

Because weeds are not adapted well to shade, consider retaining shade along roads by minimizing removal of trees and other roadside vegetation during construction, reconstruction, and maintenance, particularly on south aspects.

While the ICBEMP did not propose any substantial changes that would alter the current direction of management from its increasing reliance on chemicals, they did acknowledge that prevention efforts should be prioritized and they implied that a relationship between livestock and plant invasions should be considered by range conservation managers (ICBEMP, 2000):

If livestock grazing management is a factor in causing an area to function ‘at risk’, then that area shall be a high priority to initiate changes to livestock grazing management [p. 27]. . . . Consider using grazing management practices where feasible . . . to reduce the spread of targeted undesirable plants [p. 31].

Case example: Great Basin restoration

In the Great Basin, native sagebrush rangelands have been largely replaced by annual cheatgrass (Bromus tectorum), which burns hotter and more frequently (Mack, 1986). The Great Basin Bird Observatory breeding-bird atlas indicates that less than half the number of bird species are able to breed in the altered cheatgrass ecosystem (Christensen, 2000). Livestock forage is also reduced and desperate managers have resorted to “restoration” with alien species such as kochia (Kochia scoparia, a noxious weed in some states) and crested wheatgrass (Agropyron cristatum) to try and outcompete the cheatgrass. In the Great Basin, the BLM spends $70 an acre to counter weeds that cause devastating wildfires, but also pays an additional $71 an acre to put out fires, and another $64 per acre for rehabilitation following fires (Wolfson, 2000). In 2000, the BLM requested $17 million to begin the Great Basin Restoration Initiative on 500,000 acres of rangeland.

In denying the initiative, Congress was willing to write off 500,000 acres of Great Basin rangeland and save the taxpayers $40 million on an investment with unknown return potential. Given the available data, the inevitable conclusion is that the current agricultural system in the Great Basin is unsustainable in the long run. Addressing global desertification, Verstraete and Schwartz (1991) offer a glimmer of hope:

. . . it is not difficult to see that this situation cannot be sustained for much longer. The time is for action, not panic, and the scientific community has a definite role to play.

Due to the haste with which wildfires are often fought and “restored”, written guidelines for wildfire fighting and restoration that could have prevented this are often unavailable or ignored. Revision of the guidelines came about several years later in a Wallowa-Whitman Forest Plan Amendment (Richmond, 1992), which unfortunately was too late to correct the infestation.

A year 2000 FOIA request to the Wallowa-Whitman NF from George Wooten for “plans and management actions related to reseeding efforts for restoration and recovery following the 1988 Tee Pee fire” was nonresponsive, as the records have destroyed (Forsgren, 2000). It is unfortunate that the Forest Service lost these records, as they could be invaluable to managers developing invasive species programs.

• All planning documents must address the impacts of invasive species and proposed control measures, whenever soil disturbances are planned or are a result of planned activities.
• Prevention strategies must be built into all projects through their inception, planning and implementation.
• Planning for projects which involve soil disturbances should involve specialists trained in invasive species biology.

In order to protect public interests in guarding against alien species invasions, the legal right to citizen appeals of government proposals must be preserved. There is little cause for assurances that decisions will be given in good faith, given the poor record to date given by the BLM and Forest Service for protection from invasive species.

Recent proposed changes to the National Forest Management Act (Forest Service, 1999b) preserve these rights:

Under the proposed regulation any person would be allowed to object to a pending decision. The objection would be filed, in writing, within 30 days of public notice of the appropriate NEPA documentation. Unlike the current 217-regulation, the proposed objection process does not have a specific time limit for resolving objections. Instead, the responsible official would not be allowed to approve a plan amendment or revision under objection until a decision on the objection has been reached and documented in an appropriate decision document for the land and resource management plan. The proposed rule does not change the 36 CFR 215 appeal process for site- specific project decisions.

In addition to these provisions for public review and appeal, the Record of Decision (ROD) for Managing Competing and Unwanted Vegetation (Forest Service, 1988) responded to a 1984 court injunction that prohibited all herbicide spraying until legal insufficiencies were satisfied. As a result of court-ordered mediation, additional requirements were added to the ROD in a Mediated Agreement (1989). Signers of the Mediated Agreement were granted the right to petition the court for relief if the Region failed to comply with its requirements.

Even with these safeguards, invasive species continue to make inroads onto public lands. The specter looming before us is that the agencies who have sat idly for so long will now react to Congressional prodding with hastily prepared, costly approaches which depend primarily on widespread use of broad-spectrum herbicides, with concomitant contamination of the environment, compromised human health and preordained failure to control invasive species.

In site-specific situations, once invasive species establish themselves as integral, albeit unwanted, parts of the natural environment, further control efforts may be wasteful spending, and in addition may require destruction of the underlying ecosystem. If agencies are not held accountable to protecting ecosystem resources, they may have little restraint in spending funds on futile eradication efforts.

There are some provisions to limit wasteful government spending when weed control becomes a lost cause. For instance, when it becomes apparent that control of a species is no longer possible, state noxious weed Control Boards can remove weeds from lists requiring mandatory controls, in essence, admitting defeat, but at the same time saving limited funds for use in controlling other species. For instance, in Washington state, bull thistle (Cirsium vulgare), and Canada thistle (Cirsium arvense) have become so ubiquitous that they have been relegated Class C status, meaning they no longer require mandatory control under the state noxious weed control laws. Regardless of their “status”, Canada thistle used to be listed as the worst weed in the United State and still has the potential to aid in the loss of western rangelands (Colorado State University, 1999).

The determination to reclassify these weeds was made purely on economic grounds because control was perceived to be beyond the range of present budgets. However, such weeds should still be given consideration whenever ground disturbing activities are proposed in areas where known populations exist or are adjacent.

For some invasive species, a point may be reached where the costs of control outweigh the benefits of removal. This point occurred long ago for species like Kentucky bluegrass (Poa pratensis), Dutch clover (Trifolium repens) and common bentgrass (Agrostis stolonifera), which were originally valued additions to the pastures of early homesteaders. They were chosen for their value as forage components and their ability to dominate and thrive under a regime of heavy disturbance. These same traits caused these species to spread aggressively into native ecosystems (Almack et al., 1993) where they have choked out native species.

• Decisions must remain subject to appeal.
• Decision documents must be subject to public review and comment.
• Decisions must be subject to peer review and legislative oversight.
• Managers need to provide a documented review process for issues that involve controversy.