Introduction

Developing high quality seedlings for reclamation and restoration activities is an important component of the overall forest reclamation and restoration process (Hobbs and Norton 1996; Puttonen 1996; Brown and Amacher 1999; Suding 2011). This is particularly true when confronted with novel landscapes and soils that have been heavily altered through resource extraction or other disturbances. In such cases, the use of high quality seedlings in the reclamation process can often increase seedling survival and may promote more rapid forest establishment (Colombo 2003). In addition, the use of high quality seedlings can mitigate costs throughout the reclamation process. For example, the use of high quality seedlings can reduce the initial planting densities necessary for a given site and can later limit replanting costs associated with insufficiently meeting performance standards on seedling survival for a reclaimed site (Sullivan and Amacher 2013). Thus, identifying nursery practices which increase seedling quality is increasingly important in the overall reclamation process (Davis and Jacobs 2005).

Several environmental factors affect seedling growth and development during nursery production. Of particular importance to seedling nursery production is the light environment, which includes the photoperiod or day length, along with the quantity and quality (i.e. light spectrum) of light. These factors impact seedling growth through direct effects on photosynthesis and developmental processes. For example, seedlings grown under a shorter day length (photoperiod) typically display a delay in the time to fruiting or flowering (Jackson 2009). Similarly, seedlings grown under shade (reduced light quantity) generally display slower overall growth than comparable seedlings grown under full sun conditions (Björkman et al. 1972; Langenheim et al. 1984; Givnish 1988). In respect to light quality, the most commonly observed effect on tree seedlings is as an increase in height growth in response to a low red/far-red (R:FR) ratio (e.g., Morgan et al. 1983; Warrington et al. 1989; Ritchie 1997; Jacobs and Steinbeck 2001; Aphalo and Lehto 2001; Olsen and Juntilla 2002; Devine et al. 2007). Seedlings grown in low blue light environments or enriched in the green region of the light spectrum also commonly exhibit shade avoidance characteristics such as increased height growth, and typically exhibit a reduced investment in organs such as roots or leaf blades (Carabelli et al. 2007; Zhang et al. 2011; Pierik and de Wit 2014). In contrast, in the blue and red regions of the light spectrum, the effect of light quality on seedlings operates more towards enhanced rates of photosynthesis and greater gains in biomass (Goins et al. 1997; Lian et al. 2002; Poudel et al. 2008; Johkan et al. 2010; Wu and Lin 2012). From a nursery management perspective, the application of the blue or red region of the light spectrum during early growth may improve seedling quality by creating larger overall seedlings with greater reserves without diverting resources to height growth as occurs in the low R:FR and white light region of the light spectrum (e.g., Poudel et al. 2008; Johkan et al. 2010; Lin et al. 2013). Further, by manipulating any one of the remaining components of the light environment there may be potential to achieve desired outcomes for increased seedling quality and greater outplanting success.

Few studies have focused on the effects of the light quality, quantity, and growth environment on seedling development of boreal forest species, and in particular, how these environmental factors can influence desired characteristics that improve outplanting success in boreal environments. There is growing interest in the use of trembling aspen (Populus tremuloides Michx.) in boreal forest reclamation, as it is a fast growing early successional species with a known ability to establish after disturbance (Lieffers 2001). For aspen, recent research suggests that the seedling characteristics which best promote outplanting success include a high root-to-shoot ratio (R:S) and a high root total non-structural carbohydrate (NSC) concentration (Martens et al. 2007; Landhäusser et al. 2012a, b). These two characteristics appear to provide a benefit to aspen seedlings, particularly when grown in harsh reclamation environments, by providing seedlings with greater access to water and nutrients (high R:S) and greater reserves (high root NSC) (Schott et al. 2013). However, there is currently limited information surrounding the nursery management practices which best promote these characteristics in aspen seedlings (Landhäusser et al. 2012a, b).

A potentially effective method for producing aspen seedlings with a high R:S and high root NSC may entail predisposing them to stress (Landhäusser et al. 2012b). During periods of stress it is commonly found that plant growth decreases more than photosynthesis (Chaves et al. 2009). This often leads to a greater accumulation of NSC and greater carbon allocation to roots under environmental stresses than when conditions are optimal (Villar-Salvador et al. 1999; Landhäusser et al. 2012b). There is also strong indication that mechanical stimuli, such as wind, may lead to a greater allocation of carbon to roots and root reserves in tree seedlings exposed to, rather than sheltered from, the wind (Coutand et al. 2008). Hence, growth of aspen seedlings in the open, with common environmental stressors, may facilitate greater resource allocation to roots and root reserves than seedlings grown in protected and sheltered environments. However, most seedling nursery production occurs within controlled environments, such as a greenhouse (particularly in northern, short growing season environments), which seek to limit environmental stressors by providing more control over light, wind, temperature, and relative humidity. In open conditions, environmental stresses on plants are commonly exaggerated due to the combined increases in light intensity, wind and temperature (Niinemets et al. 1999). It is primarily the magnitude of these combined increases, and their associated effect on plant physiological processes, that directs carbon allocation towards reserve accumulation or growth and development. Thus identification of the key environmental cues or stressors (i.e. light, water, wind and nutrients) for aspen seedlings could help in devising an easily applied method of environmental stress conducive to producing high quality seedlings for use in harsh reclamation sites.

Our main objective in this study was to identify the growth environment which best promotes a high R:S and a high root NSC in aspen seedlings. To do so, we grew aspen seedlings in the field under one of six light treatments [unsheltered, full-sun (100 % full-sun); sheltered, full-sun (~90 %); sheltered, shaded (~40 %); sheltered, shaded blue (~40 %); sheltered, shaded red (~40 %); and sheltered, shaded low red/far-red ratio (R:FR) (~40 %)]. These growth light environments were meant to represent outside (unsheltered, full-sun) conditions or greenhouse and sheltered conditions (sheltered, full-sun; and sheltered, shaded: full spectrum and under a range of light qualities). The range of light qualities was selected to represent preferred light conditions for photosynthesis (blue and red light) and to compare these preferred conditions with those found in the understory (low R:FR) and the white (full-spectrum) portion of the light spectrum. Our hypotheses include that: seedlings developed under full-sun and unsheltered conditions will display a higher R:S and higher root NSC than seedlings grown under both sheltered, full-sun and sheltered, shaded conditions; seedlings developed under sheltered, full-sun conditions will exhibit a higher R:S and higher root NSC than seedlings grown under sheltered, shaded conditions; and further, that shaded seedlings grown under blue and red light will display greater biomass, a higher R:S and a higher root NSC than seedlings grown under the white or low R:FR portion of the light spectrum.

Material and methods

Plant material

In the late spring of 2012, 60 trembling aspen (Populus tremuloides) seedlings were grown at the Crop Diversification Centre (CDC North), Edmonton, Alberta, Canada (53°64′N, 113°37′W) from seed directly in 4 L pots (23 cm diameter by 10 cm height). Aspen seed was collected from a wide range of open pollinated aspen clones of the boreal mixed wood region of west Central Alberta near Edmonton, AB, Canada. Seedlings were sown in April 2012 in a mixture of peat, vermiculite and clay (2:1:1, by volume) and were established and initially grown under greenhouse conditions at the CDC North for 7 weeks. After the initial 7 week period (~June 2012), seedlings were moved from greenhouses to one of six light and sheltered or unsheltered treatments (unsheltered: full-sun and sheltered: full-sun, shaded, shaded blue and shaded red light, and shaded low R:FR) constructed in a nearby field. In both the greenhouse and field portions, seedlings were watered to field capacity every 2–3 days and fertilized biweekly with a 15-30-15 [N-P2O5-K2O] plus micronutrient solution (Agrium Advanced Technologies, Calgary, AB, Canada) at 2 g L−1.

Experimental setup

The study was setup in a randomized complete-block design with 10 blocks. Each block consisted of 6 seedlings, with each seedling corresponding to one of six light and sheltered or unsheltered treatments in each block. To construct each unique light treatment we covered wire cages (30-cm diameter by 120-cm height) with a combination of color filters and/or shade cloth. Each shelter environment was constructed with two ventilation holes to allow air to flow freely; one at the base just above the surface of each pot and one at the top of each color filter. The color filters (Rosco Laboratories Inc., Stamford, CT, USA) used are deep dye-coated polyester films, approximately 50–75 microns in thickness, which work by blocking passage of some portion of the visible spectrum (Fig. 1). For example, a primarily blue color filter will allow blue-associated light frequencies to pass, and block all others. The shade cloth used included two types: the first a black polyethylene fabric, purchased from a local fabric store, with a light reduction factor of approximately 30 %; the second a black polyethylene knitted fabric (Shade Rite, Green-Tek Inc., Janesville, WI, USA) with a light reduction factor of approximately 40 %. The blue light color filter naturally excluded ~60 % of the quantity of ambient light so we set up all other light treatments using color filters and shade cloths to match this reduction in light quantity. As a result, the red light color filter, which filtered ~10 % of incoming radiation, and the shaded (transparent color filter) treatment were matched with 30 and 40 % shade cloth, respectively, to correspond to the light quantity of the blue color filter. Similarly, the low R:FR, which also filtered ~10 % of incoming radiation, was matched with a 30 % shade cloth to correspond with the blue color filter. To evaluate the effect of sheltering on seedling growth and physiology, unsheltered seedlings were grown in wire cages without a cover (unsheltered full-sun) and with a transparent filter cover (sheltered full-sun). In both full-sun treatments (unsheltered and sheltered) the light quantity was equivalent to ambient light conditions (≥90 %; Fig. 1).

Fig. 1
figure 1

Light spectrum of experimental light treatments normalized to a clear sky reading of 1500 µmol s−1 m−2 (i.e. unsheltered conditions). Note the sum of all PAR is ~equal among the shade and color filter treatments (blue light, shaded, red light and low R:FR). (Color figure online)

Full size image

Evaluation of microclimatic conditions within shelters

To determine the effect of sheltering on air temperature and relative humidity, we installed data loggers (HOBO U23-001; Onset Computer Corporation, USA) inside four randomly selected shelters in each sheltering treatment and four data loggers in the open unsheltered treatment. In each treatment data loggers were attached to wire cages at a height between the two ventilation holes of each shelter and at a comparable height in the case of the unsheltered treatment. During the study period we measured the average daytime (0700–1800 h) air temperature (Tair) and leaf to air vapor pressure deficit (VPD). VPD was calculated from simultaneous air temperature and relative humidity readings.

Gas exchange

Net photosynthetic rates (Anet), stomatal conductance (g s) and leaf fluorescence (ΦPSII) at ambient light conditions for each light treatment were measured on fully expanded leaves with an open-flow infrared gas analyzer equipped with a red-blue light source (LI-6400XT, Li-Cor, Lincoln NE, USA). Measurements were conducted near the end of the growing season in late August/early September 2012. Leaf cuvette conditions during measurements were maintained at the light level corresponding to ambient conditions (~1400 µmol m−2 s−1 for full-sun conditions and ~600 µmol m−2 s−1 for shaded conditions), temperature of 25 °C, CO2 concentration of 400 µmol mol−1 and relative humidity >60 %.

Biomass allocation, and chlorophyll and NSC content

Seedlings were destructively harvested in October 2012. Just prior to harvest, stems were measured for height, basal diameter and terminal bud length and diameter; leaves were collected for leaf area estimates; and roots were measured for volume. At harvest, plant parts were further separated into leaves, stems, and coarse (>2 mm) and fine (<2 mm) roots, and then oven-dried for 48 h at 70 °C, after which component parts were weighed with an electronic balance. Specific leaf area (SLA: cm2 g−1) was calculated from leaf area estimates and leaf dry mass. Chlorophyll content was determined in leaves measured for gas exchange in late August/early September 2012 following the methods of Hiscox and Israelstam (1979). This method entailed punching leaf discs (each ~1 cm in diameter) out of leaves measured for gas exchange and then placing leaf discs in a vial containing 5 mL of dimethyl sulphoxide (DMSO). A 3 mL sample of this chlorophyll extract was then transferred to a cuvette, where the absorbances at 665, 649 and 480 nm were read against a DMSO blank. Once all biomass was oven-dried, stem and root (coarse and fine roots) tissue was collected for estimates of stem, and root, sugar and starch concentrations. Determination of tissue sugar and starch concentration, and analysis, from each harvest were conducted following the methods of Chow and Landhäusser (2004).

Statistical analysis

Statistical analyses were performed with R 3.0.1 (R Core Development Team 2013). In all analyses, statistical significance was assessed using α = 0.05. The effect of light treatments on microclimatic conditions, and morphological, physiological and biochemical variables was evaluated by either a t test or one-way ANOVA. Three separate analyses were performed, as the experiment was setup to test three separate effects (i.e., sheltering, light quantity and light quality). Two of the effects (sheltering and light quantity) were tested using a t test, as each individual effect included only two levels. Conversely, light quality with four levels was evaluated initially by one-way ANOVA, and when the main effect was statistically significant further testing was done with Fisher’s least significant difference (LSD) tests to identify differences among the light quality treatments. All data were tested for normality and log transformed when necessary.

Results

Shelter effects on microclimate

Mean ambient daytime air temperature (i.e. unsheltered treatment) over the study period averaged 23.0 °C, while mean ambient VPD averaged 1.46 kPa (Table 1). Conversely, mean daytime temperature of all sheltering treatments was 4.0 °C greater than the unsheltered treatment, while mean VPD in the sheltering treatments was 0.93 kPa greater than the unsheltered treatment. Between treatments this resulted in a significant difference between air temperature and VPD of unsheltered versus sheltered treatments (both p < 0.001; Table 1). In addition, VPD was 13 % greater in the sheltered, full-sun than the sheltered, shaded treatment (p = 0.031; Table 1). However, mean daytime temperature over the study period was similar among all sheltering treatments, and VPD was similar among the remaining sheltering treatments over the study period, with differences in temperature of less than 0.6 °C and differences in VPD of less than 0.13 kPa.

Table 1 Mean and standard error, plus maximum and minimum, of the daytime (0700 to 1800) microclimate (air temperature (Tair) and leaf to air vapor pressure deficit (VPD)) of Populus tremuloides within the unsheltered and sheltered (sheltered full-sun, shaded, blue light, red light and low R:FR) treatments. Different letters indicate significant differences among treatments. Lower-case bold letters (a, b) represent sheltered treatment comparisons; upper-case bold letters (A, B) represent shaded treatment comparisons; and lower-case bold (x, y, z) letters represent light quality treatment comparisons
Full size table

Sheltering effect

Sheltering had an effect on most shoot growth measures (all p < 0.05). Seedling shoot growth in the sheltered conditions was consistently greater in respect to total height, stem dry weight, leaf area, leaf dry weight and SLA compared with unsheltered seedlings (Table 2); although, there was no effect of sheltering on basal diameter (p = 0.131). For root measures, there were minimal sheltering effects. Seedlings grown without a shelter exhibited somewhat larger increases in all root growth measures (Table 3); although, the only statistically significant effect was an increase in coarse root dry weight (p = 0.046) for unsheltered compared to sheltered seedlings. This increase in coarse root dry weight for unsheltered seedlings, along with the smaller increase in fine root dry weight (Table 3), strongly contributed to an increase of ~48 % in the R:S for unsheltered compared to sheltered seedlings (p < 0.01; Table 3; Fig. 3).

Table 2 Mean and standard error of shoot characteristics, stem and leaf biochemistry, and physiological responses of Populus tremuloides under experimental light treatments (sheltered, shaded or light quality). Different letters indicate significant differences among treatments. Lower-case bold letters (a, b) represent sheltered treatment comparisons; upper-case bold letters (A, B) represent shaded treatment comparisons; and lower-case bold (x, y, z) letters represent light quality treatment comparisons
Full size table
Table 3 Mean and standard error of root characteristics and root biochemistry of Populus tremuloides under experimental light treatments (sheltered, shaded or light quality). Different letters indicate significant differences among treatments. Lower-case bold letters (a, b) represent sheltered treatment comparisons; upper-case bold letters (A, B) represent shaded treatment comparisons; and lower-case bold (x, y, z) letters represent light quality treatment comparisons
Full size table

At the biochemical level, statistically significant effects of sheltering were observed in both shoot and root components. At the stem level, total NSC and sugar concentrations were higher for the unsheltered rather than the sheltered seedlings (both p < 0.05; Table 2); although, there was no significant difference in any content parameters (total NSC, sugar or starch content) due to the 22 % larger stem mass of the sheltered compared to the unsheltered seedlings (Table 2). Conversely, unsheltered seedlings which exhibited a larger overall root mass also exhibited a larger increase in both root total NSC, and sugar, content and concentration (Table 3; Fig. 3). Root starch concentration (and content) were also higher for unsheltered seedlings, although not statistically significant at α = 0.05 (Table 3). At the leaf level, the amount of chlorophyll (a and b) per unit leaf dry weight was greater in sheltered compared with unsheltered seedlings (both p < 0.01; Table 2); although, there was no effect of sheltering on the ratio of chlorophyll a to b (p = 0.081). At the leaf physiological level, there was only an effect (p = 0.047) of sheltering on stomatal conductance (g s) with sheltered seedlings exhibiting a 25 % higher g s than unsheltered seedlings (Table 2). Lastly at the leaf level, there was no significant effect of sheltering on photosynthesis (A) or the quantum yield of photosystem II (ΦPSII) (Table 2).

Shade effect

In this analysis, sheltered, full-sun seedlings were compared to sheltered seedlings grown under a 60 % reduction in light quantity (shaded). There were few effects attributed to shade on shoot or root growth measures. For shoot growth measures, terminal bud diameter (p = 0.047) and basal diameter (p = 0.026) were greater for seedlings grown in the sheltered, full-sun compared to the shaded treatment (Table 2). In contrast, SLA was 35 % higher in the shaded than the full-sun treatment (p < 0.001; Table 2). In respect to roots, coarse root dry weight for seedlings grown under full-light was higher compared to the shaded treatment (p = 0.041). In all other shoot and root growth parameters, there were minimal observable differences between the sheltered full-sun and shaded treatment (Tables 2, 3).

In respect to leaf and whole plant biochemistry, there were few measured differences between the sheltered, shaded and sheltered, full-sun treatment. At the leaf level, chlorophyll (a and b) content per unit leaf dry weight was higher (both p < 0.01; Table 2) for shaded seedlings than under the full-sun treatment, but there was no effect of shade on the ratio of chlorophyll a to b (p = 0.311). However, this increase in chlorophyll (a and b) did not lead to an increase in photosynthetic rates for the shaded seedlings. Rather, the sheltered, full-sun seedlings exhibited a 40 % higher photosynthetic rate than the sheltered, shaded seedlings (Table 2); although, the effect of full-sun also led to a 36 % reduction in the quantum yield of photosystem II (Table 2). Lastly, there were no effects or discernible patterns between the two treatments in measured NSC components (content or concentration) at the stem or root level (Tables 2, 3).

Light quality effect

There was a minimal effect of light quality on both shoot and root growth measures. In fact, for root measures, there were no significant effects attributed to a change in light quality (Table 3). For shoot growth measures, there was only a significant effect of light quality on height growth (p = 0.003; Table 2). The greatest height growth occurred in seedlings grown under the low red/far red ratio (R:FR) and blue light, followed by seedlings grown under the red-light portion of the spectrum (Table 2). The least height growth occurred in the shaded, full-light spectrum (Table 2).

Light quality did not affect NSC measures at the stem or root level (content or concentration; Tables 2, 3). For leaf biochemistry, there was only a significant increase in chlorophyll a content per unit leaf dry weight (Table 2). However, this did not lead to an increase in the ratio of chlorophyll a to b from seedlings grown under the contrasting light qualities (Table 2). Similarly, light quality did not translate into statistically significant differences in any leaf-level physiological measures (Table 2).

Discussion

Manipulation of growth and storage in seedling stock is an ongoing process of trial and error in forest nursery management (Davis and Jacobs 2005). For example, altering the light environment is commonly used as a means to facilitate carbon transfer (either to growth or storage) in nursery practices for crop and ornamental species (e.g., Runkle and Heins 2003; Pires et al. 2013; Cope and Bugbee 2013; Olle and Viršilė 2013), whereas the applicability of this practice to tree species is less well understood. Partly this relates to identifying the attributes most closely associated with seedling quality in a particular setting. From a reclamation perspective, the key attributes for early survival and growth of planted aspen tree seedlings include large storage reserves and a high R:S (Martens et al. 2009; Landhäusser et al. 2012a, b). In the experiment presented here with trembling aspen (P. tremuloides), we found minimal differences in carbon partitioning between growth or storage for seedlings grown under light differing in both quantity (full-sun or shade) and quality (full spectrum, blue and red light or low R:FR). Rather, the largest differences in our experiment in growth and storage parameters occurred with the effect of sheltering (Tables 2, 3; Figs. 2, 3).

Fig. 2
figure 2

Box and Whisker plot of root-to-shoot ratio (R:S) of P. tremuloides seedlings grown under experimental light treatments (sheltered, shaded or light quality treatment). Box plots show median values (solid horizontal lines), lower (25th) and upper (75th) quartiles (box outline), minimum and maximum values excluding outliers (whiskers), and outlier values (open circles). Asterisk represents a significant difference for a given light treatment following a t test (shelter and light quantity treatment) or one-way ANOVA and least significant difference (LSD) post hoc testing (light quality treatment)

Full size image
Fig. 3
figure 3

Box and Whisker plot of root non-structural carbohydrate (NSC) concentration of P. tremuloides seedlings grown under experimental light treatments (sheltered, shaded or light quality treatment). Box and Whisker plots show median values (solid horizontal lines), lower (25th) and upper (75th) quartiles (box outline), minimum and maximum values excluding outliers (whiskers), and outlier values (open circles). Asterisk represents a significant difference for a given light treatment following a t test (shelter and light quantity treatment) or one-way ANOVA and least significant difference (LSD) post hoc testing (light quality treatment)

Full size image

Sheltering effect

Shelters are commonly used to protect seedlings from environmental stressors, such as wind and pests, while also permitting sufficient light for photosynthetic carbon gain and in association, seedling growth (Devine and Harrington 2008). In our experiment, sheltering was the most effective at promoting shoot growth over reserve storage. A similar response of greater gains in growth versus storage in greenhouse and sheltered rather than open conditions has been observed by numerous researchers for both coniferous and deciduous species (West et al. 1999; Pardos et al. 2003; Coutand et al. 2008). This was best represented in our study by significantly greater increases in shoot biomass (both stem and leaf mass), height and leaf area for aspen seedlings grown under sheltered full-sun conditions (Table 2).

In open full-sun conditions, on the other hand, seedlings generally experience greater exposure to environmental stressors, which ultimately can slow growth and may lead to preferential carbon allocation towards storage (Coutand et al. 2008). The carbon allocation in our study was clearly directed towards reserve storage rather than growth for the unsheltered seedlings, with larger increases between the two full-light treatments in R:S, root total NSC, and sugar, content and concentration (Table 3; Figs. 2, 3). In addition, there was a significant increase at the shoot level, in reserve storage, with a higher stem total NSC, and sugar, concentration of seedlings grown under unsheltered compared to sheltered conditions (Table 2). Similar results of enhanced allocation to reserve storage under open- and full-sun conditions have been observed by Naidu and DeLucia (1997) with oak and Landhäusser et al. (2012b) with aspen seedlings. In addition, Coutand et al. (2008) showed an increase in R:S of cherry seedlings grown without protection from wind mechanical stimuli compared with seedlings grown in environments protected from the wind. Thus, it appears that growing aspen seedlings under environmental stressors and natural light conditions common to outside environments may facilitate greater allocation towards desired seedling characteristics, such as high R:S and high root NSC, which ultimately may increase aspen seedling survival under stressful environmental conditions (e.g. Landhäusser et al. 2012b).

Shade effect

Our objective in growing seedlings under sheltered, full-sun compared to sheltered, shaded conditions was to produce larger seedlings with a greater allocation towards reserve storage and belowground biomass. We predicted a greater allocation to reserve storage, for seedlings grown in full-sun, due to an expected increase in photosynthesis, and in relation carbon gain, of seedlings at the higher (full-sun) light intensity (e.g., Givnish 1988; Montgomery 2004; Valladares et al. 2012). Further, we expected seedlings in the full-sun treatment to invest gained resources opposite to seedlings grown in the shade; for instance, greater allocation in the full-sun towards biomass rather than height growth, and to roots than shoots (e.g., Kitajima 1994; King 2003; Franklin 2008; Pierik and de Wit 2014). However, there were surprisingly limited growth or storage differences in our study between aspen seedlings grown under sheltered and shaded or sheltered and full-sun conditions. This lack of an effect on growth or storage differences in our study is in contrast to the findings of Gansert and Sprick (1998) for European beech seedlings, which displayed large increases in overall biomass, height growth and root NSC under full-sun compared to shaded conditions. Similarly, Chan et al. (2003) found much larger increases in overall biomass, height and diameter for Douglas-fir and red alder seedlings under full-sun compared to shaded conditions. However, the range in light quantity used by Gansert and Sprick (1998) and Chan et al. (2003) of ~100 and 25 % of full sun-conditions were much wider than the ~90 and 40 % of full-sun conditions used in our study. In comparable light conditions (100 and 45 % of full-sun), Logan (1965) found no significant difference in growth parameters with several maple and birch species. In contrast experiments, with P. tremuloides, focused more at the extremes of light quantity show significant positive effects of high light conditions on both growth and storage parameters (e.g., Hemming and Lindroth 1999; Landhäusser and Lieffers 2001; Osier and Lindroth 2006; Calder et al. 2011).

The much smaller range in light conditions in our study likely contributed to the limited observable differences in growth or storage for sheltered seedlings (Tables 2, 3; Figs. 2, 3). For example, numerous researchers have identified a light saturation point for P. tremuloides and for many other Populus species at or near 800 µmol m−2 s−1 (e.g., Bassman and Zwier 1991; Roden and Pearcy 1993; Landhäusser and Lieffers 2001; Roden 2003). This value is close to the conditions existing within the shaded treatments (at ~650 µmol m−2 s−1), so that the shaded seedlings were still very near the point at which the light conditions for photosynthesis of aspen is saturating.

Additionally, in sheltered environments, seedlings typically experience an increase in both temperature and VPD compared to open conditions (e.g., Boyer and South 1984; Puértolas et al. 2010; de Castro et al. 2014). An effect of air temperature on growth or storage reserves, in our study, was negligible due to this common sheltering effect, with a difference of only 0.1 °C between the sheltered, full-sun and shaded treatment (Table 1). Conversely, the negative impacts of higher VPD (Table 1) and higher radiation loads associated with the sheltered, full-sun conditions likely contributed to comparable growth between seedlings in the sheltered, full-sun and shaded treatment (Tables 2, 3), due to impacts on transpiration and heat exchange at the leaf and whole plant scale. For example, the increased VPD, and higher radiation loads naturally occurring under sheltered, full-sun conditions strongly influenced the significant reduction in the quantum efficiency of photosystem II (ΦPSII) (Table 1). This reduction in ΦPSII was likely due to the nonlinear response of photosynthesis to increases in insolation (e.g., Ögren and Sjöström 1990; Ort et al. 2011). This nonlinear response features strongly in our experimental outcome as any light intercepted above saturation (i.e. ~800 µmol m−2 s−1 for aspen) typically lowers photosynthetic efficiency in proportion to the excess light absorbed (Baker 1991; Ort et al. 2011). So that, above a certain point, the seedlings grown under sheltered full-sun conditions likely experienced diminishing returns in carbon gain. Further, any light which is not absorbed is either dissipated as heat (i.e. reduced ΦPSII) or results in photodamage (Melis 1999), thus additionally limiting any beneficial effects of increased light on growth or storage for seedlings grown under sheltered full-sun conditions in our study. Consequently, seedlings grown under sheltered full-sun conditions likely experienced not only increased losses of energy as heat but also enhanced respiration rates, which ultimately limited opportunities for carbon gain and resulted in similar growth and storage responses as seedlings grown under shaded and sheltered conditions.

Light quality effect

In crop and ornamental species, a change in light quality has been shown to lead to increases in both growth and reserve storage depending upon the region of the light spectrum utilized (e.g., Johkan et al. 2010; Samuolienė et al. 2011; Cope and Bugbee 2013). Similarly, studies with many boreal forest species have also found both enhanced growth and reserve storage when seedlings are grown under a particular portion of the light spectrum (Aphalo and Lehto 1997, 2001; Sarala et al. 2009). However, in our study, variation in light quality led to very minimal differences in either growth or NSC storage (Tables 2, 3; Figs. 2, 3). In fact, the only statistically significant difference in growth parameters was for height growth. This pattern, with seedlings grown under a low R:FR exhibiting the largest increases in height growth, is in close agreement with observations for many planted coniferous (Warrington et al. 1989; Ritchie 1997; Jacobs and Steinbeck 2001) and deciduous seedlings (Aphalo and Lehto 2001; Sharew and Hairston-Strang 2005; Devine et al. 2007; Devine and Harrington 2008). Such a response to low R:FR is thought to approximate understory sensing of competing vegetation and is considered a strategy for a given species to grow and expand beyond close neighbors (Ritchie 1997).

Notably, the increase in height for seedlings grown under low R:FR was not matched by peak responses in any other biomass component (Tables 2, 3). Rather, peak growth components undulated between the different light qualities but commonly centered upon seedlings grown under red light (Tables 2, 3). For example, peak gains in shoot components and total biomass were exhibited by seedlings grown under red light (Table 2). Similar findings of an increased allocation to growth under red light, particularly to shoot components, have been observed for numerous plant species (Wu et al. 2007; Wang et al. 2009; Samuoliene et al. 2011; Lu et al. 2012; Son and Oh 2013). This common increase in shoot growth under red light is thought to be related to the efficiency at which chlorophyll a and b absorb the red range of the light spectrum (Hopkins and Huner 2004). The blue range of the light spectrum is also efficiently absorbed by chlorophyll a and b, but is often less implicated in biomass accumulation (Wang et al. 2009; Johkan et al. 2010; Savvides et al. 2012; Son and Oh 2013). Rather, the blue range of the light spectrum is considered more important to chlorophyll formation (Son and Oh 2013), in close agreement with our finding of an increase in chlorophyll (a and b) content by both leaf weight without a concomitant increase in biomass or any biomass component for seedlings grown under blue light (Tables 2, 3).

Patterns in reserve storage were similarly equivocal and in all cases not different among light quality treatments. Although not statistically significant, the allocation in storage components such as root volume and R:S tended to be highest for seedlings grown under shaded, full-spectrum conditions. Conversely, the highest root NSC reserves and total root biomass tend to be found in seedlings grown under red light (Table 3). The differences in all storage components, however, were rather small, particularly in the case of R:S and total root biomass (Table 3; Fig. 2). A similar response of minimal storage differences was observed for seedlings of the evergreen Pinus sylvestris and two deciduous Betula species grown under blue light depletion (Aphalo and Lehto 1997; Sarala et al. 2011), and for several tropical tree species grown under differing light qualities (Tinoco-Ojanguren and Pearcy 1995), suggesting that growth is more commonly affected by changes in light quality than is storage.

In summary, the effect of light quality (blue, red and low R:FR) in our experiment operated more to direct resources towards height growth rather than to roots as storage or reserves. An increase towards height growth, particularly under red and low R:FR (Rajapakse et al. 1999; Clapham et al. 2002; Taulavuori et al. 2005; Sarala et al. 2009, 2011) is commonly observed for many plant species. Therefore, the alteration of the light environment using these two portions of the light spectrum is contraindicated when seeking to direct resources to reserves rather than growth. Conversely, height growth is typically constrained under blue light for boreal tree species (Taulavuori et al. 2010), although this common pattern does not automatically correspond to an increased allocation to reserve storage (Sarala et al. 2009). Rather, growth and developmental responses under blue light appear to be species-dependent and are potentially related to interactions with other wavelengths of light, in addition to interactions with light quantity (Cope and Bugbee 2013). As such, it appears for aspen seedlings that growth under blue light, and the associated light quantity in our experiment, provides limited benefit either in terms of increased growth or allocation to reserves (Tables 2, 3; Figs. 2, 3), and hence little contribution to the development of high quality aspen seedlings for use in reclamation activities.

Conclusion

Carefully manipulating the balance between growth and reserve storage is one particularly useful means to prepare seedlings for the harsh conditions common to many reclamation sites. This is most often accomplished in a seedling nursery, where there is wider control over environmental conditions (e.g., light, wind, water and nutrients) and therefore, theoretically, greater control over the balance between growth and storage. Based on our results, however, it appears that when these environmental stresses are removed that aspen seedlings direct resources to growth rather than reserve storage. The suggested reserve status of aspen seedlings, for successful outplanting in a reclamation setting, are for a R:S greater than 2 and root NSC greater than 30 %. These suggested benchmarks were most closely reached by seedlings grown under unsheltered, full-sun conditions. In all other cases the R:S was less than half of that provided as a benchmark and the root NSC concentration just three quarters of what is recommended for successful outplanting in a reclamation setting. We conclude from our results that full-sun and unsheltered environments with common environmental stresses on seedlings such as wind, temperature and enhanced radiation, are most effective in the development of high quality aspen seedlings for use in reclamation activities.