Tree Physiology 20, 527–533
© 2000 Heron Publishing—Victoria, Canada
Needle and stem wood production in Scots pine (Pinus sylvestris) trees
of different age, size and competitive status
PETTERI VANNINEN1,2 and ANNIKKI MÄKELÄ1
1
Department of Forest Ecology, University of Helsinki, Helsinki, Finland
2
Present address: SAIMA—Centre for Environmental Sciences, Linnankatu 11, 57130 Savonlinna, Finland
Received June 2, 1999
Summary We studied effects of tree age, size and competi-
tive status on foliage and stem production of 43 Scots pine
(Pinus sylvestris L.) trees in southern Finland. The tree attrib-
utes related to competition included foliage density, crown ra-
tio and height/diameter ratio. Needle mass was considered to
be the primary cause of growth through photosynthesis. Both
stem growth and foliage growth were strongly correlated with
foliage mass. Consequently, differences in growth allocation
between needles and stem wood in trees of different age, size,
or position were small. However, increasing relative height in-
creased the sum of stem growth and foliage growth per unit
foliage mass, indicating an effect of available light. Sup-
pressed trees seemed to allocate more growth to stem wood
than dominant trees, and their stem growth per unit foliage
mass was larger. Similarly, trees in dense stands allocated
more growth to stem wood than trees in sparse stands. The re-
sults conformed to the pipe model theory but seemed to contra-
dict the priority principle of allocation.
Keywords: allocation, growth efficiency, pipe model, priority
principle.
Introduction
Foliage and stem wood production are major components of
tree growth, yet the effects of tree age, size and competitive
status on growth allocation between these components re-
mains the subject of both empirical and theoretical contro-
versy. Allocation of growth to stem wood has most often been
described in terms of either (1) priority between compartments
(Gordon and Larson 1968, Waring and Schlesinger 1985, Oli-
ver and Larson 1996), or (2) combinations of the pipe model
theory and functional balance theory (Valentine 1985, Mäkelä
1986, Mäkelä 1990, Nikinmaa 1992, Sievänen 1993, West
1993, Perttunen et al. 1996, Bartelink 1998). According to the
priority theory of allocation (Gordon and Larson 1968, War-
ing and Schlesinger 1985, Oliver and Larson 1996), tree pro-
duction is directed hierarchically to the different parts of a
tree, according to their priority and demand. Foliage and buds
are usually assumed to have the highest priority, followed by
roots and mobile reserves, whereas protective chemicals and
Downloaded from https://academic.oup.com/treephys/article/20/8/527/1658864 by guest on 27 November 2020
stems have the lowest priority. As an application of this the-
ory, Waring et al. (1980) proposed that a growth efficiency in-
dex (GEI), defined as stem growth per unit foliage area, could
be used to quantify tree vigor. This definition is based on the
assumption that various stresses are reflected as a decrease in
stem growth relative to foliage growth. Thus, increased com-
petition should decrease the share of carbon allocated to the
stems. Similarly, if older trees suffer from decreased overall
productivity, stem growth should decline relative to foliage
growth.
The pipe model theory (Shinozaki et al. 1964a, 1964b) de-
scribes the tree as a system of conducting pipes between fo-
liage and roots. The pipes sustain metabolism between the root
and foliage systems and provide the mechanical support. The
pipe model theory implies that growth allocation occurs to
maintain a stable ratio between foliage area and the cross-sec-
tional area of stem sapwood (Valentine 1985, Mäkelä 1986).
The theory predicts a higher proportion of stem growth in sup-
pressed trees than in dominant trees because their height
growth relative to diameter growth is faster, and diameter
growth is roughly proportional to foliage growth (Mäkelä
1986, Bartelink 1998). The pipe model predicts a growth allo-
cation pattern with tree age that largely follows the pattern of
height growth, with increasing allocation to stems in young
trees, a peak and a leveling off as the tree ages (Mäkelä 1990).
The priority principle has been widely accepted as an over-
all theoretical framework for growth allocation (Waring et al.
1980, Bassow et al. 1990, Bossel 1994, Jäghagen 1997), and
several empirical studies have provided evidence to support it.
In particular, GEI and allocation to stems have been observed
to increase with increasing site fertility (Linder and Axelsson
1982, Satoo and Madgwick 1982). However, there is little
support for the priority theory of growth allocation under con-
ditions of competition and during aging. Cannell (1985),
Albrektson and Valinger (1985) and Nilsson and Albrektson
(1993) found that allocation to stems increased with increas-
ing tree height and decreasing crown ratio (crown length/tree
length), which agrees more with the pipe model predictions.
We analyzed the effects of age, size, and competition on
growth allocation between foliage and stems, and on the
growth efficiency index. The analysis was carried out on Scots
528 VANNINEN AND MÄKELÄ

pine (Pinus sylvestris L.) from southern Finland, comprising Table 2. Distribution of sample trees in sampling categories. Some
stands of different ages and density and trees of different com- sampling categories are empty because we did not include trees in
petitive status. We used the data to assess the hypotheses un- stands YD and YS in this study. In stands MD, MS and OS, there were
no trees in some subclasses.
derlying the priority principle and the pipe model theory. We
found that needle mass provided good estimates of both stem Plot Dominant trees Suppressed trees
wood and needle production, indicating that needle mass alone Total
could be used as an indicator of total photosynthetic produc- Subclass Subclass
tion in Scots pine.
1 2 3 1 2 3

Material and methods

Downloaded from https://academic.oup.com/treephys/article/20/8/527/1658864 by guest on 27 November 2020

Data collection
Data were collected at the end of the 1995 and 1996 growing
seasons from six young (Y), middle-aged (M) and mature (O)
stands in southern Finland representing two stocking densities
(dense = D and sparse = S), regenerated either naturally or by
sowing (Table 1). All stands were located on mineral soil that
represented Myrtillus-type fertility (Cajander 1949). Trees
were selected by stratified sampling from the dominant and
non-dominant crown layers (Vuokila 1980). We further di-
vided the dominant and non-dominant trees into three sub-
groups for sampling. The subgroups consisted of small,
medium and large trees with respect to diameter at breast
height, and they covered an equal share of total basal area in
the respective crown layer. More trees were sampled from the
smaller tree classes, because of greater variation among these
trees (Table 2).
A comprehensive analysis of biomass and stem dimensions
was made for the trees, as described by Vanninen et al. (1996).
The stem and branch dimensions used in the regression analy-
ses of needle mass estimates included: branch diameter at the
stem junction and at the lowest live sub-branch, running num-
ber of the whorl, stem diameter below the whorl and the dis-
tance of the whorl from the top. Ten sample branches were
taken systematically (sampling interval = total number of liv-
ing branches divided by 10) from each tree to estimate needle
mass. In 1996, to estimate needle growth, the needles in each
sample branch were further divided into age classes.
For stem analysis, a minimum of nine stem disks was taken
from heights of 10, 50, 130 cm, at the middle of the bole below
the crown base, at the crown base, and at 25, 50, 75 and 90% of
the length of the live crown from the crown base upward. Ad-
ditionally, if the distance between any two disks was more
than 3 m, an extra disk was taken halfway between the two.
Foliage mass for each branch per tree was estimated by lin-
ear regression analysis of branch diameter and position in the
crown. Tree-level foliage mass was the sum of the branch-
level masses determined by regression. To find the most con-
sistent variables for the needle dry weight models, data for all
trees were first combined and the independent variables were
selected by the best-subset-regression method. Models with
those variables were further fitted separately for each tree (Ta-
ble 3). In the case of two sample trees, it was necessary to devi-
ate from the general variables.
Needle dry weight production was estimated on the basis of
current-year and 1-year-old needles combined as described for
the estimation of total needle mass (Table 3). It was assumed
that there was negligible mortality in these age classes, be-
cause no needle scars were evident for 1-year-old needles.
The relative errors (SErel) of tree-level foliage mass (kg; Wf)
and foliage growth (kg year –1; Gf) were calculated as esti-
mated mean error (SEmean) divided by the estimated tree-level
Wf and Gf. Tree-level mean errors (SEmean) were calculated as:
n
SEmean = ns 2 + ∑ ( x k (Y T Y ) −1 s 2x k),
T
(1)
k= 1
where s 2 is the residual mean square (MSE) of the branch-
level regression model, (Y TY) –1s 2 is the covariance matrix of

Table 1. Stand and plot characteristics. Abbreviations: H100 is expected tree height at Age 100; D is diameter at breast height (1.3 m); and Hi is
5-year mean height increment.

Plot Location Regeneration Age Site index Year of Stocking Basal area Mean D Mean tree Mean Hi
method (H100) thinning (ha –1) (m 2 ha –1) (cm) height (m) (cm)
YD 61°48′ Ν, 24°19′ E Sowing 16 30 – 18727 21 4 4.1 40
YS 61°48′ N, 24°19′ E Natural 15 30 – 2584 10.2 4 5.8 53
MD 61°20′ N, 25°00′ E Sowing 41 29 ’68 2914 35 12 11.7 29
MS 61°20′ N, 25°00′ E Sowing 41 29 ’68, ’75, ’96 693 23 20 16.3 36
OD 61°17′ N, 27°00′ E Natural 71 27 ’61 1070 33 19 20.8 13
OS 61°17′ N, 27°00′ E Natural 71 27 ’61, ’78, ’94 455 19 23 22.3 15

TREE PHYSIOLOGY VOLUME 20, 2000

 ALLOCATION OF GROWTH IN SCOTS PINE 529

Table 3. Summary of the linear regression models used to estimate fo-
liage dry weight (Wf) and needle dry weight growth (Gf) per tree. The
models were of the form y = ax1 + bx2 + c, where y is either Wf or Gf, x1
is branch cross-sectional area at the lowest live sub-branch, and x2 is
x1z1, where z1 = (distance of branch from top)/(crown length). For two
trees, a third independent variable, x3 = z12, was used in addition. The
range of percentage of variation (r 2) in individual-branch needle dry
weight or dry weight growth explained by the models, as well as the
relative standard error of the foliage estimate of the whole tree (SErel ),
based on this model, are given.
Plot y R2 SErel (%)
analyzed λs. Foliage growth was measured only in stands MD,
MS, OD and OS; therefore, the analyses of stem growth were
conducted separately on this subset of the data, as well as on
the entire data set.
Tree age was measured as the number of rings in the sample
disk at a height of 10 cm. Indicators of tree size included fo-
liage mass (Wf), tree height (H), diameter at breast height (D),
crown length (L C), and height of the crown base (HB). Indica-
tors of competitive status included relative height (Hrel), crown
ratio (RC = L C/H), and height/diameter ratio (H/D). Relative
height was defined as tree height relative to maximum tree

Downloaded from https://academic.oup.com/treephys/article/20/8/527/1658864 by guest on 27 November 2020

height in the plot sample. Foliage density in the crown with re-
YD Wf 0.62–0.95 1.3–7.3
spect to crown volume (FV) and crown surface area (FS) were
YS Wf 0.67–0.87 3.5–7.8
MD Wf 0.50–0.98 1.8–9.2 also studied as indicators of competitive status. Both FV and
MS Wf 0.77–0.91 3.3–3.9 FS were calculated based on the assumption that crowns were
OD Wf 0.66–0.99 0.5–8.8 cones with length HC and radius equal to mean crown radius at
OS Wf 0.86–0.98 2.1–3.5 the height where the crown was widest. The number of mea-
MD Gf 0.48–0.98 2.3–27.5 surements for FS and FV was only 26, because crown radius
MS Gf 0.72–0.98 1.7–12.9 was not measured in five trees in the sample.
OD Gf 0.63–0.98 2.3–7.9 The Statistix program (Analytical Software Co., La Jolla,
OS Gf 0.76–0.98 1.9–3.7
CA) was used for the statistical analyses. A two-sample t-test
was used to evaluate differences in Gf and λs between plots.
Effects of age, size and competitive status were evaluated by
the regression coefficients of the branches used for model fit- correlation and regression analyses.
ting, xk is the vector of independent variables at branch k (in all
branches where the model was applied), and n is the number of
Results
branch needle estimates in the tree. The first term in the sum
represents the error attributable to summing all of the branch Stem and foliage growth
estimates, and the second term is the error attributable to the
Both Gs, Gf and their sum Gs + Gf (total growth), calculated as
use of a model in determining the individual-branch estimates.
the mean value for the 1995 and 1996 growing seasons, were
Stem wood production was determined by stem analysis
strongly correlated with Wf, with r = 0.92, 0.97 and 0.96, re-
and wood density analysis. Stem analysis was carried out with
spectively (n = 31) (Figure 1). The intercept of the regression
the WinDendro and XlStem programs (Regent Instruments,
line was insignificant in all cases. When the total data set, in-
Inc., Québec City, Québec, Canada). Annual ring width was
measured in all cardinal points and also inter-cardinal points
for the 1996 data set. Tree volume was calculated as the sum of
truncated cones based on the sample disks. Volume was con-
verted to mass piece by piece, based on measurements of
wood density of the sample disks, and assuming that the wood
density between two disks was the mean of the two density
measurements. Volume growth was converted to biomass
growth based on the mean wood density of each stem. Annual
means for needle growth and stem growth were determined
from the combined data for 1996 and 1997.

Growth allocation
Allocation of growth between stems and foliage was analyzed
in terms of stem growth, Gs (kg year –1), foliage growth Gf (kg
year –1), relative allocation to stems (λs = Gs /Gf + Gs), and a
growth efficiency index based on foliage weight (abbreviated
as GEIW to distinguish it from the original definition based on Figure 1. Mean annual stem and foliage growth in 1995–1996 in
foliage area; Waring et al. 1980). We defined GEIW as Gs /Wf, stands MD, MS, OD and OS as a function of foliage mass. Symbols:
䊉 = foliage growth, regression line y = 0.293x – 0.116, r 2 = 0.95,
where Gs (kg year –1) is stem growth and Wf (kg) is foliage
RMSE = 0.253, and Pconstant = 0.15. 䉭 = stem growth, regression line
mass. Foliage area was replaced by foliage mass because no y = 0.423x + 0.238, r 2 = 0.84, RMSE = 0.692, and Pconstant = 0.28. 䉬 =
information was available on foliage area. stem and foliage growth summed, regression line y = 0.716x + 0.122,
Because the allocation to foliage, λf , equals 1 – λs, we only r 2 = 0.92, RMSE = 0.776, Pconstant = 0.62. For all regressions, n = 31.

TREE PHYSIOLOGY ON-LINE at http://www.heronpublishing.com

530 VANNINEN AND MÄKELÄ

cluding stands YD and YS, was used for the correlation be-
tween stem growth (Gs) and foliage mass (Wf), the correlation
coefficient increased slightly (r = 0.93, n = 43).
The accuracy of the prediction of growth from foliage mass
was improved by using information about the size and position
of the tree (Table 4). For total growth, the best single addi-
tional variable was relative height. Foliage density in the
crown, measured as both volumetric density and foliage mass
over surface area, reduced the total growth predicted from fo-
liage mass, whereas increasing crown length increased the to-
tal growth predicted from foliage mass. None of the other
statistically significant (P = 0.053). Increasing crown length
or diameter at breast height also increased stem growth per
unit foliage mass. When the total data set was considered for
stem growth, relative height was no longer significant (P =
0.150), but both the surface area density of foliage and volu-
metric density of foliage were significant. For the total data
set, tree height, crown length and height to the crown base sig-
nificantly improved the prediction of stem growth from fo-
liage mass (Table 4). Tree age did not significantly affect the
dependence of stem and foliage growth on foliage mass.

Downloaded from https://academic.oup.com/treephys/article/20/8/527/1658864 by guest on 27 November 2020

variables examined improved the prediction. The prediction of
foliage growth was improved by using crown ratio. Foliage
growth increased with increasing crown ratio. Foliage density
did not improve the prediction of foliage growth from foliage
mass.
The prediction of stem growth from foliage mass was im-
proved by using the relative height of the tree. Stem growth in-
creased with increasing relative height. The surface area
density of foliage reduced stem growth per unit foliage mass.
The effect of volumetric density of foliage was similar but not
Growth efficiency index
As shown in Figure 1, GEIW was relatively constant. How-
ever, the surface area density of foliage was negatively corre-
lated with GEIW. In the larger data set, this correlation was
stronger, and there was also a significant correlation between
GEIW and the volumetric density of foliage, and between
GEIW and H/D, such that more slender trees had a larger GEIW
(Table 5). In all age groups, GEIW tended to be larger in the
dense plots than in the sparse plots, but the differences were
not statistically significant (Figure 2a).

Table 4. Single factors improving the linear regression between fo-
liage mass and total growth, stem growth and foliage growth. The re-
gression equation was y = aWf + bx + c, where y is the growth
component, Wf is foliage mass, x is the factor in question, and a, b, and
c are regression coefficients. Factors have been reported if b ≠ 0 at P <
0.05. Effects of each factor on the linear regressions are reported as
positive (+) or negative (–); r 2 = percent of variation in y explained by
the regression; RMSE = root mean square error, P = probability that b
= 0; and n = number of observations. Symbols: D = diameter at breast
height; H = tree height; Hrel = relative height; HC = crown height; HB =
height of crown base; FS = surface area density of foliage; FV = volu-
metric density of foliage; and RC = crown ratio.
Growth allocation
Growth allocation to stems, λs, was positively correlated with
GEIW, as would be expected on the basis of our definitions of
the terms. A significant correlation was also found with both
surface area density of foliage and H/D, but no other indicators
of size or position correlated with λs (Table 5).
The two-sample t-test of the four plots where λs was mea-
sured showed that allocation to stems was significantly higher
(P = 0.0296) in the dense middle-aged plot (MD) than in the
sparse plot of the same stand (MS). The sparse plot MS dif-
fered from the dense plots (MD and OD) but not from the old
sparse plot (P = 0.185) (Figure 2b).

Factor Effect r2 RMSE P n

Discussion
Total growth (Plots MD, MS, OD, OS) Accurate measurements of biomass growth by compartment
Hrel + 0.95 0.400 0.0009 31
FS – 0.94 0.397 0.0008 26
HC + 0.94 0.475 0.0127 31
Table 5. Variables correlated with growth efficiency index (GEIW)
FV – 0.92 0.525 0.0261 26
and relative allocation to stems (λs) at P < 0.05.
Foliage growth (Plots MD, MS, OD, OS)
Variable R P n
RC + 0.96 0.0465 0.0145 31
Stem growth (Plots MD, MS, OD, OS) GEIW (Plots MD, MS, OD, OS)

Hrel + 0.90 0.332 0.0008 31 FS –0.523 0.0000 26

FS – 0.88 0.3674 0.0011 26
GEIW (Plots YD, YS, MD, MS, OD, OS)
HC + 0.87 0.401 0.0150 31
D + 0.87 0.416 0.0269 31 FS –0.641 0.0000 38
FV –0.438 0.0060 38
Stem growth (Plots YD, YS, MD, MS, OD, OS)
H/D 0.39 0.0082 43
FS – 0.89 0.319 0.0004 38
λs (Plots MD, MS, OD, OS)
HC + 0.89 0.314 0.0011 43
HB + 0.88 0.369 0.0362 43 GEIW 0.72 0.0000 31
H + 0.88 0.350 0.0108 43 FS –0.47 0.0143 26
FV – 0.86 0.409 0.0420 38 H/D 0.43 0.0163 31

TREE PHYSIOLOGY VOLUME 20, 2000

 ALLOCATION OF GROWTH IN SCOTS PINE
Figure 2. (a) Growth efficiency index (GEIW) in the six plots. Error
bars indicate standard deviation. (b) Relative allocation to stems (λs)
in the four plots where foliage growth was measured. Error bars indi-
cate standard deviation.
are difficult to obtain for trees, especially for older trees, be-
cause many assumptions and compromises have to be made.
In this study, we assumed that means of needle and stem wood
production over 2 years described the allocation between the
compartments more precisely than 1-year means. However,
because Scots pine in southern Finland usually supports only
about four annual cohorts of foliage, foliage growth as we de-
fined it closely approximated foliage dry weight. Furthermore,
we used similar independent variables and the same statistical
method to estimate both foliage growth and total foliage dry
weight. Thus, a correlation between foliage growth and total
foliage dry weight is to be expected.
Despite the experimentally built-in correlation between
needle mass and needle production, needle mass alone was a
good estimator of stem wood and needle mass production in
this study. Similar results have been obtained for the depend-
ence of volume increment on foliage mass in Scots pine and
Norway spruce (O’Hara et al. 1999). These findings indicate
that foliage mass alone is strongly indicative of total
photosynthetic production. The combined production of fo-
liage and stem wood per unit foliage mass was only slightly
dependent on indicators describing the amount of available
light; e.g., relative height—an indicator of shading by neigh-
bors, and foliage density, which affects self-shading. There
TREE PHYSIOLOGY ON-LINE at http://www.heronpublishing.com
531
were no differences between the dense and sparse stands, or
between the older and younger stands. The effects of low
irradiance may have been partly offset by acclimation, be-
cause trees are able to compensate for low irradiances by in-
creasing the specific needle area (Kellomäki and Oker-Blom
1981). The finding that surface area density of foliage, FS,
gave better results than volumetric density of foliage, FV, is in
accordance with light interception models that relate light cap-
tured by the crown to its total shadow area (which is dependent
on the surface area and amount of foliage) (Oker-Blom and
Kellomäki 1982).
Downloaded from https://academic.oup.com/treephys/article/20/8/527/1658864 by guest on 27 November 2020
It has generally been assumed that productivity decreases as
trees age (Waring and Schlesinger 1985). The old trees in our
study appeared to represent old trees based on their height
growth pattern over the last five years (Table 1); however, nei-
ther total growth (Gs + Gf), nor the separate relationships be-
tween growth of foliage (Gf) and growth of stems (Gs) on the
one hand and foliage mass on the other showed an age depend-
ence. There are several possible explanations for this finding.
First, both growth and foliage mass may decline in older trees
(cf. Table 1), thereby masking the effect of age on growth.
Second, only stem and foliage growth components were mea-
sured, and there could be reductions in the branch or root
growth components. Third, increasing availability of light and
other microclimatic changes in older stands might have
masked the effect of age on growth.
Regression analysis indicated that deviations of foliage
growth and stem growth from the pattern predicted by the
amount of foliage were, to some extent, in opposite directions.
The observed correlations of GEIW and λs with H/D may pro-
vide an explanation. If (1) total growth is proportional to fo-
liage, and (2) stem growth can be predicted from foliage mass
and height, then both allocation to stems and GEIW must be
proportional to the ratio of height to foliage. Because foliage
and stem diameter are usually closely connected (Kittredge
1944, Loomis et al. 1966), it follows that λs and GEIW should
also be correlated with H/D.
These findings are in agreement with other empirical stud-
ies on Scots pine (Albrektson and Valinger 1985, Nilsson and
Albrektson 1993, O’Hara et al. 1999), and also with deriva-
tions based on the pipe-model theory (Mäkelä 1986, Bartelink
1998). Our finding that increasing competition increased allo-
cation to stems corroborates reports that height to diameter
(Ek 1971, Weiner and Thomas 1992) and height to foliage
(Albrektson 1980) ratios are larger in suppressed trees than in
dominant trees.
In a study of branch growth, Mäkinen and Colin (1998)
found that H/D explained the effects of social position more
consistently than the other competition indices examined. If
growth allocation between height and diameter is regulated by
the environment, then H/D is a cumulative index integrating
the long-term effects of the local environment. The depend-
ence of allocation on H/D could therefore be interpreted as ac-
climation to the local environment through variation of the
growth pattern.
Contrary to our results, some studies have shown that com-
532 VANNINEN AND MÄKELÄ
petition has no effect on the ratio of stem wood production to
foliage production (Waring et al. 1980, Waring et al. 1981), or
even reduces it (Clutter 1980). Nilsson and Gemmel (1993)
found that competition increased growth allocation to stem
wood in Scots pine, but had little effect in Norway spruce. The
effect of competition on stem wood allocation could be related
to the different light requirements of the two species. For ex-
ample, Scots pine actively searches for high light conditions
by investing in the growth of the upper part of the crown when
shaded.
Foliage density was the most important single factor charac-
terizing GEIW and growth allocation to stems. One interpreta-
tion of this finding is that it is a result of shading; i.e., total
growth and stem growth are reduced in parallel. This interpre-
tation would be justified if the effects of relative height on total
growth and stem growth were similar. However, the effect of
relative height in the regression analysis on the small data set
was not apparent in the larger data set, whereas the effect of fo-
liage density was more pronounced in the larger data set than
in the small data set (Table 4). Mäkelä and Vanninen (1998)
found that the density of foliage in Scots pine was strongly
correlated with the competitive status of the tree, i.e., sup-
pressed trees had sparser crowns than more dominant trees,
which is in agreement with our results on the dependence of
growth allocation on H/D.
Although the concept of tree vigor is not very precise, our
results do not support the hypothesis that GEIW increases with
increasing vigor. In contrast, the factors correlating with
GEIW, i.e., increasing H/D ratio and decreasing foliage den-
sity, are associated with decreasing tree vigor. However, our
results are not directly comparable with those obtained by
Waring and Schlesinger (1985), because we calculated the
growth efficiency index relative to foliage mass rather than fo-
liage area.
In conclusion, our results conform to derivations from the
pipe-model theory (Mäkelä 1986, Bartelink 1998), but they do
not support the priority theory, which predicts that trees with
low productivity invest a high proportion of growth to foliage.
It is possible that the relative demands of foliage and stems dif-
fer for dominant and suppressed trees and for species, reflect-
ing fundamental differences in growth adaptations.
Acknowledgments
We thank Prof. Kari Mielikäinen and Dr. Harri Mäkinen for their sup-
port. Dr. Eero Nikinmaa provided many valuable comments on an
earlier version of this paper. Thanks also to the field and laboratory
assistants and data compiler Hannele Saloseutu for their great work
during the study.
References
Albrektson, A. 1980. Relationships between tree weight fractions and
conventional silvicultural measurements. In Structure and Func-
tion of Northern Coniferous Forests—An Ecosystem Study. Ed. T.
Persson. Ecol. Bull. 32:315–327.
TREE PHYSIOLOGY VOLUME 20, 2000
Albrektson, A. and E. Valinger. 1985. Relations between tree height
and diameter, productivity and allocation of growth in Scots pine
(Pinus sylvestris L.) sample tree material. In Crop Physiology of
Forest Trees. Eds. P.M.A. Tigerstedt, P. Puttonen and V. Koski.
University of Helsinki, Finland, pp 95–105.
Bartelink, H.H. 1998. A model of dry matter partitioning in trees.
Tree Physiol. 18:91–101.
Bassow, S.L., E.D. Ford and R.A. Kiester. 1990. A critique of car-
bon-based tree growth models. In Process Modeling of Forest
Growth Responses to Environmental Stress. Eds. R.K. Dixon, R.S.
Meldahl, G.A. Ruark and W.G.Warren. Timber Press, Inc., Port-
land, OR, pp 50–57.
Downloaded from https://academic.oup.com/treephys/article/20/8/527/1658864 by guest on 27 November 2020
Bossel, H. 1994. TREEDYN3 forest simulation model. Mathematical
model, program documentation, and simulation results. Ber. For-
schungzentrums Waldökoysysteme, Reihe B, Bd. 35, 118 p.
Cajander, A.K. 1949. Forest types and their significance. Acta For.
Fenn. 56, 71 p.
Cannell, M.G.R. 1985. Dry matter partitioning in tree crops. In At-
tributes of Trees as Crop Plants. Eds. M.G.R. Cannell and J.E.
Jackson. Institute of Terrestrial Ecology, U.K., pp 160–193.
Clutter, J.L. 1980. Development of taper functions from variable-top
merchantable volume equations. For. Sci. 26:117–120.
Ek, A.R. 1971. Size–age relationships for open-grown red pine. The
University of Wisconsin, Forestry Research Notes 156.
Gordon, C. and P.R. Larson. 1968. Seasonal course of photosynthe-
sis, respiration, and distribution of 14C in young Pinus resinosa
trees as related to wood formation. Plant Physiol. 43:1617–1624.
Jäghagen, K. 1997. Impact of competition on survival, growth, and
tree characteristics of young conifers. Ph.D. Thesis, Swedish Uni-
versity of Agricultural Sciences, Umeå, Sweden. Acta Univers-
itatis Agriculturae Suecica, Silvestria 32:1–29.
Kellomäki, S. and P. Oker-Blom. 1981. Specific needle area of Scots
pine and its dependence on light conditions inside the canopy.
Silva Fenn. 15:190–198.
Kittredge, J. 1944. Estimation of the amount of foliage on trees and
stands. J. For. 42:905–912.
Linder, S. and B. Axelsson. 1982. Changes in carbon uptake and allo-
cation patterns as a result of irrigation and fertilization in a young
Pinus sylvestris stand. In Carbon Uptake and Allocation in
Sub-Alpine Ecosystems as a Key to Management. Ed. R.H. War-
ing. Oregon State University, Forest Research Laboratory. Cor-
vallis, OR, pp 38–44.
Loomis, R.M., R.E. Phares and J.S. Crosby. 1966. Estimating foliage
and branchwood quantities in shortleaf pine. For. Sci.12:30–39.
Mäkelä, A. 1986. Implications of the pipe model theory on dry matter
partitioning and height growth in trees. J. Theor. Biol. 123:
103–120.
Mäkelä, A. 1990. Adaptation of light interception computations to
stand growth models. Silva Carelica 15:221–239.
Mäkelä, A. and P. Vanninen. 1998. Impacts of size and competition
on tree form and distribution of above-ground biomass in Scots
pine. Can. J. For. Res. 28:216–227.
Mäkinen, H. and F. Colin. 1998. Predicting branch angle and branch
diameter of Scots pine from usual tree measurements and stand
structural information. Can. J. For. Res. 28:1686–1696.
Nikinmaa, E. 1992. Analyses of the growth of Scots pine: matching
structure with function. Acta For. Fenn. 235:68.
Nilsson, U. and A. Albrektson. 1993. Productivity of needles and al-
location of growth in young Scots pine trees of different competi-
tive status. For. Ecol. Manage. 62:173–187.
Nilsson, U. and P. Gemmel. 1993. Changes in growth and allocation
of growth in young Pinus sylvestris and Picea abies due to compe-
tition. Scand. J. For. Res. 8:213–222.
 ALLOCATION OF GROWTH IN SCOTS PINE
Oker-Blom, P. and S. Kellomäki. 1982. Theoretical computations on
the role of crown shape in the absorption of light by forest trees.
Math. Biosci. 59:291–311.
Oliver, C.D. And B.C. Larson. 1966. Forest stand dynamics.
McGraw-Hill, New York, 520 p.
O’Hara, K.L., E. Lähde, O. Laiho, Y. Norokorpi and T. Saksa. 1999.
Leaf area and tree increment dynamics on a fertile mixed-conifer
site in southern Finland. Ann. For. Sci. 56:237–247.
Perttunen, J., R. Sievänen, E. Nikinmaa, H. Salminen, H. Saarenmaa
and J. Väkevä. 1996. LIGNUM: A tree model based on simple
structural units. Ann. Bot. 77:87–98.
Satoo, T. and H.A.I. Madgwick. 1982. Forest biomass. Martinus
Nijhoff/Junk, The Hague, London, 152 p.
Shinozaki, K., K. Yoda, K. Hozumi and T. Kira. 1964a. A quantita-
tive analysis of plant form: the pipe model theory. I. Basic analy-
ses. Jpn. J. Ecol. 14:97–105.
Shinozaki, K., K. Yoda, K. Hozumi and T. Kira. 1964b. A quantita-
tive analysis of plant form: the pipe model theory. II. Further evi-
dence of the theory and its application in forest ecology. Jpn. J.
Ecol. 14:133–139.
Sievänen, R. 1993. A process based model for dimensional growth of
even-aged stands. Scand. J. For. Res. 8:28–48.
TREE PHYSIOLOGY ON-LINE at http://www.heronpublishing.com
533
Valentine, H.T., 1985. Tree-growth models: derivations employing
the pipe model theory. J. Theor. Biol. 117:579–584.
Vanninen, P., H. Ylitalo, R. Sievänen and A. Mäkelä. 1996. Effects of
age and site quality on the distribution of biomass in Scots pine
(Pinus sylvestris L.). Trees 10:231–238.
Vuokila, Y. 1980. Metsänkasvatuksen perusteet ja menetelmät. In
Finnish. WSOY, Porvoo, Finland, p 257.
Waring, R.H., W.G. Thies and D. Muscato. 1980. Stem growth per
unit of leaf area: a measure of tree vigour. For. Sci. 26:112–117.
Waring, R.H., K. Newman and J. Bell. 1981. Efficiency of tree
crowns and stemwood production at different canopy leaf densi-
ties. Forestry 54:129–137.
Downloaded from https://academic.oup.com/treephys/article/20/8/527/1658864 by guest on 27 November 2020
Waring, R.H. and W.H. Schlesinger. 1985. Forest ecosystems: con-
cepts and management. Academic Press, New York, 390 p..
Weiner, J. and S.C. Thomas. 1992. Competition and allometry in
three species of annual plants. Ecology 73:648–656.
West, P.W. 1993. Model of above-ground assimilate partitioning and
growth of individual trees in even-aged forest monoculture.
J. Theor. Biol.161:369–394.
Downloaded from https://academic.oup.com/treephys/article/20/8/527/1658864 by guest on 27 November 2020