Estimation of crown radii and crown projection area from stem size and tree position
Ann. For. Sci.
Estimation of crown radii and crown projection area from stem size and tree position
Rüdiger GROTE 0 1
0 Chair for Forest Yield Science, Department of Ecology and Landscape Management, TU Munich , Am Hochanger 13, 85354 Freising , Germany
1  Utschig H., Analyzing the development of regeneration under crown cover: Inventory methods and results from 10 years of observation , in: Skovsgaard J.P., Burkhart H.E. (Eds.) , IUFRO - Recent Advances in Forest Mensuration and Growth and Yield Research, Danish Forest and Landscape Research Institute/Ministry of Environment and Energy , Tampere, Finland, 1995, pp. 234-241
- This paper describes a method for crown radii estimation in different cardinal directions using tree diameter, height, crown length, and stem position within the stand as independent variables. The approach can serve for the initialisation of crown dimensions if measured crown radii are not available in order to address various research questions. Test calculations are carried out with 4 pure spruce (Picea abies L. Karst), 5 beech (Fagus sylvatica L.), and 6 mixed stands with both species. Simulated tree radii, crown projection area and canopy cover are compared with measurements and simple estimation procedures based on logarithmic and linear equations. In beech stands and dense spruce stands the estimates with the new approach are similar or superior to those obtained with the other methods. However, in sparse plots or in stands, which have experienced a recent thinning crown size of trees is overestimated.
Many ecological and economic problems in forestry today
(e.g. continuous cover forestry, wood production and quality)
are approached using crown dimensional measures. For
example, individual tree competition indices are derived from crown
area estimates [
] because crown dimension is a result of
past competition as well as an indicator of the current growth
potential . Thus, crown dimensional measures are also
used in more sophisticated single-tree models – particularly
when forest growth in uneven-aged or mixed species stands is
]. Furthermore, crown size and canopy cover
determine the probability of successful natural regeneration by
its influence on the pattern of shade, light, and rainfall on the
ground . In general, many approaches of modelling light
distribution (e.g. ), water balance (e.g. [
]), tree growth
]), and tree physiology (e.g. ) depend on
information about crown dimensions of individual trees. Possibly,
considering a more realistic crown shape will become
increasingly important also for stem quality simulation, because branch
dimension is one of the most important determinants [
Despite its importance crown extension remains difficult to
determine. It can only be measured by optical methods from
below  or from above [
], which both are subjected to a
likely underestimation of crown width due to a limited
visibility of crowns. The crown projection area can be estimated from
stem dimensions [
], but has to be thoroughly
parameterised for specific stand conditions [
], which in most cases
involves again a large number of direct measurements. Finally,
canopy cover can not be assumed to be the sum of tree crown
projection areas, because overlapping is a common
phenomenon particularly in dense, uneven-aged, and mixed stands.
The difficult measurements and the sensitivity of crown
dimension on management makes it desirable to develop
estimation procedures based on variables that are easier to measure
than crown extension itself. Thus, maximum crown radius,
which can be derived from stem diameter, has been used to
estimate crown projection area [
]. Because increasing
stand density results in increasing overestimates an adjustment
factor has been introduced that is generally derived from
overlap estimates [
]. More recently, average crown radius and
canopy cover in several types of conifer forests were
successfully estimated with regression equations that have been
derived from stem diameter, height, and/or crown length [
All of these methods are developed to give reliable results on
the stand level, which is suitable for many of the purposes
mentioned above. It is not sufficient, however, for analyses
that account explicitly for the asymmetry of crowns.
Information about asymmetric crown extension has been used e.g. for
detailed ecosystem characterisation  or the simulation of
wood quality [
], radiation distribution [
susceptibility of trees to windthrow , crown biomass , and
individual tree physiology [
]. Therefore, a method, which
estimates crown radii in various cardinal directions for every
tree in a given stand would be of great value for these research
areas. In this paper, such an approach is presented that is based
on the size of a tree and the size and position of surrounding
competitors. Also, the sum of crown projection area and
canopy cover (the total area covered by canopy) are both
calculated based on the estimated radii and results are compared
with those obtained with other methods.
2. MATERIALS AND METHODS
2.1. Stand description
In order to test the proposed method for crown radii estimation, a
number of forest stands were selected that include the most important
tree species and stand structure types in Germany. The stands consist
of trees with a coniferous (Picea abies L. Karst.) and a broadleaved
(Fagus sylvatica L.) tree species either in pure or mixed stands. All
of them belong to the network of long-term investigation plots in
Bavaria, South Germany and are maintained by the Chair for Forest
Yield Science in Freising. Thus, tree position, stem diameter, height,
height of crown base, and crown radii length had already been
measured at many trees. The plots of pure spruce (Eurach, 4 plots) and
beech (Starnberg, 5 plots) both represent different degrees of stand
density. The mixed plots (Freising, 6 plots) represent different age
classes. All plots of one site are located closely together to minimise
differences in site conditions. For a more detailed description see
Table I. The plots of pure spruce and the mixed plots are furthermore
described in connection with other investigations [
Diameter at breast height had been measured with a girth tape at
all trees. Tree height and crown base height of each tree within one
plot had been determined from height-diameter relations that are
derived from a subset of approximately 40 measured heights at each
plot. The visible crown extension in each of eight cardinal directions
had been measured by vertically looking up as described by Röhle
. Calculations are carried out with all trees within the plots, but
trees at the plot boundaries are omitted from the results. This is
necessary because in these cases no competitors at the outward side are
considered and crown radii would thus be overestimated.
2.2. Distance dependent approach
The suggested approach is based on two assumptions. The first is
that the potential horizontal crown extension is a function of stem
diameter, and the second is that the distance between the tree and its
competitors determines the actual crown dimension within the limit
of the potential crown extension. Following Arney [
are defined as trees with an overlap in potential crown extensions
(Fig. 1A). Crown radius length in a particular direction is limited by
the position of competitors within a certain angle on both sides of the
radius (Figs. 1A and 1B) and by their crown width at the height where
the maximum crown extension of a centre tree is assumed (Fig. 1C).
The method is further on referred to as ‘maximum radii estimation’
Firstly, in order to determine the height where maximum crown
width occurs, a crown shape function is required similar to those that
have been suggested by several authors during the last decades (e.g.
8, 25, 26, 29, 35
]). However, these equations require many
parameters that are not directly measurable (e.g. [
]), assume a steady
increase with canopy depth (e.g. [
]), or end with a zero-radius at
crown base height (e.g. [
]). In the current context, these properties
are considered as disadvantages. Thus, a new one-parameter equation
is used that describes crown radius at every height h (rh) as a function
of crown base height (hcr), crown length (lcr), and the maximum
radius in a particular cardinal direction (rmax). The term relH refers
to the relative height within the crown, which is 0 at crown base and
1 at the tip of the tree.
1 – relH f h
rh = rmax
1 – relH f h
f h = base
The effect of the base-term in equation (1c) is demonstrated in
Figure 2, with rmax = 1 and a crown length of 15 m. In a detailed
analysis of 12 trees, values of base were found to be 1.23 0.074 for
spruce and 2.02 0.71 for beech [
]. However, the standard
deviation can be decreased if base is derived from crown length according
to equation (1d) (1.23 0.071 for spruce and 2.08 0.345 for beech)
with ps equal to 0.018 (R2 = 0.65) and 0.0756 (R2 = 0.54) for spruce
and beech respectively.
base = 1 + ps lcr
ps: shape parameter.
To determine the maximum crown radii of one tree, the maximum
crown extension for the competitor trees j are needed but generally
not known. Thus, for a given competitor, rmax,j is calculated from the
distance to tree i (dij) and from diameter at breast height (dbh) of both
trees according to equation 2. The distance between a tree and its
competitor dij can easily be calculated from stem positions.
rmax j = dij -d---b---h--d-j---b---h---j---------. Eq. (2)
However, rmax of any tree is limited to its potential radius (rpot),
which describes the physical maximum is hardly affected by site
]. Since no open grown trees were available, rpot is
estimated from the 5% relative largest crown radii found at the trial plots.
To determine these radii, firstly all radii (with 8 radii measured per
tree) are exponentially fitted to the stem diameter at ground height do
(MS Excel software package). The 5% selected radii are the ones with
the largest positive deviation from this relation. Another exponential
fit through these radii according to equation 3 derives the parameter
pr1 and pr2. The diameter at ground height is derived from dbh by
assuming a certain diameter decrease of the bole with increasing
height (0.3 cm m–1). It is used as independent variable instead of dbh
because otherwise equation 3 would imply that small trees (< 1.3 m
height) have no crowns at all, which would restrict the generality of
the approach. Parameters are determined separately for each tree
species and for pure and mixed stands although the differences between
the relations for spruces in different stand structure types were not
significant (Fig. 3). Values for pr1 and pr2 together with the number
of radii that have been used to derive the functions are given in
rpot = pr1 dopr2
(rpot and do in m).
From rmax,j the potential crown extension of all competitor trees is
calculated for every height according to equation 1 in height steps of
0.5 m. In this calculation, rh of competing trees below the height of
maximum crown diameter is set to rmax to better account for the
influence of light competition in deeper canopy layers.
In the next step, the angle ij between the tree (i) and its
competitor (j) is calculated from tree positions (Fig. 1A). Assuming that a
branch will grow until it reaches the crown circumference of a
competitor tree, the length of each crown radius is calculated as follows
rh i = min TiS rpot h i
TiS = cos ij dij – l
l = rpot h j T T
rh,i: actual radius of centre tree i at height h; rpoth,i: potential radius of
centre tree i in height h; rpoth,j: potential radius of competitor tree j in
height h; TiS: distance between stem position and the point of
intersection; l: help variable; TjT’j: distance between competitor tree j and
a point, mirrored at the radius prolongation (T’j is described as a
‘virtual mirror tree’ in Fig. 1C).
The actual radius of rh,i is calculated as the minimum radius
determined by considering every competitor within angle s on both sides
of the radius (see illustration in Fig. 1A). Based on former
investigation results [
] and test calculations with different angles, s is
set to 45o (8 radii).
Since first calculations showed that the largest crown radii in one
direction was too often equal to the potential radii, a further restriction
was introduced to get more realistic results for rmax. As illustrated in
Figure 1B, the assumption is made that a radius can not grow beyond
the stem position of a competitor tree. Despite these limitations, it
should be recognised that an overlap between crowns can result from
the elliptical connection between two adjacent radii (see further
2.3. Other calculations
Currently, the most common estimation procedures of crown
projection area are based on linear [
] or logarithmic [
relationships between stem and crown diameter. Thus, simple calculations
are carried out using linear correlations between dbh and radius
length (rmax in dm = alin1 + blin1 dbh), and dbh and crown
projection area (A in m2= alin2 + blin2 dbh in cm) of individual trees.
Crown projection area is also calculated with a logarithmic relation to
stem cross-sectional area (lnA = alog + blog ln(dbh2 0.25)).
The parameter alog and blog are derived analytically from the same
data set as pr1 and pr2 and are also presented in Table II for each tree
species and for pure and mixed plots (not for each plot!). In order to
derive crown projection area from measured and simulated crown
radii, the area between the radii is considered as a fraction of an
ellipse . Canopy cover is calculated with a computer program that
draws the crown projection area of every tree on a grid and counts the
number of coloured pixels.
The relation between simulated and measured radii is
shown in Figures 4A–4D. The coefficient of determination
ranges from 0.2 for pure spruce to 0.45 for beech in mixed
stands. A small bias is obvious in every figure, which indicates
an overestimation of small radii and an underestimation of
large radii. This is at least partly due to radii that had been
measured with zero length, which can not be represented with
the MRE method due to the assumption made in equation 2.
Slope values with the regression line forced through the origin
are presented in Table III separately for species and sites
together with the respective R2 values. The table shows that
despite the bias positive correlation coefficients had been
obtained with the MRE method in all cases, but not with the
estimation based on the linear approach.
Figure 5 and Table IV show that MRE does not decrease
the accuracy of crown projection area estimates compared
with the fitted logarithmic (LOG) and the linear method (LIN).
The slope values obtained with every method are similar (in
average over all plots separated by species: MRE = 1.00, LOG =
0.91, LIN = 0.96) although the standard deviation of MRE is
highest (MRE = 0.25, LOG = 0.12, LIN = 0.15). The R2 values
of MRE are similar to those obtained with the LOG approach
and are higher than R2 values obtained with the linear
approach (MRE = 0.64 0.15, LOG = 0.61 0.16, LIN =
0.50 0.23, with all negative values excluded from the
average). However, crown projection area for spruce is somewhat
overestimated, particularly in the mixed plots (+4 and +28%
mean deviation from measurements for pure and mixed plots
respectively), whereas for beech it is generally underestimated
(–15 and –17%).
The goodness of fit apparently depends on the density of the
plot and of the thinning intensity that the stand has been treated
with (see Tab. I). In the plots Eurach 1 and Starnberg 2, which
are the most dense for each species, the deviation from the 1:1
line is only marginal (spruce –4%, beech +1%) and the simulated
values are closely correlated with measured crown projection
area (R2 = 0.7 and 0.8 for spruce and beech respectively). In
spruce, overestimation increases in sparser plots (up to 23% in
the sparsest plot Eurach 2), whereas for beech crown
projection area is underestimated in thinned plots but no particular
trend with the intensity of thinning is obvious.
The sum of crown projection areas within one plot is similar
to that calculated from the measurements although an
overestimation for spruce (+9%) and an underestimation for beech
(–19%) is obtained (Tab. V). Again, the simulation of the
densest plots for both species are closest to the measurements
(Eurach 1: –6%, Starnberg 2: –4%).
Table V shows canopy cover values derived from either
measured or simulated crown radii. Additionally, crown
overlap is calculated from the difference between the sum of single
tree crown projection areas and canopy cover. This
demonstrates that the overlap derived with the MRE method is
generally too small. In spruce stands, however, this underestimation
is only slight (–3%), whereas it is in average –14% for beech
stands. Mixed stands are in between (in average –6%).
Results indicate that the MRE method can be used to
estimate crown radii for beech and spruce in dense stands but has
to be applied cautiously. Although some of the variance may
be due to the high inaccuracy of crown measurements ,
crown radii of trees from sparse plots or in recently thinned
stands are generally overestimated. This is consistent with the
underlying assumption of a balanced crown extension, which
can not be expected in heavily thinned stands and which is
more likely with morphological flexible tree species like
beech than with spruce [
Future tests and improvements of the MRE approach will
focus on crown shape estimation, which is based on a quite
small sample size of trees yet. Only a larger sample provides
the possibility to establish dependencies of crown shape on
spacing and competition that have been already found in other
5, 12, 14, 32, 34
]. Further improvements could
be based on the finding that in mixed stands spruce radii are
generally over and beech radii are underestimated. This would
be mitigated if a species-specific weighing factor for the
calculation of potential spruce and beech radii is introduced in
equation 2. However, it is not clear from the limited set of test
sites to which degree the effect is due to the stand structure
rather than species-specific properties. Although they are
older, most beeches of the mixed plots are smaller than the
spruces. Thus, the assumption that crowns of small trees are
restricted by the largest extension of competitor crowns rather
than their actual extension may affect beeches more than the
spruces at these particular plots. In this case, separate crown
radii estimations for different crown layers may produce more
favourable results but simulations of differently structured
mixed stands are required to test this assumption.
Improvements in crown radii estimates will generally
positively affect crown projection area and canopy cover
estimates. Nevertheless, the good agreement of simulated and
measured canopy cover despite the underestimation of crown
projection area in beech stands shows that the estimation of
crown overlap is also subjected to errors. Again,
species-specific differences have to be considered since the predicted
overlap for spruce trees is quite close to the
measurementbased calculations. This finding strengthen the assumption
that a separate calculation of different crown layers may be
The MRE method aims not preliminary on a precise
estimate of crown projection area or canopy cover. Over all, the
logarithmic approach, which is used here as an example for
similar and sometimes more sophisticated procedures (e.g.
13, 17, 51
]), produced slightly better results and would
perform even better if parameters would have been fitted for each
plot separately. Furthermore, the estimates produced with the
MRE method seem to be more sensitive to stand density
effects than established estimation methods  – at least for
trees with inflexible crowns.
However, the author has found no other approach that
estimates crown radii for different cardinal directions. Thus, the
demand for crown asymmetry-information that has been
formulated in various fields of research (scaling, light modelling,
estimation of windthrow susceptibility, wood quality, and
crown biomass) can currently only be fulfilled with actual
measurements. Despite the scatter, the immanent bias, and the
dependency of accuracy on species and stand density, the
MRE method may thus be used as a substitute for measured
crown radii in cases where these are not available but
information about crown asymmetry is needed. While stem diameter,
tree height, and crown length are often directly measured or
can be estimated with suitable equations (e.g. [
acquisition of tree position data in the field may be more
difficult and expensive. However, also tree positions can be
ated based on stand inventory data (e.g. [
]), which may be
sufficient for many of the purposes mentioned above.
Acknowledgements: This research has been conducted within the
framework of the joint-research project ‘Growth and Parasite
Defence’, funded by the German Research Agency (DFG). The Chair
of Forest Yield Science, lead by Hans Pretzsch, supported the
research generously with the supply of basic data, collected and
processed by Martin Nickel, Leonhard Steinacker, and Martin
Bachmann. Furthermore, I’d like to thank Hans Pretzsch, Greg
Biging (Berkeley University, California), and the two anonymous
reviewers, who made valuable comments to the manuscript, as well
as Thomas Seifert, who provided yet unpublished data for the
parameter estimation of crown shape for spruce trees.
 Stadt K.J., Lieffers V.J., MIXLIGHT: a flexible light transmission
model for mixed-species forest stands, Agric. For. Meteorol. 102
 Wang Y.P., Jarvis P.G., Description and validation of an array
model - MAESTRO, Agric. For. Meteorol. 51 (1990) 257–280.
 Akça A. , Aerophotogrammetrische Messung der Baumkronen , AFZ/Der Wald 30 ( 1983 ) 772 - 773 .
 Alsheimer M. , Köstner B. , Falge E. , Tenhunen J.D. , Temporal and spatial variation in transpiration of Norway spruce stands within a forested catchment of the Fichtelgebirge , Germany, Ann. Sci. For . 55 ( 1998 ) 103 - 123 .
 Arney J.D. , Computer Simulation of Douglas-fir Tree and Stand Growth. Philosophy , Oregon State University, 1972 .
 Assmann E. , Die Theorie der Grundflächenhaltung und die Praxis der Bestandespflege bei der Rotbuche , Forst- u. Holzwirt /5 ( 1954 ).
 Baldwin J.V.C. , Peterson K.D. , Clark III A. , Ferguson R.B. , Strub M.R. , Bower D.R. , The effects of spacing and thinning on stand and tree characteristics of 38-year old Loblolly Pine, For . Ecol. Manage. 137 ( 2000 ) 91 - 102 .
 Bella I.E. , A new competition model for individual trees , For. Sci . 17 ( 1971 ) 364 - 372 .
 Biging G.S. , Dobbertin M. , Evaluation of competition indices in individual tree growth models, For . Sci. 41 /2 ( 1995 ) 360 - 377 .
 Biging G.S. , Wensel L.C. , Estimation of the crown form for six conifer species of northern California, Can . J. For. Res . 20 ( 1990 ) 1137 - 1142 .
 Bonnor G.M., The Influence of Stand Density on the Correlation of Stem Diameter with Crown Width and Height for Lodgepole Pine, For . Chron. 40/3 ( 1964 ) 347 - 349 .
 Brunner A. , A light model for spatially explicit forest stand models , For. Ecol. Manage . 107 ( 1998 ) 19 - 46 .
 Cescatti A. , Modelling the radiative transfer in discontinuous canopies of asymmetric crowns. I. Model structure and algorithms , Ecol. Modell. 101 /2- 3 ( 1997 ) 263 - 274 .
 Chen H.Y.H. , Klinka K. , Kayahara G.J., Effects of light on growth, crown architecture, and specific leaf area for naturally established Pinus contorta var. latifolia and Pseudotsuga menziesii var . glauca saplings , Can. J. For. Res . 26 ( 1996 ) 1149 - 1157 .
 Crookston N.L. , Stage A.R. , Percent canopy cover and stand structure statistics from the Forest Vegetation simulator , no. RMRSGTR-24 , Department of Agriculture, Forest Service, Rocky Mountain Research Station, Ogden, UT , 1999 .
 Deleuze C. , Hervé J.C. , Colin F. , Modelling crown shape of Picea abies: spacing effects , Can. J. For. Res . 26 ( 1996 ) 1957 - 1966 .
 Dubrasich M.E. , Hann D.W. , Tappeiner II J.C. , Methods for evaluating crown area profiles of forest stands , Can. J. For. Res . 27 ( 1997 ) 385 - 392 .
 Franz F. , Pretzsch H. , Nüsslein S. , Strukturentwicklung und Wuchsverhalten von Buchenbeständen - Ertragskundliche Merkmale des Schirmschlag-Femelschlag-Verjüngungsverfahrens im Spessart, Allg . Forst- und Jagdztg. 160/6 ( 1989 ) 114 - 123 .
 Gill S.J. , Biging G.S. , Murphy E.C. , Modeling conifer tree crown radius and estimating canopy cover , For. Ecol. Manage . 126 ( 2000 ) 405 - 416 .
 Gilmore D.W. , Equations to describe crown allometry of Larix require local validation , For. Ecol. Manage . 148 ( 2001 ) 109 - 116 .
 Goelz J.C.G. , Open-grown crown radius of eleven bottom-land hardwood species: prediction and use in assessing stocking , S.J. Appl. For. 20/3 ( 1996 ) 156 - 161 .
 Grauer A. , Auswirkungen unterschiedlicher Moorwasserstände auf das Wachstum von Fichtenbeständen im Versuch EURACH 259 . Lehrstuhl f . Waldwachstumskunde, TUM , Freising, 2000 .
 Grote R. , Foliage and branch biomass estimation of coniferous and deciduous tree species , Silva Fennica 36 ( 2002 ) 779 - 788 .
 Grote R. , Von der Baumdimension zur Biomasse und wieder zurück - Ein neuer Ansatz zur dynamischen Modellierung von Baum- und Bestandesbiomassen , in: Dietrich H.-P., Raspe S. , Preuhsler T . (Eds.), Inventur von Biomasse- und Nährstoffvorräten in Waldbeständen, Wissenschaftszentrum Weihenstephan, München, 2002 , pp. 129 - 138 .
 Grote R. , Pretzsch H. , A Model for Individual Tree Development Based on Physiological Processes, Plant Biology 4/2 ( 2002 ) 167 - 180 .
 Hasenauer H. , Dimensional relationships of open-grown trees in Austria, For . Ecol. Manage. 96 ( 1997 ) 197 - 206 .
 Honer T.G. , Crown shape in open- and forest-grown balsam fir and black spruce , Can. J. For. Res . 1 ( 1971 ) 203 - 207 .
 Ishii H. , Clement J.P. , Shaw D.C. , Branch growth and crown form in old coastal Douglas-fir, For . Ecol. Manage. 131 ( 2000 ) 81 - 91 .
 Iwasa Y. , Cohen D. , Cohen J.A.L. , Tree Height and Crown Shape as Results of Competitive Games , J. Theoret. Biol . 112 ( 1984 ) 279 - 297 .
 Kellomäki S. , Ikonen V.-P. , Peltola H. , Kolström T. , Modelling the structural growth of Scots pine with implications for wood quality , Ecol. Modell . 122 ( 1999 ) 117 - 134 .
 Kikuzawa K. , Umeki K. , Effect of Canopy Structure on Degree of Asymmetry of Competition in Two Forest Stands in Northern Japan , Ann. Bot. 77 ( 1996 ) 565 - 571 .
 Kramer H. , Zur Qualitätsentwicklung junger Kiefernbestände in Abhängigkeit vom Ausgangsverband , Der Forst- und Holzwirt 32/ 23 ( 1977 ) 1 - 7 .
 Laar van A. , The Influence of Stand Density on Crown Dimensions of Pinus radiata D. Don, Forestry in South Africa 3 ( 1963 ) 133 - 143 .
 Mäkelä A. , Vanninen P. , Impacts of size and competition on tree form and distribution of aboveground biomass in Scots pine , Can. J. For. Res. 28/2 ( 1998 ) 216 - 227 .
 Mayer R. , Kronengröße und Zuwachsleistung der Traubeneiche auf süddeutschen Standorten , Allg. Forst- und Jagdztg . 129 ( 1958 ) 105 - 114 .
 Medhurst J.L. , Beadle C.L. , Crown architecture and leaf area index development in thinned and unthinned Eucalyptus nitens plantations , Tree Physiol . 21 ( 2001 ) 989 - 999 .
 Mitchell K.J. , Dynamics of simulated yield of Douglas-fir , For. Sci. Monogr . 17 ( 1975 ) 1 - 39 .
 Monserud R.A. , Marshall D. , Allometric crown relations in three northern Idaho conifer species , Can. J. For. Res. 29/5 ( 1999 ) 521 - 535 .
 Oltchev A. , Constantin J. , Gravenhorst G. , Ibrom A. , Heimann J. , Schmidt J. , Falk M. , Morgenstern K. , Richter I. , Vygodskaya N. , Application of a six-layer SVAT model for simulation of evapotranspiration and water uptake in a spruce forest , J. Phys. Chem. Earth 21/3 ( 1996 ) 195 - 199 .
 Opie J.E. , Predictability of individual tree growth using various definitions of competing basal area, For . Sci. 14 ( 1968 ) 314 - 323 .
 Pommerening A. , Biber P. , Stoyan D. , Pretzsch H. , Neue Methoden zur Analyse und Charakterisierung von Bestandesstrukturen, Forstw . Cbl. 1/2 ( 2000 ) 62 - 78 .
 Pretzsch H. , Konzeption und Konstruktion von Wuchsmodellen für Rein- und Mischbestände . Forstliche Forschungsberichte München . Vol. 115 , Forstwissenschaftliche Fakultät der Universität München, München, 1992 .
 Pretzsch H. , Biber P. , Dursky J., The single tree-based stand simulator SILVA: construction, application and evaluation, For . Ecol. Manage. 162 ( 2002 ) 3 - 21 .
 Pretzsch H. , Kahn M. , Grote R. , Die Fichten-Buchen-Mischbestände des Sonderforschungsbereiches “Wachstum oder Parasitenabwehr?” im Kranzberger Forst, Forstw . Cbl. 117 ( 1998 ) 241 - 257 .