Relationships Between Stand Age, Stand Structure, and Biodiversity in Aspen Mixedwood Forests in Alberta

 

A joint publication of the

Alberta Environmental Centre, Canadian Forest Service, Alberta Land and Forest Services

 

Full Report in PDF Format (1.4 MB)

Executive Summary

Research Team

Contents

Recommendations

 

EXECUTIVE SUMMARY

Resource managers and the environmental community are concerned that intensive clearcut logging of Alberta's aspen-dominated boreal mixedwood forests at 60-70 year rotations may alter the age class structure of the forest landscape and result in a change in forest structure and biota. In response to these concerns, we described forest structure and composition of plant and animal communities in young (20-30 years), mature (50-65 years) and old (120+ years) aspen mixedwood stands of fire origin in Alberta. The information collected in this study will serve as a reference against which structure and biota in harvested forests can be compared.

Old stands were structurally distinct from younger seral stages. Their uniqueness was related to a combination of canopy break-up and subsequent release of understory plants, emergence of secondary canopy species (white spruce and paper birch), the accumulation of deadwood, i.e., snags and down woody material (DWM), and the presence of nonvascular communities that develop on DWM. Relative to younger seral stages, old stands had trees of many ages and had more large-canopy trees, large snags, and advanced rot-class large DWM.

Old stands had a deeper organic soil layer and greater microtopographic relief than young and mature stands. The accumulation of organic material in the upper soil horizon may be related to greater input of deadwood materials from the canopy and subcanopy or to slower decomposition rates associated with cooler soils found in old stands.

Young stands had features associated with old stands because of residual materials from the previous stand cohort that escaped combustion during the fire or because of light conditions that were similar to old stands. Attributes that escaped combustion, or were created during fires, included large canopy trees, large snags, and coarse DWM.

Relative to young and old stands, mature stands were simple in structure with a closed canopy of aspen of relatively similar age, height, and diameter. In addition, mature stands had a more open understory with fewer shrubs and saplings.

Successional changes in stand heterogeneity were observed. For example, fires produced spatially heterogeneous patterns in the densities of trees and deadwood materials in young stands. The degree of spatial heterogeneity within and among stands was lowest in mature stands. In old stands, densities of trees and deadwood materials retained spatial homogeneity within stands; however, these structures became more heterogeneous among stands.

Microclimate varied among stand age by season. During summer, young stands were colder at night and warmer during the day than other stand ages. During winter, old stands were warmer during the day and young stands were warmer during the night than other stand ages. Soil temperatures below ground varied among season and by stand age. Old stands had lower soil temperatures than other stand ages in summer while young stands had the coldest soils in winter. Photosynthetically active radiation (PAR) varied seasonally and was related to canopy and subcanopy foliation and defoliation events. Mature stands had lower levels of PAR and higher wind speeds than did young and old stands.

Herbaceous cover during summer was positively correlated with temperatures above and below the ground surface. Shrub/sapling cover and depth of organic material (DOM) were negatively correlated with air and soil temperatures during summer. Soil temperatures were highest beneath forest canopies characterized by tall sparse trees and decreased as shrub/sapling cover and DOM increased. PAR was positively correlated with tall sparse trees, shrub/sapling cover, and DOM, and negatively correlated with grass and herb cover.

Compositional changes in understory vascular plants during succession in aspen mixedwood forests followed an initial floristics, rather than a relay floristics, pattern. Species turnover (introduction/extinctions) from young to old stands were relatively uncommon. Turnovers occurred mostly among uncommon species and were independent of stand age. Approximately half of the species differed in abundance among stand age. Of the species that changed, most either increased or decreased monotonically with stand age. Successional changes and patterns in understory plants were the result of changes in the relative frequency of core species. The core group produced the different physiognomic profiles that were associated with different stand ages.

Old aspen mixedwood stands had the greatest variety of woody substrates available in different decay stages; consequently they supported the greatest number of moss, liverwort, and fungi species. High variability of habitat types in old stands may also explain why old stands had more rare nonvascular species than either young or mature stands. These species were categorized into three ecological groups: epiphytes (species commonly found on live trees); epixylics (species typical of rotting wood). and terricolous species (species that use the ground as their primary habitat). Different decay stages of DWM did not have unique nonvascular assemblages due to high intergradation of species. Epiphytes were more abundant on early decay stages, epixylics were more abundant on intermediate decay stages, and terricolous species were more abundant on advanced decay stages. The sensitive liverwort group was restricted to intermediate stages of decay. Nonvascular species assemblages on similar decay stages of DWM were different among stand ages, indicating that time, as well as structural attributes, was important in determining species assemblages.

A rich biota in aspen mixedwood forests was identified including 47 lichen species, 39 mosses, 7 liverworts, 24 fungi, 11 pteridiophytes (horsetails, club mosses, and ferns), 65 herbs, 27 shrubs, 6 trees, 3 amphibians, 76 birds, and 33 mammals.

Old stands had greater bird species richness than young stands, which had greater richness than mature stands. Variation in abundance among stand age was evaluated for 33 bird species that were detected ten or more times. Twenty-one species had their highest abundance in old stands, nine species had their highest abundance in young stands, and only three species had their highest abundance in mature stands.

Variation in bird abundance among stand age may have been caused by variation in vegetation. During succession within aspen mixedwood forests, density of live trees decreased, height of canopy trees increased, and density of large live trees, large dead trees, and large DWM increased. As a secondary pattern during succession, canopy and understory heterogeneity were greatest in old stands, intermediate in young stands, and lowest in mature stands. Many attributes of deadwood (snags and DWM) and understory plant communities were correlated with variation in canopy heterogeneity. For most species of birds, abundances were related to variation in one, or both, of these two major successional patterns of forest community attributes. Abundance of 20 bird species was correlated with successional stage, and abundance of 14 bird species was correlated with canopy heterogeneity. Ten of the 21 species that were most abundant in old stands preferred conifer-dominated mixedwood forests during the breeding season; they may have been most abundant in old stands because those stands contained many conifer trees.

Mammal species richness in summer was significantly higher in old stands than in young or mature stands. During winter, mammal species richness was higher in old and young stands than in mature stands. Variation in relative abundance among stand age was evaluated for 19 mammal species that were detected ten or more times. During summer, most species had their highest abundance in old stands (9), followed by young stands (4), then by mature stands (3). Differences in mammal abundance during winter were less obvious (2, 3, 1 species had their highest abundances in old, young, and mature stands, respectively).

Changes in mammal species abundance were associated with forest community attributes that varied in relation to successional stage (stand age) and canopy heterogeneity. Complex old stands supported more species at high relative abundance, structurally simple mature stands supported few species at high relative abundance, and young stands that were intermediate in structural complexity due to pre-fire residuals (snags, large live trees, DWM) supported an intermediate number of species at high relative abundance.

Three mammal groups were examined in greater detail because of their economic importance (ungulates) or because of their perceived sensitivity to logging practices (bats, flying squirrel). During winter, forest structures that offer thermal protection may limit deer abundance, while browse availability may limit moose abundance. For example, relative abundance of white-tailed deer during winter was significantly higher in old than in young and mature stands, a trend that may be explained by greater availability of thermal protection (white spruce) found proximal to shrubby browse in old stands. Relative abundance of moose during winter was significantly higher in young and old stands than in mature stands, a trend that may be explained by high abundance of shrub and sapling browse in young and old stands. For both white-tailed deer and moose, abundance did not differ among stand age during summer; thus, their abundance seems to be affected by stand age only in winter.

Bat abundance was evaluated using echolocation calls and feeding buzzes, while bats fitted with radio transmitters were used to locate roost trees. Adult and juvenile little brown bat, silver-haired bat of both sexes, and one adult of both hoary bat and northern long-eared bat were captured. Nine little brown bat and five silver-haired bat were radiotagged and followed to 23 roosts, all in Populus spp. trees. Mean roost tree height and DBH were 21.8 m and 41.4 cm, respectively, for little brown bat, and 22.4 m and 41.3 cm for silver-haired bat.

Myotis spp., silver-haired bat, hoary bat, and big brown bat were detected in all stand ages. Myotis spp. constituted the highest proportion of total activity (#passes/hr) in both years. During 1993, more old stands than young stands were used by Myotis spp. and all bats in the first hour post-sunset; during 1994, more old stands than both young and mature stands were used. The distribution of bat activity in the boreal forest was related to the density of trees in the forest, the presence of large gaps in the canopy, and the abundance and distribution of roost trees in stands of different ages.

Flying squirrel abundance was higher in old stands than other stand ages and was positively related to conifer and shrub/sapling densities. Although flying squirrels frequently used cavities with small entrances in large diameter aspen snags, preference for den trees, relative to availability, appeared to be based on the following characteristics: live trees, low decay stage, high percent of bark remaining, a single cavity in the tree, and a large cavity entrance.

Much of the structural complexity (residual green trees, snags, DWM) found in young stands is related to the variable manner in which fire burns through aspen mixedwood forests. Furthermore, these structures appear to be highly related to the presence and abundance of many different species of plants and wildlife. If this general connection between fire, stand structure, and biodiversity exists, then resource managers must compare structures retained by natural disturbances (e.g., wildfire) to those left by human-caused disturbances (e.g., clearcuts).

The current and dominant form of forest management in Alberta's mixedwood forests includes unstructured clearcut logging, a narrow range of rotation ages leading to a constant-age class merchantable forest, silvicultural enhancement for conifer regeneration, and the administrative separation of the hardwood and softwood landbase. These forestry practices are fundamentally different from fire as a stand-initiating event and, when implemented over periods as short as one rotation, will lead to a forest landscape that is both different and more uniform than those found presently. Changes to the forest landscape caused by current forestry practices include the loss of structural complexity (green trees, snags, DWM) in young forests, loss of older forest stages, and the likely spatial separation of aspen and conifers. The transformation of the boreal forest landscape, in turn, is likely to alter the composition and abundances of biota and the nature of ecological processes. Current forest management practices, therefore, are inadequate to ensure the ecological integrity of Alberta's boreal mixedwood forests. Results of this study offer a template for incorporating the natural dynamics of snags, DWM, and conifer regeneration into forestry operations.

Provincial government agencies and the forest industry are therefore encouraged to:

  1. accept the ecological importance of fire in maintaining stand and landscape level variability, and therefore a) implement forest harvest strategies that approximate patterns produced by fire by retaining structures (green trees, snags, DWM), b) adopt variable harvest rotation ages, and c) maintain natural disturbances in the boreal mixedwood forest to the extent permissible by society,
  2. recognize the existence of a distinctive "old-growth" stage in aspen mixedwood forests, and ensure that appropriate frequencies of older seral stages are maintained in the boreal forest landscape. The practice of sequencing the harvest of old stands early in the rotation needs to be re-evaluated in the context of maintaining appropriate frequencies of all seral stages on the forest landscape. An empirical description of fire return intervals offers a preferred model for maintaining age class structure of managed aspen mixedwood forests,
  3. recognize that mixed canopies of aspen and spruce dominate the mixedwood forests and that this interspersion provides important substrate, forage, nesting, and cover for many biota, and therefore devise land tenure, harvesting, and silviculture systems that do not encourage the unmixing of hardwood and softwood species at the scale of the stand, and
  4. recognize that many of the ecological effects of commercial forestry, including those guided by forest ecosystem management, are presently unknown, and therefore establish aspen mixedwood forest reserves of appropriate scales as ecological bench marks against which other forest land-uses can be evaluated.

The implementation of these recommendations may affect future production and availability of tree fiber for harvest from Alberta's aspen mixedwood forests. Changes to forest harvest levels will be affected by the degree to which the recommendations are adopted (i.e., how much structure is retained on cutblocks, how variable rotation ages become, how much wildfire occurs on the landscape, how much area is assigned to unharvested reserves). For example, the retention of high densities of live aspen trees on cutblocks for purposes of wildlife habitat or ecological processes will likely reduce post-harvest aspen regeneration and, hence, future fiber production. A review of the social and economic consequences of the proposed recommendations is required to arrive at sustainable management goals that meet ecological, social, and economic criteria. However, resource managers must recognize that all forest-based land-uses and social benefits (i.e., commercial forestry, hunting, aesthetics, bird-watching, etc.) ultimately depend on functional forest ecosystems; hence wise forest stewardship should be the primary and over-riding societal objective.

 

RESEARCH TEAM

J. BRAD STELFOX, Project Leader, Forest Ecologist; Alberta Environmental Centre, Vegreville, AB.
LAURENCE D. ROY, Research Scientist, Wildlife Biologist; Alberta Environmental Centre, Vegreville, AB.
PHILIP C. LEE, Research Scientist, Plant Community Ecologist; Alberta Environmental Centre, Vegreville, AB.
JIM SCHIECK, Research Scientist, Ecologist; Alberta Environmental Centre, Vegreville, AB.
SUSAN CRITES, M.Sc. Candidate, Plant Ecologist; Department of Biological Sciences, University of Alberta, Edmonton, AB.
MARIE NIETFELD, Research Scientist, Wildlife Biologist; Alberta Environmental Centre, Vegreville, AB.
JACK NOLAN, Senior Technician, Wildlife Biology; Alberta Environmental Centre, Vegreville, AB.
LISA H. CRAMPTON, M.Sc. Candidate, Ecologist; Department of Biological Sciences, University of Calgary, Calgary, AB.
LISA MCDONALD, M.Sc. Candidate, Ecologist; Department of Biological Sciences, University of Alberta, Edmonton, AB.
DAVID CHESTERMAN, M.Sc. Candidate, Geographer; Department of Earth and Atmospheric Sciences, University of Alberta, Edmonton, AB.
LEN PELESHOK, Technician, Wildlife Biology; Alberta Environmental Centre, Vegreville, AB.
KELLY STURGESS, Technician, Wildlife Biology; Alberta Environmental Centre, Vegreville, AB.
DELINDA RYERSON, Research Technician, Database Coordinator; Alberta Environmental Centre, Vegreville, AB.
DANI WALKER, Technician, Wildlife Biology; Alberta Environmental Centre, Vegreville, AB.
GREG RADSTAAK, Wildlife Database Coordinator; Alberta Environmental Centre, Vegreville, AB.
ZACK FLORENCE, Biometrician; Alberta Environmental Centre, Vegreville, AB.
HAI VAN NGUYEN, Biometrician; Alberta Environmental Centre, Vegreville, AB.
ROBERT M.R. BARCLAY, Associate Professor, Ecologist; Department of Biological Sciences, University of Calgary, Calgary, AB.
DAVE HALLIWELL, Research Scientist, Meteorologist; Canadian Forest Service, Edmonton, AB.
PETER KERSHAW, Associate Professor, Biogeographer; Department of Earth and Atmospheric Sciences, University of Alberta, Edmonton, AB.
MARK DALE, Professor, Plant Community Ecologist; Department of Biological Sciences, University of Alberta, Edmonton, AB.

 

CONTENTS

Preface
Executive summary
Research team
Acknowledgements
1. Introduction
2. Relationships between microclimate and stand age and structure in aspen mixedwood forests in Alberta.
3. Changes in forest structure and floral composition in a chronosequence of aspen mixedwood stands in Alberta
4. Changes in snags and down woody material characteristics in a chronosequence of aspen mixedwood forests in Alberta
5. Changes in understory composition for a chronosequence of aspen mixedwood stands in Alberta
6. Relationships between nonvascular species and stand age and stand structure in aspen mixedwood forests in Alberta.
7. Bird species richness and abundance in relation to stand age and structure in aspen mixedwood forests in Alberta.
8. Relationships between mammal biodiversity and stand age and structure in aspen mixedwood forests in Alberta.
9. Abundance of ungulates in relation to stand age and structure in aspen mixedwood forests in Alberta.
10. Relationships between bats and stand age and structure in aspen mixedwood forests in Alberta.
11. Relationships between northern flying squirrels and stand age and structure in aspen mixedwood forests in Alberta
12. Changes in vertebrate communities in relation to variation in forest community attributes: a comparison of bird and mammal communities
13. General summary
14. Recommendations
15. References
i. Glossary
ii. List of abbreviations
iii. Species list
iv. Stand maps

 

RECOMMENDATIONS

Intensive forest harvest (e.g., clearcut logging of 60-70 year rotations) in boreal mixedwood forests, by simplifying the structure of young stands and reducing the frequency of older stands, will not maintain abundances of flora and fauna at levels found in unmanaged forests.

In recognition that conservation of ecological processes and biodiversity requires a landscape approach embodying both commercial forests and reserve networks (Thomas et al. 1990; Franklin 1993), this chapter describes four components of aspen mixedwood forest management (structuring forests, variable harvest rotation intervals, maintaining mixedwood forests, and mixedwood forest reserves) that require implementation. These practices, although important to correcting current problems with forest management, must be complemented by a societal philosophy that recognizes the value of stand structure and landscape forest patterns in maintaining biota prior to decisions involving allocations of trees to fiber production. Such a new approach differs markedly from previous annual allowable cut (AAC) calculations in Alberta that were based on merchantable landbase, wood growth and yield projections, and optimum "fiber" rotation age.

A new forest management paradigm receiving considerable attention in North America argues that managed forests should approximate stand and landscape patterns historically created by natural disturbance regimes (Hansen et al. 1991; Hunter 1993; Swanson et al. 1993). The basic assumption of this management model is that plants and animals are adapted to natural disturbances and are more likely to maintain viable populations if human land-uses create forest structures similar to those left by fire, insect outbreaks, floods, and windthrow.

Stand-replacing fires are considered the dominant disturbance agent of the boreal forest (Rowe and Scotter 1973; Van Wagner 1978; Johnson 1992) and are largely responsible for maintaining variance in stand age, size, shape, and structure (Zackrisson 1977; Eberhart and Woodard 1987; Engelmark et al. 1993). As such, commercial forests with structural variability similar to those of fire-dominated landscapes may offer a preferred risk management strategy for protecting both biota and ecological processes in Alberta's boreal mixedwood forests.

The considerable variance in plant community structures that occurred within and between seral stages, and the variety of forest community attributes to which abundances of wildlife and plants species were correlated, indicate that reliance on a "fine filter" species-by-species approach to conserving biodiversity is unlikely to succeed. Such a species-based approach will lead to conflicting objectives involving species with disparate habitat requirements. Opportunities to resolve conflicts and manage all species are further constrained because the majority of biota in mixedwood boreal forests are invertebrates, and many of these taxa have yet to be identified or their life histories described.

If the most salient features of aspen mixedwood forests are a diverse biota associated with a diversity of forest structures, forest management approaches that propagate appropriate frequencies of structures in space and time may hold the greatest promise for maintaining the integrity of this ecosystem.

The following recommendations are logical extensions of findings from this study and are consistent with the principles of forest ecosystem management.

1. Maintain Structure in Forests

Abundances of many species of plants, birds, and mammals were positively related to residual structures (green trees, snags, down woody material [DWM]) that originated from the previous stand cohort or were created by the stand-initiating fire.

Therefore, to maintain plant and wildlife communities in young forests, logging practices should more closely approximate the amount and variability of structures created or retained by fire.

Current wood utilization standards in Alberta, however, encourage clearcut logging that removes all merchantable tree biomass from cutblocks (Alberta Timber Harvest Planning and Operating Ground Rules 1994).

A presumed advantage of retaining structures on cutblocks is the reduction in years required before post-harvest stands acquire characteristics associated with old stands. For example, retaining green trees at harvest will lead to the production of large trees, and subsequently to large snags and canopy gaps earlier than would occur in a even-age stand of trees following unstructured clearcut logging.

Clearly, fire and clearcut logging differ in the amount of wood removed from the stand and in the structure of the new stands they initiate. Logging companies, however, are reticent to retain amounts of structure similar to those found in young stands following fire, as such a proposal would require a significant reduction in the annual allowable cut (AAC) of many companies (Maser 1994). A more moderate approach involves the modification of the cutblock utilization standards that direct logging practices. Rather than requiring the forest industry to remove all merchantable wood from cutblocks, regulations should encourage retention of structure in the form of live trees, snags, and DWM. Although we do not know the precise amounts of structure required on mixedwood cutblocks to maintain wildlife populations or ecological processes, our general understanding indicates that significant increases in retention of live trees, large snags, and DWM are warranted.

As a first approximation to maintaining the ecological function of young aspen mixedwood forests, managers should consider retaining snags and DWM in amounts similar to those of pre-fire origin measured on our young stands. These residual levels can be viewed as conservative estimates, as our young stands had 20-30 years since the fire event for residual green trees to become snags, for snags to fall and become DWM, and for DWM to rot. In young stands, snags greater than 10 cm DBH had mean densities of 19/ha (13 S.E.M), while coarse DWM volumes were ~50 m3/ha (12 S.E.M.). As abundances of both snags and coarse DWM were highly variable, it is important to maintain stand-to-stand variation in these structures.

Determinations of which stands are to be highly structured and which stands are to be lightly structured will depend in large part on the amount of residual trees, snags, and coarse DWM that exist at each cutblock at time of harvest. Where possible, the distribution of DWM on cutblocks should approximate the patterns found in young forests following fires. As such, localized concentrations of DWM, such as tall slash piles near landings of whole-tree harvesting systems, should be discouraged. Stump-side delimbing, where feasible, provides a mechanism for maintaining DWM on the cutblock. Where whole-tree harvesting systems are used, slash should be redistributed throughout the cutblock using skidders. In cutblocks where soil compression is a serious concern, slash could be re-distributed during the winter season.

Recently, forest companies have intentionally retained live trees and snags in clumps within cutblocks to approximate variation created by fire (Plates 35, 36, 37). The practice of clumping residual trees, although advisable, should not be implemented exclusively until the prevalence of clumped and dispersed residual materials have been measured, and their effects on forest flora and fauna assessed. If a goal of forest ecosystem management is to approximate the variability in green trees, snags, and DWM left by fire, then a variety of forest harvest strategies may have merit, including unstructured clearcuts, structured clearcuts, patch cutting, shelterwood harvest, and selection logging.

Constraints and Research Needs

Although our project provides a general insight into the importance of DWM, further research on which forest structures to retain at harvest, and in what abundances and locations, needs to be conducted to determine a fair compromise between the long-term ecological and economic sustainability of public forests.

Although patterns of retention for green trees and snags following fire may offer a general template for harvest planning, further research is required to indicate the most appropriate frequencies and locations of each attribute on the cutblock. Proper forecasting of these structures during succession in managed forests will require research on decay rates of snags and DWM, success of green tree retention, and fall down rates of snags in post-logged forests. Stand level research is also needed on optimal frequencies of snags required to maintain populations of cavity-dwelling birds and mammals, optimal levels of DWM required to maintain decomposition processes, nonvascular populations, and soil fertility, and optimal levels and size of live trees retained to produce a spatially heterogeneous canopy and canopy gaps.

Retention of live trees and snags does not constitute a safety hazard for mechanical harvesters in aspen forests where regeneration is vegetative (Steneker 1976), but standing snags are considered dangerous when manual felling is involved or when planting and tending operations occur subsequent to harvest (Occupational Health and Safety Act 1983). Further research is required to evaluate what types and placement of residual materials are required to maintain ecosystem health while securing worker safety.

However, ignoring the ecological importance of forest structures for reasons of human safety is not an acceptable rationale if such practices compromise both forest integrity and the ability of forests to provide wood products over the next several rotations.

Managers should understand that anthropic reconstruction of fire-related structures or "old-growthness" in managed forests will be of limited value if the ecosystem processes which maintain the integrity of forests are impaired through intensive forestry.

Other potential problems with reconstruction of stand structure include: 1) many forest community attributes covary and it is unclear which are most important to retain, 2) some forest components may be time dependent and can not be rushed, and 3) stand level treatments of structure may not provide adequate habitat unless landscape level treatments are also implemented.

2. Adopt Variable Rotation Ages for Forest Harvest

Forest companies in Alberta commonly manage the merchantable forest landscape with a narrow range of harvest rotation ages to facilitate a constant forest age class frequency that provides an even timber supply (Alberta Timber Harvest Planning and Operating Ground Rules 1994). The 60-70 year rotation that is used by the forest industry for Alberta's aspen mixedwoods approximates maximum annual increment (MAI) of aspen and was chosen to maximize wood production (Kabzems et al. 1986; Peterson and Peterson 1992). Although stands of this age have favorable live tree characteristics for pulp production, they are structurally simple (closed canopy, few large trees, snags, and DWM). In general, a commercial mixedwood forest landscape would have a truncated age class and have fewer of the structures that are associated with old stands. Although data confirming this pattern are unavailable for comparisons of unmanaged and commercial boreal mixedwood forests, many species of plants and wildlife in Pacific Northwest forests have been adversely affected by intensive forestry (Ruggiero et al. 1991; Thomas and others 1993). These studies raise concerns whether recurrent logging of 60-70 year rotations can maintain viable populations of plant and wildlife species whose preferred habitat is late successional aspen mixedwood forests.

One corrective approach to this issue is for the forest industry to de-emphasize their reliance on a narrow rotation interval and a balanced age class structure of the forest, and to select a range of rotation ages that more closely approximates the variability produced by natural fire return intervals.

Research on fire frequency in boreal mixedwood forests in Alberta suggests that fires were generally frequent with a short fire return interval and created a landscape where frequency of stands declined with increasing stand age (Murphy 1985). Based on this natural pattern, a smaller portion of the landscape could be harvested at long intervals while the majority of stands would be harvested at shorter return intervals.

The practice of rushing mixedwood forests through early successional stages, using softwood planting and tending techniques, may have adverse ecological effects. For example, our study identified shrubs and aspen saplings as equal to or more important than canopy and deadwood structures to which vertebrate species abundance was positively related. The low frequency of young (0-20 years) stands in the boreal mixedwood forests of northeast Alberta, possibly the result of fire suppression, may have caused shrubs and sapling stands to be limiting to some bird and mammal populations. However, high rates of logging on mixedwood forests using natural regeneration of aspen should significantly increase abundance of both shrubs and saplings during the next few decades.

Constraints and Research Needs

Far more research is required to quantify the spatial and temporal variance in fire return intervals and stand age before society and industry can evaluate the optimum landscape age structure that meets both ecological and economic goals. The use of a fire disturbance model as a basis for harvest rotation intervals is predicated on the assumption that logging will largely replace wildfire on the mixedwood forest landscape. Such a transition appears to have occurred in many U.S. forests (Agee 1994; Johnson et al. 1994), but fires in boreal mixedwood forests may be less controllable (Attiwill 1994).

3. Maintain the Mixedwood Forest by Adopting a Mixedwood Management Model

Many species of plants and wildlife are associated with stands containing both aspen and spruce, yet forest regulations may lead to the "unmixing" of hardwood/softwood stands.

Although monocultures occur naturally in the boreal mixedwood forest, most stand canopies contain both hardwood and softwood tree species (Alberta Forest Service 1985; Peterson and Peterson 1992). Current forest management regulations in Alberta require FMA holders operating on a mixedwood landbase to regenerate softwood volumes equal to those harvested (Alberta Timber Harvest Planning and Operating Ground Rules 1994).

The most common practice for spruce regeneration involves site preparation (usually scarification), manual planting of seedlings, and subsequent stand tending to reduce competition of seedlings with surrounding plants (Smith 1986). Because much of the spruce harvested from previously unharvested mixedwood forests are solitary or from small clumps <5 ha, it is a common practice for managers to aggregate spruce regeneration at larger spatial scales (20-40 ha; Alberta Timber Harvest Planning and Operating Ground Rules 1994). This softwood regeneration strategy, coupled with short-rotation aspen logging where site preparation for conifers does not occur, is likely to alter successional trajectories and lead to spruce stands and aspen stands that are separated spatially. On mixedwood cutblocks where conifer regeneration is not encouraged by silviculture, spruce densities may decline because few cone-bearing spruce trees may be present and/or appropriate germination substrate (large DWM) may be absent (Kabzems and Lousier 1992).

Thus, current silviculture practices encourage monodominant stand canopies, and may unmix the tree species that define the mixedwood forests (McDougall 1988).

"Unmixing the mixedwood forests" may jeopardize wildlife species and ecological processes reliant on a mixed-species tree canopy. For example, the relative abundances of approximately 25% of bird and mammal species in this study were positively related to densities of white spruce. Conifer trees, by providing forage, cover, or nesting resources, may allow some species to use mixedwood forests that otherwise would not.

Alternative silvicultural strategies that encourage natural regeneration strategies for spruce in aspen mixedwood stands, are required.

One option involves leaving some spruce trees to serve as current or future seed sources. Conifers retained during harvest may experience less wind-related mortality if they are positioned at the upwind border of the cutblock (Navratil et al. 1994) or if they are protected from wind by other conifers or aspen trees (Wasel, personal communication). In addition, because large DWM provides germination substrate for white spruce (Rowe 1955), large logs, snags, and live trees should be retained on cutblocks to ensure adequate amounts of germination substrate throughout forest succession. Finally, because white spruce seedlings often grow in exposed soils (Neinstaedt and Zasada 1990), forest companies might consider planting spruce seedlings on mixedwood cutblock access roads.

Constraints and Research Needs

Management of mixedwood forests for wood production from both aspen and spruce presents several challenges, including land tenure allocations, harvest involving multiple entries, and regeneration and growth of understory conifers.

Effective management of mixedwood forests will require government and industry to re-evaluate growth and yield estimates, silvicultural practices, and regeneration standards. Although untested in the field, strategies that rely on natural regeneration of mixedwood canopies may be economically viable because of reduced scarification, planting, and tending costs, and because of greater total yield from both aspen and spruce. Conversely, mixedwood harvest systems may not regenerate conifers quickly and may result in lower AAC for softwood operators.

The forest industry is unlikely to widely adopt natural regeneration of white spruce as a silvicultural option until researchers have quantified rates of natural regeneration and growth, evaluated DWM as a germination substrate, documented spatial relationships of seed trees, and better understand spruce mortality factors of spruce such as aspen, snowshoe hares, and marsh reed grass (Calamagrostis spp).

4. Establish Aspen Mixedwood Forest Reserves as Ecological Benchmarks

Although forest companies in Alberta are exploring "forest ecosystem management" (Franklin 1989; Hunter 1990) as a strategy for balancing commodity and ecological concerns, the long-term success of this approach remains uncertain (Stanley 1995).

As such, the establishment of unmanaged forest reserves of appropriate age, size, and composition within the forest landscape remains a prudent risk management strategy for conserving biodiversity and ecological function (Franklin 1993).

Because clearcut logging in Alberta's boreal mixedwood forests has a short history (most commercial forests are less than half way through their first rotation), we have no evidence that current forest practices are sustainable over long periods of time (i.e., multiple harvest rotations). If problems are encountered with maintaining viable plant and wildlife populations, forest regeneration, or ecological processes in forests managed for wood production, the scientific community will require a reserve system of natural forests to make comparisons and conduct experiments.

Because much of the aspen mixedwood forests has been allocated for fiber production, this plant community must become adequately represented in the boreal forest reserve system. The deletion of unmerchantable forest types, riparian buffer strips, and steep riverine slopes from the merchantable landbase (Alberta Timber Harvest Planning and Operating Ground Rules 1994) may provide significant environmental benefits, but at present there have been no evaluations of how adequate these other forest types are for conserving species and ecological processes characteristic of aspen mixedwood forests.

The diversity of ages, sizes, shapes, and juxtaposition found in boreal mixedwood forest landscapes (Peterson and Peterson 1992) emphasizes the need for a reserve system that maintains reasonable frequencies of natural stands spatially and temporally positioned in the landscape (Baker 1992).

Such reserves, particularly old growth forests, may also offer cultural and aesthetic values to society (Burton et al. 1992; Maser 1994), maintain core populations of threatened species, and produce dispersing individuals or propagules that may be important to the maintenance of plants or wildlife populations in the surrounding managed landscape (Franklin and Forman 1987; Noss 1991).

Finally, reserves represent ecological benchmarks against which society can evaluate the integrity of forests managed for commercial land-uses such as forestry.

Constraints and Research Needs

Central to the issue of sustainable forestry is whether high utilization levels (e.g., conventional clearcut logging) and short rotation ages (60-70 years) can be sustained over multiple harvests. Answering this question will require lengthy research projects involving growth and yield monitoring from permanent sample plots (PSP) in both harvested and reserve systems. Such monitoring programs will provide government and industry with an ongoing mechanism to assess the viability of harvesting systems. Forest managers need to be able to detect small changes in site productivity, as minor losses to plant or wildlife productivity with each rotation may have profound consequences to forest systems when considered over meaningful ecological time (100s-1000s of years).

To be effective as ecological benchmarks, reserves need to be large enough for natural disturbances to occur in appropriate size, frequency, and intensity (Pickett and Thompson 1978; The Nature Conservancy 1982; Noss 1987; Baker 1992). Allowing natural levels of wildfire may not be possible within small reserves surrounded by land-use practices that view natural disturbances unfavorably. A protected area network based on large areas where conflicts with competing land-uses are minimal may be more likely to support natural disturbances of appropriate type and frequency. Large, "semi-natural" boreal mixedwood forests in western North America include Wood Buffalo National Park and Primrose Air Weapons Range in Alberta and Prince Alberta National Park in Saskatchewan. Even within these areas, however, natural fire regimes may not be present as fire suppression and other land-uses occur. Further discussion is required on whether reserves should be representatively established within the boreal mixedwood biome, embedded within the commercial forest, or both. Because the degree of ecological similarity between harvested and reserve forests may diminish with increasing distance, appropriate attention to geographical positioning of reserves in the landscape matrix is required (Franklin 1993).

It would be unwise to rely solely on reserve networks to conserve biological diversity, however, as reserve systems of sufficient scale to embody the range of natural variability of landform, disturbance events, biota, and ecological processes may not be attainable (Baker 1989; Wilcove 1989). Reserve networks should therefore be viewed as complementary to harvest strategies involving structured cutblocks, multiple rotation ages, and mixedwood management.

The degree of similarity between natural lands and human land-use practices, however, should influence the area of reserves required (Harris 1984; Franklin 1993). By embedding reserve networks within a matrix of land-uses that adopt ecological principles, it may be possible to create a landscape useful for both commodity production and conservation of biodiversity (Harris 1984; Hansen et al. 1991; Mladenoff et al. 1993). In landscapes manipulated intensively (e.g., conventional clearcut logging), a large amount of the area should be established as reserves to serve as ecological benchmarks and to compensate for problems caused by intensive forestry.

In contrast, forestry manipulations that are less intensive (e.g., structured clearcuts, selection logging) and that better approximate natural disturbances in intensity, size, and frequency, may offer less ecological risk and therefore require less area allocated to reserves.

Obstacles and Opportunities

Albertans are beginning to realize that consideration of ecological issues after fiber allocations have been made and after forest industry infrastructures are complete, severely compromises our ability to find solutions that offer both economic and ecological sustainability. A new forest management model, one that addresses both ecological and forestry concerns at appropriate spatial and temporal scale, is required. However, the remarkable complexity of aspen mixedwood forests, and our rudimentary understanding of this ecosystem, have been used by different interest groups to either justify the abolishment of commercial forestry or to suggest that it is premature to change current forest harvest strategies. A more reasonable approach to directing forest management practices is offered; a model based on maintaining variability in stand age, stand size, and stand structure following patterns created by such natural disturbances as fire.

The implementation of this model will require extensive scientific study of boreal mixedwood forests and an extension program that informs Albertans about issues relating to both the ecological and economic consequences of forest land-use practices. The implementation of this model will be difficult, however, as historic approaches to timber allocations have not adequately considered non-fiber attributes. Forest structure that has been historically allocated for timber or pulp may now be required as components of wildlife habitat or ecological processes. For these reasons, further allocations of the few remaining uncommitted public forests in Alberta is inadvisable and not consistent with current national and international directions in forest stewardship. In the context of forest ecosystem management principles, these unallocated forests are required as unlogged reserves, and as areas where existing forest harvest operations can expand to provide the latitude needed to implement alternative silviculture practices, structured cutblocks, a mixedwood management model, and variable rotation ages.

Studies such as the one presented in this report have repeatedly shown strong relationships between stand age, stand structure, and biota. The patterns that define these relationships can be used by resource managers to mitigate adverse effects of forestry through modification of operational practices (e.g., structured clearcuts, multiple rotation ages, mixedwood management models). Although this approach to understanding forest ecosystems, and how they respond to landuses, has merit, it must be complemented by research that focuses on fundamental ecological processes. The range of ecological processes affected by intensive forestry is far from fully known (Maser 1994), but includes nutrient cycling, soil fertility, plant germination, and competitive interactions of organisms. In particular, more research should be directed to invertebrates and fungi (Hawksworth 1991; Olson 1992), because these organisms may be critical to tree germination, growth, and mortality, and to nutrient cycling (Trappe and Luoma 1992; Franklin 1993).

Although commercial forestry is presently foremost in the public's mind as a threat to forest ecosystems, the effects of the oil and gas sector and agriculture should not be underestimated. Both of these landuses excise considerable areas, either permanently or semi-permanently, from Alberta's boreal forest each year, and cause extensive fragmentation to the remaining forest landscape.