Variation in the ratio of shoot silhouette area to needle area in fertilized and unfertilized Norway spruce trees

Transcription

1 Tree Physiology 15, Heron Publishing Victoria, Canada Variation in the ratio of shoot silhouette area to needle area in fertilized and unfertilized Norway spruce trees PAULINE STENBERG, 1 SUNE LINDER 2 and HEIKKI SMOLANDER 3 1 Department of Forest Ecology, P.O. Box 24, University of Helsinki, SF Helsinki, Finland 2 Department of Ecology and Environmental Research, Swedish University of Agricultural Sciences, P.O. Box 7072, S Uppsala, Sweden 3 The Finnish Forest Research Institute, Suonenjoki Research Station, SF Suonenjoki, Finland Received November 17, 1994 Summary We compared the range and variation in shoot silhouette area to projected leaf area ratio (SPAR) in fertilized and unfertilized (control) Norway spruce (Picea abies (L.) Karst.) trees. We measured SPAR for several view directions of 169 shoots at different depths in the crown of fertilized and control trees. There was an increase in SPAR with depth in the crown in both control and fertilized trees. In the fertilized trees, however, mean SPAR was larger overall, the increase with depth in the crown was steeper, and there was a larger variation in SPAR with inclination and rotation angle of the shoot (relative to the view direction). In particular, shoots in the lower crown of fertilized trees were rotationally asymmetrical ( flat ) and had high values of the maximum ratio of shoot silhouette area to projected leaf area (SPAR max ). Differences in SPAR between fertilized and control trees were explained by changes in shoot structure in response to fertilization and shading. Shoots of fertilized trees were larger and had more needle area than shoots of control trees. However, the ratio of needle area to shoot size was smaller in fertilized trees than in control trees, implying less within-shoot shading and, consequently, a larger SPAR. Also, the increase in SPAR with increased shading (depth in the crown) could be explained by a decrease in the ratio of needle area to shoot size. In addition, because fertilized trees had more needle area than control trees, the effect of shading at a given depth in the crown was more pronounced in fertilized trees than in control trees. Keywords: fertilization, light interception efficiency, Picea abies, shading, shoot structure, SPAR, STAR. Introduction Because light interception by a coniferous shoot is determined by its silhouette (shadow) area rather than its total needle surface area, the ratio of silhouette to total leaf area (STAR) provides a means of quantifying the light interception efficiency of a shoot (i.e., the mean interception per unit of needle area). The STAR depends on shoot structure and varies with the direction of the shoot relative to the view direction ( sun ). The mean STAR (STAR) taken over all directions in space (a spherical average) represents the efficiency of diffuse (isotropic) light interceptance by the shoot (Oker-Blom and Smolander 1988). Further, because the spherically averaged projected area of a single needle (if assumed to be a convex body) equals one fourth of its total surface area (Lang 1991), the departure of STAR from this value (0.25) is a direct consequence of the mutual shading of needles on the shoot. Thus, because STAR is a measure of the decrease in effective leaf area resulting from clumping of needles into shoots, it is a useful concept for modeling light interception by coniferous canopies (Stenberg et al. 1994a). The STAR can also be used to modify indirect estimates of leaf area index based on light penetration, e.g., radiative techniques such as the LAI-2000 plant canopy analyzer (Stenberg et al. 1994b). In STAR, the total needle surface area is used because it is a well-defined and unambiguous measure, and because it represents the light intercepting and photosynthesizing area of a needle (Smolander et al. 1994). However, often (as in this study) only the projected needle area is available. Consequently, we introduce the silhouette area to projected leaf area ratio (SPAR) to denote the ratio of shoot silhouette area to the vertically projected area of its needles when detached and laid out flat in the horizontal. The SPAR differs from the silhouette area ratio used by, e.g., Tucker et al. (1987) (SAR) and Leverenz and Hinckley (1990) (R max ) in that SAR and R max are defined as the maximum ratio (i.e., the silhouette area has been measured as the vertical projection of a horizontally lying shoot). The spherical average of SPAR is denoted as SPAR. The SPAR (or SPAR) can be converted to STAR (or STAR) by dividing the ratio of total needle surface area to projected needle area. The aim of this study was to assess the range and variation in SPAR of fertilized and unfertilized (control) Norway spruce (Picea abies (L.) Karst.) trees, as affected by shoot structure and orientation, shoot age and shading (position in the crown). Material Branches were obtained from a 30-year-old Norway spruce stand, situated at the Flakaliden Research Area in northern Sweden (64 7 N, E, 310 m asl). An irrigation and

2 706 STENBERG, LINDER AND SMOLANDER fertilization experiment had been established in the stand in 1987, when the stand was divided in plots of m and the following treatments were applied: irrigation, solid fertilization, irrigation combined with liquid fertilization, and a control (no treatment) (Linder 1990). Branches were collected in July 1993 from the 4th, 7th, 10th and 13th whorls (from the top) of two trees on a control plot and two trees on an irrigated and fertilized plot (Table 1). The sample included in total 169 second-order shoots taken from a lateral (second-order axis) at each position of the main branch axis (for structural terminology see Flower-Ellis et al. 1976). The distribution of sample shoots by whorls and age-classes was approximately the same for control and fertilized trees, except that the control data set had a somewhat larger proportion of current-year shoots than the fertilized data set (Table 1). Methods Measurements Total length, width, vertical radius and needle-free angle of the intact shoots were measured (Figure 1). Silhouette areas (SA s ) of the shoots were then measured photographically at four different inclinations (φ = 0, 30, 60 and 90 ) of the shoot axis to the plane of projection using a video system described by Oker-Blom and Smolander (1988). Each of the pixels of the video camera corresponds to 0.06 mm 2 in the plane of the shoot. At each inclination, the shoot was first measured with its upper side facing the camera ( rotation angle α = 0 ) and was then rotated twice in increments of 45 (α = 45 and 90 ), giving a total of 12 silhouette areas measured per shoot. Needles were then detached from the shoot, and the following morphological measurements were made: diameter and length of shoot axis (twig), number of needles, average needle length, and projected needle area (Table 2). Projected needle area was measured with the same video camera that was used to measure shoot silhouette areas. Twig area (projected) was calculated as twig diameter multiplied by twig length. In Figure 1. Definition of (a) shoot length and shoot width, and (b) vertical radius and needle-free angle. In (a), the shoot axis is perpendicular to the view direction and the shoot s upper side is facing the viewer (φ =0 andα = 0 ). In (b), the shoot axis is parallel to the view direction (φ = 90 ). addition, needle cover was calculated as the ratio of projected needle area to projected shoot envelope area, defined as shoot length multiplied by shoot width. Calculations The silhouette area to projected leaf area ratio (SPAR) of a shoot is defined as: SAs( φ, α ) (1) SPAR( φ, α) =, PA n where SA s (φ,α) is the shoot silhouette area for inclination (φ) and rotation angle (α), and PA n is the projected needle area. The SPAR varies with the direction (φ,α) and normally attains its maximum value (SPAR max )atφ =0 andα = 0, (shoot axis is perpendicular to the view direction with upper side of the shoot facing the viewer (camera)), and its minimum value (SPAR min )atφ = 90 (shoot axis parallel to the view direction). The mean SPAR for a given inclination (φ) was calculated as: Table 1. Sampling scheme; DBH = diameter at breast height. Tree characteristics Plot Tree height (cm) DBH (mm) No. of sample shoots Control Fertilized Distribution of shoots by whorls Plot No. of shoots Whorl 4 Whorl 7 Whorl 10 Whorl 13 Control Fertilized Number of shoots per age class Plot Current-year One-year-old Two-year-old Older Control Fertilized

3 VARIATION IN SPAR OF NORWAY SPRUCE 707 Table 2. Shoot characteristics. Characteristic Control Fertilized Range Mean Range Mean Shoot length (cm) Shoot width (cm) Shoot radius (cm) Needle-free angle ( ) Needle area (cm 2 ) Needle length (mm) Needle cover No. of needles (cm 2 ) SPAR( φ ) = SPAR( φ, α ), 3 α where the average was taken over α = 0, 45 and 90. Note that for φ = 90 there should be no variation with rotation angle (α), and the value of SPAR thus represents the mean of three replicates. The SPAR (spherically averaged SPAR) is defined as the integral: π / 2 0 SPAR = SPAR( φ)cos φdφ, and was estimated by interpolating SPAR(φ) with a quadratic spline function ( f ), i.e., a piecewise quadratic function going through the measured points and having continuous derivatives at intermediate points (e.g., DeBoor 1987). We assume that f (0) = 0, which provides an additional constraint necessary to define the function uniquely. (2) (3) SPAR between fertilized and control shoots was largest at φ = 0 (SPAR max ) and quite small at φ = 90 (SPAR min ). (Note that although SPAR max and SPAR min are not defined as SPAR(0,0) and SPAR(90,0), they were similar in our material.) Shoot asymmetry around the shoot axis, as indicated by the variation in SPAR(0,α) with rotation angle (α), was more pronounced for fertilized shoots than for control shoots (Figure 3b). For fertilized shoots, SPAR(0,α) decreased from a mean of at α = 0 (SPAR max ) to at α = 90 (side view), whereas for control shoots, it decreased from to In both data sets, SPAR, SPAR min and SPAR max all increased with whorl number (Figure 4). The increase was larger in fertilized trees than in control trees and steepest for SPAR max. In the fertilized trees, SPAR increased from a mean of (Whorl 4) to (Whorl 13) (47%), SPAR min increased from to (102%), and SPAR max increased from to 1.13 (63%). (Note that SPAR max may exceed unity because the twig is included in the silhouette area but not in the denominator of SPAR.) In the control trees, SPAR increased from a Results Frequency distributions of SPAR for the two data sets are shown in Figure 2. The mean (arithmetic) and standard deviation of SPAR were ± and ± for control and fertilized shoots, respectively. No dependency of SPAR on needle age was found. The mean SPAR in current-year, and 1- and 2-year-old needles were 0.512, and for control shoots, and 0.709, and for fertilized shoots, respectively. Directional and spatial variation in SPAR The directional variation in SPAR was described by two components: (i) variation in SPAR(φ,α)(α = 0, upper side of shoot facing the camera) with inclination angle (φ), and (ii) variation in SPAR(0, α) (shoot axis perpendicular to the view direction) with rotation angle (α) (Figure 3). The SPAR of fertilized shoots was larger and varied more with direction than the SPAR of control shoots. The mean SPAR(φ,0) of control shoots decreased from at φ =0 to at φ = 90, whereas for fertilized shoots it decreased from to (Figure 3a). Subsequently, the difference in Figure 2. Frequency distribution of SPAR in control and fertilized shoots.

4 708 STENBERG, LINDER AND SMOLANDER Figure 3. Variation in SPAR with (a) inclination angle and (b) rotation angle. The points are means for control and fertilized shoots, and the bars represent the standard deviations. mean of (Whorl 4) to (Whorl 13) (28%), SPAR min increased from to (83%), and SPAR max increased from to (36%). In relative terms, SPAR min increased more than SPAR max and, as a result, the ratio of SPAR max to SPAR min decreased from 2.5 (Whorl 4) to 1.8 (Whorl 13) in the control trees and from 3.2 to 2.5 in the fertilized trees. This implies a smaller variation in SPAR with inclination angle in the lower crown than in the upper crown. In contrast, variation with rotation angle (shoot asymmetry) increased with depth in the crown (Figure 5). In the upper crown (Whorl 4), SPAR(0,0) ( SPAR max ) was only 11% higher than SPAR(0,90) (α = 90, side view), whereas in the lower crown (Whorl 13), the difference was 35% for control shoots and 56% for fertilized shoots. Dependency of SPAR on shoot structure Because SPAR is inversely related to the degree of mutual shading of needles in the shoot, both a small needle area per shoot and a small ratio of needle area to shoot size (e.g., a small needle cover) will cause SPAR to increase. The connection between SPAR and shoot size is not straightforward; however, at a fixed needle density (ratio of needle area to shoot size), SPAR would be expected to decrease with shoot size. Shoot size, needle length and mean needle area per shoot were on average larger for fertilized shoots than for control shoots, but the number of needles per unit length of twig and the needle cover were smaller (Table 2). Both needle area and needle cover decreased with depth in the crown (Figure 6). The decrease in needle cover was similar in both fertilized and control trees, implying that the difference in needle cover between control and fertilized shoots remained approximately the same throughout the crown. In contrast, the decrease in needle area with depth in the crown was much steeper in fertilized trees, so that in the lowest whorls (10 and 13), the mean needle area per shoot was smaller in the fertilized trees than in the control trees. An increase in the ratio of shoot width to vertical radius was found in fertilized trees, but not in control trees (Figure 6). There was a negative correlation between SPAR and needle area, implying that SPAR was, on average, larger for shoots with small needle area than for shoots with large needle area (Figure 7). The correlation between SPAR and needle area was stronger for fertilized shoots (r = 0.76) than for control shoots (r = 0.54). Consequently, mean SPAR calculated as the average weighted by needle area was smaller (0.608) than the arithmetic (unweighted) mean (0.679) for fertilized shoots, whereas for the control shoots, the difference was less (0.505 versus 0.488). Needle area per shoot, as a single variable, did not explain the difference in SPAR between control and fertilized shoots. The dependency differed between the fertilized and control trees so that, for a given needle area, SPAR of fertilized shoots was higher than SPAR of control shoots (Figure 7). However, the density of needle area in the shoot envelope, expressed in terms of needle cover, was more closely linked with SPAR than needle area. There was an inverse linear relationship between SPAR and needle cover that explained most of the variation in SPAR (r 2 = 0.816) (Figure 8). The larger variation in SPAR with inclination and rotation angle in fertilized shoots compared with control shoots (Figures 3 and 5) was associated with differences in shoot geometry. Fertilized shoots were, on average, almost twice as long as control shoots (Table 2), which explained the larger variation with inclination angle in fertilized shoots than in control shoots (Figure 3a). On the other hand, the larger rotational variation of SPAR in fertilized shoots than in control shoots (Figure 3b) indicated a more flat (rotationally asymmetrical) shoot structure in fertilized shoots, which is reflected by a higher ratio of shoot width to vertical shoot radius and a larger needle-free angle (Table 2). Shoot asymmetry increased with depth in the crown (Figure 5), particularly in the fertilized trees where there was also an increase in the ratio of shoot width to vertical shoot radius (Figure 6). The increase in SPAR (SPAR, SPAR min and SPAR max ) (Figure 4) was explained by decreases in both needle area and needle cover of the shoots with depth in the crown. Because

5 VARIATION IN SPAR OF NORWAY SPRUCE 709 Figure 4. SPAR min, SPAR and SPAR max (mean and standard deviation) as a function of whorl number. Figure 5. Shoot asymmetry (variation in SPAR with rotation angle) in the upper crown (Whorl 4) and lower crown (Whorl 13). Bars represent standard deviations. the needle cover was smaller overall and was accompanied by a steeper decrease in needle area per shoot in the fertilized trees (Figure 6), the change in SPAR with depth in the crown was larger in fertilized trees than in control trees. In particular, the large values of SPAR max found in the lower crown of fertilized trees were associated with shoots of small needle area and small needle cover, combined with a large ratio of shoot width to vertical shoot radius. Discussion Effects of fertilization Changes in shoot structure as a result of fertilization caused SPAR of fertilized trees to be larger, on average, and to vary more with direction and location in the crown than SPAR of control trees. Known effects of fertilization in Norway spruce include increases in shoot and needle length, and number of needles and needle area per shoot (Flower-Ellis 1993, and see Table 2). However, because the increase in needle number was less than the effect of fertilization on shoot elongation, the number of needles per unit length of the twig was smaller in fertilized shoots than in control shoots. Thus, although shoots of the fertilized trees were, on average, larger in terms of size and needle area, they had smaller within-shoot shading and, consequently, a larger mean SPAR than control shoots (Figure 2). Because fertilization increases the total needle area of a tree, an indirect effect of fertilization is increased shading at a given position in the crown. The effect of shading on shoot structure is in some respects opposite to the effect of fertilization (e.g., shading causes a decrease in needle area and shoot size). However, both shading and fertilization resulted in an increase in SPAR. The key to understanding the connection between SPAR and shoot structure is that neither needle area nor shoot

6 710 STENBERG, LINDER AND SMOLANDER Figure 8. The dependency of SPAR on needle cover ( = fertilized shoots, = control shoots). Figure 6. Needle area, needle cover and ratio of shoot width to shoot radius (sw/sr) by whorls. Figure 7. SPAR plotted against needle area for fertilized ( ) and control ( ) shoots. size is important per se, but rather it is the ratio of these two variables that is important. In this study, needle cover was used as a measure of the density of needle area in the shoot and the consequent within-shoot shading. Needle cover decreased both as a direct effect of fertilization and in response to increased shading. In addition, shading resulted in a decrease in shoot size with depth in the canopy. The high values of SPAR, especially in the lower crown of fertilized trees, were thus associated with small needle area and needle cover, implying low within-shoot shading. Conversion between SPAR and STAR In this study, we reported values of SPAR and not STAR because only the projected needle areas of the shoots were measured. However, to assess the degree of within-shoot shading, SPAR should be converted to STAR. Because the spherically averaged ratio of silhouette area to total surface area for a single needle is 0.25, irrespective of its shape as long as it is convex (Lang 1991), this is the value of STAR for a shoot with no within-shoot shading (disregarding the contribution of the twig to the shoot silhouette area). In contrast, the maximum value of SPAR depends on needle shape, the definition of projected needle area, and the measurement technique. Accordingly, STAR is a more useful measure of within-shoot shading and light interception efficiency of a shoot than is SPAR. To convert from SPAR to STAR, we need an estimate of the conversion factor, i.e., the ratio of total surface area to projected needle area. The projected area of a needle refers to the silhouette area of a horizontally lying needle when viewed from the vertical. This is an unambiguous measure only for cylindrical or perfectly flat needles, in which case the conversion factor is π or 2, respectively. Consequently, the projected needle area is often defined as the silhouette area of the needle

7 VARIATION IN SPAR OF NORWAY SPRUCE 711 when placed with its widest cross section parallel to the plane of projection. Riederer et al. (1988) used this definition to determine a mean value of 2.74 for the conversion factor in Norway spruce. In this study, however, the position (rotation angle) of the needles was not fixed when measuring the silhouette area, because needles were allowed to lie horizontally in their natural position. If the silhouette area is averaged for a whole rotation about the longitudinal axis, the ratio of total surface area to projected needle area is π (Grace 1987). We therefore used π as a presumptive conversion factor. Estimated mean values of STAR were then for control trees and for fertilized trees, or and when weighted by the needle area. These are higher than the values reported for Scots pine (Pinus sylvestris L.) (Oker-Blom and Smolander 1988, Smolander et al. 1994); however, if the comparison is restricted to unfertilized shoots from the upper crown, the values are similar. Assessment of within-shoot shading It should be recognized that, in our definition of STAR, the shadow area of the shoot axis (the twig) is included in the silhouette (light intercepting) area, whereas only the needle area is included in the denominator of STAR. Consequently, the light interception efficiency of a shoot is potentially overestimated and the within-shoot shading is underestimated when calculated based on STAR. Ideally, the shoot silhouette area should represent the shadow area of needles only, but this would require a more advanced measuring technique. Another way of correcting the error would be to include the twig in the denominator of STAR. In a study of Scots pine, Smolander et al. (1994) obtained a difference of only 2.4% between mean STAR and the mean silhouette to total area ratio when they included the area of the twig in the denominator. A much larger proportion of twig to needle area was found for our Norway spruce shoots, with differences of 8% for control shoots and 10% for fertilized shoots. Assuming, for simplicity, that the twig increases the shoot silhouette area by 10%, the maximum value of STAR is instead of 0.25, and maximum SPAR (using π as conversion factor) would be Thus, on average, the values of for control shoots and for fertilized shoots found for SPAR would correspond to decreases of 42 and 21% in shoot silhouette area, respectively, as a result of mutual shading of needles on the shoot. In the lowest whorl of the fertilized trees, the mean value of SPAR was (6% within-shoot shading). This may, however, be an overestimate because for a few shoots in the lower crown of the fertilized trees, measured values of SPAR (and SPAR max ) were larger than is theoretically possible, even when considering the contribution of the twig. These shoots had a very small needle area (Figure 7), and the precision of the measurements of shoot silhouette area and projected needle area was apparently not sufficient. Variation in SPAR and its ecological significance The observed increase in SPAR with depth in the crown was accompanied by a smaller variation with inclination angle and a larger variation with rotation angle, reflecting the fact that shoots became shorter and flatter with increasing crown depth. As a result of increasing flatness, there was an increase, especially in SPAR max, of the more shaded shoots in the lower crown of the fertilized trees. It has been suggested that the large SPAR max (R max ) of shade-acclimated shoots is a means by which a coniferous stand can maintain a high productive leaf area index (Leverenz and Hinckley 1990). This is in accordance with the notion that an increase in SPAR implies more efficient light interception by shoots in the lower crown, where little light is available. However, the large directional variation in SPAR implies that shoot orientation and directional distribution of light at different levels in the canopy need to be considered to estimate the benefit of this phenomenon quantitatively. Specifically, a large SPAR max that was not connected with a large SPAR would be favorable only if shoots were oriented to face the direction of an assumed main light source. In contrast, a large SPAR is favorable irrespective of the directional distribution of incident light. We found a strong positive correlation between SPAR and SPAR max, i.e., shoots with high values of SPAR max also had a large SPAR. We therefore propose that the development toward a more flat shoot structure (implying a large SPAR max ) is one means to reduce overall within-shoot shading (i.e., to increase SPAR). Acknowledgments We thank Mr. Pekka Voipio for technical assistance, and Dr. Juha Lappi for computational advice and useful comments on the manuscript. References DeBoor, C A practical guide to splines. Springer-Verlag, New York, Heidelberg, Berlin, 392 p. Flower-Ellis, J.G.K Dry-matter allocation in Norway spruce branches: a demographic approach. Stud. For. Suec. 191: Flower-Ellis, J.G.K., A. Albrektson and L. Olsson Structure and growth of some young Scots pine stands. (1) Dimensional and numerical relationships. Swed. Conifer. Project Tech. Rep. 3, Swedish Univ. of Agric. Sci., Uppsala, Sweden, 98 p. Grace, J Theoretical ratio between one-sided and total surface area for pine needles. N.Z. J. For. Sci. 17: Lang, A.R.G Application of some of Cauchy s theorems to estimation of surface areas of leaves, needles and branches of plants, and light transmittance. Agric. For. Meteorol. 55: Leverenz, J.W. and T.M. Hinckley Shoot structure, leaf area index and productivity of evergreen conifer stands. Tree Physiol. 6: Linder, S Nutritional control of forest yield. In Nutrition in Trees. The Marcus Wallenberg Foundation Symp. Proc. 6: Oker-Blom, P. and H. Smolander The ratio of shoot silhouette area to total needle area in Scots pine. For. Sci. 34: Riederer, M., K. Kurbasik, R. Steinbrecher and A. Voss Surface areas, lengths and volumes of Picea abies (L.) Karst. needles: determination, biological variability and effect of environmental factors. Trees 2: Smolander, H., P. Stenberg and S. Linder Dependence of light interception efficiency of Scots pine shoots on structural parameters. Tree Physiol. 14:

8 712 STENBERG, LINDER AND SMOLANDER Stenberg, P., T. Kuuluvainen, S. Kellomäki, J.C. Grace, E.J. Jokela and H.L. Gholz. 1994a. Crown structure, light interception and productivity in pine trees and stands. In A Comparative Analysis of Pine Forest Productivity. Eds. H.L. Gholz, R. McMurtrie and S. Linder. Ecol. Bull. 43: Stenberg, P., S. Linder, H. Smolander and J. Flower-Ellis. 1994b. Performance of the LAI-2000 plant canopy analyzer in estimating leaf area index of some Scots pine stands. Tree Physiol. 14: Tucker, G.F., T.M. Hinckley, J.W. Leverenz and S. Jiang Adjustments of foliar morphology in the acclimation of understory Pacific silver fir following clearcutting. For. Ecol. Manage. 21: