Gap Characteristics and Vegetation Response in Coniferous Forests of the Pacific Northwest
Author(s): Thomas A. Spies and Jerry F. Franklin
Reviewed work(s):
Source: Ecology, Vol. 70, No. 3 (Jun., 1989), pp. 543-545
Published by: Ecological Society of America
Stable URL: http://www.jstor.org/stable/1940198 .
Accessed: 13/07/2012 12:14
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
.
Ecological Society of America is collaborating with JSTOR to digitize, preserve and extend access to Ecology.
http://www.jstor.org
June 1989
SPECIAL FEATURE-TREEFALL
GAPS AND FOREST DYNAMICS
543
however,fire(both naturaland human-set)is thelarge- the studysite and mightwell be different
in otherparts
scale disturbancemost frequentlyresponsiblefor the of the species range.
In South AmericanNothofagusforests,gap-creating
establishmentofeven-agedN. dombeyistands(Veblen
and Lorenz 1987, 1988).
disturbancesof various typesclearlyhave a major influence on forest structureand composition, and
knowledgeof disturbanceregimes is essential to exCONCLUSIONS
plaining vegetationpatterns.Despite the strikinginThe role of gaps of various sizes and originsin the fluence of coarse-scale allogenic disturbance on the
regenerationof Nothofagusdombeyivaries markedly compositionof theseforests,theinfluenceof fine-scale
along the transandeangradient.This variationchiefly autogenic processes must also be considered. For exresultsfromsite-to-sitedifferencesin associated tree ample, microsite modification by both understory
species and understorybamboos, as well as fromdif- bamboos (Veblen 1982) and canopy trees (Veblen et
ferencesin the abiotic environment.Otherwide-rang- al. 1979) may influencethe responsepatternsof Nothing tree species, such as Pseudotsuga menziesii in the ofagusspp. to gaps. The recentemphasis in forestecolPacificNorthwest(Spies and Franklin 1989), are also ogyon thepervasiveinfluencesofallogenicdisturbance
knownto responddifferently
to gaps in different
forest should not divertattentionfromautogenic processes,
types. In assessing the response of a given species to since mostvegetationchangereflectsbothtypesofprogaps, ecologistsshould considerthatthe responsemay cesses.
vary geographically.Gap response may be predicated
on the particularphysicaland biotic characteristicsof For reprintsof thisSpecial Feature,see footnote1, page 535.
Ecology, 70(3), 1989, pp. 543-545
? 1989 by the Ecological Society of America
GAP CHARACTERISTICS AND VEGETATION RESPONSE IN
CONIFEROUS FORESTS OF THE PACIFIC NORTHWEST
THOMAS
A.
SPIES
USDA Forest Service,PacificNorthwestResearch Station, Corvallis,Oregon 97331 USA
AND
JERRY
F.
FRANKLIN
USDA Forest Service,PacificNorthwestResearch Station and
of Washington,Seattle, Washington98195 USA
College of Forest ResourcesAR- IO, University
Gaps (sensu Watt 1947) in forestlandscapes assume
a wide rangeof sizes fromthe openingscreatedby the
death of singlebranches or treesto areas of hundreds
or thousandsof hectarescreatedby catastrophicwildfire.Most "gap" studieshave focusedon thedeath and
replacementof one to several canopy trees,excluding
larger sized gap processes that occur in most landscapes. However, the importanceof openingscreated
bydisturbancecannotbe adequatelyaddressedwithout
consideringboth fine-and coarse-scale gap processes.
Opportunitiesforexamininggap processesat a range
of spatial and temporalscales are excellentin the PacificNorthwestwhereextensive,uncutforestsstillexist
and stumps on many cutover areas retaina record of
previous treeages and firescars. Canopy disturbances
are ofvaried typesand spatial and temporalscales, and
intervalsbetweenstand-replacingdisturbancespoten-
tiallyexceed 1000 yron many sites (Dale et al. 1986).
In thisessay we focuson coniferousforestsdominated
by Douglas-fir(Pseudotsuga menziesii) and western
hemlock(Tsuga heterophylla).Our objective is to contrasttheroles and importanceof fine-and coarse-scale
disturbances,illustratingthe importance of various
disturbancecharacteristicsand systemcontextin controllingcommunitydynamics.
COARSE-SCALE
DYNAMICS
Wildfire,wind, and volcanic eruption are the primary agents that create canopy openings of 0.1 to
> 100 000 ha. Wildfirehas affectedthelargestarea with
the greatestfrequency.The complex fireregimeof the
regioncannot be characterizedby a singleestimateof
rotationlengthor intensity.Estimates of firerotation
lengthin the Cascade Mountains range from434 yr
544
SPECIAL FEATURE-TREEFALL
GAPS AND FOREST DYNAMICS
forstand-replacingfiresin centralwesternWashington
(Hemstrom and Franklin 1982) to 95 yr forlow- and
firesin centralwesternOregon (Morrihigh-intensity
son and Swanson, in press).
Aftercoarse-scale disturbancesuch as fireor clearcutting,total plant species richnessreaches maximum
values withina fewyears(Dyrness 1973, Halpern 1987)
and thendeclines as the forestcanopy closes, 30-40 yr
later(Long 1977). Tree species richnessis also highest
initiallyfollowingformationof largeopenings.For exin two
ample, of the 12 treespecies foundregenerating
clearcuts(Halpern 1987), 8 occurredin adjacent oldthere);
growthforests(althoughonly5 wereregenerating
theremaining4 species only occurredin nearbyrecent
clearcuts.Douglas-fir,a shade-intolerantpioneer(sensu Whitmore1975, 1982, 1989, Swaine and Whitmore
1988), typicallydominatesthe regenerationproducing
nearlypure, closed stands followingloss of early-colonizing,deciduous trees. More shade-tolerantspecies
can dominate young stands, especially in coastal and
northerlyareas. Douglas-firtypicallydoes not regeneratein small gaps in thesestands,and shade-tolerants
western
such as westernhemlock(Tsuga heterophylla),
redcedar (Thuja plicata), and Pacific silver fir(Abies
amabilis) invade and eventuallydominatethe canopy.
The size of coarse-scale events may influencethe
developmentand compositionof post-firestands. For
example, large fires -500 and 150 yr ago may have
limitedavailable seed sources of Douglas-firand westofmany
to delayfullrestocking
ernredcedarsufficiently
stands for decades (Franklin and Hemstrom 1981,
Klopsch 1985).
FINE-SCALEDYNAMICS
Fine-scale disturbances(<0.1 ha) break up the relativelyuniformcoarse-scale patches into a mosaic of
openings and canopy areas dominated by narrowcrowned, tall (70 + m) Douglas-fir or by broader
crowned,shorter(30-50 m) species such as hemlock
and redcedar.In the Cascade Range, > 70% of canopy
Douglas-firand westernhemlocktreesdie standingor
fromstem breakage (T. A. Spies and J. F. Franklin,
personal observation).Importantcauses of mortality
are root pathogens,wind, and bark beetles (Gedney
1981, Franklinet al. 1984). Mortalityrates of canopy
trees range from 0.5 to 1.0% (Franklin et al. 1984).
This translatesinto canopy-returnintervals of 100200 yr,whichare about thesame or shorterthancoarsescale events caused by fire.
Our observations suggestthat even the shade-tolerant tree species require canopy gaps to reach the
canopy in old-growthconiferousforests.Tsuga, the
most shade tolerantof the group,can regeneratewithout canopy gaps (Stewart 1986) but may eventually
requireone or moresmall openingsto surviveand grow
Ecology, Vol. 70, No. 3
into the overstory(T. A. Spies and J. F. Franklin,personal observation),as has been observed with some
shade-tolerantspecies in eastern deciduous forests
(Canham 1985, 1989). Small seedlingsand veryslowgrowingsaplingsofseveralothershade-tolerantspecies
can be found under intactcanopies.
We hypothesizethatthe natureand role of fine-scale
gaps changes over time followinglarge-scale disruption. In young Douglas-firstands with dense, closed
canopies, gaps createdby the normal thinningprocess
and pathogens(Childs 1970) are generallysmall and
ephemeral. Such gaps are closed primarilyby lateral
branch growth. Shade-tolerantspecies invading the
deeplyshaded understoryof the stand duringthistime
may have increased survivorshipand growthrates in
these small openings.As the Douglas-firstand develops, canopy heightincreases,and understorylightincreases as numerous small openings form between
crowns. Canopy gaps of one to several trees may be
less importantto invading shade-toleranttreesat this
timebecause thegeneralcanopytransmitsenoughlight
for survival and growth. Eventually shade-tolerant
species, such as hemlock and redcedar,dominate the
canopy and transmittanceof lightthroughthe canopy
declinesagain. When thishappens,canopy gaps of one
to several trees may become more importantin the
survival and growth of understoryseedlings of the
shade-tolerantspecies.
Where minimum lightlevels occur in canopy gaps
or under low-densitycanopies, seedling survivorship
and growthis probablycontrolledby micro-scalegaps
(<10 M2) in the herb/shrub,litter,and rootingzone
layers of the forest.For example, western hemlock
seedlingsare almost exclusivelyfoundon rottenwood
substratesthatare elevated above the forestfloor(Minore 1972, McKee et al. 1982). Christy(1986) found
that trenchingincreased the growthof hemlock seedlings more than removal of foliage above seedlings.
Harmon (1986) found that hemlock and Sitka spruce
(Picea sitchensis)regenerationwas inhibitedon theforest floorby competitionfrommosses and herbs.Competition with shrubs and herbs and lack of suitable
seedbeds may help explain why many gaps in oldgrowthDouglas-firforestsfillslowly,with many existingfor> 50 yrwithoutdevelopmentoftreesaplings.
Douglas-fir,unlike shade-intoleranttrees in other
forestecosystems(Brokaw 1985b, Runkle 1982), does
not reach the canopy in gaps less than -700-100O m2
in the WesternHemlock Zone (Franklinand Dyrness
1973). At least two mechanisms may be responsible
forinhibitingregenerationand growthof Douglas-fir
in mostcanopy gaps. First,lightlevels may be too low.
The combinedcanopystrataofDouglas-firand western
hemlock typicallyreduces lightin the understorybeneath intact canopy to <5% of full sunlight.Hence,
June 1989
SPECIAL FEATURE-TREEFALL
unlikeEasterndeciduous forests(see Poulson and Platt
1989), the additional light enteringthroughgaps of
even several treesmay not be enough forregeneration
of shade-intolerantspecies. Second, since most trees
die standingor bytop breakage,mostcanopygap events
do not formgaps in the lower strata of the forestor
expose mineralsoil which Douglas-firgerminantsprefer to organic seedbeds (Minore 1979). Gaps with
standingdead trees also transmitless lightthan gaps
formedby fallentrees.
The importanceof fine-scalegaps appears to differ
locallyand regionallyamong foresttypes.For example,
Douglas-firreproducesand reachesthecanopy in small
gaps on drysitesin theWesternHemlock Zone (Means
1982) and in the Mixed Coniferand Mixed Evergreen
zones in southwesternOregon (Franklinand Dyrness
1973). These forestshave shortercanopies than Douglas-fir/western
hemlockforestsand lack thedense canopies of westernhemlock. Consequently,lightlevels
in gaps may be high enough forDouglas-firregeneration. These foresttypesare also subjectedto relatively
frequent,patchy,low-intensity
fireswhichwould create
gaps in the overstoryand understoryvegetationand
expose mineral soil seedbeds favoringDouglas-firregeneration.
GAPS AND FOREST DYNAMICS
545
notmatchDenslow's prediction(1 980b) that,in forests
wheresmall gaps were more common than largegaps,
species able to regeneratein small gaps will be more
common than species able to regeneratein largegaps.
The observed pattern of species composition in
Northwesternconiferousforestsprobablyresultsfrom
mostsmall gaps beingformedby pestsand windbreakSmallage,whereaslargegaps are typicallyfire-created.
gap processes typicallycreate gaps only in the canopy
layer and release relatively little above- and belowgroundresources,whereaslargefirescreategaps in
nearlyall layersof the forest,releasingmore resources
and creatinga diversityof microsites.Shade-tolerant
species are not restrictedto small gaps because suitable
regenerationsites are available in most large,fire-created openings. Clearly, predictive models of species
composition based on gap dynamics must take into
account disturbancetypealong withgap size.
While gap dynamics at the scale of one to several
trees are important,finerscale processes at the level
of the seedling microenvironmentand coarser scale
(landscape) processesalso play importantrolesin communitydynamics.Lower vegetationallayersinfluence
treeregenerationin Douglas-firand Nothofagusforests
(Veblen 1989). At a coarser scale, large disturbances
may locally reduce seed sources of some species, affectingrates of succession and composition in the reCONCLUSIONS
generatingforest.At regional levels, variation in the
In tall coniferousforestsof thePacificNorthwest,as response of Douglas-firto small gaps illustratesthe
in temperatehardwood forests(see Lorimer 1989), importanceofcommunityand environmentin gap dyspecies composition appears to be molded by gaps of namics. SimilarfindingsforNothofagusin Chilean fora varietyof sizes and types. Large, relativelyuniform ests (Veblen 1989) suggestthatgeographicvariationin
patches createdat infrequentintervalsby fireare con- species responses to gaps may be common in many
vertedto a mosaic of canopy and gap by smaller,more forestregions.
frequentlyoccurringgap processes. In the Western
ACKNOWLEDGMENTS
Hemlock Zone, where fine-scalegaps are at least as
common temporallyand spatiallyas large gaps, most
We thankA. McKee, M. Harmon, C. Halpern, B. Smith,
tree species can regeneratein the large, fire-created D. Hibbs, and F. Swanson fortheirhelpfulcommentson an
draftof this essay.
gaps, but less than half of the tree species regenerate earlier
in smaller gaps formedby otherprocesses. This does For reprintsof thisSpecial Feature,see footnote1, page 535.