Tree Physiology 28, 331–342
© 2008 Heron Publishing—Victoria, Canada
Light interception and partitioning between shoots in apple cultivars
influenced by training
JEAN STEPHAN,1 HERVÉ SINOQUET,1,2 NICOLAS DONÈS,1 NICOLAS HADDAD,3 SALMA
TALHOUK3 and PIERRE-ÉRIC LAURI 4
1
UMR547 PIAF, INRA, Univ Blaise Pascal, F-63100 Clermont Ferrand, France
2
Corresponding author ([email protected])
3
Faculty of Agricultural and Food Sciences, American University of Beirut, Riad El Solh 1107 2020, P.O. Box 11-0236, Beirut, Lebanon
4
UMR DAP, INRA-SUPAGRO-CIRAD-UM II, Equipe ‘Architecture et Fonctionnement des Espèces Fruitières’, 2 place Viala, 34060, Montpellier,
Cedex 1, France
Received May 15, 2007; accepted September 20, 2007; published online January 2, 2008
Summary The effect of two training systems (Central
Leader with branch pruning versus Centrifugal Training with
minimal pruning, i.e., removal of fruiting laterals only) on canopy structure and light interception was analyzed in three architecturally contrasting apple (Malus domestica Borkh.) cultivars: ‘Scarletspur Delicious’ (Type II); ‘Golden Delicious’
(Type III); and ‘Granny Smith’(Type IV). Trees were 3D-digitized at the shoot scale at the 2004 and 2005 harvests. Shoots
were separated according to length (short versus long) and type
(fruiting versus vegetative). Leaf area density (LAD) and its
relative variance (ξ), total leaf area (TLA) and crown volume
(V) varied consistently with cultivar. ‘Scarletspur Delicious’
had higher LAD and ξ and lower TLA and V compared with the
other cultivars with more open canopies. At the whole-tree
scale, training had no effect on structure and light interception
parameters (silhouette to total area ratio, STAR; projected leaf
area, PLA). At the shoot scale, Centrifugal Training increased
STAR values compared with Central Leader. In both training
systems, vegetative shoots had higher STAR values than fruiting shoots. However, vegetative and fruiting shoots had similar
TLA and PLA in Centrifugal Trained trees, whereas vegetative
shoots had higher TLA and PLA than fruiting shoots in Central
Leader trees. This unbalanced distribution of leaf area and light
interception between shoot types in Central Leader trees partly
resulted from the high proportion of long vegetative shoots that
developed from latent buds. These shoots developed in the interior shaded zone of the canopy and therefore had low STAR
and PLA. In conclusion, training may greatly affect the development and spatial positioning of shoots, which in turn significantly affects light interception by fruiting shoots.
Keywords: canopy structure, central leader, centrifugal training, Malus domestica, reiteration, shoot type, spatial pattern,
tree ideotype.
Introduction
In fruit trees, fruit yield and quality depend on the light microclimate. At the orchard and tree scales, fruit yield of healthy
and well-watered trees is related to total light interception
(Jackson 1980, Palmer 1989, Robinson and Lakso 1991), because photosynthetic carbon fixation depends mainly on the
sunlight captured by a tree or an orchard. At the intra-tree
scale, fruit quality may show large changes in response to architectural position—e.g., in apple (Malus domestica Borkh.),
fruits on 1-year-old wood are smaller than fruits on older wood
(Volz et al. 1994; Lauri and Trottier 2004)—or uneven distribution of light within the crown (Tustin et al. 1988). Usually,
shade reduces fruit mass and fruit quality attributes like color
and soluble sugar and secondary metabolite concentrations
(Lakso 1980, Doud and Ferree 1980, Robinson et al. 1983,
Awad et al. 2001). Shading decreases shoot photosynthesis
and fruit temperature (Abbott 1984) and may change the light
spectrum (Awad et al. 2001). Local irradiance may also affect
flower bud initiation and development (Jackson 1980, Palmer
1989) and leaf attributes affecting photosynthetic capacity,
namely specific leaf area (Palmer et al. 1992).
Fruit growers can improve orchard light microclimate by
manipulating canopy structure. At the orchard scale, light interception depends on planting pattern, e.g., tree spacing (Palmer et al. 1992), rectangular versus square plantation design
(Wagenmakers 1991a), single versus multi-row system (Wagenmakers 1991b) and row orientation (Jackson and Palmer
1972, Palmer 1989). At the tree scale and for a given genotype,
light interception is affected by pruning and training procedures (Johnson and Lakso 1986, Rom and Barritt 1990, Ferree
et al. 1992, Wünsche et al. 1996). Many studies on irradiance
and light interception have compared training systems at the
tree scale (e.g., Palmette-Leader versus Central Leader (Elfving et al. 1990); V-trellis, Solen Y-trellis, Geneva Y-trellis
versus Slender Spindle (Hampson et al. 2002)) or the orchard
332
STEPHAN, SINOQUET, DONÈS, HADDAD, TALHOUK AND LAURI
scale (Tustin et al. 1998). However, light environment has
rarely been investigated at the shoot scale. According to Elfving and Forshey (1976) and Wünsche et al. (2000), a major
drawback of most training systems is that pruning keeps fruiting shoots inside the tree canopy, whereas vegetative shoots
are located in the outer part. This arrangement is contrary to
the main objective of tree training which is not only to maximize light interception at the tree scale but also to optimize
light penetration to the fruiting shoots, especially early in the
season (Lakso and Corelli-Grappadelli 1992). This is particularly true in apple trees where fruit growth is closely related to
light interception by spur leaves close to the flower/fruit cluster. Spur leaves appear to be the main supporters of early fruit
growth (Volz et al. 1994)
Recently, a Centrifugal Training system for apple was proposed with the aim of improving light interception by fruiting
shoots. It is based on the removal of young spurs along the
trunk, and at the base and on the underside of branches (Lauri
et al. 1997). The resulting light well and the porous mantle significantly increased total light interception and allocation to
fruiting shoots (Willaume et al. 2004) and resulted in larger
fruits (Lauri et al. 2007).
Several methods have been proposed for studying light interception by orchard trees. On the one hand, empirical studies
based on light measurements within and below the canopies
have been attempted (Ferree et al. 1992, Robinson et al. 1991).
On the other hand, modeling approaches relating canopy geometry to light interception have been developed. Most models have been based on the turbid medium analogy (Ross
1981), where trees are abstracted as simple geometric shapes
with a uniform density of leaf area (Charles-Edwards and
Thorpe 1976, Palmer 1977, Green et al. 2003, Oyarzun et al.
2007). These models compute total light interception at the orchard scale and light distribution within tree crowns. For the
latter, isolines of light transmission are computed to define the
fraction of leaf area or canopy volume where available light is
above or below a given threshold (Jackson and Palmer 1981).
Another way to compute light microclimate from canopy
structure is to use 3D plant mock-ups where plant geometry is
explicitly described (Sinoquet et al. 1998). This entails defining tree structure by the geometric attributes of all leaves on
the tree, namely size, spatial position and orientation. Plant
mock-ups can be created from field measurements on real
plants with a 3D digitizer (Sinoquet and Rivet 1997) or from
morphogenetic rules (Prusinkiewicz and Lindenmayer 1990).
Although field data acquisition is tedious, the 3D approach is
powerful, because it allows the computation of light interception at any scale.
Previously, we analyzed the influence of canopy manipulation on architecture (Stephan et al. 2007). Here, we have used
3D apple tree mock-ups to study relationships between canopy
structure and light interception at the whole-tree, shoot-type
and individual-shoot scales. We aimed to compare the responses of three apple cultivars of contrasting architectural
types, and to assess effects of changing from a classical central
leader training system to a centrifugal system. For this purpose, field-grown apple trees were 3D-digitized over 2 years.
We studied trees in a low-density orchard (714 trees ha – 1 ) to
ensure optimal architectural development of each tree according to training system. However, as previously shown by
Hampson et al. (2004), there are complex interactions between
tree density and training system in determining yield and individual fruit quality (especially fruit mass and color). Because
there is a predominant effect of tree density on yield in the
early years, whereas there is no significant effect of training
system before the sixth year in the orchard, we did not conduct
a yield analysis.
Materials and methods
Plant material and training systems
The experiment was located in the Bekaa valley, at 900 m a.s.l.
at the American University of Beirut research field (Haouch
Sneid; 33°95′ N, 36°02′ E) in Lebanon. Trees of three cultivars
were planted in 1999 in a randomized block design. The
cultivars belonged to contrasting ideotypes according to Lespinasse’s (1992) typology: ‘Scarletspur Delicious’ (Type II),
‘Golden Delicious’ (Type III) and ‘Granny Smith’ (Type IV),
hereafter referred to as ‘Scarletspur’, ‘Golden’ and ‘Granny’,
respectively. Trees were grafted on M7 rootstock. The orchard
layout was 4 × 3.5 m, with North–South orientation, and included three blocks of trees. Agricultural practices included irrigation with micro-sprinklers, standard fertilization and
spraying.
From planting, trees of all cultivars were trained to obtain a
Central Leader (Heinicke 1975). The leader was pruned from
the first year to obtain a stronger trunk made of consecutive reiterated complexes. One-third of each 1-year-old lateral shoot
was pruned to reduce competition between branches and trunk.
Only shoots competitive with the trunk were pruned. In winter
2004, six healthy trees of each cultivar (two trees per block)
were chosen at random excluding border trees. In each pair,
one tree was trained as a Central Leader (L) and the other according to the Centrifugal Training system (C) as described by
Willaume et al. (2004). In winter 2004, C-training consisted of
making a light well by removing all shoots and buds (extinction pruning) along the trunk and on the underside of each
branch. All branches along the trunk were kept. In 2005, almost no regrowth was found at sites subjected to extinction
pruning in 2004. Complementary extinction pruning was carried out to maintain the same crop load as in 2004. Extinction
pruning on C trees and shortening cuts on L trees were made
before full bloom and bud burst, respectively. In both treatments, fruit was thinned to one per flower cluster.
3D digitizing of trees and foliage reconstruction
Trees were 3D digitized at the current-year shoot level one
month before harvest in 2004 and 2005, in August or September of each year depending on the cultivar. Digitizing involves
measuring the spatial coordinates of the proximal and distal
tips of all shoots with a digitizer Polhemus Fastrak (Polhemus,
Cochester, VT) (Sinoquet and Rivet 1997). The method requires an operator to locate the digitizer pointer at each point
TREE PHYSIOLOGY VOLUME 28, 2008
MULTISCALE LIGHT INTERCEPTION IN 3D APPLE TREES
to be measured. The spatial coordinates of each point are recorded (Polhemus 1993). Data acquisition is driven by Pol95
Version 1.0 software (1999; B. Adam, UMR PIAF INRAUBP, Clermont-Ferrand) available at http://www2.clermont.
inra.fr/piaf/eng/download/download.php.
While digitizing, the operator recorded shoot type. Current-year shoots were classified as fruiting shoots (FS) or vegetative shoots (VS) according to Lauri and Kelner (2001) and
Willaume et al. (2004): the FS included the fruited or aborted
bourse and its associated bourse shoot(s). Vegetative shoots
were classified as short (SVS, < 4 cm) or long (LVS, > 4 cm)
shoots. Reiterated LVS originating from latent buds were distinguished and termed R-LVS, and LVS resulting from terminal growth of existing shoots were termed T-LVS.
In addition to shoot-level digitizing, 20 annual shoots of
each category of each cultivar were randomly chosen and digitized at the leaf level according to Sinoquet et al. (1998). For
each leaf, the pointer was set at the junction between the petiole and lamina, with the pointer oriented parallel to both the
midrib axis and the lamina plane. This setup allowed us to derive leaf orientation angles (i.e., midrib azimuth and inclination) and lamina rolling around the midrib. Leaf length and
width were also measured. Leaf samples were harvested for
individual leaf area measurement. Leaves were scanned and
the images processed with Scion software (Frederick, MD).
Allometric relationships and leaf angle distributions were
computed.
Foliage associated with each shoot in the crown was reconstructed according to the method of Sonohat et al. (2006): (1)
shoot length was computed from spatial coordinates of shoot
tips; (2) allometric relationships between shoot length, leaf
area and number of leaves were computed for each year,
cultivar and shoot type, according to the method proposed by
Palmer (1987); (3) all leaves in a shoot were assumed to have
the same area and to be regularly positioned along the shoot
axis; (4) leaf length and width were computed from an allometric relationship for leaf area and length, respectively; and
(5) leaf angles were randomly sampled in the distribution set
from digitized data at the leaf scale. This sequence of computations was implemented in software written in FORTRAN,
where inputs were the digitized tip coordinates of shoots and
the parameters of the reconstruction rules. The reconstruction
output was a collection of leaves with spatial coordinates, orientation angles and dimensions (Figure 1). The method was
previously shown suitable for computing light properties at the
tree and shoot scales for peach trees (Sonohat et al. 2006). The
method has also been used with apple trees (Willaume et al.
2004).
Derivation of tree canopy structure parameters
Canopy structure parameters were computed from the reconstructed foliage data. Total leaf area (TLA) was computed as
the sum of areas of all leaves at the scales of the whole tree and
the shoot types (FS and VS). Crown volume (V ) was computed as follows. A bounding box was defined around the tree
according to the minimum and maximum values of leaf coordinates along axes x, y and z. The bounding box was then di-
333
vided into cubic volume elements of 0.2 m called voxels.
Crown volume was finally approximated as the cumulated volume of vegetated voxels, i.e., voxels containing at least one
leaf (Figure 1). Mean leaf area density (LAD) was computed
as the ratio of TLA to V. The relative variance of LAD
(ξ)—which has previously been reported as a main contributor
to foliage clumping (Sinoquet et al. 2005)—was computed
from values of LAD in voxels:
1
ξ=
nv
nv
∑
v =1
(LAD
v
– LAD
)
2
LAD
(1)
where nv is the number of vegetated voxels, LADv is LAD in
voxel v and LAD is mean LAD in the tree crown (LAD =
TLA/V ).
The spatial distribution of leaf area was also visually assessed by dividing the canopy space into toric voxels of
squared section (0.2 × 0.2 m) that are vertically symmetric
about the tree trunk. Leaves were assigned to toric voxels according to their spatial coordinates. Such spatial division allowed us to derive 2D maps of changes in LAD as a function of
altitude and distance from the tree trunk. This was computed at
the tree scale and per shoot type (FS and VS).
Computation of light interception attributes
Light interception at both the tree and shoot scales was computed from the 3D plant mock-ups. Projected leaf area (PLA,
m2 )—also called silhouette area—and silhouette to total area
ratio (STAR, Carter and Smith 1985, Oker-Blom and Smolander 1988, Hemmerlein and Smith 1994) were computed
with VegeSTAR Version 3.0 software (2002; B. Adam, N. Donès and H. Sinoquet, UMR PIAF INRA-UBP, Clermont-Ferrand). Projected leaf area characterizes light intercepting leaf
area, whereas STAR corresponds to mean leaf irradiance relative to incident radiation. The two variables are related as
STAR = PLA/TLA.
Unlike STAR computations for conifers (Carter and Smith
1985, Oker-Blom and Smolander 1988), TLA of the apple
trees was defined as one-sided leaf area. VegeSTAR software
computes PLA by processing 3D tree mock-ups (Sinoquet et
al. 1998). Leaf area lit by a light source of a given direction
(e.g., the sun direction) is the same as leaf area seen on a tree
photograph taken in the same direction with an orthographic
camera. Therefore, PLA can be computed by counting vegetation pixels in the image. Moreover, assigning a different false
color to each shoot in the tree makes it possible to compute
PLA at the intra-tree level, namely the shoot scale.
Values of PLA and STAR depend on the direction of incident light. For directional integration over the sky hemisphere,
the sky was discretized into 46 solid angle sectors of equal area
according to the Turtle sky proposed by den Dulk (1989). Directional PLA and STAR values were computed for the central
direction of each solid angle sector. Directional values were
summed over the sky hemisphere using weighting coefficients
derived from the Standard OverCast distribution of sky radiance (Moon and Spencer 1942).
TREE PHYSIOLOGY ONLINE at http://heronpublishing.com
334
STEPHAN, SINOQUET, DONÈS, HADDAD, TALHOUK AND LAURI
Figure 1.Virtual images of representative
3D digitized trees of three cultivars of apple (Malus domestica) subjected to two
training systems. False colors were assigned to shoot types: red = bourses; blue
= bourse shoots; and green = vegetative
shoots. Crown volume was approximated
by a set of voxels. Voxel size was 0.2 m
and shows the scale of the images. Images
were synthesized with VegeSTAR software. Gray bars indicate tree trunks.
We computed PLA and STAR for the actual trees, their
crown envelopes and the corresponding homogeneous canopies. On the one hand, PLA of the crown envelope, i.e., the
projected envelope area (PEA, m2 ), was computed with VegeSTAR by replacing the leaves by the set of voxels making the
crown volume. On the other hand, the homogeneous canopy
was built by randomly locating leaves in the crown volume defined by the set of vegetated voxels. The PEA of the actual tree
crown and PLA of both actual and homogeneous (subscript H)
tree canopies were used to compute crown porosity P0 and
P0,H, respectively:
P0 = 1 −
PLA H
PLA
and P0 , H = 1 −
PEA
PEA
(2)
Foliage clumping (µ) was derived from porosity terms according to Nilson (1971):
P0 = P0µ, H hence µ =
lnP0
lnP0, H
the crown and less than 1 if foliage shows clumping.
Finally the leaf area index of the tree was defined as LAI =
TLA/PEA (Sinoquet et al. 2007).
Statistical analyses
The following variables were subjected to analysis of variance
(ANOVA): (1) canopy structure parameters: LAD, TLA at entire tree and shoot (FS, VS) scales; and (2) light interception
parameters: PLA, STAR at the shoot scale and PEA, LAI and µ
at the tree scale. A Duncan multiple mean comparison test was
performed when the cultivar effect was significant. The comparisons between training systems (C versus L) and between
shoot types (VS versus FS) were performed with the independent t-test. A 95% degree of confidence (P < 0.05) was applied to all tests. The STAR distributions of FS shoots between
training systems or cultivars, or both, were compared by the
non parametric χ2 test.
Results
(3)
Parameter µ is unity in the case of random leaf dispersion in
Tree-scale canopy structure
A qualitative analysis of the spatial distribution of LAD of total tree foliage showed that canopy volume depended on cul-
TREE PHYSIOLOGY VOLUME 28, 2008
MULTISCALE LIGHT INTERCEPTION IN 3D APPLE TREES
tivar (Figure 2), with the smallest and largest volumes for
‘Scarletspur’ and ‘Granny’, respectively. The three cultivars
showed similar gradients in LAD, with higher LAD in the central part of the tree and lower values at the crown periphery.
The C-training reduced LAD only in ‘Granny,’but had no clear
effect in the other cultivars. Crown volume of ‘Granny’ trees
was larger in C-trained trees than in L-trained trees. For all
cultivars, C-training, which resulted in a light well in the tree
canopy, reduced LAD around the trunk, and this effect was
striking for ‘Granny’.
One-way ANOVAs showed that structural parameters differed significantly between ‘Scarletspur’ and the other cultivars (Table 1). ‘Scarletspur’ TLA was about half that of
‘Golden’ and ‘Granny’ (6.8 versus 13.5 m2 ). ‘Scarletspur’
crown volume was about 1/3 that of the other cultivars (1.5
versus 4.6 m3 ). Consequently LAD was higher in ‘Scarletspur’
compared with the other cultivars. ‘Scarletspur’ also had a
higher ξ.
335
Significant changes in structural parameters occurred between 2004 and 2005, especially in TLA, LAD and ξ, where
values were higher in 2005. One-way ANOVA showed no effect of training system on canopy structure at the tree scale
(Table 1).
Tree-scale light interception
There were no significant differences in STAR between cultivars, with values around 0.30 (Table 1). In contrast, STAR
decreased significantly from 0.32 to 0.28 between 2004 and
2005. Training had no significant effect on STAR. Values of
PLA differed among cultivars, with lower PLA in ‘Scarletspur’ and similar values in ‘Golden’ and ‘Granny’. Although
PLA was slightly higher in 2005 and for C-trained trees, tree
age and training system had no significant effect on PLA (Table 1).
Determinants of light interception differed significantly
among cultivars. The only exception was LAI, where cultivar
differences were small and nonsignificant (Table 1). ‘Scarletspur’ PEA was less than half that of ‘Golden’ and ‘Granny’.
Similarly, µ was significantly lower in ‘Scarletspur’ than in the
other cultivars, indicating clumpier foliage. Temporal changes
between 2004 and 2005 were significant only for LAI and µ,
with slightly more LAI and clumpiness in 2005. There was a
trend for larger PEA in C-trained trees than in L-trained trees
(Table 1).
Intra-tree-scale canopy structure
The spatial distribution of LAD on both FS and VS is shown in
Figures 3 and 4, respectively. In ‘Scarletspur’, FS leaf area was
distributed in a smaller and lower canopy space than VS leaf
area, whereas FS and VS of ‘Golden’ and ‘Granny’ trees occupied about the same space. For all cultivars, C-training markedly reduced the number of zones with a high density of FS foliage, especially for ‘Scarletspur’ (Figure 3). In C-trained
trees, zones of higher LAD of FS were further from the tree
trunk than in L-trained trees (Figure 3). In ‘Scarletspur’, LAD
of VS showed high density in both training systems (Figure 4).
The C-trained trees of ‘Golden’ showed zones with high LAD,
but they were further from the trunk. C-Training did not induce clear changes in LAD distribution in ‘Granny’ trees.
We found significant differences in TLA between VS and
FS in ‘Scarletspur’ and ‘Golden’, with about double the leaf
area for VS compared with FS (Table 2). In contrast, ‘Granny’
FS and VS had roughly equal leaf areas. The TLA of FS was
also significantly lower than that of VS in 2004 and in Ltrained trees, whereas differences were not significant in 2005
or in C-trained trees (Table 2).
Intra-tree-scale light interception
Figure 2. Leaf area density distribution as a function of height above
the soil surface and distance from the tree trunk for three apple (Malus
domestica) cultivars subjected to two training systems, at the wholetree scale.
Values of STAR for FS and VS foliage showed no significant
differences among cultivars or between training systems (Table 2). Only STAR of VS showed a value significantly higher
in 2005 than in 2004. In contrast, STAR values were always
significantly higher for VS than for FS, with the difference
ranging from 0.04 in ‘Granny’ to 0.12 in ‘Golden’. Changes in
PLA at the shoot-type level were similar to changes in TLA
TREE PHYSIOLOGY ONLINE at http://heronpublishing.com
336
STEPHAN, SINOQUET, DONÈS, HADDAD, TALHOUK AND LAURI
Table 1. Cultivar, year and training system (TS; C = Centrifugal Training, L = Central Leader) effects on total leaf area (TLA, m2 ), silhouette to total area ratio (STAR), projected leaf area (PLA, m2 ), canopy volume (V, m3 ), leaf area density (LAD, m2 m – 3 ), relative variance of leaf area density (ξ, m2 m – 3 ), leaf area index (LAI), projected envelope area (PEA, m2) and foliage clumping (µ) at tree scale (n = number of trees) in apple
(Malus domestica) trees. Within a column, values followed by different letters differ significantly based on the Duncan multiple mean comparison
test for cultivars and the Student t-test for years and training systems (P < 0.05).
Cultivar
August 2004
‘Scarletspur’
‘Golden’
‘Granny’
August 2005
‘Scarletspur’
‘Golden’
‘Granny’
STAR
PLA
V
LAD
ξ
LAI
4.48
5.41
13.93
11.45
12.52
8.34
0.34
0.29
0.31
0.31
0.33
0.35
12.99
11.47
30.03
25.49
42.88
25.33
1.38
1.34
4.51
3.69
4.65
3.19
3.14
4.06
3.06
3.14
2.71
2.61
2.84
3.27
2.53
2.21
3.29
2.14
1.19
1.54
1.76
1.71
1.35
1.29
3.61
3.40
7.69
6.55
9.11
6.37
0.60
0.56
0.68
0.68
0.64
0.65
7.57
9.70
16.11
16.42
21.51
11.71
0.27
0.22
0.30
0.27
0.30
0.32
16.72
14.05
40.88
33.28
69.12
30.53
1.69
1.71
4.94
4.77
6.71
3.73
4.33
5.62
3.26
3.47
3.20
3.02
4.53
5.27
3.04
2.96
3.76
3.05
1.66
2.26
1.76
2.03
1.54
1.49
4.47
4.09
9.16
7.89
13.84
7.49
0.53
0.49
0.64
0.64
0.61
0.63
TS
n
TLA
C
L
C
L
C
L
3
3
3
3
3
3
C
L
C
L
C
L
3
3
2
3
3
3
PEA
µ
Mean per cultivar
‘Scarletspur’
‘Golden’
‘Granny’
12
10
12
6.79 b
14.33 a
13.52 a
0.28 b
0.30 ab
0.32 a
13.81 b
31.90 a
42.00 a
1.53 b
4.43 a
4.57 a
4.29 a
3.23 b
2.89 b
3.97 a
2.65 b
3.06 b
1.66 ab
1.82 a
1.42 b
3.90 b
7.70 a
9.20 a
0.54 c
0.66 a
0.63 b
Mean per year
2004
2005
18
16
9.36 b
13.70 a
0.32 a
0.28 b
24.85
33.70
3.13
3.87
3.12 b
3.85 a
2.71 b
3.81 a
1.47 b
1.79 a
6.13
7.74
0.63 a
0.58 b
Mean per TS
C-trained trees
L-trained trees
17
18
12.49
10.51
0.31
0.29
35.28
23.36
3.93
3.07
3.29
3.65
3.35
3.15
1.53
1.72
7.91
5.97
0.61
0.61
(Table 2): PLA values of both FS in ‘Granny’ and VS in
‘Golden’ were significantly higher than in the other cultivars.
The VS foliage generally had higher PLA than the FS foliage,
especially in the ‘Scarletspur’ and ‘Golden’ cultivars, and also
in L-trained trees. In contrast, ‘Granny’ and C-trained trees
showed no significant difference in PLA between FS and VS
foliage (Table 2).
Table 3 shows light interception by LVS at the shoot scale,
where terminal (T-LVS) and reiterated (R-LVS) shoots were
distinguished. As ‘Scarletspur’ trees mainly displayed SVS
with few LVS, they were excluded from the analysis. The
R-LVS were much less frequent on C-trained trees than on
L-trained trees (Table 3). This was directly related to training
system. However, the growth patterns of T-LVS and R-LVS
within L-trained trees differed depending on cultivar (Table 3). For ‘Golden’, R-LVS behaved similarly to T-LVS, with
higher LA, STAR and PLA values on L-trained trees than on
C-trained trees. For ‘Granny’, the only significant difference
observed was for STAR, with higher values in L-trained trees
than in C-trained trees.
Leaf irradiance distribution between individual FS is shown
in Figure 5 for years 2004 and 2005. For all cultivars except
‘Golden’ in 2004, C-training significantly improved STAR
distribution (χ2 test, P < 0.01), with fewer shaded and more
sunny shoots. For both C- and L-trained trees, shoot irradiance
distribution differed significantly among cultivars, with improved distribution increasing from ‘Scarletspur’ to ‘Golden’
and then ‘Granny’ (P < 0.01, data not shown).
Discussion
Advantages and limitations of virtual experiments
The combination of making 3D tree mock-ups with light computations based on projection methods allowed detailed study
of canopy structure and light interception attributes. First,
structural parameters at the tree scale were computed from the
databases where canopy geometry is described at the leaf
scale. These computations can be considered actual measurements of the parameters: TLA estimation was similar to the
detailed shoot sampling method proposed by Wünsche and
Palmer (1997); and crown volume was derived from accurate
localization of shoots in space. Additionally, we were able to
describe the spatial distribution of LAD in the tree crowns
(Figure 2). Only a few measurements based on 2D or 3D leaf
clipping have been reported in fruit trees (Cohen and Fuchs
1987, Palmer et al. 1992) and vineyards (Mabrouk et al. 1997).
They are usually destructive and time-consuming. With the 3D
TREE PHYSIOLOGY VOLUME 28, 2008
MULTISCALE LIGHT INTERCEPTION IN 3D APPLE TREES
337
Figure 3. Leaf area density (m2 m –3 ) distribution as a function of
height above the soil surface and distance from tree trunk for three apple (Malus domestica) cultivars subjected to two training systems, for
fruiting shoots.
Figure 4. Leaf area density (m2 m –3 ) distribution as a function of
height above the soil surface and distance from tree trunk for three apple (Malus domestica) cultivars subjected to two training systems, for
vegetative shoots.
digitizing method, we established the spatial distribution of
leaf area for FS and VS, separately (Figures 3 and 4). Second,
virtual experiments allowed us to compute light interception
properties and their determinants at several scales: STAR and
PLA of the whole tree (Table 1), per shoot type (FS and VS)
(Table 2) and at the individual shoot scale (Figure 5). Several
methods have been proposed to measure total light interception at the tree scale, e.g., arrays of sensors (Oker-Blom et al.
1991, Giuliani et al. 2000), fisheye photographs below the canopy (Wünsche et al. 1995) and photographs of the shadow cast
by the crown (Hemmerlein and Smith 1994). At the shoot
scale, video projection and photography methods have been
proposed for detached shoots (Oker-Blom et al. 1991, Niinemets et al. 2004). Field methods for estimating light interception at the shoot scale within crowns are time consuming.
Light micro-sensors can be attached to the leaf surface (Gutschick et al. 1985), but this method needs many sensors to get
unbiased values of mean irradiance (Sinoquet et al. 2001).
Wünsche et al. (1996) proposed a scanner laser method to
measure light partitioning between spurs and extension shoots,
which are closely analogous to FS and VS, respectively. Laser
beams were sent to the canopy, and the target organ was visually recorded as a spur or extension shoot leaf. The plant image
processing method we used can be regarded as an extension of
Wünsche’s work, where each pixel in the image corresponds
to a laser beam and target classification is automated by assigning false colors to each shoot type (Figure 1) or individual
shoot. The virtual computation allowed a large number of
sampled beams (the image size used in VegeSTAR is 356 ×
356 pixels, and one image is used for each of the 46 light directions used to abstract the sky radiation) and fine light partitioning among plant components. Virtual experiments also allowed us to compute parameters that are difficult to measure in
the field. This is especially the case for µ, which expresses departure from random distribution of leaves in the canopy space
(Nilson 1971). Cohen et al. (1995) inferred clumping proper-
TREE PHYSIOLOGY ONLINE at http://heronpublishing.com
338
STEPHAN, SINOQUET, DONÈS, HADDAD, TALHOUK AND LAURI
Table 2. Effects of cultivar, year and training system (TS; C = Centrifugal Training, L = Central Leader) on total leaf area (TLA, m2 ), light interception (STAR) and projected leaf area (PLA, m2 ) for fruiting (FS) and vegetative shoots (VS) (n = number of trees) in apple (Malus domestica)
trees. For each variable and each FS–VS pair, values in bold within rows indicate significant differences based on the Dependent t-test (P < 0.05).
Within a column, values followed by different letters differ significantly based on the Duncan multiple mean comparison test for cultivars and the
Student t-test for years and training systems (P < 0.05).
FS
Cultivar
TS
August 2004
‘Scarletspur’
‘Golden’
‘Granny’
August 2005
‘Scarletspur’
‘Golden’
‘Granny’
n
VS
TLA
STAR
PLA
TLA
STAR
PLA
C
L
C
L
C
L
3
3
3
3
3
3
2.34
1.06
2.80
3.32
6.33
2.72
0.27
0.27
0.26
0.24
0.30
0.29
0.69
0.28
0.75
0.79
1.91
0.77
3.50
4.46
11.22
8.21
7.09
6.22
0.36
0.33
0.31
0.31
0.31
0.30
1.24
1.40
3.21
2.44
2.22
1.89
C
L
C
L
C
L
3
3
2
2
3
3
2.08
2.70
5.28
5.06
12.86
5.28
0.33
0.22
0.25
0.22
0.30
0.28
0.67
0.58
1.30
1.09
3.81
1.40
4.25
5.40
10.83
7.20
5.95
4.96
0.38
0.36
0.51
0.35
0.34
0.37
1.54
1.78
5.09
2.41
2.01
1.65
Mean per cultivar
‘Scarletspur’
‘Golden’
‘Granny’
12
10
12
2.05 b
3.91 ab
6.80 a
0.27 a
0.24 a
0.29 a
0.55 b
0.94 b
1.97 a
4.41 b
9.44 a
6.06 b
0.36 a
0.36 a
0.33 a
1.49 b
3.19 a
1.94 b
Mean per year
2004
2005
18
16
3.10 a
5.60 a
0.27 a
0.27 a
0.86 a
1.51 a
6.79 a
6.11 a
0.32 a
0.38 b
2.07 a
2.25 a
Mean per TS
C-trained trees
L-trained trees
17
17
5.28 a
3.26 a
0.29 a
0.25 a
1.53 a
0.80 a
6.93 a
6.01 a
0.36 a
0.34 a
2.40 a
1.90 a
Table 3. Leaf area (LA, dm2 ), light interception (STAR) and projected leaf area (PLA, dm2 ) of long vegetative shoots (LVS) issued from terminal
(T-LVS) and reiterated (R-LVS) growth in 2005 as influenced by training system (TS; C = Centrifugal Training, L = Central Leader) and pruning
(T-LVS, R-LVS) for L-trees only, for ‘Golden’ and ‘Granny’ (n = number of shoots). Within a column, values followed by different lowercase letters differ significantly between training systems (Independent t-test, P < 0.05). Within a row and for each variable, values followed by different
uppercase letters differ significantly between T-LVS and R-LVS for L-trained trees (Independent t-test, P < 0.05).
Cultivar
‘Golden’
‘Granny’
TS
C
L
C
L
T-LVS
R-LVS
n
LA
STAR
PLA
n
LA
STAR
500
239
349
34
2.10 b
3.06 a
3.12 b
3.51 a
0.22 b
0.52 aA
0.29 b
0.63 aA
0.51 b
1.62 aA
0.94 b
2.29 a
24
264
28
232
2.30 b
3.09 a
4.45
3.66
0.25 b
0.42 aB
0.65 a
0.50 bB
Training effect (C-trained versus L-trained)
‘Golden’
F
1.82
P
< 0.01
‘Granny’
F
3.31
P
< 0.01
1.67
< 0.01
7.4
< 0.01
56.72
< 0.01
21.67
< 0.01
Pruning effect (T-LVS versus R-LVS) in L-trained trees
‘Golden’
F
0.67
2.72
P
0.70
< 0.01
‘Granny’
F
0.04
6.18
P
0.56
< 0.01
1.98
0.02
0.04
0.06
TREE PHYSIOLOGY VOLUME 28, 2008
21.21
< 0.01
57.31
0.18
2.30
< 0.01
14.07
< 0.01
PLA
0.58 b
1.38 aB
2.95
1.93
21.44
< 0.01
0.10
0.47
MULTISCALE LIGHT INTERCEPTION IN 3D APPLE TREES
339
Figure 5. Effect of the training
system, Centrifugal (open bars)
versus Central Leader (filled
bars), on the STAR distribution of
fruiting shoots at harvest 2004
and 2005 of three apple (Malus
domestica) cultivars: (A, B)
‘Scarletspur’, (C, D) ‘Golden’ and
(E, F) ‘Granny’. For each cultivar,
STAR distribution between training systems was compared by a χ2
non-parametric test.
ties of apple trees by simultaneously measuring leaf area and
light distribution within the tree canopy. Here we constructed a
virtual homogeneous tree canopy where leaves were randomly
distributed in the crown volume, and we derived µ by comparing the crown porosity of both actual and homogeneous tree
crowns (Equation 4). Such computations could not be made in
the field, mainly because homogeneous trees do not exist in
nature.
Unfortunately, virtual experiments also have limitations.
The main one is related to the construction of 3D plants. We
used a 3D digitizing technique (Sinoquet and Rivet 1997) that
allowed us to describe the 3D structure of the trees. However,
tree digitizing in the field is time consuming. The rate of 3D
data acquisition is about 1000–2000 points per day, if the tree
is not too big. For tall trees, tree digitizing is impossible. We
used digitizing at the leafy shoot scale associated with a reconstruction method of leaves attached to the shoot (Sonohat et al.
2006). This reduced data acquisition time, and the method has
been shown to be accurate for building 3D tree structures suitable for light interception computations (Sonohat et al. 2006).
Digitizing at the shoot scale allowed measurement of the 3D
structure of 18 trees in 2004 and 2005, which in comparison
with other studies with virtual plants, is a large number (cf.
Thanisawanyangkura et al. 1997, Willaume et al. 2004). However, the sample size was insufficient to show statistically significant combined effects of cultivar, training system and development over two years at the tree scale (Table 1). The small
number of individuals allowed only one-way ANOVAs, where
the separate effects of cultivar, training and year were analyzed. At the shoot scale, however, the number of individuals,
TREE PHYSIOLOGY ONLINE at http://heronpublishing.com
340
STEPHAN, SINOQUET, DONÈS, HADDAD, TALHOUK AND LAURI
i.e., shoots, was sufficient and analysis showed significant differences among cultivars and between training systems (Table 3 and Figure 5).
Cultivar and training effects on canopy structure and light
interception
At the tree scale, differences in structural parameters and interception properties can be related to the typology of the cultivars: Type II for ‘Scarletspur’, Type III for ‘Golden’and Type
IV for ‘Granny’ (Lespinasse 1992). Type II cultivars show
more short shoots and fewer long shoots, as shown by Stephan
et al. (2007) for ‘Scarletspur’. This leads to smaller canopy
volume, leaf area and total light interception (PLA). Moreover, the display of foliage mainly in short shoots may explain
the high foliage clumping (µ and ξ) in ‘Scarletspur’. Small
internode length, which is a characteristic of dwarf or compact
scion cultivars (Rusholme et al. 2004), has been shown to increase foliage clumping (Takenaka 1994). Leaf dispersion values were not in complete agreement with those reported by
Cohen et al. (1995): ‘Smoothee’ trees (Type III) were clumped,
whereas ‘Top Red’ trees (Type II) showed random leaf dispersion. Leaf irradiance (STAR) at the tree scale did not differ
significantly between cultivars, although values tended to be
lower for ‘Scarletspur’ and higher for ‘Granny’. The equations
relating LAI to PLA and STAR have been established and discussed previously (Sinoquet et al. 2007).
Partitioning leaf area between FS and VS revealed differences among cultivars, with much more VS leaf area compared with FS leaf area in ‘Scarletspur’ and ‘Golden’, and similar leaf area for the two shoot types in ‘Granny’ (Table 2). Although STAR values were higher in VS than in FS for the three
cultivars, the sunlit leaf area (PLA) was similar in both shoot
types for ‘Granny’ whereas it was significantly higher in VS
than in FS for the other cultivars. Because fruit yield is related
to light interception by spurs (Wünsche et al. 1996, Lakso et al.
1999), the greater sunlit leaf area in FS of ‘Granny’ likely explains the higher regularity of bearing of this cultivar (Lauri et
al. 1997, Stephan et al. 2007).
We found no significant effect of training system at the
whole-tree scale for any leaf canopy or light interception parameter or for any combination of them. However, an effect of
training was observed at the within-canopy scale. First, STAR
values of individual shoots, whether VS or FS, increased more
in C-trained trees than in L-trained trees (Figure 5). This is
likely related to the lower LAD values in C-trained trees compared with L-trained trees (Figures 2–4). Second, L-trained
trees consistently had increased VS leaf area (total and sunlit)
compared with FS leaf area, whereas leaf areas of both shoot
types were similar for C-trained trees (Table 2). From a physiological viewpoint, the greater sunlit leaf area proximal to the
fruit is known to be an important positive factor for yield
(Wünsche and Lakso 2000). We showed that the training system may significantly change the balance between the leaf areas of each shoot type (TLA) as well as their spatial distribution and access to light (PLA). However, comparison with
other studies is difficult, because few data obtained at the
individual shoot scale have been published.
Another effect of the training system was observed on the
type of long vegetative shoots that developed either from an
existing shoot (T-LVS) or from a latent bud (R-LVS). A previous analysis of the same experiment showed that pruning of
branches included in the Central Leader system led to strong
reiterative phenomena (Stephan et al. 2007). For ‘Golden’ and
‘Granny’, T-LVS of L-trained trees had significantly higher
LA, STAR and PLA than T-LVS of C-trained trees, suggesting
that at the tree scale, pruning trunk and scaffold branches stimulated growth (LA) and light interception (STAR, PLA) of individual vegetative shoots (Table 3A). We have now shown
that these shoots (R-LVS) had similar LA and significantly
lower STAR values than T-LVS on both cultivars. Although
PLA was lower in R-LVS compared with T-LVS, the difference was significant only for ‘Golden’. This could be explained by the location of R-LVS on scaffold branches below
the pruning cuts, i.e., within the canopy where there was less
light penetration. Therefore R-LVS were not expected to contribute strongly to the carbon budget of the tree although they
bear much of the leaf area of long vegetative shoots, i.e., 53%
and 88% for L-trained trees of ‘Golden’ and ‘Granny Smith’
respectively (Table 3).
The interest in adapting training to tree architecture has often been emphasized (Lauri and Laurens 2005, Costes et al.
2006, Robinson 2007). In particular, the agronomic interest in
minimal pruning strategies adapted to the cultivar (Robinson
2007) as opposed to pruning strategies based only on heading
or shortening cuts, i.e., reducing the length of current-season
or older wood of scaffold branches, has been demonstrated.
We have now shown that manipulations of tree architecture
through the training system and pruning strategies also affect
light interception. For example, we found that Centrifugal
Training enhanced light interception by fruiting shoots and reduced the amount of less fruit-feeding vegetative shoots (RLVS). These results serve to guide a further study on the influence of plant structure on photosynthesis and transpiration at
the fruiting branch and tree scales (Massonnet et al. 2008).
Our findings contribute to the overall physiological interpretation of the positive effects of minimal pruning on fruit growth
and quality, and regularity of yield across consecutive years
(Lauri et al. 2007).
Acknowledgments
We thank the American University of Beirut, and especially Dr.
Mohamed Farran, for orchard monitoring and availability, and the students Dahlia Mansour, Gebran Nassif, Marica Abi Nader and Majed
Feghali for tree digitizing.
References
Abbott, D.L. 1984. The apple tree: physiology and management.
Grower Books, London, 90 p.
Awad, M.A., P. Wagenmakers and A. de Jager. 2001. Effects of light
on flavonoid and chlorogenic acid levels in the skin of ‘Jonagold’
apples. Sci. Hortic. 88:289–298.
Carter, G. and W. Smith. 1985. Influence of shoot structure on light
interception and photosynthesis in conifers. Plant Physiol. 79:
1038–1043.
TREE PHYSIOLOGY VOLUME 28, 2008
MULTISCALE LIGHT INTERCEPTION IN 3D APPLE TREES
Charles-Edwards, D.A. and M.R. Thorpe. 1976. Interception of diffuse and direct-beam radiation by a hedgerow apple orchard. Ann.
Bot. 40:603–613.
Cohen, S. and M. Fuchs. 1987. The distribution of leaf area, radiation,
photosynthesis and transpiration in a Shamouti orange hedgerow
orchard. I. Leaf area and radiation. Agric. For. Meteorol. 40:
123–144.
Cohen, S., P. Mason and M. Meron. 1995. Canopy clumpiness and radiation penetration in a young hedgerow apple orchard. Agric. For.
Meteorol. 76:185–200.
Costes, E., P.É. Lauri and J.L. Regnard. 2006. Analysing fruit tree architecture: implications for tree management and fruit production.
Hortic. Rev. 32:1–61.
den Dulk, J.A. 1989. The interpretation of remote sensing, a feasibility study. Ph.D. Thesis, Wageningen University, The Netherlands,
189 p.
Doud, D.S. and D.C. Ferree. 1980. Influence of altered light levels on
growth and fruiting of mature ‘Delicious’ apple trees. J. Am. Soc.
Hortic. Sci. 105:325–328.
Elfving, D.C. and C.G. Forshey. 1976. Growth and fruiting responses
of vigorous apple branches to pruning and branch orientation treatments. J. Am. Soc. Hortic. Sci. 101:290–293.
Elfving, D.C., I. Schecter, R.Q. Cline and W.F. Pierce. 1990.
Palmette-Leader and Central-Leader tree forms compared for light
distribution, productivity and fruit quality of ‘McIntosh’ apple
trees. HortScience 25:1386–1388.
Ferree, D., K. Clayton-Greene and B. Bishop. 1992. Influence of orchard management system on canopy light distribution and net
photosynthesis of apple trees. J. Hortic. Sci. 68:377–392.
Giuliani, R., E. Magnanini, C. Fragassa and F. Nerozzi. 2000. Ground
monitoring the light-shadow windows of a tree canopy to yield canopy light interception and morphological traits. Plant Cell Environ.
23:783–796.
Green, S., K. McNaughton, J.N. Wünsche and B. Clothier. 2003.
Modeling light interception and transpiration of apple tree canopies. Agron. J. 95:1380–1387.
Gutschick, V.P., M.H. Barron, D.A. Waechter and M.A. Wolf. 1985.
Portable monitor for solar radiation that accumulates irradiance
histograms for 32 leaf-mounted sensors. Agric. For. Meteorol.
33:281–290.
Hampson, C.R., H.A. Quamme and R.T. Brownlee. 2002. Canopy
growth, yield and fruit quality of ‘Royal Gala’apple trees grown for
eight years in five tree training systems. HortScience 37:627–631.
Hampson, C.R., H.A. Quamme, F. Kappel and R.T. Brownlee. 2004.
Varying density with constant rectangularity: II. Effects on apple
tree yield, fruit size and fruit color development in three training
systems over ten years. HortScience 39:507–511.
Heinicke, D.R. 1975. High-density apple orchards: planting, training
and pruning. US Department of Agriculture Handbook. 458:1–34.
Hemmerlein, M.T. and W.K. Smith. 1994. Structural scaling of light
interception efficiency in Picea engelmannii and Abies lasiocarpa.
Tree Physiol. 14:1139–1148.
Jackson, J.E. 1980. Light interception and utilization by orchard systems. Hortic. Rev. 2:208–267.
Jackson, J.E. and J.W. Palmer. 1972. Interception of light by model
hedgerow orchards in relation to latitude, time of year and hedgerow configuration and orientation. J. Appl. Ecol. 9:341–357.
Jackson, J.E. and J.W. Palmer. 1981. Light distribution in discontinuous canopies: calculation of leaf areas and canopy volumes above
defined ‘irradiance contours’ for use in productivity modeling.
Ann. Bot. 47:561–565.
Johnson, R.S. and A.N. Lakso. 1986. Carbon balance model of a
growing apple shoot. II. Simulated effects of light and temperature
on long and short shoots. J. Am. Soc. Hortic. Sci. 111:164–169.
341
Lakso, A.N. 1980. Correlations of fisheye photography to canopy
structure, light climate and biological response to light in apple
trees. J. Am. Soc. Hortic. Sci. 105:43–46.
Lakso, A.N. and L. Corelli- Grappadelli. 1992. Implications of pruning and training practices to carbon partitioning and fruit development in apple. Acta Hortic. 322:231–239.
Lakso, A.N., J.N. Wünsche, J.W. Palmer and L. Corelli-Grappadelli.
1999. Measurement and modeling of carbon balance of the apple
tree. HortScience 34:1040–1047.
Lauri, P.É. and J.J. Kelner. 2001. Shoot type demography and dry
matter partitioning: a morphometric approach in apple (Malus ×
domestica). Can. J. Bot. 79:1270–1273.
Lauri, P.É. and F. Laurens. 2005. Architectural types in apple (Malus
× domestica Borkh.). In Crops: Growth, Quality and Biotechnology. Ed. R. Dris. World Food Limited, Helsinki, Finland,
pp 1300–1314.
Lauri, P.É. and C. Trottier. 2004. Patterns of size and fate relationships
of contiguous organs in the apple (Malus × domestica) crown. New
Phytol. 163:533–546.
Lauri, P.É., É. Térouanne and J.M. Lespinasse. 1997. Relationship between the early development of apple fruiting branches and the regularity of bearing—an approach to the strategies of various
cultivars. J. Hortic. Sci. 72:519–530.
Lauri, P.É., X. Crété and G. Ferré. 2007. Centrifugal training in
apple—appraisal of a two-year experiment on cv. ‘Galaxy’ in
southeast France. Acta Hortic. 732:391–396.
Lespinasse, Y. 1992. Breeding apple tree: aims and methods. In Proc.
Joint Conference of the E.A.P.R. Breeding and Varietal Assessment Section and the E.U.C.A.R.P.I.A Potato Section. Eds. F. Rousselle-Bourgeois and P. Rousselle. INRA, Landerneau, France and
Ploudaniel, France, pp 103–110.
Mabrouk, H., H. Sinoquet and A. Carbonneau. 1997. Canopy structure and radiation regime in grapevine. II. Modeling radiation interception and distribution inside the canopy. Vitis 36:125–132.
Massonnet, C., J.L. Regnard, P.É. Lauri, E. Costes and H. Sinoquet.
2008. Unraveling the relative contributions of foliage distribution
and leaf functions to branch capacities in two apple cultivars. Tree
Physiol. In press.
Moon, P. and D.E. Spencer. 1942. Illumination from a non-uniform
sky. Trans. Illumin. Engineer. Soc. 37:707–726.
Niinemets, Ü., A. Cescatti and R. Christian. 2004. Constraints on
light interception efficiency due to shoot architecture in broadleaved Nothofagus species. Tree Physiol. 24:617–630.
Nilson, T. 1971. A theoretical analysis of the frequency of gaps in
plant stands. Agric. Meteorol. 8:25–38.
Oker-Blom, P. and H. Smolander 1988. The ratio of shoot silhouette
area to total needle area in Scots pine. For. Sci. 34:894–906.
Oker-Blom, P., M.R. Kaufmann and M.G. Ryan. 1991. Performance
of a canopy light interception model for conifer shoots, trees and
stands. Tree Physiol. 9:227–243.
Oyarzun, R.A., C.O. Stöckle and M.D. Whiting. 2007. A simple approach to modeling radiation interception by fruit-tree orchards.
Agric. For. Meteorol. 142:12–24.
Palmer, J.W. 1977. Diurnal light interception and a computer model of
light interception by hedgerow apple orchards. J. App. Ecol. 14:
601–614.
Palmer, J.W. 1987. The measurement of leaf area in apple trees. J.
Hortic. Sci. 62:5–10.
Palmer, J.W. 1989. Canopy manipulation for optimum utilization of
light. In Manipulation of Fruiting. Ed. C.J. Wright. Butterworths,
London, pp 245–262.
Palmer, J.W., D.J. Avery and S.J. Wertheim. 1992. Effect of apple tree
spacing and summer pruning on leaf area distribution and light interception. Sci. Hortic. 52:303–312.
TREE PHYSIOLOGY ONLINE at http://heronpublishing.com
342
STEPHAN, SINOQUET, DONÈS, HADDAD, TALHOUK AND LAURI
Prusinkiewicz, P. and A. Lindenmayer. 1990. The algorithmic beauty
of plants. Springer-Verlag, New York, 228 p.
Robinson, T.L. 2007. Recent advances and future directions in orchard planting systems. Acta Hortic. 732:367–381.
Robinson, T.L. and A.N. Lakso. 1991. Bases of yield and production
efficiency in apple orchard systems. J. Am. Soc. Hortic. Sci. 116:
188–194.
Robinson, T.L., E.J. Seeley and B.H. Barritt. 1983. Effect of light environment and spur age on ‘Delicious’ apple fruit size and quality.
J. Am. Soc. Hortic. Sci. 108:855–861.
Robinson, T.L., A.N. Lakso and Z. Ren. 1991. Modifying apple tree
canopies for improved production efficiency. HortScience 26:
1005–1012.
Rom, C.R. and B. Barritt. 1990. Spur development of ‘Delicious’ apple as influenced by position, wood age, strain and pruning.
HortScience 25:1578–1581.
Ross, J. 1981. The radiation regime and architecture of plant stands.
Dr. W. Junk, The Hague, 342 p.
Rusholme, R.L., S.E. Gardiner, H.C.M. Bassett, D.S. Tustin,
S.M. Ward and A. Didier. 2004. Identifying genetic markers for an
apple dwarfing gene. Acta Hortic. 663:405–409.
Sinoquet, H. and P. Rivet. 1997. Measurement and visualization of the
architecture of an adult tree based on a three-dimensional digitizing
device. Trees 11:265–270.
Sinoquet, H., S. Thanisawanyangkura, H. Mabrouk and P. Kasemsap.
1998. Characterization of the light environment in canopies using
3D digitizing and image processing. Ann. Bot. 82:203–212.
Sinoquet, H., X. Le Roux, B. Adam, T. Améglio and F.A. Daudet.
2001. RATP, a model for simulating the spatial distribution of radiation absorption, transpiration and photosynthesis within canopies:
application to an isolated tree crown. Plant Cell Environ. 24:
395–406.
Sinoquet, H., G. Sonohat, J. Phattaralerphong and C. Godin. 2005.
Foliage randomness and light interception in 3D digitized trees: an
analysis from multiscale discretisation of the canopy. Plant Cell
Environ. 28:1158–1170.
Sinoquet, H., J. Stephan, G. Sonohat, P.É. Lauri and P. Monney. 2007.
Simple equations to estimate light interception by isolated trees
from canopy structure features: assessment with 3D digitized apple
trees. New Phytol. 175:94–106.
Sonohat, G., H. Sinoquet, V. Kulandaivelu, D. Combes and F. Lescourret. 2006. Three dimensional reconstruction of partially 3Ddigitized peach tree canopies. Tree Physiol. 26:337–351.
Stephan, J., P.É. Lauri, N. Donès, N. Haddad, S. Talhouk and H. Sinoquet. 2007. Architecture of the pruned tree: impact of contrasted
pruning procedures over 2 years on shoot demography and spatial
distribution of leaf area in apple (Malus domestica). Ann. Bot.
99:1055–1065.
Takenaka, A. 1994. Effects of leaf blade narrowness and petiole
length on the light capture efficiency of a shoot. Ecol. Res. 9:
109–114.
Thanisawanyangkura, S., H. Sinoquet, P. Rivet, M. Crétenet and
E. Jallas. 1997. Leaf orientation and sunlit leaf area distribution in
cotton. Agric. For. Meteorol. 86:1–15.
Tustin, D.S., P.M. Hirst and I.J. Warrington. 1988. Influence of orientation and position of fruiting laterals on canopy light penetration,
yield and fruit quality of ‘Granny Smith’ apple. J. Hortic. Sci.
113:693–699.
Tustin, D.S., W.M. Cashmore and R.B. Bensley. 1998. The influence
of orchard row canopy discontinuity on irradiance and leaf area distribution in apple trees. J. Hortic. Sci. 73:289–297.
Volz, R.K., I.B. Ferguson, E.B. Hewett and E.W. Wooley. 1994. Wood
age and leaf area influence fruit size and mineral composition of
apple fruit. J. Hortic. Sci. 69:385–395.
Wagenmakers, P.S. 1991a. Planting systems for fruit trees in temperate climates. Crit. Rev. Plant Sci. 10:369–385.
Wagenmakers, P.S. 1991b. Simulation of light distribution in dense
orchard systems. Agric. For. Meteorol. 57:13–25.
Willaume, M., P.É. Lauri and H. Sinoquet. 2004. Light interception in
apple trees influenced by canopy architecture manipulation. Trees
18:705–713.
Wünsche, J.N. and A.N. Lakso. 2000. The relationship between leaf
area and light interception by spur and extension shoot leaves and
apple orchard productivity. HortScience 37:1202–1206.
Wünsche, J.N. and J.W. Palmer. 1997. Comparison of non-destructive
methods of estimating leaf area in apple tree canopies. Acta Hortic.
451:701–707.
Wünsche, J.N., A.N. Lakso and T.L. Robinson. 1995. Comparison of
four methods for estimating total light interception by apple trees of
varying forms. HortScience 30:272–276.
Wünsche, J.N., A.N. Lakso, T.L. Robinson, F. Lenz and S.S. Denning. 1996. The bases of productivity in apple production systems:
the role of light interception by different shoot types. J. Am. Soc.
Hortic. Sci. 121:886–893.
Wünsche, J.N., J.W. Palmer and D.H. Greer. 2000. Effects of crop
load on fruiting and gas-exchange characteristics of ‘Braeburn’/M26 apple trees at full canopy. J. Am. Soc. Hortic. Sci. 125:93–99.
TREE PHYSIOLOGY VOLUME 28, 2008