BioEnergy Research

, Volume 1, Issue 3, pp 229–238

Assessment of Canopy Structure, Light Interception, and Light-use Efficiency of First Year Regrowth of Shrub Willow (Salix sp.)

  • Pradeep J. Tharakan
  • Timothy A. Volk
  • Christopher A. Nowak
  • Godfrey J. Ofezu

DOI: 10.1007/s12155-008-9023-9

Cite this article as:
Tharakan, P.J., Volk, T.A., Nowak, C.A. et al. Bioenerg. Res. (2008) 1: 229. doi:10.1007/s12155-008-9023-9


According to the light-use efficiency model, differential biomass production among willow varieties may be attributed either to differences in the amount of light intercepted, the efficiency with which the intercepted light is converted to aboveground biomass, or both. In this study, variation in aboveground biomass production (AGBP) was analyzed in relation to fraction of incoming radiation intercepted (IPARF) and light-use efficiency (LUE) for five willow varieties. The plants were grown in a short-rotation woody crop (SRWC) system and were in their first year of regrowth on a 5 year old root system. The study was conducted during a two-month period (June 15th–August 15th, 2001) when growing conditions were deemed most favorable. The objectives were: (1) to assess the relative importance of IPARF in explaining variation in AGBP, and (2) to identify the key drivers of variation in LUE from a suite of measured leaf and canopy-level traits. Aboveground biomass production varied nearly three-fold among genotypes (3.55–10.02 Mg ha−1), while LUE spanned a two-fold range (1.21–2.52 g MJ−1). At peak leaf area index (LAI), IPARF ranged from 66%–92%. Nonetheless, both IPARF and LUE contributed to AGBP. An additive model combining photosynthesis on leaf area basis (Aarea), leaf mass per unit area (LMA), and light extinction coefficient (k) produced the most compelling predictors of LUE. In a post-coppice willow crop, the ability to maximize IPARF and LUE early in the growing season is advantageous for maximizing biomass production.


Aboveground biomass Canopy structure Light interception Light-use efficiency Willow (Salix sp.) 



Light-saturated photosynthesis per unit leaf area μmol m−2 s−1


Aboveground biomass production Mg ha−1


Light-saturated photosynthesis per unit leaf mass nmol g−1 s−1


Leaf compensation irradiance


Fraction of incoming photosynthetically active radiation intercepted


Total incoming photosynthetically active radiation intercepted MJ m−2


Infrared gas analyzer


Light extinction coefficient for Beer’s law


Leaf area index


Leaf mass per unit area g m−2


Light-use efficiency g MJ−1


Leaf nitrogen concentration expressed on a per unit area g m−2


Leaf nitrogen concentration expressed on a per unit mass g kg−1


National Oceanic and Atmospheric Administration


Photosynthetically active radiation μmol m−2 s−1


Incident photosynthetically active radiation on top of the canopy μmol m−2 s−1


Photosynthetically active radiation below the canopy μmol m−2 s−1


Photon flux density μE m−2 s−1


Short-rotation woody crop


State University of New York, College of Environmental Science and Forestry


The existence of a strong relationship between light interception and crop growth is well established in agricultural crops [14, 39], short-rotation woody crops (SRWC) [5] and forest trees [27, 29]. This relationship is commonly analyzed using the light-use efficiency model [4, 30], which comprises two conceptual components: fraction of incoming photosynthetically active radiation (PAR) that is intercepted (IPARF) and the efficiency with which the intercepted radiation is converted into biomass, called light-use efficiency (LUE (ɛ); grams of biomass produced per megajoule of intercepted photosynthetically active radiation). Enhanced productivity of SRWC have been linked to early canopy development leading to high levels of light interception, and high photosynthetic capacity resulting in high carbon gain rates [6]. Previous studies have reported extensive variation in biomass production potential among willow varieties in Europe [21, 26] and North America [18]. From the light-use efficiency model, it can be theorized that the variation in AGBP among the different varieties may be related to differences in IPARF, LUE, or both.

There is some debate regarding the importance of LUE as a determinant of genotypic variation in biomass production. For example, Cannell et al. [5] found similar LUEs in a 1-year old poplar genotype and a 1-year old willow genotype, grown in containers under favorable conditions that varied nearly two-fold in biomass production. [19] and [1] have reported similar conclusions that under non-limiting growing conditions, LUE may be constant for a single species. When LUE refers to total dry mass production, it is likely to be a conservative value for C3 crops growing under similar temperature and radiation environments, especially on longer time scales. [1, 5, 19] and [5] have further reported that large improvements in biomass production might be realized mostly by focusing on differences in light interception (leaf area duration) and amount of biomass allocated to aboveground tissue, rather than on aspects related to LUE.

Others contend that the rates of canopy photosynthesis and possibly LUE can be varied by manipulating canopy structure [9, 37]. A comparison of two native and three-hybrid poplar genotypes in a high-density plantation in their third year of growth showed that LUE varied nearly two-fold [11]. Biomass production was closely related to LUE and was unrelated to IPARF, which only varied by 5% among the genotypes. Studies conducted on conifers have shown that inter-species and intra-species differences in biomass production were linked to variation in both IPARF and LUE [27, 28]. While studies of light-use efficiency in shrub willows are scarce, Bullard et al. [3] outlined some effects of varying planting densities on light-use efficiency in willows for three growing seasons. Under optimal conditions, change in planting density from 10,000–111,000 plants ha−1 had a significant effect on LUE of Salix viminalis (1.55–2.55 g MJ−1) and Salix × dasyclados (1.34–1.84 g MJ−1) respectively.

Thus, overall analyses of light-use dynamics in SRWC, and in willows in particular, are far from conclusive. We examined the relative importance of IPARF and LUE in explaining variability in biomass production, and identified the key drivers of variation in IPARF and LUE from the measured leaf- and canopy-level attributes. Analysis of this kind will provide information for selection of useful varieties and design of management practices that will facilitate high biomass production in competitive environments [20].

Intercepted radiation is regulated by the amount and orientation of foliage, and the duration during which the foliage is deployed [4, 19]. Given the variation reported in leaf area index (LAI) and leaf demography in willow [6], it can be hypothesized that significant variation in IPARF can be found among willow varieties. Variation in LUE between genotypes growing in a favorable environment, on the other hand, may be attributed to variation in leaf photosynthetic capacity, which is related to a combination of leaf structural and biochemical properties [4, 29, 35]. For example, differences in leaf nitrogen and leaf mass per unit area (LMA) are known to influence variation in leaf photosynthetic capacity within a single species and among different species [13, 35]. Both of these traits have been shown to vary among willow varieties [33]. Thus, efforts to understand LUE should include an assessment of attributes related to photosynthesis and foliage morphology and biochemistry. Canopy structure is a major feature that influences light distribution over the foliage surface within the canopy [6, 9, 34]. An “ideal” canopy would “optimize” light environment by distributing radiation throughout the canopy in such a way that all leaves were exposed to intermediate, nearly saturating quantum flux densities. The optimization of light environment is aided by corresponding variability in canopy structure. Light extinction coefficient (k), a measure of light attenuation within canopies, decreases as leaf angles increase. Low k in mid-and-upper canopy regions result in gradual light attenuation and deeper penetration of light, especially at the high leaf area indices observed in high-density SRWC plantations [9, 12]. This would maximize the rate of photosynthesis per unit of light intercepted by each leaf. Varieties can also differ in leaf size and shape and their arrangement on the branches, and stem and branch architecture- measured in terms of canopy width, number of stems, and branching patterns [42]. These morphological differences affect the energy and gas exchange processes and the development of large leaf areas.

The present study was conducted on five willow varieties known to have different aboveground biomass production and stem and foliar morphological characteristics. The objectives of this study were two-tiered: (1) to assess the relative importance of IPARF and LUE in explaining variation in aboveground woody biomass production (AGBP), and (2) to determine which of the select leaf and canopy-level constitutive traits were the key drivers of variation in IPARF and LUE. We hypothesized that variation in AGBP would be more closely related to IPARF than to LUE in this system with a rapidly developing canopy and that any LUE variation we did observe would be closely related to traits that maximize canopy photosynthesis, including those that govern intra-canopy light distribution.

Materials and Methods

Study Design and Site Conditions

The study was conducted at SUNY-ESF’s existing genetic selection trial [41] established at Tully, New York (42 47′ 30″ N, 76 07′ 30W). The soil was a well-drained Palmyra gravelly silt loam (Glossoboric Hapludalf) [17]. The trial, which included 32 shrub willow varieties and eight hybrid poplar genotypes, was established in late April 1997 on approximately 0.4 ha, as a randomized complete block design with four blocks. Individual plots were planted with 48 willow cuttings of 0.6 m by 0.9 m spacing. Cuttings were hand planted with 25 cm long dormant unrooted cuttings, flush with the soil surface. The plants were coppiced in the winter of 1997, harvested in 2000 at the end of the first rotation, and began their second rotation in the spring of 2001. Details of the site preparation, plot establishment and maintenance are presented in Tharakan et al. [41]. In late May 2001, the trial was fertilized with sulfur-coated urea at 120 kg N ha−1.

The study focused on measurements taken during the 2001 growing season on five shrub willow varieties (Table 1). These varieties had diverse morphological and growth attributes in the first rotation [41], and hence it was hypothesized that they would differ in aspects related to radiation capture and use. The study was conducted over the period when growing conditions were deemed to be most optimal (June 15th –August 15th). This time period was chosen to avoid the difficulty of separating variation in growth rate resulting from differences in growth duration related to differences in phenology (i.e. time of leaf flush and abscission) [11]. During this period, precipitation at the site (240 mm) was comparable to the 30 year mean, while growing degree-days (base 10°C) exceeded the 30 year average by 15% [31]. The study was conducted under near-optimal growing conditions to minimize genotype by environment interactions in biomass productivity and its determinants [4, 7].
Table 1

List of willow varieties used in the light-use efficiency (LUE) analysis





Salix purpurea



Salix purpurea

Ontario, Canada


Salix eriocephala (erio) 28×Salix eriocephala 24

Ontario, Canada


Salix dasyclados

Ontario, Canada


Salix sachalinensis


aDenotes the place where the collections or crosses were made rather than the geographical or botanical origin. For example, S. purpurea was imported from Europe in colonial times and has since been naturalized to Ontario, Canada and the Northeastern U.S.A

Aboveground Biomass Production (AGBP)

Allometric relationships relating stem diameter to stem dry weight, and stem diameter to foliage mass, were developed to estimate the net biomass gain of aboveground tissues. In early August, 30 stems representing the entire diameter range of each variety were selected from across the four replications and their diameters were recorded at a height of 5 cm from the ground prior to harvesting them. Post-harvest stem and foliage were separated and bagged. They were then dried at 65 °C to a constant weight. Variety-specific relations between stem biomass and diameter (y = axib+ei; r2 = 0.94–0.99) were subsequently used to estimate biomass gain for the center four stems in each plot. Similarly variety-specific relations between foliar biomass and diameter (y =axib+ei; r2 = 0.86–0.96) were used to estimate foliage biomass for the center four stems in each plot. Finally, measurements were averaged across the four individuals in each plot and converted to an area basis and summed to estimate AGBP (g m−2).

Canopy Structure, Light Interception, and Light Use-Efficiency

Leaf area index (LAI; projected leaf area and branches per unit ground area) was measured on a weekly basis during the study period using a LAI 2000 plant canopy analyzer (LI-COR Inc. Lincoln, NE, USA) [8]. A measurement cycle consisted of a reference measurement in a clearing, away from the canopy. This was followed by eight below-canopy readings and another reference measurement. The fish-eye lens of the instrument was covered with a view cap with a 45° opening to ensure that the measurements were not influenced by the surrounding plots or by the operator (LI-COR Inc. Lincoln, Nebraska). All measurements were taken at 0.5 m aboveground, either early in the morning or late afternoon to allow for totally diffuse light conditions, on cloudless or uniformly overcast days. To estimate canopy averages for leaf mass per unit area (LMA, g m−2), 15 undamaged leaves were selected and harvested from throughout the canopy in each plot in early August. Following leaf area estimation using a LI-COR 3100 leaf area machine [25], the harvested leaves were dried to constant weight at 65 °C. Leaf mass per unit area (LMA) was calculated as leaf dry mass/area. The leaf samples were then ground and total nitrogen (foliar N) was estimated by acid-base volumetry after Kjeldahl mineralization [2]. Nitrogen concentrations were calculated both on mass (Nmass, g kg−1) and area basis (Narea, g m−2).

In mid-July, canopy averages for leaf angle were estimated in all the plots based on measurements taken at increments of 1 m from the base to the top of the canopies using a protractor inclinometer [32]. Measurements for S566 were only taken at two canopy positions because the canopy depth was limited. The midrib angle in relation to the horizontal was measured on 10 leaves per increment [9]. Leaf angles were then averaged across height increments in each plot. Subsequently, canopy widths were measured in each increment by taking two measurements in the N-S and E-W direction with a meter stick and averaging them. Measurements were then averaged across height increments in each plot.

The total amount of photosynthetically active radiation intercepted by the canopy during the study period (IPART, MJ m−2) was estimated in each stand by using the equation:
$${\text{IPAR}}_{\text{T}} = \,{\text{Total}}\,{\text{PAR x}}\,{\text{IPAR}}_{\text{F}} $$
where, IPARF is the fraction of incident PAR intercepted by the canopy. Total PAR for the study period was measured on site (approximately 75 m from the trial) using a LI-190SA quantum sensor [23] that was attached to a data logger. Data was logged at 1 min intervals and averaged every 10 min. Fraction of incoming photosynthetic active radiation (IPARF) was calculated as:
$${\text{IPAR}}_{\text{F}} \, = \,1 - \left( {{{{\text{PAR}}_{\text{B}} } \mathord{\left/{\vphantom {{{\text{PAR}}_{\text{B}} } {{\text{PAR}}_{\text{A}} }}} \right.\kern-\nulldelimiterspace} {{\text{PAR}}_{\text{A}} }}} \right)$$
where, PARB was the PAR measured below the canopy and PARA was the incident PAR on top of the canopy. Both measurements were taken simultaneously at weekly intervals between 10.00–12.00 h on uniformly sunny or cloudy days. Photosynthetic active radiation above the canopy (PARA) was measured using a LI-190SA quantum sensor [23] that was positioned in a clearing away from all canopy interference. Photosynthetic active radiation below the canopy (PARB) was measured using a LI-191SA line quantum sensor that was calibrated to the LI-190SA quantum sensor. Measurements were taken at three points in each plot along a diagonal transect. At each sampling point, two measurements were taken by orienting the sensor first in the N–S direction and then along the E–W direction. All measurements were taken at approximately 10 cm from the ground. Weekly measurements of PARA and PARB were then used to calculate weekly values for IPARF and IPART. Weekly values of IPART were summed to yield IPART for the study period. The LUE (g MJ−1) of each plot was calculated by taking the quotient of AGBP and IPART for the study period. The canopy light extinction coefficient (k) for Beer’s law was estimated using the relationship between IPARF and LAI [30].
$${\text{IPAR}}_{\text{F}} \, = \,1 - e\,^{ - k\,x\,{\text{LAI}}} $$
where, k was determined from the regression of log (1- IPARF) against LAI. Canopy width and k were used as basis for comparisons of intracanopy light distribution.

Photosynthesis Measurements

Net photosynthesis was measured once in late July using a LI-COR 6200 photosynthetic system [24] equipped with a 0.25 L chamber. Readings were corrected for leaf area whenever the chamber was not completely filled. The infrared gas analyzer (IRGA) was calibrated daily and checked periodically throughout the day. All observations were made on days with bright, uniform conditions, between 9.30–13.00 h, to eliminate the possibility of any late-afternoon decline of photosynthetic capacity. Measurements were taken on two healthy, mature leaves, selected from the top third of the canopy of two plants per plot [11]. Leaf orientation was maintained while being enclosed in the chamber and all observations were made when the photon flux density (PFD) was above 800 μE m−2 s−1, which is saturating for many willow varieties [38]. We restricted measurements to ambient CO2 concentrations (330–360 μl l−1) and humidity levels (35–60%). Leaves were enclosed in the chamber for less than 90 s to prevent excessive rise in leaf temperature [24]. After the light-saturated photosynthetic rate (Aarea, μmol m−2 s−1) was measured, each leaf was collected, sealed in a bag and kept out of the sun until it was brought back to the lab to measure leaf area using a LI-COR 3100 leaf area meter [25]. The leaves were then dried to constant mass at 65 °C and weighed to determine LMA, which was used to calculate light-saturated photosynthesis per unit leaf mass (Amass, nmol g−1 s−1).

Statistical Analyses

Univariate analysis of variance (ANOVA), conducted using PROC GLM [36], was used to assess differences in AGBP, peak IPARF, LUE and all leaf and canopy traits among clones. Tukey’s mean studentized range test was used to determine significant mean separations among varieties at a critical level of ∝ = 0.05. Relationships among AGBP, peak IPARF, LUE and canopy traits were assessed by regression analysis conducted using Statistica version 6 [40]. Scatter plots of the variables and residual plots were used to identify suitable models. Except relationships between AGBP and IPARF, and between LAI and IPARF, which were curvilinear and hence described using a polynomial regression model, all other relationships were characterized using simple linear regression. The plots were the experimental units in all regressions. Step-wise multivariate regression was used to examine the relationships between LUE and all combinations of measured canopy traits.


Biomass Production, Light-use Efficiency, and Leaf and Canopy Traits

Estimates of accumulated AGBP during the two months varied nearly three-fold (3.55–10.02 Mg ha−1; Table 2). Among the varieties, Pur12 and SX61 had the highest AGBP while 94012 and S566 were grouped at the lower end (Table 2; Fig. 1). Salix dasyclados (SV1) had intermediate AGBP. Light-use efficiency during the same period spanned a two-fold range (1.21–2.52 g MJ−1; Table 2). Again, Pur12 and SX61 had the highest values. As expected in a rapidly developing canopy, LAI varied throughout the study period with SX61, SV1 and Pur12 having the highest LAI in mid August (Fig. 2). Peak LAI in mid-August ranged from 2.74–3.74 (Table 2). The proportion of radiation intercepted also varied from mid-June to mid-August (Fig. 2). At peak LAI, IPARF ranged from 66.9–92.4% (Table 2).
Fig. 1

Relationship between biomass production, IPARF (top) and LUE (bottom) for five willow varieties in their first growing season post-coppice

Fig. 2

Seasonal progression in leaf area index (top) and IPARF (bottom) for five willow varieties in their first growing season post-coppice

Table 2

Mean net gain in aboveground biomass production (AGBP), light-use efficiency (LUE), peak light interception (IPARF), and select foliage and canopy characteristics








AGBP (Mg ha−1)b






LUE (g MJ−1)


















LMA (g m−2)






Leaf angle (°)












Crown width (cm)






Amass (nmol g−1s−1)






Aarea (μmol m−2 s−1)






Nmass (g kg−1)






Narea (g m−2)






aAbbreviations are: IPARF-fraction of incoming photosynthetically active radiation (PAR) intercepted by the canopy at peak LAI; LAI-projected leaf area index on per unit area. The values presented here are peak values measured during the study; LMA-canopy average for leaf mass per unit area; Nmass-leaf nitrogen concentration expressed on a per unit mass; Leaf angle-angle of the leaf mid-rib from the horizontal; k-light extinction coefficient; Aarea-and Amass-light-saturated photosynthesis per unit leaf area and leaf mass in the upper canopy respectively

bMean values for a given variable followed by the same letter do not differ significantly at α = 0.05, according to Tukey’s studentized range test

Canopy averages for LMA were greatest in Pur12 (81.7 g m−2) and lowest in SV1 (69.1 g m−2) (Table 2). The differences in other canopy structural traits were also evident. Average leaf angle generally increased with canopy height (Fig. 3). Average leaf angle ranged from nearly planophile in SV1 (14°) and SX61 (−28°, drooping leaves), to more plagiophile in Pur12 (25°), S566 (38°) and 94012 (40°). Correspondingly, k varied from 0.68–0.95 (Table 2). Crown width also varied considerably, with 94012 and SX61 having the smallest value (64.17 cm), and the largest value (114.38 cm), respectively. Light-saturated photosynthesis both on leaf mass (Amass) and leaf area (Aarea) basis, showed similar patterns with highest value in Pur12 and lowest value in S566 (Table 2). Light-saturated photosynthesis ranged from 111.1–197.2 nmol g−1 s−1 by leaf mass and 8.3–14.9 μmol m−2 s−1 by leaf area. Canopy Nmass (20.90–32.30 g kg−1) and Narea (1.65–2.58 g m−2) also varied significantly among the varieties.
Fig. 3

Variation in leaf angle with height increment for five willow varieties

Determinants of AGBP, IPARF, and LUE

Across genotypes, AGBP was strongly and positively related to peak IPARF and LUE (Table 3; Fig. 1). Aboveground biomass production showed a quadratic relationship with IPARF, which in turn was positively related to LAI, crown width, and k. At >90% light interception, three varieties (SX61, Pur12, and SV1) exhibited increased aboveground biomass (>7 Mg ha−1). While the quadratic relationship with LAI and the linear relationship with k were strong, the linear relationship with crown width was weak, albeit significant. Regressing IPARF with combinations of the variables LAI, k, and crown width did not significantly enhance the model fit.
Table 3

Results of regression analyses between aboveground biomass production (AGBP), light interception (IPARF), light-use efficiency (LUE), and measured foliage and canopy traits (n = 20)

Variable ya

Variable x


























Crown width











































Crown width






aVariable definitions are the same as in the Table 2

bThe relationship between the variables were either linear (y=a+bx) or quadratic (y=a+bx+cx2)

Light use efficiency was positively related to photosynthesis expressed on both leaf area (Aarea) and leaf mass (Amass) basis, with the relationship with Aarea being the stronger of the two (Table 3). Light-use efficiency was also positively related to canopy averages of foliar nitrogen concentrations expressed on mass-basis. In terms of canopy characteristics, LUE was positively related to k, canopy averages of LMA, Aarea, Amass, and crown width. Variable selection indicated that LUE was most strongly related to the additive combination of Aarea, Amass, and LMA, which explained 86% of the variation (p < 0.01). All the variables showed a positive influence and the contribution of each of the variables was highly significant (p < 0.05; LUE=−4.286 + 0.316(Aarea) + 0.036(LMA)−0.013(Amass).


We assessed the effects of IPARF and LUE on the AGBP and several constitutive leaf- and canopy-level attributes. The findings reveal that IPARF is a strong determinant of AGBP in willow varieties in the first year of growth after harvest when the canopy is developing rapidly. High LUE was associated with high leaf photosynthetic potential (high Aarea and LMA), rather than with aspects related to intra-canopy light distribution (canopy width) and maximum light interception capacity (k).

Varietal Differences

The mean values for LUE measured in this study (1.21–2.52 g MJ−1) were similar to some published values obtained for willow. For instance Cannell et al. [5] obtained LUE value of 1.58 g MJ−1 for Salix sp. in their first year growth. Similarly LUE values for Salix viminalis and Salix × dasyclados ranged from 1.55–2.55 g MJ−1 and 1.34–1.84 g MJ−1 respectively [3]. The large variability in LUE observed mirrors the large variation reported for Populus spp [11] and for willow [3]. Cannell et al. [5] have suggested that significant differences in LUE and hence biomass allocation can be attributed to differences in carbon allocation. Since belowground biomass accumulation was not assessed in this study, it is uncertain to what extent the variation in LUE was attributed to differences in carbon allocation between aboveground and belowground components. Given the non-stressed growing conditions (i.e., fertilization, average precipitation, and near optimal temperatures), it can be presumed that the genotypes approached their full productive potential and adopted favorable allocation patterns [11, 20]. Foliar nitrogen concentration (Nmass) measured in this study (20.90–32.30 mg g−1) was similar to the normal to optimal range reported in Kopinga and van der Burg [16]. Thus, the observed behavior of these varieties may be typical for high-density plantings on productive sites, and the differences in LUE among these varieties may indicate inherent genetic variation arising from differential adaptation to such conditions.

Determinants of AGBP, IPARF, and LUE

The amount of radiation intercepted was most strongly related to the amount of foliage (LAI), but also was related to k, and crown width. In high-density willow plantings, the influence of canopy architecture on the mean fractional canopy interception is usually much less important than the LAI [6]. An additive model combining Aarea and LMA produced the best predictors of LUE, supporting the hypothesis that a suite of photosynthesis and light distribution traits would be closely related to LUE. Leaf mass per unit area relates to leaf thickness and density [10]. Leaves with high LMA often develop in high light and are associated with high mesophyll cell density. Studies have shown that sunlit canopy leaves tend to maximize photosynthetic capacity by combining high LMA with high Nmass to maximize Narea (Narea = Nmass * LMA), which is closely correlated with Aarea in leaves within plant canopies [10, 13]. Given the positive association between LUE and both Aarea and LMA in these willow varieties, it can be said that LUE was enhanced by intrinsically high photosynthetic rates combined with leaf structural traits that were adaptive under “high light” conditions.

The positive relation between k and LUE seen in this study is contrary to the negative relationship that has been observed in high-density plantations of other tree species and agricultural crops [11, 22, 29]. In dense canopies with high LAI, crops with erect leaves (high leaf angles and low k values) are considered to have a considerable yield advantage over those with horizontal leaves [9, 11]. Steep leaf angles in the uppermost canopy leaves followed by a gradual decrease in leaf angles along the vertical profile facilitates gradual light attenuation and better light distribution within the canopy, thereby maximizing canopy carbon gain [13]. However, orientation of foliage assumes less importance in sparse canopies that result from small-sized leaves or low LAI.

A coppiced canopy such as the one in this study can be considered growing in an open, “high light” condition with limited self-shading [11]. The potential for self-shading is further limited in a developing willow stand owing to small individual leaf size [41], the relatively low leaf area on each of the multiple stems per willow plant, and the relatively large crown width resulting from its multiple stem habits. Under such conditions, deeper penetration of light is achieved within the canopy even at the high k values seen in this study (0.5–0.7). In fact, high k at low levels of canopy competition maximizes light interception and thus the potential for energy harvesting [11, 13]. An analysis of the spatial variation in LMA provides some evidence of the “high light” environment prevailing in these willow canopies. In contrast to the spatial pattern of decreasing LMA from top to bottom in canopies characterized by strong light gradients [10] the varieties in this study showed very little variation in LMA through the depth of the canopy). The high positive association between LUE and LMA (Table 3), which is considered adaptive only under “high light” conditions further supports this explanation.

In a post-coppice context, the ability to rapidly maximize both IPARF and LUE in shrub willow is highly advantageous for maximizing biomass production. While this finding is contrary to the suggestions by Cannell et al. [5], it is in line with other studies, which reported similar variation in LUE among varieties and species [3, 11, 15]. However, it must be emphasized that the LAI were only moderately high, and severe competition had not developed. In a highly competitive environment, the relative importance of IPARF and LUE may shift in the direction of LUE as has been reported in older poplar plantations [11]. It should be noted that the gas exchange measurements in this study represent a point in time and may not be indicative of the entire growing season, which could vary due to seasonal variation in incident radiation. Nonetheless, the significant variation in photosynthetic rates seen among the clones is noteworthy. Even small differences in leaf-level photosynthesis of willow varieties when integrated over the entire canopy over time, can result in substantial differences in seasonal carbon gain [9].

This study corroborates the theory that at low levels of intra-canopy competition, higher light interception capacity, rather than intra-canopy light distributions are more important for maximizing LUE [22, 34]. Further research is needed to determine whether this trend is specific to willow due to its aforementioned canopy and foliar characteristics, or whether this is an artifact of the light environment in a young canopy.

Implications for Willow Breeding and Management

The variability in traits may provide useful information for breeding, selection and management. From our analysis, a large amount of photosynthetically active radiation was intercepted in mid-July at an optimal leaf area index for each variety (Fig. 2). Above this optimal point, further increase in leaf area index does not noticeably increase the energy intercepted. In addition, maximum radiation interception was significantly correlated with maximum biomass accumulation. Optimization of LAI would therefore increase light interception but further increase in LAI will negatively contribute to light interception. The analysis further demonstrates that breeding varieties with larger leaf surface area and leaf mass per unit area would maximize LUE. In these cases, the importance of maintaining large leaf area must be associated with an efficient utilization of the intercepted radiation, rather than the maximization of the interception of the radiation. High leaf area or light interception is not necessarily a good indicator of a plant’s ultimate potential following canopy closure. Traits that maximize the photosynthetic rates must balance losses from respiration.

While the results presented here highlight the adaptive value of a specific combination of traits, it must be borne in mind that alternative combinations of foliage amount, orientation, and aspects related to gas exchange may be adaptive to other environments and cultural conditions. Light-use efficiency is very sensitive to environmental variables that affect photosynthesis and balance with respiration (such as light, temperature, soil-water and humidity) [4]. In areas that have frequent moisture stress, high leaf area and high k during the growing season may be less advantageous since the condition enhances energy loading resulting into excess leaf temperature, water loss, respiratory loss, and photo-inhibition, all of which adversely affect LUE [19]. In such environments, the combination of high Nmass and LMA, and low k are more successful [9, 42]. In addition willow varieties with small leaves and numerous upright branches such as the S. purpurea have advantages in such environments. Smaller leaves are more efficient at regulating heat balance through convective cooling than larger leaves, which depend more on transpirative cooling [42].

A high LUE can be attained at appropriate planting density-the density that combines efficient capture of radiant energy early in the growing season while minimizing intra-specific competition [3]. Increased planting density leads to low light levels in the lower canopy. Thus, a high LAI early in the growing season, foliage orientation and other canopy architectural characters may assume a more important role, as has been seen in poplar [11]. A high k may not confer any yield advantage after canopy closure and may instead promote self-shading. Thus, a willow variety such as 94012 that appears to be maladapted to an open, “high light” growing environment due to its low k values and relatively low LAI, may perform relatively well under denser canopy conditions. As the amount of radiation intercepted per unit of leaf area decreases, a reduction in the LMA serves to maximize the radiation intercepted per unit of foliage mass or nitrogen, and thus the efficiency of nitrogen and light utilization. This may help offset the effect of increased self-shading [10, 43]. Additionally, in dense canopies, the degree of acclimatization of photosynthetic metabolism to light availability in the canopy interior (Ic, leaf compensation irradiance) will become more important [11].

Over the rotation period, as canopy density increases and variable stress environments are experienced, the ability to maximize LUE is a function of the extent of plasticity that might exist in the traits discussed above. Will et al. [43] reported that in Populus taeda and Populus elliottii plantations, LMA decreased with stocking density, and this modification in leaf morphology was the main mechanism by which trees in the different density treatments improved canopy light interception. The degree of plasticity exhibited will determine the optimal planting density range, the ultimate rotation length, and also the kind of sites on which a particular clone can be expected to perform well. For instance, if traits such as leaf angle and LMA are relatively fixed within a clone, then the associated optimal growing conditions, including density range, may be quite narrow for each clone. Conversely, if key traits are more plastic in some clones, then their planting density range, rotation length, and planting sites range would correspondingly be greater [11]. Information on trait plasticity in shrub willow is meager and inconclusive. Bullard et al. [3] reported a significant increase in LUE with increased planting density in two willow varieties. However, large sample variation precluded them from reporting any significant trend in corresponding k values.


Variation in biomass productivity among willow varieties in a post-coppice context was related to both the amount of light intercepted and the efficiency with which the intercepted light was converted to dry matter. Light-use efficiency, in turn, was most strongly related to “high light” interception capacity coupled with high intrinsic photosynthetic potential and foliar structure that was conducive to maximum carbon gain in “high light” conditions. The relative importance of these traits may vary under alternative environmental conditions, density, and competition regimes. Future studies with shrub willows should aim to examine genotype trait plasticity and the resulting LUE across diverse growing environments, stocking densities, and over a complete rotation.


The authors are grateful to the Biomass Feedstock Development Program of the US Department of Energy under contract DE–AC05–00OR22725 with the University of Tennessee-Battelle LLC (Subcontract number 19X–SW561C), USDA CSREES, and the New York State Energy and Research Development Authority (NYSERDA) for funding this study. Special appreciation is extended to C. Dattler and A. Millar for assistance in the field and the laboratory.

Copyright information

© Springer Science+Business Media, LLC. 2008

Authors and Affiliations

  • Pradeep J. Tharakan
    • 1
  • Timothy A. Volk
    • 2
    • 3
  • Christopher A. Nowak
    • 2
  • Godfrey J. Ofezu
    • 2
  1. 1.International Resources GroupWashington DCUSA
  2. 2.Dept. of Forest and Natural Resources ManagementState University of New York, College of Environmental Science and Forestry (SUNY-ESF)SyracuseUSA
  3. 3.Dept. of Forest and Natural Resources ManagementSUNY-ESFSyracuseUSA

Personalised recommendations