Effect of stock type characteristics and time of planting on field performance of aspen (Populus tremuloides Michx.) seedlings on boreal reclamation sites
- First Online:
- Cite this article as:
- Landhäusser, S.M., Rodriguez-Alvarez, J., Marenholtz, E.H. et al. New Forests (2012) 43: 679. doi:10.1007/s11056-012-9346-4
- 299 Views
Aspen (Populus tremuloides Michx) has great potential as a reclamation species for mining sites in the boreal forest, but planting stock has shown poor field performance after outplanting. In this study we tested how different aspen seedling characteristics and planting times affect field outplanting performance on reclamation sites. We produced three different types of aspen planting stock, which varied significantly in seedling size, root-to-shoot ratio (RSR), and total non-structural carbohydrate (TNC) reserves in roots, by artificially manipulating shoot growth during seedling production. All three stock types were then field-planted either in late summer, late fall, or early spring after frozen storage. Seedlings were outplanted onto two reclaimed open-pit mining areas in the boreal forest region of central and east-central Alberta, Canada, which varied significantly in latitude, reclamation history, and soil conditions. Overall, height growth was better in aspen stock types with high RSR and TNC reserves. Differences in field performance among aspen stock types appeared to be more strongly expressed when seedlings were exposed to more stressful environmental site conditions, such as low soil nutrients and moisture. Generally, aspen seedlings planted with leaves in the summer showed the poorest performance, and summer- or fall-planted seedlings with no shoot growth manipulation had much greater stem dieback after the first winter. This indicates that the dormancy and hardening of the stem, as a result of premature bud set treatments, could improve the outplanting performance of aspen seedlings, particularly those planted during summer and fall.
KeywordsGrowth and carbon allocationNursery stockRoot carbohydrate reservesRoot-to-shoot ratioSeedling quality
Trembling aspen (Populus tremuloides Michx.) is a widely distributed tree species native to North America. Its natural range extends from Alaska in the west, across Canada into the north eastern United States and at higher elevations in the western and south-western areas of the United States and into northern Mexico (Little 1971; Perala 1990). This early successional, fast-growing species grows over a wide range of climatic and soil conditions and is considered relatively drought tolerant as compared to other boreal forest species (Lieffers et al. 2001). This makes it an ideal species for reclamation of disturbed mine sites in the boreal region, where there is a limited number of native tree species available for the reclamation and restoration of disturbed sites (Macdonald et al., in press). Currently, most aspen reclamation programs in the boreal forest region of Canada use nursery-grown seedlings. However, the outplanting success of aspen seedlings has been limited on boreal sites, where aspen seedlings often suffer from transplant shock and exhibit several years of slow growth after outplanting (Van den Driessche et al. 2003; Martens et al. 2007).
There is a lack of knowledge in identifying seedling characteristics and planting techniques that could be beneficial to improving aspen seedling establishment (Puttonen 1997). In the past, aspen seedling quality was assessed similarly to conifer planting stock (Sutton 1979) based upon height, root collar diameter (RCD), and terminal bud size (Chavasse 1980; Thompson 1985; Navarro et al. 2006). The current choice of aspen seedling attributes might not be adequate for assessing seedling quality and subsequent field performance since it has been recognized that seedling height and RCD are not always associated with field performance of broadleaf seedlings (Jacobs et al. 2005; Martens et al. 2007; del Campo et al. 2010). For example, there is a negative correlation between seedling initial height and field performance in northern red oak seedlings (Quercus rubra L.) (Thompson and Schultz 1995) as well as a negative relationship between survival and initial height in sawtooth oak (Quercus acutissima Carruth.) seedlings (Hashizume and Han 1993). Martens et al. (2007) observed that, after the first growing season, naturally-regenerating aspen seedlings were short in stature but had very high root-to-shoot ratios (RSR) and total non-structural carbohydrate (TNC) root reserves. However, these much shorter natural seedlings outperformed the nursery-produced seedling stock in height and root growth in the following growing season (Martens et al. 2007). The importance of TNC reserves in roots has also been shown for conifers (Grossnickle 2005) and aspen root cuttings (Snedden et al. 2010), which grew more new roots upon transplanting when higher reserves were present. TNC reserves in seedlings are also important for maintaining respiration from the time of lifting until restarting photosynthesis after outplanting (Marshall 1985). Since the balance between the water absorbing roots and the transpiring leaf area of newly planted seedlings is important (Haase 2008), large seedling stock with proportionally lower RSR and TNC reserves might cause poor outplanting success on reclamation sites, where a larger root system and greater root reserves may become critical to seedling survival.
The selection of a planting time for aspen on reclamation sites is generally based upon local climatic conditions and/or operational constraints such as planter availability. In northern boreal climates, fall-lifted and spring-planted aspen seedlings are generally stored frozen at −3 °C for up to 7 months. In the spring, site conditions are generally moist (shortly after snowmelt), and seedlings have a full growing season for growth and root system development before the following winter. Summer planting uses aspen seedlings that have set a terminal bud and stopped height growth and/or are top-pruned if too tall; however, these seedlings are not fully dormant. These seedlings can be considered “hot”-planted because they still have green leaves at the time of planting and can potentially re-flush if conditions are favorable. Summer planting also allows for the expansion of the root system during the remainder of the growing season after planting and results in a more intimate contact of the root system with the adjacent soils, potentially providing an advantage to the seedling the following spring (Taylor and Dumbroff 1975; Good and Corell 1982). Seedlings planted in the fall are generally considered to be fully dormant, which eliminates the need for prolonged winter storage, but here seedlings do not have time to develop much contact between the root system and the soil prior to winter. Different planting times have been investigated for conifers (Barber 1989; Dierauf 1989; Adams et al. 1991) and temperate hardwoods (Seifert et al. 2006), but the effect of planting time for aspen in a boreal climate has not been tested.
The objectives of this study were to determine the effect of planting time and different stock type characteristics on the establishment and growth of regionally-sourced aspen seedlings. In particular we were interested in the impact of seedling and root system size and the root reserves on the early outplanting performance of aspen on reclamation sites.
Materials and methods
Planting stock production
The aspen seed used in this study was collected from two open-pollinated seed sources near Edmonton, AB, Canada (53°34′N; 113°31′W; elevation 668 m asl) and near Fort McMurray, AB, Canada (56°43′N; 111°22′W; 370 m asl). Government regulations require different seed lots to be used for seedling stock production as the planting sites are in different seed zones to assure seedlings are suited for the climate and conditions in the respective reclamation areas. Seeds were sown mid-May 2008 into 615A Styroblock™ (Beaver Plastics Ltd, Acheson, AB, Canada) containers with a cavity volume of 340 cm3. Seedlings were grown in a mixture of peat and vermiculite (9 to 1 by volume) and were germinated under greenhouse conditions at a commercial tree nursery (Smoky Lake Forest Nursery, Smoky Lake, AB, Canada 54°6′N; 112°28′W; 598 m asl). During germination the greenhouse had a mean temperature of 21 °C, with a minimum of 18 °C and maximum of 28 °C; relative humidity was maintained at greater than 70 %. After seeding, styroblocks were irrigated using multiple automated mists per day for 4 weeks. Fertilization began 4 weeks after seeding using a standard nursery protocol used for growing commercial aspen seedling stock. The fertilizer solution consisted of 83 ppm of N, 76 ppm of P, 160 ppm of K, and chelated micronutrients; fertilizer was applied with every watering. Seedlings were moved to outside conditions after 8 or 11 weeks, depending on the stock type produced. Once outside, the fertilization regime was changed to 54 ppm N and 95 ppm K, while P remained at 76 ppm; this was done to limit height growth and the potential of re-flush in the seedlings that had set bud (see below). Seedlings continued to be fertilized with every watering cycle over the next 12 weeks.
Prior to moving seedlings to outside conditions, the styroblocks had been assigned to three different shoot treatments to create stock types with different seedling characteristics. The three stock types were labeled after the treatments used to terminate shoot growth: blackout, shoot growth inhibitor, and control with no artificial shoot growth termination. Generally, nursery-grown aspen seedlings for boreal forest climates are sown in early May, grown over the summer, and lifted in late summer or fall, and stored frozen over the winter months to be planted the following spring. Depending on the planting time, premature bud set can be artificially induced in aspen by shortening day length by means of blackout cloths, or by using a chemical shoot growth inhibitor (Rietveld 1988; Landhäusser and Lieffers 2009). Naturally, shoot growth terminates in aspen seedlings in early fall as a result of shortened day length and cooler nights. In an earlier study, the blackout and shoot growth inhibitor treatments had been found to be two reliable methods to induce premature bud set in aspen (Landhäusser et al. 2012). In the blackout treatment, bud set was induced after 8 weeks of growth. These seedlings were moved outside and subjected to an artificial shortening of day length for 7 consecutive days, by covering them with a black plastic tarp for a portion of the day to shorten the photoperiod to 8 h from the ambient 17 h. The same treatment was repeated again 2 weeks later. In the shoot growth inhibitor treatment, premature bud set was induced after 8 weeks of growth by treating seedlings with the plant growth regulator paclobutrazol (Bonzi®, Syngenta, North Carolina, USA). Paclobutrazol is absorbed by roots and shoots and inhibits gibberellin biosynthesis (Hedden and Graebet 1985) reducing internode expansion and apical dominance. This growth regulator was applied to the roots by soaking the styroblocks in a water bath with a concentration of 5 mL of Bonzi per L of water (0.02 g of paclobutrazol/L of water) as recommended by the manufacturer. Seedlings in the untreated control were moved outside the greenhouse after 11 weeks, to continue to grow and harden off naturally in the fall.
Planting time and planting sites
For the summer outplanting treatment, a third of the seedlings were lifted after the third week of August (14 weeks since seeding). The remaining seedlings stayed outside until another third of the seedlings were lifted at the end of September for the fall outplanting (18 weeks after seeding). The remaining third of the seedlings was lifted in November and stored frozen at −3 °C until the following spring (May 2009).
Winter (September (previous year) to April)/summer (May to August) precipitation and mean air temperatures for 2008, 2009, and 2010 at Warburg and Fort McMurray planting sites (AgroClimatic Information Service 2012)
Air temp. (°C)
Air temp. (°C)
Air temp. (°C)
To characterize the soil conditions at both research sites, 12 random soil samples were collected from the top 20 cm of the capping soil across each site and four samples were randomly pooled and analyzed as a single sample (total of n = 3 soil samples per site). Soil texture was estimated using a graduated cylinder and hydrometer (Carter and Gregorich 2008). Soils were analyzed for K+, Na+, Mg2+, and Ca2+ concentrations using the 1 M NH4OAc method (Page 1982), for NO3− and NH4+ using the 2N KCl method (Jones 2001), and for PO43− with the Kelowna method (Carter and Gregorich 2008). Total N and total P were analyzed with the Kjeldahl digestion method (Carter and Gregorich 2008).
Average nutrient concentrations (standard deviation) of capping soil collected at Warburg and Fort McMurray planting sites
To conform to agricultural regulations pertaining to the spread of noxious weeds, plastic mulch blankets commonly used in tree plantations to suppress competition (90 × 90 cm) (Arbortec Industries Ltd, Mission, BC, Canada) were placed around the seedlings at the Warburg site in May 2009 after the spring-planting treatment had been completed. Due to the dimensions of the plastic mulch, seedlings had been planted at a slightly tighter spacing at Warburg (0.85 × 0.85 m) than at Fort McMurray (1 × 1 m) where no plastic mulch was required. The use of plastic mulch in Fort McMurray is not required, as the spread of noxious agronomic weed species is currently not considered a critical issue in this area. All seedlings were planted by hand.
Prior to planting, each site was divided into 72 plots, which were randomly assigned to one of 9 treatment combinations of three planting times (summer, fall, spring) and three stock types (blackout, growth inhibitor, control) and replicated 8 times. Each plot contained 16 seedlings (subsamples) of the same treatment combination; the plot was considered the experimental unit in this study. In total 1,152 seedlings were planted at each site. Both sites were surrounded by a 2 m buffer planted with aspen seedlings to ensure seedlings growing at the periphery of each site experienced similar conditions to those growing in more central locations.
Average (standard deviation) of pre-planting characteristics of aspen seedling stock grown from Edmonton and Fort McMurray seed sources combined
Root volume (mL)
Root TNC (%)
After planting, initial height (height at time of planting) was measured on all field-planted seedlings. In the spring of 2009 prior to bud flush, two seedlings of each stock type planted in the summer and fall of 2008 were excavated to determine whether root growth had occurred during the previous partial growing season. Shoot dieback was determined on each seedling once bud flush had occurred in the spring. At the end of the first and second full growing seasons (August 2009 and 2010), total seedling height (measured from ground level to the highest terminal bud) was measured. After the first and second growing season, seedling mortality was also assessed.
To gain more detailed information on seedling growth and growth partitioning in the first growing season, two randomly selected seedlings of the 16 seedlings planted in each plot were excavated (total 144 seedlings). Shoot growth, RCD, and root and stem dry mass were measured. RCD growth and root growth were estimated by subtracting the average initial RCD and root mass measured on the 10 seedlings prior to planting from the RCD and root mass determined from the excavated seedlings at the end of the first growing season (2009).
Experimental design and data analysis
Initial planting stock characteristics were combined for Warburg and Fort McMurray, as seedling stock type characteristics were not different between the two seed sources (data not shown). Initial characteristics were analyzed using a two-way ANOVA with three stock types and three lifting times as the fixed factors. The field study was designed as a completely randomized 3 × 3 factorial design, with the three stock types (blackout, growth inhibitor, control) and three planting times (spring, summer, fall) as the fixed factors. Each treatment combination was replicated 8 times at each planting site (Warburg and Fort McMurray). Planting sites were analyzed separately as seed source, reclamation operations, and planting procedures were very different between sites (see above) and a comparison of these geographically distant locations was not an objective of this study; however, qualitative comparisons were made between the two sites when reasonable. The effect of stock type and planting time on shoot dieback, RCD increment, and root growth was tested with a two-way ANOVA. Because height growth was measured in 2009 and 2010, height growth was analyzed as a three-way ANOVA with planting time, stock type, and measurement year as main factors. Prior to analyses, the variables were examined for normality (Shapiro–Wilk test) and homogeneity of variances (Levene test). Only shoot dieback at Warburg did not conform to homogeneity of variances and was subsequently log 10 transformed. All analyses used the GLM procedure of SAS (SAS 9.2, SAS Institute, Cary, NC, USA). Data presented in graphs are the non-transformed means. A significance level of α = 0.05 was used for all analyses. When significant treatment effects were detected differences among means were determined using LSD multiple comparisons. Linear regression analysis was used to relate field performance of all nine treatment combinations to average initial seedling stock morphological and physiological characteristics.
Stock type characteristics prior to planting
The control stock type was twice as tall as seedlings of the blackout and growth inhibitor stock types regardless of planting time (P < 0.01) (Table 3). However, the control stock type also tended to have the lowest root volume (Table 3) and root mass (data not shown) particularly when lifted in the summer. The control stock type had a root volume similar to seedlings treated with growth inhibitor only in the spring, which resulted in a significant planting time by stock type interaction term (P < 0.001) (Table 3). Generally, summer-planted stock types had lower root volumes, root dry mass, and root-to-shoot ratio (RSR) compared to fall- and spring-planted seedlings; however, differences in RSR among stock types became larger in the fall- and spring-lifted seedlings resulting in a significant interaction term (P < 0.001) (Table 3). Overall, the seedlings treated with the growth inhibitor had the highest RSR, followed by the blackout stock type and the control stock type (P < 0.001). Root TNC in fall- and spring-planted seedlings was high among all stock types, while summer-planted stock types had the lowest root TNC, particularly in the control stock type. This also resulted in a significant interaction between planting time and stock type (P < 0.001) (Table 3). Generally, the TNC concentration in stems was about half of those in the roots, and stem TNC concentrations showed little difference among stock types (data not shown).
Field performance at Warburg
Field performance at Fort McMurray
Impact of initial seedling characteristics on field performance
Differences in the growth performance of aspen seedlings in response to the different planting times and stock types can be related to initial seedling characteristics, particularly root TNC concentration and RSR (Fig. 5), but they were also likely influenced by planting site conditions, such as fertility, soil texture, and climatic conditions. Seedlings with high TNC reserves and RSR (e.g., fall- and winter-planted and blackout and growth inhibitor-treated stock types) grew the best, indicating that RSR and TNC reserves are important characteristics to consider for describing aspen seedling quality. This might become more important for seedlings planted on sites with potentially limiting resources such as Fort McMurray, where both TNC reserves and RSR showed stronger relationships with height growth than at Warburg (Fig. 5). At Warburg, the above seedling characteristics might not have played such a prominent role due to the potentially higher resource availability. Apart from the initial transplant check (stem dieback, see below), seedlings at Warburg were exposed to more suitable soil moisture and nutrient conditions over the first growing season, which resulted in much better aspen growth than at Fort McMurray. These improved growing conditions were likely in part due to the use of the plastic mulch, but also due to the higher soil nutrient conditions (Table 2). The high nutrient content was likely a result of the alfalfa crop, which had occupied the site prior to becoming a research site, and had been incorporated into the soil. Secondly, in addition to reducing competition, the plastic mulch also reduced water evaporation (Allen et al. 1998; Mamkagh 2009) and increased soil temperature, which might have amplified N mineralization (Truax and Gagnon 1992). The greater percentage of sand in the soil at Fort McMurray may also have negatively influenced growth due to reduced water holding capacity and faster drainage.
Carbohydrate reserves have been found to be important in newly establishing seedlings (Farmer 1978; Wilson and Jacobs 2006) as an energy source between planting and the restart of photosynthesis (Marshall 1985; Carlson and Miller 1990; Landhäusser and Lieffers 2002). Generally, the lifting and planting of nursery seedlings during the summer coincides with a phenological stage of low carbohydrate reserves in the growing seedlings (Kozlowski and Pallardy 2002). During summer planting, aspen seedlings had green leaves and in general their TNC reserves were low as compared to fall- and spring-planted seedlings. This could have exposed them to greater water stress after outplanting, reducing their ability to photosynthesis and accumulate additional reserves for the winter and the following growing season (Rietveld 1989; Carlson and Miller 1990; Kozlowski 1991; Martens et al. 2007).
Although the control and growth inhibitor seedlings had relatively similar (although still statistically significantly different) root TNC reserves, the growth-inhibitor stock type still had twice the height growth at the Fort McMurray site. This response is interesting and might indicate that either the paclobutrazol affects other physiological mechanisms related to stress tolerance, or that the size of shoot tissue relative to the root system (RSR) could have an effect on root performance (Wan et al. 2006; Percival and AlBalushi 2007).
There was a very strong indication that the RSR played a significant role in the ability of aspen seedlings to establish and grow in the following year. The stock types that had a much smaller shoot relative to its root mass (high RSR) at the time of planting showed consistently better height growth regardless of planting time and location. A larger root system relative to the shoot should increase the capacity to supply more water and to remobilize reserves to the shoot, which initially carries fewer leaves, and, in combination with high TNC reserves relative to the shoot size, this could result in greater shoot growth over the growing season (Galvez et al. 2011; Landhäusser et al. 2012). Further, the smaller shoots of the growth inhibitor and blackout seedlings could potentially be beneficial in coping with the greater transpirational demands imposed on the root system when outplanted on harsh sites (Ritchie 1984).
Shoot dieback regardless of stock type and planting time was much more severe at Warburg than at Fort McMurray. However, only the summer- and fall-planted control seedlings at Warburg were severely affected. This could have been the result of seed source selection or site conditions. Seedlings planted at the Warburg site were produced from a more southern seed source; therefore, seedlings might require stronger queues for stem hardening to occur during seedling production than seedlings grown from a more northern seed source. Bud dormancy and cold hardening in plant populations are known to be influenced by latitude and altitude, where longer night lengths are required to induce dormancy in more southern populations (Dormling et al. 1968; Heide 1974; Ledgard and Norton 1988). During the early stages of stem hardening, and after bud set, seedlings are able to withstand air temperatures close to 0 °C and can withstand stem cavitation (Levitt 1980). During that time, seedlings still maintain cambial activity (Timmis and Worrall 1974), root growth (Day and Butson 1989), and reserve accumulation (Landhäusser and Lieffers 2003). However, if stems are not woody enough, drought or frost might affect the planted seedlings more severely. At Warburg the summer and fall stock types were planted into a recently rototilled field and the freshly disturbed soil likely provided poor root contact, which increased the risk of drought. At Fort McMurray, these stock types were planted into soil that had last been disturbed a year prior to planting, which allowed it to settle. In addition, shortly after planting in the fall, a killing frost event occurred at Warburg (−6 °C) which was less severe at Fort McMurray (−2 °C). Future research further exploring artificially induce bud and shoot dormancy during seedling production should also consider seed provenance locations in their treatments. Regardless, this shoot dieback at the Warburg appeared not to be a significant impediment for subsequent seedling growth, as these seedlings grew well in the following growing seasons, likely as a result of the high resource availability at this site.
The results of this study suggest that aspen seedling planting stock with high initial RSR and TNC reserves will perform much better on reclamation sites. This effect appears to become more prominent under resource-limiting site conditions. Summer seedling stock, which had the lowest TNC and RSR, performed the poorest, particularly on the harsher reclamation site at Fort McMurray. Artificially inducing bud dormancy increases TNC reserves and RSR in seedlings and these are linked to improved seedlings outplanting performance.
We thank Brad Pinno, Richard Caners, and two anonymous reviewers for their suggestions on improving the manuscript. We are grateful for the field assistance provided by Kim Stang, Jacklyn Burko, Kate Melnik, Tyana Rudolfsen, Candace Serben, Tory Cullen, Ryan Sherritt, Julia Wachowski, Jordana Fair, and Stefan Schreiber. Assistance with sample analyses and TNC measurements was provided by Pak Chow. We especially thank George Greenhough, Dan Kuchmak, Rob Vassov, and Francis Salifu for their logistic support. This research was supported by grants from Natural Sciences and Engineering Research Council of Canada (NSERC), Capital Power, Shell Canada, Suncor Energy, and Syncrude Canada.