Forest Science,Vol. 33, No. 2, pp. 538-547.
Copyright1987by the Societyof AmericanForesters
Twenty-four Years of PonderosaPine
Growth in Relation to CanopyLeaf Area
and UnderstoryCompetition
R. OREN
R. H. WARING
S. G. STAFFORD
J. W. BARRETF
ABSTRACT.With datafrom a long-termexperimentin whichinitial stockingof treesand
understoryvegetationwere controlled,we testedthe hypothesisthat standgrowthis
directly correlatedwith the developmentof the forest canopy;e.g., growth is proportionalto canopyleaf area.The hypothesis
wasgenerallysupported.Standsof ponderosa
pine without understoryvegetationmorerapidlydevelopedcanopyleaf area than stands
with an understory.At very low leaf area indices,wood productionper unit of leaf area
was more efficient, particularly in stands where the understory vegetation had been
removed.Once the leaf area index exceeded2.0, however,the efficiencyof wood production remained stable and was comparablefor standswith and without understory
vegetation.FOR. ScI. 33(2):538-547.
ADDITIONAL KEY WORDS. Pinus ponderosa, sapwood area, water use.
SILVICULTURISTSHAVE LONGATFEMPTEDto increase stand growth by reducing stockinglevels and by improvingthe availability of soil resources.
Ecologistshave suggestedthat silviculturaltreatmentsmight be more appropriately evaluated if standswere comparedat equivalent canopy densities rather than at equivalent ages (Miller 1981, Waring 1983). Canopies
may be comparedon the basisof their projectedleaf area or on the amount
of solar radiation they absorb (Waring 1983).
We were able to evaluate the effect of stockingcontrol and removal of
understoryvegetationon the growth of ponderosapine (Pinus ponderosa
Dougl. ex Laws) at equivalentcanopydensitiesby analyzingdata collected
over 24 years (Barrett 1982). Changesin canopy density were expressedas
projectedsurfacearea of foliage(m2foliage/m2 ground).This leaf areaindex
(LAD was estimated by a linear correlation with sapwood basal area
(Waring and others 1982),which could be appliedbecauseonly a few trees
had died in the courseof 24 years, and the remainingtreeshad yet to form a
significantamountof heartwood.
Study Area
The study area is in the USDA Forest Service Pringle Falls Experimental
Forest (43øN, 121øW,1,350 m above sea level), about 60 km southwestof
Bend, Oregon.The experimentalsite is an east-facingslopethat receives
The authors are, respectively,AssistantProfessor,Schoolof Forestry and Environmental
Studies,Duke University, Durham, NC 27706;Professorof ForestScienceand AssociateProfessor,Departmentof Forest Science,Collegeof Forestry,OregonStateUniversity,Corvallis,
OR 97331; and ResearchForester,Bend Forest SciencesLaboratory, USDA Forest Service,
Bend, OR 97701.This work wasdonewhenR. Orenwasa graduatestudentat OregonState
University.They extendthanksto P. Cochran,for commentson an earlierdraft of the manuscript, and to two anonymousreviewers.This is Paper2064, Forest ResearchLaboratory,
Oregon State University, Corvallis, OR. Manuscript receivedNovember 4, 1985.
538/ FOREST SCIENCE
about60 cm of precipitationannually,85% of which falls betweenOctober
and April. From January to March, the snowpack is about 60 cm deep
(Franklin and Dyrness 1973).
The 30 rectangular0.08-haplots, eachsurroundedby a 10-mbuffer strip,
were establishedin 1957.At initiationof the experiment,the selectedstand
contained old-growth ponderosapine (about 50/ha) growing above 40- to
70-year-old suppressed trees (about 17,000/ha) (Barrett 1982). These
youngertrees in the understoryaveraged5.0 cm diameterand 2.5 m height.
Treeson the site are projectedto reach 24 m in heightat 100years (Barrett
1982).
Ground vegetationconsistedmainly of bitterbrush(Purshia tridentata
[Pursh] D.C.), varnish-leaf ceanothus (Ceanothus velutinus Dougl. ex
Hook.), and green-leafmanzanita(Arctostaphylos
patula Greene).Soilsdevelopedfrom dacite pumiceextendto a depthof 80 cm, where a buried soil
profile can be found in an older ash layer (Barrett 1982).
Methods
In 1958, all of the old-growthpineswere removed,and five stockinglevels,
each replicatedsix times, were establishedin the understory.Tree densities
for the five levels were 2,470, 1,235, 618, 309, and 153 trees/ha. All treatments were randomized and extended into the surroundingbuffer strips.
Within each stockingtreatment, half of the plots were kept clear of understory vegetationby herbicidesand mechanicalmeansfor the durationof the
experiment. Beginningin the fall of 1959, and during every fourth growing
seasonthereafter until 1979, diameter at breast height (dbh), total height,
and heightto the baseof the live crown were measuredon all trees (Barrett
1982).
In the summerof 1983, we measureddbh and height of all trees to the
base of the live crown. In addition, we measuredtotal heightof at least 12
trees per plot at the two lower stockingdensitiesand of at least 25 percent
of the trees at higherdensities.
ESTIMATING LEAF AREA
Leaf areaof ponderosapine was estimatedwith the assumptionthat it hasa
linear relationshipwith sapwoodcross-sectional
area at the baseof the live
crown, where a squarecentimeterof sapwoodsupports0.25 m2 leaf area
(Waringet al. 1982,Larssonet al. 1983).Taperin the sapwoodarea between
breast height and the base of the live crown was predictedfrom a relationship with breast-heightdiameter, tree height, and distancefrom the crown
base to the treetop (r2 = 0.97, P < 0.001) that was developedfrom data
collected from 120 trees (Cochran 1979).
In 1981, we confirmed that less than 3% of the total wood cross-sectionat
breast height consistedof heartwoodby coring 10 trees in the 2,470/haand
153/hastockingtreatments. This knowledgeallowed us to estimatethe leaf
area of all trees from diametermeasurementsand knowledgeof bark thickness for each year from the time treatmentsbegan in 1959 until 1979. In
calculatingleaf area, we accountedfor additionalformation of heartwood
with data from the samplingof wood coresin 1983.
ESTIMATING GROWTH
Tree volume was calculatedfor each tree from an equationdevelopedespecially for the study area (DeMars and Barrett, USDA Forest Service, Forest
SciencesLaboratory,Bend, OR, unpublished).Growth was determinedby
taking the differencesbetween all consecutivemeasurementsand interpo-
JUNE 1987/539
lating valuesfor the periodsbetween measurements.Tree height, required
to estimatevolume, was not measuredon all trees in 1983;therefore, specific heightequationshad to be developedfor each combinationof stocking
and controlled and uncontrolled understory vegetation. The equations,
which were basedon measurements
from 35 to 175trees, explainedbetween
70 and 85% of the variationin height(P < 0.001).
Growth in wood biomasswas determinedfor each plot by multiplying
volume growth by wood specificgravity. The latter value, estimatedfrom
core samplestaken at breastheightfrom 5 to 19treesper plot, dependingon
stocking,rangedfrom 260 to 310 kg/m3. To estimatehow much stemwood
was producedper unit of foliage, we divided the incrementin volume or
biomassby the estimatedleaf area of each tree. Averagegrowth efficiency
was determinedfor each plot. Stand stem growth was calculatedby summing the growth of individual trees in a givenplot, then convertingthe value
to growth per hectare.
To removeheteroscedasticity
andmakethe relationships
easierto analyze
statistically,the reciprocalof growthefficiencywas regressedagainstLAI.
This transformationgave a linear relationshipwith correlation coefficients
above0.77. The analyseswere performedon the complete24-yeardata set
(n = 90) and againon all but the last 4-year period(n = 75). In this way, the
estimatesof growth and canopydevelopmentmadewith equationsderived
from earlier measurementscould be comparedwith values from the last
sampling.
To account for possiblegrowth differencesassociatedwith tree size, we
added information about mean stem biomassand mean leaf area per tree.
BecauseLA! was important in predictinggrowth efficiency,we examined
the rate of increase in LAI
as a function of current LAI.
The accuracy of predicted changesin leaf area, as well as growth efficiency and standgrowth, was assessed:first, by comparingthe coefficients
of the relationships generated from the entire study with those derived
without the last set of measurementsand, second,by comparingprojected
valueswith valuestaken duringthe last measurementperiod. Standgrowth
was predictedin two ways: by calculatingcontinuousaccumulationas the
productof predictedLAI and predictedgrowthefficiency,and by describing
growth as a function of LA/.
Results
Over the 24 years of the experiment,diameterincreased71% (5 cm) on trees
growingat the higheststockinglevel and 171% (12 cm) on trees growingat
the lowest stockinglevel. With control of understoryvegetation,mean diameter at the higheststockinglevel increased63% (5 cm) and at the lowest
stockinglevel 300% (24 cm).
Basalarea increasedfrom 5.1 to 31.5 m2/haat the highestlevel and from
0.35 to 8.8 m2/haat the lowest level over the sameperiod. With no understory,it increasedfrom 3.8 to 33.59 m2/haand from 0.35 to 12.9 m2/haat the
two stocking extremes.
Tree growth efficiencydecreasedrapidly as the canopydeveloped(Figure
1). The main effect of understoryremoval was expressedat low leaf area
indices.For example, at LAI 1.0, understoryremovalincreasedgrowth per
unit of leaf area 30% over that for plots with an understory.
Althoughthe relationshipbetweengrowthefficiencyand canopyleaf area
differed significantly(Table 1, P < 0.01) amongsome4-year measurement
540/FOREST
SCIENCE
UNDERSTORY
NO UNDERSTORY
o
0.8.
0-o
0.4LU 0.2-
o
LEAF AREA INDEX (m2/m 2)
FIOURI•1. Averageannualgrowthefficiency(stemvolumeproduced/unitof foliagearea) in
relation to leaf area index (foliagearea/groundarea). Data are for measurementsmade at six
4-year intervals, the last shownby filled circles.
periods, there was no pattern (Figure 2); thus, poolingdata for the entire
24-year period seemedjustified (Figure 1).
Inclusion of the amount of leaf area per tree significantlyreduced the
amountof unexplainedvariationin the independentvariable(1/GE, Table1)
in all eight equations. The greatest negative effect on growth appeared
where understoryvegetationwas present. Leaf area and stembiomasswere
closelycorrelatedbecauselive crown ratioswere similaramongtreatments
(0.72-0.83). In such cases, both variables are functions of stem diameter.
The presenceor absenceof understoryvegetationdid not alter the slope
of the relationshipbetweenthe reciprocalof growthefficiencyandLAI. The
intercept was significantlylower, however,on plots without an understory,
indicatingthat growth efficiency at low LAI was improved by understory
removal (Table 1).
The relationshipof LAI increaseand currentLAI can be describedwith a
parabolic curve (Figure 3), becauseLAI increasesover time in a form describedby a sigmoidcurve. The equationsin Table 1 bestfit the data up to
LAI 2.0. Above LAI 2.0, plots varied considerably,probablyindicatingsite
differencesin water availabilityrather than in stocking(Barrett 1970).When
understoryvegetationwas removed, leaf area initially increasedmore rapidly than it did when the understorywas present(Figure 3, Table 1). Relative
changesin LAI for the entire data set can be describedequally well by a
peakingfunction (Jensenand Homeyer 1970):
ALAi
i=
g(LAI/LAI,.)
-1,.o ]
0.9
-0.2167)/0.7813
ALAIm•x
where
ALAI i = net annumincrementof LAI at LAI i,
JUNE 1987/541
TABLE 1. Regressionequationsfor predictingstemwoodgrowth efficiency(GE),
leaf area increment(•AI), and standstemwoodgrowth(G• and G2)as a function of
stand LAI and mean tree leaf area (LA).
Annual
Independent
variables
Dependent growth Understory
variable
units vegetation
1/GE
I/GE
1/GE
1/GE
ZILAI
Gi
G•
G2
G2
dm3/m2
kg/m2
dm3/m 2
kg/m2
m2/m2
m3/ha
T/ha
m3/ha
T/ha
Intercepta
LAI
LA12
LA
n
r2
+
2.12'**
1.39Ns
75 0.79
-
1.19
1.50
75
+
2.04**
1.61Ns
90 0.80
-
1.05
1.67
90
0.87
0.87
+
7.16'**
5.45Ns
75 0.77
-
4.33
5.58
75
+
7.18'*
5.87 Ns
90 0.77
-
3.96
6.11
90
0.86
+
3.06***
1.28 Ns
0.02***
75
0.88
-
2.29
1.45
0.01
75
0.91
+
1.77'**
1.36Ns
0.02***
90
0.90
-
0.95
1.45
0.01
90
0.92
+
10.29'**
0.85
4.98Ns
0.09***
75 0.85
-
8.34
5.35
0.03
75
0.91
+
6.06**
4.85 Ns
0.09**
90
0.89
-
3.61
5.35
0.04
90
0.90
+
0.0002 NS 0.1649'**
-0.0259***
75 0.83
-
0.0012
-0.0386
75
+
0.0096 Ns 0.1420'**
-0.0187'**
90 0.82
-
0.2420
0.1706
-0.0256
90
+
0.41'*
1.65Ns
-0.14 Ns
75 0.94
-
0.80
1.68
-0.17
75
+
0.55**
1.38Ns
-0.08 Ns
90 0.93
-
0.93
1.49
-0.12
90
+
0.11'*
0.50Ns
-0.06 NS
75 0.90
-
0.23
0.44
-0.04
75
0.91
+
0.16'*
0.39 Ns
-0.02 •qs
90
0.89
-
0.27
0.37
-0.03
90
0.89
+
0.41'**
1.93Ns
-0.20 Ns
- 0.008Ns 75 0.90
+
0.76
0.51'**
1.92
1.73Ns
-0.20
-0.14 Ns
-0.005
-0.008 Ns
75
90
0.93
0.97
-
0.87
1.71
-0.15
-0.006
90
0.94
0.2205
0.73
0.69
0.91
0.90
+
0.11'**
0.60**
-0.08***
-0.003***
75 0.94
-
0.22
0.50
-0.05
-0.001
75
+
0.14'**
0.50 Ns
-0.04**
-0.003***
90 0.93
-
0.26
0.43
-0.03
-0.001
90
0.92
0.92
a Significantdifferencesbetweentreatmentswith ( + ) andwithout( - ) understoryvegetation
are indicated:*'0.01, **'0.001, NS not significantat 0.05. Regressions
are derivedfrom analyses of 6, or all measurementperiods using the total 24-year data set (n = 90), and from
analysesof the samedata set but without the last period (n = 75).
Note: Adjustedr 2 is reportedfor multiplelinear regressionmodelswith variablesincludedin
the equationsat significancelevel 0.01. G• and G2 refer, respectively,to equationswith and
without the variable mean tree leaf area.
542
NO UNDERSTORY
UNDERSTORY
•: E QS'
O•
L9• 0.6.
LLIZ
n" • 0.4.
Q2'
5
o
6
o
LEAF AREA INDEX (m2/m 2)
FIO•JRE 2.
Least squaresfit of averageannualgrowthefficiency(stemvolumeproduced/unit
of foliage area) in relation to leaf area index (foliagearea/groundarea). Data taken at six
4-year intervalsare shownby separatelines numberedsequentially.
•LAImax = maximumrate of LAI accumulation,and
LAI m = LAI at ALAIma•.
With an understory,r2 = 0.81; with no understory,r2 = 0.69.
Predictedchangein LAI duringthe last 4-year periodwas within 30% of
measuredvalues for plots without an understoryand within 45% of those
with an understory.Much of the errormayreflectreal variationfromplot to
plot (Figure 3). Predictionsof growth efficiencyin units of volumeor biomasswere within 20 and 30%, respectively,of measuredvalues,and estimates of leaf area within 15%.
NO UNDERSTORY
UNDERSTORY ß
E o.4-
•
Q3-
•
0.2-
z
0
o
o
ø
•
DO o o q•
o•Oo•Oo
• ß ..
g
0 o O• •eo•e
0
ß
0
0
0
o
z
CURRENT LEAF AREA INDEX (m2/m •)
FIOURE
3.
Annual changein the leaf area index (foliagearea/groundarea) in relationto the
currentleaf areaindex.Data are frommeasurements
madeat six 4-yearintervals.
JUNE 1987/ 543
Productionis a product of leaf area index and growth efficiency.The most
rapid changein productionoccurredat the early periodsbecausegrowth
efficiencyactuallyincreasedalongwith LAI (Figure 4). The rate of increase
in production decreased and eventually stabilized as increasingLAI was
balancedby a proportionaldecreasein growth efficiency(Figure 2). LAI
appeared to be approachingmaximum at the denseststockinglevels, as
shownby very slow incrementsin leaf area (Figure 3) and by the first signs
of density-relatedmortality of smaller trees. Eventually, as the trees become larger, maintenancerespirationis expectedto further reduce growth
rates (Whittaker 1975, Waring and Schlesinger1985). Growth estimated as
the productof LAI and growthefficiencywas within 20 and 30%, respectively, of measuredvolume (Figure 4) or biomassincrement, and when estimated as a function
of stand leaf area was within
10% of the measured
values (Table 1).
Discussion
The exponentialdecreasein growthefficiencyobservedwith increasingLAI
(Figures 1,2) is similar to that observedin lightly managedforests of Pinus
contorta Dougl. (Mitchell et al. 1983),Pinus sylvestrisL. (Waring 1985), and
Pseudotsugamenziesii (Binkley and Reid 1984). In intensively managed
forestsof Pseudotsuga(Waring et al. 1981)and Pinus ponderosa(Larsson
et al. 1983), where frequent thinning allows lower crowns to be maintained
and to contribute substantiallymore to net carbon uptake (Helms 1971,
Linder and Axelsson 1982), growth per unit leaf area decreasesnearly linearly with increasingLAI.
The rapid decreasein growth efficiencyobservedbetween0 and 2.0 LAI
probably is not related to a comparablereductionin photosynthesis,al-
ZO-
NOUNDERSTORY
ß
UNDERSTORY
ß
o
ß
o
ß
o
o
o
oß
ß
o
o oß
T 4.0'
o
o
3.oOf
•
o
oOO%
o
oOo
¸
O0
o
ß
ß
0
oß
oo
o
o
ß
o
LEAF AREA INDEX (m2/m 2)
FIGURE4. Averageannual standvol.ume-growth
in relation to leaf area index (foliagearea/
groundarea). Data are from measurementsmadeat six 4-year intervals.
544/FOREST
SCIENCE
thoughirradiancemay be reducedby 40 to 50% in that range of canopy
cover (Kira and Shidei 1967, Jarvis and Leverenz 1983). Once irradiance
dropsbelow 50% of that at the top of the canopy,shadedfoliageis equally
efficientover a broadrangein LAI (Nygrenand Kellomaki1983).Even in a
standofPicea abieswith LAI 11.6, for example,uptakeby shadedfoliagein
the canopyaccountedfor 30% of the annualtotal (Schulzeet al. 1977).
At comparableLAI below 1.5, plotswith an understoryusedthe available
water more rapidly than plots without an understory.Althoughthe transpiringsurfaceof the understoryvegetationwas not determined,it used
water equivalentto a pine canopywith LA/of 0.7 when it was fully developed (Figure 5). Thesefindingshelp explainwhy growthefficiencyamong
treatmentsdifferedonly at low LA/. The increasein growthefficiencybetween the first and third periodsat a comparableLAI of 1.0 (Figure2) suggests,however,that rootsdid not fully reoccupythe coarse-textured
pumice
soil for 12 years (Hermann and Petersen1969, Youngbergand Cochran
1982). Once they fully occupiedthe soil, the relative allocationof photosynthateto stemsmay also have increased.
A decreasein growth efficiencyat a givenLAI throughoutthe studywas
relatedto an increasein averagetree leaf area (Table1) and couldbe a result
of leaf area accumulatingfaster than crowns expand. For example, trees on
lightly stockedplots had 10 times the foliagebut only 4 timesthe crown
volume of trees growingat denserstockingwith similarLA/. Suchdifferenceswouldpermittreesat higherstockinglevelsto photosynthesize
more
efficiently than those growingat lower stockingbecauselight would pene-
ß
ß
ß
75-
ß
ß
ß
&ß
o
o
50-
o
o
o
NO UNDERSTORY
25-
UNDERSTORY
o
0.5
2YR
8YR
O
/•
ß
ß
I.O
LEAF AREA INDEX (m•/m •)
Fioum• 5. Percentageof availablesoil water usedby the end of the growingseason,as compared with water usedby an adjacentuncut stand,in relationto the leaf area index (foliage
area/groundarea) of ponderosapine plotswith and withoutunderstoryvegetation2 and 8
years after initiationof the experiment(calculatedfrom Barrett 1970).
JUNE 1987/545
trate more efficiently.In addition, a better groundcover providedby the
distributionof a givenLAI on moretreesreducesthe amountof lightthat is
not interceptedand is unavailablefor photosynthesis.
Once forests approachtheir maximumLAL differencesin growth efficiency are often between75 and 85 g wood producedper squaremeter of
foliageannually,evenin contrasting
environments
wherethe amountof leaf
area or duration of display differ considerably(Waring 1985, Waring and
Schlesinger
1985).This studyemphasizes
the ideathat the majoreffectsof
silviculturaltreatmentsmay be best assessedby comparingtree responses
at a common,relatively low leaf area index (Waring 1983, 1985).
A temporaryimprovement
in resourceavailability,suchas that observed
in this study,significantlyincreasesthe initial rate of canopydevelopment
(Miller 1981, Waring 1983).A sustainedimprovementin the availabilityof
scarceresources,althougheconomicallyoften a questionablealternative,
shouldincreasethe maximumleaf area, while maintaininghigh growth efficiency by shiftingcarbohydrateallocationaway from roots (Waringand
Schlesinger1985).
Literature
Cited
BnV,•ETT,J. W. 1970.Ponderosa
pine saplings
respondto controlof spacingandunderstory
vegetation.USDA For. Serv. Res. Pap. PNW-106. 16 p.
BnV,•ETT,J. W. 1982.Twenty-yeargrowthof ponderosa
pinesaplings
thinnedto five spacings
in centralOregon.USDA For. Serv. Res. Pap. PI',FW-301.
18 p.
BINKLEY,D., andP.REID. 1984.Long-termresponses
of stemgrowthandleafareato thinning
andfertilizationto a Douglas-firplantation.Can. J. For. Res. 14:656-660.
BROUWER,
R. 1962.Nutritive influenceson the distributionof dry matterin plants.Neth. J.
Agric. Sci. 10:399-408.
COCHP,
AN, P. H. 1979.Response
of thinnedponderosa
pineto fertilization.USDA For. Serv.
Res. Pap. PNW-339.8 p.
EMMINGHAM,W. H., and R. H. WArnNO.1977. An index of photosynthesis
for comparing
forest sitesin westernOregon.Can. J. For. Res. 7:165-174.
FRANKLIN,J. E, and C. T. DVm•ESS.1973. Natural vegetationof Oregonand Washington.
USDA For. Serv. Gen. Tech. Rep. PNW-8.
HELMS,J. A. 1971.Environmentalcontrolof net photosynthesis
in naturallygrowingPinus
ponderosaLaws. Ecology 53:92-101.
HERMANN, R. K., and R. G. PETERSEN.1969. Root developmentand height incrementof
ponderosa
pinesin pumicesoilsof centralOregon.For. Sci. 15:226-237.
JA•WS, P. G., andJ. W. LEVERENZ.1983.Productivityof temperatedeciduousandevergreen
forest.P.223-280 in Encyclopedia
of plantphysiology,
Vol. 12D,O. L. Lange,P.S. Nobel,
C. B. Osmond,and H. Ziegler(eds.). Springer-Verlag,
New York.
Kmn, T., andT. SHmEL 1967.Primaryproductionandturnoverof organicmatterin different
forest ecosystemsof the westernPacific.Jpn. J. Ecol. 17:70-87.
EARSSON,
S., R. OREN,R. H. WARING,and J. W. BARRETT.1983.Attacksof mountainpine
beetleas relatedto tree vigorof ponderosapine. For. Sci. 29:395-402.
LINDER,S., and B. AXELSSON.1982.Changesin carbonuptakeand allocationpatternsas a
resultof irrigationandfertilizationin a youngPinussylvestrisstand.P. 38-44 in Carbon
uptakeand allocationin sub-alpineecosystems
as a key to management.
Proc. IUFRO
workshop,R. H. Waring(ed). For. Res.Lab, Oreg.StateUniv., Corvallis.
MILLER,H. G. 1981.Forestfertilization:someguidingconcepts.Forestry54:157-167.
MITC}•ELL,R. G., R. H. WAm•G, and G. B. PlYMAN.1983.Thinninglodgepolepine increases
tree vigor andresistanceto mountainpinebeetle.For. Sci. 29:204-211.
NYGREN,M., andS. KELLOM'.X.
KI 1983.Effectof shadingon leaf structureandphotosynthesis
546/FOREST
SCIENCE
in young birches,Betula pendula Roth. and B. pubescensEhrh. For. Ecol. Manage. 7:119132.
Rn-I^, S. J., and G. S. CAMPBELL.1985. Estimating•vater fluxes in Douglas-firplantations.
Can. J. For. Res. 15:701-707.
SCHULZE, E.-D., M. I. FUCHS, and M. FUCHS. 1977. Spacial distributionof photosynthetic
capacity and performance in a mountain spruceforest of northern Germany. I. Biomass
distributionand daily CO2 uptake in differentcrownlayers. Oecologia29:43-61.
WARING, R. H. 1983. Estimating forest growth and efficiency in relation to canopy leaf area.
Adv. Ecol. Res. 13:327-354.
WARING, R. H. 1985. Irabalanced forest ecosystems: assessmentsand consequences. For.
Ecol. Manage. 12:93-112.
WARING,R. H., K. NEWMAN, and J. BELL. 1981. Efficiency of tree crownsand stemwood
production at different canopy leaf densities. For. 54:130-137.
WAmNG, R. H., and W. H. SCHLESINGER.1985. Forest ecosystems:conceptsand management. Academic Press, Orlando, FL. 338 p.
WARING,R. H., P. E. SCHROEDER,
and R. OREN. 1982. Application of the pipe model theory
to predict canopy leaf area. Can. J. For. Res. 12:556-560.
WmTrAKER, R. H. 1975. Communitiesand ecosystems.Ed. 2. MacMillan, New York. 385 p.
YOUNGBERG,C. T., and P. H. COCIm•N. 1982. Silviculture on volcanic ash soils in America.
P. 331-345 in Silvicultureunder extreme ecologicaland economicconditions.IUFRO Div.
I Meet., 1980. Thessaloniki, Greece.
JUNE 1987/547