Ecological and financial assessment of late-successional reserve management

30 Pages·2001·1.6 MB·English

by  Hummel, Susan Stevens

document
Historic, archived
Do not assume content reflects current
scientific knowledge, policies, or practices.

USDA United States Ecological and Financial
Department of
Agriculture Assessment
fl9f
Forest Service
Pacific Northwest Successional Reserve
Research Station
Management
Research Note
PNW-RN-531
October 2001
Susan Stevens Hummel, R. James Barbour, Paul F.
Hessburg, and John F. Lehmkuhl7
Abstract This paperdocuments methodsforassessingthe potential effectsofvariable-intensity
managementin late-successional reserves(LSRs)and providesanexample(theGotchen
LSR)fromtheCascade RangeineasternWashington.TheGotchen LSRstudyinvestigates
changes inforestvegetation associatedwith silviculturaltreatments, and howdifferent
treatmentcombinations mayaffect landscape patterns, LSR habitatobjectives, fire
hazard, andthecharacteristicsand valueofwood removed overspaceandtime. The
studycontributestothe conceptual and technical developmentofa decision-analysis
tool, the northeastern Cascades landscapeanalysis, management, and monitoring
system (NOCLAMMS), forland management. Landscapeevaluation ofthe Gotchen LSR
revealsthatsincethe 1930s, foreststructures have become more homogeneous; area
and average patch sizeofyoung, multistoried foreststands havedecreased; and spatial
patternsoflate-successionalforesthavechanged. Thesechangesaltervegetation
response to disturbances like fires, insects, and diseases, and suggestthat different
structures and patterns may bettersupport LSR objectives overspace and time. Study
resultsaid in identifying candidatetreatmentareas, in developing prescriptionsto
maintain orrestore desired stand structures and patterns, and in understanding the
financial commitmentnecessaryfordifferentmanagementactions. Silviculturaltreatments
areapplied by usingtheforestvegetationsimulator(FVS).Thefinancialevaluationof
ecosystem managementactivities (FEEMA) software is used to calculate net revenues
associated with differenttreatments. Results from one stand illustrate these methods.
Keywords: Forestreserves, northern spotted owl, restoration silviculture, habitat
management, western sprucebudworm, fire hazard.
'Susan Stevens Hummel is a research forester and R. James
Barbour is a research forest products technologist, Forestry
Sciences Laboratory, P.O. Box 3890, Portland, OR 97208-3890;
Paul F. Hessburg is a research plant pathologist and John F.
Lehmkuhl is a research wildlife biologist, Forestry Sciences
Laboratory, 1133 N. Western Avenue, Wenatchee, WA 98801-1229.
—
Introduction Anetworkoflate-successional reserves (LSRs)within the range ofthe northern spotted
owl (Strixoccidentalis caurina) extends east ofthe Cascade Range in the Pacific North-
western United States. This research considers howachieving LSRobjectivesovertime
maybe influenced bycurrentforestconditions, administrative boundaries, disturbance
dynamics, and costs ofsilvicultural treatments. Ourstudy area, the 15,000-acre Gotchen
LSR, illustrates the problem ofmaintaining a specific successional stage within fixed
boundarieswithoutclearguidanceonthe legal roleofactivemanagement.Anoverviewof
issues inforestreserve establishmentand managementsetsthe contextforourresearch
questions and is followed by a description ofthe study area, methods, some initial
findings, and an example.
Background on Land classification terms are intertwined with land use. Theway land is classified
Forest Reserves influences both how itis managed overtime (Burgi 1999) and how it is valued (Abbottet
al. 1994). Ambiguity in classification terms can create confusion in land management.
The land classification term "reserve" hasa historyofdifferentinternational, national, and
local definitions (Congressional Record 1907, Fagergren 1998, Green and Paine 1997,
IUCN 2000, Schama 1995). Thesedifferencesaffectfederalforestreserves inthe Pacific
Northwestern United States. Forexample, the Northwest Forest Plan (NWFP) established
late-successional reserveswithin the range ofthe northern spotted owl butdid not
specify which historical definition was to be used.
Adebate aboutthe role ofpassivevs. active management in LSRs currentlyexists, in
part, because land classified as a reserve may, depending on thedefinition, be eligible
foreithertype ofmanagement.Adding tothe debate, documents underlying the NWFP
contain untested assumptions aboutforestdevelopmentand the costs ofLSR manage-
ment, particularlyon the eastern slopeofthe Cascade Range (FEMAT 1993). Weexplore
these assumptions by investigating the potential ecological and financial effects of
passivevs. active managementin the Gotchen LSR.
Issues in Reserve Forest ecosystems are dynamic in space and time (Harper 1977, Koslowski et al. 1991,
Management Watt 1947). Plants and animals continuously reproduce, grow, and die. Some plants,
DynamicSystems such astrees, can liveforcenturies, whereas others have annual life cycles. Overtime,
differentcombinationsofplantand animal life historiesand architectures createcommunity
structuresthatsupportdifferentgroups oforganisms and processes. In a forest, for
example, the animals using early-successional communities differfrom those in later
serai stages (Harris et al. 1982). The role ofdead plants like snags and logs also differs
as the community changes (McComb and Lindenmayer 1999). How plants and animals
—
arearrayedwithin a particularforeststand the horizontal and vertical structure
influenceswhatotherplantsand animals maythrivethereovertime (Spies 1998). Similarly,
the relation ofthatforeststand with neighboring land creates influential structures and
patterns(Franklinand Forman 1987, Reinersand Lang 1979,Turner1989).A"landscape"
refers to groups ofsmaller land units and, like these smaller units, changes continuously.
Landscape dynamics have important implicationsforthe potential risks associated with
differenttypes and intensities offorest managementand land use.
Land use patterns resulting from human activities (e.g., roads, cities, and logging) are
superimposed overlandscape patterns resultingfrom disturbanceevents(e.g., landslides
and wildfires) and biophysical factors (e.g., soils and pathogens). Ifland use patterns
differsignificantlyfrom landscape patterns, changes may occurin soil productivity, fire
regimes, and species extinction rates (Christensen et al. 1996). The risk ofthese
2
changes is uncertain but potentially high, and thus an ecosystem-based approach to
land managementhas been adoptedforfederalforests inthe PacificNorthwestern
United States (Anon. 1999). In ecosystem-based management, the primaryobjective is
maintaining ecosystem structure, organization, and function ratherthan producing
commodities (Woodleyetal. 1993). Currently, there is no agreementon either howto
assess the risks ofecosystem change or howto manage forests for ecosystem
function, butdifferentapproaches havebeen proposed. Theyinclude:
Using historical disturbance patterns to guide land use decisions (Cissel et al. 1999)
Using rangeandvariation ofhistorical vegetation patterns (Morgan etal. 1994,
Swanson etal. 1994, Hessburg etal. 1999a, Landres etal. 1999)
Establishing reserves to protect key ecosystems, organisms, and processes (e.g.,
Diamond 1975).
Static Boundaries Forestreserve boundariesgenerallyarefixed around an areachosenforspecific
reasons. Late-successional reserves, forexample, wereestablishedtoaddress
biological and social concerns aboutthe persistence ofspecies associated with late-
successional forests. Fixing administrative boundaries around dynamicecosystems
may, paradoxically, undermine reserveobjectives, especially if:
Reserve size is inconsistentwith key processes and organisms (Everett and
Lehmkuhl 1996, 1997; Pickettand Thompson 1978),
• Forestswithinthe reserveweresignificantlymodified before boundaryestablishment
(Peluso 1992), or
• Anonmodifiedforestreserve isan "island"surrounded byheavilymodified land
(Curranetal. 1999).
Norton (1999) reviewsthe biogeographicoriginsof, and argumentssurrounding, reserve
designs includingthe relative meritsofseveral small reservesconnected bycorridorsvs.
largereserves.
Potential Role of Human activitiespermitted inforestreservesdifferbycountryandwith reserveobjectives
Management (Parviainen etal. 2000). Relations betweencurrentconditionsand reserveobjectives
influence howforestreservesareestablished and maintained. Passive managementmay
be all that is needed ifthe successional dynamicsofcurrentforestconditions are consistent
with the conditions and processes desired forthefuture. In contrast, active management
could be required toachievefutureobjectives. If, forexample, a species like ponderosa
pine (Pinusponderosa Dougl. ex Laws.) is desired but no seed source is available on
site, orifseeds are available butconditionsforgermination, establishment, and growth
are not,thenfavorableconditionsforregeneration andcompetitiveadvantagewould need
tobecreated. Hence, the roleofactive managementdependson reserveobjectivesand
theextenttowhich currentand predicted future conditions supportthese objectives at
appropriate scales. Ifconditions and objectives are inconsistent, or ifmethodsto make
them consistentare not available orare too costly, then the risk increases that reserve
objectives maynotbe met. Failureto achieve reserveobjectivescouldweaken support
for a system of reserves, especially ifthe system restricts or eliminates access to
publicland.
3
There are manyfederal and state parks, wilderness areas, and reserves in the Pacific
Northwestern United States. The NWFPcreated additionaltypesofreservesonfederal
land, including riparian reservesand LSRs. Itisuncertainwhethertheobjectivesforthese
reservescan beachieved, given currentecological and regulatoryconditions. This
uncertainty is documented in the Forest Ecosystem ManagementAssessmentTeam
(FEMAT) reportand inthe record ofdecision (ROD)forthe NWFP (USDAand USDI
1994). The FEMAT report noted, forexample, the presence ofyoung stands in LSRs
and asked, "Should these young stands be silviculturallytreated to acceleratetheir
attainment ofa condition that mimics late-successional forest conditions ... or should
there be no silvicultural treatment . . . underthe assumption thatsuch stands will evolve,
given enough time, intothe desired habitatconditions?" (FEMAT 1993). Thetopicwas
thus leftopen with the idea thatscientificstudies could develop and investigate testable
hypotheses (USDAand USDI 1994).Anotherassumption ofFEMAT recast in this study
as a testable hypothesis, is that silvicultural treatments could both reduce fire hazard
and maintain and develop late-successionalforestconditions.
Late—Successional The NWFP established LSRson 30 percentofthefederal land areawithin the rangeof
Reserves and The n°rthem spotted owl (USDAand USD1 1994). The ROD, whichjointlyamendedthe
Northwest Forest planning documentsof19 nationalforestsand seven BureauofLand Management
pian Districts, adopted seven land allocations. Objectivesforthe LSR land allocation areto
"protectand enhance conditions oflate-successional and old-growth forestecosystems,
Definition which serve as habitatforlate-successional and old-growthforest related species
includingthe northern spotted owl (USDAand USD1 1994)."Theseobjectivesare similar
tothoseofInternational Union ofConservation ofNature (IUCN)category IV(Habitat/
Species ManagementArea), which suggestglobal standardsto ensurethe maintenance
ofhabitats and meetthe requirementsofparticularspecies (IUCN 2000). The NWFP
objectives may, arguably, be clear, butthe methods to achieve them are not, especially
on the eastern slopeofthe Cascade Range, where both disturbance dynamics and
previous land use patterns complicate maintaining a specificserai stagewithin fixed
boundaries (FEMAT 1993).
Issues The rangeofthe northern spotted owl extends eastofthe Cascade crest, where dense,
multilayered canopies and down wood create suitable habitat (Buchanan etal. 1995,
Gaines etal. 1997, Thomas et al. 1990) but also increase susceptibility to fire (USDA
and USD1 1994) and to insectsthatcause defoliation (e.g., western spruce budworm
(Chohstoneuraoccidentalis))(a\so see Brookes etal. 1987). Stand replacementfires are
partofthe historical disturbance regime (Agee 1990, 1993, 1994; Camp etal. 1997), yet
such fires may be socially unacceptable because ofactual or perceived riskto human
life orproperty, wildlife habitat, riparian systems, and tourism. Due in partto previous
federal fire suppression policies, mixed-coniferforestscurrently have more smallto
intermediateshade-toleranttreesthanwould havesurvived underhistoricalfire regimes
(Covington etal. 1994, Hann etal. 1997, Huffetal. 1995, Lemhkuhl etal. 1994). This
increased biomass is distributed on manysmall-diametertrees, ratherthan on fewer,
largerones (GAO 1999). Silvicultural treatments in such conditions are often expensive,
owing to the high costofremoving small trees and the generally low marketvalue of
wood removed. Theseconditionsare prevalentthroughoutthe interiorWest(Everettetal.
1997), and budget allocation methods in the National Forest System may mean that
stands with the highestfire hazard are left untreated (GAO 1999).As stated in FEMAT
(1993), "some habitatmodification activities in Late-Successional Reserveswill generate
4
enough revenueto payforthemselves Otherswill notand need to be supported by
. . .
appropriatedfunds."This research evaluateswhatsilviculturaltreatments mayfall in
which categoryand the potential subsidies needed forhabitat management.
Land managers responsibleforLSRson the eastslopeofthe Cascade Rangeare
concerned about potential risks to existing late-successional habitat associated with
passivemanagementandabouttheecologicalandfinancial implicationsofactivemanage-
ment. The northern spotted owl isfederally listed as an endangered species, and thus
the primary concern in the LSRs is forowl habitat. Accordingly, wefocus on owl habitat
but recognize thatthe requirements ofotherspecies associated with late-successional
forest ecosystems also must be considered.
The NWFP acknowledges a potential need foractive managementin LSRs byspecifying
thatsilvicultural systemsallowedwithinthem should "prevent large-scaledisturbances
byfire, wind, insects, and diseases thatwould destroy or limitthe ability ofthe reserves
tosustain viableforestspecies populations" (USDAand USD1 1994). The NWFP also
explicitly recognizesthe challenges associated with maintaining suitable owl habitateast
ofthe Cascade Range, wherethere is an "increased riskoffire . . . duetoxericconditions
and the rapid accumulation offuels asthe aftermath ofinsectoutbreaks and drought"
(FEMAT1993). Managementguidelinesareoffered inthe RODthatpermit"risk-reduction"
activities in LSRson the eastern slope ofthe Cascade Range, provided such activities
are"clearlyneeded"and assure"long-term maintenanceofhabitat" (USDAand USDI
1994).Thereislackofagreement, however, among managementand regulatoryagencies,
advocacy groups, and citizens on acceptable levels ofriskand on the best management
practicestoprovide habitatsoverthe longterm atscalesappropriateforspeciesassociated
with late-successionalforests. Thesituation seems urgentbecause passive management
ofcurrent conditions may instead result in the "large-scale disturbances" the ROD seeks
toavoid.Activemanagement, however, raisesbothtechnical and philosophicalquestions.
"On one side ofthe debate arethose who, cognizantofpastsuccesses, believethat
managementcan and will produce desired results. On the otherside arethosewho,
cognizantofpastfailures, are morecautious. They believethatproofshould precede
anysilvicultural activitiesin reserves" (FEMAT 1993). Silviculturaltreatmentscould have
mixed effects overthe shortterm (decades), longterm (centuries), and differing spatial
scales.Actionstaken nowto avoid orminimize large-scale disturbances might have
directorindirectconsequencesfor NativeAmerican treaty rightsorhabitatsofrareor
endangered species (USDAand USD1 1994). Nonfederal landownersareconcerned
aboutpotential implicationsofpassive managementon adjacentprivatepropertyrights
and uses. Lackofinformationaboutthepotentialeffectsofpassiveandactivemanagement
fuels the debate. This research addressesthetechnical, ratherthan the philosophical,
aspectsofLSR managementbyconsidering thefollowing general questions:
• Whatarethe potential consequences ofpassive LSR managementontheeastern
slopeofthe Cascade Range?
Howmightan understanding oflandscape pattern-process interactionsguide LSR
management?
• Whathuman activities supportLSRobjectives?
• Dotheactivitiesgenerate revenuesthatoffsetthe costofhabitat managementand
monitoring?
5
Gotchen Late- This study addresses thesefourgeneral questions in the Gotchen LSR by (1) charact-
Successional erizing recentchanges in vegetation structure and composition at both stand and LSR
Reserve Study levels; (2)evaluating associated changes infireandwestern spruce budworm hazard; (3)
testing ifsilvicultural treatments can simultaneously reducefire hazard, maintain existing
Objectives
late-successional habitat, and recruitfuture habitatattributes; and (4) calculating net
revenuesassociatedwith differentcombinationsofpassiveand active management.
Analysesofboth Gotchen LSRvegetation conditions and patterns ofpotential responses
toalternativetreatmentscontributetothe northeastern Cascades landscape analysis,
management, and monitoring system (NOCLAMMS), currentlybeing developed bythe
authors. The Gotchen LSR is the pilotarea for NOCLAMMS, a decision tool thatenables
land managerstosimultaneouslyanalyzetheeffectsofalternative managementstrategies
on several resources. Thegoal ofNOCLAMMS isto integrate information on landscape
pattern and process interactions, key habitats and species, and economics.
Thethree-step NOCLAMMS process usesthe USDA ForestService ecosystem
managementdecision support(EMDS) software (Reynolds2000). The EMDS isthe
integrativeframeworkforstructuring ecological and managementdata. Thefirststep in
the iterative process uses EMDStoevaluatewhethercurrentvegetation patterns (e.g.,
late-successionalforest)fallwithin a rangeoflandscape referencevariation (Hessburg
1999a, 2000a). Current conditions in the Gotchen LSR illustrate inputs to this first step.
Ifcurrent conditions are outside this range, the second step uses the data, processes,
models, and outputdescribed in this paperto develop silvicultural treatments and to
evaluatetheirpotential contribution tofuture landscape patterns. Thefinal step uses
EMDSto evaluate ifthesefuture patterns meetthe objectives established forowl habitat,
fireand budworm hazard, and referencevariation. Netrevenuesareestimatedfortreat-
mentcombinationsthat meetthese objectives to evaluate the potential costofhabitat
management. These methodscould beadaptedforsimilaranalyses in otherlocations
where information on a rangeofpotential costsassociated with developing and
maintaining specificforest structures is desired.
Study Area The Gotchen LSR is in south-central Washington (T. 7 N., R.10 E.) in the Gifford Pinchot
National Forest. The northern boundaryofthe LSR isthe MountAdams Wilderness
Area, the eastern istheYakama Indian Reservation, the southern is a mile inside the
KlickitatCounty line, and thewestern boundaryfollows theWhite Salmon River (fig. 1).
The Gotchen LSRwas included in planning areas ofboththe NWFP (USDAand USDI
1994) and the InteriorColumbia Basin Ecosystem Management Project (ICBEMP)
(Quigley andArbelbide 1997) butexcluded from the analysis ofthe Eastside Forests
Scientific Society Panel (Henjum et al. 1994). In this study, the LSR administrative
boundariesdefinethe landscape analysis unit.
Mostofthe Gotchen LSRfalls within the grand fir (Abiesgrandis) series (Franklin and
Dyrness 1988, USDA ForestService 1997). Elevation rangesfrom about3,000to6,000
feetand estimated annual precipitation from 35to 65 inches (Topik 1989). There are six
documented spotted owl activitycenters inthe Gotchen LSR (USDA ForestService 1997).
6
7
Conditions in LSRs on the eastside ofthe Cascade Range are generallydrierthan those
on thewestside (FEMAT 1993, USDAand USDI 1994), which meansthat increasing
stand density also increases competitive stress and makes all trees more susceptible to
insects and pathogens (Anderson etal. 1987, Edmonds etat. 2000, Seidel and Cochran
1981).Acontinuingwestern spruce budworm outbreakbegan in the Gotchen LSR in
1994 (Willhite 1999). Budworm defoliation may affectowl habitatsuitability both directly,
by reducing canopycover, and indirectly, by reducing habitatavailabilitythrough
increased tree mortalityand fire hazard.
Methods In thefirst phase ofthis study, we characterize changing vegetation patterns in the LSR
and investigate relations among these changing conditions, late-successional habitat,
and vulnerabilitytofire and budworm disturbanceatboththe stand and landscape scales.
Phaseoneaddressesthefirstgeneral question onthe potential consequencesofpassive
management. Some ofthe results from phase one are included in this paper. In the
second phase, wewill addressthethree remaining general questions (see "Issues") by
evaluating changes in untreated vs. treated stands overspace and time and assessing
the potential effects ofassociated landscape patterns on owl habitat, fire and budworm
hazard, and characteristics ofwood removals. Resultsfrom the second phasewill be
reported infuture publications.
—
Phase 1: Characterizing recent change We will use empirical and simulation
methodsto investigate stand-level structural and compositional changes in owl habitat
associated with budworm defoliation inthe currentoutbreak.Agrid ofrandomlyallocated
sample plots, installed in 1992 priorto the currentoutbreak, will be remeasured in 2000.
Plots are arranged in a systematic sampling grid nearthe Smith Butte owl activity center.
Each plot consists ofthree nested subplots and a down wood transect. The largest plot,
in which data will be collected on overstorytrees and snags >10 inches diameterat
breast height (d.b.h. or4.5feet), is a 40 basal areafactorvariable radius plot. Insidethe
variable plotaretwofixed plots: one 1/20th-acre and the other 1/50th-acre. In thefixed
plots, data on understory trees, snags <10 inches d.b.h., shrubs, herbs, and conifer
regeneration will becollected.A330-footdownwood transectwill include measurements
on dufflayerdepth and wood in all size classes (Brown 1974).
Theforestvegetation simulator(FVS) (formerly PROGNOSIS Stage 1973)will be used
with the stand visualization system (SVS) (McGaughey 1997) to portray structural and
compositional characteristics ofthe plots in 1992 and 2000. Thefire and fuels extension to
FVS (FFE-FVS) will be used to model changes in fire hazard between thetwo meas-
urementperiodsforthreefirebehaviorparameters: surfacefirebehavior,torching potential,
and active crownfire potential. The potential contribution ofdifferentstand attributesto
fire hazard will be modeled by using the FVS-FFE extension. Comparisons of 1992 and
2000 plotdata will be completed in 2001 and reported elsewhere. Resultswill be used to
calibrate the FVS base model and extensionsforlocal conditions to improve subsequent
model projections in phase 2.
Currentvegetation inthe LSRwill be mapped and characterized atthe mid-scale
(1:12,000)byusing photo-interpretation methodsand vegetation classifications
developed in the InteriorColumbia Basin Ecosystem Management Project(ICBEMP)
(Hessburg etal. 1999b). Vegetation patches resulting from photo interpretation will be
populated with both interpreted and derived attributes in ageographical information
system (GIS). Minimum patch size will be 10 acres. Two ofthese attributes, structural
8