Boulder Creek on the Okanogan National Forest

Case study of a weed management action gone wrong

This web page follows a series of letters to various government agencies as a case study of a noxious weed management action--Boulder Creek on the Okanogan National Forest--and what went wrong with plans. Preferred and recommended actions available to managers were ignored, in preference for more harmful actions, which impacted treated areas beyond acceptable limits.

[Letter of Aug 25, 1999, to agencies, requesting a review of serious problems caused by improper actions.]
August 25, 1999
George Wooten, Jr.
23 Aspen Lane
Winthrop, WA 98862

To: All interested parties
Re: Herbicide spraying on Okanogan National Forest, 1999.
Sirs / Mmes:

I would like to request a site visit with representatives from the National Marine Fisheries Service (NMFS) and US Forest Service (USFS) to investigate the way in which riparian herbicide spraying is affecting riparian habitat along Boulder Creek and other Forest Service Roads. The 1999 spraying of Boulder Creek has resulted in apparent riparian tree mortality, foliage removal and unsightly browning along Boulder Creek road and in the Creek itself. I believe that the Okanogan National Forest (ONF) has not adequately analyzed the effects caused by this chemical application, including: (1) loss of habitat for T & E fish species, including Upper Columbia River steelhead and bull trout; (2) direct effect to T & E species; (3) indirect effects to T & E species, including grizzly bear and gray wolf; (4) direct and indirect effects to fish (including proposed westslope cutthroat), wildlife, riparian vegetation, and sensitive plants. The Methow Valley Ranger District (MVRD) is required to analyze these effects under the Endangered Species Act (ESA) and the National Environmental Policy Act (NEPA). This document refers to the following decision documents: Biological Evaluations and Noxious Weeds Date: July 12 1999, (BE) by Jennifer Molesworth, and the 1997 ONF Integrated Weed Management Environmental Assessment (EA) and Decision Notice (DN). The BE has not corrected the earlier inadequate analysis presented in the out- of date and inaccurate EA, as detailed below.

My experience in this matter stems from visits I made to Boulder Creek in July 1999, and from my training as a biochemist, botanist and wetland scientist. During my visits I noticed that both sides of the roadsides had been sprayed in a continuous stripe for miles up Boulder Creek (Treatment site #27, Bromas Creek / First Butte), going through several riparian areas that come within 50 feet of the road edge. The stripe consisted of a completely killed zone of vegetation, primarily native plants and including shrubs and trees.

My site visit indicates that treated roadsides have been denuded of their prior vegetative cover, which was primarily native species, thus increasing the amount of sediment delivered into the riparian system. This is contrary to the claims made in the EA, justifying the need for the spraying program. Rather than improving the habitat for fish and wildlife, the increased sedimentation impairs the quality of the riparian system, increases the amount of fine sediments into the channel and contributes to the embeddedness that plagues downstream fish redds. If siltation and embedding of redds is detrimental to listed fish species, then quite possibly the no effect determination of the BE is in error, and a take or harm may occur.

The herbicide treatment on Boulder Creek has caused the removal of beneficial native shrubs and trees along the road and within the adjacent riparian zone. This will result in a lowered recruitment of large woody debris (LWD) into the stream channel. Forest Service biologists are critically aware of the necessity of adequate stream side buffers that provide shade, cover and structural habitat for fish and maintain the continuous input of allochtonous material necessary to keep streams productive. Forest Service regulations (36 CFR §219.27 (e)) require that special attention be given to land and vegetation approximately 100 feet from the edge of all perennial streams, lakes, and other bodies of water. In a conversation I had on on August 23 with Christina Bauman, noxious weed specialist on the MVRD, I mentioned that the road appeared to have been continuously striped with herbicide along both sides clear up to the First Butte Lookout (Bromus Creek) Road, killing all vegetation within the striped zone, even across riparian areas. She replied that, on the contrary, hand wicking and spot spraying were used in riparian areas, but that in any case, the incidental killing of non-target species was an intended consequence of the spraying program. Such a management goal is arguably suboptimal on merit alone, however the BE specified hand wicking and spot spraying would occur in riparian areas as a mitigation measure for surfactant, and presumably, herbicide, entry into the water, not loss of vegetation. The loss and degradation of riparian habitat that occurred as a result of the spraying went unacknowledged in the BE, leading once again to the conclusion that the no effect determination is in error, and a take may occur.

Due to its phytotoxicity, the Environmental Protection Agency (EPA) compiled a list of over 70 species jeopardized by the use of glyphosate (US EPA, 1986). The BE did not disclose whether the herbicide removal of vegetation would affect any listed species, or if a survey for listed species occurred along the riparian corridor. In prior visits to the area, however, I have collected specimens from unusual species of willows (Salicaceae) which do not appear on list of species known from the ONF, but which should have been identified in any surveys for threatened, endangered or sensitive (TES) plants. This indicates TES surveys were not done prior to the application as specified in the EA. During my visit to the area in 1999, I noticed several of these willows in the riparian zone had been killed or defoliated by broadcast herbicide treatment, again in apparent violation of specifications in the EA and BE. I would like an independent botanist from an agency outside the USFS to review this issue with MVRD personnel to determine if any listed plant or other species were harmed. As a broad spectrum herbicide, glyphosate has documented phytotoxicity to a wide array of organisms, including lichens (Brown, 1995), nitrogen-fixing bacteria (Tu, 1994, Carlisle et al, 1986, Moorman et al, 1992, Martensson, 1992), beneficial mycorrhizal fungi (Estok et al, 1989, Chakravarty and Chatarpaul, 1990, Chakravarty and Chatarpaul, 1990, Sidhu and Chakravarty, 1990, and Chakravarty and Sidhy, 1987). These species are all integral components of the ecosystem which has been affected by spraying on Boulder Creek. The Carlisle study found that the rate of glyphosate degradation correlates with the soil respiration rate, an estimate of microbial activity. Glyphosate has been found to inhibit growth (at 50ppm) of 59% of randomly selected soil bacterial, fungal, actinomycete, and yeast isolates; of nine herbicides tested, glyphosate was the second most toxic. The no effect determination of the BE is likely in error, and a take or harm may have already occurred.

The EA and BE state that "herbicide label specifications will be followed". The label on Monsanto glyphosate formulation Rodeo© requires that it be applied in combination with an approved surfactant such as Ortho X-77 Spreader© (alkylarylpolyoxyethylene [AAPOE, a substance related to POEA, or polyethoxylated tallow amine, described below], isopropanol, water, and fatty acids). If the label requirements were followed then the BE is incorrect in stating that surfactants, "will not be used within the riparian buffers". In my August 23 conversation with Christina Bauman, she informed me that the Monsanto glyphosate formulation Rodeo© was used within 100 feet of riparian areas, without a surfactant. Either the BE is incorrect about following label requirements, or the BE is incorrect about the use of surfactants within the riparian buffer. It obviously can't be both. Either way, this is a serious matter. The ambivalence of official decision documents indicate to me that the effects analysis is flawed and the DN should not have been approved without consultation by a qualified chemist or biochemist. The gravity of this violation is that the Okanogan NF and Wenatchee NF weed control programs denied well-informed public comments raised about these points and sprayed vast tracts of land this year, willfully ignorant about the environmental effects. Such actions could be construed as arbitrary and capricious if this were reviewed by the courts.

At the time of this writing, I haven't determined whether or not the surfactant was used with the Rodeo© application. I am asking that your agency meet with the Forest Service staff and myself about this project, get the facts straight, and consider if there has been a taking. If the label directions to use a surfactant were ignored by the applicator, then there would be less chemical entry into the riparian ecosystem, assuming this is a one-time application, and the weed problem would be solved. But without the surfactant, the effectiveness of Rodeo© is lessened and the application becomes a waste of money.

According to a September 7, 1998, memo to acting director of the Washington State Noxious Weed Board, Lisa Lance, from the Ad Hoc Coalitioin for Willapa Bay, Nahcotta, WA 98637, a similar story of the failure of agencies to knowledgably use glyphosate has played out in attempting to control Spartina alternifolia (cordgrass) in Willapa Bay in Pacific County. In order to appease environmental pressure, The Nature Conservancy, in concert with some zealous Washington Congressmen, authored an EIS to control Spartina through hand-wicking with glyphosate. When this treatment didn't work, the 6-inch thick EIS was essentially ingored, and the chemical rewicked, and when this didn't work the chemical was sprayed from helicopters, and when this didn't work the helicopters returned, ad infinitum. I worked for the Forest Service when the EIS for Spartina control was released, and I remember the elation of the staff upon hearing that a safe herbicide was finally discovered that could be used in riparian ecosystems. Perhaps I have some of the details wrong, after all, nobody wants to admit failure, but the damage is that the program has since become institutionalized by a Forest Service all too anxious for a quick-fix which doesn't exist.

Weeds are not directly causing fish mortality, but indirectly through habitat loss. In fact, weeds are a symptom of poor forest health brought on by years of roading, logging and livestock grazing on the ONF. Because of the rugged terrain, most MVRD roads are located along streams and rivers, where they have become festering sores in the fragile glacial soils. The ONF has been largely unsuccessful at establishing beneficial vegetation along roadsides, and relies primarily on forage grass plantings for road maintenance. Orchard grass, Kentucky bluegrass, timothy and intermediate and crested wheat are planted in riparian sites, replacing shrub-dominated communities with grasses amenable to management. These grasses then attract cattle to preferentially graze along the roadsides, converting them into linear pastures. The increased disturbance provides a seedbed for weed seeds travelling on and in cattle and cars, and the trampling scarifies and plants the seeds, even providing hoof-prints to hold water and aid the weeds during spring germination. The problem of sedimentation and erosion affecting T & E fish in Boulder Creek has not been improved by the MVRD's weed treatment program-it is worse. The damage wrought by management to riparian habitats is amply documented by scientists on the Wenatchee and Okanogan National Forests (Everett, 1994, Johnson, 1994, Wissmar, 1994). What is lacking is resolve on the part of the agencies to deal with the causes of the problem, and change the way cattle grazing occurs along streams and rivers.

The inescapable conclusion is that the cause of weed spread and the damage to riparian ecosystems are both being caused by uncontrolled cattle trampling and grazing. NMFS and the USFS need to reconcile conflicting goals and assign a fair percentage of damage to riparian habitat where it belongs-livestock grazing along forest roads. But while it is the duty of the USFS to examine and correct the cause of the weed problem, it is the duty of NMFS to attend to the problem of T & E species impacts, which is why I am writing this letter, asking for an investigation into the damage caused on the MVRD through the weed management program. It is unfair that NMFS expects one hundred percent assurance that fish are not harmed through delivery of water through the Skyline irrigation canal, when the USFS is allowing far more damage to occur through its weed management program, with only a nod and a wink to the cattlemen.

Returning to the subject of the spraying incident on Boulder Creek, for the moment let's assume that the applicator had the common sense to follow the label directions, since label violations could result in license forfeiture. In that case, the action to go ahead with the spraying plan appears to be an arbitrary and capricious decision, as the Forest Service did not avail itself of recent documented effects from the chemical threats posed to ecosystems by glyphosate, Rodeo©, or X-77 Spreader©. The EA listed as references for glyphosate toxicity were Folmar et al (1979) and Newton et al (1984), yet these appear not to have been read, for the BE to have garnered a no effect.

The first reference found that 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 ug/kg. Roundup is four times more toxic to rainbow trout fry and fingerlings than to larger fish. Also, significant increases in stream drift of midge larvae was observed after the 2.0 mg/L of Roundup treatment. The toxicity of Roundup to rainbow trout and bluegill increased 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 appeared 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).

In the 1984 report by Newton, glyphosate residues and metabolites were evaluated in forest brush field in Oregon coast range. Concentrations were higher in sediment than in water and persisted longer. Early stream-bottom samples reflected concentratrions found in the streamwater, but later samples showed that even the water concentrations occurring below the detection limit may contain enough glyphosate to contribute to adsorption by sediments. Of particular interest was the author's finding that residues (of glyphosate) in animals may remain detectable for several months. Concentrations in viscera were always higher than those in the remainder of the animal. This indicates that the BE did not even consider the one available reference they had that indicated measurable levels of herbicide can and do become detectable in fish:

"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."

Since the publication of Newton in 1984, reports of the toxicity of glyphosate formulations on fish and wildlife have become widespread in the scientific literature. The Forest Service was given ample information through scoping on the EA to avail themselves of this literature, yet they chose to ignore it through the implementation of the project, in violation of National Environmental Policy Act (NEPA) procedures, and Endangered Species Act (ESA) consultation procedures.

Glyphosate is acutely toxic to fish (Servizi et al, 1987). This report summarizes the acute lethality of Garlon 4, Roundup and a surfactant (MON 0818) contained in the formulation of the latter to sockeye salmon, rainbow trout and coho salmon. This study is important because it establishes the evidence that the combined effect of glyphosate and the surfactant POEA is more than additive (synergism). MON0818 (surfactant) was tested separately from glyphosate and found to be much more toxic than the latter. Further testing showed that the combined effect of glyphosate and MON0818 were "more than additive and raises doubt that the LC50's reported for Roundup in reconstituted water are applicable to natural waters.

Acute toxicities of Rodeo©, with X-77 Spreader© per label recommendations, vary from 120 to 290 ppm (Mitchell et al, 1987). Effects on pink and chum salmon are more serious than on coho and chinook (Wan et al, 1989).

Sublethal effects of glyphosate on fish include erratic swimming, labored breathing, altered feeding, migration and reproduction, and increased likelihood of being eaten (Morgan, J.D., et al, 1991, Liong et al, 1988).

Clearly the "no effect" determination of the BE occurred without substantial review. I would like to request NMFS review of this oversight so that the correct assignment of effects be given to the herbicide program, including incidental takings or harm to listed fish species.

The effects of glyphosate are not limited to fish and plants. Connor and McMillan (1990) compared moose forage resources on control and on herbicided cutovers near Thunder Bay, Ontario. Available moose browse, on control areas, was four times greater, and browse utilized was 32 times greater, than in treated areas after one growing season post-spray. Winter moose presence was almost two times greater on untreated than treated areas after one growing season and similar at two growing seasons post-spray. An important issue that needs to be considered is the effect of spraying sizable contiguous areas, substantially affecting food supplies within the home ranges of a number of moose, and the eventual effect on moose populations. The Chewuch watershed is just beginning to show a comeback in moose, and the effect of this ill-conceived herbicide spray program needs to be assessed. Other species besides moose, for example lynx and grizzly bear, use the Boulder Creek area. A possibly grizzly bear and cub sighting was made by Charlie Dewberry in 1998 on the sprayed section of Boulder Creek, however the BE did not analyse the effects of herbicide on listed mammals, and should not have been approved, for this omission.

Martinez and Brown (1991) found that the surfactant POEA (in doses of 1.03g/kg ) has serious pulmonary toxicity, but not quite as serious as the full formulation, Roundup which produced 100% death in rat subjects within 24 hours. Recent cases of poisoning in suicide attempts using a 41% glyphosate concentration have produced life-threatening symptoms (in humans) and a DEATH RATE of 10-20%, in spite of animal studies which show that the active ingredient, glyphosate, has a toxicity of approximately 5g/kg. This is important because gilled species are also more susceptible to surfactants (Rankin et al, 1982).

A number of studies show detrimental effects from glyphosate on birds (Cox, 1991, 1995a, 1995b). A study by MacKinnon and Freedman (1993) examined the effects of silvicultural glyphosate use on breeding birds near Stewiacke, Nova Scotia. Densities of most common breeding species decreased on all treatment plots, including the reference plot. The White-throated sparrow and the common yellow decreased significantly, as did also the alder flycatcher, especially on the spray plots. The BE did not analyse the effects of herbicide on birds, and should not have been approved.

Glyphosate causes a magnitude of untoward effects to the entire aquatic food chain. Austin et al (1991) found that glyphosate could potentially act as a phosphorus source and thus could stimulate undesirable eutrophication of waterways, especially those used by salmon or trout. The paucity of published research on the action of glyphosate on aquatic species composition, bioaccummulation, food chain relationships, further recommends caution in the application of this herbicide, which unfortunately has gone unheeded by the ONF and WNF. Buhl and Faerber (1989) found that Roundup treatment of a Maine clearcut caused an 89 percent decline in the numbers of the midge Chironomas riparius. This species is an important food resource in the food chain.

In studies commissioned by the Department of Environmental Protection in Southwest Australia, Bidwell and Gorrie (1995) showed that tadpoles (which respire with gills) were many more times sensitive to the full formulation than adult frogs, and were considerably more sensitive to the formulation (Roundup 360) than to technical grade glyphosate. Some surfactants affect aquatic organisms by damaging the gill membrane. LC50 values for adult frogs indicate there may be very little safety margin between concentrations in shallow water and concentrations lethal to frogs. Frogs are exposed to herbicides through runoff or overspray from treated areas adjacent to permanent or temporary wetlands.

Goldsborough and Brown (1988) assesses the effects of the Roundup formulation of glyphosate on short term carbon assimilation by periphytic (growing in fresh water) algal communities collected from six small forest ponds. The six study ponds were located in boreal forest near Lake Winnepeg in Manitoba. The short term photosynthetic rates of intact periphyton communities were significantly impacted by glyphosate additions. Glyphosate concentrations less than .89mg/L had no effect on short term algal photosynthesis, while the EC-50 value (glyphosate level resulting in 50% inhibition of carbon fixation) lies between 8.9 and 89mg/L.

Also there has been no description of the effect from the carriers so-called "inert ingredients" in the herbicide forumulations used, nor on the cumulative effects of combining the herbicide glyphosate with picloram. In a chapter from their book, Colborn and Clement (1992) examine the issue of increased exposure to pesticides (including herbicides and fungicides) and the difficulties encountered in obtaining any meaningful measurements of these exposures for women, children, even embryos in vivo. (Most standards of exposure are geared to the adult male.) Exposure (to active and so-called "inert" ingredients) through treated food and contaminated water are now combined with exposures through rainwater, snow, household dust, yard soil and indoor air. The timing of exposure, even to a small amount, can have a profound effect on an embryo. No studies showing cumulative exposure from all pathways could be found. There are many reproductive and endocrinological effects associated with herbicides and fungicides. The reassessment and reregistration of many older pesticides (using new techniques) mandated by law in 1988 had not been conducted in 1990. False claims about the safety of pesticides, combined with flaws in the federal registration process, raise serious concerns about increased exposure to environmental chemicals, when there is lack of information on their reproductive and endocrinological effects, synergy, bioaccumulation, and continual low-dose exposure.

Certain chemicals in the environment are estrogenic Arnold et al (1996). Low potencies of these compounds when studied singly may indicate little effect on biological systems. However, combinations of two weak estrogen-mimicking chemicals were 1000 (!) times as potent as any chemical alone. This synergistic interaction of chemical mixtures with the estrogen receptor has profound environmental implications for the ONF, which hasn't considered the synergistic or cumulative effects of their project on listed species.

The ONF does not have the means to carry out such complicated analyses in its EAs and BEs, let alone even read the recent literature. It is not reasonable to expect that the ONF or WNF produce a reasonable description of the effects of weed control on the National Forests. Both documents should be withdrawn and the real job of protecting salmon through control of mismanagement begun.

Unfortunately the detrimental effects on the aquatic food chain have already occurred. It is unfortunate that the job of documenting these effects should fall on citizens rather than paid agency personnel. I have neither the time nor financial resources to document the extent of harm to T & E species which has been caused by the MVRD on Boulder Creek, and throughout the ONF and WNF by this unfortunate circumstance. I would appreciate the assistance of NMFS in helping to resolve my concerns about fish, and USFS about wildlife.

It was dismaying to visit Boulder, Wolf and McFarland Creeks only to find out that the MVRD had sprayed them without public notice. After spending countless hours commenting on the scoping and EA documents, the MVRD apparently owes nothing to the public. The BE was not distributed, no warning signs were posted along the road side, and the promise to contact sensitive individuals (of which I am one) made in the EA was not kept. The herbicides sprayed in McFarland Creek caused me to have intense headaches all day long. The herbicide formulations used there appear to have been dispersed in a solvent carrier, which was highly irritating to work in. Part of my income comes from sales of natural scents and oils, however the sprays used by the MVRD in McFarland Creek impair my ability to sense smells. Obviously if I can smell these chemicals, then fish, with their highly developed sensory organs, can sense, and possibly avoid, them and the entire Methow River system. Avoidance of pesticide polluted rivers by anadromous fish has been documented (Ewing, 1999).

Almost every road in McFarland Creek was completely saturated with herbicide, from the ridge tops to the Methow River. In private talks with people in the area, it appears that the Okanogan National Forest is not acting under a Federal Program, but rather under pressure from the Okanogan County Noxious Weed Control Board. Almost to confirm this, at the bottom of the McFarland road, I photographed the Okanogan County spray truck spraying herbicide directly into the Methow River. Although the sprayed cover many miles, but the infestations are not being controlled, only hidden.

To address the native tree and shrub mortality caused by the herbicide spraying, the following quote from Karen Ripley, DNR Resource Protection Division, upon discovering the extent of ponderosa pine mortality caused by the Department of Transportation (DOT) roadside spray program in fall, 1998, is relevant:

"Such damage must be prevented in the future. Noxious weeds, road surface integrity, visibility, ditch maintenance and vegetation control are all very critical and consequential road management issues. As manager of DNR's Forest Health program, I am interested in solutions to these issues which also are compatible with healthy roadside forest trees.

If a roadside chemical is being applied which is damaging to the trees, a substitute chemical or practice must be located. There is an especially critical need to find suitable ways of managing roadsides that are adjacent to major waterways such as Peshastin Creek and the Okanogan and Methow Rivers.

Next Steps: It. is important to pursue the larger goal that such tree injuries not be repeated."

My hope is that NMFS will also take such impacts to their resource to heart when they contact the Forest Service about these issues and concerns.

Sincerely yours,
George Wooten

References