1M '1’? Er'sfl'TL. icon This is to certify that the thesis entitled DOES CLEARCUT HARVESTING EMULATE THE EFFECTS OF NATURAL DISTURBANCE ON THE DEVELOPMENT OF STAND STRUCTURE IN PINUS BANKSIANA FORESTS OF NORTHERN LOWER MICHIGAN? presented by Susan Spaulding has been accepted towards fulfillment of the requirements for the Master of degree in Forestry Science ,7w ‘57 Major Professor’s Signature ’ /0-/7‘- 0;; Date MSU is an Aflinnative Action/Equal Opportunity Employer PLACE IN RETURN BOX to remove this checkout from your record. To AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE 5/08 KzlProj/Achres/CIRC/DaIeDuevindd DOES CLEARCUT HARVESTING EMULATE THE EFFECTS OF NATURAL DISTURBANCE ON THE DEVELOPMENT OF STAND STRUCTURE IN PINUS BANKSIANA FORESTS OF NORTHERN LOWER MICHIGAN? By Susan Spaulding A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree Of MASTER OF SCIENCE Forestry 2008 ABSTRACT DOES CLEARCUT HARVESTING EMULATE THE EFFECTS OF NATURAL DISTURBANCE ON THE DEVELOPMENT OF STAND STRUCTURE IN PINUS BANKSIANA FORESTS OF NORTHERN LOWER MICHIGAN? By Susan Spaulding I used two chronosequences to study the structural development of jack pine (Pinus banksiana) stands in northern lower Michigan following either clearcut harvesting and replanting or stand-replacing wildfire. The chronosequences were used to quantify 1) aboveground biomass, 2) stem density, 3) snag density and volume, 4) coarse woody debris volume, and 5) forest floor mass, and within—stand patchiness. Total aboveground biomass and stem density in the harvest sites closely resembled that of the wildfire- regenerated sites across all stand ages. In contrast, within-stand patchiness Of live stem density was much higher in the wildfire sites for the first 20 years of stand development. Harvest sites possessed very little dead wood at young ages, but snag volume and CWD gradually increased. Young wildfire sites contained the largest amounts of dead wood; those levels declined dramatically in the first 20 years. Both chronosequence recovery patterns merged by 40 years post-disturbance where they began a steady increase in total dead wood. Forest floor mass was very high in young harvest stands compared to young wildfire stands, but by approximately 10 years there is no longer a discemable difference. The differences between the chronosequences were driven mainly by recovery in the youngest (<20 yrs) sites, after which many of the attributes were very similar. This has important implications for management that is intended to replicate natural recovery across stand development, and not just at stand maturity. TABLE OF CONTENTS CHAPTER] LITERATURE REVIEW: STAND STRUCTURAL DEVELOPMENT AND MANAGEMENT IN BOREAL FOREST SYSTEMS DOMINATED BY STAND Introduction” ....1 Stand Structure . .........1 Disturbance Dynamics and Stand Development In BOreaI FOrest Systems ............ 2 Disturbance In the boreal forest .................................................................. 2 Recovery after fire in Pinus banksiana forests ........................................... 3 Management that Emulates the Natural Fire Regimes in Wildfire Systems .......... 4 Goals and Mechanisms" 4 Benefits of replacing wildfire with harvest .................................... .5 Challenges of replacing wildfire with harvest .................................. 6 Northern Lower Michigan Jack Pine System7 Managing for Kirtland's warbler........................ 8 Management Implications ........................................................... 8 Rationale .......................................................................................... 9 CHAPTER 2 DOES CLEARCUT HARVESTING EMULATE THE EFFECTS OF NATURAL DISTURBANCE ON THE DEVELOPMENT OF STAND STRUCTURE IN PIN US BANKSIANA FORESTS OF NORTHERN LOWER MICHIGAN? ........................ 10 Introduction .................................................................................... 10 Materials and Methods ..................................................................... 14 Study system and experimental design .......................................... l4 Plot layout and field measurements...............................................15 Data analysis .......................................................................... 18 Results ........................................................................................... 19 Standing biomass .................................................................. 19 Snags ................................................................................. 20 Coarse woody debns2l Forest floor ............................................................................ 23 Discussion ..................................................................................... 24 Management Implications ................................................................. 30 iii LIST OF TABLES Table 1. Regression statistics for changes in stand structure over wildfire- and harvest- origin chronosequences. Linear or polynomial models predict each structural parameter (y) as a function Of time since stand-replacing disturbance (x). The first column of probability values are for the lack-Of-fit hypothesis test for different regression models between wildfire and harvest chronosequences. Where the null hypothesis Of significantly different regression equations was accepted, statistics for a single, pooled model are presented. Parameters for which the null hypothesis was rejected are noted in bold, and statistics for treatment-specific models are presented .......................................... 41 Table 2 Stand attributes. Age refers to the age of the stand in 2005. % Jack pine refers to jack pine basal area/total stand basal area ..................................................................... 44 iv LIST OF FIGURES Figure 1. Locations and year of stand origin Of study sites in northern Lower Michigan. Closed circles indicate harvest origin and Open circles indicate wildfire origin ............ 33 Figure 2. Snag density (A) and volume (C) Of jack pine and patchiness Of snag density (B) and volume (D) along chronosequences of both harvest and fire origin. Symbols are as described for Figure 2 .................................................................................................... 34 Figure 3. Aboveground biomass (A) and stem density (C) of jack pine and patchiness Of aboveground biomass (B) and stem density (D) along chronosequences of both harvest regime and fire origin. Open circles represent fire origin stands and closed circles represent harvest regime stands. Dashed lines represent regression lines for fire-ori gin stands, solid lines regression lines for harvest-ori gin stands, and dotted lines represent pooled regression lines ....................................................................................................... 35 Figure 4. Density Of snags >5cm dbh (A) and patchiness of snag density for snags >5cm dbh (B) along chronosequences of both harvest and fire origin. Symbols are as described for Figure 2 ....................................................................................................................... .36 Figure 5. Coarse woody debris volume (A) variation of coarse woody debris volume (B) along chronosequences of both harvest regime and fire origin. Symbols are as described for Figure 2 ........................................................................................................................ 37 Figure 6. Volume of coarse woody debris along chronosequences Of both harvest regime and fire origin for decay classes 1 (A and B), 2 (C and D), 3 (E and F) and 4 (G and H.) Symbols are as described for Figure 2 ............................................................................. 38 Figure 7. Total dead wood volume (A) and variation of total dead wood volume (B) along chronosequences of both harvest regime and fire origin. Symbols are as described for Figure 2 ........................................................................................................................ 39 Figure 8. Forest floor mass (A) and variation Of forest floor mass (B) along chronosequences of both harvest regime and fire origin. Symbols are as described for Fig 3 .......................................................................................................................................... 4 Chapterl: Literature Review: Stand Structural Development and Management in Boreal Forest Systems Dominated by Stand-replacing Fire Regimes Introduction Understanding structural dynamics in forest stands managed for specific ecological goals is crucial for understanding the effects of the management plan on the ecosystem, and for quantifying the success of the plan. Structure is of particular importance because of its effects on temperature (Binkley 1984), moisture (Tanskanen et a1 2006, Simard et al 2001), light availability (Hart and Chen 2006), soil formation (Prescott and Laiho 2004, Tinker and Knight 2000), plant species composition (Houseman and Anderson 2002), and animal Species composition (Converse et al 2006, Lindenmayer and McCarthy 2002, Carey and Harrington 2001). Stand structure is affected by many disturbances, both small— and large-scale, including insects and other pathogens, wind damage, and especially in boreal systems, fire (Attiwill 1994). In addition, stand structure is highly variable throughout stand development, so it is important to understand stand dynamics over time. Understanding the effects of management on structure in forests is critical; this knowledge can be used for accomplishing long-term management goals that take into account broad ecological interactions. Stand structure Stand structure describes the vertical and horizontal physical components of a stand, both living and dead. These include live herbaceous and woody plants at all stages Of their life cycle, and dead wood in the form of standing dead trees (snags), downed dead wood (CWD), and smaller pieces of plant matter that make up the forest floor. The vertical component of stand structure is the combination of volume, mass and density of plant material. The horizontal component of stand structure is the spatial distribution, or patchiness, of those measurements. Patchiness is important in a system for two main reasons: 1) It provides a mix Of foraging, nesting, and resting habitat for wildlife species, and 2) Increased diversity of habitats can support more species over time (Vanbergen et a1 2007, Lindenmayer and McCarthy 2002). Disturbance Dynamics and Stand Development in Boreal Forest Systems Disturbance in the boreal forest In boreal forests, the main types of natural Stand disturbance are insect or other pathogen damage, wind damage, and fire (Dordel et al 2008, Schulze 2005, Attiwill 1994). Some sources of disturbance, like armillaria root rot (caused by the fungus Armillaria ostoyae), can create small patches of dead or damaged trees a few acres in size (Worrall 1994). Others, such as outbreaks of infestation by mountain pine beetle (Dendroctonus ponderosae) and wildfire, frequently destroy thousands of acres per event. By killing trees and creating openings, all of these types of disturbance cause changes in the forest structure (Jenkins et al 2008, Wright et al 2002, Tinker and Knight 2001 , Sturtevant et a1 1997). The boreal forest system is dominated by fire-adapted trees that utilize fire as the main mechanism of regenerating new stands. Some trees, like jack pine (Pinus banksiana) and the closely related lodgepole pine (Pinus contorta) are especially good at promoting fire and reestablishing new stands after a fire. These trees hold on to their lower dead branches, creating a "fire ladder" that facilitates the spread of a fire. The trees also have a high resin content, which makes them highly flammable. Their cones are mainly serotinous, and require exposure to flame to open and release seeds (deGroot et a1 2004). In addition they grow very quickly, and can take advantage of openings created by fire faster than most other species. Recovery after fire in Pinus banksiana forests Stand-replacing fire has immediate effects on structure in stands dominated by jack pine. While most of the trees are killed, the majority of the aboveground biomass remains intact; mainly the foliage and smaller branches are consumed by fire, leaving most of the bolewood behind (Brais et al 2005). Many of the fire-killed trees fall immediately or within days of the fire, becoming CWD (Gore 1985). The remaining dead trees form a forest of snags (Hutto 2006). Forest floor may be partially or completely consumed, depending on the speed and intensity of the fire. The aboveground portion of plants in the understory is generally consumed, but the root system of many plants survives (Hart and Chen 2006, Degrandpre et al 1993). Existing CWD on the ground also burns, often smoldering longer than the surrounding finer branches and creating "log burnout" areas (Rhoades et a1 2004); these areas are sections of forest floor that burn hotter during the fire, resulting in death of all surrounding plants and loss Of all the organic matter in the soil, leaving only mineral soil behind. Fire also has long-term impacts on forest structure in these systems. The standing forest of snags becomes a source of CWD as it falls (Wright et a1 2002, Tinker and Knight 2001). In addition, smaller dead branches from these snags contribute directly to the forest floor (Brais et al 2005). Immediately after the fire and for the first few years, most of the CWD belongs to decay classes 1 and 2, the most intact groups. Over time, this initial pulse of CWD decomposes into decay classes 3 and 4 and eventually becomes soil organic matter (Brais et a1 2005). Germination of new jack pine seeds begins soon after fire, primarily in the first few years. The quick germination and rapid growth of jack pine give it an advantage over other tree species that might germinate after a fire. Management that Emulates the Natural Fire Regimes in Wildfire Systems Goals and Mechanisms There has been a shift in recent years toward management strategies that more closely emulate natural ecological regimes (Lindenmayer et a1 2006, Hutto 2006, Bergeron et a1 1999, Angelstam 1998, Hunter 1993). This thinking follows the assumption that conditions that are closest to the natural conditions are most conducive to 1) maintaining the natural biota and 2) long-term sustainability. This type of management is typically used in systems where the natural history of the region is contrary to human interests, for example, areas where large wildfires frequently occur. The goals of this type of management are to accomplish objectives " 1" and "2" from above while simultaneously addressing human concerns and forest resource needs. Management plans that emulate natural systems seek to maintain the following components within natural boundaries: 1) Species composition, 2) Rate of ecosystem processes and functions, and 3) Structure. Aiming to fulfill all of these together is the best way to create a sustainable plan (Lindenmayer and Recher 1998, Attiwill 1994). In fire-prone systems in which wildfire is unacceptable, there are two current methods of maintaining a forest ecosystem that is similar to the natural system: 1) prescribed burning, and 2) harvest as a surrogate for fire followed by replanting. Prescribed burning as a management strategy is difficult to implement in many forest systems for several reasons. First, there is no financial benefit to prescribed burning, in fact there is a perceived financial loss due the opportunity cost of not harvesting the stand. Also, the cost of prescribed burning is great due to the extensive manpower, training, and physical resources required for implementation of a controlled burn. Lastly, fire is a potential risk to human property and in extreme cases, human life. For these reasons, harvest followed by planting is often the preferred plan. Benefits of replacing wildfire with harvest Replacing a natural wildfire regime with harvest and planting is a successful strategy on many levels. Harvesting large stands of mature jack pine in an area reduces the risk of wildfire in that area by reducing the fuel load. This in turn reduces the risk to nearby intact stands as well as other property (Chapin et al 2008). Reducing the risk of fire also reduces the danger to firefighters who would try to control the wildfire. Harvest is financially beneficial, as jack pine is a species commonly used for lumber and pulpwood. In addition, harvests and plantings can be planned and controlled. There is no ambiguity about when planted stands will be ready for harvest, and it is possible to create stands in whatever combination of stand dimension, age, species and density you require for your management objectives. It is also easy to monitor the effects of this management strategy, since stands may be assigned to different versions of a harvest and planting plan. Challenges of replacing wildfire with harvest One of the challenges of tailoring a management plan to emulate a natural system is deciding which parameters of the original system are most important to reproduce. There are some key differences between harvest and fire. First, wildfire is stochastic, sometimes killing all trees in an area, sometimes leaving pockets of living trees; fire also has irregular borders (Kashian et al 2004). Depending on fuel load and weather conditions, fire may consume all or part of the forest floor and understory. Generally, fires later in the summer (when conditions are drier and fuel loads are higher) burn the forest floor more completely, exposing more mineral soil and creating more germination sites for seedlings (Kembell et a1 2006). Typically, jack pine stands regenerated by fire have some combination of treeless openings and forested areas of varying density. This is difficult to replicate in a harvested system, which tends to be much more regular. The forest of snags and high levels of CWD created in jack pine stands that have recently burned are very different from levels in a harvested stand (Sturtevant et al 1996, Hansen et al 1991,). Instead, a harvest removes most of the wood, leaving behind only smaller pieces of slash and sometimes a few existing snags (Tinker and Knight 2001). Snags have been identified as a key habitat element in forested systems, especially for cavity-nesting birds (Hutto 2006). CWD serves as habitat for small mammals and invertebrates, and contributes to soil development as a source of nutrients and organic matter (Brais et a1 2005). Management activities that change levels of dead wood in these systems can affect all of these parameters. Lindenmayer and McCarthy(2002) compared structural responses in a forest system containing stands naturally regenerated by fire and regenerated by harvest and planting. There were fewer, smaller pieces of dead wood, both standing and down in the harvested stands. Harvested systems also lacked the older living trees that sometimes survive a fire. Hansen (1991) found that natural forest disturbance maintains structural complexity, promoting plant and animal diversity. Both studies found that intensively managed forest systems are skewed toward younger stands and contain very few (if any) old living trees. Increasingly, recommendations are being made to adapt harvest plans to safeguard these existing natural elements and to restore natural attributes that have been lost (Bergeron 1999, Angelstam 1998, Hunter 1993). Northern Lower Michigan Jack Pine System The highly managed jack pine system of northern lower Michigan provides a unique opportunity to observe the effects of a management plan that emulates a natural fire regime. This system is unique because it provides the only significant breeding habitat for the Kirtland's warbler (Dendroica kirtlandii), a federally-endangered bird. Because of this special designation, this area has been intensively managed with the primary goal of maintaining habitat for this bird (Probst et a1 2003). The publicly managed land in this system is interspersed with homes and private property, so wildfire suppression is highly desirable. The fire return interval estimates in this system before settlement range from 30-80 y (Simard and Blank 1982, Whitney 1986). This has been replaced with a harvest rotation followed by planting of 50-60 y. Managing for Kirtland's warbler The Kirtland's warbler nests in very dense young stands of jack pine approximately 1.4 m tall (Probst and Weinrich 1993). They build their nests at the edge of openings in stands over 80 acres in size. The management plan in this system has incorporated these requirements and implements large-scale harvest planted with regular openings within the stand and very densely planted jack pine trees. Since its implementation, this strategy has been very successful at increasing the number of breeding warblers in the area (Solomon 1998). Management Implications There is a dichotomy in modern conservation biology between managing for preservation of a single species (for example, an endangered species or an indicator species) and managing for broader ecosystem characteristics (Lindenmayer and N038 2006, Simberloff 1998). Single species management makes the assumption that management that preserves the species of interest will automatically preserve additional characteristics of the ecosystem that are most important. The ecosystem approach acknowledges that there are numerous species in an ecosystem, even those that are common, that perform critical functions and must also be preserved. Single species conservation is attractive because it has been very successful at increasing populations of endangered species (Probst et a1 2003). It is also easier to implement by focusing on the requirements of and effects on a single species than by coordinating full ecosystem biodiversity. Certainly, aggressive plans to preserve endangered species are sometimes necessary for the short-term goal of preventing imminent extinction; however, the additional ecosystem effects Of such a plan often go unmeasured. Any single-species management plan needs to consider the large-scale ecosystem effects and issues of sustainability. Franklin (1993) advocates conversion to management that includes landscape and ecosystem approaches, as well as an approach that addresses a wide array of spatial scales. Rationale The objective of this study is to examine temporal patterns of stand structure in a system managed for a single species, the Kirtland's warbler, and to compare this to structural development under a natural disturbance regime. The results of this study will provide insight into ecosystem-level effects of replacing fire with harvest in this system and how well it replicates natural forest structure. Chapter 2. Does clearcut harvesting emulate the effects of natural disturbance on the development of stand structure in Pinus banksiana forests of northern Lower Michigan? Introduction According to Lindenmayer and Recher’s (1998) definition of sustainable forest management, structure is one of three important components that need to be maintained within the bounds of normal disturbance regimes, along with species composition and the rate of ecological processes and functions. Of particular interest is whether management that emulates natural disturbance regimes will allow sustainable forest management without long-term degradation or disruption of the system (Attiwill 1994). Forest structure is an ecosystem property that is highly affected by disturbance and is critical because it, in turn, affects many ecosystem elements, including herbaceous plants (Houseman and Anderson 2002, Abrams and Dickman 1982), insects (Heliola et a1 2001), birds and mammals (Converse et al 2006), and nutrient cycling (Brais et al 2005, Laiho and Prescott 1999). Stand Structure is the vertical (volume, mass and density) and horizontal (spatial distribution) pattern of all the dead and living components of the stand. Together, these components form the internal shape of a stand, comprised of patterns of size, density and distribution of live and standing dead trees (snags) and amounts and patterns of coarse woody debris (CWD) and forest floor detritus. A single stand may contain some openings, some very dense areas, and both large and small diameter trees of varying heights; or the stand may be very homogenous, containing similarly-sized trees of the same age and size with no openings. Quantifying the structural attributes of a stand can 10 help us predict its future development, as well as understand current and future habitat suitability for many species. These elements of forest structure undergo dramatic changes in composition and volume as a result of a major disturbance (Brassard and Chen 2006, Harper et al 2006, Kashian et a1 2004, Wright et a1 2002, Tinker and Knight 2001). Some structural changes due to disturbance are immediate, such as reduction of woody biomass during a harvest; while other effects occur over time, such as the reallocation of snag biomass to the forest floor as CWD when trees fall in the years following a fire. Therefore, it is important to examine structural changes throughout stand development to understand disturbance effects. Patterns of Standing biomass, snags, CWD and forest floor detritus have strong influences on the biota within a stand. Areas of dense vegetation offer cover for small and large mammals. Conifer snags can hold seed-bearing cones for several years after death and provide a seed source, as well as shade, for new recruits. Snags are also very important as habitat for numerous bird species and small mammals, especially larger snags. CWD provides habitat for numerous small mammals and insects, and as it decomposes over time, CWD adds nutrients to the soil, and contributes to soil development (Laiho and Prescott 2004, Tinker and Knight 2000). Forest floor detritus holds moisture and nutrients (Simard et a1 2001), provides sites for mineralization of N (Hazlett et al 2007, Westbrook et al 2006, Yermakov and Rothstein 2006), regulates temperature (Matsushima and Chang 2006), and provides habitat for small mammals, insects and other invertebrates (Hannam et a1 2006). In addition, habitat complexity is highly correlated with total relative abundance of small mammals in the forest understory (Carey and Harrington 2001). 11 Development of Silvicultural systems that mimic natural disturbance, under the assumption that the biota of a forest are adapted to a natural disturbance regime, is particularly challenging in areas where even-aged stands of shade intolerant trees were historically maintained by stand-replacing wildfires (Franklin et al. 2002). Clearcut harvesting followed by seeding or planting has been the traditional approach to managing forests characterized by a severe fire regime, and Should aim to replicate natural systems by emulating disturbance frequency, size and distribution, and residual organic matter (Hunter 1993, Bergeron et al. 1999). While a good surrogate for fire in many ways, harvesting also has some notable differences. Fire is a stochastic process, Often resulting in spatially irregular patterns of regeneration in the new stand (Kashian et al 2004). Post- wildfire stand composition can vary as a function of these irregularities, creating pockets Of very dense trees adjacent to large areas with no trees at all (Charron and Greene 2002). In contrast, clearcutting followed by planting creates a uniform stand, with evenly spaced recruits in regular rows. Furthermore, while wildfire and harvesting both kill trees, only a small fraction of aboveground biomass is actually consumed by fire (primarily foliage and small branches; Stocks 1989). Fire leaves behind large amounts of legacy structure in the form of CWD and snags (Franklin et al. 2002), while whole-tree harvesting removes virtually the entire overstory, leaving little legacy structure in young stands. Prior to European settlement, jack pine (Pinus banksiana) forests of northern Lower Michigan were regenerated by stand-replacing fires with return interval estimates ranging from 30-80 y (Simard and Blank 1982, Whitney 1986). This historic disturbance regime has been largely replaced by one of intensive harvesting and planting. An unusual aspect about this system is that the driving goal behind this intensive 12 management is conservation of biodiversity - not maximization of timber production. These forests provide the only significant breeding habitat for the federally endangered songbird, the Kirtland’s Warbler (Dendroica kirtlandii). The Kirtland’s Warbler requires dense Stands of young jack pine approximately 1.4m tall (Probst and Weinrich 1993) for breeding and nesting. Forest managers actively suppress wildfires but employ whole-tree clearcutting and planting on a 50-60 year rotation as a means of regenerating young jack pine stands. This shift to a harvest-dominated disturbance regime has been a great success in terms of promoting warbler populations (Solomon 1998); however, it raises the possibility that intensive management, narrowly focused on maximizing suitable habitat for a single endangered species, could have the unintended consequence of decreasing overall diversity through simplification of forest structure. Because some structural changes due to disturbance are immediate (e.g. reduction of live biomass), while other effects occur over time (e.g. reallocation of dead wood from snags to CWD), it is important to examine structural changes throughout stand development to fully understand the effects of a shift in disturbance regime. I set out to investigate the consequences of disturbance regime changes in northern Lower Michigan jack pine forests by examining stand structural features along two, parallel chronosequences: one of stands regenerated by wildfire the other by harvesting and planting. 13 Materials and Methods Study system and experimental design In order to investigate dynamics of forest structure as stands develop following harvesting vs. wildfire, I developed two, parallel chronosequences, one of wildfire-origin stands and one of plantations ranging in age from 4-69 years since stand-destroying disturbance. The stands were located within four counties in northern lower Michigan (Crawford, Oscoda, Roscommon and Ogemaw) across an area of approximately 130 km2 (Figure 1). Potential stands were identified using information on fire and harvest history of public lands through the United States Department of Agriculture Forest Service (USDA-F8) and Michigan Department of Natural Resources (MDNR). Only stands classified as pure jack pine were included. Harvest stands were only included if the harvest was followed by planting of jack pine. Stands with records that indicated regeneration by seed or additional treatments after harvest such as roller chopping of slash were excluded. Potential stands were then visited and eliminated if any of the following were discovered: I) Stand had been partially harvested, 2) Preceding stand was not jack pine 3) Stand had been salvage-logged after fire, 4) Fire had not been completely stand-replacing (for example, strips of unburned original stand remained) 5) Soil cores taken to two- meter depth revealed soil type other than sand (for example, clay bands or gravel layer), 6) Predominance of species indicating higher water or nutrient availability, such as trembling aspen (Populus tremuloides), 7) Stand had significant topographical variation, 8) Stand was other than described in the records regarding my original criteria for stand selection. Planting as the mode of stand origin was confirmed by the presence of furrows 14 and trees growing in rows. Fire as the stand origin was confirmed by the lack of furrows or tree rows, and also by the presence of charcoal. If there was any ambiguity about the origin of a stand, the stand was excluded. By using these standards, I hoped to minimize variation in factors other than stand age and origin that could confound my analysis. Stands for this chronosequence comparison study were selected with the goal that the only differences among individual stands would be mode of origin and age. An important, potentially confounding factor that could not be controlled is that the harvest and planting regime employed during the years of origin of the three oldest plantations (1936-1955) is different than that of the more recent stands. The current management strategy put into place in the 19803 to promote Kirtland’s warbler calls for very dense stands (approximately 1452 trees per acre) with regular openings (1/4 acre opening for every acre planted); this differs from less dense stands planted previously for timber that have no planned Openings (Phil Huber, USDA Forest Service Wildlife Biologist for the Mio Ranger District). Exact planting densities for stands planted prior to warbler management are unavailable, but are estimated between 681 and 1210 trees per acre. I included these older stands because they were the only option to provide insight into later-stage stand dynamics; however, data from these stands need to be interpreted with caution. Plot layout and field measurements Once a stand was accepted, an approximately 10 ha square site was located within the boundary at least 50 m from any edge of the stand and on a north/south axis. The shape of three of the stands required fitting the site along a northwest/southeast axis with a 15 Slightly rectangular shape (1950 fire, 1936 harvest and 1999 fire; Figure 1). Stratified random sampling was used for plot selection within each site to assign 20 locations for sampling plots. Four equal quadrats were created in each site and each quadrat was assigned 5 plot centers using a random number table for selection of X and Y coordinates on a UTM grid. Each plot was circular and 8m in diameter. In the event that two plot boundaries overlapped, a new plot center was randomly chosen for the second plot. Some plots in the harvest sites fell within openings, some fell within planted rows, and some overlapped both; all sites were included in analysis. Within each plot I measured the diameter at breast height (DBH, 1.37m) and total tree height (using a laser hypsometer) for all standing trees, live and dead greater than 2m high, and I counted and measured the height Of all tree seedlings and saplings less than 2 m in a 1/4 section of each plot. I measured the length and diameters at each end for all pieces of CWD larger than 5 cm in diameter on at least one end. For pieces of CWD that extended outside of the plot, I measured length and diameter at the plot border. I assigned each piece of CWD to one of four decay classes: I) Newly downed wood with most or all bark still present, 11) Debris with sloughing bark that held its original shape and was structurally sound (could not be crushed), 111) Wood that still held its shape but could be crushed, and IV) Wood that had lost its structure and was variously flattened. For decay classes [-111 one diameter measurement was made with calipers at each end, whereas for decay class IV, two orthogonal measurements were taken at each end, a height and width. These two measurements were averaged to estimate diameters for volume calculations. To calculate volume, I used the formula for the frustrum of a cone (Eq. 1): l6 volume = 1/3J'El (r2 + rR + R2) [1] where l is the total length of the piece, r is the radius of the small end, and R is the radius of the large end (Robertson and Bowser 1999). To quantify forest floor mass, I placed a square metal frame (37 X 37 cm) on the south edge of each plot, and collected all recognizable plant litter (Oi + Oe horizons) within the frame. Samples were initially air- dried in a warm greenhouse until they could be oven-dried at 65 degrees C for at least 48 hours and weighed. The weight of each forest floor sample was then scaled to a kg per ha estimate. Aboveground biomass of jack pine trees was calculated using the allometric biomass estimator determined by Perala and Alban (1994; Eq. 2): W = 0.3837*D1'95"‘H'8369 [2] where D is dbh in cm, H is height in m, and the constants 1.95 and 0.8369 are specific to upper Great Lakes jack pine. Snag volume was determined by first calculating amount of taper in dead wood (using the CWD measurements of decay classes 1 and 2 from all plots in all sites combined; Eq. 3), y=0.0122x [3] where y is the difference in diameter at each end of a piece of CWD (cm) and x is the length of the piece in cm. The same equation was used for snags in all sites. This taper calculation was used to calculate the diameters at the base and apex of each snag from the measured DBH and total height. The volume of each snag was calculated using the formula for volume of the frustrum of a cone (see Eq. 1). 17 Data analysis For each structural parameter examined (standing biomass, stem density, snag density, snag volume, CWD volume, CWD mass, total dead wood volume and forest floor mass), I calculated stand level means based on measurements from all 20 plots within each site as metrics of vertical structure. I used the coefficient of variation (CV) for each parameter, calculated from the 20 plots within each stand, as a metric to describe horizontal structure, or within-stand spatial heterogeneity. I used polynomial regression to compare changes in vertical and horizontal structure over time, and between wildfire and harvest chronosequences. I used lack—Of-fit tests (Neter et al. 1990) to choose between linear, quadratic and cubic regression functions for each parameter. I also used a lack-of-fit approach to test the null hypothesis that the relationship between each structural parameter and stand age did not differ between wildfire and harvest chronosequences. In this case I used the lack of fit test to compare sum-of-squares error of two separate regression equations vs. a single, pooled equation. In the analysis of coefficient of variation of aboveground biomass, two fire sites were excluded (stand origin 1998 and 1999) because there were no standing trees >2m in height on any of the plots within the sites. One harvest regime site (stand origin 1998) was excluded from the regression of coefficient of variation Of snag density (Figure 2B) and snag volume (Figure 2D) because there were no snags present at any of the plots. All regression analyses were performed using JMP IN statistical software (SAS Institute Inc) and Si gmaPlot (2002 Systat Software Inc). Significance was accepted at or = 0.05. 18 Results Standing biomass Aboveground biomass of jack pine increased steadily with time since disturbance, beginning near zero at 4 years and reaching its peak of 105Mg/ha by age 69 (Figure 3A). There was no difference between wildfire and harvest chronosequences in the relationship between aboveground biomass and stand age (P = 0.084). Aboveground biomass was highly patchy in the youngest ages (mean CV = 435% for ages 4-7) but patchiness decreased rapidly to 100% by age 17, approaching an asymptote around 60% by age 69 and forming an inverse J -curve (Figure 3B). There was no difference between wildfire and harvest chronosequences in the pattern of within-stand patchiness over time (P = 0.790). Live-Stem density of wildfire-regenerated jack pine generally decreased with stand age, although this pattern was not statistically Si gnificant due to high variability among young fire-origin stands (Figure 3C; Table 1). Density in harvest regime sites decreased linearly from its highest level in the youngest stands; however, it is important to note that the 3 oldest harvest stands were likely planted at lower density than the others, potentially exaggerating density declines with age. Although among-Stand variability of stem density in young fire-ori gin stands was higher than harvest-ori gin stands (Range 390-6000 vs. 400-2600), they centered around similar means resulting in no significant difference between wildfire and harvest chronosequences (P = 0.592). Within-stand patchiness of stem density was high in fire sites from age 5 to at least age 20, decreasing with increasing stand age until about 40 years of age in an inverse j- shaped curve (Figure 3D). The pattern of within-stand patchiness was significantly 19 different for harvest-ori gin stands (P < 0.001), which exhibited relatively uniform stem density with no change across the chronosequence. Patchiness of live stem density became similar between treatments at approximately age 40, from which point forward stands from the two treatments were indistinguishable. Snags Snag density in harvest stands (Figure 2A) increased from 0 snags/ha at age 4 to 250 snags/ha by age 50, at which point the density of snags remained steady. In contrast, there was no significant trend with age for snag density in the fire-ori gin stands, which exhibited both higher density and higher variability among stands. Nevertheless, the lack-Of-fit test still indicated significant differences in patterns of snag density over time between wildfire and harvest chronosequences (P = 0.010). Of particular note was the difference in snag density in the youngest age class, where all 3 of the youngest wildfire stands had much higher numbers of snags than the harvest sites. The three youngest fire sites (ages 5, 6 and 7) had very high numbers of snags (mean of 252 snags/ha) compared to the three youngest harvest sites (ages 4, 6 and 7 with a mean of 3 snags/ha). There is no clear difference between treatments in the stands approximately 15-20yrs of age. Two of the Oldest wildfire sites had the highest snag density; at nearly 700 snags/ha they had 3 times the number of snags that similar aged harvest sites had. Temporal patterns of within-stand patchiness of snag density (Figure 23) were significantly different between wildfire- and harvest-ori gin stands (P<0.001).. Patchiness within harvest regime sites was high at the youngest ages, with a CV of 447% in sites aged 4, 6 and 12 (the remaining youngest site, age 7 was not included in the analysis 20 because it contained no snags in any plots within the site). Patchiness rapidly dropped to a CV of approximately 100% by age 50 to form an inverse J-curve. In contrast, the patchiness of snag density within fire-ori gin stands followed a sigmoidal curve through stand development, with lower initial heterogeneity (CV = 152% at age 4) and then patchiness similar to older harvest-ori gin Stands by ca. age 40. Snag volume increased linearly with time since disturbance with no significant difference between wildfire and harvest chronosequences (P =0.105; Figure 2C). Within- stand variation in snag volume followed patterns nearly identical to variation of snag density with Si gnificant differences between wildfire and harvest chronosequences (P < 0.001; Figure 2D). To understand snag dynamics further, I examined only those snags >5cm dbh in all sites (Figure 4). This size is the predicted minimum size preferred by black-backed woodpeckers due to the high insect prey abundance (Nappi et al 2003). The snag density of these larger snags is significantly higher across all ages of wildfire sites compared to harvest sites (P=0.029), beginning at 100-150 snags/ha in the younger wildfire stands compared to no snags/ha for at least the first 10 years of the harvest stands. Snags of this category rise steadily in both types of stands at the same rate, as evidenced by parallel regression lines. Coarse woody debris Coarse-woody-debris volume followed distinctly different temporal patterns with stand development in wildfire- and harvest-ori gin chronosequences (P < 0.001). Coarse- woody-debris volume (Figure 5A) was highest in jack pine sites of wildfire origin during 21 early stand development (as high as 62 m3/ha at 6 years post fire, with a mean of 49 m3/ha in sites aged 5, 6 and 7 years) and then declined with stand age until about 40 years of age in an inverse J -shaped curve. Harvest regime sites, on the other hand, initially contained a much lower volume of CWD (with a mean of only 12 m3/ha in sites aged 4,6 and 7 years) than the wildfire sites at those ages. Wildfire-ori gin chronosequences display a decline of CWD volume in the first 20 years, while harvest sites remain low but relatively stable; both chronosequences exhibit an increase beginning around 40 years with merging trajectories. Within-stand patchiness of CWD volume in wildfire-ori gin stands (Figure 5B) displayed no significant pattern with stand age. The temporal pattern of CWD patchiness was significantly different in harvest-origin stands (P = 0.001), where it followed a hump-Shaped curve, with CVs around 100% shortly after stand initiation, an increase to a peak CV of 175%, and displaying lowest variation after 60 years (80%). I separated CWD by decay classes across stand age to detect any trends. Both the harvest and wildfire sites had very little CWD of decay class 1 at the younger ages, but their levels had risen in the oldest stands. Young fire stands had large amounts of CWD in decay class 2. No other differences between the chronosequences were apparent. (Figure 6) Total dead wood volume (snag + CWD volume) was calculated to understand dead wood structure in the stands without the confounding effects of the timing of snag transfer to CWD in each individual site (Figure 7). While a large pulse of snags generally falls a few years after a fire, some snags fall earlier or later, or in different stages of decay, or only break away partially, leaving part of the snag behind (Passovoy and Fulé 2006). By examining all dead wood together, we can see a snapshot of each site 22 at one time for a total volume measurement. The difference between harvest- and wildfire origin chronosequences was highly significant for total dead wood (P < 0.001). The pattern was similar to that of CWD volume, but with the addition of snag volume, the difference between young wildfire and young harvest stands was exacerbated. Total dead wood volume became similar in wildfire and harvest chronosequences around age 40 and appears to be on a common trajectory from that point forward. Patchiness of total dead wood volume was not significantly different between harvest and wildfire chronosequences (P = 0.068; Figure 7B). The highest within-stand patchiness occurred early in stand development in both treatments, decreasing linearly over time through the Oldest Stands (P =0.005). Forest floor Changes in forest floor mass over the course of stand development differed significantly between wildfire- and harvest-ori gin chronosequences (P = 0.006; Figure 8C). Forest floor mass was lowest in the fire origin sites early in stand development, then increased asymptotically to a high of 10Mg/ha at stand age 39. In contrast, there was no significant pattern of forest floor mass through stand development in the stands originated by harvesting. There was no difference between harvest- and wildfire-ori gin stands in the temporal pattern of forest floor patchiness (P = 0.658). Patchiness followed a shallow U- Shaped curved, with the highest variation within stands occurring early in stand development in both fire and harvest sites, and the sharpest decline occurring by year 20. 23 Discussion I found that whole-tree harvesting and planting produced jack pine stands that were largely indistinguishable from wildfire-ori gin stands in terms of biomass and live-stem density, but which differed markedly in terms of patchiness and woody debris. Harvest and planting has been demonstrated as an effective method of creating preferred Kirtland’s warbler habitat (Probst and Weinrich 1993), as well as productive forests for harvest, but has caused differences in other ecosystem components at the stand and landscape scale. Stem density has been identified as a factor that is strongly related to stand function (Turner et al 2004). Early in stand development within-stand patchiness is much higher in wildfire sites. Despite the planting pattern of these plantations, which involves regularly spaced openings designed to emulate variable patterns associated with fire, variation was found to be much lower in the plantations early in stand development. At the landscape level, wildfire-regenerated stands are stochastic in regeneration, dependent on many factors such as preceding stand age and cone crop, as well as climatic factors that affect seedling establishment and survival in the first growing season. They are also stochastic at the stand level relative to stands created by harvest and planting. Patchiness in density creates increased structural complexity in a stand. Sparse areas have greater light availability, promoting shade intolerant herbaceous species, while dense areas experience canopy closure sooner. Through stem exclusion, canopy closure adds snags to the system. These snags initially contribute dead foliage and fine litter, and eventually whole tree boles to the ground, affecting soil development through the 24 addition of nutrients and also organic matter, which helps retain water. All of these factors are critical components of ground cover dynamics, which in turn affect many other organisms, including invertebrates and mammals. Snags are crucial in forested ecosystems for many organisms, especially cavity nesting birds. Early in stand development in wildfire sites, large numbers of fire-killed trees provide habitat for birds and small mammals, as well as for wood-eating organisms. As predicted, snag density in young harvest regime stands is very low (or zero, especially for larger snags) due to recent removal for harvest and low snag density in the preceding stands (Figure 2A.) Greater numbers of snags in these wildfire sites at early ages supports previous findings (Clark et a1 1998). There are also greater numbers of snags in wildfire stands at the older ages, although this result is confounded by the lower density planting of the harvest sites at those ages. The current generation of plantations is likely to produce higher numbers of snags as they age, relative to the oldest stands in this chronosequence due to their higher planting density. Many studies measuring the effects of reduced snag densities in conifer stands due to postfire salvage logging have shown strong effects on woodpecker populations; these studies reinforce the importance of postfire snags in fire systems. Nappi et al (2003) showed direct correlations between these larger, less deteriorated snags found immediately after a fire, and wood-boring beetle larvae holes. The holes strongly predicted foraging activity by black-backed woodpeckers. Hutto (2006) stresses that the current management guidelines for 6- 10 snags/ha in most conifer harvest systems are inadequate, due to much higher snag requirements of fire specialist bird species. Not only are the recommended densities initially too low for these birds, but the lower 25 densities leave the existing snags more vulnerable to wind throw, so the lifespan of the snags is shorter (Mast and Chambers 2006). Examination of CWD dynamics between wildfire and harvest regime chronosequences further enhances the importance of the stands less than 40 years old in assessing differences between the two treatments (Figures 4A and 4C.) Very high volumes of CWD in the younger wildfire stands gradually drop to their lowest levels around 40-60 years, with an indication of rising levels as the stands become very old. This pattern is additionally strengthened when total dead wood (snags + CWD) is examined (Figure 7A); the u-shaped pattern of the wildfire sites merging with the gradual increase Of harvest regime sites from their initially lowest levels is very clear. Once again, spatial composition of these stands is very different in early development, becoming more similar over time. The relationship of CWD and snag recovery to stand origin becomes clearer when total dead wood volume (CWD volume plus snag volume, Figure 7A) is examined. The result is a much tighter fit to a recovery scheme where stands of different dead wood amounts merge to a similar trajectory around 40 years. Within the stands at different age groups, most of the stands with lower snag volume had higher CWD volume, and vice versa, the result being a much better understanding of snag recovery dynamics. There are many factors that affect when snags fall and become CWD, such as localized stem density, size of snags, and presence of surrounding live trees and stochastic wind factor, and although each stand varies considerably in snag volume, these volumes contribute to a clear recovery pattern of overall dead wood. This further supports the conclusion that 26 this harvest for fire emulation strategy does not produce a similar stand until around 40 years after disturbance. These data match well with the theoretical model of CWD dynamics after stand- destroying disturbance (Sturtevant et al 1997 .) The theoretical model is composed of two stages: 1) An initial stage in which residual CWD from the preceding stand decays and declines, followed by 2) An accumulation stage in which new CWD is added to the forest floor from the regenerated stand and levels increase. In this case, wildfire stands initially have much higher CWD levels due to immediate fire effects and newly downed trees in the first few years after disturbance. Harvest regime stands initially have much lower CWD levels. Both treatments experience an initial decline in CWD, followed by a nearly identical pattern of CWD accumulation after age 40 (Rothstein et al 2004). These differences in CWD in stands less than 40 years Old hold important implications for sustainable management of these stands. Levels of CWD in these two treatments are not similar until approximately 40 years, which is very close to the harvest cycle of 50-60 years currently employed. Relative to CWD, the method of management designed to emulate natural disturbance actually creates stands that are different for most of their development. Repeatedly clearcut stands will never reach the CWD levels present in a naturally regenerated stand in the first 30 years following a wildfire. The only significant pattern in patchiness of CWD except for volume of CWD in harvest regime sites. Patchiness was highest in intermediate sites and lowest early and late in stand development. Overall, it appears that patchiness may be lower in wildfire sites early in stand development, but this may also be driven by higher patchiness in 27 harvest regime sites caused by fewer actual pieces of CWD per plot, as many plots contained no pieces at all. Forest floor serves as habitat for invertebrates and also as an important source of mineralizable nutrients in this ecosystem that is already low in organic matter (Yermakov and Rothstein 2006). Mass of the forest floor in the wildfire chronosequence is lowest after a fire has burned away much of the forest floor, increasing to around double the initial levels around 40 years, then gradually leveling off or declining. This makes sense, as the CWD in the system contributes to the forest floor in the early post-disturbance years; when CWD reaches its lowest level (near 40 years) and has little to contribute, the levels of forest floor plateau and begin to decline slightly, receiving further contributions from standing biomass instead of CWD. The harvest regime stands do not show a significant overall pattern, but the youngest stands have a greater amount (two to three times as much) of forest floor as the burned stands. There is no difference between the chronosequences by approximately year 15. Slash from logging leaves behind large amounts of fine litter during the harvest, but while that litter decays there is no major source to contribute to further development of the forest floor, unlike in the fire sites so levels drop rapidly. The early pulse of litter due to harvesting is quickly lost to decomposition while at the same time in the fire sites the early loss of forest floor is quickly regained by additions of fine litter and CWD from the fire-killed trees in the site. The structure of the jack pine forests in this study is similar as the stands age, but come from very different conditions at the beginning of stand development. Organisms adapted to the conditions commonly found in young fire-regenerated stands will encounter different habitat scenarios in young harvest sites, if they depend primarily on 28 large amounts of dead wood (down or standing) or specific forest floor conditions. Many invertebrates depend on dead wood for food, shelter, and locations to lay eggs. Many species in turn depend on these species, such as birds and rodents. Woodpeckers are a good example of a Species that requires insect-harboring wood for survival, of which snags are a large component. Although the emulation of natural disturbance as a management strategy is a reasonable objective, there are considerations to keep in mind. In this system, the management strategy has been successful at replicating some components of a fire- re generated stand. For other components, however, the differences are great, especially in the younger stands. It is important to regard these issues at a landscape scale. Consider Jves Bergeron et al's model (1999) comparing forest age-class distributions of two scenarios: 1) a forest system regenerating naturally by fire and 2) a system managed to replicate the same process. Even though the harvest rotation is the same as the fire return interval, the stand-age distribution between the systems is vastly different. The fire system has many more young stands than other ages, but also has some very old stands, much older than the average time since disturbance. At the landscape level, there exists a broad spread of stands of all ages. In contrast, the managed system maintains an even distribution of all cohorts up to the harvest interval, yet excludes the older stand ages completely, skewing the age distribution to the younger stands, relative to the wildfire stands. Over long periods of intensive harvesting, old stands may virtually disappear, along with the increased vertical structure and age diversity typically associated with older forests. With this type of management strategy, at the landscape level, species dependent on the very old stands may lose habitat due to harvesting, while 29 concurrently, the species that rely on several key structural features of the youngest stands will effectively experience habitat loss due to the differences between plantations and fire-regenerated stands at these young ages. So while these emulation management strategies can be very successful at maintaining habitat and diversity for many species, it is important to continue to monitor all components and effects of these strategies to ensure the best possible conditions for all species involved. Management Implications Replacing wildfire with harvest and planting has strong repercussions on the ecosystem compared to the natural wildfire-regenerated system. Most notably, the legacy wood present after a fire is largely gone. CWD is greatly diminished, especially in the early years after disturbance. The snag biomass and density after a harvest, even one that follows the common snag retention guidelines, is greatly reduced compared to post-fire conditions. Not only are there fewer snags onsite, the lower densities also mean that those snags will fall sooner due to higher susceptibility to windthrow (Mast and Chambers 2006). We have also shown that snag density differences between wildfire and harvest persist over time (Figure 4). The largest differences in legacy wood occur in the early years following disturbance. Hutto (2006) has demonstrated that for fire following species, the most stringent habitat requirements are in these early years, the years during which harvest and wildfire stands are most different. Differences also occur at the landscape scale as stand age becomes homogenized and very old stands (older than the harvest rotation) are slowly phased out. Recommendations for management that successfully emulates natural wildfire disturbance needs to address these issues. 30 Snag retention in harvest systems is a contentious issue because post-fire snag dynamics are so different from post-harvest snag dynamics; in order to truly recreate post-fire snag dynamics in a harvest system, you would have to harvest without actually removing any of the trees. One option that is more viable is to leave some patches of snags in a harvested system by girdling groups of trees. These patches would mimic post-fire dynamics by replicating snag density, and would also prolong the life of the snags by minimizing risk of windthrow. These snags should be desirable "high quality" snags as defined by Nappi et al (2003), of larger size and low deterioration. At the very least, stands in fire systems that have burned due to wildfire should not be salvage cut (Hutto 2006, Nappi et a1 2003). An effort should be made to maintain CWD levels in the harvest system that are closer to early post-fire levels. Leaving more snags in the system will help to increase the CWD levels as the snags fall. If possible, some intact cut trees should be left on the forest floor, not just the smaller slash that is commonly left. As with snags, larger diameter trees are also valuable as sources of CWD (Lindenmayer and McCarthy 2002) and this should be considered when selecting trees for dead wood retention. Some consideration should also be given to the dynamics of living trees in the harvest system. Leaving some large trees intact in the stand will simulate the trees that Often survive a fire, usually the larger diameter trees (Lindenmayer and N035 2006, Lindenmayer and McCarthy 2002). Managers should also consider leaving some "no-fire refugia" sections completely unharvested if they are in areas that typically may not have burned in the natural system, for example a wetland area (Angelstam 1998). Even in 31 systems that naturally regenerate with fire, there are almost always some survivors or sections of survivors that persist into the development of the succeeding stand. lastly, management should address these issues at the landscape scale. Large- scale harvest management skews stand age toward younger stands (Lindenmayer and McCarthy 2002, Hansen 1991); over time there are fewer really old stands and more evenly-distributed numbers of cohorts in all other ages. Management that allows a more natural distribution of stand ages by leaving some old stands would more closely reproduce post-fire landscape conditions. Also, since harvest and fire stands are so different early in stand development and more Similar as they age, extending harvest cycles for some stands would allow them to be more similar to the natural systems they are made to emulate for a longer amount of time. Managers of fire systems need to consider the unique nature of burned forests and make some adjustments to protect the sensitive early successional ecosystem and mediate the long-term effects of management. 32 05:. o mLowEQEHHIuHI VN 9. NF 0 m o 00 o 3(2m00 ZOZEoomoa 0 n8— .mme . 88—. l 4 fl 0 F O Q 820mm. 8me m8? 0 08m 0 o mmm— owe . 0 m8. m3. omon§ Ea A3< .n 953% 35 3 om< vcmflm cc end. 285 cm 00 CV ON 0 ow ow ow ON 0 _ - b 0 fl 0 O m w 00—. D. S m. rCON a. o w I. .8 s W .0 u. 8.. w e . u. W... e ) 0 18m m/m. < 000 .N 83wE 8m Doom—80c 8 08 288%. .535 8m 28 8032 Son mo acumen—082850 wee? Amy in EomA mwmcm new EEO—o wwam Ho 80:38am 28 3c not EomA mwmam mo baa—DO .v PEER 36 3 09a. 058m 3 09‘ “Exam 8 8 8 o 8 8 8 8 o I . . om . . 0 Mo . 0 row a .8 W 9 :8 m 83.0.. 68% U. 824.. 0 .8Pu wow—.h/w < o .8 8m 8 .N 0.5mE ooh “00:58.0 8 08 238% .598 05 .08 0:88 808: 58 mo 8803008880 mac—m Amv 0829 $520 8003 0800 no 8832, A 2.50.0 8003 0800 .m 9:83 37 Figure 6. Volume of coarse woody debris along chronosequences of both harvest regime and fire origin for decay classes 1 (A and B), 2 (C and D), 3 (E and F) and 4 (G and H.) Symbols are as described for Figure 2. 60 60 A B 504 50‘ ,2 404 g 40‘ 5 35 70 0.30 0-30 m 0 5 2 20 20‘ 10+ 0 1° . W ...... 9’13 o : fi , , w o-———a>v "‘5" v o 20 4o 60 80 o 20 40 60 Stand Age (y) Stand A98 (Y) 60 C 601 V D 50. 50. 0 f 404 g 404 8.3(H 2'301 M 2 0') o 20‘ 2 . 20+ 0 O 10 10‘ ° .. O 0 O o -é , f w O G . or ' o 20 40 60 80 o 20 4° 6° Stand Age (y) SW16 A96 (Y) 60 E 60 501 50+ F 0 f 401 2 40‘ 8.30. 2304 0 "’2 "’2 zol 20* ‘9 o o 0 o 10* . . . . 10* o O o o t '. . 0 ° 0 o o 40 60 80 o 50 4b 50 Stand Age (3’) Stand Age (y) so G 60 H 50‘ 50 €40 1 2 401 o in? ”Q30 ‘ Q30 2 F) 2 20 ‘ 20 o 10 4 o 10* 0° 0 O o M. o o o 1 1 , o «P—Q v V O Y 0 20 40 60 80 O 20 4O 60 Stand Age (y) Stand A90 ()0 38 E 09. 08.0 S 09. 08.0 8 8 ow 0w. o o 8 8 9. cm 0 .. O ...............U o . ,8. m 5...: m. ...... w. aaaaaaaa 0 O a ...... . w 0 com w d o m: . .08 M U. 9 u" .8. m m o 00 W § .N 0SwE .00 00000000 00 0.0 2008.3 .Ew00 05 0:0 0E_m0.. 80.5... .000 .0 00000300000020 wcofi Am: 0520.» 0003 0000 :30. ..0 .5000? 0.... A5 0520., 0003 0000 30H .h 0.53% 39 S 09.. 08.0 ON 888° 88 P‘— é (%) uoneue/uo tumomeoo 3 809.220 8 0.. on o . l _ o o .m A. W 0.x. 6 O// \\ . o. .w 0 Illll\\ . Id 0 O u- o B . .0. .N 0.:wE 00.: 00000000 00 000 2008.3 .5300 05 0:0 0800.0: 8030: 500 00 000:0..00000030 w:0_0 Amy 80:. 0000 8000.. 00 00000? 0:0 9.0 80.: 0000 80am .w 009w?“ 4O voooo 3.er wobmm + 0&0: - N~.o u .A 00.00. 3.6 080.0.» w00m 35o momwo m. ...m + KN. ..w. - N000 .0 000000> .ooo.ov owvoo 0.08m - £0.80 + N0 08.088 0008 0. 80w0000 :0. 80.8008 00.8083. .— 0.00... 41 mmvoo 90.3 m .o0.~ + 0.003."... 80:00.0 obwoo 0 0n. Q30 .0 08:.0> 38o mvbco ...00..o-v<0m..m0 + 0.069. 08.0....» mmooo NUA. 9.50 .0 088.0.» Noooo 000W o 0806 - xiooo .I. .. 00.000 890 .0Q Q30 .0 088.0> Naooo movo 00% + xhwwvd - N mmooo 0.036 .088 + 803.. + N<80m~o.o- H .. 08.2.3 0oo.o .808 000.. 80:0.. 088.0> 0003 mmooo mmamo madam + .3800- u .. 00.000 wood 0000 .08. .0 :0..0...0> obooo wwowo bow. .. + xmowmo - N20 02 0.0 0m :25 mum: m 2025 9. E :2 :z mam: 00 5280 .0, S oz :2 mum: 8 32020 :8 co DZ DZ mum: on 50:00.0 :0: mm :2 :2 02: 0. S2020 2.0 E :z :2 mum: 2 52020 3 am OZ DZ MZQ N . 30:00.0 30. :z :z 02 mum: 0 3280 .38 :z 02 :z mum: 0 52020 50 :2 oz :2 020 .2 52020 88 0:.0 0.00% 00 A80. 30:0 .0 .30 :02: 0:... .0 0000 00.000550 0m0 Ewto no 0008 060: .0000 .0000 0:03 000000000 .0000 0:.n. 0.00_. o. $02.00 00.: 0.00.. 00 .38 :. 0:03, 0:0 ...0 0m0 05 00 E000: 0w< 0030.000 05% N 030,—. References Abrams, M.D., Dickman, DJ. 1982. Early revegetation of clear-cut and burned jack pine sites in northern lower Michigan. Canadian Journal of Forest Research. 60 (6): 946-954 Angelstam, P.K.. 1998. Maintaining and restoring biodiversity in European boreal forests by developing natural disturbance regimes. Journal of Vegetation Science. 9(4):593-602. Attiwill, RM. 1994. The disturbance of forest ecosystems: the ecological basis for conservative management. Forest Ecology and Management. 63(2-3):247-300. Bergeron, Y., B. Harvey, A. Leduc, and S. Gauthier. 1999. Forest management guidelines based on natural disturbance dynamics: Stand- and forest-level considerations. Forestry Chronicles. 75(1): 49-54. Binkley, D. 1984. Does Forest Removal Increase Rates of Decomposition and Nitrogen Release. Forest Ecology and Management. 8(3—4):229-233. Brais, S., Sadi, F., Bergeron, Y., Grenier, Y. 2005. Coarse woody debris dynamics in a post-fire jack pine chronosequence and its relation with site productivity. Forest Ecology and Management. 220(1-3):216-226. Brassard, B.W., Chen, H.Y.H. 2006. Stand structural dynamics of North American boreal forests. Critical Reviews in Plant Sciences. 25(2): 1 15- 137. Carey, A.B., Harrington CA. 2001. Small mammals in young forests: implications for management for sustainability. Forest Ecology and Management. 154(1-2):289-309 Chapin, F.S., Trainor, S.F., Huntington, 0., Lovecraft, A.L., Zavaleta, E., Natcher, DC, A. McGuire, D., Nelson, J.L., Ray, L., Calef, M., Fresco, N., Huntington, H., Rupp, T.S., DeWilde, L., Naylor, R.L. 2008. Increasing wildfire in Alaska's boreal forest: Pathways to potential solutions of a wicked problem. Bioscience. 58(6):531-540. Charron, I., Greene, DP. 2002. Post-wildfire seedbeds and tree establishment in the southern mixedwood boreal forest. Canadian Journal of Forest Research. 32 (9): 1607— 1615. Clark, D.F., Kneeshaw, D.D., Burton, PJ. 1998. Coarse woody debris in sub-boreal spruce forests of west-central British Columbia. Canadian Journal of Forest Research. 28(2):284-290. Converse, S. J., White, G. C., Farris, K. L., Zack, S. 2006. Small mammals and forest fuel reduction: National-scale responses to fire and fire surrogates. Ecological Applications. 16(5): 1717-1729. 45 Degrandpre, L., Gagnon, D., Bergeron, Y. 1993. Changes in the Understory of Canadian Southern Boreal Forest After Fire. Journal of Vegetation Science. 4(6):803-810. De Groot, WJ., Bothwell, P.M., Taylor, S.W., Wotton, B.M., Stocks, BJ., Alexander, ME. 2004. Jack pine regeneration and crown fires. Canadian Journal of Forest Research. 34(8): 1634-1641. Dordel, J ., Feller, M.C., Simard, S.W.. 2008. Effects of mountain pine beetle (Dendroctonus ponderosae Hopkins) infestations on forest stand structure in the southern Canadian Rocky Mountains. Forest Ecology and Management. 255(10):3563-3570. Franklin, J .F., Spies T.A., Van Pelt, R., Carey, A.B., Thomburgh, DA., Berg, D.R., Lindenmayer, D., B., Harmon, M.E., Keeton, W.S., Shaw, D.C., Bible, K., Chen, J. 2002. Disturbances and structural development of natural forest ecosystems with Silvicultural implications, using Douglas-fir forests as an example. Forest Ecology and Management. 155(1—3): 399-423. Franklin,J. 1993. Preserving Biodiversity: Species,Ecosystems,or Landscapes? Ecological Applications. 3(2):202-205. Gore, A.P., Johnson, EA., Lo, HP. 1985. Estimating the time a dead tree has been on the ground. Ecology. 66(6): 1981-1983 Hannam, K.D., Quideau, S.A., Kishchuk, BE. 2006. Forest floor microbial communities in relation to stand composition and timber harvesting in northern Alberta. Soil Biology & Biochemistry. 38(9): 2565-2575 Hansen, AJ., Spies, T.A., Swanson, FJ., Ohmann, J .L. 1991. Conserving Biodiversity in Managed Forests. Bioscience. 41(6):382-392. Harper, KA., Bergeron,Y., Drapeau, P., Gauthier, S., De Grandpre, L. 2006. Changes in spatial pattern of trees and snags during structural development in Picea mariana boreal forests. Journal of Vegetation Science. 17(5):625-636. Hart, 8A., Chen, H.Y.H. 2006. Understory Vegetation Dynamics of North American Boreal Forests. Critical Reviews in Plant Science. 25(4):381-397. Hazlett, P.W., Gordon, A.M., Voroney, RP. 2007 . Impact of harvesting and logging slash on nitrogen and carbon dynamics in soils from upland spruce forests in northeastern Ontario. Soil Biology & Biochemistry. 39(1):43-57 Heliola, J ., Koivula, M., Niemela, J. 2001. Distribution of carabid beetles (Coleoptera, Carabidae) across a boreal forest-clearcut ecotone. Conservation Biology. 15(2):370- 377. 46 Houseman, G.R., Anderson, RC. 2002. Effects of jack pine plantation management on barrens flora and potential Kirtland's warbler nest habitat. Restoration Ecology. 10(1): 27-36. Huber, P. 08 September 2008. E-mail Interview. Wildlife Biologist, USDA Forest Service, Mio Ranger Station, Mio, MI. Hunter, M.L. Jr. 1993. Natural fire regimes as spatial models for managing boreal forests for biological diversity. Biological Conservation. 65(2): 1 15- 120. Hutto, R.L. 2006. Toward Meaningful Snag-Management Guidelines for Postfire Salvage Logging in North American Conifer Forests. Conservation Biology. 20(4):984- 993. Jenkins, MJ., Hebertson, E., Page, W., Jorgensen, CA., 2008. Bark beetles, fuels, fires and implications for forest management in the Intermountain West. Forest Ecology and Management. 254(1): 16-34. Kashian, D.M., Tinker, D.B., Turner, M.G., Scarpace, FL. 2004. Spatial heterogeneity of lodgepole pine sapling densities following the 1988 fires in Yellowstone National Park, Wyoming, USA. Canadian Journal of Forest Research. 34:2263-2276. Kashian, D.M., Barnes, B.V., Walker, W.S. 2003. Landscape ecosystems of northern Lower Michigan and the occurrence and management of the Kirtland's warbler. Forest Science 49(1): 140-159. Kashian, D.M., Barnes, B.V. 2000. Landscape influence on the spatial and temporal distribution of the Kirtland's warbler at the Bald Hill burn, northern Lower Michigan, US .A. Canadian Journal of Forest Research. 30(12): 1895— 1904. Kemball, KJ., Wang, G.G., Westwood, AR. 2006. Are mineral soils exposed by severe wildfire better seedbeds for conifer regeneration? Canadian Journal of Forest Research. 36(8): 1943- 1950. Laiho, R., Prescott, CE. 2004. Decay and nutrient dynamics of coarse woody debris in northern coniferous forests: a synthesis. Canadian Journal of Forest Research. 34(4):763-777. Laiho, R., Prescott, CE. 1999. The contribution of coarse woody debris to carbon, nitrogen, and phosphorus cycles in three Rocky Mountain coniferous forests. Canadian Journal of Forest Research. 29(10): 1592- 1603. Lindenmayer, D.B., Noss, RF. 2006. Salvage Logging, Ecosystem Processes, and Biodiversity Conservation. Conservation Biology. 20(4):949-958. 47 Lindenmayer, D.B., Franklin, J .F ., Fischer, J. 2006. General management principles and a checklist of strategies to guide forest biodiversity conservation. Biological Conservation. l3 l(3):433-445. Lindenmayer, D, McCarthy, M.A.. 2002. Congruence between natural and human forest disturbance: a case study from Australian montane ash forests. Forest Ecology and Management. 155(1-3):319-335. Lindenmayer, D.B., Recher, HF. 1998. Aspects of ecologically sustainable forestry in temperate eucalypt forests - beyond an expanded reserve system. Pacific Conservation Biology. 4:4-10. Mast,J.N., Chambers,C.L. 2006. Integrated approaches, multiple scales: Snag dynamics in burned versus unburned landscapes. The Professional Geographer. 58(4):397—405. Matsushima, M., Chang, S.X. 2006. Vector analysis of understory competition, N fertilization, and litter layer removal effects on white spruce growth and nutrition in a 13- year-old plantation. Forest Ecology and Management. 236(2-3):332-341 Nappi, A. Drapeau, P., Giroux, J ., Savard, J .L. 2003. Snag Use by Foraging Black- Backed Woodpeckers (Picoides arcticus) in a Recently Burned Eastern Boreal Forest. The Auk. 120(2):505—51 1. ' Neter, J ., Wasserman, W., Kutner, M.H. 1990. W. Homewood (IL): Richard D. Irwin, Inc. 1181p. Passovoy, D.M., Fule, P2. 2006. Snag and woody debris dynamics following severe wildfires in northern Arizona ponderosa pine forests. Forest Ecology and Management. 223(1-3):237-246. Perala, DA., Alban, DH, 1994. Allometric biomass estimators for aspen-dominated ecosystems in the upper Great Lakes. Research Paper NC-314. St. Paul, MN: North Central Forest Experiment Station, Forest Service, US. Department of Agriculture. Probst, J .R., Donner, D.M., Bocetti, C.I., et al. 2003. Population increase in Kirtland's warbler and summer range expansion to Wisconsin and Michigan's Upper Peninsula, USA. Oryx. 37(3):365-373. Probst, J .R., Weinrich, J. 1993. Relating Kirtland's warbler population to changing landscape composition and structure. Landscape Ecology. 82257-271. Rhoades,C.C.,Meier,AJ.,Rebertus,A.J.. 2004. Soil properties in fire-consumed log burnout openings in a Missouri oak savanna. Forest Ecology and Management. 192(2- 3):277-284. 48 Roberston, PA., Bowser, Y.H. 1999. Coarse Woody Debris in Mature Pinus ponderosa Stands in Colorado. Journal of the Torrey Botanical Society, 126(3): 255-267. Rothstein DE, Yermakov ZY , Buell AL 2004. Loss and recovery of ecosystem carbon pools following stand-replacing wildfire in Michigan jack pine forests. Canadian Journal of Forest Research. 34(9): 1908-1918 Schulze, D., Wirth, C., Mollicone, D., Ziegler, W. 2005. Succession after stand replacing disturbances by fire, wind throw, and insects in the dark Taiga of Central Siberia. 0ecologia. 146:77-78. Simard, D.G., Fyles, J .W., Pare, D. 2001. Impacts of clearcut harvesting and wildfire on soil nutrient status in the Quebec boreal forest. Canadian Journal of Soil Science. 81(2): 229-237 Simard, AJ., Blank, R.W. 1982. Fire history of a Michigan jack pine forest. Michigan Academician. 15:59-71. Simberloff, D. 1998. Flagships, umbrellas, and keystones: Is single-species management passe in the landscape era? Biological Conservation. 83(3):247-257. Solomon, BD. 1998. Impending recovery of Kirtland's warbler: Case study in the effectiveness of the Endangered Species Act. Environmental Management. 22(1):9-17. Stocks, BJ. 1989. Fire Behavior in Mature Jack Pine. Canadian Journal of Forest Research. 19(6):783-790. Sturtevant, B.R., Bissonette, J .A., Long, J .N ., Roberts, D.W. 1997. Coarse woody debris as a function of age, stand structure, and disturbance in boreal Newfoundland. Ecological Applications. 7(2):702-712. Tanskanen, H., Granstrom, A., Venalainen, A., and Puttonen, P. 2006. Moisture dynamics of moss-dominated surface fuel in relation to the structure of Picea abies and Pinus sylvestris stands. Forest Ecology and Management. 226: 189-198. Tinker, D.B., Knight, DH. 2001. Temporal and spatial dynamics of coarse woody debris in harvested and unharvested lodgepole pine forests. Ecological Modeling. 1412125-149. Tinker, D.B., Knight, DH. 2000. Coarse woody debris following fire and logging in Wyoming lodgepole pine forests. Ecosystems. 3(5):472-483. Turner, M.G., Tinker, D.B., Romme, WH. Kashian, D.M., Creighton, ML. 2004. Landscape patterns of sapling density, leaf area, and aboveground net primary production in postfire lodgepole pine forests, Yellowstone National Park (USA). Ecosystems. 7(7):751-775. 49 Vanbergen, AJ., Watt, A.D., Mitchell, R., Truscott, A., Palmer, S.C.F., Ivits, E., Eggleton, P.T., Jones, H., and Sousa, J .P. 2007. Scale-specific correlations between habitat heterogeneity and soil fauna diversity along a landscape structure gradient. 0ecologia. 153(3):713-725. Westbrook, C J ., Devito, Kl, Allan, CJ. 2006. Soil N cycling in harvested and pristine boreal forests and peatlands. Forest Ecology and Management. 234(1-3):227-237. Whitney, G.G. 1986. Relation of Michigan presettlement pine forests to substrate and disturbance history. Ecology. 67: 1548—1559. Worrall, JJ. 1994. Population structure of Armillaria species in several forest types Mycologia. 86(3):401-407. Wright, P., Harmon, M., Swanson, F. 2002. Assessing the effect of fire regime on coarse woody debris. Gen. Tech. Rep. PSW—GTR-181. Forest Service, US Department of Agriculture. Yermakov, Z., Rothstein, DE. 2006. Changes in soil carbon and nitrogen cycling along a 72-year wildfire chronosequence in Michigan jack pine forests. 0ecologia. 149(4):690-700. 50 IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII lllllllllllzlllllMlllllllllll 062 52