New Forests

, Volume 44, Issue 1, pp 23–38

Effects of soil moisture and species composition on growth and productivity of trembling aspen and white spruce in planted mixtures: 5-year results

Authors

    • Ontario Ministry of Natural ResourcesOntario Forest Research Institute
  • Ken J. Greenway
    • Alberta Sustainable Resource Development
Article

DOI: 10.1007/s11056-011-9297-1

Cite this article as:
Man, R. & Greenway, K.J. New Forests (2013) 44: 23. doi:10.1007/s11056-011-9297-1

Abstract

Aspen (Populus tremuloides Michx.) and white spruce (Picea glauca (Moench.) Voss) were planted 0.5 m apart in intimate mixtures in 5 × 4 m plots, with two moisture regimes—irrigation versus control—and five species compositions—pure aspen (Aw100), mixed aspen and spruce (Aw83Sw17, Aw50Sw50, Aw17Sw83), and pure spruce (Sw100), replicated six times. Fifth-year assessments indicated that irrigation increased individual tree growth (height, RCD, crown width), plot leaf area index (LAI), and wood biomass. Increased aspen composition reduced the availability of soil moisture and consequently the growth of individual trees. With increased aspen composition more growth was allocated to stem in aspen and to foliage in white spruce. Comparatively, aspen responded more to irrigation and thus their growth is more dependent on precipitation than that of spruce. Among the three growth variables assessed, height responded more to irrigation in both species. Equal mixtures and aspen-dominated mixtures in control plots had higher productivity in terms of total wood biomass in both absolute and relative terms. The implications of these findings are discussed in relation to managing aspen and white spruce mixedwood forests under increasing drought expected as a result of climate change.

Keywords

Root collar diameterHeightCrown widthLeaf area indexBiomassIrrigationDrought

Introduction

Trembling aspen (Populus tremuloides Michx.) and white spruce (Picea glauca (Moench) Voss) are widely distributed across Canadian boreal forests (Rowe 1972) and often grow in mixtures of various species compositions depending on location and stage of natural succession (Lieffers et al. 1996; Chen and Popadiouk 2002). These mixed-species forests are traditionally managed almost exclusively for single species, either white spruce or aspen. Over recent decades, appreciation of these mixedwood forests has increased, largely for non-timber values such as species, structural, and habitat diversity (Chen and Popadiouk 2002; Scholes and Biggs 2005) and resistance/resilience to natural disturbances (Su et al. 1996; Jactel et al. 2005).

Trembling aspen and white spruce occupy different niches in terms of growth rate, phenology, and requirements for light, nutrients, and moisture, and interact through inter- and intra-specific relationships, which can positively or negatively affect individual tree growth and stand productivity (Man and Lieffers 1999a; Lecomte et al. 2009). The positive interactions that may lead to reduced competition between the two species and, thus, greater productivity, include differences in shade tolerance, canopy level, phenological and successional development, as well as root distribution and nutrient requirements. Aspen may also positively affect white spruce growth by reducing environmental extremes at time of seedling establishment (Groot and Carlson 1996; Man and Lieffers 1999b), improving litter decomposition and nutrient cycling and therefore long-term site potential (Man and Lieffers 1999a; MacDonald 1995; Lecomte et al. 2009), and lowering the risk of pest damage such as from white pine weevil (Taylor et al. 1996).

MacPherson et al. (2001) compared aspen stands in northern Alberta with and without spruce understory and found that mixed-species stands had 10.5% greater periodic annual biomass increment and 10% greater standing biomass. These results are consistent with the calculations by Man and Lieffers (1999a) based on yield tables from Alberta’s phase 3 forest inventory. Possible productivity gains by mixed aspen and white spruce were also suggested by Kabzems and Senyk (1967), based on mean annual increment on average sites in Saskatchewan, and by Wang et al. (1995) in northeastern British Columbia based on model simulations with FORCAST. These findings are, however, restricted by variations in aspen clone configuration, site conditions, and stand history among natural stands (MacPherson et al. 2001).

In future, many parts of Canadian boreal forests will likely become drier, especially in summer, due to warming temperatures and/or decreasing precipitation (Kellogg and Zao 1988; Barrow et al. 2004). In response to increasing drought, on an individual tree basis, both aspen and white spruce are expected to show decreased growth (Hogg et al. 2005; Hogg and Wein 2005) and increased mortality (Candau et al. 2002; Hogg et al. 2002, 2005). Comparatively, white spruce has stronger stomatal control over water loss than aspen (Jarvis and Jarvis 1963) and may be more resistant to drought. At stand level, however, the effects of interactions between aspen and white spruce in mixtures of different species composition and soil moisture conditions, in terms of individual tree growth and stand productivity, are largely unknown.

In this paper, we present the fifth-year results of a mixed aspen-white spruce plantation in Alberta, established using a ‘replacement series’ design in which species in mixture are substituted for one another in different proportions, with total density kept constant (Kelty and Cameron 1995). The overall objective of this study was to examine the growth of aspen and white spruce in planted mixtures of varying species composition and soil moisture conditions, and to evaluate productivity of tree mixtures in relation to monocultures. The design of this study, however, may provide species interactions that differ from those in a mixedwood structure where overstory aspen density is maintained and white spruce in the understory adds to stand productivity (MacPherson et al. 2001).

Methods

Experimental design and treatments

The study site is located at the Alberta Research Council’s research station near Vegreville, Alberta. The soil is Dark Gray Chernozem with medium texture and good drainage (Soil Classification Working Group 1998).

A randomized complete block design, replicated six times, was used to deal with the effects of a possible difference in soil depth and moisture along an east-facing slope. Ten treatments, resulting from combinations of two moisture regimes (irrigation and control) and five species compositions from pure aspen (Aw100), to mixed aspen and spruce (Aw83Sw17, Aw50Sw50, Aw17Sw83), to pure spruce (Sw100), were randomly assigned to treatment plots within each block. Each treatment plot was 5 × 4 m, surrounded by metal sheets inserted vertically 60 cm deep to reduce root growth between plots and lateral movement of moisture and nutrients. The soil was ripped to a depth of approximately 30 cm before planting and Propex 3919 landscape fabric was used to hinder development of competitive vegetation.

White spruce and aspen seedlings (provided by Smoky Lake Forest Nursery, AlPac seedlot and K & C Silviculture, Weyerhaeuser Drayton Valley seedlot) were planted in early June 1999. The 1-year-old aspen seedlings had been trimmed to approximately 15 cm tall and cold stored until planting. The white spruce seedlings, about 25 cm tall at the time of planting, were started from seed, field grown for 1.5 years and cold stored until the time of outplanting. To promote rapid crown closure and early competition, seedlings were planted on a 50 cm grid so that a total of 63 seedlings were planted in each plot, with seedlings in the first row planted 25 cm from plot edge. In the mixed-species plots aspen and spruce were interspersed. The 25 seedlings in the plot centre were used for growth measurements. A handful of forest soil from a natural mixedwood stand containing mycohrrizae and plant starter fertilizer (10-52-10; Plant-Prod® water soluble fertilizer) were applied to each seedling at the time of transplanting. Regular watering was carried out in the first year to improve seedling establishment.

Moisture conditions were differentiated between irrigated and control plots in the second year. Because of high rainfall in the spring, irrigation was not carried out until summer. Plots were watered every 2 weeks, with approximately 5 mm of water delivered each time. Manual irrigation was continued for the entire growing season from early May to late October of the third year, following the same intensity and frequency. In the fourth year, manual irrigation was replaced with a drip semi-automatic system installed at individual tree bases, and a total of 180 mm water was added to the irrigated plots. Due to the extremely dry weather in 2002, approximately 50 mm water was manually added to the non-irrigated control plots to reduce the risk of massive mortality from the drought. Irrigation was discontinued in the fifth year due to a wet summer.

Weeds and aspen suckers inside the plots were hand-cleared. Planted seedlings that died in the first summer and winter were replaced in spring of second year to maintain plot density and species composition, with spare seedlings of the same stock reserved in a nearby planting. To reduce the risk of winter desiccation, each year a snow fence was set up on the north side of the plots in late October and removed in late April. The entire study area was fenced to prevent deer browsing and rabbit damage.

Data collection and analysis

Data collected included soil moisture, individual tree growth of 25 trees in plot centre, biomass of selected individual trees, and leaf area index.

Soil moisture was monitored in all treatment plots during the first 4 years, with single diode TDR probes inserted vertically at plot centre at three depths: top (0–20 cm), middle (20–40 cm), and bottom (40–60 cm). Rod length was 16.5 cm. Volumetric soil moisture content readings were taken every 2 weeks between early May and early October except for a 2-month delay in the first year. Two readings were taken each time and averaged for each probe using the TDR unit (Moisture Point TK-917, Environmental Sensors Inc., Victoria, BC).

Annual growth measurements were conducted in late September and early October. Measurements included total height, RCD (root collar diameter), and crown width averaged from two directions, south-north and east–west, for all live trees. In the fall of year 5 (2003), for each species one measurement tree, close in size to the plot mean, was felled at the ground surface and leaves, branches, and stems were separated to determine oven dry mass. Plot wood biomass (branches and stems) was calculated using the biomass values of sampled individual trees and number of live trees in each plot.

Leaf area index (LAI) was determined annually for 4 years starting in the second year. All treatment plots were assessed in mid-August using LAI 2000 Plant Canopy Analyzer. Readings were taken at five locations within each plot, four in the corners and one at the plot centre. At each location, one above canopy reading was followed by three below canopy readings at 5 cm above ground surface, and the LAI values by location were averaged by plot.

Proc Mixed with repeated option (in SAS 9.2) was applied to examine treatment effects on plot level LAI, tree height, RCD, crown width, and height:RCD ratio by species. Measurements in different years were treated as repeated factors. The first-order autoregressive covariance structure was used since measurements between adjacent years are correlated with one another (Littell et al. 1996). Prior to statistical analysis all individual tree measurements were averaged to generate plot-level means. Proc Mixed without the repeated option was applied to determine treatment effects on soil moisture at three depths by each measurement date, biomass proportions among foliage, branch, and stem for individual trees, and plot wood biomass.

In addition to total wood biomass, the productivity of different species mixtures was also compared on a relative basis, as outlined by Kelty (1992). The relative wood biomass production for aspen or white spruce was defined as the species biomass in mixed species plots, divided by the species biomass in pure species plots. The relative wood biomass total (RBT) is the sum of relative biomass production of both aspen and spruce in the mixed-species plots.

Results

Soil moisture

A period of severe drought occurred during the experiment. Based on data from a nearby weather station in Vegreville, total annual precipitation was 305 mm in 1999, 386 mm in 2000, 238 mm in 2001, 201 mm in 2002, and 450 mm in 2003, compared to 374 mm, the 30-year normal from 1971 to 2000 (Environment Canada weather station data). Soil moisture was similar at different depths, ranging from 15 to 34% for the top layer (0–20 cm), 15–33% for the middle layer (21–40 cm), and 15–33% for the bottom layer (41–60 cm) during the first 4 years. Mean soil moisture generally decreased with depth and over time (from year 1 to 4) as well as within years from the beginning to the end of growing season (Fig. 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs11056-011-9297-1/MediaObjects/11056_2011_9297_Fig1_HTML.gif
Fig. 1

Mean soil moisture by tree species composition and soil depth for 4 years after establishment of aspen–white spruce mixtures in Alberta

Soil moisture decreased with increasing aspen composition/density in the plots, particularly in the first 2 years when moisture levels were relatively high (P < 0.05 for most of the top and middle layer measurements; Fig. 1). Mean soil moisture averaged over the 4 years and three soil depths were 18% for aspen-dominated and pure aspen plots (Aw83Sw17 and Aw100), 19% for spruce-dominated and equal mixed plots (Sw83Aw17and Aw50Sw50), and 20% for pure spruce plots (Sw100). Differences in soil moisture content among the five species mixtures decreased with increased soil depth and reduced soil moisture over time (Fig. 1).

Mean soil moisture difference between the irrigated and control plots was 0.5%, averaged over the 4 years and three soil depths. A significant difference between the irrigated and control plots was generally observed in third year (2001) when 2–3% higher moisture content was frequently recorded in irrigated plots at middle and bottom layers. No significant interaction effect was evident between irrigation and species composition on soil moisture at any depth during the experiment.

Tree growth and allometry

Overall mortality was low, about 0.5% for aspen and 1.5% for white spruce by the end of fifth year. Average aspen trees were 11 mm in RCD, 75 cm in total height, and 39 cm in crown width by the end of first year. The largest increment occurred in the second year, with an average increase of 13 mm in RCD, 113 cm in height, and 57 cm in crown width. Aspen growth slowed down from the third to fourth year, particularly crown width, with an obvious bounce-back in the fifth year (Fig. 2).
https://static-content.springer.com/image/art%3A10.1007%2Fs11056-011-9297-1/MediaObjects/11056_2011_9297_Fig2_HTML.gif
Fig. 2

Root collar diameter (RCD), total height, crown width, and height:RCD ratio (least square mean ± S.E.) by species composition and measurement year in aspen–white spruce mixtures for 5 years after establishment in Alberta

White spruce RCD followed the same pattern as that in aspen, with the largest increment, 6 mm, in the second year, and the smallest, 2 mm, in the fourth year (Fig. 2). Mean total height of spruce trees was 40 cm by the end of first year, followed by increments of 14, 21, 17, and 16 cm from second to fifth year. Mean crown width was 20 cm in the first year and increased 15 cm in the second year, 16 cm in the third year, 12 cm in the fourth year, and only 3 cm in the fifth year.

Except for crown width of white spruce, both aspen and white spruce growth decreased with increased aspen composition (Table 1; Fig. 2). The effect of species composition on growth generally started in the second year (significant interaction between year and composition; see Y*C in the Table 1). By year 5, mean aspen trees were 48% bigger in RCD, 24% taller, and 56% larger in crown width as aspen composition decreased from 100 to 17%, while the corresponding increases in white spruce were 53, 32, and only 3% with aspen composition decreasing from 83 to 0%.
Table 1

Summary of analysis of variance results (P values) for growth and leaf area index (LAI) for aspen–white spruce mixtures established in Alberta

Treatment

Root collar diameter (RCD)

Total height

Crown width

Height:RCD ratio

LAI

Aspen

Spruce

Aspen

Spruce

Aspen

Spruce

Aspen

Spruce

Plot

Irrigation (I)

<0.01

0.18

<0.01

0.01

0.03

0.04

<0.01

0.03

<0.01

Composition (C)

<0.01

<0.01

<0.01

<0.01

<0.01

0.12

<0.01

<0.01

<0.01

I * C

0.41

0.12

0.79

0.47

1.00

0.13

0.34

0.09

0.63

Year (Y)

<0.01

<0.01

<0.01

<0.01

<0.01

<0.01

<0.01

<0.01

<0.01

Y * I

<0.01

<0.01

<0.01

<0.01

<0.01

<0.01

<0.01

<0.01

<0.01

Y * C

<0.01

<0.01

<0.01

<0.01

<0.01

0.41

<0.01

<0.01

<0.01

Y * I * C

0.03

0.34

0.93

0.03

0.27

0.85

0.62

0.50

0.71

The overall effect of irrigation on growth was significant, except on spruce RCD (Table 1). The examination of significant year by irrigation interactions suggested that both aspen and spruce growth started to differ between irrigated and control treatments by year four and the differences grew from fourth to fifth year (Fig. 3). By the end of fifth year, the irrigated aspen trees averaged 47 mm in RCD, 370 cm in height, and 125 cm in crown width, compared to 42 mm, 304, and 111 cm for the non-irrigated control trees. By year 5, irrigated spruce trees were 24 mm in RCD, 116 cm in height, and 67 cm in crown width, while the non-irrigated control spruce were 23 mm, 100, and 60 cm, respectively. Irrigation increased the total 5-year growth of individual aspen by 12, 22, and 13% for RCD, height, and crown width, and white spruce growth by 4, 16, and 12%, respectively.
https://static-content.springer.com/image/art%3A10.1007%2Fs11056-011-9297-1/MediaObjects/11056_2011_9297_Fig3_HTML.gif
Fig. 3

Root collar diameter (RCD), total height, crown width, and height:RCD ratio (least square mean ± S.E.) by irrigation and measurement year in aspen–white spruce mixtures for 5 years after establishment in Alberta

The height:RCD ratio generally increased with aspen composition (P < 0.01) for both aspen and white spruce, except for the first year when trees in different species compositions showed similar values (P < 0.01 for year by composition interaction, see Table 1 and Fig. 2). Over time, the height:RCD ratio in aspen increased from year 1 to 2 and oscillated afterwards, while the ratio in white spruce dropped from year 1 to 2, but increased gradually from year 2 to 5 (Figs. 2, 3). Both aspen and spruce height:RCD ratios were increased by irrigation (P < 0.05), also starting in the fourth year.

On an individual tree basis, about 50% or more of aboveground biomass was allocated to foliage in white spruce, while similar proportion of biomass was allocated to stem in aspen (Fig. 4). With the increase in aspen composition, white spruce allocated an increased proportion of aboveground biomass to foliage (P < 0.01), but less to branches (P < 0.01) and stem (P = 0.02). In aspen, however, with the increase in aspen composition more biomass was allocated to stem (P < 0.01) and less to branches (P < 0.01). The proportion of aboveground biomass was not affected by irrigation or the interaction between irrigation and species composition.
https://static-content.springer.com/image/art%3A10.1007%2Fs11056-011-9297-1/MediaObjects/11056_2011_9297_Fig4_HTML.gif
Fig. 4

Biomass proportions among foliage, branches, and stem of individual trees (least square mean ± S.E.) at year 5 by species composition for a aspen and b white spruce

LAI and plot wood biomass

On average, irrigated plots carried more LAI than control plots and the differences increased over time (P < 0.01 for irrigation and irrigation by year interaction; Table 1). By year 5, irrigated plots had 27% more LAI than control plots (3.18 vs. 2.72). In year 2, plots with high spruce composition had less LAI than those with high aspen composition, but the trend gradually disappeared. By year 5, however, plots with high spruce composition carried more LAI than those with high aspen composition (Fig. 5).
https://static-content.springer.com/image/art%3A10.1007%2Fs11056-011-9297-1/MediaObjects/11056_2011_9297_Fig5_HTML.gif
Fig. 5

Leaf area index (LAI) (least square mean ± S.E.) by species composition and measurement year in aspen–white spruce mixtures in Alberta

Plot wood biomass by year 5 was generally higher in irrigated than control plots and the treatment effect was greater for aspen than for spruce (P < 0.05 for biomass components of both aspen and spruce; see Fig. 6). The plot wood biomass total (combined branches and stem of aspen and spruce) was high in aspen-dominated mixed plots (P < 0.01), 1.95 kg/m2 in Aw83Sw17 and 1.92 kg/m2 in Aw50Sw50, followed by 1.88 kg/m2 in pure aspen plots, 1.42 kg/m2 in spruce-dominated mixed plots, and 1.09 kg/m2 in pure spruce plots. The plot branch biomass also appeared to be greater in mixed plots, even though the treatment effect was only marginally significant (P = 0.06).
https://static-content.springer.com/image/art%3A10.1007%2Fs11056-011-9297-1/MediaObjects/11056_2011_9297_Fig6_HTML.gif
Fig. 6

Plot wood biomass by species composition and irrigation for branches, stem, and total for aspen and white spruce 5 years after establishment in Alberta

On a relative basis, wood biomass total (RBT) for the three mixtures from low to high aspen composition was 0.96 (Aw17Sw83), 1.10 (Aw50Sw50), and 1.06 (Aw83Sw17). The corresponding values by irrigation treatment were 0.95, 0.98, and 0.97 for irrigated plots, and 0.95, 1.28, and 1.20 for non-irrigated plots. The increased RBT of mixtures in control plots resulted mainly from relatively greater aspen wood biomass (Fig. 6).

Discussion

Soil moisture

A period of severe drought during the course of the experiment substantially reduced soil moisture (Fig. 1) to a level close to the permanent wilting point of medium-textured soil (Larcher 1983) and likely imposed stress on planted aspen and white spruce trees, based on the dependency of growth upon moisture (both annual precipitation and irrigation) and the observed growth reduction and mortality loss in mature aspen trees in the affected areas (Hogg et al. 2008). At all three depths, soil moisture generally decreased with increased amounts of aspen in mixtures (Fig. 1), largely due to fast growth of aspen (Comeau et al. 2005), which requires more light, water, and nutrients than spruce (Peterson and Peterson 1992; Hangs et al. 2003). Lack of measured differences in soil moisture between irrigated and control plots were possibly due to (1) inadequate water added to the irrigated plots relative to the soil moisture deficit caused by the severe drought, and (2) greater growth of trees in the irrigation treatment which increased the demand for water and reduced soil moisture levels. We also noticed that 15 mm of water, delivered through the semi-automatic system over a week, generally stayed in the zone within 10–20 cm of the tree’s base. It is likely that the TDR probes, installed between trees, underestimated moisture values in irrigated plots, except for manual broadcast watering in control plots in the fourth year.

Tree growth

Other than competition for light and nutrients, increased moisture stress could also be responsible for the decreased individual tree growth in both species with the increase in aspen composition in mixtures (Figs. 1, 2). In plots with high aspen composition, growth seems to be restricted or stagnated, as evidenced by LAI saturation from year 3 to 4, which then decreased from year 4 to 5 (Fig. 5). The reduction of LAI by drought has also been reported elsewhere and may take several years to recover (Le Dantec et al. 2000; Bréda et al. 2006). With the increase in aspen composition and therefore competition, individual aspen and white spruce trees showed different biomass allocation strategies to maximize growth potential (Fig. 4). Aspen is shade-intolerant and prioritizes stem growth (Messier et al. 1999), while slow-growing white spruce is able to photosynthesize under reduced light conditions (Man and Lieffers 1997). The optimum for spruce may therefore be to maximize foliage production to capture more understory light (Messier et al. 1999). Both aspen and white spruce allocated more growth to height than to diameter (higher height:RCD ratio) when aspen proportion was high to improve access to light.

Irrigation improved tree growth through enhanced availability of soil moisture, and likely nutrients as well (Kramer and Green 2000; Hu and Schmidhalter 2005; Sardans et al. 2008). Between the two species, aspen responded more to irrigation, in terms of individual tree growth and plot LAI and wood biomass, similar to observations by Hogg and Wein (2005) in the Yukon. Other than poorer stomatal control of water loss, and therefore higher vulnerability to drought (Jarvis and Jarvis 1963), aspen might also need more water than spruce since they are in the upper canopy in mixtures and are generally larger, and thus ultimately subject to greater water stress. In the dry climate of Canada’s prairie provinces, aspen growth is strongly influenced by drought (Hogg et al. 2002, 2008). Periodic drought, coupled with defoliation by forest tent caterpillar, is suggested as the key factor responsible for extensive dieback and decline of aspen stands (Candau et al. 2002; Hogg et al. 2005, 2008). The projected increase of drought with climate change (Kellogg and Zao 1988; Barrow et al. 2004) may affect aspen more than on white spruce, especially in mixedwood stands of overstory aspen and understory spruce. Among the three growth variables assessed, height had the largest response to irrigation in both species (Fig. 3) and may therefore be more sensitive to drought.

Productivity

Species of different niches can be more productive in mixtures than in monocultures through reduced competition and increased facilitation (Vandermeer 1989; Kelty 1992). Both processes are possible in aspen and white spruce mixture (Man and Lieffers 1999a), but the reduced competition through shade tolerance separation usually plays an important role in mixedwood productivity gains in stratified mixtures of fast-growing shade-intolerant species over slower-starting shade-tolerant species (Smith et al. 1997; Vandermeer 1989; Kelty 1992). In a short-term small-scale planting trial, however, the reduced competition for moisture and possible nutrients may also be critical. In this study, RBT exceeds one only in equal mixes and aspen-dominated mixtures in non-irrigated control plots. Enhanced productivity resulted primarily from the improved growth of aspen, possibly due to increased moisture and nutrient availability (high aspen plots had significantly lower available K and P at year 2; R. Man, unpublished data). Fewer large healthy aspen in mixed plots produced similar branch and stem biomass as more small stagnant trees in pure aspen plots (Fig. 6). The reduced aspen density in mixed plots likely increased aspen canopy gaps and therefore understory light level, which enhanced the growth of white spruce and contributed to plot-level productivity (MacPherson et al. 2001). Further improvement of aspen growth in spruce-dominated mixtures did not compensate for loss of aspen biomass due to reduced density. This suggests that productivity gain in aspen and white spruce mixtures is more likely when soil moisture, and possibly nutrients, are limiting, similar to what Man and Lieffers (1999a) found in Alberta using yield tables where calculated relative total volume production of mixedwood stands increased with the decreases in site class. The same level of productivity gain may occur in irrigated plots with the increase in tree size and demand for resources over time. These results support the findings of others that mixtures of ecologically compatible species may not necessarily be more productive, and that stand attributes and site conditions are important considerations in mixedwood productivity (Kelty 1992; Man and Lieffers 1999a; Chen et al. 2003).

The magnitude of productivity gain in control plots in this study is comparable to those calculated by Man and Lieffers (1999a) using volume data from yield tables and the findings by MacPherson et al. (2001) from natural stands in Alberta. The productivity gain, however, may be smaller if foliage and roots are included. In this study, soil moisture decreased with increased aspen composition (Fig. 1; Table 1), which would cause a shift in growth allocation to roots (Fitter and Hay 1987; Ericsson et al. 1996; Prior et al. 1997; Chan et al. 2003) to increase water uptake. On the other hand, competition for light could increase biomass allocation of white spruce to foliage, even though the effect on aspen is limited (Fig. 4).

Ecological and silvicultural implications

The results from this study suggest that when soil moisture is limiting aspen and white spruce in mixtures are likely to be more productive than single-species stands. The enhanced productivity is achieved through the reduction of upper canopy density, a productive mixedwood structure suggested by Kelty (1992), which reduces competition for moisture and nutrients among aspen trees and improves understory light conditions for spruce. In this mixture, trembling aspen and white spruce are separated not only by shade tolerance, but also by drought tolerance. White spruce should also benefit from the reduced environmental extremes created by the aspen canopy resulting in improved establishment (Groot and Carlson 1996; Man and Lieffers 1999b). The reduction of aspen density and stocking in mixed stands may be achieved through natural or artificial thinning (Rice et al. 2001), harvesting (Alban 1991; Comeau et al. 2005), or modified tending treatments (Lecomte et al. 2009; Man et al. 2010).

Drought is expected to reduce all growth parameters in both aspen and white spruce. Comparatively, height may be more sensitive to drought than diameter and crown width, as a result of increased hydraulic limitation with soil moisture deficit (Ryan and Yoder 1997), as reported by Hogg and Hurdle (1995) who noted reduced aspen height along a moisture gradient in the western boreal forest. A greater response of height growth to drought, relative to that of diameter, has been reported in many other tree species (Broadmeadow and Jackson 2000; Merchant et al. 2007; Wang et al. 2008). Increasing drought with climate change will require adjustment of height-diameter and site index relationships in current growth and yield models. On the other hand, the lower sensitivity of white spruce to soil moisture suggests that, in the boreal mixedwood area where aspen frequently shows dieback and mortality from soil moisture deficit (Hogg and Hurdle 1995; Hogg et al. 2008), the extent of white spruce may increase if its regeneration is secured.

Caution is required, however, when management practices are oriented towards the development of productive aspen and spruce mixtures. First, species interactions in intimate mixtures may differ from those in strip mixtures, and the productivity gain associated with strip mixtures may be less. Second, species interactions are temporally and spatially dynamic, as are productive compositions. Under drier conditions, productive mixtures may contain less aspen (<50%), particularly with their increased size and requirements for resources over time. The span of this study was short and did not provide dynamics of species interactions as may occur during stand development. Third, this study was carried out in the southern boundary of the western boreal mixedwood where weather is generally dry and restricts tree growth (Hogg and Hurdle 1995). In the central zone of the western boreal mixedwood and the eastern boreal forest where moisture supports higher aspen composition/density, productivity gains in mixtures may be relatively less due to reduced light transmission through the aspen canopy (Comeau et al. 2005) and therefore reduced additional productivity contribution from spruce.

Acknowledgments

This study was supported financially by Alberta-Pacific Forest Industries Inc. and the Alberta Research Council. The authors thank Marie Gorda, Amar Varma, Dave Kelsberg, and Ed Korpela for help with plot set up and data collection, G. Grover for advice on treatment selection, and Smoky Lake Forest Nursery and K & C Silviculture for providing seedlings. Lisa Buse, Jim Rice, and two anonymous reviewers provided valuable suggestions for improving an earlier draft of the manuscript.

Copyright information

© Springer Science+Business Media B.V. 2011