Dendroecological and Ecophysiological Analysis of Gap Environments in Mixed-Oak
Understoreys of Northern Virginia
Author(s): D. A. Orwig and M. D. Abrams
Reviewed work(s):
Source: Functional Ecology, Vol. 9, No. 6 (Dec., 1995), pp. 799-806
Published by: British Ecological Society
Stable URL: http://www.jstor.org/stable/2389977 .
Accessed: 11/04/2012 13:25
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 [email protected].
British Ecological Society is collaborating with JSTOR to digitize, preserve and extend access to Functional
Ecology.
http://www.jstor.org
Functional
Ecology 1995
9,799-806
Dendroecological
and ecophysiological
analysisofgap
in mixed-oakunderstoreys
environments
ofnorthern
Virginia
D. A. ORWIG* and M. D. ABRAMS
ThePennsylvaniaState University,
School ofForestResources,4 FergusonBuilding,University
Park,PA
16802, USA
Summary
1. Ecophysiological,morphologicaland dendroecologicalattributes
were examined
in species growingin understorey
in mixedand newly formedgap environments
Quercusforestsof northern
Virginia.
2. Gap environments
werecharacterized
byhigherphotosynthetic
photonfluxdensity
fromnon-gapcontrolplots in theirsoil moisture,
(PPFD) but were not different
predawnleafwaterpotential(v), leaftemperature
or leafto airvapourpressuredeficit
differences
existedbetweenregeneration
(VPD). In addition,fewsignificant
responses
at gap and controlsites.
3. Plantsin gaps experiencedhigherratesof netphotosynthesis
(A) and leaf conductanceto watervapourdiffusion
control
than
plants,
although
parametersdid not
(gwv)
differ
betweenearlyvs late successionalspecies.Gap plantsalso producedleaves with
smallerleaf area and greaterleaf thickness,leaf mass per area (LMA) and stomatal
densitythancontrolplants.
4. All gap speciesexperiencedacceleratedheightgrowthcomparedto non-gapplants,
withearly successionalSassafras albidum (Sassafras) and Liquidambarstyraciflua
(Sweetgum)exhibitingthehighestrates.In contrast,averageradial growthin small,
different
between habitattype, although
understoreytrees was not significantly
to themostrecentgaps.
Quercusalba (WhiteOak) did respondsignificantly
5. Many smalltrees(6-12 cm d.b.h.)weresurprisingly
old - up to 100 yearsin age.
Despite severesuppressionperiodsin thesetrees,theyexperiencedreleaseeventsduring theirlifetime,
althoughnotnecessarilyto themostrecentgaps. This maypartially
betweenecophysiologicaland growthparameters.
explainthelack ofcorrelation
radialgrowth,successionalstatus,winddisturbance
Key-words:Photosynthesis,
FunctionalEcology (1995) 9, 799-806
Introduction
?C1995British
Ecological
Society
tion, studies examiningspecies in contrastinglight
environmentsincludinggaps have reportedaltered
Windis oneofthemostfrequent
causesoflarge-and
leaf morphologyin sun leaves owing to acclimation
andis
small-scale
disturbance
affecting
eastern
forests
such as thickerleaves, higherstomataldensityand
inthedetermination
in
an intrinsic
factor
ofstructure
smallerleafarea (Carpenter& Smith1981; Abrams&
manyforesttypes(Bormann
& Likens1979;Foster
Kubiske 1990). The degreeof physiologicaland phe1988).In general,
creation
ofsingleandmultiple-tree
notypicresponsemay varybetweenearlyvs late sucleadsto a widearrayofmicroenvicanopyopenings
cessional species. HigherA, gwvand respiration
rates
ronmental,
ecophysiological
andcommunity
changes.
and more xerophyticleaves have been recordedin
The few studieswhich examinedphysiological
early vs late successional species (Bazzaz 1979;
responsesto speciesgrowingin temperate
canopy
Abrams 1988; Koike 1988; Kloeppel, Abrams &
rates(A)
gapshavereported
increased
photosynthetic
Kubiske 1993), because late successionalspeciesmay
andleafconductances
(gwv)(Wallace& Dunn1980; exhibitmorelimitedphysiologicalplasticity(Bazzaz
Reichet al. 1990;Ellsworth
& Reich1992).In addi& Carlson 1982).
Few studieshave examinedcanopy gap processes
*
Presentaddress:HarvardForest,HarvardUniversity, and colonizationdynamicsfollowingwindthrowin
MA 01366-0068,
USA.
Petersham,
oak forests(Clebsch & Busing 1989)
second-growth
799
800
D. A. Orwig &
M. D. Abrams
?C1995British
EcologicalSociety,
Functional
Ecology,
9,799-806
formulafor an ellipse (a = 7rLWI4,whereL is the
lengthof the longeststraightline thatwill fitin the
gap and W is thelengthof thelongestline in thegap
thatis perpendicular
to L) (Runkle 1992). Gaps were
definedas canopyopenings(>15 M2) createdby the
death of one or more trees (gap-makers)and were
delimitedby verticallyprojectingthecanopyopening
to thegroundsurface.Adjacentnon-gapplotsofequal
area wererandomlyestablishedwithinthesame stand
at a randomdistancesufficiently
faraway to exclude
the lighteffectsof thegap (i.e. > 20 m). Each establishedwindthrow
plotwas inventoried
forgap-maker
species and damage type,while bothplot typeswere
sampledforsaplingand seedlingdensitiesand shrub
and herbaceouscover(%).
Growthresponsesforthe last 3 years were determinedfora subsampleof saplingsencounteredin the
plotsby measuringthedistance(cm) betweenterminal bud scars(cf.Runkle1992). In addition,treecores
were obtained0.5 m above thegroundfromsaplings
and small trees(d.b.h. < 18 cm) whenpresentwithin
gaps and control plots to examine radial growth
and to ascertainthe
responsesto recentgap formation
extentof past disturbances.All cores were dried,
mounted,sanded and aged witha stereomicroscope.
weremeasuredto thenearAnnualgrowthincrements
Materialsand methods
est 0.01 mm with a tree-ringmeasuring device
(Regents InstrumentsInc., Quebec, Canada) and
STUDY SITE
recordedusing the MacDendro microcomputer
proThe study was conducted within Chancellorsville gram.Cores were examinedforrecentabruptgrowth
Battlefield,one of fourCivil War battlefieldsassocireleases accordingto Lorimer& Frelich(1989). The
with
ated
the Fredericksburgand Spotsylvania
10 oldest overstoreyQ. alba cores takenpreviously
NationalMilitaryPark,located near Fredericksburg, fromthe sampledsites (cf. Orwig & Abrams 1994a)
VA (380 15 'N, 77 0 37 'W). Mean annualtemperatures were used to develop a ring-width
index or disturbance history.Cores were visuallycross-dated,meaaverage 13 0C, while winterand summerdaily mean
are 2 0C and 23 0C, respectively(Elder
suredas above, standardizedvia linearregressionand
temperatures
divisionof each measuredringwidthby thevalue of
1985). Precipitation
averages 102 cm annuallyand is
well distributed
theyear.The studyarea is
thefittedline (Fritts& Swetnam1989), and averaged
throughout
located withinthe PiedmontPlateau physiographic to construct
thesitechronology.
and
consists
of
province
gentlyrollingtopography,
elevationsof60-90 m above sea level (US Geological
GAS EXCHANGE AND PLANT-WATER RELATIONS
Survey 1984) and deep, well-drainedsilt loam soils
Three gaps rangingin size from77 to 171 m2 and
(Elder 1985). Local overstorey vegetation was
assessed priorto the wind disturbanceand consisted
threenearbycontrolplots of equal dimensionswere
ofa mixtureof Quercusalba L. (WhiteOak), Quercus
selected forecophysiologicalstudy.Each of the six
coccinea Muenchh. (Scarlet Oak), Liquidambar
plotscontainedfourto sevenunderstorey
seedlingsof
L. (Sweetgum)and Pinus virginianaMill.
styraciflua
Q. alba, Sassafras albidum(Nutt.)Nees (Sassafras),
and Nyssa sylvaticaMarsh. (Blackgum). Duringthe
(VirginiaPine) (Orwig& Abrams1994a).
1993 growingseason,predawn(0500 solar time)and
diurnal(every3 h from0730 to 1800 solar time)leaf
GAP SAMPLING
water potential(v) measurementswere conducted
with a pressure chamber (PMS InstrumentCo.,
Duringa 2 week periodin July1990, several severe
thunderstorms
with winds in excess of 100 km h-'
Corvallis,OR) on one leaffromfiveseedlingsofeach
resulted in many single-treewindthrowsof P.
species for each habitattype.Diurnal gas-exchange
virginiana and various Quercus species within measurements
were made fourtimes(9 and 29 June,
14 Julyand 10 August) duringrelativelycloud free
ChancellorsvilleBattlefield(NOAA 1990). During
mid-summer1993 (the thirdgrowingseason followconditionswitha portablephotosynthesis
unit(LCAing thedisturbance),
3, ADC Ltd, Herts.,UK). Following the final sam26 gaps withina 20 ha portionof
plingdate of 1993, leaf-structure
the battlefieldwere measuredfor area by using the
measurements
were
and thereis a scarcityof information
on theage structureof saplings and small understorey
trees (Ross,
Sharik& Smith1982) and theirabilityto persistand
respond to canopy gaps. We also know very little
abouthow sympatric
species of differing
successional
status respond physiologicallyto gap formation.
Therefore,thisstudyassessed the role of windthrow
disturbanceon thedynamicsof Quercus (Oak) stands
of northernVirginia by linking ecophysiological,
morphological and dendroecological attributesof
speciesgrowingin understorey
and newlyformedgap
environments.
We addressedthe followinghypotheses in thisstudy:(1) newlycreatedtreefallgaps will
experiencehigherlevels of light,soil moisture,and
space availabilitythancontrolplots; (2) plantsin gap
will exhibitphenotypicplasticityto the
environments
lightenvironment
resultingin greaterphotosynthetic
and conductanceratesthancontrolplants;(3) physiological responsesto canopy gaps will varybetween
earlyvs late successionalspecies, withearlysuccessional species exhibitinggreaterstimulation;(4) leaf
structure
and physiologywill be relatedto observed
growthratesin gap and non-gapplantsand between
earlyvs late successionalspecies.
801
Dendroecological
analysisofgaps
made on fullyexpandedleaves fromfourindividuals
of each species fromeach plotfollowingthemethods
of Abrams & Kubiske (1990). All statisticalmean
comparisonswere accomplished using analysis of
varianceor generallinearmodels proceduresin SAS
(SAS Institute
Inc. 1985).
Resultsand discussion
TREEFALL AND GAP CHARACTERISTICS
A totalof45 treesweredamagedduringtheJuly1990
withinthe26 gaps examined.The domiwindstorms
nant mode of treefallwas bole-snap (31 trees) followed by uprootedtrees and snags (nine and five
trees,respectively).Pinus virginianaand Q. coccinea
comprised89% of damaged trees and trees which
of downed treessugwere uprooted.The orientation
geststhatprevailingwindswerefromthenorth-west,
as 88% of stemsfell in a south-easterly
(95-180?)
direction.Nearly90% of sampledgaps were formed
by one or two gap-makersand threegap-makerswas
themaximumnumberencountered.Canopy gap size
rangedfrom31 to 196 m2witha mean of 95 m2(? 9
was 38.9 cm (?
m2) and themeand.b.h.ofgap-makers
1.2 cm, range 18.7-55-8). Gap area was correlated
with numberof gap makers (R = 0.49, P < 0.01,
Spearmanrankcorrelation)whilethenumberof gapmakerswas negativelycorrelated(R = -0 38, P <
0-05, Spearmanrankcorrelation)withaverage d.b.h.
ofgap-makers,
gaps may
suggestingthatmultiple-tree
have resultedfromlargetreesknockingdown smaller
wereclassifiedas hinge
trees.All uprootsencountered
treefallsaccording to Beatty & Stone (1986) and
areas rangedfrom4 to 19 m2withan averpit/mound
age of 10 ?2 M2.
MICROENVIRONMENTAL
AND ECOPHYSIOLOGICAL
RESPONSE
?C1995British
EcologicalSociety,
Functional
Ecology,
9,799-806
Soil moistureand predawnleaf-waterpotentials(v)
between gap and controlhabitatswere not signififorany of the samplingdates (Table
cantlydifferent
1), which is consistent with some gap studies
(Ellsworth& Reich 1992; Kloeppel et al. 1993) but
not others (Abrams 1986; Abrams & Mostoller
1995). Mean diurnal v was significantlylower
(more negative) in gap plants and differedamong
species (Table 1). Quercus alba maintainedthelowest values in gap and controlplots (-1.54 ? 0.06 and
-1 29 ? 0 05 MPa, respectively)while S. albidum
exhibitedthehighestleaf v (-0.66 ? 0 04 and -0 65
? 0.04 MPa, respectively).Gaps received significantlyhigherPPFD than controlsites on all samand
plingdates and times;however,leaf temperature
VPD were not significantly
different
betweenhabitats.
existedamong
No significant
(P < 0.05) differences
species' daily gas-exchangemeans withinhabitatsor
Table 1. Mean seasonal microenvironmental
and pooled
diurnalgas-exchangeparameters
forspecies growingin gap
and controlplots followingdisturbancein mixed-Quercus
forestsof northern
Virginia
Parameter*
Gravimetric
soil
moisture(%)
PPFD (gmol m-2s'l)
Predawn
v (MPa)
' (MPa)
Meandiurnal
Leaftemperature
(0C)
VPD (kPa)
A(gmol
m-2 s'l)
(molM2
Amax
m s)
gwv(mmolm-2 s-1)
Gap
19.4 ? 1.3a
479? 24a
0. 19 ? 0.01a
Control
18.9 ? 1.4a
110? 6b
-0 21? 0.Ola
-1.03? 0.04a -0 91 ? 0 03b
28.5? 0.2a
28.3? 0.2a
1.24? 0.03a
1.30? 0.04a
5.10 ? 0.18a
1148 ? 1.01a
284 ? Ila
1.63? O.lob
6.43? 0.34b
202 ? 7b
* VPD = leafto airvapourpressuredifference;
A andAmax=
mean and maximumnetphotosynthetic
rate;gwv= stomatal
conductanceto watervapour.Values of thesame parameter
followedbythesame letterare notsignificantly
different
(P
< 0.05, t-test).
withinplots on any samplingdate,therefore
analysis
of variance was performedon pooled species data
betweenhabitattype.Plants growingin gaps exhibited significantly
highermean and maximumphotosyntheticrates,as well as higherconductancerates,
thanshadedunderstorey
plants(Table 1). Highergashave
exchangeratesin gap vs non-gapenvironments
been previouslyreported(Bazzaz 1979; Wallace &
Dunn 1980; Abrams1988; Ellsworth& Reich 1992).
However,severalstudiesalso reportedspecies differences in bothhighand low lightenvironments
which
were not observed here (Bazzaz & Carlson 1982;
Walters,Kruger,& Reich 1993; Abrams& Mostoller
1995). One potentialexplanationforthe lack of sucin thisstudyis the greaterthan
cessional differences
expectedphysiologicaland morphologicalplasticity
of N. sylvaticaobservedin thisstudy.These findings
are consistentwiththereputation
of N. sylvaticaas a
droughttolerant,gap-facultativespecies that can
utilizesmallgaps (Latham 1992; Orwig&
effectively
Abrams1994b).
All threespecies adjustedmorphologicallyto gap
environments
by producingsignificantly
(P< 0-05, ttest) smallerleaf area and greaterstomataldensity,
leaf thicknessand leaf mass per area than control
plants. Sassafras albidum exhibitedthe largestleaf
areas (66-103 cm2), N. sylvaticahad thelargestguard
cell lengths(32.9-33.6 gim)and Q. alba displayedthe
highest stomatal densities in both habitat types
(482-584 mm2) (P < 0 05, Tukey's multiplecomparison test).In bothhabitatsN. sylvaticahad thethickest
leaves (127-165 gim)while Q. alba had the thinnest
leaves (90-110 gim).This suggeststhatbothearlyand
late successionalspecies can displaya highdegreeof
plasticityin responseto changingenvironmental
conditions(Carpenter& Smith1981; Abrams& Kubiske
1990).
802
D. A. Orwig&
M. D. Abrams
Table 2. Summaryof 3 yearaverageheightgrowthrates(cm year-'+ SE) followingdisturbanceof saplings(n) sampledin 26
Virginia.Maximumand minimumvalues are average 1 yearvalues
gap and controlplotswithinmixed-oakforestsof northern
(cm ? SE)
Control
Gap
Species
(n)
(3 year)
Max
Min
(n)
(3 year)
Max
Min
Acer rubrum
Liquidambarstyraciflua
Nyssasylvatica
Quercusalba
Sassafras albidum
16
19
20
5
16
15.2?1.6a
30.0?5.2c
15.5 ? 1.6a
223?26
383?64
18.1 ?1?9
305?83
46.3 ? 6 1
8 1?09
212?44
12 9 ? 1 4
219?64
24.8 ? 3 9
15
12
27
2
4
9.9? 1.3b
122?36b
13 9+1-8
160+41
13 6 + 1 4
133+33
17 1 + 4 4
7.7?1 3
75?27
8.3 ? 1.2
95?20
10.2 ? 2 8
26.0?6.9ac
33.7 ? 4.3c
11.0 ? 1.2b
111? 2.3ab
13 6? 35 b
(P < 0.05, t-testbetweenhabitatsand
different
Values in a row or a column withthe same letter(s)are not significantly
Tukey's multiplecomparisontestamongspecies).
HEIGHT AND RADIAL GROWTH RESPONSE
Growthparameterswere measuredon a largernumber and wider array of species than physiological
data. In contrastto gas-exchangedata,thereweredifferencesin heightgrowthamongspecies and between
habitats. All species exhibited accelerated height
growthin gap vs controlplots (Table 2). In addition,
early successional S. albidum and L. styraciflua
exhibited3 year average heightgrowthvalues near
30 cm in gaps while late successionalA. rubrumand
N. sylvaticaaveraged approximately15 cm. Height
amongspecies
growthwas notsignificantly
different
in controlplots. All species achieved average maximum 1 yearheightgrowthratesof at least 18 cm in
gaps followingdisturbance,and several individuals
of S. albidumand L. styracifluadisplayedratesnear
1 m per year. It should be noted thatrelatingtree
growthto photosynthesis
has been problematicowiing to thecomplexityof carbonallocationpatternsin
plants (e.g. reproduction,above- vs below-ground
biomass, and leaf chemistry)(Canham & Marks
1985; Fitter& Hay 1987)
A totalof 93 saplingsand small understorey
trees
rangingin age from24 to 108 yearsold werecoredin
gap and controlplots.In bothhabitats,A. rubrumand
L. styracifluadisplayed the highest radial growth
rates,while Q. alba and N. sylvaticaexperiencedthe
lowest rates (Table 3). Mean radial growthwas not
betweengap and controlplots
different
significantly
However,Q.
forany species priorto gap formation.
alba was the only species respondingto recentgaps
withacceleratedradial growth(Table 3). Our results
contrastwith those found in northernhardwood
forests,whereshade tolerantAcer saccharumMarsh.
(Sugar Maple) and Fagus grandifolia Ehrh.
(American Beech) saplings exhibited accelerated
radial growthas a resultof canopy gaps (Canham
1988; Tryon,Lanasa & Townsend 1992; Poage &
Peart 1993). Moreover,we foundthatearly successionalL. styraciflua
and S. albidumdid notrespondto
recentgaps. Radial growthresponsesmaynotemerge
until several years following disturbancein suppressed trees,because theytypicallyuse photosynthatefor foliage productionimmediatelyfollowing
release, followed by heightand thenradial growth
(Oliver& Larson 1990).
of saplingsand small trees(n) sampledin 26 gap and controlplots followingdisturTable 3. Radial growthcharacteristics
Virginia
bance withinmixed-oakforestsof northern
Control
Gap
Species
A. rubrum
L.styraciflua
N. sylvatica
Q. alba
S. albidum
(C 1995British
EcologicalSociety,
Functional
Ecology,
9,799-806
(n)
5
11
9
21
2
Average
growth
(cm ? SE)
%
%
%
Major Moderate Recent
release release release
0.90?0-13a
20
9
33
5
60
18
67
52
0
0
24
0
0
0
0.83?0.08a
0.57 + 006b
0.53+ 003b
072
0 10ab
0
(n)
5
12
7
9
5
%
Major
release
%
%
Moderate Recent
release release
0.69?012ab
053 + 003b
0-61+ 003ab
10
8
29
0
40
33
86
33
0
0
0
0
0 65+ 003ab
0
10
0
Average
growth
(cm ? SE)
0.83?0.
la
different
(P < 0.05, t-testbetweenhabitatsand Tukey's
Values in a rowor columnwiththesame letter(s)are notsignificantly
multiplecomparisontestamongspecies).
majorand moderatereleasesweredefinedaccordingto Lorimer& Frelich(1989). A recent
The percentageofcoresexhibiting
release is at least a 100% increasein the mean radial growthof the 3 yearsfollowinggap formationcomparedto the mean
growthofthe 10 yearspriorto thedisturbance.
803
Dendroecological
analysisofgaps
Table 4. Summaryof seedlingdensities(numberha-' ? SE) and groundlayervegetationsampledin 26 gap and controlplots
withinmixed-oakforestsof northern
Virginia
Gap*
(n = 26)
Species
Acer rubrum
Amelanchierarboreum
Castanea pumila
flexopaca
Liquidambarstyraciflua
Liriodendrontulipifera
Nyssasylvatica
Pinus virginiana
Prunusspp.
Quercusalba
Quercuscoccinea
Quercusfalcata
Quercusstellata
Rhamnuscathartica
Sassafras albidum
Others
Totals
1918 ? 585a
1297 ? 354a
138 ? 89a
442 ?229a
3919 ? 2035a
1336 ? 552a
2360 ? 466ab
911 ? 420a
1753 ?899a
1325 ? 238a
5214 ? 910a
552 ?156a
248 ?159a
Herb cover (%)
Shrub cover (%)
1451 ? 603a
572 ? 384a
0
245 ? 245a
2416 ? 1210a
5040 ? 2093b
1259 ? 678ab
1209 ? 620a
0
392 ? 269a
1352 ? 664b
0
0
Gap mound
(n = 9)
1453 ? 473a
985 ? 709a
294 ? 294a
780 ? 524a
1604 ? 807a
2298 ? 1570a
495 261l
1875 ? 1324a
1223? 1015a
844 ? 397a
3270 ? 1421ab
297 ?152a
7535 ? 1434a
1642? 148
3611 ? 726a
0
0
347 ?347a
5108 ? 2135a
0
28848? 3326a
13539 ?3841b
22294 ?5545ab
Oa
8.23 ? 0.5a
7.37 ? 3.0a
44.9 ? 4.2a
Richness
Gap pit
(n = 9)
0
4.22 ? 0.9c
4.37 ? 2.4a
3.06 ? 1.5b
5.78 ? 10b
1.67 ? 0.4a
16.38 ? 6.9b
Control
(n = 26)
4444 ?908b
1684 ? 376a
83 ? 61a
276 ? 95a
3353 ? 1705a
814 ? 327a
3436 ? 797a
497 ? 238a
1642 ?723a
4206 ? 993b
2470 ? 435ab
277 ?98a
276? 113a
55 ? 55a
7645 ? 1905a
1564? 145
31139 ?3419a
7.85 ? 0.5ab
6.14 ? 2.7a
41.4 ? 3.9a
* Excludingpitand mounddensities.
different
Values in a rowwiththesame letterare notsignificantly
(P< 0.05, Kruskal-Wallistest).
(Fig. 2a). In contrast,a 92 year-oldQ. alba experienced an earlymoderaterelease in 1911 followedby
an extendedperiod of low growthuntila moderate
A. rubrumand N.
A high percentageof understorey
release
in 1974 (Fig. 2b). This treedid not respond
sylvaticain bothenvironments
experiencedmajorand
to the gap after1990. Following a major
favourably
moderatereleases lastingat least 10 years (Table 3),
in
release
1960,
a 50 year-oldA. rubrumdisplayed
which is consistentwith radial growthresponses
low
radial
growth
for15 years,experienceda moderreportedin overstoreyindividuals of these shadeate
release
in
and exhibiteda spurtofgrowthfol1984
tolerantspecies (Mikan, Orwig & Abrams 1994).
A N. sylvaticatreethatbecame
lowing
1990
(Fig.
2c).
To assess whetherthesereleaseswereassociatedwith
established
the
during
early 1900s experienced a
or localized tree-bypreviousstand-widedisturbances
in
major
growth
release
1921 which lasted for 20
treeeventswe examinedthehistoricaldevelopmentof
A
years(Fig. 2d). periodofdeclininggrowthfollowed
thestudyarea (Fig. 1). A moderatereleaseobservedin
untilit was releasedagain in 1963. Figure2e shows a
thesitechronologyduring1894 was associatedwitha
75 year-oldL. styracifluathathad early favourable
concomitantperiod of Quercus and N. sylvatica
radial
growthfollowedby a 30 yearsuppressedperiod
recruitment
seen in the overstoreyand understorey
and
then
a moderatereleasein 1959. ManyL. styracirecruitment
figures.A second pulse of understorey
trees
exhibiteda characteristic
flua
patternofhighiniconsistingprimarilyof Quercus and L. styracifua
2
mm
tial
and thendeclining
growth
exceeding
year-'
occurredfromthe late 1940s to the early 1960s and
radial
growth
over
time
with
no
recent
gap response
coincidedwitha periodof graduallyincreasingringtheir
(Fig.
2f).
successional
Despite
early
status,L.
widthindex.These resultsprovideevidenceofcontinstyraciflua
and
alba
endured
of
Q.
protracted
periods
uous understorey
recruitment
sincethe1890s. In addiThis
suppression.
indicates
that
most
underanalysis
basal area and stem
tion,therelativelylow overstorey
storeytrees were capable of experiencingrelease
densityof thisxeric site may have resultedin condievents
nature
duringtheirlifetime.The asynchronous
tionsthatpermittedunderstorey
persistenceof these
of
release
dates
that
disturbances
were
localsuggests
species.
ized
and
20
30
occurred
at
to
year
intervals.
Examples of radial growthindicateindividualistic
patternsamongtheovertoppedspecies (Fig. 2). A 60
year-oldQ. alba exhibitedlow initialradial growth
GAP REGENERATION
<0.5 mm year-', a moderaterelease in 1950 and a
dramatic increase following the 1990 disturbance Threeyearsafterthewind disturbancetherewere no
DISTURBANCE
DEVELOPMENT
?C1995British
EcologicalSociety,
Functional
Ecology,
9,799-806
FREQUENCY
AND HISTORICAL
3
804
D. A. Orwig&
M. D. Abrams
,,
2
2
n= 10
Overstorey
1894
-
0
n 80
n=
80
5050
F4040
-~~~
~
0
I
~~~~Overstorey
S
ran] R
~ 4iZ
~
~
Q. alba
B~~1
~0 L. styracfflua
+ N.sylvatica
Q. coccinea
Q. falcata
0 S. albidum
A A. rubrum
Other
X[Q
U
~0
E
co
20-
O
IN
10-
n = 93
Gap-understorey
16a
E
+
ci)
40.
1850
L. tulipiferaand P. virginiyears. Light-demanding
ana benefitedmost fromthese disturbedmicrosites,
which did not containthe prominentshrubspecies
Aiton (Blueberry)
such as Vacciniumangustifolium
and Gayluccacia baccata (Wang.) K. Koch
(Huckleberry),common in gap and control sites.
Because shrubcover was negativelycorrelatedwith
gap seedlingdensity(R = -0 67, P <0.001) itmaypredespitegap
clude significantseedlingestablishment
formation
(Huenneke1983).
I
1900
1950
Yearoforigin
2000
andunderstorey
age-diameoverstorey
Fig. 1. Present-day
and present
terrelationships
showingdateof recruitment
indexofthe10
diameters
of sampledtreesandring-width
standsinnortholdestoverstorey
Q. alba inmixed-Quercus
ernVirginia.OtherspeciesincludeCaryaglabra (Mill.)
Sweet (PignutHickory),Liriodendron
(Tulip
tulipifera
Poplar),flex opaca Ait. (AmericanHolly), Q. stellata
Wanghen.
(PostOak) and Q. velutinaLam. (BlackOak).
The asteriskindicatesa moderate
release,as definedby
Lorimer
& Frelich(1989).
Conclusions
in gap and non-gap
This studyevaluateddifferences
leafphysiology,
leaf morphology,
microenvironment,
heightand radial growth,and populationdynamics.
we foundsurprisingsimiFrom thesemeasurements
VPD and soil moisture
laritiesin leaf temperature,
between gap and non-gap environments.
Similarly,
differences
amongspecies in
we foundno significant
gas-exchange parametersin either habitat despite
clear leaf morphologicaldifferencesamong species
and betweenlightenvironments.
EarlysuccessionalS.
albidum and L. styracifluaexhibitedrapid height
growthin gaps and maintainedmoderateheightand
radial growthin controlplots, whereas late successionalN. sylvaticaexhibiteda highdegreeof physiological and morphologicalplasticityin response to
The dendroecologicaltechniquesused
gap formationin this study increased our understandingof gap
dynamicsand understoreydevelopmentin secondand individualtree-ring
growthforests.Age structure
chronologiesfrom this site provided evidence of
small-scale disturbancesat 20 to 30 year intervals,
since the
continuousunderstorey
species recruitment
1890s and remarkablepersistenceof certainspecies,
suchas N. sylvaticaand Q. alba, whichwerestill< 10
cm d.b.h. after70-90 years. This multi-disciplinary
approachappearsto be a promisingtool forexamining successionaltrendsand gap eventsin a varietyof
foresttypes.
Acknowledgements
The authorsthankChad Spencer and JayWestover
forassistingin fielddata collection.In addition,we
thankJohnKarish and Mike Johnsonforproviding
co-ordinationand technicalsupportto conductthis
study.This researchwas financiallysupportedby the
NationalParkService (CA-4000-9-8004,supplemental agreements1 and 14).
(C 1995British
EcologicalSociety,
Functional
Ecology,
9,799-806
significantdifferencesin average sapling densities
between gap and controlplots. Nyssa sylvatica,L.
andA. rubrumexhibitedhighsaplingdenstyraciflua
sities in both plot types,while S. albidum densities
were significantlyhigher in gap habitats. Total
seedlingdensitiesand speciesrichnessweresimilarin
lower in
gap and controlplotsbut were significantly
References
within-gapmicrosites(pits and mounds) (Table 4).
Sassafras albidum, Q. coccinea and L. styraciflua Abrams,M.D. (1986) Physiologicalplasticityin waterrelations and leaf structureof understoryvs open-grown
densitieswereabundantacrossall micrositeswhereas
Cercis canadensis in northeasternKansas. Canadian
A. rubrumand Q. alba densitieswere significantly
JournalofForestResearch 16,1170-1174.
higher(P < 0 05, Kruskal-Wallistest)in controlplots.
Abrams,M.D. (1988) Comparativewaterrelationsof three
Our data indicaterapidcolonizationandhighseedling
successionalhardwoodspecies in centralWisconsin.Tree
densitieson pit and mound micrositesafteronly 3
Physiology4, 263-273.
4
(a)
Q. a/ba
Q. alba
(b)
31911
2-
1974
1950
0
cE 2
(C)
A. rubrum
N. sylvatica
(d)
-
~~~1921
1959
1960
1900
1925
1950
1975
2000
1925
1963
1950
1975
2000
radial growthpatternsof overtoppedtreesgrowingin gaps in
Fig. 2. Characteristic
mixed-Quercusof northernVirginia. Large arrowheadsindicate the 1990 windstorms,while arrowsand asterisksindicatethestartof majorand moderatereleases,
as definedby Lorimer& Frelich(1989).
respectively,
( 1995British
EcologicalSociety,
FunctionalEcology,
9,799-806
charAbrams,M.D. & Kubiske,M.E. (1990) Leaf structural
acteristicsof 31 hardwoodand conifertreespecies in centralWisconsin:influenceof lightregimeand shade tolerance rank.ForestEcologyand Management31, 245-253.
gas
Abrams,M.D. & Mostoller,S.A. (1995) Drought-year
sucand nitrogenin contrasting
exchange,leaf structure
cessionaltreespecies in highand low lightenvironments.
TreePhysiology15, 361-370.
Bazzaz, F.A. (1979) The physiologicalecologyofplantsuccession. Annual Review of Ecology and Systematics10,
351-371.
accliBazzaz, F.A. & Carlson,R.W. (1982) Photosynthetic
in thelightenvironment
of earlyand
mationto variability
late successionalplants.Oecologia 54, 313-316.
Beatty, S.W. & Stone, E.L. (1986) The varietyof soil
micrositescreated by tree falls. Canadian Journal of
ForestResearch 16, 539-548.
Bormann,F.H. & Likens,G.E. (1979) Patternand Process
in a ForestedEcosystem.Springer,New York.
of
Canham, C.D. (1988) Growthand canopy architecture
shade-tolerant
trees:theresponseofAcer saccharumand
Fagus grandifoliato canopygaps. Ecology 69, 786-795.
Canham,C.D. & Marks,P.L. (1985) The responseof woody
plants to disturbance:patternsof establishmentand
growth.The Ecology of Natural Disturbanceand Patch
Dynamics (eds S. T. A. Pickett & P. S. White), pp.
197-216. AcademicPress,New York.
Carpenter,S.B. & Smith,N.D. (1981) A comparativestudy
of leaf thickness among southernAppalachian hardwoods. Canadian JournalofBotany59, 1393-1396.
Clebsch,E.E. & Busing,R.T. (1989) Secondarysuccession,
gap dynamics,and communitystructurein a southern
Appalachiancove forest.Ecology 70, 728-735.
Elder, J.H.,Jr(1985) Soil Surveyof SpotsylvaniaCounty,
Virginia.USDA and Virginia AgriculturalExperiment
Station,Washington,DC.
Ellsworth,D.S. & Reich, P.B. (1992) Waterrelationsand
gas exchangeofAcer saccharumseedlingsin contrasting
naturallightand moistureregimes.Tree Physiology10,
1-20.
Fitter, A.H. & Hay, R.K.M. (1987) Environmental
Physiologyof Plants, 2nd edn. Academic Press, San
Diego.
community
organiFoster,D.R. (1988) Disturbancehistory,
zationand vegetationdynamicsof theold-growthPisgah
New Hampshire,USA. Journalof
forest,south-western
Ecology 76, 105-134.
Fritts,H.C. & Swetnam,T.W. (1989) Dendroecology:A
tool for evaluatingvariationsin past and presentforest
environments.Advances in Ecological Research 19,
111-188.
responseto gaps caused
Huenneke,L.F. (1983) Understory
by the death of Ulmusamericana in centralNew York.
BulletinoftheTorreyBotanical Club 110, 170-175.
Kloeppel, B.D., Abrams, M.D. & Kubiske, M.E. (1993)
Seasonal ecophysiologyand leafmorphologyof foursuccessional Pennsylvania barrens species in open and
understoryenvironments.Canadian Journal of Forest
Research23, 181-189.
perforKoike, T. (1988) Leaf structure
and photosynthetic
mance as relatedto the forestssuccession of deciduous
broad-leavedtrees.Plant Species Biology3, 77-87.
Latham,R.E. (1992) Co-occurringtreespecies changerank
in seedling performancewith resources varied experimentally.Ecology73, 2129-2144.
Lorimer,C.G. & Frelich,L.E. (1989) A methodologyfor
in
estimatingcanopydisturbancefrequencyand intensity
dense temperateforests.Canadian Journal of Forest
Research 19, 651-663.
Mikan, C.J., Orwig, D.A. & Abrams, M.D. (1994) Age
structureand successional dynamicsof a presettlementoriginchestnutoak forestin thePennsylvaniaPiedmont.
BulletinoftheTorreyBotanical Club 121, 13-23.
Oliver,C.D. & Larson,B.C. (1990) ForestStandDynamics.
McGraw-Hill,New York.
Orwig, D.A. & Abrams, M.D. (1994a) Land-use history
(1720-1992), composition,and dynamics of oak-pine
forestswithinthePiedmontand Coastal Plain of northern
Virginia. Canadian Journal of Forest Research 24,
1216-1225.
Orwig, D.A. & Abrams,M.D. (1994b) Contrastingradial
growthand canopy recruitment
patternsin Liriodendron
tulipiferaand Nyssa sylvatica:gap-obligatevs gap-facultativetreespecies. Canadian JournalofForestResearch
24, 2141-2149.
Poage, N.J.& Peart,D.R. (1993) The radialgrowthresponse
of Americanbeech (Fagus grandifolia)to small canopy
hardwoodforest.BulletinoftheTorrey
gaps in a northern
Botanical Club 120,45-48.
Reich,P.B., Abrams,M.D., Ellsworth,D.S., Kruger,E.L. &
Tabone, T.J. (1990) Fire affectsecophysiologyand communitydynamicsofcentralWisconsinoak forestregeneration.Ecology71, 2179-2190.
Ross, M.S., Sharik,T.L. & Smith,D.W. (1982) Age-structurerelationshipsof tree species in an Appalachian oak
forestin southwesternVirginia.Bulletin of the Torrey
Botanical Club 109, 287-298.
Runkle, J.R. (1992) Guidelines and sample protocol for
sampling forest gaps. USDA Forest Service Pacific
NorthwestResearchStationPNW-GTR-283.
SAS InstituteInc. (1985) SAS User's Guide: Statistics,
Version5 edn. Cary,NC.
806
D. A. Orwig&
M. D. Abrams
(C 1995British
EcologicalSociety,
Functional
Ecology,
9,799-806
Tryon,E.H., Lanasa, M. & Townsend,E.C. (1992) Radial
growthresponseofunderstory
sugarmaple (Acersaccharum) surrounding openings. Forest Ecology and
Management55, 249-257.
US Geological Survey(1984) SpotsylvaniaQuadrangle. 7
1/2"Series. Reston,VA.
Wallace, L.L. & Dunn,E.L. (1980) Comparativephotosynthesis of three gap phase successional tree species.
Oecologia 45, 331-340.
Walters,M.B., Kruger,E.L. & Reich, P.B. (1993) Growth,
biomass distribution
and CO2 exchangeof northern
hardwood seedlingsin highand low light:relationshipswith
successional statusand shade tolerance.Oecologia 94,
7-16.
Received 15 November1994; revised2 May 1995; accepted
3 May 1995