Simply put, monitoring can be defined as education. Such education may reinforce what one already knows or provide new information, which encourages a change. In the context of natural resource management, high-quality monitoring programs can provide early-warning signs of unsustainable practices (Franklin et al., 1999).

Monitoring is imperative in order to determine if public land management decisions are being implemented as intended, if such decisions are effective in achieving goals and objectives, what types of impacts and effects such decisions may have, and to verify that the various assumptions that were made during planning were and are still valid. Monitoring analysis results are best used to evaluate actions taken and then determine whether activities should be changed, restricted, or stopped altogether.

Monitoring can make major contributions to ecological research programs if it produces research-quality data, and can also highlight important phenomena or spatial and temporal patterns that need scientific attention (Franklin et al., 1999).

To date, there is a lack of emphasis on monitoring public land management decisions that affect invasive species and a lack of emphasis on monitoring activities taken to “treat” invasive species. In the case of projects that involve commercial thinning or road reconstruction, for example, there is inadequate acknowledgment and disclosure of how such projects affect invasive species’ spread.

Actions taken to “treat” invasive species, particularly those involving the use of chemicals, often result in a number of other indirect effects, such as soil and water quality, which are also rarely monitored.

Indeed, monitoring rarely occurs for projects that “treat” weeds, usually due to a lack of funding. Even when decisions specify monitoring as a required measure for likely impacts, treatment activities are likely to proceed to completion before monitoring is begun, at which point funds may well be exhausted, and monitoring requirements rendered moot. The end result is that there is no accountability to the public regarding whether Forest Service activities (particularly herbicide treatments) are spreading weeds and degrading public lands or whether they are even effective.

• Project monitoring procedures should be funded separately from other project actions.
• Baseline monitoring must be performed prior to project implementation.
• Monitoring should include these critical components: (1) measurement of the extent of invasive species populations; (2) measurement of the effectiveness of treatments; and (3) assessment of the extent of non-target impacts resulting from treatments.

Quantitative information on the effects of forest management practices is ultimately essential to assess the long-term sustainability of a particular practice (Franklin et al., 1999).

With respect to activities involving the treatment of noxious weeds, the Mediated Agreement (1989) requires the Forest Service to monitor not only site-specific post-treatment conditions, but also to monitor the impacts to human health from using herbicides. Site-specific post-treatment information is to be used to aid in future project planning. The Agreement also included programmatic objectives to design a noxious weed monitoring plan for land management activities.

With respect to activities on public lands that do not involve the treatment of weeds, but may result in the spread of invasive species, it is important to recognize that the National Forest System is legally bound to monitor the effects of its activities. The National Forest System Land and Resource Management Planning Authority states in Sec. 219.7:

(f) A program of monitoring and evaluation shall be conducted that includes consideration of the effects of National Forest management on land, resources, and communities adjacent to or near the National Forest being planned and the effects upon National Forest management of activities on nearby lands managed by other Federal or other government agencies or under the jurisdiction of local governments.

Proposed NFMA regulation (Forest Service, 1999b) call for increased monitoring on public lands. In a statement of principles regarding the new regulations, the Committee of Scientists (1999) identified the four types of monitoring, and the circumstances for their application:

Four types of monitoring can be considered:

1) Implementation monitoring asks the question, have the management standards and guidelines been used as anticipated to guide strategic and operational decisions?

2) Effectiveness monitoring asks, are the standards and guidelines producing the desired future conditions as anticipated at both the large-landscape and small-landscape planning levels?

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

4) Anticipatory monitoring asks, what factors (human induced or natural stressors) could compromise the attainment of sustainability in the near and long terms?

In all these instances, three key factors contributed to success. . . . particular aspects of the biology of the target species . . . sufficient resources . . . widespread support both from the relevant agencies and the public.

Case Example: Colville NF lack of monitoring

The Colville NF response was, “Besides the water quality monitoring reports previously described, there are no other reports available. Field monitoring is occurring, however, no written reports of effectiveness or other environmental effects reports besides water quality monitoring have been done”.

It is important to recognize that a total of approximately 122 herbicide application records were supplied in response to the FOIA request for 1999 alone. Despite such a high number of treatments, the complete record for the Forest response to the FOIA returned only 1 water quality monitoring report and 4 vegetation monitoring transects, leaving the remaining treated areas without monitoring.

Every year, the Colville NF is obligated to report the results of their annual monitoring in order to assess whether their activities have complied with the Colville Forest Plan. These Reports are titled “Annual Monitoring and Evaluation Reports”. As of the last Report published, which was for fiscal year 1997, there was no reporting of annual monitoring of noxious weeds on the Forest since the Forest Plan was signed in 1989. In fact, there is not even a monitoring item specific to noxious weeds in the Report, which would allow the Forest to gauge how their project activities are affecting the spread of invasive species on public lands. No one has any idea of how effective noxious weed treatment projects have been or whether funds were well spend.

One could legitimately question whether the Colville NF is in compliance with existing regulations, however, their performance is not exceptional. In response to a FOIA request from Kettle Range Conservation Group for monitoring results on the Methow Valley Ranger District, the response was that no monitoring occurred at all on this District, which is one of the largest and most heavily infested in region.

• Invasive species treatments should monitor site-specific, post-treatment conditions for all affected resources.
• Monitoring should be used to determine the effectiveness of treatments and whether project implementation was performed as planned.
• Monitoring should be included in all projects with invasive species impacts, not just "weed" management projects. For example, monitoring of impacts to invasive species' spread should be occurring in projects involving road maintenance, fire fighting, livestock management, and timber sales.
• Monitoring invasive species needs to be periodically repeated on public lands.
• Monitoring must follow a consistent protocol, with written records maintained in a permanent archive.

Monitoring reports for vegetation management projects must be available to policy-makers, program managers and the public.

Monitoring of the effects of past management decisions is critical for managers to assess the need and direction of future programs. Verstraete and Schwartz (1991) identified the critical roles of monitoring:

Monitoring the environment . . . plays a number of crucial roles and must be pursued to:

1) establish a baseline against which future observations can be compared;

2) document the spatial and temporal variability of the relevant environmental parameters;

3) identify the regions at risk of further degradation, and the nature of the processes at work;

4) provide the data needed to build and validate the mathematical models of the environment that are needed to understand and predict the evolution of these ecosystems;

5) support policy decision making in such tasks as prioritizing the target areas for relief and conducting cost-benefit analyses of various remedial actions or feasibility studies, as well as support field activities geared towards minimizing further degradation or reclaiming affected areas; and

6) evaluate the effectiveness of these policies, plans and remedial actions.

At intervals established in the plan, implementation shall be evaluated on a sample basis to determine how well objectives have been met and how closely management standards and guidelines have been applied. Based upon this evaluation, the interdisciplinary team shall recommend to the Forest Supervisor such changes in management direction, revisions, or amendments to the forest plan as are deemed necessary. Monitoring requirements identified in the forest plan shall provide for:

(1) A quantitative estimate of performance comparing outputs and services with those projected by the forest plan;

(2) Documentation of the measured prescriptions and effects, including significant changes in productivity of the land; and

(3) Documentation of costs associated with carrying out the planned management prescriptions as compared with costs estimated in the forest plan.

(4) A description of the following monitoring activities:

i. The actions, effects, or resources to be measured, and the frequency of measurements; ii. Expected precision and reliability of the monitoring process; and iii. The time when evaluation will be reported.

(5) A determination of compliance with the following standards:

i. Lands are adequately restocked as specified in the forest plan; ii. Lands identified as not suited for timber production are examined at least every 10 years to determine if they have become suited; and that, if determined suited, such lands are returned to timber production; iii. Maximum size limits for harvest areas are evaluated to determine whether such size limits should be continued; and iv. Destructive insects and disease organisms do not increase to potentially damaging levels following management activities.

Within the Bureau of Land Management, it is acknowledged that comprehensive monitoring programs are necessary to evaluate management activities, control noxious weeds, and demonstrate BLM compliance with applicable laws, regulations, and policies (BLM Weed Team, Asher et al. (http://www.blm.gov/nhp/main/WP7weedplan.html). The BLM report discusses the need for comprehensive monitoring programs:

Monitoring and research are essential to provide information necessary for long-term planning and decision-making. For example, monitoring and research will help determine if: 1) BLM is achieving the management objectives established in land use and activity plans, 2) certain projects or management actions are having the desired effect, 3) species-specific control methods are effective, and 4) BLM should change its management. Monitoring and research also allows BLM to base its noxious weed management program on sound ecological knowledge of noxious weeds and their relationships to management actions.

Monitoring information should be collected on treatment sites to determine effectiveness, the effects on nontarget species, and subsequent species that invade the treated site. Established infestation sites not currently being treated should be monitored for growth rates, rates of spread, population structure, and the environmental conditions that support the noxious weed invasion.

An important part of conducting a monitoring program and, specifically managing data, concerns making data available on a timely and comprehensive basis to a wide range of interested parties (Franklin et al., 1999). Unfortunately, insuring high-quality data management is a frequently unrecognized part of a monitoring program (ibid).

• Monitoring information results should be periodically evaluated and the evaluation summaries transmitted to regional and national offices.
• A summarized account of monitoring results should be readily available for inspection at all supervisory offices and higher. Information gathered during the procedure is strategically important for future accounting needs and should be stored in a safe place, and kept on hand for many years, both on the districts as well as in regional offices, in both raw and summarized formats. The type of information contained on monitoring accounts should include:

1) Date, site description and cross-reference number for mapping purposes
2) Applicator name
3) Application method
4) Time of application
5) Chemical used
6) Additives and carriers used
7) Mix concentration
8) Extent of area treated (as general descriptions and specific locations)
9) Rate of active chemical application (quantity per area)
10) Rate of mix formulation (quantity per area)
11) Total chemical amount applied (quantity per application)
12) Weeds present before and after treatment
13) Field notes
15) Efficacy of treatment
16) Experienced costs of the project (direct and indirect)
17) Residue analysis as appropriate (e.g., water quality, soil quality monitoring)
18) Analysis of unintended effects as appropriate (e.g., non-target vegetation effects)
19) Accidents, spills, drift encountered (reported and experienced)
20) Human health effects as appropriate (e.g., hazards and symptoms experienced)
21) Worker complaints

Monitoring procedures should be carefully designed to provide useful information about project outcomes.

The outcome of monitoring is important enough that standardized procedures should be followed by all National Forest units when monitoring is performed. Monitoring is often complex and costly, however a wealth of source material is available to aid in design.

Franklin et al. (1999) assert that the development, operation, and interpretation of credible natural resource monitoring programs can only be achieved with extensive scientific involvement. The authors contend that results of scientific research and scientific expertise are needed in at least four major aspects of monitoring: (1) Design of monitoring programs, including the selection of parameters and development of the sampling design – where, when, and how to sample as well as details of the statistical design; (2) quality control; (3) interpretation of results; and (4) periodic assessments of the effectiveness of the monitoring program (“adaptive management”).

The type of monitoring required will to some extent dictate the range of available monitoring procedures. The need to design monitoring carefully is stated in the National Forest System Land and Resource Management Planning Authority. Sections 6 and 15, 90 Stat. 2949, 2952, 2958 (16 USC 1604, 1613); and 5 USC 301 (47 FR 43037, Sept. 30, 1982, Sec. 219.19 (6), Fish and wildlife resource):

Population trends of the management indicator species will be monitored and relationships to habitat changes determined. This monitoring will be done in cooperation with State fish and wildlife agencies, to the extent practicable.

Monitoring forms had not been designed to note the following common ingredients of failure: inappropriate design, including insufficient hydrology as a result of design or construction oversights; slopes steeper than 20%, and plants specified for inappropriate habitat; compacted soil without organics; and lack of maintenance.

If a project’s scope, and hence monitoring needs, are extensive, then a rigorous, standardized approach to monitoring should be required as a prior condition of project approval and continued funding.

Case example: Boulder Creek on the Okanogan NF

In choosing Alternative C to use herbicides to treat over 5,000 acres of weeds with herbicides, the Okanogan NF (1997, p. 109) explained that with this alternative,

Water quality and the sediment regime would not be affected by noxious weeds.

This statement naively implies that the treatment will be 100% effective and have no impacts to water quality at all. A monitoring program should have been used to show the actual effects on water quality and sediment regime, before making the claim, however the project was implemented without even performing baseline monitoring.

In an independent review of the project, photographs of areas treated in the project found that herbicides killed vegetation along riparian areas, resulting in decreased rain-intercept ability and likely increased erosion (Wooten, 1999d; Photo 4, p. 3).
 
 

• Require a rigorous, standardized approach to monitoring as a prior condition of project approval and continued funding
• For all projects with invasive species impacts, allocate a percentage of implementation funds toward monitoring.

Scoping for the National Invasive Species Council Research, Information Sharing, Documentation and Monitoring Working Group (2000) stated a need to, “Identify research and monitoring that address real needs, fill key information gaps, and address limitations.”

Franklin et al. (1999) recommend that a monitoring program should include a definition of objectives, selection of the critical response variables, and design of a sampling scheme which will fulfill stated objectives. The following steps in developing a high-quality monitoring program are given:

A) Initially, it is important to identify which parameters are likely to be sensitive indicators of important ecological conditions, e.g., which are ecologically meaningful.

B) Once parameters have been selected, the next challenge is development of a sampling design – formalizing the answers to where, when, and how in a statistically robust design. Temporal issues involve decisions about what sampling intervals will be used. For example, some monitoring is appropriate continuously, others components are sampled most efficiently at regular intervals or on an event basis. The design of the monitoring program needs to reflect the multiple temporal scales.

Our region has already seen dramatic fish kills due to pesticides. In 1996, a weed killer used in an irrigation canal made its way into Bear Creek of the Rogue River Basin, killing off thousands of steelhead trout and scores of coho salmon. When salmon are so near extinction, the loss of each one makes recovery efforts that much harder.

• Determine what important questions need to be answered during projects with invasive species impacts on public lands.
• Include monitoring procedures in planning documents that will ask and answer important questions about proposed actions, both prior to and after project implementation.

Interpreting the ecological significance of a change in a monitored parameter is not a trivial issue and may in fact be the most challenging element of an operational monitoring program (Franklin et al., 1999). Substantial knowledge is often required to interpret the significance of observed changes in parameters.

• Insure that personnel involved in a monitoring program have necessary technical qualifications, or receive training for such procedures.
• Monitoring reports must include the names of all participants involved and their qualifications for accomplishing the job.

To be statistically valid, designs for treatment monitoring should attempt to either randomize the samples or assure that samples are representative of the entire treated area. Monitoring of treatments should include both treated and untreated samples, as experimental and control groups, respectively (Franklin et al., 1999). Samples should include enough observations to insure results are significant.

Monitoring procedures should use reproducible methods of measurement and replicate sampling to insure statistical validity. Consistent recording procedures and timing of visits should be part of the procedures. Monitoring should use quantitative measurements such as frequency, cover, density, etc., and plant identification should be carried to the level of the species.

Analyses should be performed that summarize the observations and should include a description of the implicit assumptions in the methods along with calculations of central tendency, data dispersion, and significance of results. Evaluations should be subject to independent professional reviews. Evaluations of monitoring results should be used as a series of approximations which will be modified periodically to adjust initial parameters toward fulfillment of objectives, or as new parameters are identified, or monitoring objectives change (Franklin et al., 1999). All stakeholders need to be a part of this process (ibid.).

When obstacles arise in the performance of monitoring procedures, alternate methods may be found in the literature. The use of indicator species as representative indicators of a wide array of environmental effects may be appropriate or necessary, as for example, surveys for spotted owls as indicators of prime old growth habitat. The use of bacterial or macroinvertebrate biosensors capable of detecting low concentrations of chemicals has been demonstrated to have wide applicability.

Hoof et al. (1992) were able to use the cyanobacterium Synechoccus to detect the herbicide atrazine rapidly at 50 micrograms/L concentrations, while Servizi (1987) used the microorganism Daphnia to demonstrate toxicity to the herbicide glyphosate.

Case example: Boulder Creek on the Okanogan NF in Washington

Spotted knapweed in the area was originally found on Forest Service land near the mouth of Boulder Creek. One of the first herbicide treatment projects undertaken in the Pacific Northwest, following lifting of the Region 6 injunction against their use in Region 6, was at Eightmile Ranch, when the amount of spotted knapweed was still limited. In the approximately half dozen years since repeatedly treating Eightmile Ranch with herbicides, the infestation of spotted knapweed has now spread many miles in all directions onto adjacent private lands, and further into the National Forest, e.g., up Boulder Creek. Control of spotted knapweed by the Forest Service has been a failure.

Following the use of herbicides at Boulder Creek, the seeds in the soil will remain fertile and will require additional treatments again soon. It is unfortunate that an adequate monitoring program for spotted knapweed in this area does not exist, because it would have helped inform managers of the effectiveness (or lack thereof) of treatments.

• Monitoring procedures should be performed for invasive species programs that incorporate valid statistical and sampling methods.
• Monitoring results should be evaluated for calculations of central tendency, data dispersion, and significance of results.
• Monitoring procedures should use reproducible methods of measurement and replicate sampling to insure statistical validity. Consistent recording procedures and timing of visits should be part of the procedures.
• Monitoring should use quantitative measurements such as frequency, cover and density, and plant identification should be carried to the level of the species.
• Use of a map-based system such as a Geographic Information System (GIS) should be used to map invasive species locations.

An accurate inventory and map of existing populations of invasive species is critical to invasive species management. Land managers must inventory sites regularly in order to identify any small or outlier populations that could easily be eradicated (Moody and Mack, 1988). It would be foolish to try and control invasive species in a small area surrounded by uncontrolled weeds, yet this is presently what is happening on most National Forests.

National Forests should base their management on the measurement of damage, action and safety thresholds. A related part of monitoring includes recording incident reports involving harm to human safety. Baseline surveys used to determine the extent of weed populations are a requirement of adequate monitoring. The Okanogan NF has based a projected $1 million program primarily on outdated surveys made by personnel driving along Forest roads (George Wooten, personal observation while working as a botanist on the Okanogan NF). Botany surveys for TES species, which could be used to provide valuable location information for noxious weeds, do not include standardized procedures for assessing the extent of noxious weed populations, and these surveys are seldom used in invasive species projects anyway (ibid.)

Environmental Assessments can distort the scale of an infestation if not performed for a representative area. Quantitative measurements of invasive species require more than just road miles, but should also include estimates of the species, density, number of individuals, area, and control costs.

Many weeds can find a niche in secluded areas, where their presence is hidden from surveyors. For example, riparian habitats with annual flooding disturbance regimes are continual sources of reinfestation. Yet the practicality of surveying or treating these widely scattered satellite populations is exceedingly low. If the actual extent of noxious species was truly known to decision-makers, the futility of proceeding in the face of large-scale invasions without better planning and baseline monitoring might be more apparent.

• Monitoring should include adequate baseline data and experimental controls.
• Monitoring of standards and guidelines for damage, action and safety thresholds should be performed prior to and during project implementation.
• Monitoring should include investigating and filing incident reports whenever human harm was reported or suspected.
• Monitoring should be performed over representative areas of a National Forest, using surveys that inventory more than roadsides.
• Quantitative measurements of invasive species should incorporate meaningful measures, including species identification, density, number of individuals, extent occupied, distance from roads, and potential control costs.
• Inventories should be performed periodically to confirm changes in the extent of target populations.

Documentation must be provided for the accomplishment of all mitigation measures, conditional procedures and stipulated agreements. The satisfaction of such legally binding accomplishments should be a condition of project continuance.

The Mediated Agreement (1989), stipulates that the Forest Service interpret noxious weed survey data for use in determining environmental impacts, and then follow the established uniform method set forth in the vegetation management FEIS (Forest Service, 1988, IV, 77-78) to ask, “What changes can be made in the project design to mitigate potential and current weed problems?” Because some Forests are doing a poor job of monitoring, the Forest Service is in violation of the Mediated Agreement (1989), which stipulates that the Forest Service must interpret noxious weed survey data for use in determining environmental impacts.

• Legally required mitigation measures, conditional procedures and stipulated agreements must be satisfied as a condition of continued project funding.
• Implementation monitoring should be incorporated into all mitigation measures, conditional procedures and stipulated agreements.