guidelines for measuring the physical, chemical, and biological
TRANSCRIPT
United States Department of
Agriculture
Forest Service
Rocky MountainForest and RangeExperiment Station
Fort Collins,Colorado 60526
General TechnicalReport RM-146
Guidelines for Measuring the Physical,Chemical, and Biological Condition of
Wilderness Ecosystems
Douglas G. Fox
J. Christopher BernaboBetsy Hood
..
This file was created by scanning the printed publication.Errors identified by the software have been corrected;
however, some errors may remain.
Fox, Douglas G.; Bernabo, J. Christopher; Hood, Betsy. 1987. Guidelines formeasuring the physical, chemical, and biological condition of wildernessecosystems. USDA Forest Service General Technical Report RM-146, 48 p.Rocky Mountain Forest and Range Experiment Station, Fort Collins, Colo.
Guidelines include a large number of specific measures to characterizethe existing condition of wilderness resources. Measures involve theatmospheric environment, water chemistry and biology, geology and soils, andflora. Where possible, measures are coordinated with existing long-termmonitoring programs. Application of the measures will allow more effectiveevaluation of proposed new air pollution sources.
Keywords: Monitoring, Wilderness, Baseline Conditions, Air Pollution,Atmospheric Deposition
The use of trade and company names is for thebenefit of the reader; such use does notconstitute an official endorsement or approval ofany service or product by the U.S. Department ofAgriculture to the exclusion of others that may besuitable.
USDA Forest ServiceGeneral Technical Report RM-146 November 1987
Guidelines for Measuring the Physical, Chemical,and Biological Condition of
Wilderness Ecosystems
Douglas G. Fox,Rocky Mountain Forest and Range Experiment Station1
andJ. Christopher Bernabo,
Betsy Hood,Science and Policy Associates, inc.
‘Research reported here was funded by the Rocky Mountain Forest and Range ExperimentStation under a contract with Science and Policy Associates, Inc. The Station’s headquartersIs In Fort Collins, In cooperation with Colorado State Universlty. Supervision was provided byDouglas G. Fox, Chief Meteorologist and Project Leader for The Research Work Unit, Effectsof Atmospheric Deposition on Natural Ecosystems In the Western United States.
Foreword
This report is the product of an effort topoll the scientific community about the mostappropriate techniques to be used to measure thecondition of wilderness ecosystems. Thesetechniques recognize the constraints imposed bythe statutory designation of wilderness. Theyare focused on monitoring needed to support theair resource management responsibilities of theForest Service and other managers of Class Iareas, as mandated by the Clean Air Act.
This report was prepared as part of acontract effort between the Rocky Mountain Forestand Range Experiment Station, and Science endPolicy Associates, Inc. of Washington, D.C. SPAcrafted a process that included a large group ofscientific talent (listed at the end of theGuidelines) organized to develop a concensusproduct with an ever-widening group of interestedparties. These Guidelines specifically resultfrom a formal public review of earlier drafts.Review comments and the responses to them areavailable from the Rocky Mountain Station.
Readers should keep in mind that wildernessmonitoring is complex and controversial.Improvements are likely to result only throughexperience with the application of theseguidelines in diverse locations over the breadthof ecosystems that populate the Wilderness systemin the US. Toward that end the Rocky MountainStation is continuing to develop and recordexperiences with the application of theseGuidelines. Three specific examples are worthmentioning:
1. The Wyoming State Office of the USDIBureau of Land Management is applying theGuidelines to selected wilderness study areas inthe western part of the State and evaluatingtheir utility. This work was initiated in 1987and will be ongoing for 5 years.
2. The Idaho National EngineeringLaboratory, a national laboratory under theDepartment of Energy, is conducting a 2 yeartechnical review and critique of the Guidelinemethods. INEL work is focused on the BridgerWilderness.
3. The Atmospheric Deposition Effectsresearch unit at the Rocky Mountain Station isconducting continued long term study ofwilderness ecosystems using both directstress/response and general biogeochemicalprocedures. A focus of this research is to
provide Federal land managers and regulators withtools to discharge appropriate and effectivemanagement of air resources as one of themultiple natural resources of wilderness.
Thus, we recognize that the guidance providedin this report will need periodic review. It islikely that versions of these Guidelines will beupdated every 5 years.
The Need for Guidelines
Guidelines for determining current conditionsof sensitive resources in Wilderness ecosystemshave several purposes. FLMs and regulators needimplementable measures to determine ifsignificant changes are occurring in Wildernessareas in order to comply with the law andeffectively steward these resources. Air qualitydecisions must be made now; they cannot awaitfull scientific understanding or development ofideal measurement end monitoring techniques.Information concerning current conditions alsowill be valuable in fulfilling FLMs’ broaderstewardship functions for these special areas.
Guidelines are essential to the FLMs’ airquality and management missions as well as to theprocess of sound scientific research.Standardized methods are crucial so thatcomparable data are produced from differentstudies and sites. Guidelines help ensurereproducible results and document the proceduresused so that future efforts can be related to olddata. Uniformity of technique also is criticalfor appropriately extrapolating results.Scientifically credible protocols provide theneeded basis for making sound regulatory, legal,and management decisions.
One final note: These protocols wereoriginally developed to apply to alpine andsubalpine ecosystems in areas where the airquality is considered to be clean. Followingtheir development it became clear that many ofthe measures recommended were more generallyapplicable. Nevertheless, caveats restrictingthe protocols to alpine and subalpine conditionsare widespread in the document. The reader iscautioned to use the protocolsaccordingly.
Douglas G. Fox
i
Protocol Design.............................................................The Decision Process........................................................General Guidelines for Sampling and Analysis...................................................Long-Term Monitoring: Measuring and Basic Sampling Design...............References...............................................................
WORK GROUP PARTICIPANTS........................................................
Contents
FOREWORD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Organization of Document . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
REGULATORY AND MANAGEMENT CONSTRAINTS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Clean Air Act Context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Wilderness Act Context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
ATMOSPHERIC ENVIRONMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Warm Season Measurements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Cold Season Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Field Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Laboratory Sample Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Support Needs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
VISIBILITY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Photographic Visibility System. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Transmissometer Measurement System. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
SOILS AND GEOLOGY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .List of Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Field Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Laboratory Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Support Needs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
AQUATIC CHEMISTRY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .List of Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Field Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Laboratory Sample Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Support Needs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
AQUATIC BIOLOGY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Salmonid Fish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Macroinvertebrates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
PLANTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Constraints and Philosophy of Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Page
i11122
33455577
1013141616181821212222222323252525252527282828282929293335373737384040414648
Guidelines for Measuring the Physical, Chemical,and Biological Condition of
Wilderness Ecosystems
Douglas G. Fox, J. Christopher Bernabo, and Betsy Hood
INTRODUCTION
BackgroundThis report presents appropriate scientific
protocols to measure current conditions of airquality related values (AQRVs) in Wildernessareas. These protocols are intended to beguidelines for quantifying the existing status ofAQRVs, monitoring for changes from these existingconditions, end subsequently, evaluating whetherthe changes are naturally occurring or the resultof man-caused air pollution/chemical deposition.Certain regulatory and management requirementshave constrained the development of the protocolspresented in this report. These constraints areexplained fully in the first section.
The scientists and contributors thatdeveloped this document (listed at the end ofthese Guidelines) were divided into six workgroups, with the Work Group Leaders responsiblefor the development of protocols for the fivetechnical areas: 1) Atmospheric Environment, 2)Soils and Geology, 3) Aquatic Chemistry andBiology, 4) Vegetation, and 5) Regulatory andManagement Constraints. A sixth group,Government Applications, was added shortly afterthe project began.
A meeting was held in Fort Collins, CO inJanuary 1986 to bring scientists developing theprotocols together with some of the federal lendmanagers (FLMs) and regulator who will be theusers of the guidelines. Discussions of draftlists of measures resulted in substantialprogress toward consensus. After furtherinternal review of the lists, the Work Groupsbegan to prepare their draft protocols.
A second meeting was held in June 1986 to 1)provide en opportunity for the project team torefine the draft protocols; 2) widen the sphereof participation in reviewing and refiningprotocols by involving additional key externalusers end scientists; and 3) aid in a smoothtransition to the larger consensus developmentmeeting, the Public Review. Consensus buildingamong diverse stakeholders and other interestedparties is an important component of thisproject. Involving key industry, state, federal,end environmental groups prior to the publicreview meeting enabled the Work Group Leaders toaddress emerging technical concerns early in theprocess.
The final teak of this phase of the projectwas the Public Review of the draft document. Aspart of this process, a Public Review Meeting,announced in the Federal Register, was held inBoulder, CO in December 1986. The purpose ofthis meeting was twofold: 1) to educate theparticipants on FLMs’ needs and regulatory andmanagement constraints, end 2) to allow review
end discussion on key technical issues, and todevelop consensus on these issues. The commentsreceived during this meeting, and writtencomments from the comment period following themeeting, were addressed by the Work Group Leadersduring the final revision of this document.
Protocol DevelopmentSeveral basic project assumptions were
discussed and clarified at the June meeting toprovide guidance in the preparation of theseprotocols:
1. The protocols are being developedinitially for application to high-elevationwestern areas.
2. The measurements and protocols will beused by FLMs end air quality permittingauthorities specifically for the protection of“air quality related values” of nationalWilderness areas designated as class I areasunder the Clean Air Act.
3. The measurements are not intended to be aresearch project, but will be conducted to fillspecific resource management regulatoryinformation needs.
4. In most cases, the land manager’s/permitting authority’s needs will bemet by the measurement of change in the mostsensitive component of the ecosystem. Thedetermination of whether a change is adverse isthe responsibility of the FLM.
Not all possible attributes can be measured,and the list must be parsimonious and practical.An attribute should have ecological significance,and should be likely to change as a consequenceof air quality effects. Ideally, it shouldchange only in response to changes in air qualityend nothing else; clearly an impossibility! Theattribute measured and the method of measurementmust be defensible to a consensus of thescientific community. Non-destructive methodsare preferable, not only because of Wildernessregulations but also because repeatedmeasurements of the same organism or assemblageis advantageous. Attribute variables that can bemonitored with low frequency should be givenconsideration over those that require manymeasurements at intervals of less then one year.Only those attributes that can be readilymeasured with high accuracy should be considered.
ProblemThere are several important reasons for
systematically establishing guidelines formethods end techniques for monitoring currentAQRV conditions and tracking future conditionchanges. These reasons include the following:
1. To provide clarity internally for the
1
Forest Service in Wilderness management,planning, operations, and decisions, andexternally to States and prospective new airpollution sources about what the FLM considersnecessary in monitoring the current conditions ofcertain AQRVs and potential air pollution-causedchanges in those AQRVs
2. To reduce the period of time needed tocomplete AQRV impact analyses during the airquality permit review process;
3. To provide a standardized approach toAQRV impact analysis so that analyses bydifferent parties can be compared, and so thatanalyses in different permitting cases can becompared with one another;
4. To provide a framework for due processwith respect to both the Forest Service’s AQRVimpact analyses and findings and its broaderWilderness protection mandate, thus enhancing thedefensibility of such analyses and findings inregulatory and judicial proceedings; and
5. To the extent feasible, to minimizeconflicts over technical issues surrounding AQRVsampling, monitoring, and measurement, thuslimiting disagreement where possible to valuejudgments about whether a projected AQRV effectis considered an “adverse impact.”
General IssuesGuidelines of several types are needed.
First, however, it must be established whatshould be measured to gauge man’s impacts onecosystems. Techniques and sampling and analyticprocedures then must be determined that areappropriate to the physical, regulatory, endmanagement constraints of Wilderness areas.These constraints include rugged, remote, oftenhigh-altitude settings subject to extremephysical conditions; Wilderness Act statutory andrelated regulatory prohibitions; Clean Air Actrequirements for permitting; and managementconstraints such as budget limitations.
Major problems only touched on in theseguidelines, but still requiring furtherconsideration, are:
1. Sampling intensity and location are keyvariables that must be determined, given thedegree of natural variability and physical limitson practical measurements. Not only in-depthknowledge of the natural systems, but alsostatistical design considerations bear on thisissue. Technical approaches must be developedfor the most practical end representative ways tomake the required measurements in Wildernessareas. Guidelines are needed for both on-siteand laboratory analysis so that sources of errorcan be minimized. A major challenge is designingsampling schemes that can adequately representthe diverse physical, chemical, and biologicalvariables.
1. Guidelines also are needed for datareduction, analysis, and archiving. Thesepost-measurement treatments of data and samplesare an important consideration for developingresults that will still be useful in the distantfuture. A quality assurance plan should bedeveloped along with the other protocols toensure reliable and meaningful results. Qualityassurance/quality control (QA/QC) is essential to
characterize adequately the sources of error andthe inherent uncertainties in the data collected.
2. Major concerns are how to address theinevitable tradeoffs between what measurementsideally are desirable scientifically end what isactually possible under the physical and legalconstraints imposed by high-elevation Wildernesssites and limited resources. It could be arguedthat not enough is known even to determine whatto measure, when, or how. This approach is not aluxury that FLMs can indulge. The task at handis the art of the possible; the immediate goal isto determine the best possible approach, fullydocument it, and then proceed to use it knowingit is not ideal.
The guideline protocols presented within thisdocument are not intended to represent all thatcan be measured within Wilderness ecosystems.Conversely, not all of the measurements suggestedhere may be necessary for a given site orsituation. These protocols are presented se areasonable list of measurements for establishingcurrent conditions in alpine and subalpine areasto aid in detecting changes in the future. Inaddition, a mechanism must be provided for theintegration of data collected on aquaticchemistry and biota, catchment soils, vegetation,and atmosphere. This integration will becritical to maximize confirming evidence formeasured effects.
The high degree of scientific uncertaintyabout how atmospheric chemicals influence naturalecosystems means no single widely accepted viewexists on many issues. Consensus building mustbe part of the entire process so that thegreatest degree of scientific credibilitypossible can be achieved. Part of the purpose ofthis project is to educate the research andtechnical community on what the FLM needs are,the reasons the FLM cannot wait for idealapproaches to be developed, and the legal andmanagement constraints under which the FLM mustact. The best current scientific judgment mustbe made, discussed, end agreed on to accomplishour goals.
Organization of Document
The first section of this document presentsthe paper prepared by Work Group 5, Regulatoryand Management Constraints. The followingsections present each set of protocols developedby Work Groups 1 through 4. The members of theWork Groups, including those in Work Group 6,Regulatory Applications, are listed on page 48,at the end of these Guidelines.
REGULATORY AND MANAGENT CONSTRAINTS
This chapter briefly explains constraints ondevelopment and application of scientificguidelines for the measurement and analysis of“air quality related values” (AQRVs) inwilderness areas. These constraints are imposedby the Clean Air Act (CAA) and the Wilderness Act(WA), the physical location of remote areas,weather, altitude, and other such factors. Themeasures included in this document have been
2
shaped and limited in their approach by the constraints Identified here. Therefore, thischapter is intended to assist readers indeveloping a full understanding and appreciationof the possible constraints on the developmentand implementation of the guidelines.
Clean Air Act Context
The Clean Air Act Amendments of 1977 includeda program for prevention of significantdeterioration of air quality, generally referredto as the “PSD” program. In part, this PSDprogram was intended to safeguard the air qualityrelated values (AQRVs) of Wilderness areas andNational Parks which the statute designates as“Class I areas.” This “Class I“ designationallows only very small “increments” of newpollution above already existing air pollutionlevels within the area, and subjects each sucharea’s AQRVs to special protection considerationsunder the Clean Air Act.
Under the CAA, the appropriate Federal LandManager (FLM) is charged with an “affirmativeresponsibility” to protect the AQRVS of Class Iareas from adverse air pollution impacts. In thecase of the Forest Service, the FLM’s affirmativeresponsibility to protect AQRVS has beendelegated to the Regional Forester level.
The FLM’s “affirmative responsibility” isimplemented, in part, through the PSD new sourcereview process, a preconstruction review andpermitting program for major new or expandingsources of pollution. Any major facility seekinga new source permit for location or expansion ina Clean Air area must meet several requirements,among them the Class I and/or II increments, theso-called AQRV “adverse impact test,” and theBest Available Control Technology (BACT)evaluation. In the PSD permitting process, theFLM determines whether a proposed source’semissions will have an adverse impact on Class IAQRVS.
New source permit applicants submit plans tothe permitting authority, who examines theproposed location of the facility, its generaldesign, projected air pollution emissions, andpotential impacts. When a proposed source’semissions may have an impact on a Class I area,the permitting authority (EPA, or the State, ifEPA has delegated PSD authority to that State)alerts the Federal Land Manager. The FLM thenconducts an “adverse impact determination” toassess the impact the projected pollution levelincreases would have on the Class I area. Theapplication review process may take as little as30 days or, with complex or controversialprojects, possibly longer than 1 year. The FLM’sadverse impact determination must be completedwithin this period.
Wilderness Act Context
Legal Direction for Managing WildernessCongress established the National Wilderness
preservation System in 1964 “to secure for theAmerican people en enduring resource ofwilderness.“ The Wilderness Act describes thebasic purpose of wilderness, defines the
wilderness resource end character. andestablishes management direction to preserve anenduring wilderness resource. This direction isthe foundation for the implementing regulations,found in 36 CFR 293, 36 CFR 291, and ForestService policy in FSM 2320.
The preservation of wilderness charactermeans striving to preserve “untrammeled” naturalconditions and “outstanding opportunities forsolitude.” This meaning applies to allwilderness management activities, includingresource monitoring of all kinds. Minimizing theeffects of human use or influences on naturalecological processes is the most importantprinciple of wilderness management. To clarifymanagement direction, the Act spells out specificprohibitions, while allowing only minimumnecessary exceptions:
Except as specifically provided for in thisAct, and subject to existing private rights,there shall be no commercial enterprise andno permanent road within any wilderness areadesignated by this Act and, except asnecessary to meet minimum requirements forthe administration of the area for thepurpose of this Act (including measuresrequired in emergencies involving the healthand safety of persons within the area), thereshall be no temporary road, no use of motorvehicles, motorized equipment or motorboats,no landing of aircraft, no other form ofmechanical transport, and no structure orinstallation within any such area. [Section4(c)]Forest Service wilderness q anagers must be
the leaders in demonstrating that wildernessmanagement tasks (including monitoring of airpollution impacts on resources) can be donewithout structures, installations, or the use ofmotorized equipment. The exception is to begranted only when it is clearly shown there is noother feasible way to gather information.
Criteria for Considering Exemptions toProhibitions
Measurement protocols that require exemptionsto the prohibitions against structures,installations, and motorized equipment inwilderness areas are not likely to be consideredfavorably. The criteria for consideringexemptions are found in the Forest Service Manualin the following sections:
Structures - 2324.3. This section setscriteria that are intended to limitstructures to “those actually needed formanagement, protection, end use of thewilderness for the purposes for which thewilderness was established.” This sectionalso requires documentation of need forstructures, schedules for their removal, andsets specific standards for materials andsiting.Research - 2324.4. While “encouragingresearch in wilderness that preserves thewilderness character of the area,” thissection requires that research proposals bereviewed “to ensure that research areasoutside the wilderness could not providesimilar research opportunities" and “to
3
ensure that research methods are compatiblewith wilderness values.” Further, itrequires specific use stipulations in theapproval document. Requests for exemptionsto wilderness access/user restrictions cannotbe based on economic costs to, or theconvenience of, the researchers.Motorized Equipment and Mechanical Transport- 2326. In an effort to “exclude the sight,sound, and other tangible evidence ofmotorized equipment or mechanical transportwithin wilderness,” this section lists thespecific criteria for exemption fromprohibitions on the use of motorizedequipment and mechanical transport inwilderness.
In all likelihood, then, a request for anexemption is likely to be refused unless it canbe demonstrated unequivocally that the data to begathered under the exemption are absolutelynecessary, and all possible alternatives to theexemption have been considered.
In conclusion, the Wilderness Act and theForest Service regulations require the use ofscientific protocols and measurements thatprotect wilderness values. This means that themeasurements either must be easily obtainablewithin the wilderness by primitive means, or beobtained from representative sites outside thewilderness.
Some Specific Constraints Considered in ProtocolDevelopment
The following are examples of specificfactors and issues that have constrained andshaped the development of protocols.
1. No guidance is given in the CAA as to howmuch advance notice the FLM must be given by thepermitting authority to allow proper assessmentof potential AQRV impacts of a proposed newsource of air pollution. Although the FLM mayhave more than a year, in practice he may have aslittle as 30 days for conducting this analysis.Thus, FLM may not be free to begin a monitoringstudy after he is presented with the permitapplication. To be useful in the permittingprocess, data must have already been collectedunder the protocols. Moreover, the data must
have been gathered over a sufficient time periodto establish meaningful current conditions of theresource in question.
2. The types of AQRV measurement andanalysis that can be performed may be seriouslyconstrained by certain physical and environmentalfactors in the alpine and subalpine setting.These factors may include weather, season, animaldamage, remoteness, and lack of power.
3 .The AQRV analyses likely will beconstrained by a lack of skilled personnel andfunding. In general the protocols call formonitoring efforts that are simple and cheap, usecurrent state-of-the-art methods and equipment,and do not push the boundaries of technology.
4. Because the results of AQRV analyses areto be used in the new source permitting process(and, potentially, in subsequent judicial review)a premium is placed on the reliability of theresults and the subsequent ability to make anddefend “yes” or "no" decisions concerning whethera proposed source will cause an adverse impact.
5. The ranges of uncertainty in determiningpotentially measurable changes in AQRVS (or indetermining the significance of any given change)as the result of proposed source emissions shouldbe clearly identified and described. Theimplications of such uncertainty should bedescribed adequately for nontechnical decisionmakers.
ConclusionThe constraints of the Clean Air Act, the
Wilderness Act, management considerations, andthe physical and environmental factors seriouslylimit the types of AQRV measurement and analysisthat may be performed in wilderness areas.Guidelines were developed within theseconstraints to insure realistic and feasibletechniques. Some compromises have been necessarybetween “ideal” or “preferred” AQRV measurementand analysis techniques and those which aredeemed “adequate” for management and regulatorypurposes. Therefore, the guidelines may be lessthan “state-of-the-art.” Nonetheless, theseguidelines and techniques are intended to bescientifically sound and accurate enough forreliable determinations of the current conditionof the area’s air quality related values.
4
Atmospheric Environment
Purpose
A major objective of this atmosphericcomponent of these guidelines is to establish areference for assessing the impact of airbornepollutants on sensitive ecosystems. To meet thisobjective, this guideline includes measurementmethodology for the ambient concentration ofcertain gases and aerosols, and for theconcentration of pollution-related ions inprecipitation and snow pack. Dry and wetdeposition of pollution-related material can beinferred from these ambient measurements.
Dry deposition fluxes can be computed bymultiplying the ambient concentration of thepollutant above a surface by its depositionvelocity, which is assumed to vary with landsurface type, time of day, season, and severalother factors. Meteorological measurements willtherefore accompany the ambient concentrationmeasurements. This approach represents a highlyempirical parameterization that relies heavily ona relatively sparse data base of dry depositionmeasurements.
Wet deposition can be estimated bymultiplying the precipitation-weighted ionconcentration by the total amount ofprecipitation, the latter measured by standardmeteorological means. During the cold season,snow pack measurements may be necessary for bothwet end dry deposition estimates.
Because of the difficulties of makingaerometric measurements within a Wilderness, oneor more sites will be established at the boundaryof the Wilderness, where the inflow and outflowof pollution-related material can be monitored.Within the Wilderness area, passive monitoringtechniques can be used such as measuring thetotal amount of precipitation, measuring thetotal snow pack depth, retrieving representativesnow pack samples for laboratory analysis, andestablishing a detailed inventory of land surfacetype.
The proposed protocols will provide estimatesof airborne pollution material. Compliance withpollutant regulations or with allowable airquality increments under prevention ofsignificant deterioration (PSD) regulations isnot being examined. These guidelines will not ofthemselves establish air quality baselines forpermitting new sources under the Clean Air Act.
The guidelines make maximum use of existingprocedures and methodologies that have been, orare being, field tested as part of a nationalnetwork.
Several critical assumptions have been madeduring the development of the atmosphericcomponent of the guidelines, These include:
1. Procedures and measurement methodologiesto evaluate compliance with existing standardsfor criteria pollutants are not discussed. Theconcentrations for gases and particles within aWilderness area are expected to be well withinexisting standards. A notable exception may be
ozone; its continuous measurement is thereforerecommended at all sites as required forcompliance testing. Should the need evolve forcompliance testing based on preliminaryassessments, taking into account results frommodel calculations and other efforts, then themethodologies published in the Federal Registerby the EPA will serve as protocols.
2. Models can be used for guidance inselecting regionally representative sites.
3. The meteorological and aerometricmeasurements should be expanded spatially andtemporally if the representativeness of thesemeasurements is in doubt. Aircraft sampling overthe Wilderness area and vertical profiles formeteorological data are powerful tools fordocumenting regional air quality.
4. Although large particle depositionresults in significant chemical input toecosystems, we assume that such particles areprimarily from local natural sources and hencenot man-caused. Thus no measures are suggestedat this time.
Warm Season Measurements
Gases and AerosolsTable 1 summarizes the aerometric parameters
that are measured in this protocol. Ourknowledge of trace gas and aerosol exchangebetween the atmosphere end the earth’s surfaceis limited to a small number of gases (mainlyozone, N02, HNO3 S02, and NH3 ). Thisguideline suggests dry deposition may be
Table 1.-Aerometric Measurements
QuantitativeTime detection
Analyte method resolution limit (QDL) Desired accuracy
O3 uv photometry,Automatic hourly 5ppb Lgr of QDL or 10%
NO21
filter pack day/night2 0.1ppb Lgr of QDL or 20%(TEA-impregnated (12 hr each)filter following for up to 1 wkTeflon and nylon) average
Mass,SO4 filter pack day/night2 0.2ppb Lgr 0f QDL or20%(Teflon) (12 hr each)
for up to 1 wkaverage
Inorganic filter pack day/nlght2 0.2ppb Lgr of QDL or 20%nitrate (Teflon, nylon) (12 hr each)
for up to 1 wk
Total filter pack day/night2ammonia (oxalic acid-
0.2ppb Lgr of QDL or 20%(12 hr each)
impregnated filter for up to 1 wkfollowing Teflon) average
SO2filter pack day/night2 0.2ppb Lgr of QDL or 20%(K2CO3-impregnated (12 hr each)filter following for up to 1 wkTeflon) average
1Optional.2With the option to measure over a 24-hr period only.
5
calculated from measured ambient airconcentrations. Actual deposition is inferredfrom concentration data end deposition velocitiesthat have been determined for specific gases endsurfaces. The following trace materials aremeasurement candidates:
Sulfur dioxide (S02) is a key primarypollutant of concern in Wilderness. Drydeposition of S02, especially to moistsurfaces, is considered a major sink, perhaps themajor sink of this species. Uptake of S02 byvegetation end ecosystems is an acid-producingprocess.
Nitrogen dioxide (N02) is phytotoxic andhence of direct concern in Wilderness. This gasis relatively insoluble in water, but is highlyreactive with biological materials. Chamberstudies as well as field measurements indicatethat N02 can dry deposit at moderate rates.
Ozone (O3) is formed in the atmosphere as aproduct of reactions involving hydrocarbons andnitrogen oxides. This connection with nitrogenoxides alone establishes its importance toWilderness. Ozone has been demonstrated to bephytotoxic. Dry deposition is a major sink of
O3; ozone is probably the best studied airpollutant for this reason. All of these reasonsplace O3 as high priority for measurement.
Nitric acid (HNO3) is the final product ofatmospheric oxidation of nitrogen oxides. It isa strong acid, highly soluble in water. Nitricacid is thought to be dry-deposited at a highrate, governed by atmospheric turbulence. Thehigh acidity of HNO3 as well as its role(often the case with nitrogen compounds) as aplant nutrient, establish the importance ofcharacterizing its dry deposition.
Ammonia (NH3 ) is not considered a directpollutant. Sources of NH3 are principallybiological: animal wastes, fertilizer, etc.NH3 is of interest in the Wilderness areacontext because it is one of the principalatmospheric bases available to neutralizeatmospheric acids. However it can contribute tosoil acidification when taken up by vegetation.Reactions of NH3 with aerosol H2SO4 resultin gas-to-particle conversion that in turnaffects the deposition and fate of NH3 .Additionally, NH3 as an available nitrogenspecies, is a nutrient to nitrogen-poorecosystems. Little is known about dry depositionvelocities of NH3 , but they may be large inview of the high solubility of NH3 at acidic toneutral PH.
Aerosol particles are a prime cause ofvisibility reduction as well as the means bywhich acidic material is delivered and hence areof considerable concern in Wilderness. Most ofthe sulfate and nitrate associated withatmospheric particles is found on particles of0.05 to 5 micron diameter, as a result ofgas-to-particle conversion. Because theseparticles can travel over long distances (becauseof their low gravitational settling velocities),they are good indicators of distant pollutionsources, particularly of sulfur dioxide.
Wet Deposition (Quality end Quantity)Table 2 lists the parameters to be measured
in precipitation. Wet deposition is a majorpathway for the transport of nitrogen end sulfurcompounds to the earth’s surface. Wet depositioncombined with dry (gases and aerosols) representstotal deposition. Thus precipitation quality andquantity is of major importance in determiningpollution impacts on ecosystems. Because of theremote location of Wilderness areas from majorpollution sources, a significant fraction of thetotal deposition of pollution-related materialwill be delivered as “wet” deposition.
Several national wet deposition networks havebeen in operation for several years. Theinstallation, operation, and subsequentlaboratory analyses are well established and areadapted for this protocol. All of the analyseslisted in table 2 can be conducted at a centrallaboratory. The only variables measured at themonitoring site after precipitation collection(or within a few hours driving distance from themonitoring site) are precipitation quantity,field pH, and conductivity.
Meteorological MeasurementsTo assess total deposition, both wet and dry,
a series of meteorological measurements arerequired. Meteorological variables to bemeasured at a representative site (outside theWilderness area) during the warm season include
Table 2. --Summary of analytes-analysis metheds, and detection limits forprecipitation
1Defined as the minimum value that is likely to be detected by the stated method when applied to actual precipitation samples. This value islarger than the minimum detection limit that is achievable in the laboratoryfor pure standard solutions.
2Defined as the maximum difference between the measured and the true valueof the quanity in question.
3Ion chromatography.4Automated wet chemistry.5Graphite furnace/ Atomatic absorption.6Inductive coupled plasma spectroscopy.
6
temperature, pressure, precipitation, wind speed,wind direction, surface wetness. and relativehumidity. Precipitation quantity is aparticularly important meteorological variable;wet acidic deposition calculations are verysensitive to estimated precipitation amounts.
Precipitation data measured at a singleWilderness area station during the warm seasoncannot be representative of precipitation overthe entire area. It is important to considerother mechanisms to collect precipitation datawithin the Wilderness. Because topography andground cover vary widely in these areas it is notpossible to estimate how many collection pointsmight be required in general. It may be possible
to estimate the amount of precipitation over theWilderness area from these combined data, but thereliability of any such effort will depend uponthe intensity and frequency of the sampling.Questions such as sampling design and reliabilityand representativeness of the data are notaddressed in this document.
Wind data are collected at the same locationto provide information necessary forunderstanding the variability in the aerometricand precipitation chemistry data, as thisvariability often can be attributed directly tosources upwind. Pressure and temperature dataare needed for calculating volume flow rates foraerometric samples. Temperature and humiditydata may be needed, along with land use data,vegetation cover data, surface wetness, and otherparameters, to estimate dry deposition rates.Meteorological data are also useful for modellingstudies that may be required in support of dataassessment.
The instruments proposed for this Wildernessarea program are listed in table 3. Theaccuracies specified by the manufacturers arenoted. This equipment has been field tested andhas been routinely used in many monitoringprograms. No recommendations are made for aspecific manufacturer; other instruments may haveequally satisfactory performance characteristics.
Cold Season Measurements
Table 2 lists the analyses to be conducted onsnow pack samples. The snow pack provides anaccumulation and integration of deposition eventsof natural and anthropogenic water-soluble andparticulate inputs. Total deposition ofpollution-related material averaged over theentire cold season (wet and dry) can be estimatedfor selected sites within the wilderness area,provided that snow melt occurs only during thenormal spring melt period and not intermittentlyduring the winter. Because of theinaccessibility of most parts of the wildernessarea during the cold season, seasonal totaldeposition may be the only measurement parameterobtained during this season.
Depending on the accessibility of the warmseason monitoring site, measurements of gases andaerosols should continue during winter. Themeasurement of “wet” precipitation by use of thewet-only sampler depends on the ability of thesampler to operate effectively under existing
7
Table 3.—Recommended meteorlogical equipment list, warm season
Parameter Measurement Method Range Accuracy
lAt sites where a significant fraction og the precipitation is in theform of snow, an Alter-type windshield will be added.
weather conditions. Meteorological measurementsshould continue throughout the cold season.
Requirements
Sampling Program for Warm Season
In principle, the monitoring sites cannot belocated within the wilderness area. The numberof required sites depends on the wilderness areaunder investigation. The equipment measurementsdetailed here are on a per station basis. Afurther assumption is that a qualified centrallaboratory(ies) will be responsible for preparingall required materials (filters, collectionbottles, shipment containers, etc.) for the fieldsites and analyzing the exposed filters andcollected precipitation samples. The equipmentneeds for such a central laboratory are notdetailed here, but table 4 presents an overviewof the expected concentrations in samples as afunction of various analytical techniquesavailable for the analysis of both impregnatedfilters (from filter pack) and precipitationsamples (rain and snow). Table 5 summarizes theinstrumentation requirements for the monitoringsite. All equipment or support items except thefilter pack are readily available from severalmanufacturers. All equipment and fieldprocedures for precipitation and snow packsamples should be (to the maximum extentpossible) identical to existing national or Stateprograms to insure maximum data compatibility.
No standardized equipment for filter packsystems is in use today. Both EPA and EPRI areplanning the deployment of filter pack systems in1987 as part of a nationwide gas sampling networkto measure ambient concentrations. The equipmentdiscussion that follows assumes such networkswill be implemented, and filter pack equipment(with impregnated filters) or annular denuderswill become commercially available for thiswilderness area monitoring protocol.
The filter pack air sampling method uses
Table 4.-Expected concentrations of rainwater contaminants, laboratory detection limits for selected constituents in rainwater and filter packs (min. detection level, MDL, in ug/ml, precision in %) .
can be monitored. When the airstream is drawn
selective filters within the filter pack tocollect specific pollutants over a 12-hour to7-day period, depending on protocol. The filterpacks are sheltered in a sample head, which ispermanently attached on a support pole at aheight of 7 meters.
With this pack of sequential absorbingfilters, the average concentration (12 hr up toone week) of fine particles (S04, NO3 ,NH4)
H N O3 , N O2 , N H3 and gases such as SO2,
through a size selective inlet, particles of aspecified size range can be captured. Variouschoices exist for the selection of filter media,absorbent, flow meters, and size selectiveinlets. Although filters are not considered anequipment item, they are discussed because theyare an essential part of the system.
Particles. Teflon membrane filters will beused to collect the particles before the airstream encounters other filters. A cyclone isincluded prior to all filters to remove particleslarger than approximately 2 um, and a shortlength of Teflon tubing is used as a transitionflow reactor for the flow before encountering thefilter (Knapp, et al. 1986, Durham, et al 1986).Teflon has been shown to quantitatively passHNO3 (Golden et al. 1983), although nitrateparticles collected on the filter may volatize(Appel et al. 1984). Although these filters areanalyzed only for sulfate and nitrate as part ofthe wilderness protocol, they are selected andsectioned so that more extensive chemicalanalyses can be performed on them at some laterdate. (X-ray fluorescence analysis for elementalspecies might be useful, for example.) This maybe useful for visibility considerations.
Nitric Acid and Ammonia. Nylon membrane
filters are used to capture nitric acid(HNO3). The specificity of nylon for HNO3capture has been demonstrated in both laboratory(Miller and Spicer 1975, Spicer et al. 1978) andfield studies (Spicer et al. 1982). Nylon doesnot remove NO2 or PAN but may absorb N2O5at high humidities. When located downstream ofthe Teflon filter, it also absorbs any HNO3 andsome of the SO2 that may be volatized from theparticulate collection. Nylon membrane filters(Nylasorb) have been used to trap nitric acidquantitatively (Spicer 1979). Recently , citricacid coated glass-fiber filters have beenrecommended for collection of ammonia (NH ) asa backup in the filter pack system. 3This allowsthe collection of any ammonia formed fromammonium nitrate particles collected upstream(EPA 1987).
Nitrogen Dioxide. A glass-fiber filterimpregnated with triethanolamine (TEA) absorbsnitrogen dioxide (NO2). The TEA filter willmeasure time-averaged low concentration NO2(Levaggi et al. 1973, Durham and Ellestad 1984,Knapp, et al. 1986).
Sulfur Dioxide.impregnated cellulose fiber filter has been shownto be an effective trap for sulfur dioxide(SO2) (Hugen 1963). A glass-fiber filterimpregnated with triethanolamine (TEA) has beenused to collect S02 (Knapp, et al. 1986).
Various options are available for a systemthat passes samples of the atmosphere throughthese filters. For the wilderness area,application of a heated Teflon-coated cyclonethat removes particles larger than 2 umaerodynamic diameter is proposed. Figure 1illustrates one of these systems (EPA 1987).
The cyclone assembly is housed in aninstrument shelter. The cyclone inlet isprotected from precipitation but able to sampleair directly. A minimum length of Teflon-coatedpipe is used to direct the sample streams to thefilter packs, located inside the shelter. Themass flowmeters and pump are located in aseparate pump box. When replicate sampling is
A K2CO3-glycerol
necessary, a second complete filter pack systemand shelter can be co-located.
Filters required for filter pack samplingmust meet the following requirements: 1 )mechanical stability, 2) chemical stability, 3)low flow resistance, 4) good retention withoutclogging, 5) low and consistent blank values forthe species being measured and those which mightadditionally be measured. and 6) reasonable costand availability.
EPA is currently developing protocols for atransition flow reactor (TFR} filter pack asillustrated in figure 1. Interested readers arereferred to the authors of this protocol for
further detailed information (Dr. Jack Durham,Atmospheric Sciences Research Laboratory, Officeof Research and Development, US EnvironmentalProtection Agency, Research Triangle Park. NC27711) .
Ozone. Ambient ozone (O3 ) concentration ismeasured with a W photometric type instrumentsuch as a Dasibi Environmental 1OO3-AH ozonemonitor or the equivalent model TECO 49P. TheDasibi W absorption photometer measures theamount of ultraviolet radiation absorbed by ozonein a sample of ambient air. The quantity oflight absorbed is proportional to theconcentration of ozone in the air sample. Ozoneconcentration readings are digitally displayed onthe front panel over the range of 0.000 to 1.000ppm. An analog output of O-1 VDC also isconnected to the data logger.
Gas is continually supplied to the samplechamber by a self-contained pump and handlingsystem. The intensity of the W beam traversingthe sample cell is attenuated in proportion tothe ozone concentration in the sample. Thesignal is electronically processed forpresentation by the readout system and output tothe data logger. Two reference subsystemsprovide a high degree of stability by correctingfor source intensity, optical path transmittance,and detector response changes. Self zeroing andinterference removal are accomplished bycomparison of sample and reference readings. Ifthe operating parameters of the analyzer arewithin specifications. no span or zero driftoccurs and the analyzer is self-calibrating.
Cold Season -- Snowpack SamplingSeveral standardized tools are available for
sampling snow cover. Table 6 summarizes theproperties of snow samplers used in NorthAmerica.
Tests suggest that a sharp “Federal sampler”(or equivalent) is suitable for use in ❑ ost typesand depths of snow cover. Cooperative testing byNorth American agencies through the Western SnowConference is continuing in an attempt to developa standard metric sampler that will provideaccurate and repeatable measurements for deep andshallow snow covers (Fames et al. 1980).Currently, the “Standard Federal” is thepreferred choice throughout the western U.S. andCanada. Experience indicates that, in deep snowpacks (> 4-5 m depth) with numerous ice lenses,the Standard Federal corer is not sufficientlyrobust for repeated coring during a single fieldtrip. This is especially true for coring incold, continental snow packs such as those foundin the Rocky Mountain region. In such cases, aMcCall corer should be used. Cross-calibrationto Standard Federal core sampling efficiency hasbeen reviewed by Fames et al. (1980).
For very dense, deep snowpacks, a combinationof a core and a snowpit may be necessary, sinceit may be impossible to extract the coring tool(Dozier, pers. comm.). Snow pits also allow muchmore detailed examination of the snow. Theyprovide the only practical method of determininglayer structure, ice lense structure, and snowmicrostructure available at the present time.Details of snow pit observations are found inPerla and Martinelli (1976) and Jones (1983).
9
Whether or not the increased information is worththe increased work of digging pits depends on theintended use of the information.
Field Procedures
Sampling Site Selection Criteria -- Warm SeasonRegional considerations. --Because of the
difficulties in operating monitoring sites withinthe wilderness area, “representative” locationswill be chosen for sites at the periphery of thearea. “Representative” in this context refers tothe climatology of the region and to the synopticscale air mass flowing over the wilderness area.Both of these overall meteorological parametersshould be assessed in conjunction with man-madepollution sources in a roughly 500-mile zonesurrounding the wilderness area to determineapproximate locations for potential monitoringsite(s). Ideally, such en assessment shouldyield monitoring site(s) that can characterizethe flow of pollution-related material into endout of the wilderness area. The number ofmonitoring sites must be established on acase-by-case basis, end obviously depends on thesize of the wilderness area, the complexity ofterrain, the acceptable level of uncertainty,etc. Determining an appropriate number of sitesis not en easy task. TAPAS models (Fox, et. al.,1987) are available to aid in this task.
Local considerations. --The most importantcriterion is the availability of electricalpower, because most of the atmospheric samplersrequire at least 110 V electricity. Within theconstraints set by the availability of power, thesampling site should be as close to the remotearea of interest as possible. The temporalvariation of atmospheric concentrations ofinterest is probably much greater than thespatial variation, particularly in backgroundlocations; however, very little data areavailable to confirm this speculation.
The selection of the atmospheric samplingsite also should depend on potential localsources of pollution. Potential local sourcesinclude home chimneys, vehicular traffic,auxiliary diesel generators, and local industrialactivities. Seasonal changes in activitiesproducing potential pollution also should benoted.
1 0
Because of the size of some of the equipmentand the need for servicing on a year-round basis,the site should have reasonable access: oneshould be able to drive close (0.5 km) to thesite during various weather conditions. Ideally,the site should be located where year-round staffare available to service the equipment and changethe sampling heads.
The site should be protected from animals andunauthorized human entry. The first line ofdefense is to locate the site out of normalview. Other ways to protect the site includefencing, signing, or locating at a site with apermanent resident on officially protectedproperty such as a ranger station, a universityfield station, etc. Experience indicates thatthe most effective protection of a site is tokeep it out of everyday view.
Ideally, the site should have a completeground cover to minimize resuspended materialsand dust from the local area, but should never beunder or near overstory vegetation. As a rule ofthumb, the diameter of the opening should beabout 10 times the average height of thesurrounding overstory vegetation. The siteshould be located so that sampling will reflect,as accurately as possible, the chemicalconstituency of air masses of fairly largecirculation.
Because many sampling sites will not have allof the desirable attributes, some compromisesmust be made. To evaluate tradeoffs, asystematic decision-making process should beused. For example, site criteria are dividedinto those which must be met, and those which aredesirable, Ranking is based first on “musts”,then “wants”; a final decision is made by a groupof experts in atmospheric sampling.
Site criteria are summarized in table 7. Theinstruction manual issued by NADP (NADP 1984a)provides further information regarding theestablishment of a wet deposition site using thewet/dry precipitation collector.
Sampling Site Selection Criteria for Snowpack --Cold Season
Total deposition of pollutant-relatedmaterial accumulated in snow over the entire coldseason will be monitored at sites within thewilderness area. The number of snow cores to besampled varies with the size of the wildernessarea, the extent of ecologically sensitiveregions, the complexity of the terrain, and otherfactors, but a minimum of five samples should betaken at each site. These should be selected tocollect maximum deposition.
The snowpack should be sampled at maximumaccumulation, but before spring melt starts.These ideal conditions are not always met. TheCascade and Sierra Nevada Mountains have a warmsnowpack with temperatures usually near 0°C(Smith 1974) . Because of air temperaturevariations, some melting of the snowpack mayoccur during the winter, and depending ontemperature conditions, rain may fall on thesnowpack and percolate through it. Suchpercolation, if it continues through the entiredepth of the snowpack, can leach soluble materialfrom the snow in concentrations disproportionate
to those in the snowpack (Johannessen et al.1980) . In addition, atmospheric conditions underwhich the snow was deposited, the degree and typeof metamorphism the snow has undergone, and theintensity of rain and/or melt events all caninfluence the rate at which impurities can beremoved from the snow (Shockey and Taylor 1984).Thus, the snowpack cannot be assumed toaccumulate and hold all atmospheric depositionduring the life of the snowpack.
To lessen the possibility of rain and meltingimpacts, sampling sites should be located abovethe freezing level for the particular geographicregion under consideration. Whenever possible,
11
the sampling sites should be located in thesouthwestern part. of meadows or in open areaswhere shading minimizes surface melting fromsolar radiation. At lower elevations, thetemperature of the snowpack may reach 0°C andtherefore endanger its integrity. Selectiveleaching of ions from the snowpack can beidentified by setting out waterproof boxes ofabout 2 X 2 m and lined with polypropyleneplastic in the fall. Snow cores collectedoutside the boxes can be compared with cores plusmalt water from inside the boxes during the laterwinter sampling period.
Sample Collection Procedure -- Warm SeasonFilter pack. --Every seven days (when samples
are removed from the precipitation collector),the filter pack is removed and a new filter packis installed. The following is a preliminarydescription of the procedures to be followed.The final protocol will depend on the selectionmade for the national programs.
1. Check the flow rate as indicated on thesampler control module digital readout. Obtainthe actual flow rate from the calibration sheetthat corresponds to the indicated value from thesampler. Record this value as the “OFF FlowRate” in the log book and sample record sheet.
2. Note the time on the data loggerdisplay. Record this time as the “OFF time” inthe log book and sample record sheet.
3. Lower the sample head by releasing thecam lock at the base of the tower and slowlyfeeding out the line tied to the tower upright.
4. Remove the filter packs by pulling up onthe quick-connect fittings collar. Place thefilter packs in zip lock bags.
5. The quick-connect fittings sealthemselves off when no filter is installed: thiswill check the system for leaks. After oneminute, check to see that no flow is indicated onthe control module digital readout.
6. Install the new filter packs by pushingthem into the quick-connect fittings in the baseplate. The “sample” filter pack should beinstalled in the fitting marked SAMPLE and theblank filter pack in the fitting marked BLANK.
7. Raise the tower by pulling on the lineattached to the end of the tower upright. Securethe upright into the tower base plate and engagethe cam lock.
8. Note the time indicated on the datalogger display. Record this time as the “ONtime” in the log book and sample record sheet.
9. Check the flow as indicated on thecontrol module digital readout. If necessary,adjust the flow to the value corresponding to aflow rate of 1.5 liters per minute from thesampler calibration sheet. Record this value asthe “ON flow rate” in the log book and samplerecord sheet.
Filter pack handling and shipment.--Afterremoval, the complete filter pack is sealed onboth ends with plastic screw caps, placed insidea zip lock bag, tagged, end shipped inside apadded box to a central analytical laboratory.The following information should be recorded inthe station log book and on the sampleidentification tag, which will be attached to the
zip-lock bag containing the filter pack: filternumber, side ID number, start date and time, endstop date and time.
Standard Operating Procedures (SOPS) havebeen developed for all phases of the fieldsampling collection by EPA and AtmosphericEnvironment Service, Canada.
Precipitation. --An SOP for the measurement ofwet deposition exists for all major nationalnetworks. The SOP for NADP/NTN will be adoptedhere. Operational steps including bucketchanging and weighing, sample storage, fieldlaboratory analysis (pH end conductancemeasurement ), shipment and maintenance aredetailed in the NADP site operation manual (NADP1982) . .
In summary, the NADP/NTN protocol is thefollowing:
1. An aerochem Metric Model 301 wet/dryprecipitation collector collects precipitationsamples, and a Belfort recording rain gaugemeasures daily precipitation amounts.
2. Samples are collected weekly.3. The sample is weighed at the site to
determine total precipitation volume. Thesoil-contaminated portion of sample is carefullyremoved.
4. A 20 ml aliquot is removed for laboratoryand pH measurements.
5. A form is filled out by the site operatordescribing the sample and the collectioncharacteristics (see fig. 2).
6. The sample is mailed in the sealedcollection container along with the samplereporting form to a central laboratory foranalysis.
Because of the dilute nature of precipitationsamples, handling procedures must be followedcarefully to prevent contamination. Theseprocedures are presented in detail in theNADP/NTN manual. This plan is adopted for theoperation of the monitoring sites, with theexception of those sections which referspecifically to liaison with the NADP/NTN CentralAnalytical Laboratory (CAL).
Ozone. --SOPS exist for all aspects of ozonemeasurement, calibration, and preventive fieldmaintenance. They are detailed in, and part of,the owner’s manual supplied by the manufacturer(either TECO or Dasibi). The Mountain CloudChemistry Standard Operating Procedures can beused for further guidance.
Snowpack Collection Procedure -- Cold SeasonThe snow sampler is lowered vertically into
the snowpack with a steady thrust downward. Asmall amount of twisting aids in driving the tubeend cutting thin ice layers, but considerableforce end twisting of the sampler with a drivingwrench may be required to penetrate hard layersof ground ice. Penetration to extract a soilplug helps to prevent the loss of the snow corefrom the tube, and a trace of soil or litter inthe cutter indicates no loss has occurred. Aquick comparison of the length of the snow coreagainst measured snow depth will show whether acomplete core has been obtained. The amount ofcompaction of the snow core during sampling willdepend on snow conditions. If the snow core
12
Figure 2. --NADP/NTN field observer report form.
becomes blocked or frozen in the tube, preventingsnow from entering, the core should be discardedand another sample taken. Snow may freeze in thetube when the snow temperature is below 0°C andthe air temperature is above 0°C. When a goodsnow core has been obtained, the sample isweighed in the tube and the combined weight (inwater equivalent units) is read directly with aspring balance. The tare weight of the tube issubtracted to obtain the snow water equivalent.
Figure 3 illustrates a convenient format forrecording snow survey information in the field.Such a form also provides documentation of anyproblems encountered while surveying that mayaffect the accuracy of the survey and theInterpretation of the results.
All snow samples should be double-bagged inpolypropylene (after dirt or soil has beencarefully removed from bottom), heat sealed, endkept frozen by mechanical refrigeration untilthey are analyzed in a designated chemicallaboratory.
Laboratory Sample Analysis
Filter PackDevelopment of an SOP currently is being
funded by EPA and EPRI as part of theimplementation of a dry deposition network. TheSOP will describe the processing of filtersamples from initial acceptance testing throughlaboratory analysis of the filter extracts. Theacceptance criteria end the manner in whichacceptance testing is conducted will be specifiedin the SOP. All filters that pass acceptancetesting are then weighed, packaged, numbered, andsent to the sampling sites.
Upon receipt at the laboratory, each filteris weighed and a certain fraction (specified bythe SOP) is reweighed separately. All filterspassing quality acceptance tests are chemicallyanalyzed by laboratory processes analogous tothose used for precipitation samples.
The preliminary analytical procedures forextracting and analyzing filter pack samples are
Figure 3. --Format for recording snow surveyinformation.
contained in the following SOP’s (U.S. EPA) :1. EMSL/RTP-SOP-QAD-531, Extracting and
Analyzing Dry Deposition Samples, May 14, 1985.2. EMSL/RTP-SOP-QAD-503, Analysis of Anion
Samples by IC, January 30, 1985.
PrecipitationAll sample aliquots arrive at a central
laboratory in special containers on a weeklybasis. SOPS exist for all phases of analysis andquality control. This wilderness area protocolincludes the collection and analysis proceduresdeveloped by the Illinois State Water Survey andoutlined in Peden et al. (19861.
SnowpackSamples should be analyzed at a central
laboratory. To ensure uniform handling, themelting, filtration, and bottling processesshould follow strict protocol (Shockey and Taylor1984). During melting, the overall temperatureof the sample should never exceed 4°C. Thisprocedure will minimize any bacterialdeterioration of the nutrient constituents.Immediately after thawing at the centrallaboratory, the samples should be filtered,
preserved, end bottled.Because low, near-detection-limit
concentrations of solute are expected in thesnowpack from remote regions, extraordinary caremust be taken in sampling, processing, andanalysis, as well as in collection of manysamples, if a valid picture of total depositionis to be obtained. Such standard operatingprocedures must be developed as part of a qualityassurance plan. The procedures for analysis ofthe melted snow are analogous to those used forprecipitation samples.
Support Needs
Data Collection from Continuous MonitorsMeasurements made by continuous monitors are
collected on a data logger at each site withback-up by strip chart recorder. The loggershould be a Campbell Scientific Model CR21X/L orequivalent. The collection of continuous data isdesigned to: 1) compute accurate averages orsums by regular sampling of the data channel, 2)allow checking of data on a regular basis to spotdeviations from expected operation, and 3)retrieve the dataefficiently.
The data acquisition system is illustrated infigure 4. One-hour averages (1-hr averages) arecomputed in the data logger from scans made at5-second intervals. The variables scanned arewind speed and direction, temperature (ambientand shelter), relative humidity, pressure,precipitation amount, and ozone concentration.Status channels will indicate when calibrationsor zero/span checks occur, position of the lid onthe precipitation bucket, and site service by adoor alarm (checked to assure that routine sitevisits are being performed by the sitetechnicians). The averages are stored in thedata logger and on magnetic media for laterretrieval. The data are available for inspectionand retrieval both on-site and remotely bytelephone if a telephone link is convenientlyavailable.
Flags will be placed on variables that haveyielded less than 200 valid observations for a 60-minute period because this average isconsidered potentially invalid. Magnetic mediawill be collected on a weekly basis (or daily bytelephone polling if available). Strip chartsserve as the final backup if the data loggershould fail.
Strip charts will be changed every 2 weeksand archived for referral as needed. Stripcharts are necessary if compliance with standardQA/QC procedures are considered essential. Thepolled data should be examined by experiencedpersonnel to detect any instrument problems andsuspect data that need to be checked or validatedat the site. The data should be processed at adesignated processing center and checked forrange validity, rate of change, and otherautomatic checks programmed into the data archivesystem. Flags are assigned to suspect data. Allflagged data should be examined by an experienceddata technician. The technician reviews the❑ agnetic media or strip charts as necessary tovalidate questionable data. The data are
14
Figure 4. --Schematic of data logging for meteorological and associatedmeasurements.
summarized each month. Quality control isachieved by checking in the field end at thecentral location. Protocols for QA/QC areavailable from several field programs.
Quality AssuranceThe aerometric, meteorological,
precipitation, end snow core data collected bythe site personnel should be of a qualityconsistent with the requirement of impactassessment. Archiving this goal will require awell-conceived quality assurance plan andrigorous adherence to this plan throughout alloperational phases of wilderness measurement(NADP 1984b).
Quality assurance can be divided into twotypes of activities: quality control end qualityauditing.
Quality control consists of a set of
mandatory procedures to be followed during thedesign, collection, and analysis phases of ameasurement program. These procedures aredesigned to insure that the data from the programmeet a predetermined set of performancecriteria. Quality control activities alsoprovide the information needed to determine theuncertainty in the measurements, i.e., precisionand accuracy. Quality control is therefore anongoing activity performed by the personsactually making the measurements.
The performance of instrumentation endlaboratory procedures may be evaluated bycomparison with NBS-traceable standards or by theanalysis of blind samples. The performance ofdata processing procedures is tested byindependently processing representative sets ofmeasurements by en auditor. Whenever possible,existing QA/QC protocols and SOPS are to be used.
Data Analysis
The monitoring protocol outlined for thewilderness area provides aerometric andprecipitation data sets with known accuracy andprecision. The accuracy and precision ofanalytical measurements of air or precipitationsamples are evaluated, in principle, in thefollowing manner: 1) accuracy is determined byanalyzing EPA and NBS reference samples, unknownreference samples, and spiked samples; 2)precision is determined by analyzing replicatedfilter or precipitation samples.
The aerometric measurements are reported asconcentration in parts per billion (ppb) byvolume and in micrograms per cubic meter(ug/m3), averaged over a time period of sevendays. (The protocol is still open as to thesampling mode, i.e., separate day-night samplesaveraged over one week or one week totalaverage.)
The precipitation measurements (warm season)are reported as concentration in micromoles perliter (umole/1) averaged over the weekly samplingperiod. The total precipitation volumeaccumulated during one week is reported inmilliliters (ml). Precipitation amount isrecorded separately by the rain gauge asmillimeters per day. The snowpack samples arereported as concentration averaged over theentire accumulation time. The total amount ofsnow accumulated is presented in snow depth,water equivalent, and total volume of water(liters).
From these primary data, one can derive dryand wet deposition data. In addition, the airquality at the boundary of the wilderness areacan be established. Air quality within thewilderness area can be estimated by combining thelocally measured meteorological parameters withthe synoptic scale air flow obtained fromstandard weather stations.
Concern over the deposition of acidicsubstances has led to an awareness of limitationsin the current ability to monitor drydeposition. At present, relatively few programsare designed to produce dry deposition fluxestimates, in contrast to the existence ofseveral networks that produce wet depositionfluxes. The delay in setting up dry depositionmonitoring networks is due primarily to thescientific uncertainty of the necessarymeasurements. No unequivocally accepted methodexists for monitoring dry deposition.
Because it is difficult to measure fluxes atthe surface itself, dry deposition rates areusually inferred from data obtained in the airabove the surface. The critical assumption inthis approach is that fluxes measured above thesurface are the same as those at the surface, anassumption that depends on the homogeneity of thesurroundings.
The deposition velocity, v , if known,provides a convenient method for deriving thedeposition flux, F, from measurements ofconcentration in air, C: F = vd C. Thiscalculation is the basis for the inferential or“concentration-monitoring” method. However, thedeposition velocity is not fixed for each
pollutant species and surface of interest. Inreality, values of v are site-specific andtime-varying. For this reason, knowledge of theland use and vegetation cover within thewilderness area is essential to associate“appropriate” situation-dependent depositionvelocities with the measured, ambient pollutantconcentration.
Since the air quality parameters have beenmeasured at another location (at the periphery ofthe wilderness area), the dry deposition flux asderived in this protocol can only be used as arough guideline to indicate the influx ofpollutants. As a rough guess, the dry depositionfluxes estimated on the basis of this protocolmay be accurate only to a factor of two, whereasthe concentration values are significantly moreprecise.
On the other hand, wet deposition and totaldeposition (snow pack) may be obtained withuncertainties less than 50$, particularly iflocal precipitation amount is known. Wetdeposition is derived as the product ofconcentration and rainfall amount (warm season).Total deposition accumulated over the cold seasonis obtained directly from snow depth (waterequivalent) and measured concentration ofpollutant material in the snow pack.
References
Anlauf, K. G.; Fellian, P.; Wiebe, H. A.; Schiff,H. I.; Mackay, G. I.; Braman, R. S.;
Gilbert, R. 1985. Comparison of three methodsfor measurement for atmospheric nitric acidand aerosol nitrate and ammonium. AtmosphericEnvironment 19: 325-333.
Appel, B. R.; Tokiwa, Y.; Haik, M. 1981. Samplingof nitrate in ambient air.AtmosphericEnvironment 15: 283-289.
Appel, B. R.; Tokiwa, Y.; Haik, M.; Kothny, E. L.1984. Artifact particulate sulfate andnitrate formation on filter media.Atmospheric Environment 18: 409-416.
Appel, B. R. 1986. California Department ofHealth Services, Berkeley, California.Personal communication.
Barrie, L. A.; Vet, R. J. 1984. The concentrationand deposition of acidity, major ions andtrace metals in the snowpack of the easternCanadian Shield during the winter of1980-1981. Atmospheric Environment 18:1459-1469.
Cadle, S. H.; Countess, R. J.; Kelley, N. A.1982. Nitric acid and ammonia inurban and rural locations. AtmosphericEnvironment 16: 2501-2506.
Chan, W. H.; Tomassini, F.; Loescher, B. 1983. Anevaluation of sorption properties ofprecipitation constituents on polyethylenesurfaces. Atmospheric Environment 17:1779-1785.
Dozier, J. Dept. of Geography, University ofCalifornia, Santa Barbara, CA.
Durham, J. L.; Ellested, T. G. 1984. A prototypeconcentration monitor for estimating acidicdry deposition. Presented at Air pollution
16
control association 77th annual meeting; 1984June 24-29; San Francisco, CA.
Durham, J. L.; Ellestad, T. G.; Stockburger, L.;Knapp, K. T.; Spiller, L. L. 1986. Atransition-flow reactor tube for measuringtrace gas concentrations. Journal of the AirPollution Control Association 36: 1228-1232.
Environmental Protection Agency. 1987. Protocolfor the transition-flow reactor concentrationmonitor. Atmospheric Sciences ResearchLaboratory, Office of Research andDevelopment, U.S. Environmental ProtectionAgency, Research Triangle park, NC 27711.June 9, 1987. 26 p.
Fames, P. E.; Goodison, B. E.; Peterson, N. R.;Richards, R. P. 1980. Proposed metric snowsamplers. Proc. 48th Annual Meeting, WesternSnow Conference.
Fassel end Knisely. 1974. Annalen der Chemie 46:1110.
Fox, D. G.; Dietrich, D. L.; Mussard, D. E. [andothers]. 1987. The topographic air pollutionanalysis system. In: Zannetti, P., ed.Envirosoft 86: proceedings of theinternational conference on development andapplication of computer techniques toenvironmental studies; 1986 November; LosAngeles: Computational Mechanics: 123-144.
Galloway, J. N.; Thornton, J. D.; Norton, S. A.;Volchok, H. L.; McLeon, R. A. N. 1982. Tracemetals in atmospheric deposition: a reviewend assessment. Atmospheric Environment 16:1677-1700.
Golden, P. D.; Kuster, W. C.; Abitton, F. C.;Fehsenfeld, F. C.; Connel, P. S.; Norton, R.B.: Huebert, B. J. 1983. Calibration andtests of the filter-collection method formeasuring clean-air, ambient levels of nitricacid. Atmospheric Environment 17: 1355-1364.
Grosjean, D. 1982. The stability of particulatenitrate in the Los Angeles atmosphere.Science of the Total Environment 25: 263-275.
Guiang, F. S.; Krupa, S. V.; Pratt, G. C. 1984.Measurements of S(IV) and organic anions inMinnesota rain. Atmospheric Environment 18:1677-1682.
Hugen, D. A. 1963. The sampling of sulfur dioxidein the air with impregnated filters. Annalidi Chimica 28: 349.
Johannessen, M.; Skartveit, A.; Wright, R. F.1980. Proceedings, International conferenceon ecological impact of Acid Precipitation.SNSF Project, 1432, As-NLH, Norway, 224-225.
Jones, E. B. 1983. Snowpack ground-truth manual.NASA CR 170584, Goddard Space Flight Center,Greenbelt, Maryland 20771, 39 PP.
Knapp, K. T.; Durham, J. L.; Ellestad, T. G.1986. pollutant sampler for measurements ofatmospheric acidic dry deposition.Environmental Science Technology 20: 633-637.
Levaggi, D. A.; Siu, W.; Feldstein, M. 1973. Anew method for measuring average 24-hournitrogen dioxide concentrations in theatmosphere. Journal of the Air PollutionControl Association 23: 30-33.
MAP3S/RAINE. 1982. The MAP3S/RAINE precipitationchemistry network: statistical overview forthe period 1976-1980. Atmospheric Environment16: 1603-1631.
Miller, D. F.; Spicer, C. W. 1975. Measurement ofnitric acid in smog. Journal of the AirPollution Control Association 25: 940-942.
National Atmospheric Deposition Program. 1982.NADP instruction manual--site operation.Bigelow, D. S., ed. Natural Resource EcologyLaboratory, Colorado State University, Ft.Collins, CO. January.
National Atmospheric Deposition Program. 1984.NADP/NTN instruction manual--site selectionand installation. Bigelow, D. S., ed. NaturalResource Ecology Laboratory, Colorado StateUniversity, Ft. Collins, CO. July.
National Atmospheric Deposition Program. 1984.NADP quality assurance plan--depositionmonitoring. Prepared by NADP QualityAssurance Steering Committee. 39 p.
Peden, M. E. et al. 1986. Development of standardmethods for the collection and analysis ofprecipitation. u.S. EPA Contract No. CR81O780-01-1. Illinois State Water Survey,Analytical Chemistry Unit. March 1986.
Perla, Ronald I.; Martinelli, M. Jr. 1975.Avalanche handbook. Agric. Handb. 489.Washington, DC: U.S. Department ofAgriculture. 238 p.
Shaw, R. W., Jr.; Stevens, R. K.; Bowermaster, J.[and others]. 1982. Measurements ofatmospheric nitrate and nitric acid - thedenuder difference experiment. AtmosphericEnvironment 16: 845-853.
Shockey, M. W.; Taylor, H. E. 1984. A method ofpretreatment of snow samples to ensure sampleintegrity for analytical testing. 27th RockyMountain Conference, Denver, CO, 19.
Smith, J. L. 1974. Advanced concepts andtechniques in the study of snow end iceresources. Washington, DC: National Academyof Sciences: 76-89.
Spicer, C. W.; Ward, G. F.; Gay, B. W. 1978.Further evaluation of micro-coulometry foratmospheric nitric acid monitoring.Analytical Letters All: 85-95.
Spicer, C. W. 1979. Measurement of gaseous HN03by electro-chemistry and chemiluminescence.In: Stevens, R. K., ed. Current methods formeasure atmospheric nitric acid and nitrateartifacts, EPA Report 600/2-79-051.
Spicer, C. W.; Howes, J. E.; Bishop, T. A.;Arnold, L. H.; Stevens, R. K. 1982. Nitricacid measurement methods: an intercomparison.Atmospheric environment 16: 1478-1500.
17
Visibility
Visibility (including site visual range,contrast, color, plume blight) in Class I areasis protected under the provisions of the CleanAir Act of 1977, which stipulates that thevisibility within Class I areas is not to bedegraded and, if possible, is to be brought backto pristine levels. Only visual scenes withinClass I area boundaries are protected; “integralvistas” are not protected.
Visibility measurements can be made byseveral techniques. Since monitoring equipmentis not permitted within Forest Service Wildernessboundaries, the Forest Service has adopted thepolicy of monitoring visibility from locationsadjacent to the boundaries. It is assumed thatthe measurements taken from the nearby sitelooking into and across the Wilderness arerepresentative of the Class I visual sirquality.
A primary goal of visibility monitoring is toquantify how well the image forming informationin a vista is transmitted through the atmosphereto en observer some distance away. This requiresan understanding of atmospheric extinction (thescattering and absorbing properties of theatmosphere that influence the transmission oflight) .
Three primary operational electro-opticalmonitoring techniques are available: integratingnephelometers (Charlson et al. 1967);teleradiometric techniques using natural targets(Maim and Molenar 1984, Johnson et al. 1985); andtransmissometers (Maim et al. 1986, Maim andTombach 1986). Each method has advantages anddisadvantages. For monitoring near wildernessareas where access and manpower are limited andpower is generally unavailable, the mostsuccessfully applied technique has beenphotography. Photography has therefore beenselected as a practical end economicalmeasurement method.
Transmissometry techniques are currentlyplanned for several Forest Service Wildernessareas as part of the IMPROVE program (InteragencyMonitoring of Protected Visual Environments,Joseph et al. 1986). However, currenttransmissometer systems are experimental end havepower, data collection, installation, cost,service, and logistics requirements that makethem impractical at most wilderness sites.
Photographic Visibility System
The photographic technique was first proposedby Steffans (1949) end was later refined byHoffer et al. (1982) and Johnson et al. (1985).Photography offers simplicity and economy in dataacquisition with the added advantages of 1)quality assurance of the measurements during datareduction sod analysis, and 2) a 35mm slidearchive available for future analysis andreference.
The primary electro-optical measurement oftarget/sky horizon contrast is made by
microdensitometric analysis of the 35mm slides.This technique emulates teleradiometermeasurements. The technique is an indirectmeasurement of the visual air quality because itdepends on the film media to accurately depictvisual conditions. Sampling is limited todaylight hours.
The color slides can provide the followinginformation:
1. The general condition of the sky andterrain features.
2. The relative color of the sky and terrainfeatures, as well as the presence of layered oruniform haze.
3. A target/sky horizon contrast that isreducible to standard visual range under optimalconditions.
4. Slide archives that provide an easilyinterpreted and relatively permanent visualrecord of conditions within the wilderness.
System ComponentsA primary photographic monitoring system
includes the following components:1. Rugged, reliable 35mm camera body with
automatic film winder. The camera’s automaticexposure meter must be designed so that it is ononly during the actual time of exposure end notcontinuously operating.
2. 135mm lens with UV filter.3. Databack capable of imprinting the day
and time the exposure was taken on the film.4. Battery powered programmable timer
capable of triggering the camera at least threetimes per day.
5 . The complete system must be able behoused in a small, stand-along environmentalclosure, and operate within the ambienttemperature range of -30° to 130°F unattendedfor at least 10 days.
Figure 5.--Field installation of automaticphotographic visibility monitoring used atmany Forest Service and Park Service sites(photo courtesy of Air Resource Specialists,Inc. , Fort Collins, CO)
18
A commercially available system (from AirResource Specialists, Inc.) that meets all of theabove criteria is shown in figures 5 end 6.Systems are currently operating in over 25 ForestService sites. A variety of cameraconfigurations can be fabricated to meet specificsite requirements. For example, some existingForest Service sites operate dual camera systemsthat take two exposures per day and are servicedmonthly.
Photographic Siting CriteriaThe overall configuration of the monitoring
site depends on the characteristics of the siteand target. In most cases, the site will be inen undeveloped location with a quality view. Thelocation should be reasonably accessible andsecure year-round. The monitoring view should beselected by personnel experienced in photographicexposure techniques and familiar with thepractical aspects end limitations of slidemicrodensitometry. The monitoring site endtarget should be selected so that as much of thesight path as possible runs through thewilderness.
The view must contain at least one horizonvisibility target with as many as possible of thefollowing characteristics:
1. Large--subtend at least 0.1 degree ofsolid angle (approximately 20% of the size of thefull moon.)
2. Easily identifiable on topographic mapsof the area.
3. Dark--preferably covered with coniferousvegetation.
4. Distance--preferably in the range of 40%to 60,% of the expected standard visual range.General guidelines are: 30 to 70 kilometers inthe western U.S.; 10 to 40 kilometers in theeastern U.S.
5. Number of targets--at least one qualitytarget is required; two or three targets atvarious distances are preferred.
6. Elevation angle--the site and targetshould be at approximately the same elevation.
Figure 6. --A closeup view of photographicvisibility monitoring enclosure (photocourtesy of Air Resource Specialists, Inc.).
The observer-target elevation angle should bewithin ±1°.
7. Targets should be located in the centerof the camera view finder (center 30% of theslide) .
8. For evaluation of regional air quality,the observer-target sight path should not beaffected by local sources of visual airpollution.
9. Target should be selected to be es freeof snow during the winter months as possible.Standard visual range values cannot be calculatedfor snow-covered targets.
10. Avoid exceptionally bright or darkforeground objects that would adversely affectthe camera’s ablility to accurately meter themonitoring view.
11. Sun angle--it is best to orient thetarget to avoid the sun shining directly into thelens.
System and Operation CostsSystem end operational costs depend on site
and sampling requirements. The approximateequipment cost for a single camera site, fullyoutfitted to include 35mm camera, 135mm lens, uvfilter, databack, programmable timer, batteries,environmental enclosure, internal locks,sunshield, monitoring hardware, mounting post,tripod head, cabling, documentation chart,instruction manuals, and lens cleaning supplies,would be about $2,100.
Operational costs depend on the samplingfrequency end analytical services. An averagecost of contracted services that includes allfilm, film processing, data analysis, reporting,end archiving is $5,250 per year (3 photographsper day, 365 days per year). For first-timesites, a one-time site initialization charge ofapproximately $1,000 is also charged to cover thecosts of preparing site specifications endperforming inherent contrast analyses. On-siteservicing by local personnel to change film andverify system performance is required every 10days, end on en average amounts to two to threeman-weeks per year including travel time.
It is also suggested that sufficient backupequipment be maintained to ensure continuousnetwork operations.
Field Service ProceduresRoutine operations and sampling.--Local
personnel will serve as the site operators, endwill be responsible for the routine operation ofthe camera systems. Automatic cameras will takethree photographs per day at 0900, 1200, end 1500local time. Kodachrome ASA 25 color slide filmwill be used. This film was chosen for its finegrain end excellent color reproductionqualities. For consistency, all film will bedeveloped at the Los Angeles Kodak laboratory.Photographs will be taken using the automaticexposure capabilities of the camera.
At many sites, access limits monitoring tosnow-free periods. A number of existing ForestService sites currently operate for limitedperiods, such as from late June or July throughSeptember.
Site visit/servicing protocols.--Film should
1 9
be changed every 10 or 11 days, based on threeshots per day and 36-exposure film rolls. A sitevisit by the field operator will generallyinclude the following:
1. General site/system inspection2. Remove camera; remove and replace film
(fill out ID label)
5.
3 .window
4 .
6 .7 .8 .
Inspect and clean camera lens and box
Check batteries and databackPhotograph film documentation boardReplace and align cameraCheck camera and timer settingsComplete Visibility Monitoring Status
Assessment Sheet9 . Close and lock camera shelter10. Mail film and Status Assessment SheetDetailed protocols and maintenance procedures
for camera systems have been applied throughoutexisting Forest Service networks for severalyears.
Collection, reduction end analysis ofphotographic data. --The major steps in thehandling of photographic data are summarized infigure 7.
All film collected at the sites must bemailed as soon as possible to a CentralProcessing Facility. All rolls will be loggedand forwarded to Kodak for processing. Allreturned slides will be identified by a site codeand consecutively numbered. Any missing orinconsistent samples will be noted and correctiveaction taken.
For qualitative analysis, the conditionobserved on each slide will be assigned anidentification code. These codes identifyweather conditions, observed hazes or plumes, andvisibility target illumination conditions.Appropriate qualitative summaries can be preparedfrom these codes. For example: In 20 percent ofdata the visibility target is obscured by clouds;layered hazes were observed on 30 percent of 0900observations.
The basis for quantitative analysis is themeasurement of the contrast (in the 550 nmwavelength) between and sky end selected terrainhorizon features. This contrast measurement canbe reduced to yield a standard visual range valuein kilometers. This quantitative measurement isrelated directly to the site path between theobserver and the target. Only the conditionswithin the path are quantified in this type ofanalysis .
Reporting. --The results of qualitative andquantitative analyses can be reported in avariety of formats. Most results will besummarized by monitoring season. Example reportproducts could include:
- Site specifications summary, including:site and target constantsdata, and data collection statistics
summaries.- Qualitative slide condition code summaryand statistics.- Slide and scene contrast listing for eachslide.- SVR listings for each day, time, and targetwas well as 0900. 1200. and 1500 dailyg e o m e t r i c m e a n S V R .
Figure 7. --Steps in the handling of photographicdata.
- Monthly plots of daily maximum, minimum,and geometric mean SVR.- Seasonal standard visual range summaries,statistics, and plots.All original slides will be archived by site
in a Central Processing Facility. Allqualitative and quantitative results will bearchived in both digital and hard copy formats.
Quality assurance. --All applied procedureswill follow fully documented quality assuranceprocedures. Procedures have been established andoperationally applied to account for: filmquality; film handling, processing, archiving,and storage; camera operation; scanning; datahandling; analysis; and reporting.
2 0
Transmissometer Measurement System
Transmissometers are a direct method ofmeasuring atmospheric extinction.Transmissometers consist of a constant-outputlight source transmitter and a computer-controlled photometer receiver. The twoindividually housed components must be separatedby a line of sight distance of approximately .5to 10 km, depending on the average extinctioncoefficient. The irradiance at the 550 nmwavelength from the transmitter can be determinedby measuring the light loss from the transmitterto the receiver. Data are collected on loggingsystems end strip chart recorders.
Several transmissometer installations areplanned for Forest Service wildernesses as partof the IMPROVE program. These initialexperimental installations will provide further .insights into the practical application oftransmissometry for monitoring visual air qualityin wildernesses. The initial experimentalsystems are costly to purchase end install.Power can be provided by solar panels at somesunny locations; line power will be required atthe receiver at many sites. Both ends of thetransmissometer must be serviced weekly by fieldoperators. Trained technicians must visit thesite to replace components and calibrate at leastevery six months.
Siting transmissometers near wildernesses maybe difficult since neither end of the system canbe installed in the wilderness. Sight paths mustgenerally be elevated to reduce the effect ofturbulence caused by surface heating on the lightbeam.
The advantages of the transmissometer arethat the system directly measures atmosphericextinction, both day and night. Continuousmeasurements can be averaged for selectedsampling periods. Disadvantages include highcost, power requirements, sheltering,installation, and servicing logistics. Ingeneral, it is recommended that camera systems beoperated along with the transmissometer tocorrelate measured extinction with visualconditions, end as a quality assurance reference.
References
Charlson, R. J.; Horvath, H.; Pueschel, R. F.1967. The direct measurement of atmosphericlight scattering coefficient for studies ofvisibility and pollution. AtmosphericEnvironment 1: 469.
Hoffer, T. E.; Schorran, D. E.; Farber, R. J.1982. Photography as a technique forstudying visual range. Science of the TotalEnvironment 23: 293-304.
Johnson, C. E.; Maim, W. C.; Persha, G.; Molenar,J. V.; Hein, J. R. 1985. Statisticalcomparisons between teleradiometer-derivedend slide-derived visibility parameters.Journal of the Air Pollution ControlAssociation 35: 1261-1265.
Joseph, D. B.; Maim, W. C.; Mestsa, J. C.;Pitchford, M. 1986. Plans for IMPROVE. APCASpecialty Conference on VisibilityProtection. Grand Teton National Park,Wyoming.
Maim, W. C.; Walther, E. G.; O’Dell, K.; Klein,M. 1981a. Visibility in the southwesternUnited States from summer 1978 to spring1979. Atmospheric Environment 15: 2031-2042.
Maim, W.; Kelley, K.; Molenar, J.; Daniel, T.1981b. Human perception of visualair quality(uniform haze). Atmospheric Environment 15:1875-1890.
Maim, W. C.; Moelnar, J. V. 1984. Visibilitymeasurements in national parks in the westernUnited States. APCA Journal 34(9).
Maim, W. C.; Tombach, I. 1986. Review oftechniques for measuring atmosphericextinction. APCA Specialty Conference onVisibility Protection. Grand Teton NationalPark, Wyoming.
Maim, W. C.; Persha, G.; Stoker, R.: Tree, R.;Tombach, I. 1986. Comparison of atmosphericextinction measurements made by atransmissometer, integrating nephelometer,and teleradiometer with natural andartificial black targets. APCA SpecialtyConference on Visibility Protection. GrandTeton National Park, Wyoming.
Steffens, C. 1949. Measurement of visibility byphotographic photometry. Ind. Eng. Chem. 41:2396 -2399 .
21
Soils and Geology
Soil functions in the ecosystem in roles thatare important to the productivity and diversityof the terrestrial end aquatic biota. Inaddition, it has a self-contained biota, and isan efficient trap or collection system for manyatmospheric contaminants. A careful descriptionand a set of quantitative measurements of thesoil are essential to estimating the sensitivityor stability of the ecosystem, end to determiningits response to atmospheric input. The goals ofthis section are to provide a guide in selectingareas to sample, and to suggest methods for usein the field and laboratory to accomplish thefollowing:
1. Characterize the soil-geologic resource.end evaluate its sensitivity to internal changeand its ability to buffer the aquatic system.This information will be used to assist inevaluating the susceptibility or vulnerability ofthe soils to change due to changes in airquality.
2. Determine the present condition of soilsin terms of pH. nutrient ions, metal load, etc.as a reference value to measure future changesagainst. This information will be used tomonitor the system for evidence of change viarepeated measurements.
One important factor that must be consideredis response time; while soil end geologicfeatures significantly influence the aquatic endterrestrial biological systems, their response toair quality change is likely to be slow anddifficult to measure in a time span of a fewyears.
List Of Measures
The assumed sensitivity of high elevationsystems is partially due to the expectation thatmuch of the area may be bare rock, the geologicmaterial is little weathered or perhaps resistantto weathering, end the soils are coarse endshallow, providing little buffer capacity. Areconnaissance survey of the geology is necessaryto focus the limited resources for sampling onsites that are most likely to be susceptible tochange. The soil physical end chemicalproperties recommended for measurement hereshould not be considered limiting. Othermeasures will be appropriate when the particulararea or air source suggests them. but thoseincluded in table 8 will be adequate andreasonable in cost for moat areas. If advancedanalytical techniques for soil extracts are used,the content of additional metals, for example,might be available at little or no extra cost.
Useful measurements such as bulk density andpermeability should be taken where soils containfew stones end allow the extraction of intactvolumes with coring devices. Such measurementsmay be prohibited by equipment needs or excessivetime requirements in most alpine areas. While
these additional physical measurements are usefulin characterization, they are not of the highestpriority since they are not likely to besensitive indicators of atmospheric changes.Bulk density concentrations in soil are to beconverted to mass-per-area basis.
Table 8 lists the measures to be used in boththe initial characterization and in the periodicsampling. The soil description, mineralogy,particle size analysis, end reconnaissancegeologic survey would not be repeated. Table 9lists the detection levels end laboratoryprecision of the measurements.
2 2
Requirements
Initial characterization of the geology endsoils, including writing descriptions, requiresthe services of two highly trained individuals--asoil scientist and a geologist--both with fieldexperience. Subsequent field sampling of soilscan be accomplished by technicians.Approximately one day per site is required forthe geologist and soil scientist, assuming thegoal is to characterize an area of less then 3km2. Subsequent periodic sampling can beaccomplished in one day per site.
Field equipment to be transported into thearea is listed below. All equipment needed forthe soil and geologic characterization andsampling is transportable on one pack animal.
Soil augerSampling tubeSpadeKnifeTopographic maps end aerial photographsField pH kitSoil color bookNotebookMeasuring tapePolypropylene sample bottles (1 L)Cloth end plastic bagsCompassAbney level
Soil coring device with removable ringsFrame. 20 cm X 20 cmCameraGround clothRock hemmer
Field
Site Selection ProcessCharacterization
The site selection
Procedure
and Geologic
process willresponsibility of the land manager,
be theand quite
specific for each region. We suggest ahierarchical approach working from very largeland areas, such as the entire wilderness area,down to selecting the landscape units (smallwatersheds, for example) that will be
23
characterized and sampled. All of the availableinformation from numerous disciplines should beintegrated in the landscape unit selectionprocess. Information on geology, soils,aquatics, vegetation, and general knowledge ofthe area, including public concerns and proximityto known or potential contamination sources, willbe useful.
After landscape units are selected, plots,lakes, end streams are selected for moreintensive characterization and long-termmonitoring. Each watershed or landscape unitselected is mapped for bedrock and surficialgeology on a reconnaissance scale. The soils aremapped at the same scale and differentiated atfamily or subgroup level, or mapped by soilassociations if recognized soil series areavailable. Since these mapping procedures aresomewhat subjective and require a trainedscientist following established procedures commonto the discipline, they are not included here.
Sampling StrategyConceptually, a few carefully chosen plots
can be considered, representative of sensitiveareas, even though they may not be trulyrepresentative of the entire area of concern. Afew plots thoroughly characterized for soil andvegetative properties can be monitored for changemore easily end result in higher probability ofdetecting adverse impact then a sampling schemedirected at whole watersheds or whole wildernessareas.
Selection of plots begins with a map survey(using aerial photographs) and the selection ofspecific landscape units thought to be the mostsensitive. The sampling plots themselves shouldbe representative of major soil-geologic-vegetative types within the landscape unitchosen. Whenever possible, the sample plotsshould be coordinated with those for aquatics endvegetation. In addition, the choice of locationshould also consider proximity to pollutionsources, public concerns, and area coverage.
Location of plots. --Due to size, diversity,and spatial variability, permanent plots will beestablished in selected geology-soil-vegetationassociations within major watersheds surroundinglakes and/or streams that are being sampled.This strategy will provide the opportunity formore precise characterization of smell,representative segments of the area under study,rather than attempting to characterize currentconditions throughout the area. Because of thelack of any prior classification scheme in which“associations” in these regions are recognized,on-the-spot classifications and descriptions willusually be necessary.
Size of plots. --Alpine mosaics, whetherrecognized es patches or es changes alonggradients, are usually finer grained than even atenth of a hectare. Therefore, plots will bynecessity often include more than one vegetation“type” in order to provide adequate size forrepeated sampling without destruction of thesite. Although some flexibility in size isnecessary to accommodate differences in fineness
and complexity of the soils-geologic-vegetativetypes, we recommend a target size of 1000 m2 .
Sample Location, Number, and FrequencyThe reconnaissance geologic survey maps and
aerial photographs will be used on-site to selectplots for detailed characterization. These plotswill be limited in number, depending on the sizeend complexity of the watershed or other geologicunit selected. The plots should be the same asthose used in the vegetation studies. Plotsshould be permanently marked at four corners withflush markers located on photographs by referenceto prominent landmarks.
For the initial characterization, a singlesoil pit is dug at the edge of the plot fordetailed descriptions, and approximately l-litersamples are collected, horizon by horizon, to adepth of 1.5 meters or to a limiting layer.Borings (which minimize disturbance) are takenaround the periphery of the plot to ensure thatthe pit is representative of most of the soilover the plot.
Because one goal is to measure temporaltrends, the emphasis is on surface soil samples.Deep samples are needed for completecharacterization of the site. Within eachpermanent plot, 12 samples each of the Ohorizon(s), if present, and top 2 cm of surfacemineral horizon are taken at each sampleinterval.
Samples taken to detect changes should becollected on a 5- to 10-year cycle. Morefrequent sampling is unlikely to show changes,even in severely contaminated areas.
Sample Collection ProcedureSample pH should be determined immediately in
the field by the dye technique (table 8).Organic horizons can be placed directly intobottles or plastic bags after separation of themineral soil. The O horizon samples should becollected from within a known surface area bycutting around the inside of a 20 cm X 20 cmframe. Moist mineral soil samples gently crushedand passed through a 2 mm stainless steel screenshould be placed in 1 L polypropylene bottles fortransport. Bottles should be permanentlynumbered. A record of bottle number, samplelocation, depth, date, and remarks should be keptand es much information as possible also recordedon the bottles. An estimate of percentage ofmateriel above 2 mm (that screened out) should berecorded. Samples of the coarse material shouldbe taken in cloth bags for mineralogicalanalysis. Initial samples taken from the pitshould include undisturbed core samples,representative of each horizon, for bulk density,pore space, end permeability tests.
Field Storage and HandlingSoil samples for pH. extractable sulfate, end
nitrogen should be maintained as moist and coolas possible until they reach the laboratory,where the sample can be split. An aliquot forthe above analyses is stored at 4°C end theremainder is air-dried for physical end chemical
24
analyses. Air-dried samples can be storedindefinitely at room temperature.
Hydrologic SamplingIn remote areas, the sampling of springs and
seeps offers the best approach to sampling soilwater. Sites are permanently marked forsequential sampling, and subjected to analysesprescribed for lakes and streams. Alternatively,tube lysimeters could be installed for periodicsampling if groundwater measurements areessential, but the procedures are not describedhere.
Laboratory Analysis
Only the initial samples will be subjected to
bulk density, permeability, mineralogical, andtextural classification.
The characterization samples and the surfacehorizons collected repeatedly should be analyzedfor the chemical properties shown in table 8.The methods to be used are found in thereferences cited in table 8 and listed in table10.
Support Needs
Because the soil sampling should support thevegetation analyses and the aquaticcharacterization. their locations must becoordinated.
Aquat ic Chemis t ry
Purpose
The objective of this section is to provideguidance for determining current chemicalcharacteristics of surface waters in alpine andsubalpine wilderness areas. The proposedguidelines are essentially limited to sampleacquisition, stabilization, and analysis.
Development of a detailed sampling programfor application in all potential study areas isnot feasible. Area-specific information,including the expected nature of potentialimpacts, and the spatial and temporal variationin measurable response parameters, must beconsidered. Although design issues cannot bepre-specified, in general a two-stage strategy isrecommended for determining current chemicalcharacteristics of surface waters. Stage I woulddetermine the presence end spatial distributionof sensitive surface water systems. The level ofeffort required at Stage I will depend on theamount of existing information. Stage II wouldthen involve the selection of sensitive system(s)for more intensive study end longer-termmonitoring. If information on historical ratesof deposition is required, sediment coring couldbe a component of Stage II, and would requireadditional protocols.
The following protocol is proposed as apractical approach to obtaining an initialcharacterization of current chemistry of lakesand streams in remote high-elevation wildernessareas. In conjunction with a sampling designappropriate to local conditions, these attributesand this sampling and analysis methodology shouldmeet a Stage I objective of establishing therange and distribution of chemically differingaquatic systems. A sufficiently reliable basiswould be provided for establishing classes ofaquatic systems according to sensitivity ofresponse types. Depending upon further
definition of objectives, a more rigoroussampling and analysis protocol could beimplemented for the Stage II sampling program toidentify temporal variation end trends.
Land managers may view the methods ofinvestigation described within this protocol asgeneral guidance rather than detailedrequirements. Methods actually employed woulddepend upon the potential AQRV impacts, theresources available for the task ofcharacterization, and the specific managementobjectives.
List of Measures
Major Ions in Water (including Al)Table 11 provides a general list of
measurements important for characterizing surfacewater composition. The list includes the majormineral species commonly present in low ionicstrength surface waters, and the basic parametersassociated with nutrient status and biologicalproductivity. The list of measurements may beconsidered optimal, though not all-inclusive,depending upon case-specific conditions. Someconstituents, including aluminum fractions,dissolved organic carbon, fluoride, and ammonium,are commonly present at very low concentrations.After confirmation of low concentrations, thesemeasurements could be deleted, or obtained on aless frequent basis then other constituents.Dissolved oxygen and transparency measurementswould only be obtained when mid-lake sampling isconducted.
Table 11 also lists the recommended methodsof sample analysis. Accuracy end precision goalsare listed in table 12. Sample container andpreservation requirements,are listed in table13. Sample holding times are listed in table14. Specific analytical procedures are providedby reference to the National Surface Water Survey
2 5
(NSWS) , a regional-scale survey of stream andlake chemistry conducted by the U.S. EPA. TheNSWS provides the most credible endwell-documented protocol for determination ofchemistry in natural, low ionic strength waters.This project includes a Western Lake Surveyconducted in cooperation with the ForestService. The results of this work should providean improved perspective on sampling and analysismethods as well as on the degree of precision endaccuracy attainable.
Of specific interest is the comparative studyof pH measurements, including the development anduse of a closed system pH determination.Preliminary results favor this method based onprecision and accuracy levels. Inclusion of thismethod in the proposed protocol would eliminatethe need for in situ pH determinations, whichwould greatly expedite the sample collectionprocess. The NSWS is ongoing, with importantmethodological findings still emerging. Themethods proposed may not fully reflect thecurrent information status; the proposed protocolshould be considered tentative and subject tomodification as results of the NSWS end othercurrent studies supporting methods assessmentbecome available.
The recommendation for use of NSWS methodsshould be viewed se guidance rather thanprescribed requirements. Any alternate methodsemployed should attain equivalent levels of
precision and accuracy, or be otherwisejustified. Detailed documentation of any methodsused should be considered a necessity.
Trace Metals in SedimentsTrace metals in sediments can provide
historical deposition records. Studies in RockyMountain National Park and in the Wind RiverRange indicate a history of atmospheric input ofsome heavy metals (most notably Pb) over a+100-year period. Collecting this record is aviable research goal to determine the history ofthe input of atmospheric pollutants that arepreserved (not necessarily in proportion toatmospheric flux) in lake sediment or peat.(There are probably few ombrotrophic bogssuitable for such studies.) Briefly, we suggestthe following in conjunction with Stage IIsampling if information is needed on historicaldeposition rates.
1. Coring of lake sediment from selectedlakes.
2. Abbreviated analysis of sediments toestablish approximate chronology and comparepre-1800 to modern chemical-biologicalcharacteristics.
3.Sample intervals: 1-2, 5-6, 10-11,15-16,20-21, 25-26, 30-31, 35-36, 40-41 cm.
4. Chemical parameters: 210Pb activity,
H20 and organic content, major metals (Cd, Mg,K, Na, Al, Fe, Mn), trace metals (Pb, Zn, Cd, Cu,B), polycyclic aromatic hydrocarbons, charcoal,end soot.
5. Biological parameters: diatoms,chrysophytes, end pollen.
It will also be possible to evaluate the fluxof anthropogenic material (atmosphericpollutants) to the sediments.
Coring of lake sediment is difficult, end iseven more difficult given the constraints imposedin wilderness areas. If sediment coring isdeemed necessary, a well-developed protocol willbe necessary.
Trace Metals in WaterWe recommend that only labile end total
aluminum be determined in the surveys becausealuminum is biologically important, it is verysensitive to changes in pH. and natural levelscan be measured with fair precision and accuracy.
We do not recommend that other metals bemeasured because none have been identified ashaving effects on biotic systems at naturallevels, natural variability is high end thereforetrends will be difficult to discern, and tracemetals are difficult (and expensive) to collectand analyze at natural levels.
Requirements
ManpowerThe level of effort required to collect the
recommended samples has two components: 1)traveling to the sampling site. and 2) samplingthe water end sediment. The amount of timerequired to get to the sampling sites will besite-specific and cannot be prescribed except tosay that enough travel time should be allocatedso that field operators can perform careful workat the sampling site.
The amount of time required to take the watersamples is approximately 0.5 day, althoughindividual sites may require longer. To takesediment cores, en additional 0.5 - 1 day isnecessary.
EquipmentBackpacksSampling containers for each location: 1
1000-ml LDPE (lowdensity polyethylene) bottle, 1 125-ml
LDPE bottle, and 2 60-mlsyringes
pH and conductivity metersIce and coolers for samplesRaft when taking mid-lake samplesVan Dorn bottle, dissolved oxygen meter, end
Secchi disk when taking mid-lake samples
27
Field Procedures
Sampling StrategyStage I in a two-stage sampling strategy
would be conducted to determine the distributionof sensitive lake and stream systems. The levelof effort required at this stage will depend uponthe value of existing data. Existing data wouldbe adequate if determined sufficient to identifythe most sensitive landscape units and place thewatersheds in sensitivity classes. synopticsurvey data should be obtained that reflect thespatial scale and distribution of responsecontrolling landscapes. Records of existingsurface water chemistry, as well as maps ofsurficial materials (soils and geology) endvegetation, should be employed in identifyinglandscape units and in determining additionaldata collection needs. Even where surface waterscan be lumped into general sensitivityclassifications, a more detailed survey focusedon individual classes may be useful foridentifying the most sensitive class members.
Stage II would require selection of the mostsensitive class, or classes, of lakes and streamsfor monitoring and more intensive study.Ideally, one small, well-defined watershed wouldprovide an excellent study area for measurementsof aquatic chemistry end biology, end alsovegetation end soils. These measurements areeasier to perform, interpret, and subsequentlymonitor when they are made within the same knownwatershed.
Location of Sampling Sites at Lakes and StreamsIf mid-lake samples are to be taken, lakes
should be sampled 1.5 meters below their surfacein the middle of the lake. Samples of the outletand a major inlet also should be sampled at alocation with appreciable water flow (i.e., nostagnant pools).
Streams should be sampled mid-stream in areasof appreciable flow.
Number of Water Samples end Frequency of SamplingTwo to four aliquots should be taken at each
site, depending upon the suite of analyses. Eachaliquot has its own preservation end treatmentprotocol (table 13). This sampling protocolgenerally is adopted from the NSWS. Our protocoldiffers from the NSWS protocol in that it doesnot recommend filtering (because of contaminationproblems end the lack of large amounts ofsuspended particulate in the high elevationsurface waters) end sampling treatment isminimized. This recommendation does not precludefiltration warranted by site conditions (e.g.,high turbidity) or emphasis on a specificparameter (e.g., concentration of dissolvedphosphate).
At least 10 percent of all sampling andanalysis should be done in duplicate to providean indication of the uncertainty associated withthe sampling and analysis procedures. Additionalquality assurance protocols should be used asdescribed in the quality assurance plansdeveloped for the NSWS (see Drouse et al.1986). At a minimum, quality assurance should
include analysis of sampling replicates, blanks,and NBS traceable reference standards.
Sample frequency should include 3-4 samplesper year taken between early spring end earlyautumn. The exact date of the first sample willdepend upon field conditions.
Water Sample Collection ProceduresThe lake and stream water sampling procedures
of the NSWS (1986) should be used. In general.,lake and stream samples should be collected withLDPE sampling containers that have beenacid-washed end copiously rinsed with deionizedwater. At the sampling site, the bottles end thecaps should be rinsed 3 times with the lake orstream water before taking the sample. Thesyringes should be rinsed three times prior tosample collection. Samples should be placed onice immediately after collection.
Field MeasurementsWe recommend that specific conductance
measurements be taken each time samples arecollected. Dissolved oxygen and transparencywould be measured on site when mid-lake samplingis conducted. In situ pH measurement (pHl intable 11) can b taken if the closed systemmeasurement (PH3) method is not used. Werecommend closed system pH measurement because ofthe low precision associated with in situ opensystem measurements. If in situ pH1 openmeasured in lieu of the closed system pH3 , werecommend that en open system lab measurement(PH2 ) also be made.
Water Sample Storage and TransportImmediately after collection, the samples
should be packed in an insulated container with arefrigerant, and thereafter maintained in thedark at approximately 4°C until analyzed.
Laboratory Sample Analyses
The recommended methods for analysis of watersamples are listed in table 11. (Note precedingstatement concerning use of alternate methods.)These methods are described by Hillman et al.( 1986). The appropriate units of measure arelisted in table 12.
Support Needs
No biological measurements are required tosupport the chemical measurements.
In addition to the physical measurementsalready mentioned (e.g., transparency),additional physical measurements are watertemperature and, in the case of streams, waterflow .
Data Analysis
Data analysis that would be required inaddition to standard statistical analyses todetermine data quality, depend on the objectiveof the study and how well it is realized in thesampling design and, thus, cannot be prescribedhere.
28
Aquatic Biology
Purpose
The primary objective of monitoring aquaticbiota in wilderness areas is to provideinformation on the status of sensitive biologicalcommunities over time. Men-induced change orlack of change can then be inferred from datapatterns. This general goal must be consideredwithin the context of limited monitoringresources (constraints of time, money, access tosampling sites) which, for example, may precludethe study of seasonal biological dynamics.Cause-and-effect relationships betweenanthropogenic disturbances end biologicalresponses also cannot be delineated (except forcatastrophic change) in complex ecologicalsystems by monitoring alone. Moreover, thesensi- tivities of specific aquatic biota todifferent anthropogenic stresses generally arepoorly known in subalpine and alpine regions.Consequently, the monitoring activities describedhere represent our best approximation of aminimal array of sensitive components that, ifmonitored with reasonable intensity andfrequency, will provide an estimate of the healthof aquatic communities in remote wildernessregions.
Although sampling error undoubtedly will behigh (i.e., high variance among replicate endtime series data), we have little alternative toaccepting high error short of simplepresence/absence surveys or no monitoring atall. Although present monitoring methods may becrude, it is obvious that the alternatives areundesirable from a resource managementperspective.
The biological components that constitute aminimal set for detecting change in alpine andsubalpine waters are Chlorophyll a, salmonidfisheries, and macroinvertebrates. Chlorophyll a(tables 11 end 12, Aquatic Chemistry) is the bestreadily measurable attribute of phytoplanktonbiomass (primary producers) related to thetrophic status of surface waters, which may bestimulated (fertilized) or reduced by atmosphericdeposition. Macroinvertebratea end salmonid fishare sensitive to many types of anthropogenicdisturbances; are not typically ephemeral in agiven lake or stream (although specificlife-cycle periods are ephemeral): can bequantitatively monitored by routine fieldpractices: and in the case of fish, are highlyvalued components of wilderness surface waters.
Monitoring of other aspects of lower trophiclevels (such as phytoplankton end zooplanktoncommunity structure) presently is not recommendedas part of the minimal set for the followingreasons:
1. these lower trophic levels have highnatural variability, both seasonally andyear-to-year;
2. we do not understand the complex offactors that drives the shifts in community
structure or biomass during early stages ofanthropogenic disturbance (e.g., lake and streamacidification) especially in low-productivitysystems;
3. they have low or unknown sensitivity tochange during early stages of anthropogenicdisturbance; and
4. archiving of biological samples forfuture analysis is not likely to be feasible.
However, anthropogenic stress affects a broadvariety of aquatic organisms. The wildernessresource manager should consider expanding theminimal monitoring program suggested in thisprotocol as resources permit. Studies that maybe particularly useful include studies ofattached algae, macrophytes, zooplankton, andamphibians (particularly salamanders).
As previously noted, these protocols areintended to be general guidelines rather thandetailed field manuals or research directives.Although we recognize that alternative monitoringapproaches such as remote sensing of the trophicstatus of wilderness lakes have great potentialto increase the accuracy and geographic extent ofthe monitoring effort, such developmentalapproaches are not discussed here.
Salmonid Fish
The following protocol is constrained bylimitations of access, time, manpower, andtransport of equipment inherent in sampling ofhigh-elevation lakes in remote wildernessregions. Such limitations dictate that only asmall set of basic data be collected tocharacterize fish stocks. Further assumptionsused to develop this protocol are described inthe following paragraphs.
The general objective of this protocol is tocorrelate independent fishery variables withchanges in surface water quality. However,quantitative assessment of fishery stocks inremote lakes for long-term trend analysis is notwell developed. Most monitoring ofhigh-elevation fisheries has been biased towardgeneral management goals that do not require ahigh degree of accuracy or precision oftechnique. For example, the efficiency ofsampling effort using specific gear is a functionof fish species, standing stock, seasonalbehavior, habitat, and morphometric features oflakes, end is not quantified for high-elevationsalmonid fisheries. Little guidance on thequantification of alpine salmonid fish stocks canbe derived from existing data. Consequently,collection of unbiased fishery data in wildernesslakes is unlikely (Thornton et al. 1986).
Because the availability of sampling gear ateach lake will be limited, sampling for targetfish species will be emphasized. This maypreclude complete characterization of the fishcommunity in some lakes. However, many alpineend subalpine lakes did not historically contain
29
fish, and existing fish populations have beenestablished by stocking.
Therefore, many high-altitude fisheriescontain only one or two species of introducedsalmonids. For example, only a few cutthroatpopulations were found historically in thehigh-elevation lakes of the Wind River Mountainsin Wyoming. Other fish species associated witholigotrophic conditions are not commonly found.For example, sculpins, suckers, and date are notfound in high-elevation lakes in Wyoming due tohigh gradient streams, although sculpins areoccasionally found in mid-elevation lakes.Speckled date end long-nose suckers are found insome alpine lakes in Colorado but theiroccurrence is not common; both of these specieshave probably been introduced.
The lakes selected for long-term monitoringshould, if possible, also be used for chemicalmonitoring (see Aquatic Chemistry section).However, monitored lakes must be capable ofsustaining fisheries over periods of many years(e.g., probability of winter kill must be low)and the lakes must not exhibit a high degree ofheterogeneity of fish habitats (e.g., selectedlakes should be relatively circular and notcontain coves) which tend to developsub-populations of fish. Lake morphometry shouldbe relatively uniform, conducive to randomdispersal of active fish populations. Harvest byangling should be insignificant compared tonatural sources of mortality, end be relativelyconstant year to year. The potential forover-harvesting by fishermen must be consideredduring lake selection.
Ideally, lake fisheries that are monitoredshould be sustained by natural reproductionbecause early life stages of salmonids are verysensitive to changes in water quality. This maybe an impractical constraint, however. becausemany lake fisheries in alpine wilderness regionsare maintained by periodic stocking. Inaddition, the availability of spawning habitathas been found to be positively correlated withpopulation strength and size of individual fishin wilderness lakes (Hudelson et al. 1980).
Salmonid fisheries are assumed to be ofprimary interest. Such fisheries are to besampled once during a sampling year over a 2-3day period using equipment that can betransported to the site. The field crew shouldconsist of 2-3 individuals using an inflatableraft. A requirement for non-destructive samplingin wilderness lakes will be adhered to es much aspossible, but complete elimination of samplingmortality is difficult.
Experience has shown that absolute measuresof fish stocks are difficult and time-consumingin the alpine (e.g., mark-recapture techniquesrequire at least one week of sampling). Thus,the specific objectives here are to quantifyrelative indices of fishery status. Theseinclude (beyond presence of a specific fishery)the following:
1. catch per unit effort (CPUE),2. population age structure,3. condition factors,4. growth and mortality rates, end5. absence of year classes or week year
classes.
Field SamplingSampling design.--Collection of
representative, random samples of individualsfrom a fish stock optimally should be based on astratified random sampling design where stratarepresent different habitat types. In alpinelakes that have restricted sampling area (due tosteep morphometry, boulder fields, or othermorphometric features), site-specific judgementon net placement may be based on experience withcollecting mobile salmonids if experiencedictates that a representative sample of fishfrom the total stock will be collected.
Sampling frequency.--It is generallyrecommended that the frequency of fish samplingbe related to the potential rate of change ofsurface water chemistry (Lambou et al. 1985).For intensively monitored lakes where closeobservation of fishery status is desired in theevent that such lakes undergo rapid change (e.g.,dilute headwater lakes), fish should be sampledat 1- to 2-year intervals. For lakes notexpected to exhibit rapid chemical change (e.g.,larger or less dilute subalpine lakes), fishshould be sampled at 3- to 4-year intervals(Lambou et al. 1985). Fish should be sampledonce during mid to late summer so thatyoung-of-year may be observed. This time willvary from late July to late August, dependingupon whether fish spawn in fall or spring.
Sampling intensity. --Sampling intensityshould involve both minimal sampling effort andminimal sample sizes. Minimal sampling effortshould be expended at each monitored lakeaccording to the recommended number of net setsfor lakes of given sizes (see Net Placement).Minimal sample size should be collected accordingto the following protocols.
For each monitored species, 100-15O fishshould be collected, with 150 being preferred.Some field experience suggests, however, that asfew as 30 captured individuals may be adequate tocharacterize fisheries with limited stock size(Remmick 1984). Other recommendations includesampling at least 10 fish per 2 cm size lengthover the size range of maximum accuracy forindividual fish statistics (see discussionbelow) . Lambou et al. (1985) recommend that atleast 60 fish evenly distributed across sizeclasses be measured for developing fisherystatistics (see also Thornton et al. 1986).
It should also be noted that stressed fishpopulations contain the fewest individuals andrequire the greatest effort to achieve populationestimates of known variance.
Field ProceduresSampling gear. --Because collecting unbiased
fishery data is difficult, more than one type ofsampling gear should be used. However,experience has shown that monofilament gill netsare effective in collecting most fish speciesfound in remote Rocky Mountain lakes (e.g.,Hudelson et al. 1980) and are easilytransported. Trap nets are presently beingdesigned that are more portable than previouslyavailable (e.g., modified Alaska trap net; ALSC1985). Such nets may become useful in the futureto supplement gill netting, although portabletrap nets did not prove effective for surveying
30
3
fish stocks in small Maine lakes (Haines et al.1985) .
Swedish gill nets (standard 150 ft length, 6ft depth with 5 panels of 1/2, 3/4, 1, 1 1/2 inchbar mesh) should be used because they areespecially effective in capturing mobile salmonidspecies. These nets are capable of capturingfish es small as 7-8 cm total length. Thus, thegill netting should be effective on age I fishand older, but should not catch young-of-year.
Net placement. --Analysis of trends in fisherystatus based upon results of gill netting willonly be as reliable as the reproducibility ofsampling technique for each monitored lake.Location of nets, orientation along the bottom inrelation to shoreline, time of placement endcollection, and season of placement must bestandardized for each lake. Because lakesampling programs will be site specific,standardization must be within a given lake endnot necessarily between lakes.
Fish captured by nets operated in a similarmanner and time each sampling year should providereasonably comparable estimates of populationcharacteristics (Hubert 1983). Close adherenceto standard collection procedures for each lakewill minimize total variance in catch statisticsdue to unknown or uncontrollable biotic endabiotic factors.
Each lake has a unique morphometry, and netplacement must be carefully considered accordingto lake characteristics and target species.Generally, gill nets set along the bottom inshallow waters not exceeding 5-7 m depth willcapture a representative sample of the total fishstock if crepuscular activity periods aresampled. Salmonids in alpine lakes in the RockyMountain region are typically found in relativelyshallow waters, end in general are closelyassociated with the benthos as the primary foodresource (e.g., golden trout are found in closeproximity to sediment). Nets should be placedperpendicular to the shoreline in shallow wateror at 45° angles in deep water, with the smallmesh nearest shoreline. If the initial samplingeffort yields few or no fish, the samplingstations should be moved and the sampling effortrepeated.
A rough guideline for number of nets to useis as follows:
Number of 150ft. Swedish gill
Lake size nets required
Less than 10 acres 110-25 acres 225-50 acres50-100 acres 4Each additional 100 acres add 1 net
Three gill nets set in different habitat maybe the maximum number effectively operated by afield crew of 2-3 persona.
Sampling mortality should be minimized inwilderness lakes, especially in those withrelatively small fish populations. Thus, gillnets should be deployed by midday in
high-elevation lakes where foraging by salmonidsis more or less continuous. These nets should betended every 1 1/2 to 2 hours to minimize capturemortality. The nets should be run until afterdusk to bracket one crepuscular activity period.If fish are abundant, and some sampling mortalityis acceptable, one net set over night ❑ ay beuseful to sample larger, night-feedingindividuals. Overnight sets where the dusk (onehour before sunset) end dawn (one hour aftersunrise) activity periods are bracketed generallywill capture a representative sample of fish, butsuch long sets may produce mortalities above 50%in captured fish.
Fish processing. --Fish in unproductivehigh-elevation lakes generally grow very slowlyin the older age classes due to food limitation.Generally, rapid growth occurs in such lakes onlyto approximately 20-25 cm in total length. Thus ,older fish are more difficult to age accuratelyby scale readings. Obtaining accurate weights inthe field on fish smaller then 12-15 cm also isdifficult due to variations in water content endfrom wind effects on weighing devices.Additionally, the more numerous individuals inthe younger age classes place large demands on alake ecosystem to provide habitat end food.Therefore, it is appropriate to obtainweight-length measurements for all fish captured,unless sub-sampling is required due to largenumbers of individuals captured. But fish in the12-25 cm total length range only should be usedto determine the population parameters discussedbelow.
Handling time should be minimized once fishare removed from the net. Fish should be kept ina live-car attached to the side of the raftduring handling, end be released as soon aftercapture and measurement as possible. Handlingmortality should be recorded if observed.Procedures for reducing handling mortality havebeen reviewed by Stickney (1983).
Fish collected should be carefully removedfrom the gill net (using a small polished hook tominimize damage to fish end technician),identified to species, measured to nearest mm,weighed to nearest gm (by volume displacement forfish less then 50 gm end by spring scale for fishgreater then 50 gin), and scale samples taken (onleft side just posterior to and below dorsal finend above the lateral line).
Recording of field data.--Field datarecording should be standardized, and include thefollowing:
- lake;- sampling date;- gear type:- net location, shoreline orientation,
depth, placement time, and collectionintervals;
- species, weight (gin), total length (cm),and location of scale collection sitefor each individual fish collected;
- observations of parasites, wounds,deformities, or other abnormalities; and
- capture mortality end injury.Lake temperature at sampling location and other
31
pertinent chemical condition information such asdissolved oxygen or pH should be recorded onfield sheets, if measured.
Additional fish surveys.--A reconnaissance-level, qualitative assessment of reproductivesuccess can be made by looking for juvenile fishin shallow habitat end in shoreline cover. Smallhand seines and dip nets may be used to findyoung-of-year, but capture success may belimited. Small baited hardware cloth minnowtraps have not proved effective in capturingyoung-of-year salmonids (T. Haines, personalcommunication ). Alternatively, trapping ofdrifting fry in outlet streams during samplingfor macroinvertebrate drift may be possible.
Laboratory ProceduresUse of scales to age individuals from wild
salmonid populations during the early rapidgrowth phases of ages I-IV is the most accurate,non-destructive technique available. However,aging by scale analysis is not as accurate forstocked fish due to possible interruption ofgrowth in the period immediately followingstocking. Stocking may produce a growth checksimilar to en annulus. Fall-spawning speciessuch as lake trout produce a better annulusduring age O in comparison to species that spawnin late spring such as cutthroat and goldentrout. The latter sometimes do not get a goodscale growth before the first winter, and produceen indistinct first annulus.
Technicians should be carefully trained usingscales from the monitored species that are fromfish of known age end comparable growthenvironment. Random recounting of approximately5% of fish scales by a second trained technicianis appropriate es a quality assurance check. An80-90% comparison of scale readings between twotechnicians typically is good (Thornton et al.1986) . If possible, fish scales should be agedby the same personnel over the life of themonitoring effort to minimize error.
Supporting DataAs noted above, the fisheries monitored
should be in lakes whose water chemistry is alsobeing monitored. Better resolution of ongoingenvironmental change or stability will resultfrom such integrated studies.
Data AnalysisCatch Per Unit Effort (CPUE) .--Catch per unit
effort is a relative measure of populationstrength. Theoretically, it should be linearlyproportional to the abundance of fish stock:catch = capture efficiency x fishing effort xfish abundance. However, capture by passivefishing techniques is a function of fishmovement. Consequently, CPUE is not dependentfully on stock size (Hubert 1983), end hasfrequently been found to be non-linear (Bannerotand Austin 1983). Fishery biologists have longrecognized variability in CPUE results fromproblems of gear efficiency due to interactingbiotic end abiotic factors that affect fishmovement.
Common problems with CPUE studies includespatial correlation among sampling units, inverse
non-linear relationships between captureefficiency and population abundance, end skewedfrequency distributions for CPUE with zero catchbeing most frequently recorded. (See review inThornton et al. 1986) . The high variabilityamong units of sampling effort (catch per nethour or per net night) may result in poorstatistical resolution of stock means or patternsof population fluctuations (Bannerot and Austin1983, Thornton et al. 1986).CPUE will therefore provide onlysemi-quantitative estimates of fish abundance.
Transformed catch data and the relativefrequency of zero CPUE have been demonstrated tobe the best indicators of population abundance.Catch data should be reported as catch per nethour or per net night (means end variances) foreach species captured (Bannerot and Austin 1983).
Population age structure.--Evaluation ofpopulation age structure depends on obtaining arepresentative sample of the overall population.As noted above, passive fishing techniques usinggill nets tend to produce skewed data, with olderend younger fish being less efficientlycaptured. Additionally, aging of fish by readingof scales is most accurate in the age I-IVclasses for alpine salmonid fisheries. Thus ,age-frequency data for y-o-y end older ageclasses generally will be qualitative. Agefrequency distribution within the I-IV ageclasses will be most quantitative.
Captured fish should be aged by scalereadings using standard techniques (Lagler 1956,Jerald 1983). As with CPUE, age frequency datamay be transformed to achieve independence ofvariance and mean.
Condition factors. --Weight and length arequantitative attributes of individual fish thatcan be easily measured in the field.Relationships between weights end lengthsindicate the relative abundance of food endrelative quality of habitat for growth.Condition factors for each age class I-IV shouldbe calculated. For example,
CF = (W X 105) / L3
where: CF is condition factorW is weight in poundsL is total length in inches
Condition factors are typically reported inEnglish units. For comparison with existing datafrom State game end fish departments, it probablyis best to continue using English units (Remmick1984.)
Relative condition factors or analysis ofcovariance among weight/length data for specificsubgroups captured (for example, groupedaccording to same species, sex, year-class, andphysiological condition relative to spawning)also may be used to assess the general growthenvironment (Anderson and Gutreuter 1983: W.Nelson, Colorado Div. of Fish and Wildlife,personal communication).
Growth and mortality rates.--If sampling hasbeen random among age classes, determination ofgrowth end mortality characteristics will revealimportant characteristics of the fish stock for aparticular lake. The dependent variables most
32
commonly related to water and habitat qualityinclude average annual growth increment,instantaneous growth rate, average length of agiven age-class, and instantaneous age-classmortality. Again, age classes I-IV should beused to provide the most accurate calculations.
The variable stocking rates of differentage-classes in some wilderness lakes will affectaccuracy when estimating fish mortality fromcatch data. Thus, the mathematical structureused to estimate mortality will depend upon thespecific conditions related to annual recruitment(Everhardt and Youngs 1981). For those lakefisheries not sustained by natural reproduction,knowledge of stocking rates (species, numbers,sizes, dates) should be obtained from appropriatefishery management groups. Adjustment of catchdata by weighting according to stocking recordsshould follow general recommendations inEverhardt and Youngs (1981). If stocked lakesare extensively monitored, marking of eachstocked year-class is feasible and should beconsidered to improve the accuracy of populationparameter estimates.
Average annual growth increment betweengrowth intervals of 1-2 cm can be determined byback-calculation techniques using length at agei, determined by scale analysis (Lagler 1956,Whitney end Carlander 1956, Carlander 1981).Fish growth rates also should be calculated usinglength-weight regression assuming allometricgrowth: W = aLb , where a and b are growthcoefficients.
Instantaneous growth rate is calculated asthe difference between natural logarithms ofweight for consecutive age groups (Everhardt andYoungs 1981). If annual fish surveys areconducted, instantaneous age-class mortality (Z)may be calculated from the slope of theregression of age versus frequency:
Ni = Ni-1 ez where Ni is numberof individuals captured of ith age for aparticular year-class.
Missing or weak year classes. --Observationsof missing or weak year classes in a fish stockmay indicate changing habitat conditions ordensity-independent mortality, resultingprimarily from weather. Missing or weak yearclasses are common occurrences in wild fishpopulations: gear inefficiency and poor agedeterminations preclude accurate estimates ofolder age class strength. Thus, for a givenlake, the occurrence of missing or week yearclasses in catch data is neither a definitivecharacteristic of the fishery, nor necessarily enindicator of anthropogenic change related toatmospheric deposition. However, patterns ofsimilar age-frequency distributions (number andsequence of low frequency year classes) amongmonitored lakes in the same region may beindicative of regional conditions or change(Thornton et al. 1986).
Detection GoalsPresently, there are no standard references
that quantitatively define detection goals forchanges in the indices discussed forhigh-elevation salmonid fisheries. Comparableanalyses of fishery data from remote lakes using
stock assessment by gill netting are beingconducted by the U.S. Environmental ProtectionAgency as part of the National Surface WaterSurvey. Results of these analyses, which areexpected in several years, will help definerealistic detection goals for some salmonidspecies with restricted stock sizes and forlimited sampling.
It must be stressed that, due to inherentvariability in gear efficiency between lakes,fishery statistic cannot be compared betweendifferent monitored lakes, but only within agiven lake over a sequence of years.
Based upon reasonable age determination of 1-to 4 year-old fish, changes may be detected inrelative growth and mortality indices on theorder of 20-50% for specific year-classes, butachievement of these goals is uncertain.Estimates of changes in absolute characteristicsof an alpine fish stock will be less precise.Changes in the neighborhood of only 2-3x will beconsidered valid on the basis of reasonableadherence to assumptions of random sampling offish stock end independence of variances andmeans. Estimates of age will be particularlytroublesome. In general, aging by scale readingproduces estimates of higher mortality rates thanactually are present due to poor aging of olderfish (on the order of 10-20% too high; Jerald1983) . Precision of age determinations should beestimated according to Chang (1982).
Additional protocols for sampling fish may befound in Armour et al. (1983) and from the EPANational Surface Water Survey (Fabrizio et al.1987. Hagley et al. 1987).
Macroinvertebrates
Aquatic macroinvertebrates are animalswithout backbones that live in streams end lakes,and are big enough to be seen without amicroscope when in advanced stages ofdevelopment. They have been observed to besensitive to low pH conditions in lakes endstreams (Napier and Hummon 1976; Parsons 1968:Warner 1973; Nichols and Bulow 1973; Tomkiewiczand Dunson 1977; Witters et al. 1984; Havas andHutchinson 1982; Bell 1971; Hall and Ide 1987;Hall and Likens 1985; Singer 1981, 1984).Tolerances of aquatic invertebrate species varyaccording to their specific anatomical,behavioral, end physiological adaptations (Hynes1972 ) . Since some macroinvertebrates are moretolerant to acid conditions then others (Parsons1968, Warner 1973, Bell 1971, Robak 1974, Eilerset al. 1984, Hall and Likens 1980, Sutcliffe andCarrick 1973), they may be used as a functionalpart of the warning system established to monitorpossible effects of air pollution inhigh-elevation ecosystems.
The following equipment and procedures arebeing used by federal and State agencies inwestern regions of the United States, and mayprovide a common basis for collection ofmacroinvertebrate data.
EquipmentModified Surber net (see fig. 8)Standard 8“ diam. full height 250 micron mesh
Tyler sieve
3 3
Figure 8.--Modified Surber sampler for use inmacroinvertebrate sample collection (fromUSDA 1985).
8 OZ., wide-mouth plastic bottles with stripof masking tape attached
for identification dataPreserving solution: ethyl alcohol plus one
cup 10% formalin pergallon of 70% alcohol
Hip bootsSaturated salt-water solutionTwo aluminum bread pansWaterproof gloves (scuba diving or rubber
electrical typesrecommended )
Laundry pen (waterproof marker)Long-handled 250 micron mesh kick netWhite plastic tray--2.5” x 10” x 12”Fine-pointed forceps
Sampling Station and Site SelectionStations should be established in the inlet
and outlet streams of the lake ecosystem to bemonitored. In a literature review, Eilers et al.(1984) observed that, with few exceptions, taxawere found in lakes at lower pH values than weretaxa in the same order or family in streams. Thiswas mainly because species in streams mayexperience more short-term PH depression. Thesampling station should be in a riffle area havingunimbedded rubble substrate (3-12" rocks) ifpossible. If rubble substrate is not found,whatever substrate is present in the riffle shouldbe sampled. Most of the macroinvertebrate speciespresent will be found in the rubble substrate,which has been called the “breadbasket” of thestream (Bell 1969, Cairns et al. 1971, Hart andBrusven 1976, Pennak 1977).
Number and Frequency of SamplingQuantifying stream benthic macroinvertebrates
is difficult due to the spatial and temporalvariation in species abundance (Needham andUsinger, 1956). As a minimum, three randomsamples should be taken at each station.Ideally, sites should be sampled with as manysamples as possible. Following therecommendations of Elliott (1977). enough samplesshould be taken such that the standard error isequal to or less than 20 percent of the mean.The needed number of samples should beestablished for each site for each sampling
season. Under ideal conditions with good access,samples should be taken at least monthly.Year-to-year variation in benthic communitycomposition requires that a site should bemonitored for several years.
Modified Surber Net SamplesThe modified Surber square foot sample net is
recommended for use, because it performs betterthan alternative sampling devices. The 3-ft-longnet and 18-in-high upper frame (fig. 8) reducesthe backwash problem often experienced with theoriginal Surber net (USDA 1985).
Sampling ProcedureStreams. --The foot-square modified Surber
frame is placed over the gravel-rubble substratein the stream with the net downstream. As therocks within the frame are scrubbed, themacroinvertebrates are carried into the net bythe flowing water. The substrate underlying thegravel-rubble is also stirred to a depth of 3-4in (7-10 cm), if possible.
After the water drains from the net, the netis inverted into en aluminum pan containing asaturated salt-water solution. As the salt wateris poured into a second pan, the organicmaterials thus floated are caught in a 250-micronsieve. The salt water is then poured back intothe first pan, the contents again vigorouslystirred, and the floating materials and specimensare poured for a second time into the sieve. Thesample may require sieving two, three, or moretimes. It is imperative that the pan material beinspected carefully so that non-floating benthosare hand picked and collected. Clams, snails,and cased caddisflies will not float and must behand picked and added to the sample. Large clamswill also not float into the net, and should besampled from within the Surber frame by hand.
The sample in the sieve is then washed fromthe sieve pan into the sample bottle with analcohol solution. Enough alcohol should be addedto the sample bottle to cover the sample.
Lakes. --Macroinvertebrate sampling within alake should include qualitative lake-shoresamples for sensitive indicator mayfly, stonefly,caddisfly, and amphipod species. These samplescan generally be taken with a long-handled kicknet used in a sweeping motion through vegetationor over the lake bottom substrate.
Portions of the net contents can be placed ina white tray with a small amount of water in thebottom for detection and removal of invertebratefauna with forceps. If the net contains plantq aterials, put more water in the tray andvigorously wash the plants in the tray. Thewater and its contents are poured into the 250micron sieve and then transferred to the samplebottle. The sample data on the bottle shouldinclude the words “Qualitative Lake Sample”.
Identification of Taxa CollectedThe samples collected should be sent for
identification by qualified persons. Taxa shouldbe keyed to the highest taxon possible: family ,genus, or species, depending on the group.Voucher collections should be maintained andidentifications should be checked and verified.
3 4
Once the identity and number of individuals ineach sample are known, a variety of analysismethods are available.
Data EvaluationsThe resultant data (numbers of individuals of
each taxon per sample) may be analyzed by anumber of methods. Analysis may include the useof indicator species, community composition,synthetic “biotic indices,” biomass, abundance,species richness, species diversity, andfunctional group analysis. The aim is toquantify existing conditions and identify andinterpret changes in the stream benthiccommunity. No one method will suffice. It ismost important that the data be collectedproperly with adequate sampling end accuratetaxonomic determinations. The method of analysisis not as important as the data quality itself.
Helpful guidelines for sampling and dataanalysis of stream benthos can be found in Plattset al. (1983). Sources of materials, and anextensive taxonomic literature review of insects,can be found in Merrit and Cummins (1978).Pennak (1978) provides a scholarly guide to theidentification of fresh-water invertebrates andHynes’ (1972) tome remains a classic introductionto the ecology of streams. A comprehensivereview of statistical methods can be found inElliott (1977) and a review of the use of indicesis found in Washington (1984).
ReferencesAdirondack Lakes Survey Corporation (ALSC). 1985.
Standard operating procedures: fieldoperations, 2nd ed. Ray Brook, NY: ALSC.
Anderson, R. O.: Gutreuter, S. J. 1983. Length,weight, and associated structural indices.In: Nielsen, L. A.; Johnson, D. L., eds.Fisheries techniques. Bethesda, MD: AmericanFisheries Society: 283-300.
Armour, C. L.; Burnham, K. P.; Platts, W. S.1983. Field methods and statistical analysesfor monitoring small Salmonid streams. U.S.Fish and Wildl. Serv. FWS/OBS-83/33. 200 p.
Bagenal, T.. ed. 1978. Methods for assessment offish production in freshwaters, 3rd ed.Blackwell Scientific Publications. 365 p.
Bannerot, S. F’.; Austin, C. B. 1983. Usingfrequency distributions of catch per uniteffort to measure fish-stock abundance. TransAmerican Fish Society 112: 608-617.
Bell. H. L. 1969. Effect of substrate types onaquatic insect distribution. MinnesotaAcademy of Sci. Journal 35: 2-3.
Bell, H. L. 1970. Effects of pH on the life cycleof the midge Tanytarsus dissimilis. CanadianEntomologist 102:(5): 636-639.
Bell, H. L. 1971. Effect of low PH on thesurvival and emergence of aquatic insects.Water Research 5: 313-319.
Cairns, J., Jr.: Grossman, J. S.: Dickson, K. L.;Herriks, E. E. 1971. The recovery of damagedstreams. ABS Bulletin 18: 79-105.
Carlander, K. D. 1981. Caution on the use of theregression method of backcalculating lengthsfrom scale measurements. Fisheries 6: 2-4.
Chang, w. Y. B. 1982. A statistical method ofevaluating the reproducibility of agedetermination. Canadian Journal of Fisheryand Aquatic Science 39: 1208-1210.
Cummins, K. W.; Lauff, G. H. Lauff. 1968. Theinfluence of substrate particle size on themicrodistribution of stream macrobenthos.Hydrobiologia 34: 145-181.
Drouse, S. K.; Creelmen, L. W.; Hillmsn, D. C.;Simon, S. J. 1986. Quality assurance plan forthe National Surface Water Survey. PhaseI-Eastern Lakes Survey. Las Vegas, NM: U.S.Environmental Protection Agency.
Eilers, J. M.; Lien, G. J.; Berg, R. G. 1984.Aquatic organisms in acidic environments: Aliterature review.
Elliott, J. M. 1977. Some methods for thestatistical analysis of samples of benthicinvertebrates, 2nd ed. Freshwater BiologicalAssociation Scientific Publication No. 25.157 p.
Everhardt, W. H.; Youngs, W. D. 1981. Principlesof fishery science, 2nd ed. Ithaca, NY:Cornell University Press. 349 p.
Fabrizio, M. C.; Taylor, W. W.; Baker, J. p.1987. National surface water survey phaseII--Upper Midwest Lake survey field trainingand operations manual--Part I Fish surveys.U.S.E.P.A. Corvallis, Envi. Res. Lab.,Corallis, OR 97333. 110 P.
Hagley, C.; Merritt, G.; Baldigo, B.; Kinney, W.L. 1987. National surface water survey phaseII--Upper Midwest Lake survey field trainingand operations manual--Part II. EPA fieldactivities. U.S.E.P.A. Corvallis, Envi. Res.Lab. , Corvallis, OR 97333. 57 P.
Hagley, C. A.; Knapp, C. M. Mayer, C. L.; Morris,F. A. 1986. The national surface water surveystream survey (pilot, middle-Atlantic PhaseI, Southeast screening, and middle-Atlanticepisode pilot) field training and operationsmanual. U.S. Environmental ProtectionAgency. Las Vegas, NV.
Haines, T. A.; Jagoe, C. H.; Pauwels, S. J. 1985.A comparison of gear effectiveness for fishpopulation sampling in small Maine lakes.Corvallis, OR: Report submitted to U.S.Environmental Protection Agency.
Hall, R. J.; Ide, F. P. 1987. Evidence ofacidification effects on stream insectcommunities in central Ontario between 1937and 1985. Canadian Journal of Fishery Quat.Science 44: (in press).
Hall, R. J., Likens, G. E. 1980. Ecologicaleffects of experimental acidification on astream ecosystem. In: Drablos, D; Tollan, A.,eds. Ecological impact of acid precipitation.Sanderfjord, Norway: SNSF Project: 375-376.
Hall, R. J.: Likens, G. E. 1985. Experimentalacidification of a stream tributary toHubbard Brook. EPA/600/M-85/011, May 1985.U.S. EPA, Environmental Research Brief.
Hart, D. S.; Brusven, M. A. 1976. Comparison ofbenthic insect communities in six Idahobatholith streams. Melandria 23: 1-39.
Havas, M.; Hutchinson, T. C. 1982. Aquaticinvertebrates from Smoking Hills, N.W.T.:
3 5
effect of pH end metals on mortality.Canadian Journal of Fisheries end AquaticScience 39: 890-903.
Hillman, D. C.; Potter, J. F.; Simon, S. J. 1986.Analytical methods manual for the NationalSurface Water Survey, Eastern Lake Survey(Phase I--Synoptic Chemistry). U.S.Environmental Protection Agency, Las Vegas,NV.
Helm-Hansen, O.; Reinmann, B. 1978. Chlorophyll adetermination: improvements in methodology.Oikos 30: 438-447.
Hubert, W. A. 1983. Passive capture techniques.In: Nielsen, L. A.; Johnson, D. L., eds.Fisheries Techniques. Bethesda, MD: AmericanFisheries Society: 95-122.
Hudelson, R.; Boyer, G.: McMillsn, J. 1980. Highmountain lake and stream survey of theBridger Wilderness area: 1969-1975. WyomingGame and Fish Department, Fish Division,Cheyenne, WY. Compl. Rep. F-1-R-8 toF-1-R-12.
Hynes, H. B. N. 1972. The ecology of runningwaters. University of Toronto Press. 555 p.
Interagency Task Force on Acid Precipitation.1982. Annual report to the President andCongress. National Acid PrecipitationAssessment Program, 722 Jackson Place, NW,Washington, DC.
Jerald, A., Jr. 1983. Age determination. In:Nielsen, L. A.; Johnson, D. L., eds.Fisheries techniques. Bethesda, MD: AmericanFisheries Society: 301-324.
Lagler, K. F. 1956. Freshwater fishery biology,2nd ed. Dubuque, IA: W. C. Brown CompanyPublishers. 421 p.
Lambou, V. W.; Wiener, J. G.; Biddinger, G. R.;Oglesby, R. T. 1985. Recommendations forbiological monitoring in lakes sensitive toacidification. Las Vegas, NM: U.S. Environ-mental Protection Agency, EnvironmentalMonitoring Systems Laboratory. 35 p.
Merritt, R. W.; Cummins, K. W., eds. 1978. AnIntroduction to the aquatic insects of NorthAmerica. Iowa: Kendall/Hunt Publishing Co.441pp.
Morris, F. A.; Peck, D. V.; Hillman, D. C.;Cabble, K. J.; Bonhoff, M. B.; Pierett, S. L.1986. The National Surface Water Survey.Eastern Lakes Survey-Phase I. U.S.Environmental Protection Agency, Office ofResearch and Development, Washington, DC.
Needhem, P. R.; Usinger, R. L. 1956. Variabilityin the macrofauna of a single riffle inPresser Creek, California, as indicated bythe Surber sampler. Hilgardia 24(14):383-409.
Napier, S. J.; Hummon, W, D. 1976. Survival ofmayfly larvae under acid mine conditions.Int. Revue ges. Hydrobiol. 61: 677-682.
NSWS. 1985. Analytical methods manual for theNational Surface Water Survey. U.S. Environ-q ental Protection Agency, Environmental.Monitoring Systems Laboratory, Las Vegas, NV.
NSWS. 1986. National Surface Water Survey:Eastern Lakes Survey--Phase 2, ResearchPlan. Prepared by Thornton, K.; Baker, J.P .; Rechew, K. W.; Lenders, D. H.; Wigington,
P. J., Jr. U.S. Environmental ProtectionAgency, Office of Research and Development,Washington, DC.
Nichols, L. E., Jr.; Bulow, F.J. 1973. Effects ofacid mine drainage on the stream ecosystem ofthe East Fork of the Obey River, Tennessee.Journal of Tennessee Academy of Science 48:30-39.
Parsons, J. D. 1968. The effects of acid stripmine effluents on the ecology of a stream.Archiv fuer Hydrobiologie 65: 25-50.
Pennek, R. W. 1978. Fresh-water Invertebrates ofthe United States, 2nd ed. New York, NY: J.Wiley & Sons. 803 p.
Pennek, R. W. 1977. Trophic variable in RockyMountain trout streams. Archiv fuerHydrobiologie 80: 253-285.
Platts, W. S.: Megahan, W. F.; Minshall, G. W.1983. Methods for evaluating stream,riparian, and biotic conditions. INT-138:U.S. Department of Agriculture, ForestService.
Remmick, R. 1984. Fisheries characteristics offour acid rain monitoring lakes in theBridger Wilderness. In: Air quality end aciddeposition potential in the Bridger endFitzpatrick Wildernesses. Ogden, UT: U.S.Forest Service, Intermountain Region. p.187-207.
Robak, S. S. 1974. Pollution ecology offreshwater invertebrates. In: Art, C. W.,Jr. ; Fuller, S. L. H., eds. New York, NY:Academic Press: 313-376.
Singer, R., ed. 1981. Effects of acidicprecipitation on benthos. Proceedings of aRegional Symposium on Benthic Biology, NorthAmerican Benthological Society, Hamilton, NY.
Singer, R. 1984. Benthic organisms. In:Altshuller, A. P.; Linthurse, R. A., eds. Theacidic deposition phenomenon and its effects,Vol II. Effects Sciences. U.S. EnvironmentalProtection Agency, EPA-600/8-83-016BF.
Stickney, R. R. 1983. Care end handling of livefish. In: Nielsen, L. A.; Johnson, D. L.,eds. Bethesda, MD: Fisheries Techniques:85-94.
Sutcliffe, D. W.; Carrick, T. R. 1973. Studies onmountain streams in the English LakeDistrict. Freshwater Biology 3: 437-462.
Thornton, K.; Baker, J. P.; Reckhow, K. H.;Landers, D. H.; Wigington, P. J., Jr. 1986.National Surface Water Survey, Eastern LakeSurvey--Phase II. Research Plan. U.S.Environmental Protection Agency, Office ofResearch end Development, Washington, DC.
Tomkiewica, S. M., Jr.; Dunson, W. A. 1977.Aquatic insect diversity end biomass in astream marginally polluted by acid strip minedrainage. Water Research 11: 397-402.
U.S. Department of Agriculture, Forest Service.1985. Aquatic macroinvertebrate surveys,chapter 5. Fisheries Habitat Surveys Hendbk.R-4 FSH 2609.23. U.S. Department ofAgriculture, Forest Service, IntermountainRegion.
Warner, R. W. 1973. Acid coal mine drainageeffects on aquatic life. Reprint from Ecol.and Reclamation of Devastated Land. New York,NY: Gordon and Breach. Sci. Pub. Inc. 1: 538.
3 6
Washington, H. G. 1984. Diversity, biotic andsimilarity indices. A review with specialrelevance to aquatic ecosystems. WaterResearch 18(6): 653-694.
Whitney. R. R.; Carlender, K. D. 1956.Interpretation of body-scale regression forcomputing body length of fish. Journal of
These guidelines have been prepared to assistthe Federal Lend Manager (FLM) in designing ameasurement program capable of determiningcurrent, and monitoring future, responses ofvegetation to atmospheric pollution. They aregeneral guidelines rather than specific methods.They are flexible, and guide the FLM in thedevelopment of a program suitable for a specificpermit and wilderness area.
These guidelines are built on the ongoingprograms of several regulatory agencies thatassess the effects of pollution on native plants(see Bennett 1984, 1985), but they represent anew synthesis and approach to the problem. Theguidelines are based on the acceptance of theprinciple that changes should be detected in “themost sensitive part of the ecosystem.” They arepresented as the most efficient and parsimonioussteps to make decisions involving present orfuture effects of atmospheric pollutants.
A primary goal is to obtain measurements ofplant air quality related values (AQRVs) within ashort period (one growing season), which can beused in the permitting process. However, iffunding and the management policy of the Class Iarea under consideration allow, a secondary goalwould be to establish subsequent trends endchanges via long-term measurements.
Constraints and Philosophy of Approach
Wildlife Management 20: 21-27.Witters, H.; Vangenechten, J. H. D.; Van
Puymbroeck, S.; Vanderborght, O. L. J. 1984.Interference of aluminum and pH on theNa-influx in aquatic insect Corixa punctata.Bulletin of Environmental Contamination andToxicology 32: 575-579.
Plants
Biological AQRVs worthy of measurement areinherently more difficult to identify thanabiotic AQRVS. Further, the significance oftheir measured value as an indicator of airquality is often ambiguous, The following pointsprovide the background for this comment:
1. No finite list equivalent to that ofcriteria pollutants exists for organisms andcommunities.
2. Because of the lack of adequate controlsand experimental design, any field observation orsampling will lead only to correlations andinference, not to an established cause and effectof a pollutant on a biological AQRV unlesschronic levels of pollution are present.
3. There are no known functional attributesthat respond only to specific changes in airquality. The majority, if not all, functionalattributes also will be affected by severalpollutants and by natural environmental factors.
Synergism will occur between the variouscontrolling factors. Further, some symptoms mayhave more than one cause, including those otherthan pollution.
4. Species end individuals vary in theirresponse to environmental stress because ofgenetic or ecotypic variability. Again.interactions can complicate the picture. Forexample, stresses can be mitigated or amplifiedby temporal patterns of the plants or byinvolvement with pathogenic organisms.
5. Even the functional attributes that maybe reasonably related to changes in air qualityare poorly described and quantified.Furthermore, even if the dose-responseexperiments have been done, extension to thefield is tenuous. An overall ignorance of thenorm makes many attributes, especiallyphysiological ones, of little value.
Measurements in these western high-elevationand montane Class I areas present furtherdifficulties because of the following:
1. These areas tend to be large,inaccessible, diverse, end spatiallyheterogeneous.
2. The air quality history or current statusmay be poorly known.
3. The effects of air quality on the nativeplants, communities, or biotic systems are notfully known.
The protocol has two key elements. The firstis the decision to deal with the population levelof the plant system rather than higher attributesof the ecosystem. The second key element is theflexibility and general nature of theseprotocols, which do not provide exact samplingand measurement schemes. Plants, rather thanvegetation, provide the operational perspective,which focuses on the presence end performance(health) of individual plant species and theirpopulations, rather than on attributes ofvegetation, communities, or ecosystems. Thescope of measurement is further reduced by therestriction only to known sensitive taxa andtheir sensitive organismic systems. Thecomplexity of higher-order ecological units suchas community and ecosystems and the difficulty ofmeasuring change in these units, provide amplereason for a population perspective to takeprecedence over a vegetational or total systemperspective.
Nonvascular plants such as mosses and lichens
3 7
are considered because of their known sensitivityto air pollution end their established usefulnesses indicators. Algae, bacteria, and fungi arenot included because they would require specialmethods that would be beyond any realistic budgetfor the task. The flexibility of the protocolsis justified on the basis that unique programshave to be designed by the FLM end contributingexperts to meet the unique characteristics ofeach permit, each area, end each flora.
There are two parts to the protocols: 1) aflow diagram or decision tree with text explain-ing the step-by-step process for designing aspecific program, and 2) some general guidelinesfor sampling and analysis procedures. Althoughthe decision regarding permit denial or approvalis beyond the scope of these protocols, theposition of the permitting decision process isshown in the flow diagram.
Protocol Design
The flow diagram in figure 9 shows the datasets required at the start of the design process,end the decisions necessary to design ameasurement program. The required data are shownin lettered boxes, and the processes enddecisions, here called steps, in numbered boxes.
A number of data sets are required to decidewhat, when, and where to measure the flora.
‘ These may be available in the literature orspecific archives pertaining to the Class I landunder consideration. If not available, theyshould be compiled by the FLM. Availableinformation seldom will be adequate, endpreparation of this information will be a mandateand prescription for collection. Thesecompilations may require assistance fromexperts. The data sets required for each Class Iarea are described in the sequence of thelettered boxes of figure 9. This protocol isdemanding of time end effort; there are no shortcuts .
Floristic List
A complete list of vascular plants, lichens,and mosses is needed. This list should show foreach species a commonness rating (abundant,frequent, rare) and distributional information(habitat, soil preferences, vegetationassociations ). Fairly good lists of vascularplants are available for many Class I areas. Ifthese are not available, local floras, herbariumcollections, and consultation with localsystematic botanists can supply fairly completelists that include estimates of commonness.Available lists of lichens and mosses are rarelyavailable, and their compilation would not beeasy without field surveys by specialists.Distributional information seldom may bepre-compiled but can usually be derived from suchsources as florae, plant ecology dissertations,and plant community descriptions from similarnearby regions. This list is meant as a guide inthe selection of taxa to be studied, and not as adefinitive list against which future losses canbe detected.
Land Cover MapA map of land cover units at a scale between
12,00 and 100,000 would be satisfactory. Theland cover units should be based primarily onvegetation assemblages. Each vegetation unitshould be described by species content endabundance, and may be additionally defined on thebasis of other attributes such as geology, soil,and habitat. Maps of species distribution, ifavailable, would be particularly useful,especially for a rare species whose distributionis difficult to interpret from a vegetation map.
Adequate land cover and/or vegetation mapswill only be available for a few areas. Forestinventory maps, soil surveys, and geology mapsmay be more readily available end may providebackground data for land cover maps. Most ClassI areas will have reasonable aerial photocoverage, and a skilled aerial photo interpreter,with the help of a local plant ecologist, canproduce adequate land cover maps overlaid on USGStopographic maps (1:24,000 or 1:63,360). Thecomprehensive method of Kuchler (1967) isrecommended. The description of vegetation interms of its composition is critical to theproposed protocol. According to the Kuchlermethod, this information is gathered as thegroundtruth of the map is verified and revised.
Relative Sensitivity TablesA fundamental assumption of monitoring air
pollution effects on plants is that not allspecies are sensitive to a given pollutant.Therefore, candidate test species must besensitive to a given pollutant. Furthermore, weagree with Cairns (1986) that there will also beno single reliable most sensitive indicatorspecies for specific pollutants. Therefore,several test species must be selected. Lists ofrelative sensitivity of plants to specificpollutants may be found in the literature(Applied Science Associates 1976, Davis andWilhour 1976). The EPA Criteria Documents foreach pollutant are a useful source of lists.National experts are useful contributors ofinformation at this point. When information isnot available for the actual species on the studyarea, related taxa or growth-forms (which often,but not always, have similar sensitivities to apollutant) might be considered.
Air Quality InformationThis is required to identify the probable
pollutants of concern. Sources of these datawill be any previous monitoring, the directmonitoring provided as part of the total protocol(Atmospheric Environment Section of this report),end from the permit application itself. The mostdesirable data set would contain temporal endspatial distribution of pollutants, and also thefrequency concentrations of each pollutant overthe area. Atmospheric modeling is often the bestsource of this distributional information.Research is ongoing on species sensitivity.
List of Responsive AttributesIt is necessary to measure only those
38
. Figure 9.--Flow diagram for usein designing a measurementprogram to assess effects ofpollutants on plant AQRVs.
attributes of sensitive species that clearly showdiagnostic responses to pollutants. Injury
and Blauel (1980), and Thompson, et al. (1984).Table 15 illustrates the type of information
atlases illustrating damage and stress symptoms required for each pollutant of concern.are a good source of clues as to appropriateattributes to measure; for example, Jacobson and Sampling and Analysis ProceduresHill (1970), U.S. Forest Service (1973), Malhotra For each species and attribute, there must be
39
an appropriate method of assessment of pollutanteffect. The fifth column of table 15 lists suchmethods. Handbooks with acceptable methods mustbe found or developed end refined as necessary.These methods and sources are discussed later inthis section.
Changes in Non-Plant AORVsThis information is included in the decision
tree to illustrate a complete permittingprocess. Procedures are outlined in othersections of these protocols to measurenon-floristic AQRV’s.
The Decision Process
With the assembled data and information setson hand, the measurement methods can beselected. The flow diagram (fig. 9) illustratesthis process with numbered steps. At step 1, thepotential study species are identified bycomparing the list of the flora (A) with the listof sensitive species given in the sensitivitytables (B). At step 2, the potentialpollutants--those which may increase tounacceptable levels--can be derived from the airquality information (D), and at step 3, a matchof this information with a list of species withknown sensitivity to the appropriate pollutant(C) helps to set priorities-for studies in thearea of concern and leads to the identificationof sites at which measurement could be made.
Sites where the sensitive species are presentmay be determined from the land cover and speciesdistribution maps (B). Site selection shouldalso consider predicted or known pollutionpatterns (D). If pollutant distribution data arenot available, then the sensitive species must bemonitored over its entire range. If the areasover which elevated pollutant levels are expectedto occur overlap with the distribution ofsensitive species, these areas should beintensively monitored. Those areas which areless likely to be impacted should also be
monitored. Final placement of study plots orlocation of samples will depend on which plantattributes are to be monitored.
Step 4 decides which measurement are to bemade. Table 15 provides an example of data forset E, and lists--for two pollutants and somesensitive taxa--those attributes which areresponsive or readily affected by the pollutant.For example, needle length and needle retentionin Pinus contorta are affected by ozone levels.Similarly, step 5 decides the methods ofappropriate measurement. For example, for Pinusfoliage, the method of Stolte end Bennett (1986)using large-scale random sampling would be a goodcandidate.
Step 6 involves final selection of methodsand measurements from among a candidate listderived in steps 4 end 5. Final selection willdepend on many considerations: for example, thepresence of sensitive species end attributes, theavailability of effective methods, availablelabor and time, end coordination with otherongoing monitoring programs. Often the permitdecision must be made in a very short timeframe: less than a growing season. In such acase, the FLM would need to proceed to step 7.
An important decision included in step 8 iswhether only short-term studies are mandated orwhether long-term monitoring can be attempted.(Stategies for long-term monitoring sites arediscussed later in this section.) Fromscientific and protection points of view,long-term studies (step 8) are desirable. Theywill provide the most reliable assessment ofeffects and also easier decisions in the case offuture new permit applications and permitreviews. Therefore, it is recommended that,within constraints of funds and time, step 8 begiven serious consideration. This step can bedone concurrently with short-term assessments, orto build on short-term data sets gathered duringexpediently made permit assessments.
Following the surveys and analyses (steps 7and/or 8) the FLM must review the information anddetermine, as per step 9, the health and risk ofthe plant components of the class I area. Otherevidence (G) will be brought to bear on this bythe FLM during the PSD permitting decisionproper.
General Guidelines for Sampling and Analysis
Sampling and analysis will depend on thespecies and attributes selected in the search forsensitive systems, and also upon whetherlong-term monitoring or only short-term surveystudies are being conducted. Within reason, werecommend that both monitoring and survey methodsbe as similar as possible--similar with regard tosample size and permanent marking or accuratelocation of sampling points or plots. Samplesize should always be adequate. Methods manualsshould provide information on sample size, but ifnew methods are developed, a competentstatistician should be consulted.
Sampling along gradients of airflow, and thusalong possible gradients of pollution, providesbetter, although never incontrovertible, evidencefor cause and effect. If long-term sampling is
4 0
initiated, the pollutant can be regarded as anindependent variable, and the effects of otherfactors such as naturally fluctuating climaticfactors can be taken into account. Furtherinformation on cause and effect develops as thedatabase grows over time, and when reliableco–measurements of physical and chemical factorsare collected. The re-sampling of permanentplots and tagged individual plants over timereduces the problem of spatial variability. Mostgrowth and physiological activities decline inlate summer and early fall; sampling at this timewill be the sum total of the current season’sgrowth end can reduce some of the seasonal endtemporal variability.
In most cases, potential gradients of airflow or pollution will not be known or pronouncedenough to suggest where to locate sample points.Therefore, we generally favor the randomplacement of plots or transects within the systemwhich contains the species of concern. Permanentmarking is to be preferred and, if necessary,stubbornly lobbied for in those Class I areaswhere managers may object. There are markerlessmethods but they are costly end not alwaysreliable. Methods of marking with minimum impactare discussed in Zedaker and Nicholas (1986). Werecommend that each study plot be thoroughlydescribed by the methods of Walker et al.( 1979). These descriptions form a necessarydatabase end give clues as to factors controllingplant stress other than pollutants. Photographyof plots and individual plants is a valuablesupplement to plot description. Photographs canrecord for posterity what the observer has notyet learned to spot, or what does not seemimportant at the time.
Table 15 illustrates the types end methods ofmeasurement which could be used for the twopollutants--ozone and sulfur dioxide. The tablefurther illustrates the kinds of considerationsthat FLMs and experts will need to make indeciding en appropriate measurement and itssampling method. The annotations concerningozone and sulfur dioxide effects help illustratethe process of method selection.
Ozone does not leave a residue within theplant to be measured, whereas sulfur dioxide maybe retained as sulfate or some other sulfurcompound. Therefore, tissue chemistry is notuseful for ozone detection; assays of products ofoxidation from ozone injury are not appropriatein field techniques since the products areephemeral. Lichens and bryophytes have not beenshown to be sensitive to ozone and would not,therefore, not be used in an assessment of ozoneeffects. Some pines, however, are sensitive toboth sulfur dioxide and ozone, and assessmentmethods could be combined (see Stolte and Bennett1986). We recommend the Milkweed measurementmethod of Bennett and Stolte (1985) as a model ofen assessment method.
Analysis of tree rings for accumulated tracemetals (Berish end Ragsdale 1985) and analysisfor reduced growth resulting from poor airquality (Nash et al. 1975) are attractive methodssince they have the potential to show previousregimes of effects of pollutants on growth.Generally, however, we caution against tree ring
analysis since it is technically demanding endexpensive. Also, cross-contamination of treerings is possible and any reduced growth effectscan seldom be related to specific pollutants.
Perhaps the most difficult part of the plantprotocols will be the determination of thesignificance of observed plant responses, andwhat the continued or ultimate consequences orthose responses imply for the plant population.It is upon these prognoses that the permittingdecision will rest.
Long-Term Monitoring: Measures and BasicSampling Design
When a specific pollutant is not identified,or for such concerns as acid deposition that maycause ecosystem-level effects, it is recommendedthat a basic and long-term monitoring effort beconducted.
The original list of the flora (data set A)should be field checked. Special care should betaken to have correct identification andherbarium archiving to avoid incorrectconclusions about future losses of species.
Table 16 lists the principal attributesrecommended for measurement. Tables 17 and 18and figures 10, 11, and 12 provide some sampleforms and scales for plot description by thereleve method (Walker et al. 1979). A releve is
4 1
literally a “picture” of a plot or stand. Thesetables and scales were designed for an Alaskanresearch program, and serve only as a startingpoint for the FLM and must be modified for eachnew area. The basic sampling design calls for
several equivalent landscape units, such as smallalpine basins, to be surveyed and mapped. Theseunits should be selected, wherever possible,along known or predicted airflow paths wheregradients of pollution might be expected.Permanent plots that contain the growth formsbeing studied should be established within eachbasin. These growth forms are fruticose andfoliose lichens, evergreen plants, and trees.Individual plants or plant parts must bepermanently tagged or able to be reliablyre-identified.
RequirementsEstimates of person-day and field equipment
requirements are given in tables 19 and 20.These estimates may be rather low, and may needto be increased for sites with pooraccessibility.
Field MethodsThe initial regional survey of plot
establishment. should be done during the firstgrowing season. Plots should be accessible but
away from main trails and access points. Plotsand plants are tagged as inconspicuously andunobtrusively as possible. Botanical samplingcan be conveniently done in late summer and earlyfall after plots and plants have been selectedand appropriately worked, recorded, andpositioned. In subsequent years, field work canbe restricted to resampling the attributes with afrequency of about three to ten years.
Survey and Plot EstablishmentLandscape Unit Mapping.--Aerial survey, local
interviews, and small-scale topographic maps areused t-o select candidate landscape units. Thisselection should be made in cooperation withSoils end Geology and Aquatic Chemistry andBiology. Small, well-defined valleys containingforests and meadows with surrounding uplandswould provide the required sample plots.Reconnaissance visits are required to make finalunit selections. Several (at least five)landscape units should be selected along knownairflow paths. Additional (up to five) unitscould be located in a cluster at the center ofthe airflow path. These additional units wouldserve to establish spatial and other naturalvariation of attributes.
Geobotanical mapping of each landscape unitusing the Kuchler (1967) comprehensive method ispreferred. Preliminary landform and vegetationboundaries and classification are made on acetateoverlays of suitable aerial photographs. Colorinfrared photography at 1:60,000 works well.
Following field checking end appropriateupdating, boundaries can be easily transferred toa 1:24,000 USGS topographic base map.
Each final map unit should be characterizedfor vegetation and landform cover. The method ofWalker et al. (1979), which uses the relevetechnique (Westhoff and Maarel 1978). is quickend appropriate (tables 17 and 18, figs. 11 and12) .
Voucher collections should be made when rareplants or plants of uncertain identity areencountered. Standard herbarium techniquesshould be used. Figure 11 gives the needed plantcollection information.
Soil information can be readily added tothese geobotanical maps.
Mapping of each basin will require about 0.25men days of aerial photo interpretation and 4person days of field effort. Final plantdeterminations and drafting should be done in thelaboratory.
Large plot location and description. In eachlandscape unit, large 50 X 50 m permanent plotsshould be located and established. These shouldcontain good stands of fruticose and folioselichens, evergreen plants (both shrubs andtrees), end mature forest trees. In each unit aminimum of five plots should be set up to provideaccess to the required taxa. All taxa might becontained in a single plot. Plots should bestaked, tagged, located precisely on thegeobotanical maps, end reference sightings madeto prominent terrain features. A series ofoblique photographs should record general aspectand vegetation of each plot. Each plot should bedescribed with the releve method, which willprovide a complete inventory of flora andestimates of individual species abundance. Covervalues should be recorded for shrubs, herbs, andground layers. Trees should be recorded byspecies, number, and dbh (diameter at breastheight).
Set-up and description of each plot shouldtake about 0.25 person-day.
Attribute Sampling end Monitoring
Lichens. Ten small 20 X 25 cm microplotsshould be set up in each of the five larger (50 X50 m) macro-plots to record ground and rocklichen communities. The plots can be marked withsmall stainless steel stakes or small holes inrock surfaces using a star-rock drill andhammer. The basic method is that of Hale(1982) . Vertical whole plot photographs andoblique aspect photographs should be taken foreach plot. A complete as possible listing oflichens is made. Voucher collections should bemade outside of the microplots. A large handful(5-10 g) of each common fruticose or folioselichen is collected from within each 50 X 50 mplot for elemental analysis (see table 16).Samples should not be collected into brown kraftpaper bags because of the possibility of sulfurcontamination. Similarly, zinc contamination mayresult from using ordinary plastic bags.Synthetic bags made of materials such as tyvekare preferable. The bags should “breathe” endthe lichens should be dried in the bag. Foreign
material (other lichens, insect cases, moss,etc.) should not be removed from air-driedspecimens. Specimens should not be oven-dried orwashed. Samples should be ground in a Wiley millto pass a 20-mesh screen.
Setting up and recording 10 microplots ineach macroplot and sampling for elemental contentof lichens takes about 0.5 of a man day.
Evergreen foliage. Evergreen species shouldbe sampled within the macroplots available.Evergreen shrubs or trees are best; evergreenherbs are of dubious value. When possible, thesame species should be used throughout awilderness area. Ten individual plants should betagged and their locations recorded in amacroplot. Ten individual branches should alsobe tagged on each plant to allow for replicationand future repeated measurements. Photographs ofeach individual plant should be taken.
Each branch is examined against standardcolor charts and photographs of leaf damage fromknown pollution effects, and scored for signs ofnecrosis (flecking, tip-die-back, etc.) andchlorosis. Further literature and methoddevelopment is needed here. A good start isMiller and McBride (1975).
Samples of leaves of each age class (about 20grams fresh weight) should be taken fromneighboring unmarked plants. Dead and livingleaves should be collected separately. Fivesamples per macroplot are recommended.
A record should be made of leaf numbers perage class for each marked branch.
Tree wood. Methods of tree coring and woodtrace element analysis are well worked out byBerisch end Ragsdale (1985). Within the 5macroplots, 10 trees of each dominant speciesshould each be cored 3 times at breast height.TWO of the three cores are used for elementanalysis in 5-year increments, and the remainingcore is used for growth analysis. In subsequentyears, only short cores will be necessary. Eachtree should be photographed, tagged, and itslocation recorded. Coring of 20 trees takesabout 0.5 man day.
Acknowledgement
Work Group 4 would like to thank Dr. James P.Bennett of the National Park Service, Air andWater Quality Division, for much helpful advice,discussion, and reference sources.
References
Allen, S. E.; Grimshaw, H. M.; Rowland, A. P.1986. Chemical analysis. In: Moore, P. D.:Chapman, S. B., eds. Methods in plantecology. London, England: BlackWellScientific Publications: 285-344.
Applied Science Associates, Inc. 1976. Diagnosinginjury caused by air pollution. Office Airand Waste Management, U.S. EnvironmentalProtection Agency.
Bennett, J. P. 1984. Air pollution effects onvegetation: monitoring and research. In: Airpollution effects on parks and wildernessareas. Mesa Verde National Park, NationalPark Service.
Bennett, J. P. 1985. Regulatory uses of S02
effects data. In: Winner, W. E.; Mooney, H.A .; Goldstein, R. A., eds. Sulfur dioxide endvegetation: physiology, ecology, and policyissues. California: Standford UniversityPress.
Bennett, J. P.; Stolte, K. W. 1985. Usingvegetation biomonitors to assess airpollution in National Parks: milkweed survey.Natural Resources Report Series No. 85-1,National Park Service. 16 pp.
Berish, C. W.; Ragsdale, H. L. 1985.Chronological sequence of element con-centrations of wood of Carya spp. in thesouthern Appalachian Mountains. CanadianJournal of Forestry Research 15: 477-483.
Cams, J., Jr. 1986. The myth of the mostsensitive species. Bioscience 36(10):670-672.
Davis, D. D.; Wilhour, R. G. 1976. Susceptibilityof woody plants to sulfur dioxide andphotochemical oxidents. EPA-600/3-76-102.Corvallis, OR: Environmental ResearchLaboratories, Environmental ProtectionAgency.
Fox, Douglas G. 1986. Establishing abaseline/protocols for measuring air qualityeffects in wilderness. In: Lucas, Robert C.Proceedings--national wilderness researchconference: current research. Fort Collins,CO, 23-26 July, 1985. Gen. Tech. Rep.INT-212, Ogden, UT: U.S. Department ofagriculture, Forest Service, IntermountainResearch Station: 85-91.
Hale, M. E., Jr. 1982. Lichens as bioindicatorsand monitors of air pollution in the FlatTops Wilderness Area, Colorado. Final Report:Forest Service Contract No. OM RFPR2-81-SP35. Rocky Mountain Range endExperimental Station, Fort Collins, CO.
Jacobsen, J. S.: Hill, A. C., eds. 1970.Recognition of air pollution injury tovegetation: a pictorial atlas. Pittsburg, PA:Air Pollution Control Association.
Kuchler, A. W. 1967. Vegetation mapping. NewYork, NY: Ronald Press. “
LeBlanc, F.; Rae, D. N. 1975. Effects of airpollutants on lichens and bryophytes. In:Mudd, J. B.; Kozlowski, T. T., eds. Responsesof plants to air pollution. New York, NY:Academic Press: 237-272.
Malhotra, S. S.; Blauel, R. S. 1980. Diagnosis ofair pollutant and natural stress symptoms onforest vegetation in western Canada.Information Report NOR-X-228. Edmonton,Alberta: Canadian Forestry Service.
Miller, P. R.; BcBride, J. R. 1975. Effects ofair pollution on forests. In: Mudd, J. B.,Kozlowski, T. T., eds. Responses of plantsto air pollution. New York, NY: AcademicPress: 195-235.
Nash, T. H., III; Fritts, H. C.: Stokes, M. A.1975. A technique for examining non-climaticvariation in widths end annual tree ringswith special reference to air pollution.Tree-Ring Bulletin 35: 15.2.
Stolte, K. W.; Bennett, J. P. 1986. Standardizedprocedures for establishing permanent pineplots end evaluating pollution injury onpines. Air Quality Division, National ParkService, P.O. Box 25287, Denver, Colorado80225.
Thompson, C. R.; Olszyk, D. M.; Kats, G.;Bytnerowicz, A.; Dawson, P. J.; Wolf, J.;Fox, C. A. 1984. Air pollutant injury onplants of the Mojave desert. Rosemead, CASouthern California Edison Co.
U.S. Department of Agriculture, Forest Service.1973. Air pollution damages trees. UpperDarby, PA: Northeastern Forest ExperimentStation.
Walker, A. A.; Webber, P. J.; Komarkova, V. 1979.A large scale (1:6000) vegetation mappingmethods for northern Alaska. Boulder, CO:Institute of Arctic and Alpine Research,Plant Ecology Laboratory, University ofColorado. 49 p.
Westhoff, V.; van der Maarel, E. 1978. TheBraun-Blanquet approach. In: Whittaker, R.H., ed. Classification of plant communities.2nd ed. The Hague: Junk, 287-399.
Wenmore, C. M. 1983. Lichens and air quality inVoyageurs National Park. Final Report,Contract No. CX 0001-2-0034: Denver, CO:National Park Service, Air Quality Division.13 p. + 21 figures.
Zedaker, S. M.; Nicholas, N. S. 1986. Qualityassurance methods manual for siteclassification and field measurements.Corvallis, OR: U.S. EPA and USDA ForestService Response Program, CorvallisEnvironmental Research Laboratory. 90 p.
Work Group Participants
Douglas G.
Organization
Fox - Project Leader, USDA-ForestService, Rocky Mountain Station, Ft. Collins.co
J. Christopher Bernabo - Project Manager, Scienceend Policy Associates, Inc., Washington, DC
Betsy Hood - Project Staff, Science end PolicyAssociates, Inc., Washington, DC
Work Group 1 - Atmospheric
Volker Mohnen - leader, Dept. AtmosphericSciences, SUNY-Albany, NY
David Dietrich, Air Resources Specialists, Inc.,Ft. Collins, CO
James Galloway, Dept. Environmental Sciences,University of Virgins, Charlottesville, VA
James Gibson, Natural Resources Ecology Lab,Colorado State University, Ft. Collins, CO
Thomas Hoffer, Desert Research Institute,University of Nevada, Reno, NV
William Reiners, Dept. Botany, University ofWyoming, Laramie, WY
Steven Connolly, Jellinek, Schwartz, Connolly &Freshmen, Inc., Washington, DC
Richard Fisher, USDA-Forest Service, Watershedand Air Management, WO, Ft. Collins, CO
Richard A Sommerfeld, USDA-Forest Service, RockyMountain Station, Ft. Collins, CO
Douglas G. FOX
Work Group 2 - Soils and GeologyWilliam McFee - leader, Dept. Agronomy, Purdue
University, West Lafayette, INJames GallowayArthur Johnson, Dept. Geology, University of
Pennsylvania, Philadelphia, PASteve Norton, Dept. Geological Sciences,
University of Maine, Orono, MEWilliam ReinersWilliam (Toby) Henes, USDA-Forest Service, R-3,
Albuquerque, NMCharles Troendle, USDA-Forest Service, Rocky
Mountain Station, Ft. Collins, CORay Herrmann, USDI-National Park Service, Water
Resources Division, Ft. Collins, CO
Work Group 3 - Aquatics
James Galloway - leaderJames GibsonWilliam McFeeFrank Senders - coleader, Wyoming Water Research
Center, University of Wyoming, Laramie, WYSteve NortonAlan Galbraith, USDA-Forest Service, Bridger-
Teton NF, Jackson, WYFred Mangum, USDA-Forest Service, R-4, Provo, UT
Frank Vertucci, USDA-Forest Service, RockyMountain Station, Ft. Collins, CO
Work Group 4 - Plants
Patrick Webber - leader, Institute of Arctic endAlpine Research, University of Colorado,Boulder, CO
William ReinersArthur JohnsonWilliam McFeeBarry Johnston, USDA-Forest Service, R-2, Denver,
coPaul Miller, USDA-Forest Service, Pacific
Southwest Station, Riverside, CAAnna Schoettle, USDA-Forest Service, Rocky
Mountain Station, Ft. Collins, CO
Work Group 5 - Regulatory
Steven Connolly - leaderChris BernaboVolker MohnenWilliam McFeeJames GallowayFrank SandersPatrick WebberPaul Barker, USDA-Forest Service, Recreation
Management, WO, Washington, DCDouglas FOXDennis Haddow, USDA-Forest Service, R-2, Denver,
CO
Work Group 6 - Applications
Douglas FoxLarry Svoboda - co-chair, EPA, Region 8, Denver,
coJames ByrneJohn Clouse - co-chair, State of Colorado. Air
Quality Division, Denver, CODennis HaddowLee Lockie, State of Arizona, Air Quality,
Phoenix, AZAl Riebau, USDI-Bureau of Lend Management,
Wyoming State Office, Cheyenne, WYHal Robbins, State of Montana, Air Quality,
Helena, MTChris Shaver, USDI-National Park Service, Air
Quality Division, Denver, COKent Schreiber, USDI-Fish end Wildlife Service,
Kearneysville, WVRandy Wood, State of Wyoming, Environmental
Quality Division, Cheyenne, WY
4 8
RockyMountains
Southwest
GreatPlains
U.S. Department of AgricultureForest Service
Rocky Mountain Forest andRange Experiment Station
The Rocky Mountain Station is one of eightregional experiment stations, plus the ForestProducts Laboratory and the Washington OfficeStaff, that make up the Forest Service researchorganization
RESEARCH FOCUS
Research programs at the Rocky MountainStation are coordinated with area universities andwith other institutions. Many studies areconducted on a cooperative basis to acceleratesolutions to problems involving range, water,wildlife and fish habitat, human and communitydevelopment, timber, recreation, protection, andmultiresource evaluation
RESEARCH LOCATIONS
Research Work Units of the Rocky MountainStation are operated in cooperation withuniversities in the following cities.
Albuquerque, New MexicoFlagstaff, ArizonaFort Collins, Colorado*Laramie, WyomingLincoln, NebraskaRapid City, South DakotaTempe, Arizona
* Station Headquarters 240 W Prospect St., Fort Collins, CO 80526