Forestry Advance Access originally published online on April 29, 2009
Forestry 2009 82(3):323-342; doi:10.1093/forestry/cpp013
| ||||||||||||||||||||||||||||||||||||||||||||||||
Development and long-term evaluation of harvesting patterns to reduce windthrow risk of understorey spruce in aspen–white spruce mixedwood stands in Alberta, Canada
Canadian Wood Fibre Centre, Canadian Forest Service, Natural Resources Canada, 5320-122 Street, Edmonton, Alberta T6H 3S5, Canada
* Corresponding author. E-mail: dan.macisaac{at}nrcan.gc.ca
 |
Summary |
|---|
 Top Summary IntroductionMethodsResults and discussionFundingConflict of Interest StatementAcknowledgementsReferences |
|---|
The Hotchkiss River Mixedwood Timber Harvesting Study developed new approaches to harvesting systems for western Canada's boreal mixedwood forests. Conventional harvesting equipment was used to test 11 harvesting and silvicultural systems over a 530-ha site. These include one- and two-pass modified uniform shelterwoods, two- and three-pass strip shelterwoods, two-pass alternate strip shelterwoods and four-pass progressive strip shelterwoods. These were used to test varying levels of wind protection designed to protect and minimize wind damage to understorey (immature) white spruce residuals following harvest of the aspen overstorey. Ongoing monitoring of wind dynamics and associated windthrow patterns since 1992 have provided clear management practice guidelines for reducing windthrow of immature spruce and residual aspen following harvesting. There are thresholds related to spruce height and distance from aspen residuals, beyond which windthrow damage increases significantly. The influence of topography, timing of harvest and the spatial configurations of multiple harvests on windthrow dynamics have been clarified through this research. These results have provided valuable information to help forest planners to utilize harvesting and silvicultural systems that best reduce windthrow damage to understorey spruce, following harvest of overstorey aspen.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResults and discussionFundingConflict of Interest StatementAcknowledgementsReferences |
|---|
Boreal mixedwoods occupy about one-third of the productive land base in the boreal forests of the Prairie Provinces and northeastern British Columbia. Many of these stands are characterized by immature white spruce (Picea glauca (Moench) Voss) under a mature trembling aspen (Populus tremuloides Michx.) canopy. Over the past two decades, these forests have become a valuable economic resource in the region. Conventional practice had been to clearcut these stands and allow aspen and balsam poplar (Populus balsamifera L.) trees to regenerate by sucker or seed. The main method to protect the immature white spruce has been to avoid single trees during tree-length harvest and skidding. Due to harvest and skidding damage, only a few trees are left after harvest and most of those blow down soon after (L. Brace, personal communication).
Increases in the commercial use of these forests, coupled with public demand to maintain mixedwoods for a variety of non-timber purposes, are challenging traditional methods of managing mixedwood stands on the forest landscape. There is a need to develop mixedwood strategies to manage both species and for an innovative approach to both management planning and harvest operations.
Previous Canadian Forest Service studies have shown that harvesting systems utilizing conventional harvesting equipment (feller-buncher with grapple skidders) could be used to protect up to 50–60 per cent of the immature residual white spruce during harvest of overstorey hardwood trees on mesic sites (Navratil et al., 1994
); however, they did not address the need for layout and harvesting strategies to mitigate windthrow (blowdown). As a result, the Hotchkiss River study was established to demonstrate and evaluate the effectiveness of 11 methods of harvesting overstorey aspen trees in mixedwood forests while protecting understorey white spruce trees from wind damage and maintaining mixedwood forests on the landscape.
Risk of wind damage after harvest to individual white spruce trees is influenced by many factors such as local wind regimes, topography, tree morphology, spruce density, amount of root damage during harvest, effective screening by adjacent stands or windbreaks and support from residual aspen adjacent to the harvest strips (Coutts and Grace, 1995; Mitchell, 1995; Ruel, 1995). At the stand level, slenderness coefficient (SC) (also called height–diameter ratio and defined in equation (1)) of individual boreal white spruce is best predicted with density and basal area of all species combined, average height of aspen trees and site index (Wang et al., 1998).
 | (1) |
Also at the stand level, Scott and Mitchell (2005) consider fetch to be one of the most useful variables in modelling windthrow (where ‘fetch’ is defined as the uninterrupted distance the wind travels across an opening before hitting an edge (Scott and Mitchell, 2005; cf Burton, 2001).
One of the objectives of this study was to determine the effects of harvest pattern, specifically the effects of screening by the adjacent stand or windbreaks and the effects of residual aspen canopy in strips or scattered within the harvested areas on windthrow and the implications on standing volume.
One of the main tenets of understorey protection tested in this study is that white spruce trees will become more windfirm after release (as long as there is enough nearby wind protection to keep the spruce from blowing over). One simple approach to monitor this effect is to measure the SC of the trees before and after release. This SC has been used to evaluate the potential windthrow risk of trees (Peltola and Kellomaki, 1993; Mitchell, 2000) and is used in designing harvesting systems to mitigate windthrow. Trees that have a SC of greater than 90 are deemed to be at a very high risk of windthrow (Mickovski et al., 2005). Determining the effects of our harvest treatments on SC is also an objective of this study.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResults and discussionFundingConflict of Interest StatementAcknowledgementsReferences |
|---|
Site description
The research site is located 35-air km north-west of Manning, Alberta, within the Lower Foothills Natural Subregion (Beckingham et al., 1996) (57° 07' N, 118° 03' W). The composition, density, distribution and volume for principal species within the project area were characteristics of regional mixedwoods at overstorey age 80–100 years. At project establishment in 1992, trembling aspen (Populus tremuloides Michx.) and balsam poplar (Populus balsamifera L.) formed the 90-year-old overstorey 
25 m tall (700–1000 stems ha–1, 250–350 m3 ha–1), with a 60-year-old white spruce understorey (300–1500 stems ha–1, 27–120 m3 ha–1). Immature white spruce (Picea glauca (Moench) Voss) had an average density of 774 stems ha–1 with a mean height of 10.6 m and mean diameter at 1.3 m breast height (d.b.h.) of 11.6 cm. The dominant ecosite is ‘e’ (low-bush cranberry (Viburnum edule (Michx.) Raf.)); the ecosite phase is ‘e2’ (aspen–white spruce–lodgepole pine (Pinus contorta Loudon.)) (Beckingham et al., 1996). These sites are glaciolacustrine, moderately well drained on mid-to-upper slopes, grading to glaciolacustrine, imperfectly drained and organic over glaciolacustrine and very poorly drained on some of the more moist areas. Slopes are less than 7 per cent. Additional information can be found in Navratil et al. (1994).
Treatment description and experimental design
A two-stage harvesting and tending model (Brace and Bella, 1988) was adopted as a basis for harvest prescriptions at the Hotchkiss River site. The model assumes a first harvest at year 60 after stand establishment, when aspen trees are aged 60 or less and spruce under the aspen canopy are mostly aged 40 or less. The first harvest removes all aspen and some of the larger white spruce, creating growing space for aspen root sprouting and the (formerly) understorey spruce. The second harvest is planned to occur at year 120. At that time, both the new aspen crop and the released spruce will be mature, and spruce seed sources available. This two-stage harvesting system should (1) maintain mixedwood forests; (2) improve the aspen and spruce growth and yield and (3) promote the advanced reproduction of white spruce for the new stands.
Eleven treatment blocks plus four control areas totalling 530 ha in area were assigned at random in 1992. (Treatment block names follow the general format ‘silvicultural system-number of entries-width of entry-width of residual’; ‘E’ represents entry. In the strip shelterwood treatments (SS-2E-PS-100, SS-3E-TS-100 and SS-3E-TS-200), ‘PS’ represents permanent shelterwood, ‘TS’ represents temporary shelterwood, ‘S’ at the end of the code represents shelterwood portion of the treatment (50 per cent removal) and ‘A’ represents 100 per cent removal portion of the treatment.)
The harvesting patterns were designed to provide varying degrees of wind protection to residual immature spruce following aspen harvest. Table 1 describes each of the harvesting patterns, a subjective ranking of the degree of wind protection afforded by components of the treatment and the difficulty of harvesting using the method. The number of entries into a block and the amount and distribution of residual aspen canopy were factors considered in the subjective assignment of harvest difficulty for conventional harvesting (using feller-bunchers and grapple skidders) of a treatment. Figure 1 gives side view representations of three selected harvesting patterns (AS-2E-20-5R, SS-3E-TS-200 and AS-2E-150-10R) for clarification. (Each harvesting treatment was applied to one block, except for the one-entry clearcut and avoidance system (CCA), which was applied to two blocks. In this paper, results are presented by treatment block.) A ground photo is shown of the alternate strip treatment with two entries each 20-m wide and a 5-m aspen residual every 40 m after the second entry (AS-2E-20-5R) (Figure 2). Figure 3 is an air photo of the 150-m wide alternate strip treatment (AS-2E-150 (first entry), AS-2E-150-10R (second entry)).
|
|
|
|
In all harvested blocks, machine corridors were oriented north-south to provide maximum wind protection, as this alignment is perpendicular to the prevailing westerly wind direction, as verified by ongoing wind monitoring on site since 1992 (Flesch and Wilson, 2000
).
The first-entry harvest was conducted in fall and winter 1993–1994. The second-entry harvest was conducted in fall and winter 1998–1999. The third-entry harvest was conducted in the winter of 2003–2004. After a technical review, the second- entry prescriptions for the 20-m alternate striptreatment (AS-2E-20), and the 100-m and the 150-m alternate strip treatments (AS-2E-100, AS-2E-150) were modified to prevent unacceptable (>5–10 per cent) levels of windthrow. These treatments experienced extreme amounts of windthrow after the first entry that would only have become worse after the second entry. Additional wind buffer residuals were added to the second-entry treatment prescriptions in these blocks (AS-2E-20-5R, AS-2E-100-10R, AS-2E-150-10R) to ensure adequate wind protection for the second-entry harvest and to test other treatment hypotheses. Additional information on the harvest methods, including logging diagrams, equipment specifications and site maps, can be found in MacIsaac and Sauder (2001), MacIsaac and Krygier (2004) and Sauder and MacIsaac (2004).
Field sampling
Treatment areas were sampled using two to four 5-m wide belt transect lines established across the proposed harvest area the year before treatment. The start location of each transect was selected randomly. Each transect ran perpendicular to the machine corridors across the width of the treatment block. A total of 29.2 km of transects were established (21.0, 6.5 and 1.7 km of transect in the first-, second- and third-entry areas, respectively).
A numbered tag was attached to all overstorey deciduous trees and conifers taller than 1.3 m within the belt transects. For all trees, d.b.h. was measured, while heights were taken from a subset of trees (sufficient to generate site-specific height–diameter regression equations for each species). Following each harvest entry, tree condition (logging damage, natural damage, natural lean, logging lean, insect damage, disease, dead top, free growing, frost damage, standing dead, windthrow, harvested) and vigour class (very good, good, medium, poor) were measured for each tree in the transects. As well, remeasurements were taken on trees from previous harvest entries. Because of the staggered harvest entries, first-entry trees were measured at harvest and 5 and 10 years following harvest; second-entry trees were measured at harvest and 5 years following harvest and third-entry trees were measured at harvest. As of the third-entry assessment (10 years after first entry and 5 years after second entry), 5981 live white spruce trees were assessed in the 29 km of belt transects. As well, following each harvest, the location of all disturbances (e.g. start and end of machine corridors, landings, seismic lines) and visual and stream buffers were also recorded.
Windthrow was assessed in all treatments 1 year after harvest. Each machine corridor in the harvested area was walked and any windthrow in the residual strip between machine corridors was tallied by species. Only trees that were not damaged during logging were counted. Windthrow from the first 25 m from a landing was tallied separately from the rest of the strip.
After the second- and third-entry harvesting, a random sample of machine corridors was selected from the first- and second-entry areas for assessment. Any new windthrow in these previous entries was differentiated from windthrow that occurred the year after harvest by the presence of needles and fine branches on the crowns.
Hourly wind data (average and maximum) have been collected at the Hotchkiss site adjacent to treatment PS-4E-50 since project establishment in 1992.
Analysis methods
Regression analysis was done to determine the best regression of height vs d.b.h. A separate regression was done for each species and measurement year. Linear and non-linear models were tested, along with a variety of transformations, using the software package Tablecurve 2D. The best regression models were based on simple, biologically explainable model functions. For the smaller trees the equation
 | (2) |
for Natural Region 11 (Lower Foothills region) from page 65 from Huang (1994a) was used. These chosen regression equations were then used to estimate height of all live trees with a d.b.h. but no field height was recorded. This was only needed for the first two measurement periods. In the last measurement, height was recorded for all trees.
Volume equations for the Lower Foothills Natural Region of Alberta from Huang (1994b) were used to calculate individual tree volumes. Diameter and height class determination was done using height and diameters from the year of initial establishment. Trees were assigned to 5 cm centered diameter classes and 2-m centered height classes.
Windthrow density and per cent windthrow were calculated for each machine corridor and for each treatment by species groups (i.e. spruce, aspen/balsam poplar) and by the distance from the landing (i.e. <25 m from landing, > 25 m from landing, both distances combined). For the windthrow analysis, it was necessary to determine the length, width and area of each corridor to calculate windthrow density per hectare. Corridor lengths were measured from aerial photos of a known scale. Machine corridors were included in the total strip widths. Strip widths were obtained from the design protocols. The average post-harvest residual density from the belt transects was used to calculate windthrow per cent for all strips.
SC (height/d.b.h.) of windthrow trees was calculated for those trees windthrown in the belt transects where height and diameter data were available and not from those windthrown in the machine corridors as described above.
Due to the setup of the experiment, statistical tests were, for the most part, not conducted. Where statistics were used, non-parametric tests were employed, where required, as they did not require as rigorous assumptions about the underlying population structure compared with parametric tests. This included the use of Kruskal–Wallis tests (Conover, 1980) to compare values in multiple categories and t tests under the assumption of unequal variances to compare mean values from two groups. Spearman's rank correlation coefficient (
) was used to determine if there were any correlations at the treatment level between SC in the first-entry areas following harvest and 10 years later and stand variables of (1) white spruce density pre-harvest; (2) white spruce density post-harvest; (3) deciduous spruce density pre-harvest; (4) deciduous density post-harvest; (5) white spruce height post-harvest and (6) white spruce height 10 years after harvest and fetch. Aspen height was not used in this analysis because it was relatively uniform across all treatments. One-way analysis of variance and Duncan's multiple means test were used to test for differences in SC between treatments.
|
Results and discussion |
|---|
|
TopSummaryIntroductionMethodsResults and discussionFundingConflict of Interest StatementAcknowledgementsReferences |
|---|
Windthrow
Cumulative percentage windthrow response is shown graphically in Figure 4. The most obvious trend is the wide range in windthrow between treatments. The clear-cut avoidance treatments had cumulative windthrow of 69.6 and 50.6 per cent, for the CCA1 and CCA2 treatments, respectively. At the other end of the scale, three of the four controls had windthrow rates of less than 2.3 per cent. Figure 4 clearly shows that any level of wind protection resulted in a lowering of the windthrow rate as all the harvesting treatments had less than half the level of windthrow of the clear-cut avoidance treatments. This may indicate that irregularly structured stands may be more windfirm, as shown by Mason (2002).
|
The rate of windthrow varied greatly between the first and second 5-year intervals in the first-entry harvest areas. There was a strong initial pulse of windthrow within the first 6 months after harvest (Figure 4). This was due to the fact that the months following the first-entry harvest were the most windy of the first 6 years of the study (Flesch and Wilson, 2000
). During the first 5 years, there was ongoing windthrow, as shown by the cumulative increase in windthrow between 6 months and 5 years (Figure 4). However, during the second 5-year interval, the rate of windthrow was markedly less. This may be due to several factors, including (1) by year 5, the released spruce had become more windfirm; (2) in general, this second 5-year period was less windy than the first (with the notable exception of an extreme wind event on 3 January 2003, mentioned below) and (3) modifications were made to the second-entry harvest in AS-2E-20-5R, AS-2E-100-10R and AS-2E-150-10R, to reduce windthrow, because of what was seen during the first entry for these treatments. While per cent windthrow was an important factor when determining the severity of windthrow damage, the amount of volume loss must also be considered. In most of the treatment blocks, this amount was only a small portion of the pre-harvest volume (Table 2). However, in some treatments, the volume of windthrow exceeded 10 m3 ha–1, over a 10-year period, almost 35 per cent of the post-harvest volume (e.g. in AS-2E-150). Windthrow volume was also a function of stem density as is shown by the high volume loss in the 100 per cent aspen removal areas in SS-3E-TS-200. This treatment had a high pre-harvest density and also a higher water-table than some other areas of the study.
|
Results from the windthrow rate in the second-entry harvest areas show a consistently low level of windthrow (Figure 5). Initial windthrow rates were no more than 1 per cent in the 6 months following harvest. Five years later, the maximum cumulative rate was only 6.2 per cent. As with the first-entry areas, the 50-m wide alternate and progressive strip treatments also had less windthrow than the 100-m wide strips cut in AS-2E-100. The widest alternate strip cut treatment (AS-2E-150-10R) had a surprisingly small rate of windthrow. This may have been due to the effectiveness of the 10-m wide residual retained on the west edge of every third machine corridor. In terms of total volume lost to windthrow in the second-entry areas, the numbers were quite low, with a maximum of only 2.7 m3 ha–1 in AS-2E-100-10R.
|
The windthrow rate in the second-entry areas 5 years after harvest was similar to the low windthrow rate in the second 5-year period (10 years after harvest) in the first-entry areas. For the first few years following the second-entry harvest, it was considered that the second-entry wind protection prescriptions ‘had not been tested’. In other words, there were no wind events strong enough to potentially cause the high rates of windthrow seen right after the first-entry harvest. The windiest days during this period had wind speeds no higher than 65 km h–1 (18 m sec–1). However, on 7 January 2003 (four winters after harvest), there was an extreme regional windstorm that lasted for at least 1 day. Peak winds were measured at 103 km h–1 (29 m sec–1). The previous maximum wind gusts on site were in early May 1999, when winds reached 82 km h–1 (23 m sec–1). This 103 km h–1 event would be seen to test the range of wind protection designed in the second-entry areas at the Hotchkiss site. The resulting pattern of windthrow in the second-entry areas was not strongly related to the silvicultural system; many other factors that influence windthrow may have played a contributing role in the weak trends.
While windthrow was ongoing in all the treatments, in general, the amount decreased after the third-entry harvest from previous time periods, indicating that the understorey spruce are becoming more windfirm, similar to findings by Busby et al. (2006). While there is variation between treatments, the maximum cumulative windthrow following the second- and third-entry harvest was only 8 per cent. This is relatively small compared with 10-year cumulative windthrow of up to 30 per cent after the first-entry harvest.
Cumulative 10-year windthrow was significantly negatively correlated with post-harvest white spruce density ( = 0.5009, P = 0.0245), post-harvest deciduous density ( = –0.7983, P = <0.0001) and total post-harvest density of all species ( = –0.6421, P = 0.0023). This indicates the importance of retaining not just some overstorey aspen but also a sufficient number of understorey spruce, in order to reduce windthrow after harvest.
Fetch
The amount of windthrow of understorey white spruce was a function of the fetch in each treatment (defined as distance from the edge of residual canopy to the west of the nearest retained spruce) (Figure 6). Windthrow rate increased past a distance of 2.5 times the canopy height (in this case, 75 m) from a protective upwind edge or uncut residual. Between the 75- and 100-m distance, the per cent windthrow nearly doubled from 17 to 33 per cent. This finding validates the predictions on windthrow risk based on a model developed by Flesch (1999). It is strengthened by the fact that it is based on cumulative data from more than half of the treatment blocks at the Hotchkiss site (the remaining blocks, e.g. shelterwood treatments, were not used in this analysis because they did not have a defined fetch as part of the treatment).
|
The results at the individual treatment level also validate this finding, as the windthrow rate increased consistently from the 50-m wide strips in AS-2E-50 to the 150-m wide strips in AS-2E-150 (Figure 4). There was a significant positive correlation between the fetch and 10-year cumulative windthrow (
= 0.6482, P = 0.0020), which is in agreement with results using more complex fetch indices developed by Scott and Mitchell (2005).
Windthrow effects going into the forest residual from clear-cut strips was not investigated, but other studies indicate a 10- to 30-m influence zone (Mascarua-Lopez et al., 2006).
Understorey height
The threshold height above which white spruce was most susceptible to windthrow was 7.5 m based on all treatments combined (excluding controls) in the first-entry harvest area (Table 3). The cumulative 10-year windthrow was 58 per cent higher in the 7.5- to 10-m height class compared with the 5- to 7.5-m class. Trees in other classes above and below that height threshold also showed a corresponding difference in windthrow risk. The only exception was for spruce over 15 m in height. This might be because the larger trees have become more windfirm due to greater sway movement prior to harvest.
|
For the second-entry areas, there was little difference in windthrow between height classes except for the 0–2.5 m class (Table 3). This difference could have been a result of damage to these smaller trees during harvest that made them more susceptible to windthrow. We observed that leaning trees did not recover as well once they started to lean, even if the lean was only at a 15° angle: 50 per cent of the trees with logging lean in the first entry were windthrown 10 years later and 15 per cent in the second entry were windthrown 5 years later. While not observed, this might be partly due to additional snow loading on leaning trees, in addition to wind effects.
Slenderness coefficient
Mean pre-harvest SC values ranged from 96.6 to 108.8 over all treatment areas and the controls, which meant that, on average, every treatment block would be considered to be near or at a risk of windthrow based on a threshold of 100 (Navratil, 1996) (Table 4). Analysis of variance indicated a significant difference in post-harvest SC between treatments (Degrees of Freedom = 19, Sums of Squares = 24 527, Mean Square = 1290, F-ratio = 8.96, P-value = <0.0001). Duncan's multiple means test showed seven significant groupings (P < 0.05), with a range of SC from 84.8 to 100.0, based on undamaged post-harvest white spruce.
|
For all treatments, there was a reduction in SC over the 10-year period. This was in contrast to the controls, which had an increase in SC over time. There was a linear decline in SC for the first and second 5-year periods, with a cumulative 10-year reduction of between 2.8 and 31.3 units, depending on the treatment. With this trend towards lower SC values, many of the treatment blocks were approaching the threshold where the probability of windthrow based on an SC threshold of 70–80 was reduced. There tended to be a slight acceleration in the rate of SC change between the first and second entry. This may be due to greater diameter growth for those trees in the treated areas. As expected, the shelterwood treatments (SC-PB-10R, AS-2E-20, AS-2E-20-5R) and those with a high proportion of retained spruce (SS-3E-TS-200-S) exhibited lower rates of SC change (Table 4) because of the higher density of stems in these treatments.
While SC can be used to rate a stand as to its overall windthrow risk, how well did it predict the risk of windthrow for individual trees? In other words, was there a difference in SC in windthrown trees vs those standing after 10 years? In Table 4, the pre-harvest SC of trees that were subsequently windthrown (second column from the right) was compared with the SC of trees that were not windthrown. The results are mixed. The statistic in the right-hand column shows that for 10 of 17 blocks, the SC of windthrown trees was not significantly different than the SC of the non-windthrown trees (at an alpha level of 0.05). This could imply that the initial SC was not a good predictor of an individual tree's potential to blow down. However, in five of the seven significant differences, the SC of the windthrown population was greater than that of the non-windthrown population. This supports the suggestion by Navratil (1996) that the trees with larger SC do in fact tend to blow down, more than those with a smaller SC; hence, SC could be used as a predictor of windthrow risk of individual trees. One reason for the conflicting results is that these mixedwood stands are quite variable in stand structure and composition (MacIsaac and Krygier, 2004).
When averaged over an entire stand, is pre-harvest SC a better predictor of a stand's risk to windthrow than an individual tree's risk to windthrow given that the effects of local tree conditions are diminished as sample size increases? While there was a tendency for a positive block-level correlation between SC right after harvest and 10-year cumulative post-harvest windthrow, the relationship was not significant (P = 0.5694). This may be due to two factors: the inherent variability in stand structure and composition and the overriding influence of post-harvest fetch on windthrow results (both mentioned previously). However, the response of SC to stand-level conditions is illustrated by the highly significant positive correlation between 10-year cumulative windthrow and total post-harvest density ( = 0.8615, P = <0.0001). This was one of the stand-level variables deemed as most influencing SC by Wang et al. (1998).
Wang et al. (1998) note that windthrow may be influenced by many factors interacting with each other. These factors include tree characteristics, site conditions and local wind characteristics. Our finding that a tree's initial SC was not a good predictor of windthrow risk could be related to local factors around each of the windthrown trees such as the density of surrounding aspen and spruce, crown depth and width, local topography, rooting depth and local soil moisture conditions. Unfortunately, a detailed analysis of conditions local to each windthrown tree was not conducted as part of this study.
Treatment blocks with a non-significant difference in SC between windthrown and non-windthrown trees were the ones with a greater rate of SC change after harvest.
The analysis of SC confirms that there was a shift in allocation to diameter increment rather than height growth following overstorey removal.
Extrapolation of results
One of the study limitations was that it was conducted on one site only. However, there are lessons that have been learned from the Hotchkiss study that can be applied throughout much of the western boreal forest. Mixedwood conditions of aspen and white spruce similar to those at the Hotchkiss site can be found from northeastern British Columbia through to Central Ontario. Further east, other species mixtures become more prevalent in the boreal forest.
One difference going across the boreal region is that white spruce tend to have a lower SC in Manitoba than in Alberta. This, along with the fact that winds in the lower Alberta foothills can be quite severe, especially nearer the mountains (Navratil, 1995), may lead one to believe that the level of wind protection applied at the Hotchkiss site may not be required further east. However, it is important to note that windthrow is a response, not to ‘average’ wind conditions, but to periodic ‘severe’ wind events. There are recent examples in Ontario, where regional-level winds associated with synoptic weather systems have brought down large swaths of timber (D. Pitt, personal communication). For these reasons, it is reasonable to assume that the harvesting systems designed at the Hotchkiss site have application across the western boreal forest, as far east as Ontario. As proof of this, harvesting patterns from this study have been successfully employed in demonstration sites and operationally in British Columbia, Alberta, Saskatchewan and Manitoba, with much the same results as at the Hotchkiss site (MacIsaac et al., 1999; MacIsaac and Sauder, 2001).
Conclusions
Fetch, or more specifically, the distance from the edge of uncut canopy to the west to the post-harvest understorey white spruce, played an important role in the likelihood of windthrow of understorey white spruce due to the prevalence of westerly winds. The width of the strip in which the overstorey aspen are removed should not exceed 2.5 times the height of the overstorey trees if understorey spruce are to be protected from windthrow. Based on these results and the operational experience of our industry partners, the preferred treatment in terms of reduction of windthrow and operational efficiency is a single-harvest entry with 35 m of overstorey removal and 5 m of uncut overstorey residual using machine corridors on 20-m centres (10–15 per cent aspen retention). This pattern is based on the alternate strip two-entry 20-m treatment with a 5-m residual every two corridors (AS-2E-20-5R). It had the lowest 10-year cumulative windthrow rate, excluding the partial shelterwood portion in SS-3E-TS-100 (Figure 4), which is more complicated. By retaining the 5-m residual, the volume lost to windthrow was only one-eighth of that in the similar treatment (AS-2E-20), where all wind protection was removed in the second entry (Table 2). Five metres of retained overstorey is sufficient to protect understorey white spruce in an adjacent down-wind opening no more than 2.5 times the height of the residual.
The alternate strip treatment on which the recommendation above is based was conducted over two entries. As such, there may be some question as to whether the same amount of windthrow protection would be afforded to the recommended treatment which occurs over a single entry. Given that our findings show that the amount of windthrow increased past a distance of 2.5 times the canopy height (in this case, 75 m) from a protective upwind edge or uncut residual, the understorey trees in our recommended treatment should be safe given that there is a permanent residual every 35 m.
As is typical for boreal mixedwoods, the stands at the Hotchkiss site were quite variable over short distances, especially in terms of white spruce density, and to a lesser extent, in height and d.b.h. This has an impact on recommended prescriptions. How finely should forest planners stratify the treatment areas when designing silvicultural and harvest prescriptions? The results and experience from this and associated research studies indicate that the key variables to use in the stratification process are white spruce height and density, rooting depth (moisture related) and topography (even a slight slope can have an impact on local wind behaviour and thus on windthrow) (Ruel et al., 2001). A range of modifications around the recommended prescription is possible, depending on local stand and ground conditions.
Based on the results of our research, avoidance of immature residual white spruce using the clear-cut system (CCA) is not recommended for propagation of mixedwoods, especially on more moist ecosites and in stand conditions where white spruce trees are taller than 7 m, due to associated high windthrow rates. Research results showed that a large portion of the immature white spruce trees could be protected during an understorey protection harvest. The important thing is to ensure that these white spruce trees remain standing over time through the proper design of wind buffer residuals. For example, if there is undulating topography on a site, to reduce windthrow risk, aspen buffer residuals should be retained on upwind slopes, not just on slope crests (otherwise, the upslope winds penetrate beneath the canopy).
Our work showed that the threshold height above which understorey white spruce are most susceptible to windthrow was 7.5 m, with cumulative 10-year windthrow 58 per cent higher in the 7.5–10 m height class compared with the 5–7.5 m class. Trees did not recover as well once they started to lean, even if the lean was only at a 15° angle. For example, 50 per cent of the trees with logging lean in the first entry were windthrown 10 years later and 15 per cent of leaning trees in the second entry were windthrown 5 years later.
Pre-harvest SC values ranged from 97 to 109. SC values declined linearly for the first and second 5-year periods, with a cumulative 10-year reduction of between 2.8 and 31.3 units, depending on the treatment. SC in the non-harvested controls increased over time. If we can keep the understorey spruce trees standing, then over time, our results indicate that they should become more windfirm using the range of wind protection systems tested in this study.
Our results show that the use of SC as a predictor of an individual tree's potential to be windthrown is uncertain. We found that initial SC was, in the majority of cases (12 of 17 blocks), not a good predictor of an individual tree's potential to blow down. However, in 5 of the 17 blocks, there was a significant increase in the SC of windthrown compared with non-windthrown trees. The use of SC averaged at the block level had more promising results. While there was no significant correlation between block-level SC and subsequent windthrow, SC averaged at the treatment block level showed a significant positive correlation with total tree density.
Windthrow dynamics can be very complicated due to the range of factors involved (Ruel, 1995; Stathers et al., 1994). Predictions of windthrow is even more challenging in the spatially and structurally heterogeneous mixedwood stands such as those at the Hotchkiss River site. Recent wind-tunnel studies, model development, statistical approaches (e.g. logistic regression), decision support systems and hazard rating are helping to elucidate the importance of the effect of various factors on long-term windthrow dynamics (Stathers et al., 1994; Mitchell, 1995; Gardiner et al., 2005; Mickovski et al., 2005; Scott and Mitchell, 2005). This new knowledge will assist in better managing these dynamic systems.
|
Funding |
|---|
|
TopSummaryIntroductionMethodsResults and discussionFundingConflict of Interest StatementAcknowledgementsReferences |
|---|
Daishowa-Marubeni International Ltd; Manning Diversified Forest Products Ltd; Forest Resource Improvement Program and the Alberta Mixedwood Management Association.
|
Conflict of Interest Statement |
|---|
|
TopSummaryIntroductionMethodsResults and discussionFundingConflict of Interest StatementAcknowledgementsReferences |
|---|
None declared.
|
Acknowledgements |
|---|
|
TopSummaryIntroductionMethodsResults and discussionFundingConflict of Interest StatementAcknowledgementsReferences |
|---|
This is a cooperative project involving the Canadian Forest Service (CFS), Daishowa-Marubeni International Ltd (DMI), Manning Diversified Forest Products Ltd. (MDFP), the Forest Engineering Research Institute of Canada, Alberta Sustainable Resource Development, Land and Forest Division (SRD-LFD) and the Mixedwood Management Association. Without their involvement, this project would not have been successful. We would specifically like to thank Al Dumouchel and Steve Luchkow of DMI, J. P. Bielech of MDFP, Tony Sauder of FP Innovations—FERIC Division, former CFS employees Stan Navratil, Stan Lux, Cam Rentz and Lorne Brace and former SRD employee Paul King for their dedication to and involvement in the project. Laura Chittick, Ryan Cheng, Jocelyn Montgomery, Virginie Cayer, Martin Blank and Denise Wherry assisted with the field data collection. Laura Chittick assisted with analysis and production of tables and figures. Their assistance is gratefully acknowledged.
|
References |
|---|
|
TopSummaryIntroductionMethodsResults and discussionFundingConflict of Interest StatementAcknowledgementsReferences |
|---|
Beckingham JD, Corns IGW, Archibald JH. Field Guide to Ecosites of West-Central Alberta (1996) Edmonton, Alberta: Natural Resources of Canada, Canadian Forest Service, Northwest Region, Northern Forest Centre. Spec Rep. 9.
Brace LG, Bella IE. Understanding the understory: dilemma and opportunity. In: Management and Utilization of Northern Mixedwoods—Samoil JK, ed. (1988) Edmonton, Alberta: Canadian Forest Service, Northern Forest Centre. 69–86.
Burton PJ. Windthrow patterns on cutblock edges and in retention patches in the SBSmc. In: Proceedings of the Windthrow Researchers Workshop, Richmond, BC, 31 January–1 February 2001 (2001) 9–31. (compiled by S.J. Mitchell and J. Rodney).
Busby PE, Adler P, Warren TL, Swanson FJ. Fates of live trees retained in forest cutting units, western Cascade Range, Oregon. Can. J. For. Res. (2006) 36:2550–2560.[CrossRef]
Conover WJ. Practical Nonparametric Statistics (1980) 2nd edn. New York: Wiley.
Coutts MP, Grace J. Wind and Trees. 1 (1995) Cambridge, UK: Cambridge University Press.
Flesch T. Wind Simulation of the Hotchkiss River Mixedwood Management Project: Evaluating Windthrow Risk in Different Harvesting Shelterwood Cuts, Strip Cuts and Clear Cuts Unpublished Contract Report by J.D. Wilson and Associates (1999) Edmonton, Alberta: Natural Resources Canada, Canadian Forest Service.
Flesch TK, Wilson JD. Extreme Winds at the Hotchkiss River Mixedwood Management Demonstration Project. Unpublished Contract Report to Daishowa-Marubeni International Ltd. J.D. Wilson and Associates (2000) Edmonton, Alberta. 10.
Gardiner B, Marshall B, Achim A, Belcher R, Wood C. The stability of different silvicultural systems: a wind-tunnel investigation. Forestry (2005) 78:471–484.
Huang S. Ecologically Based Individual Tree Volume Estimation for Major Alberta Tree Species. Report #1. Individual Tree Volume Estimation Procedures for Alberta: Methods of Formulation and Statistical Foundations (1994a) Edmonton, Alberta: Alberta Environmental Protection, Land and Forest Service, Forest Management Division.
Huang S. Ecologically Based Individual Tree Volume Estimation for Major Alberta Tree Species. Report #3. Summary of Equations and Estimated Coefficients for Ecologically Based Individual Tree Volume Estimation in Alberta (1994b) Edmonton, Alberta: Alberta Environmental Protection, Land and Forest Service, Forest Management Division. Publication No. T/292.
MacIsaac DA, Krygier RF. White Spruce Understory Protection Research at Hotchkiss River, Alberta. Tenth Year Re-measurement and Third-Pass Assessment (2004) Edmonton, Alberta: Natural Resources Canada, Canadian Forest Service, Northern Forestry Centre. 112. Report to Daishowa-Marubeni International Ltd. Manning Diversified Forest Products Ltd. Forest Resource Improvement Association of Alberta and Alberta Mixedwood Management Association.
MacIsaac DA, Sauder EA. Harvesting and silviculture systems to protect immature white spruce and enhance deciduous regeneration in boreal mixedwoods. In: Proceeding of the Society of American Foresters National Convention, Society of American Foresters, Bethesda, MD (2001) Washington, DC. 253–259.
MacIsaac DA, Lux S, Sidders D, Edwards I. Hotchkiss River mixedwood timber harvesting study. For. Chron. (1999) 75(3):435–438.
Mascarua-Lopez LE, Harper KA, Drapeau P. Edge influence on forest structure in large forest remnants, cutblock separators, and riparian buffers in managed black spruce forests. Ecoscience (2006) 13:226–233.[CrossRef][Web of Science]
Mason WL. Are irregular stands more windfirm? Forestry (2002) 75:347–355.
Mickovski SB, Stokes A, van Beek LPH. A decision support tool for windthrow hazard assessment and prevention. For. Ecol. Manage. (2005) 216:64–76.[CrossRef]
Mitchell SJ. The windthrow triangle: a relative windthrow hazard assessment procedure for forest managers. For. Chron. (1995) 71:446–450.
Mitchell SJ. Stem growth responses in Douglas-fir and Sitka spruce following thinning: implications for assessing wind-firmness. For. Ecol. Manage. (2000) 135:105–114.[CrossRef]
Navratil S. Minimizing wind damage in alternative silviculture systems in boreal mixedwoods Publication No. Fo42–91/124/1995E (1995) Edmonton, Alberta: Canadian Forest Service and Alberta Land and Forest Service.
Navratil S. Silvicultural systems for managing deciduous and mixedwood stands with white spruce understory. In: Silvicultural of Temperate and Boreal Broadleaf-Conifer Mixture—Comeau PG, Thomas KD, eds. (1996) Victoria: B.C. Ministry of Forests. 35–46.
Navratil S, Brace LG, Sauder EA, Lux S. Silvicultural and Harvesting Options to Favor Immature White Spruce and Aspen Regeneration in Boreal Mixedwoods (1994) Canadian Forest Service, Northwest Region, Northern Forest Centre. Edmonton, Alberta. Inf. Rep. NOR-X-337.
Peltola H, Kellomaki S. A mechanistic model for calculating windthrow and stem breakage of Scots pines at stand edge. Silva Fenn. (1993) 27:99–111.
Ruel J.-C. Understanding windthrow: silvicultural implications. For. Chron. (1995) 71:434–445.
Ruel J.-C., Pin D, Cooper K. Windthrow in riparian buffer strips: effect of wind exposure, thinning and strip width. For. Ecol. Manage. (2001) 143:105–113.[CrossRef]
Sauder EA, MacIsaac DA. Using alternative harvesting techniques to protect the boreal mixedwood understorey: evaluation of harvesting productivity and costs. Forest Engineering Research Institute of Canada, Vancouver, B.C. Advantage (2004) 5(12):44.
Scott RE, Mitchell SJ. Empirical modelling of windthrow risk in partially harvested stands using tree, neighbourhood, and stand attributes. For. Ecol. Manage. (2005) 218:193–209.[CrossRef]
Stathers RJ, Rollerson TP, Mitchell SJ. Windthrow Handbook for British Columbia Forests (1994) Victoria: B.C. Ministry of Forests. Research Program Working Paper 9401.
Wang Y, Titus SJ, LeMay VM. Relationships between tree slenderness coefficients and tree or stand characteristics for major boreal mixedwood forests. Can. J. For. Res. (1998) 28:1171–1183.[CrossRef]
Received 10 August 2007.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||