Tree Physiology 21, 1257–1267 © 2001 Heron Publishing—Victoria, Canada

Canopy structure and light interception in Quercus petraea seedlings in relation to light regime and plant density
1 2

UMR INRA-ENGREF Ressources forêt-bois, Equipe Croissance et Qualité des Bois, 54280 Champenoux, France UMR PIAF, INRA-Université Blaise Pascal, Site de Crouelle, 234 avenue du Brezet, 63039 Clermont-Ferrand Cedex 2, France

Received October 16, 2000
Downloaded from at University of Novi Sad on April 18, 2012

Summary Foliage structure was measured on 1- and 2year-old Quercus petraea (Matt.) Liebl. seedlings grown in 100 or 18% sunlight at a planting density of 2.8 or 25 plants per m 2. A three-dimensional digitizing device was used to acquire the spatial position and orientation of all leaves within the seedlings and of all seedlings within the plot. The data were used to obtain (1) quantitative information on canopy structure, including leaf area index (LAI), seedling leaf area, number of leaves, leaf area density and leaf orientation; and (2) structural information on foliage arrangement from virtual images to estimate light interception by individual seedlings (STAR) and light partitioning among seedlings. During the second year, shading significantly reduced total leaf area and number of leaves but increased individual leaf area. The STAR was greater for seedlings in shade than in full sunlight because of the more horizontal orientation of leaves. Leaf area density was unaffected by the full sun treatments, and changes in leaf area dispersion had no effect on light-interception efficiency. No plant density effect was observed during the first year. During the second year, only the high plant density treatment induced mutual shading between seedlings, resulting in greater competition for light among seedlings in the full sun treatment than in the shade treatment. The small treatment-induced changes in light interception indicate that Q. petraea has low morphological plasticity of foliage structure compared with other species. Keywords: 3-D digitizing, foliage structure, initial spacing, leaf arrangement, leaf area, morphological plasticity, shade, STAR.

Introduction Currently, foresters favor natural regeneration of oak stands over artificial regeneration because of the economic and ecological advantages. Natural regeneration of managed oak stands involves the gradual removal of shelter wood as the stand undergoes progressive regeneration. Removal of shelter wood over a number of years insures continued availability of seed and a degree of protection of seedlings from full sunlight, frost and waterlogging. It also allows continual harvesting over an extended period. However, these silvicultural opera-

tions result in highly heterogeneous light conditions at the forest floor. Improved control of the light microclimate, by manipulating shelter wood and seedling densities, may enhance seedling vigor, morphology and growth rates. European oaks require high solar irradiances, especially at the seedling stage (Barry-Langer and Nebout 1993). However, Quercus petraea (Matt.) Liebl. seedlings can grow at low irradiances and high densities (Barry-Langer and Nebout 1993), enabling them to survive for long periods in the understory or when overtopped by other species during stand regeneration. Quercus petraea responds vigorously to release from overstory competition. Keeping seedlings in shade conditions can prevent the occurrence of forks (Chaar et al. 1997, Collet et al. 1998), but it also results in reductions in seedling growth and productivity. Several studies have examined the effects of irradiance on carbon acquisition and biomass increment of oak seedlings (Jarvis 1964, Farmer 1975, Ziegenhagen and Kausch 1995). Changes in seedling morphology in response to changes in irradiance have also been analyzed (Axelsson et al. 1979, Gottschalk 1994, Welander and Ottosson 1998). However, the effects of competition between seedlings have been studied mainly in terms of carbon or water balance (Collet et al. 1996, Jinks and Mason 1998) but rarely in terms of plant architecture (Astre 1999). Consequently, there is little information about the effects of irradiance on canopy structure during the early stages of development when competition has pronounced effects on the social status and vigor of seedlings. Our objective was to analyze the effects of irradiance and plant density on the morphology of Q. petraea seedlings and their capacity to intercept light. Quercus petraea seedlings were grown at two planting densities in either 100 or 18% sunlight. We assessed treatment-induced modifications in foliage display and their consequences on light interception by 1- and 2-year-old seedlings. The light interception properties of individual seedlings and mutual shading within the canopy were quantified by the silhouette to total area ratio (STAR; OkerBlom and Smolander 1988), i.e., the ratio of projected leaf area to total leaf area. From three-dimensional (3-D) representations of foliage structure, the effects of leaf size, leaf orientation and spatial distribution of leaves within the plant were quantified to assess the efficiency of the foliage structure for light interception. Light partitioning between seedlings was

At the seedling scale. To minimize edge effects. (1998). The seedlings were watered when necessary. (1) At the leaf scale. Ltd.oxfordjournals. VT) as described by Polhemus (1993) and Sinoquet et al. Powys. The length of each leaf was image) to determine individual leaf area (A. the two outer rows were not measured. corresponding to the azimuth ψ of the midrib. y and x. and sown in the field during May 1998.. TREE PHYSIOLOGY VOLUME 21.1258 FARQUE. when they had completed their vegetative growth and before the onset of leaf abscission. Angle α was defined as the angle between the leaf normal and the vertical axis (de Wit 1965) and was calculated as (Polhemus 1993): α = arccos(cos θ cos φ). r 2 = 0. With a single digitized record per leaf. Figure 1. 2001 . Total height was measured after Year 1 and Year 2. The leaf normal azimuth angle ϕ was measured but is not presented in the figure. a set of three angles of rotation around axes z. Leaf orientation was characterized by Euler angles. the Netherlands). cm). Thus. the device allowed measurement of the spatial coordinates of the junction point between the blade and the petiole and leaf orientation.K. Foliage display was measured during October 1998 and 1999 on 1. altitude 73 m).). all leaves were digitized and a distinction was made between leaves attached to the trunk and leaves attached to the branches. The two light regimes were 100 (full sun) and 18% (shade) of external photosynthetically active radiation (PAR). n = 1465. a subplot of five neighboring seedlings was chosen for canopy structure  cos ψ sin θ cos φ − sin ψ sin φ  (2b) Because Equation 2b gives a value between –π/2 and π/2. Each treatment included one block of 80 seedlings in staggered rows. Cochester. cm2 ). supported by a 13 × 24 × 3-m aluminum infrastructure. Downloaded from http://treephys. 6°14′ E. Shading was provided by neutral polyester–aluminum nets (OLS. variables calculated from digitized data were leaf area. midrib elevation θ. Blois. total leaf area and leaf area index (LAI) were calculated by summing leaf area of all of the seedlings per plot. corresponding to full sunlight and to the average shade below a shelter wood after the first cutting. There were two light regimes and two planting densities in a nested design. Geometrical measurements of foliage structure The location and orientation of each leaf were measured with an electromagnetic 3-D digitizer (Fastrak. U.8 plants per m2 (low density. SINOQUET AND COLIN estimated as the effect of the surrounding vegetation on seedling light interception. individual leaf area was estimated from leaf length for all seedlings as: A = at University of Novi Sad on April 18. i. 1°19′ E. Each leaf was scanned and images were analyzed with NIH software (http://rsb. HD) and 2. LD). Materials and methods Experimental design Effects of light availability and plant density on oak seedling morphology and canopy structure were observed in a semicontrolled field experiment located close to the INRA Research Centre at Nancy.e. and leaf length (H.284 H 2. total leaf area and number of leaves were calculated by summing individual leaf areas. altitude 237 m).. Ludvig Svensson. the elevation θ of the midrib and the angle of rotation φ of the leaf blade around the midrib (Figure 1). France (47°35′ N. Allometric relationships between leaf length and blade area were calibrated from a set of leaves collected from the remaining trees in each treatment. For each subplot. Polhemus. nor plant density nor seedling age had affected the allometric relationship.and 2-year-old seedlings. Germination was about 98%. Inc.nih. Sessile acorns were collected in autumn 1997 from Russy forest. respectively. lamina rotation around the midrib φ) and leaf normal inclination angle α. It was computed as (Polhemus 1993):  sin ψ sin θ cos φ − cos ψ sin φ  ϕ = arctan  . 2012 (2a) Angle ϕ was defined as the angle between the projection of leaf normal onto the horizontal plane and the North direction. At the plot scale. Starting from the base of the seedling. Geometrical measurements of leaf orientation: the three Euler angles (midrib azimuth ψ. Incident PAR was measured with SKP215 quantum sensors (Skye Instruments. leaf normal inclination angle α and the leaf normal azimuth angle ϕ. all measurements were made in the same reference frame.. The soil is a stagnic luvisol without mineral restriction.87. Statistical analysis showed that neither solar irradiance. northeastern France (48°44′ N. The planting densities were 25 (high density. Leaf area density was calculated as the amount of leaf area within the crown projection area. an offset of π must be added if leaf normal points South. Hellevoetsluis. This enabled both the location and orientation of seedlings within the plot and the leaves within the seedling to be recorded. For each treatment.

Values of STAR were thus decomposed into three terms to separate the effects of leaf inclination (Pearcy and Yang 1996).LIGHT INTERCEPTION BY SESSILE OAK SEEDLINGS 1259 Reconstruction of plant images Computer graphics rendering techniques were used to compute hidden sides of foliage and then to estimate the visible or projected leaf area (Foley 1995). each leaf was represented by a single geometrical model consisting of 13 coplanar polygons. 1200 and 1400 h and at zenith on June 21 in Nancy. Images of projected foliage areas on an orthogonal plane in the direction of observation. leaves or soil. with S = S ∑ L. i. S i′ binom (5) where L′ is projected area of leaf l according to its orientation with respect to the light direction (see Ross 1981) and S ′ binom i is the theoretical projected seedling foliage area assuming a random dispersion of leaves in the crown projection area C i′ Figure 2. and S is total leaf area of the respectively. The polygon around the projection of foliage area defined the crown projection area. i. a virtual camera using parallel rays. S is total leaf area of the seedling and L is area of leaf l. which was scaled proportionally to leaf length and blade area. This technique of image synthesis results in the calculation of a projected image of the canopy onto a plane orthogonal to the direction of the rays.povray. Images were calculated for four directions: sun position at 1000. Sun elevation was thus 56°. The STAR of surrounded seedlings quantified the reduction in light interception caused by the surrounding vegetation. Calculation of foliage images was derived from these points of contact (Figures 2A and 2B). l (4) where S i′ digit is projected leaf area of seedling s measured on a plant image where neighboring seedlings have been removed. 2000) by comparing virtual and real photographs of the canopy.. leaf inclination. The value 1 – STAR i provided a measure of self-shading within the seedling at University of Novi Sad on April 18. (Rakocevic et al. Projected foliage area for a given direction was then estimated by counting the pixels of the projected surfaces on the plant images. the comparison between the full sun and shade treatments was based on the vertical direction. Light interception by isolated seedlings The STAR for an isolated seedling (STARi) was defined as: Downloaded from http://treephys. Values of STAR were used to estimate light interception by single seedlings and by surrounded seedlings in the canopy. leaf area density and leaf area dispersion in the crown projection area.e. Oker-Blom and Smolander 1988) as: STAR = S′ . TREE PHYSIOLOGY ONLINE at http://heronpublishing. 1400 h and at zenith (also called the vertical direction). The capacity of the foliage to intercept light is likely to depend on leaf size. (B) Vertical projection of the whole plot. S (3) where S′ is projected foliage area. Because radiation was almost completely diffuse under the shading nets. (Sinoquet and Rivet 1997) and Trifolium repens L. The STAR of single seedlings was used to evaluate the effect of the spatial arrangement of leaf area within the seedling on light interception. Projected area is the foliage seen on an image taken from the direction of the sun with an orthographic camera. leaf area density and dispersion (Planchais and Sinoquet 1998) in the crown projection area: STAR i = ∑ L′ S ′ binom i l S ∑ L′ l S i′ digit . 65°. 2012 STAR i = S i′ digit . Light interception by the foliage was quantified from silhouette to total leaf area ratios (STAR. Calculation of direct light interception from images Light interception by each seedling was assessed by measuring the projected leaf area on plant images. was noted. Each ray was followed individually and the contact with the nearest surface.e. The black area is the visible part of the foliage of a given seedling shaded by surrounding seedlings (shown in grey). Foley 1995) with POV-Ray software (http://www. 55° and 90°.com .oxfordjournals. Computations were made by means of a ray-tracing method (Ariès et al. (A) Isolated seedling according to four directions of observation: sun position at 1000. A large sample of parallel rays from an infinite point light source was projected onto the artificial canopy built from digitized data. 1200. For the foliage reconstruction. 1993. Geometric measurements recorded with the 3-D digitizing device and the plant image reconstruction were previously evaluated with Juglans regia L. Dauzat 1993.

2012 Results ∑ L′ l Leaf area development During Year 1. considering plant density nested with light condition (GLM procedure 1989. plant density had no effect on total leaf area. The results given in Table 1 and Figure 3 are for trunk and branch leaves combined. There were no significant differences in leaf area density between treatments. 1997). leaf area density did not differ between treatments. The first ratio on the right-hand side of Equation 5 accounted for the effect of leaf inclination on light interception. which was derived from the plant images. SINOQUET AND COLIN (Thanisawanyangkura et al. Light interception by seedlings surrounded by neighbors Light interception by a seedling surrounded by its neighbors was defined as: STAR s = S ′ digit s . Figure 3 presents the distributions of the three Euler angles (Figures 3A–C). shaded seedlings had larger individual leaf areas. If long branches emerged from the seedling. which implicitly assumes that intra-leaf shading could occur. Cary. We can also write STAR s as: TREE PHYSIOLOGY VOLUME 21. C ′ was calcui lated as the union of all the polygon surfaces. i  C i′    l (6) STAR s = S ′ digit S ′ digit S ′ digit s i = STAR i s digit . is defined as the area of the smallest convex polygon including the leaf area projection (Figure 2A). Total seedling height was similar in all treatments. total height and leaf area density at the plant scale in either the full sun or shade treatment (Table 1). The value of S i′ binom was calculated according to a binomial law (Nilson 1971):   L′   S ′ binom = 1 − ∏ 1 −   C i′ . Leaf orientation Inclination of leaves on branches was not correlated to branch inclination and was similar to the inclination of leaves on the trunk. The ratio: S i′ binom ∑ L′ l accounts for the effect of leaf area density. The ratio S ′ digit/S i′ binom characterizes the effect of the i non-randomness of leaf location within the crown. number of leaves. especially at low density (LD). In contrast. total leaf area and number of leaves per plant were greater in full sun seedlings than in the shaded seedlings. where S i′ beer is computed from Beer’s law:   L′   S ′ beer = 1 − exp  − ∑   C i′ . respectively (Nilson 1971). it does not account for mutual shading between leaves. 1991). Crown projection area was larger in the full sun treatment. the effect of leaf size on light interception efficiency was assessed by the ratio S i′ binom/ S i′ beer . In at University of Novi Sad on April 18. Values greater and less than 1 accounted for regular and clumped leaf dispersion. Over the 2-year study period. As in Year 1. the difference in total leaf area per plant was related to the larger individual leaf area in the shade treatment. Downloaded from http://treephys. Equation 9 allowed us to estimate the rate of light competition between seedlings. because S i′ binom is the projected area in the case of random shading between leaves. respectively. Statistical analysis Effects of the experimental treatments on foliage features and light interception were analyzed by a classical analysis of variance.oxfordjournals. In Year 2. S S ′ digit S′ i i (9) Crown projection area C i′ . even for seedlings in the full sun + high density (HD) treatment where total leaf area was as large as in the full sun + LD treatment and crown projection area was greatly reduced. However.. Because there was no significant treatment difference in number of leaves. SAS Institute. total leaf area per plant in the shade treatments was larger than in the full sun treatments. Inc. but leaves were smaller than in Year 1. i  l C′    i (7) The value of S i′ beer was computed based on the assumption of infinitesimal leaf size (Nilson 1971). foliage packages were defined on the image and the smallest convex polygon containing each package was drawn. which were directly measured with the 3-D is the sum of the projected areas of inclined leaves. Littell et al. LAI values were higher in shaded plots than in full sun plots. 2001 . computed from a plant image including neighboring seedlings. leaf inclination was independent of the time when the leaves expanded and did not significantly change with the development of the bud in the leaf axil (data not shown). providing more space for leaf display. where there was no spatial limitation. Crown projection area and total height were also greater in the shaded plots than in the full sun plots. S (8) where S ′ digit is the visible part of the projected area of seedling s s within the plot. Furthermore. Because: where the two ratios represent the light interception by the isolated seedling (STARi) and the shading effect from neighbors. In this case. NC. S i′ binom explicitly accounts for leaf size (Equation 6) and does not allow intra-leaf shading. During Year 1. total height. whereas the opposite was true during Year 2 (Table 1).1260 FARQUE.

the second letter to plant density.17 aa 53.31 ± 0. values for the leaf area dispersion effect were close to 1 (Figure 5).18 578.6 ab 11.3 aa 1. Parameter Shade HD First year LAI Total leaf area (cm2) Number of leaves Crown projection area (cm2) Leaf area density (m2 m –2 ) Total height (cm) Individual leaf area (cm2) Leaf inclination angle (°) Second year LAI Total leaf area (cm2) Number of leaves Crown projection area (cm2) Leaf area density (m2 m –2 ) Total height (cm) Individual leaf area (cm2) Leaf inclination angle (°) 0.001) for seedlings in all treatments. Moreover. However. HD.93 321.1 aa 21. Leaf area overlapping in the vertical direction.3 ± 0. when data for Years 1 and 2 were pooled.3 ab 24.7 aa 663.6 aa 21.0 ± 22.0 ± 88. P < 0.9 ± 75.26 ± 0. indicating that leaf area dispersion was close to a random dispersion. the HD and LD data were pooled in Table 2 and Figure 4.8 ± 0.7 aa 246.6 aa 428.61 1327.3 aa 12.4 bb 32. LD).9 ± 6.0 ± 21. In addition.7 aa 1.5 ba 30.43 aa 46. Light interception by isolated seedlings Plant density had no significant effect on seedling STARi: it did not affect either seedling geometrical structure (Table 1) or projected leaf area for the vertical direction (data not shown).0 ± 17.2 aa 54. whereas the influences of leaf inclination and leaf area dispersion were small.3 ± 224.6 ± 275. which are more commonly used to characterize canopy structure and light interception (de Wit 1965. was unaffected by the light treatments.6 ± 70. However.0 ± 2.2 aa 12.38 821.5 aa 16. Canopy structure parameters of Quercus petraea seedlings in two light regimes (shade and full sun) and two plant densities (high density.oxfordjournals. reflecting the more horizontal orientation of foliage in the shaded seedlings.1 aa 16. Ross 1981).6 ± 7.0 aa 1.3 ba LD 0.6 ± 2. 2012 digitizing device. for seedlings in all treatments..1 aa 17.05): the first letter relates to light regime.19 ± 0.1 aa 25.2 aa 45.4 ± 5. Hence. the highest STARi values were found in seedlings with small total leaf areas (Figure 4) and little leaf overlapping. STARi was significantly higher (8%) in shaded seedlings than in full sun seedlings (P = 0.3 ± 0. whereas STARi values for the vertical direction were similar in Years 1 and 2 (Table 2).7 ba 1.5 ba 6.7 ± 1010.5 aa Full sun HD .0 ± 1.4 ± 41.5 aa 195.3 ± 38.8 ba 1297.3 ± 2.23 ± 0. However.4 aa 2.6 aa 76.65 ± 0.2 bb Downloaded from http://treephys.0 ba 18.7 ± 39. TREE PHYSIOLOGY ONLINE at http://heronpublishing. 87% of the leaf area of 1-year-old seedlings and 57% of the leaf area of 2-year-old seedlings had positive midrib elevation angles in the full sun treatment. Values are means per plant ± SE (n = 5).5 ± 0.2 ± 0. integrating leaf area density and dispersion effects. the leaf inclination effect was larger in full sun seedlings than in shaded seedlings.5 ± 4.9 ba 24.8 ± 147.2 ± 20.1 ± 10. Mean absolute value of midrib elevation angle and mean leaf normal inclination angle (Table 1) were slightly higher in Year 2 than in Year 1 (Table 1) and leaves appeared more erect in full sun seedlings than in shaded seedlings (Figure 3B).5 aa at University of Novi Sad on April 18.8 ± 2.31 ± 0.5 ± 241. i.8 ± 262.6 aa 98. Total leaf area and projected leaf area of full sun seedlings increased markedly between the first and the second year of growth. Within a row.7 bd 0.52 1672. as well as leaf normal inclination and azimuth angles (Figures 3D and 3E).7 ± 0.0 ba 241.6 ba 23.0 ± 13.1 bb 1.3 aa 13. Lamina rotation angle distributions (Figure 3C) were also planophile.9 bb 30. For seedlings in all treatments. Individual leaf area and leaf inclination angle were calculated by averaging values of all leaves from the same seedling. The STARi values were higher in shaded seedlings than in full sun seedlings in both Year 1 and Year 2.8 ± 6. leaves on shaded seedlings tended to hang down: 62% of the leaf area of 1-year-old shaded seedlings and 68% of the leaf area of 2-year-old shaded seedlings had negative midrib elevation angles.2 ± 123.3 aa 15.02 ± 0. and mean angles were close to zero in all treatments.0 ba 9.05 172.6 aa 6. different letters denote significant differences between treatments (ANOVA.42 ± 0.70 aa 49. Plant density treatment had no clear effect on mean leaf angles. and low density. indicating that light interception and mutual shading within leaves were mostly independent of leaf size.0 ± 2.1 ± 69.3 ± 42. but the differences were not significant. Midrib and leaf normal azimuth angle distributions (Figures 3A and 3D) were constant across seedling age and treatments.7 ± 10.0275) (Table 2 and Figure 5).44 161.LIGHT INTERCEPTION BY SESSILE OAK SEEDLINGS 1261 Table 1.5 ± 0.9 ± 1.0 ± 220.6 ± 4.8 ba 138.4 ± 0.26 aa 60.4 aa LD 0. As shown in Figure 4.1 aa 24.5 ± 405. high values of vertical STARi were associated with the small total leaf areas of shaded seedlings.4 ba 1.2 ba 734.e.0 aa 9.45 ± 0.9 ± 10. except for LAI which was calculated for each plot (n = 1). In contrast. STARi was mainly determined by the leaf area density effect. the ratio S i′ binom/ S i′ beer was close to 1 (1.3 ab 0.5 aa 1.0 ± 11.8 ba 2.09 288. Midrib elevation and leaf normal inclination angle distributions (Figures 3B and 3E) were close to planophile distributions defined by fairly horizontal leaves (de Wit 1965).7 bc 3.0 ± 7. Figure 5 shows that.

org/ at University of Novi Sad on April 18. leaf normal azimuth angle (D) and leaf normal inclination angle (E) for Quercus petraea seedlings grown in two light regimes (shade and full sun) at two planting densities (HD and LD) at the end of the first and the second growing season. midrib elevation angle (B). Distribution of midrib azimuth angle (A). TREE PHYSIOLOGY VOLUME 21. the theoretical planophile distribution (de Wit 1965) is shown.oxfordjournals. 2001 .1262 FARQUE. 2012 Figure 3. For the leaf normal inclination distribution. lamina rotation around the midrib (C). SINOQUET AND COLIN Downloaded from http://treephys.

54 ± 0.8 a 154. Total leaf area. During the 2-year study.1 a 700. which tended toward a regular dispersion (i.1 a 166. projected leaf area and light interception efficiency (STARi) for the vertical direction of seedlings in the full sun and shade treatments for each year of growth.6 a 362.05).and 2-year-old Quercus petraea seedlings in the shade and full sun treatments. 1200 and 1400 h for 1. The HD and LD data were pooled because they were not significantly different.LIGHT INTERCEPTION BY SESSILE OAK SEEDLINGS Table 2.45 ± 0.9 ± 113. P < 0. but only for the seedlings in the HD treatment. suggesting that the vertical direction provides good estimates of light interception properties. Shade First year Total leaf area (cm2) Projected leaf area (cm2) STAR i Second year Total leaf area (cm2) Projected leaf area (cm2) STAR i Full sun 1263 305. The HD and LD data were combined because they did not differ significantly (data not shown). the ratio was equal to 1 (Figure 6). Light-interception efficiency (STARi) and the effects of leaf inclination. the sun direction integrated leaf area dispersion in both the horizontal and vertical directions.9 ± 29. Mean STARi for the three sun directions was Figure 5. Light interception by seedlings surrounded by neighbors Because there was no competition for space between seedlings during Year 1. mutual shading between seedlings and competition for light was Figure 4. Values of STARi for the vertical direction as a function of total leaf area. close to the vertical STARi.46 ± 0. This difference can be related to the significantly larger values of mean leaf inclination angle (Table 1) and the lower (nonsignificantly) leaf area density (Table 1). In the HD treatment (Figure 6). Intermingling of crowns occurred during Year 2. The 1-year-old seedlings in the full sun + LD treatment exhibited lower STARi values (Figure 5) than seedlings in the other treatments. leaf area density (except in 2-year-old seedlings) and leaf area dispersion contributed to the higher STARi values in shaded seedlings (Table 3). means followed by the same letter are not significantly different (ANOVA.3 ± 721. values greater than 1) in shaded seedlings (Figure 5).8 b 63. Results of statistical analyses are given in Table 3. projected leaf area for surrounded seedlings was reduced by 32% (± 11%) in the full sun treatment and 23% (± 8%) in the shade treatment compared with the projected leaf area for isolated seedlings. The treatment difference between slopes was significant (P = 0. leaf inclination. for 1.54 ± 0.2 a 1500.0 a 0.3 ± 61..0 ± 153.1 ± 339.oxfordjournals. leaf area density and leaf area dispersion at 1000. 2012 Computations for the sun direction at 1000. During the day. maximal interception was generally observed at 1200 h. At a given plant density.0001).4 a 0.1 a Downloaded from at University of Novi Sad on April 18. especially for the 1-year-old seedlings. whereas only leaf inclination had a significant effect for the vertical direction. Compared with the vertical direction.e. TREE PHYSIOLOGY ONLINE at http://heronpublishing. 1200 and 1400 h on June 21 in Nancy confirmed that shaded seedlings had a greater light interception efficiency than full sun seedlings (Table 3 and Figure 5).5 b 0. The STAR ratios calculated for the vertical direction of projection are presented for comparison.9 ± 229.7 b 641. Within a row.7 ± .5 b 0.and 2-year-old Quercus petraea seedlings in the shade and full sun treatments.

leaf inclination angle and leaf overlapping. petraea seedlings. which was nested in the full sun treatments. In each treatment. Significance level was 0.0001 0. Kappel and Flore 1983.0008 0. Gottschalk 1994.oxfordjournals.0001 0.and 2-year-old Quercus petraea seedlings in the shade + high planting density (HD) and full sun + HD treatments measured by comparing the projected foliage area of a seedling with and without surrounding neighbors.0031 0. However. and the effects of leaf inclination.0032 0. on June 21 at Nancy.0123 0. the degree of shading by neighbors was represented by the difference between the slope (1:1) and the slope of the values in each treatment. petraea seedlings. Gottschalk TREE PHYSIOLOGY VOLUME 21. each factor was analyzed by grouping the other factors. Our results suggest that 18% shade is already limiting for leaf area development in 2-year-old Q. Light interception by 1. the soil was drier than under the nets. ramification and branch length and thus leaf area development and lateral expansion.0001 0. Notwithstanding the reduction in incoming radiation.0001 ns ns 0. Means and variations are given in Figure 5. except plant density. The vertical direction was analyzed separately from the sun directions. seedling leaf area and number of leaves were greater in full sunlight than in 18% sunlight.0001 0.0003 0. P-Values of analysis of variance for variables STARi. 2001 .0001 0. In Q. despite frequent irrigation in the full sun treatment. this trend was not observed for the 1-year-old seedlings.05. 2012 Leaf inclination effect Leaf area density effect ns ns – ns ns – Leaf area dispersion effect higher in seedlings in the full sun treatment than in the shade treatment. These more favorable growth conditions could partly account for the greater leaf area development in the shaded seedlings (Madsen and Larsen 1997). Ziegenhagen and Kausch 1995). The sun direction factor included the three directions of the sun at 1000.0001 Vertical direction ns ns – 0.0001 ns 0. oak seedlings utilize carbohydrates stored in the cotyledons and leaf area development is almost completely insensitive to solar irradiance (Welander and Ottosson 1998. because of the larger crown projection area of seedlings in the full sun treatment. where total leaf area was greater in shaded seedlings than in full sun seedlings.0180 – ns ns – ns ns – Two-year-old Sun directions 0. Intermingling of crowns occurred during the second year of growth but no morphological response to competition was visible. Impacts of shading and plant density on canopy structure and light interception Figure 6. because the shade nets reduced evaporation and.1264 at University of Novi Sad on April 18. 1200 and 1400 h. in 3-year-old Q. Variables Factors One-year-old Sun directions STAR i Light Density (Light) Sun direction Light Density (Light) Sun direction Light Density (Light) Sun direction Light Density (Light) Sun direction 0. For each year. No plant density effects were observed on foliage development. calculated for 1.and 2-year-old Quercus petraea seedlings. LAI. mean leaf size increased in response to shading in 1. leaf area density and leaf area dispersion. respectively. as in many species. growth conditions were more favorable in the shade treatment than in the full sun treatment.0008 Vertical direction ns ns – 0. Ss′ and S i′. Ke and Werger 1999). petraea seedlings.and 2year-old seedlings (cf. Light interception is affected by modifications to canopy structure in response to low irradiance. During the first year after seedling emergence. Self-shading can be modified through leaf size and form.0067 ns – Downloaded from http://treephys.0098 ns 0. Many authors have reported that maximal leaf area in Quercus seedlings occurs at around 20–50% of full sunlight (Jarvis 1964. bud production. However. Lower irradiances should produce a negative effect on leaf area development.0035 0.0452 0. SINOQUET AND COLIN Table 3. Astre (1999) observed that increasing plant density reduced shoot growth.0001 0. Discussion Impact of shading and plant density on leaf area development During the second year of growth.0004 ns ns 0.

petraea seedlings. thereby maximizing light capture at low irradiances. 1997).LIGHT INTERCEPTION BY SESSILE OAK SEEDLINGS 1265 1994. Values of STAR in deciduous species are compared with STAR × 2 values in coniferous species.62 0.30 0. STAR values increased in response to shading. and then eroded the image. In Q. petraea leaves. however. It is. the magnitude of leaf inclination angle and leaf overlapping was greater in F. In addition. Planchais and Sinoquet 1998). petiole length ranged from 0. Takenaka (1994) underlined the importance of petiole length in light capture. sunlit seedlings produced lateral shoots that increased leaf clumping in the oblique directions. allowing comparison with the other values presented. 2012 Table 4. Shading did not modify leaf area density or the dispersion effect in the vertical direction. therefore. Our results suggest that leaf overlapping in . L. in all species. 1997. petraea seedlings.242 0.36 0. petraea has a minor effect on light-interception efficiency at the seedling stage. 1999 This study Planchais and Sinoquet 1998 0. Thanisawanyangkura et al. Abies lasiocarpa (Hook. sylvatica (Planchais and Sinoquet 1998) than in Q. however. Table 4 compares STAR values of young trees of different species in contrasting light environments.54 0. Quercus petraea Fagus sylvatica 1 2 3 % Light 75 20 75 20 75 20 901 231 100 18 100 5 Scale Shoot Shoot Shoot Shoot Shoot Shoot Shoot Shoot Seedling Seedling Branch Branch STAR STAR × 2 0. to determine whether gaps are included in the canopy volume. petraea. data on Fagus sylvatica trees measured at the branch level are also included. These findings suggest that the assumption of infinitesimal leaf size used in Beer’s law is likely to be valid. Collet et al.) (Thanisawanyangkura et al.28 0. an important parameter related to the ability of the plant to compete in low-light environments (Niinemets 1998a). Farque.46 Reference Carter and Smith 1985 Stenberg et al. we can conclude that Q. We note. because of the scarcity of STAR values for deciduous trees. Picea abies (L.50 3 0. Thus. In contrast. Differences in the range of STAR values among species probably reflect differences in morphological plasticity to light. This is a crucial issue because space occupation is a main determinant of resource capture.) Karst. unpublished data). which was especially important at the 1-year-old stage. As pointed out by Nilson (1992). Beech branches were sampled at different heights in the canopy so the mean STAR values are representative of the whole plant. Moreover.oxfordjournals. our comparison between projected leaf area computed from Beer’s law and the binomial law showed that changes in leaf size did not modify self-shading within the canopy and did not affect light interception of Q.e. Pinus concorta Dougl. Species Picea engelmannii Parry ex Engelm. unpublished data). but concluded that petiole length is insufficient to reduce aggregation of leaf area around the stem and improve light interception efficiency. This supports the general observation that leaves in sunlit environments are more erect than leaves in shade environments (McMillen and McClendon 1979. because STAR for coniferous shoots refers to shoot silhouette area divided by all-sided needle area. TREE PHYSIOLOGY ONLINE at http://heronpublishing. we lack objective methods to define the volume occupied by a single plant. Nevertheless. sylvatica shoots and branches was about 30%. whereas Planchais and Sinoquet (1998) derived C i′ from plant image processing: they dilated the binarized image until gaps within the plant projection were filled.24 0. shading decreased leaf clumping in the oblique directions. Niinemets and Kull 1994.. pet- Downloaded from http://treephys. even in the case of small individual plants. Vertical STAR values.26 0. The increase in STAR in shaded Q. but the variation was independent of the light regime (L. Few differences in leaf overlapping were observed between shaded and sunlit seedlings. ex Loud. whereas it was 8% for our Q.5 to 10 mm. Measured values of STAR in different tree species.) (Planchais and Sinoquet 1998) and cotton (Gossypium hirsutum L. i. STARmax values. The shade-induced increase in STAR in F.) Nutt. that our clumping analysis in isolated seedlings requires a value for crown projection area. Niinemets 1998b). Farque. According to these authors. (1997) defined C i′ as an ellipse fitting plant area projection. However. This phenotypic plasticity allows construction of an extensive foliar display from a given biomass investment in leaves. The data in Table 4 show that. Similar results have been observed in beech (Fagus sylvatica L.46 3 at University of Novi Sad on April 18. indicating that seedling development may have more influence on leaf overlapping than irradiance. In our study C i′ was defined as the smallest convex polygon enclosing the leaf area projection (Figure 2A). C i′ . petraea seedlings was mainly explained by the more horizontal leaf inclination (cf.84 Calculated from indices of canopy openness. when the seedlings had few branches and foliage was clumped around the trunk (Harmer 1989. Planchais and Sinoquet 1998). leaf size increases in parallel with decreases in leaf mass per area. clumping generally increased with decreasing leaf size.

.) and Douglas-fir (Pseudotsuga menziesii Franco) in containers. Manage. Ke. Ann. 82: 1342–1442. Colin and C. Bot.. Sinoquet. specific leaf weight. H.E. For. leaf chlorophyll content. according to Ballaré (1994).. In particular. R. J. SINOQUET AND COLIN Astre. C. C. Woodland canopy structure and the light response of juvenile Quercus lobata (Fagaceae). Klockare and C... 21: 373–381.. Res. G. 16 p. Le Goff and F. N. Sci. Photosynthesis of leaf canopies. C. 79:1038–1043. Flore. I. This increment was mainly accounted for by a decrease in leaf inclination and was not associated with changes in leaf overlapping.). and morphology of young peach trees. H. Ecol. petraea seedlings exhibited shade avoidance responses owing to neighbor proximity. 1975. Simulated plants and radiative transfer simulations. variations in morphological plasticity seemed to parallel the shade tolerance of the species. Editions du Perron. INRA. F. 1993. 1993. and W. Ann. Holmes. R. 54: 65–81. C. T. 2012 Conclusion Although Q. Alleur-Liège. The effect of mineral nutrients on growth. Influence de la densité de plantation sur les composantes de la croissance en hauteur de jeunes chênes sessiles (Quercus petraea (Matt. Hortic. For. such as shoot elongation or reduced branching. Feiner and J. pp 271–278. 1998.H. Collet. Nouguier and B. and J. Geometrical canopy modelling in radiation simulation studies. K. Low irradiance increased light interception by 8%. F. R. and B. Effect of shade on photosynthesis. Report No.A. Montpellier II. P. Hughes. Bernier. Modifying the microclimate around young oaks through vegetation manipulation: effects on seedling growth and branching. Because changes in leaf arrangement and structure also have consequences for the water and carbon balance of the plant. M.L.P. Addison Wesley. Jarvis. Plant Physiol. Height growth. New York. J.... Caldwell and R. 20:579–586. 1999. 14:735–749. Collet.) Liebl. Wageningen. Influence of shoot structure on light interception and photosynthesis in conifers. Ballaré. C. In the hardwood species. if our Q. 1997. 1994. raea exhibits a lower morphological plasticity to light than F.. Can. 1175 p. these responses had no effect on foliage structure and light interception capacities. Forestry 62:383–395. Dreyer. Adam for their computer support. pp 159–173. petraea seedlings could help to predict seedling development during natural regeneration in mixed stands.M.W.T.F. petraea is considered a shade-tolerant species. 1997. G. seedlings exhibited low morphological plasticity in foliage structure in response to shading. INRA. Am.. 1994. Bonhomme and H. Prunus serotina and Acer rubrum seedlings. Scots pine (Pinus sylvestris L. Academic Press. J. R.) Liebl. Werger.K. although foliage structure of our Q. For. S. Plant. For. Collet. L. J. Many morphological changes in foliage structure tend to saturate at a particular irradiance. In Crop Structure and Light Microclimate. Bot. 1995.G. Bernier. and POV-RayTM. F. Computer graphics. and J. this study provides the basis for further analyses on the structure–function relationships in Q. Sci. 45:387–392. Downloaded from http://treephys. C. apical dominance and branching in Quercus petraea (Matt) Liebl. Collet.K. Le chêne. F. ex Loud. Ecol. Colin and F. Gottschalk. In Exploitation of Environmental Heterogeneity by Plants. 108:541–544. L. Barry-Langer. Growth and assimilation rate of juvenile northern red oak: effects of light and temperature. the range of STAR values) was lower. Harmer. Plant density did not modify foliage structure or light interception during the 2-year study. 1985. VarletGrancher. Garnier for their technical assistance. 110: 249–262. Sundqvist. Ningre and H. 1998. Carter. it is possible that our low irradiance treatment (18% of full sunlight) was too high to effect changes in canopy structure. petraea seedlings. TREE PHYSIOLOGY VOLUME 21. 1983. Varlet-Grancher. Effects of environmental factors on the shoot development of Quercus petraea seedlings. Fagaceae). Eds. F. Also. Smith. Axelsson. Growth dynamics and water uptake of two forest grasses differing in their growth strategy and potentially competing with forest seedlings.M. Jinks. Physiol. Jr. Guehl. 97:119–131. a knowledge of the combined effects of morphological and physiological plasticity to light (leaf photosynthesis.J. at University of Novi Sad on April 18. Thus. pp 1–57. C. flushing. plants perceive radiation reflected from nearby vegetation as an indicator of neighbor proximity and respond morphologically even before they are directly shaded. Foley. Morphological development. Agric. A. Prévot and P.A. 1964.e. Oak seedlings grown in different light qualities. Manage. Univ. stomatal conductance) on carbon acquisition and water use of Q. C. M. 2001 . Nevertheless. Thesis. Effect of seedling density on the growth of Corsican pine (Pinus nigra var. Soc.oxfordjournals. leaf growth and crown development of Quercus rubra. J. de Wit. Farmer.1266 FARQUE. Different responses to shade of evergreen and deciduous oak seedlings and the effect of acorn size. References Ariès. France. Frochot. Acta Oecol. 663. petraea seedlings was unaffected by plant density during the 2-year study. A methodological approach. 52:545–571. 1989. Pearcy. Quercus velutina. and M.D. R. 1993. Chaar. However. Sci. shoot elongation and branch development of young Quercus petraea grown under different levels of resource availability. 1979. J. because numerous processes are determined by light interception. Dauzat. In Crop Structure and Light Microclimate. pp 73–110. B. Kappel. Sci. sylvatica and the coniferous species (except Pinus contorta Dougl. 74: 1555–1561. Ecol. Monestiez. Carter and Smith 1985) because the effect of changes in foliage structure on light interception (i. Ningre for their helpful comments and suggestions on the manuscript. Bonhomme and H. Shade. San Diego. The adaptability to light intensity of seedlings of Quercus petraea (Matt. we note that the study terminated before competition between plants had developed. Frochot and J. C.Sc. R.A. maritima Melv. 1996. 1999. For. Sinoquet. Acknowledgments Thanks to E. Am. Montpellier. 1995. especially in seedlings. Eds. Mason. Canta and B. Nebout. 55:407–423. 604 p. 1965. Eds. Tree Physiol. Light gaps: sensing the light opportunities in highly dynamics canopy environments. van Dam. A.

1998. canopy openness. 1998a. Rivet.oxfordjournals. A theoretical analysis of the frequency of gaps in plant stands. 2012 TREE PHYSIOLOGY ONLINE at http://heronpublishing. Ross. 70:1–10. Homewood. Downloaded from http://treephys. P. and H. 1994. Stenberg. and W. Manage. Measurement and visualization of the architecture of an adult tree based on a three-dimensional digitising device. Ottosson. Takenaka. For. Res. Bot. 2000. G. H. Meteorol. H. 82:203–212. Pearcy.LIGHT INTERCEPTION BY SESSILE OAK SEEDLINGS Littell. The radiation regime and architecture of plant stands. For. Sinoquet. Nilson. Junk Publishers. 2nd Edn. McClendon. A three-dimensional crown architecture model for assessment of light capture and carbon gain by understory plants. Bot. Springer-Verlag. Linder. Sinoquet. T. A.J.H. 1997. T.) as corresponding to shoot morphology and leaf anatomy.) with respect to canopy density. Sci. Radiative transfer in nonhomogeneous plant canopies. Ecol. NC. B. Adjustment of foliage structure and function to a canopy light gradient in two co-existing deciduous trees. Spector. revision F. and B. T. SAS System for linear models. Agric. Variability in leaf inclination angles in relation to the petiole morphology.T. Shoot structure. A. Sinoquet. Trees 11:265–270. Meteorol. J. Productivity of young shaded oaks (Quercus robur L. Colchester. and H. Cary. Niinemets. 86: 519–526. 72:97–108. Oecologia 108:1–12. and J. W. 1998b. Varlet-Grancher. Assessing the geometric structure of a white clover (Trifolium repens L. Kull. 329 p. Kausch. For.. 134:1–11. Kasemsap. H. Ann. 1996. Characterization of the light environment in canopies using 3D digitising and image processing. Ann. 391 p. Meteorol. Foliage determinants of light interception in sunny and shaded branches of Fagus sylvatica (L. Leaf orientation and sunlit leaf area distribution in cotton. 1127 p. 107:117–126. R. Ü. Plant Cell Environ. Richard D.. Ü and K. H. Ed. Welander. S.. 1997. 1994. Vol. 1997.. Sinoquet.. Ecol.) canopy using 3-D digitising. Ü. Rivet.. and P. R. Plant Ecol. 1998. Irwin. 97:95–105. Bot. Agric. P. 1991. Thanisawanyangkura.C. Ecol. Leaf angle: an adaptive feature of sun and shade leaves. P. soil moisture and soil carbon content. Manage. H. 1985. Jallas. 1971. 1992. Are compound-leaved woody species inherently shade-intolerant? An analysis of species ecological requirements and foliar support costs. Christophe and C.G. 3SPACE FASTRAK User’s manual. I. G. 89:241–253. For. Mabrouk and P. Oker-Blom. Gaz. and light interception in Norway spruce. J. The ratio of silhouette area to total needle area in Scots pine. 9: 109–114. 1. SAS Institute. and W. N. Manage. Kangas. The Netherlands. Ecol.H. 1995. Ecol. Trees 12:446–451. 86:1–15. M. 158 p. 140:437–442. S. 1998. For. Smolander.. 1981. P. VT. Effects of leaf blade narrowness and petiole length on the light capture efficiency of a shoot. Yang. H. Madsen. Berlin. Sinoquet. For. Niinemets. pp 59–88. 1999. 1979. 8:25–38. Freund and P. Manage.C. 1993. and J.W. Wasserman and M. 1998. R. Rakovecic. 3rd Edn. Neter. Natural regeneration of beech (Fagus sylvatica L. Planchais.). The influence of shading on growth and morphology in seedlings of Quercus robur L. McMillen. For. Nilson. Applied linear statistical models. P. and Fagus sylvatica. Niinemets. Ziegenhagen. Kutner. Cretenet and E. Agric. . In Advances in Bioclimatology. Thanisawanyangkura. Leaf weight per area and leaf size of 85 Estonian woody species in relation to shade tolerance and light availability. at University of Novi Sad on April 18. Larsen. 1267 Polhemus Inc. Smolander and S. 22:1133–1142. 34:894–906.

Sign up to vote on this title
UsefulNot useful