AN ABSTRACT OF THE THESIS OF
Tiffany A. Brett for the degree of Master of Science in Wildlife Science presented on July 21, 1997. Title: Habitat Associations of Woodpeckers in Managed Forests of the Southern Oregon Cascades.
Abstract approved: John P. HaY(s
Abstract approved: WilliamC. McComb
I studied nest-site characteristics and habitat relationships for three species of primary cavity-nesting birds--hairy woodpecker (Picoides villosus), northern flicker (Colaptes auratus), and red-breasted sapsucker (Sphyrapicus ruber)-- over spatially heterogeneous landscapes in managed forests of the Southern Oregon Cascades during 1995 and 1996. The study was conducted on the Diamond Lake Ranger District of the Umpqua National Forest. I found 163 nests--68 nests of red-breasted sapsuckers, 63 of northern flickers, and 32 of hairy woodpeckers. I evaluated characteristics of nest trees, habitat within a 0.04-ha circle surrounding nest trees, and the surrounding landscape and compared these to characteristics of randomly selected plots. Analysis revealed statistically significant differences between nest trees and random trees among species. Variables significantly influencing the probability of use of a tree were state of decay, diameter at breast height, and tree top condition (broken or intact). Red-breasted sapsucker nests were associated with large diameter (mean80.5 cm) trees with broken tops. Hairy woodpeckers used large diameter (mean80.4 cm) trees in decay class 3 for nesting. Northern flickers were associated with large diameter trees (mean =77.7cm) with broken tops. I investigated habitat associations at the landscape level using GIS coverages and FRAGSTATS spatial analysis to generate indices quantifying landscape structure for each study site. Higher numbers of woodpecker nests were found in landscapes with lower proportions of mature closed-canopy forest (trees> 53 cm, canopy closure >40%),and possessing greater overall habitat complexity. I also detected a significant association between all three species of woodpecker and proximity to edge. Red-breasted sapsuckers exhibited the most dramatic affinity for edge habitat, choosing to use nest trees located on a habitat edge more often than would be expected if selection was due to chance alone (P <5 x1(f9). Nests of northern flickers and hairy woodpeckers were also significantly associated with habitat edges (P <0.004 and P <0.04 respectively). When managing for woodpeckers within managed forest landscapes, large snags (>80 cm) should be retained. Woodpecker needs for nesting habitat may not be met if retained snags do not represent a diversity of decay classes. Snags representing a variety of tree species and decay classes should be left following management activity. Additionally, management plans should incorporate the use of a variety of methods for stand regeneration. Using prescriptions such as shelterwoods, commercial thinning, and partial cuts, overall landscape diversity will increase, enhancing habitat conditions for nesting woodpeckers. Habitat Associations of Woodpeckers at Multiple Scales in Managed Forestsof the Southern Oregon Cascades
by
Tiffany A. Brett
A THESIS
submitted to
Oregon State University
in partial fulfillment of the requirements for the degree of
Master of Science
Presented July 21, 1997 Commencement June, 1998 Master of Science thesis of Tiffany A. Brett presented on July 21, 1997
APPROVED:
Co-Major representing Wildlife Science
I understand that my thesis will become part of the permanent collection of Oregon State University libraries. My signature below authorizes release of my thesis to any reader upon
Tiffany A. Brett, Author -Acknowledgements
My thesis work was funded by DEMO (Demonstration of Ecosystem Management Options), a cooperative research project initiated in Oregon and Washington in 1994. I am grateful to the organizers of the project, and most especially the technical support from the Diamond Lake Ranger District. This thesis was inspired and encouraged by many, I could not possibly mention them all here. To my major advisors Bill McComb and John Hayes, I offer my sincere thanks in allowing me to 'explore' while keeping me on track with their timely encouragement and advice. I am especially grateful to John for his patient guidance throughout the analysis and writing of the thesis, his thoughtful advice and support were invaluable.I would like to thank the other members of my committee as well; Dan Edge for his tireless cheer, and Susan Stafford, not only for her statistical advice, but also for her encouraging words. I thank Sandy Lewis for her tireless attention to bureaucratic details, and her constant smile. I am grateful to Logan Norris, department head and supporter, for his unending optimism with regards to the DEMO project. I would be remiss without thanking Carol Chambers, who had enough faith in me to choose me as a research assistant on the DEMO project in 1994. Carol was always there to offer professional and personal advice whenever it was needed. Gody Spycher, database whiz and fellow lover of Rilke, was always willing to share bits of wisdom on database management, and would help out with SAS problems in a pinch. His cheery disposition got me through some of the rough spots in the analysis. I would have been lost had it not been for the GIS wizardry of George Leinkaemper. He patiently guided me through the painful introductory stages of Arc Info, and in the process turned me on to the wonders of spatial analysis with his always-present enthusiasm. Thanks also to Barbara Marks, who aided in analysis with FRAGSTATS. My work would not have been possible without the assistance of these two people. I am grateful for the support of my friends and colleagues, I would not have made it through this difficult time without their empathy and understanding. Special thanks to friends and fellow classmates Jennifer Weikel, Nobi Suzuki, and Jonathon Brooks, who gave invaluable advice along the way. Thanks to Matthew Hunter notonly for his endless enthusiasm for all creatures winged and otherwise, but also for his thoughtful comments which inspired ideas during the later writing stages. To fellow Wisconsinite Joan Hagar I am grateful, as someone with whom I could share my love for birds, and who was always willing to 'lend an ear' when I needed one. I am indebted to Louise Yount and Shelley Church, who were always there for me when the going got tough, and who have become very dear friends along the way. I would like to acknowledge and thank fellow Demo researchers Rebecca Thompson, and Jeff Feen, with whom I shared the joys and frustrations of field work. Jeff, despite being a 'mammal guy' has become a dear friend during the last two years (even though he remains a staunch advocate of Fords and flying squirrels). I would like to thank those who have worked hard to help me collect the data during the past two years; including Wayne Harris, Sue Danver, and the Swift sisters Roberta and Stephanie. Thanks to Lyndia Hammer for her endless enthusiasm and to Anthony Floyd for his loyalty and hard work. I owe a great deal to Anthony, who not only shared my love of feathered critters, but also stuck out two field seasons, putting up with the hard work and delightful spring weather of the Oregon Cascades-cheerfully for the most part. I owe gratitude to many forest service employees at the Diamond Lake Ranger District who generously provided technical support throughout the duration of my study. To Rick Abbott, die-hard supporter of DEMO, I thank for accommodating all DEMO researchers, and dealing with problems we encountered along the way. I am grateful to Kim Vieira, GIS guru on the district, for her patient direction in my explorations of GIS and Arc-Info. I am grateful to friends and a fellow bird-enthusiasts Keith and Mary Graves. Keith, as district recreation specialist/biologist, gave freely of his advice and support throughout the study. I am most grateful to my parents Frank and Chris Church, who encouraged me in all of my explorations. Thanks to my father, for cultivating in me a love and respect for nature as a young girl, and for sharing his awe and wonder in things natural, wild, and free. He is responsible for 'planting the seed' that blossomed into my passion (some might call it obsession) for birds.I will always be grateful to my mother for helping me to see beauty in all things, and for her tireless support and undying faithin me. Finally, I would like to thank my husband Joe, who besides being wonderfully supportive throughout, also gave freely of his time to instruct me in the use of the GPS satellite equipment, and helped me to create the map coverages I needed for analysis. I especially thank Joe for his constant love and patience, and for sharing his love and knowledge of the Diamond Lake area with me. TABLE OF CONTENTS
Page CHAPTER 1- HABITAT ASSOCIATIONS OF WOODPECKERS AT MULTIPLE SCALES IN MANAGED FORESTS OF TIlE SOUTHERN OREGON CASCADES
CONCEPTUAL BACKGROUND 1
RESEARCH NEED AND JUSTIFICATION 5
SnagManagement 7 Woodpeckers of the Southern Oregon Cascades 7
OBJECTiVES 8 LITERATURE CITED 9
CHAPTER 2- NEST-SITE CHARACTERISTICS OF WOODPECKERS IN MANAGED FORESTS OF THE SOUTHERN OREGON CASCADES 13
INTRODUCTION 13 STUDY AREA 15 METHODS 19 Study-Site Selection 19 Bird Sampling and Nest Searches 19 Habitat Sampling 20 Data Analysis 22
RESULTS 23 Characteristics of Nest Trees 24 Multivariate analysis 26
DISCUSSION 32 Diameter 32 Tree Condition and Decay 34 Tree Species 34
CONCLUSION 35 LITERATURE CITED 36 TABLE OF CONTENTS (continued)
Page CHAPTER 3- HABITAT ASSOCIATIONS OF WOODPECKERS 41 AT TILELANDSCAPELEVEL IN MANAGED FORESTS OF THE SOUTHERN OREGON CASCADES
INTRODUCTION 41 OBJECTIVES 42 STUDY AREA 43 METHODS 46 Study Site Selection 46 Bird Sampling and Nest Searches 47 Creating a GIS Layer 48 Statistical analysis 48
RESIJLTS 52 FRAGSTATSanalysis of Landscape Indices 52 EdgeAnalysis 56
DISCUSSION 58 Species-Specific Patterns 58 Edge Effect 60
CONCLUSION 61 L1TERATEJRE CITED 62
CHAPTER 4- CONCLUSIONS AND IMPLICATIONS FOR MANAGEMENT FOR WOODPECKERS IN TILE OREGON CASCADES 66
LARGE DIAMETER SNAGS 66 EDGE 67 NATURAL AGENTS OF DIVERSITY 68 RESEARCH RECOMMENDATIONS 69
LiTERATURE CITED 70 TABLE OF CONTENTS(continued)
Page
BIBLIOGRAPhY 73
APPENDICES 81
Appendix A Maps of eleven study sites showing habitat patch type and spatial arrangement, and location of woodpecker nests at each site 82
Appendix B.Weight file used in FRAGSTATS analysis to characterize vegetation characteristics of 200-ha landscapes where woodpecker nests were located on the Diamond Lake Ranger District 94
Appendix CClass level indices generated in FRAGSTATS to 95 characterize eleven 200-ha landscapes on the Diamond Lake Ranger District, Umpqua National Forest, southern Oregon Cascades, 1995-1996 LIST OF FIGURES
Fi2ure Page
1. Location of study sites for woodpecker habitat association study on the Diamond Lake Ranger District Umpqua National Forest, Southern Oregon Cascades 17
2. Diameter of woodpecker nest trees and random trees. 25
3. Tree-top condition for trees used for nesting by hairy woodpeckers (HAWO), red-breasted sapsuckers (RBSA), and northern flickers (NOFL) and for random trees. 25
4. Decay class for nest trees of hairy woodpeckers (HAWO), red-breasted sapsuckers (RBSA), northern flickers (NOFL) and random trees. 28
5. Treespecies groups of nest trees for hairy woodpeckers (HAWO), red-breasted sapsuckers (RBSA), northern flickers (NOFL) and random trees. 28
6. Location of study sites for woodpecker habitat association study on the Diamond Lake Ranger District Umpqua National Forest, Southern Oregon Cascades. 45
7. Association between the log of mean forest patch size and density of red-breasted sapsucker nests over landscapes. 53
8. Association between proportion of mature forest and density of northern flicker nests over landscapes. 54
9. Association between proportion of mature forest and density of hairy woodpecker nests Over landscapes. 55 LIST OF TABLES
Table Pa2e
1. Definition of habitat characteristics measured in plots centered around woodpecker nest trees and random trees. 21
2. Number of nests of woodpeckers found during 1995-96 breeding seasons in 11 study-sites located in the Southern Oregon Cascades. 24
3. Mean values nest tree characteristics for hairy woodpeckers, northern flickers and red-breasted sapsuckers. 27
4. Results of multiple logistic regression comparing tree characteristics for nests of red-breasted sapsuckers compared to random trees. 29
5. Results of multiple logistic regression comparing tree characteristics nests of hairy woodpeckers compared to random trees. 30
6. Results of multiple logistic regression comparing tree characteristics nests of northern flickers compared to random trees. 31
7. Patch type classification system used for FRAGSTATS analysis to classify vegetation in 200-ha study sites. 49
8. Definitions of class level habitat variables generated by FRAGSTATS to describe landscape characteristics in 200-ha sites. 51
9. Location of nests of red-breasted sapsuckers, hairy woodpeckers and northern flickers with respect to edge within landscapes. 56
I 0. Comparison of distance from an edge between woodpecker nests and random locations, after all points located directly on an edge were removed 57 PREFACE
Timber harvest in the Cascade region of the Pacific Northwest dates back to1827 with the construction of the first water-driven sawmill in the region (Hams1984). The Pacific Northwest region is known for its ability to produce timber, and ismarveled for expansive conifer-dominated forests and trees of massive proportions. Timberharvest during the last 50 years has created landscapes comprised of patchworkconfigurations of older forest interspersed with younger seral-stage stands (Franidin and Forman 1987). The forest fragments created acquire characteristics of habitat islands in proportion to their size and length of time since isolation (Harris 1984). This practice has led to a severe reduction in the amount of low elevationold-growth forest, with the total acreage amounting to 10% of its original volume. In addition, the continuous nature of the forest prior to logging has been altered. Large-scale changes in landscape pattern have occurred within the last century, yet we have limited understanding of the consequences of altered landscape pattern on wildlife species and ecosystems. We need to reexamine whether traditional silvicultural methods of timber harvest are compatible with emerging concerns of biodiversity protection and overall forest ecosystem integrity. In the past, scientists have approached problems through studies that examine organisms or processes using the reductionist philosophy rooted in mechanistic science. Mechanistic theory was developed and used with great success by pioneers such as Francis Bacon, René Descartes, and Isaac Newton. It is based upon the premise that any system or process can be understood by dissecting it to its smallest components. Investigation of the individual pieces, according to the theory, will lead to an understanding of the whole (Capra 1982). One of the major disadvantages of using the reductive, mechanistic approach is its severe limitations when applied to complex biotic systems. In many cases, study designs are oblivious to the importance of scale in examining wildlife habitat associations. The manner in which we make observations directly influences the conclusions we can draw from our observations, complete disregard for the effects of scale can lead to false or incomplete conclusions. Realizing the importance of scale, many researchers have begunexploring other methods of research design to study wildlife populations. I developed my research questions based on a perceived need tomodify our perspectives in research, to move beyond stand and organism-levelstudies towards studies that examine dynamic systems and their related interactions.Forests have become more intensively managed to meet a number of demandsincluding timber production, recreation, and maintenance of habitat for wildlife species. Forest managers need information to help direct them in development and implementation of management plans to satisfy these demands and maintain functioning forest systems. Because management strategies are addressing problems that occur at the landscapelevel, it is appropriate and necessary that we attempt to examine processes at this scale. My thesis is arranged into three chapters; in Chapter One, I present the contextual framework and justification for my research, to provide relevant information to help the reader interpret the results in a meaningful way. To understand and appreciate the problem, one must have an idea of the multiple dimensions in which it exists. In Chapter Two, I present and discuss the results of the study at the smallest scale of resolution, habitat associations of woodpeckers at the nest tree level. In Chapter Three, I describe and discuss the results of the landscape level analyses, which considered habitat associations of woodpeckers on a broader scale, examining relationships between nesting woodpeckers and characteristics at the patch and landscape dimensions. I conclude with a brief chapter integrating the results of both the nest tree leveland landscape level analyses, and present management recommendations based upon my findings. McGarigal (1993) asserted: "there has been a great deal of theoretical and conceptual work on populations or spatially complex environments, but few empirical studies to confirm or refute these ideas." My goal in undertaking this research project is to ask the bigger question, to look beyond current perspectives and attempt to examine wildlife habitat interactions at multiple scales. I do not intend to produce a universal solution to questions regarding habitat management for maintenance of viable woodpecker populations. Rather, I wish to approach that question from a different angle, and perhaps shed light on how to frame appropriate questions and design research projects that examine wildlife-habitat relationships in a contextual way.
LITERATURE CITED
Capra, F. 1982. The turning point: science, society and the rising culture. Bantam. New York, N.Y. 464 pp.
Franklin, J.F., and R.T. Forman. 1987. Creating landscape patterns by forest cutting: ecological consequences and principles. Landscape Ecology 1: 5-18
Harris, L.D. 1984. The fragmented forest. University of Chicago Press, Chicago, I.L. 2llpp.
McGarigal, K. 1993. Relationship between landscape structure and avian abundance patterns in the Oregon Coast Range. Ph.D. Dissertation. Oregon State University, Corvallis. 273 pp. Habitat Associations of Woodpeckers at Multiple Scales in Managed Forests of the Southern Oregon Cascades
CHAPTER 1 Introduction
CONCEPTUAL BACKGROUND The Cascades region of the Pacific Northwest represents a bioregion with a great diversity of habitats supporting a rich vertebrate community. The forests of the Pacific Northwest are home to 251 species of breeding birds, 125 species of mammals, 26 species of amphibians, and 28 species of reptiles (Harris 1984). In recent decades, there has been increasing concern over the way in which our public lands have been managed, specifically questioning whether management activities are compatible with the conservation of unique flora and fauna. Alterations in landscape structure due to timber harvest have had negative effects on some wildlife populations and alarming declines in populations of many avian species have been documented. The marbled murrelet (Brachyramphus marmoratus), a sea-bird that nests in old-growth trees, was listed as threatened in 1994. The northern spotted owl (Sirix occidentalis caurina), another old-growth associate, was added to the list of threatened species in 1991. The decision to list the owl has had a profound effect, throwing the field of natural resource management into a quagmire in an attempt to resolve conflicts between resource extraction-based industry and wildlife management of threatened species. These issues continue to challenge scientists and land managers, who are grappling with the complex problem of managing natural resources to meet public demands while protecting habitat for indigenous species. Public involvement in forest management issues has increased, and with it has come pressure to change forest management policy and practices. Demands on federal forest lands have steadily increased, both in the form of recreational pressure and 2
demand for timber. Timber harvest is the dominant land management activity in the federal forests of the Pacific Northwest (Edgerton and Thomas 1977). For the past 40 years, harvest has been conducted using even-aged managementstrategies. This method has been followed primarily for economic and silvicultural reasons, because it allowsfor relatively easy forest regeneration, slash disposal and road building. The resulting landscape is characterized by patchwork configurations of mature forest interspersed with even-aged regenerating stands (Franidin and Forman 1987). The openings created by timber harvest disrupt the continuity of the forest and leave behind a patchwork mosaic, reducing connectivity between forest patches (Harris 1984). With the recent shift to create management strategies addressing ecosystem-level processes, we need to consider the ecological consequences of forest fragmentation (Harris 1984). It is imperative that we utilize silvicultural methods that retain overall ecosystem integrity, and preserve the inherent biodiversity of forest ecosystems. The foundation for understanding ecological processes on a broad scale is built upon the evolving field of landscape ecology. This branch of ecology is arelatively new scientific field. It focuses on the ecological structure, function and change of a landscape (Forman and Godron 1981, Naveh and Lieberman 1984, Forman and Godron, 1986). Detecting relationships at a landscape level poses a challenge for scientists, and requires a major change in perception from traditional plot-level analyses. Wiens (1976) asserted that ecologists have been slow in recognizing this need because humans are more likely to perceive and study such phenomena on anthropocentric scales that are in accordance with their own experience. In designing studies, we frequently select scales that appear to be appropriate to observe phenomena, though that scale may be completely irrelevant to the organism being studied. Often in field studies, observations are conducted over short time periods and on small spatial scales, as a matter of convenience, despite the fact that this can result in a skewed interpretation of organism response. As Wiens (1976) states, "we need nonarbitrary operational ways of defining and detecting scales." Though a noble objective, it is difficult to determine how to implement this in study design. The crucial step in the design stage of a study is to define habitat patch from the perspective of the organism. Environments can be patchy over landscapes, creating 3 patterns of heterogeneity that in turn influence vertebrate population abundance and distribution (McGarigal 1993). The term patch must be defined within the context of the question(s) of interest and the organism(s) being studied. Many definitions of patch are found within the ecological literature. A resource patch is defined as "a nonlinear surface area differing in appearance from its surroundings" (Forman and Godron 1986). Wiens (1976) defined a patch as:
"relatively discrete areas of spatial or temporal domain of relatively homogeneous environmental conditions where patch boundaries are distinguished by discontinuities in environmental character from their surroundings of magnitudes that are perceived by the animal under consideration"
Patches are dynamic and inherently possess unique qualities such as vegetation association, size, shape, interior-to-edge ratio, and boundary characteristics. Patches exhibit an internal structure, referred to as patch hierarchy, in which heterogeneity occurs at successively finer scales. In any study of wildlife-habitat associations, patch dynamics must be considered because they influence species diversity and abundance. Our ability to detect patterns within a patch are a function of both extent and grain. Extent is the coarsest scale of heterogeneity to which an organism responds, and is determined by the home range of the organism (Kotliar and Wiens 1990). For practical purposes extent is usually restricted to the area encompassed by the study. Grain represents the individual units of heterogeneity within a patch to which a species responds (Wiens 1989). Within the context of a research study, these terms must be defined as the upper and lower limits of resolution of a study. Grain response may be either fine or coarse, and should be defined from an organism-centered perspective. Grain size is determined by an organism's perceptual abilities. A fine-grained response occurs when an organism does not perceive any heterogeneity within the landscape beyond a sufficiently fine scale of resolution, i.e. 'grains' are used in direct proportion to their availability in the environment. Conversely, a coarse-grained response is that level of patchiness at which patterns of use by an 4
organism are observed, in other words, 'grains' are used disproportionate to their availability. Using the analogy of a foraging woodpecker to illustrate this concept, consider that the home range of the woodpecker is 12-hectares. Within the 12-ha patch, the woodpecker may perceive the entire area as homogeneous with respect to stand structure, no behavioral response to stand structure is observed. The woodpecker could be said to exhibit a fine-grained response with regard to stand structure at that particular scale. If small groups of snags are found within the 12-ha patch that contain high densities of insects that are desired food items, the woodpecker may select these trees for foraging disproportionate to availability. This behavior illustrates a coarse-grained response to snag distribution throughout the stand. Patches are critical in determining the relative stability of a population in a landscape. den Boer (1981) hypothesized that a population will be more persistent, and therefore less at risk of extinction, in a landscape characterized by heterogeneity versus one of a similar population occupying a homogeneous environment. Populations within a heterogeneous landscape are hypothesized to be less susceptible to natural catastrophes because even if populations in local patches go extinct, individuals from other patches remain to recolonize the vacant patches (den Boer 1981). Van Home (1983) asserted that patches with superior habitat function as 'source' patches. Source patches are crucial because they are thought to provide dispersers to the overall population, enhancing genetic diversity and providing stability to the population. Patches with inferior habitat may be 'sinks', or patches where organisms can exist, but do not persist over time due to decreased reproduction (Van Home 1983). An understanding of patch dynamics is important when examining the effects of fragmentation on wildlife species, because fragmentation essentially restructures the inherent patch hierarchies and can decrease the diversity of patch hierarchies in the remaining habitat (Wiens 1976). By determining how patch hierarchies affect foraging behavior, nest-site selection, and population dynamics, we may be able to better predict the consequences of habitat fragmentation on wildlife populations. When investigating habitat associations of an organism over a landscape, temporal aspects must be considered as well. Data collected represent that landscape at a 5
very discrete place in time as well as in space. For example, the rateof decay and fall of snags within a stand could drastically alter the quality of that stand from theperspective of a nesting woodpecker. These dynamics and their effects on the organism of interest can only be captured by studies that span a longer time period. Studiesconducted over a
1 -to2-year period may represent only a 'snapshot' of reality, and may not represent 'normal' conditions. This may lead the researcher to false conclusions. Taking into consideration heterogeneity that exists over landscapes, I designed a study to examine habitat associations of primary cavity-nesters, or woodpeckers, in the Oregon Cascades. I measured characteristics at multiple scales of resolution that I thought may be influencing response of woodpeckers over a patchy landscape. To define patch structure from the perspective of a woodpecker, foraging, nesting and roosting habitat requirements should be examined. Due to logistical constraints, I chose to focus my study on associations for nesting habitat, and the conclusions I draw must be framed within that context.
RESEARCH NEED AND JUSTIFICATION Primary cavity-nesters create habitat for a suite of secondary cavity-users. About 53 species of birds and small mammals in Oregon use cavities excavated by woodpeckers, including the mountain bluebird (Sialia cirrucoides), saw-whet owl (Aegolius acadicus), violet-green swallow (Tycinata bicolor), and northern flying squirrel (Glaucomys sabrinus) (Bull 1 986a). In addition, recent studies have suggested that cavities provide critical roosting habitat for several species of bats (Neitro et al. 1985, Hunter 1990, Ormsbee 1996). Woodpeckers play a role in maintenance of forest health, by suppressing insect levels (Otvos 1965). Woodpeckers also initiate and promote decay and nutrient cycling by excavating in dead trees and down woody material (Bull 1977, Mannan et al. 1980). They feed primarily on insects found in decaying trees, reducing the loss of wood products to pest outbreaks (Otvos 1965). Carpenter ants (Camponotus spp.), bark beetles (Dendroctonus spp.), woodboring beetles (Buprestidae, Cerambycidae.) and Douglas-fir tussock moths (Orgyia pseudotsuga) are common prey for woodpeckers (Bull 1977). 6
Dead trees, or snags, are a primary resource requirement of woodpeckers (Mannan et a!. 1980). Woodpeckers select larger diameter snags for foraging, which may be due in part to the abundance of insects found in larger snags (Mannan et al. 1980, Madsen 1985, Nelson 1989, Mariani and Manuwal 1990, Weikel 1997). Woodpeckers have been shown to select the tallest and largest snags available in all forest types and to place nest cavities at the largest possible diameter above the base of the tree (Mannan et al. 1980, Raphael and White 1984, Madsen 1985, Mellen 1987, Nelson 1989). Higher nest location may enhance protection from predators (Nilsson 1984), and weaker excavators, such as the northern flicker, choose cavities near the top because decay conditions have sufficiently weakened that portion of the tree to allow excavation (Nelson 1989). Bull (1 986c) found that woodpeckers prefer dead trees> 35 cm in diameter for nesting. She defined selection criteria for nest trees based on tree decay, condition, species, diameter, height and location. Mannan (1977) found that in general, cavity- nesters foraged on and nested in Douglas-fir snags averaging more than 60 cm in dbh and with less than 100% bark cover. Most were> 15 m tall, usually with broken tops, few branches, and with decayed sapwood and heartwood. Snags of this type were found primarily in forests over 70 years old, and consequently, populations of cavity-nesters were concentrated in forested stands 100-200 years old or greater (Mannan 1977). Nelson (1989) confirmed these findings, showing that densities of cavity-nesting birds are higher in mature and old-growth forests than in young closed-canopy forests of the Oregon Coast Range. Researchers have recommended that habitat plans for cavity-nesters provide tall, large diameter snags >50 cm dbh within harvest units. Because both species richness and density have been positively associated with snag density (Schreiber and de Calesta 1992), recommendations provide for maintenance of 3 to 11 large snags per acre depending on site conditions and geographic location. When and where snag densities are inadequate to maintain viable populations of cavity-nesters, habitat should be supplemented by recruiting green trees into the snag population through fungus innoculation, girdling, or topping. 7
Snag Management During the past 10-15 years land management agencies have identified snags as an important limiting resource for woodpeckers in environmental impact statements (Neitro et a!. 1985). Previously, snags were removed as additional sources of wood fiber because commercial forestry practices emphasized maximum utilization of the resource and also because snags and defective trees were seen as potential hazards to workers involved in tree felling operations (Haapanen 1965, Jackman 1974, Neitro et al. 1985). Recently, guidelines have been established to ensure that some snags are retained as wildlife habitat. When snags must be created artificially, snag management can be costly. Some land management agencies, including the USDA-Forest Service have begun creating snags for wildlife using a variety of techniques; including topping green trees using explosives or chainsaws, girdling the tree at the base, or inoculating the tree with a fungus to promote heart rot (Greg Filip, pers. comm, October 15, 1995.) Each method shows promise, but more follow-up monitoring needs to be done to assess the success of these methods. Chambers et al. (1997) monitored snags artificially created by topping in the Oregon Coast Range to detect usage by cavity-nesters. They found considerable use of snags by woodpeckers in the first five years following creation. Bull and Partridge (1986b) reported that topping trees with a saw was the most cost-effective method, producing trees that stood longer and were used most often by cavity-nesters. Explosives and herbicides also were tested, but did not appear to be as effective as topping.
Woodpeckers of the Southern Oregon Cascades The northern flicker (Colaptes auratus), red-breasted sapsucker (Sphyrapicus ruber) and hairy woodpecker (Picoides villosus) have been described as habitat generalists, utilizing a wide range of forest types (Madsen 1985, Nelson 1989). Madsen (1985) found that flickers and hairy woodpeckers generally selected nest sites in open stand conditions, adjacent to or on the edge of forested areas. Pileated woodpeckers (Dryocopus pileatus) however, have specific habitat requirements and occupy a narrower niche within forest ecosystems. They have been found to nest in areas with high tree and snag density, and require snags with comparatively larger diameters for nesting (Bull 8
1977, 1980, 1986a, 1986c). Madsen (1985) found pileated woodpeckers in unmanaged stands dominated by old-growth characteristics (e.g., large trees, snags, downed woody material) or in managed stands where efforts had been made to retain snags. Comparatively little is known about the nesting and foraging habitat requirements of the black-backed woodpecker (Picoides arcticus). Nests of black-backed woodpeckers have been documented in recent burn areas (McClelland 1977). Bull (1980) found this species predominantly in open ponderosa pine (Pinus ponderosa) forests in eastern Oregon. On the Diamond Lake Ranger district of southwestern Oregon, nests have been found in both live and dead lodgepole pine (Pinus contorta) trees (Keith Graves, pers. comm., May 15, 1994). Significant gaps remain in our knowledge of specific requirements needed by individual species of woodpeckers with regard to snag size and decay class (Mannan et al. 1980). In order to develop and implement plans that will succeed in maintaining habitat for woodpeckers, it is critical that we identify what habitat characteristics are used by woodpeckers.
OBJECTIVES Within the context outlined above, I developed specific study objectives. The primary objective of this study was to estimate nest-site characteristics of woodpeckers at multiple scales in managed forests of the Southern Oregon Cascades. Steps in this process that contributed to the overall objective included:
•To estimate nesting habitats of resident woodpeckers including the red-breasted sapsuckers, black-backed woodpeckers, northern flickers, hairy woodpeckers, and pileated woodpeckers at the nest-tree, patch, and landscape levels. •To determine whether an association existed between percent mature forest and abundance of woodpecker nests over a landscape, using spatial pattern analysis and regression analysis. 9
•To develop models to determine habitat associations of woodpeckers at the nest-tree, patch, and landscape level. •To develop guidelines that can be used in management prescriptions to maintain habitat for woodpeckers.
LiTERATURE CITED
Bull,E.L. 1977. Habitat utilization of the pileated woodpecker, Blue Mountains, Oregon. M. S. Thesis. Oregon State University, Corvallis. 58 pp.
Bull, EL. 1980. Resource partitioning among woodpeckers in northeastern Oregon Ph.D. Dissertation. University of Idaho, Moscow. 109 pp.
Bull, E.L. 1986a. Ecological value of dead trees to cavity-nesting birds in northeastern Oregon. Oregon Birds 12: 9 1-99.
Bull, E.L. 1986b. Methods of killing trees for use by cavity-nesters. Wildlife Society Bulletin 14:142-146.
Bull, E.L., S.R. Peterson, and J.W. Thomas. 1986c. Resource partitioning among woodpeckers in northeastern Oregon. USDA Forest Service Research Note, PNW-444. 19 pp.
Chambers, C.L., T. Carrigan, T.E. Sabin, J.Tappeiner, and W.C. McComb. 1997. Use of artificially created Douglas-fir snags by cavity-nesting birds. Western Journal of Applied Forestry 12:93-97. den Boer, P.J. 1981. On the survival of populations in a heterogeneous and variable environment. Oecologia 50: 3 9-53.
Edgerton, P.J. and J.W. Thomas. 1977. Silvicultural options and habitat values in coniferous forests. Paper presented at the Nongame Bird Habitat Management in Coniferous Forests of the Western United States, Portland, Oregon. Feb. 7-9, 1977. pp56-65.
Filip, G. Personal Communication. October 15, 1995. Forest pathology expert. Oregon State University, Corvallis. 10
Forman, R.T. and M. Godron. 1981. Patches and structural components for a landscape ecology. Bioscience 31: 73 3-740.
Forman, R.T. and M. Godron. 1986. Landscape ecology. John Wiley and Sons, New York, N.Y. 619 pp.
Franklin, J.F., and R.T. Forman. 1987. Creating landscape patterns by forest cutting: ecological consequences and principles. Landscape Ecology 1: 5-18
Graves, K. 1995. Personal Comninication. March 15, 1997. USDA Forest Service Natural Resource Manager, Diamond Lake Ranger District, Umpqua National Forest.
Haapanen, A. 1965. Bird fauna of the finnish forests in relation to forest succession. I. Am. Zoo!. Fenn. 2:153-196
Harris, L.D. 1984. The fragmented forest. University of Chicago Press, Chicago, I.L. 211 pp.
Hunter, M.L. Jr. 1990. Wildlife, forests and forestry: Principles of managing forests for biological diversity. Prentice-Hall, Inc. Englewood Cliffs, N.J.
Jackman, S.M. 1974. Woodpeckers of the Pacific Northwest; their characteristics and their role in the forest. M.S. Thesis, Oregon State University, Corvallis. 147 pp.
Kotliar, N.B., and J.A. Wiens. 1990. Multiple scales of patchiness and patch structure: a hierarchical framework for the study of heterogeneity. Oikos 59: 253-260.
Madsen, S.J. 1985. Habitat use by cavity-nesting birds in the Okanogan National Forest, Washington. M.S. Thesis. University of Washington, Seattle. 113 pp.
Mannan, R.W. 1977. Use of snags by birds, Douglas-fir region, western Oregon. M.S. Thesis. Oregon State University, Corvallis. 113 pp.
Mannan, R.W., E.C. Meslow, H.M. Wight. 1980. Use of snags by birds in Douglas-fir forests, western Oregon. Journal of Wildlife Management 44: 787- 797.
Mariani, J.M. and D.A. Manuwal. 1990. Factors influencing brown creeper (Certhia americana) abundance patterns in the southern Washington Cascade Range. Studies in Avian Biology 13:53-57. 11
McClelland, B.R. 1977. Relationships between hole-nesting birds, forest snags and decay in western larch, Douglas-fir forests of the northern Rocky Mountains. Ph.D. Dissertation. University of Montana, Missoula. 495 pp.
McGarigal, K. 1993. Relationship between landscape structure and avian abundance patterns in the Oregon Coast Range. Ph.D. Dissertation. Oregon State Corvallis. 273 pp.
Mellen, T.K. 1987. Home range and habitat use by pileated woodpeckers. M.S. Thesis. Oregon State University, Corvallis. 96 pp.
Naveh, Z. and A. S. Liebermann. 1984. Landscape ecology, theoiy and application. Springer-Verlag, New York, N.Y. 356 pp.
Nelson, S.K. 1989. Habitat use and densities of cavity-nesting birds in the Oregon Coast Ranges. M.S. Thesis. Oregon State University, Corvallis. 157 pp.
Neitro, W.A., V.W. Binkley, S.P. Cline, R.W. Mannan, B.G. Marcot, D. Taylor, F.F. Wagner. 1985. Snags. Pages 129-168 In Brown, E.R. tech. ed. Management of wildlife and fish habitats in forests of western Oregon and Washington. Part 2 - Appendices.USDA Forest Service, Pacific Northwest Region, Portland, OR.
Nilsson, S.G. 1984. The evolution of nest-site selection among hole-nesting birds: the importance of nest predation and competition. Ornis Scand. 15: 167-175
Ormsbee, P. 1996. Selection of day roosts by female long-legged myotis (Myotis volans) in forests of the central Oregon Cascades. M.S. Thesis. Oregon State University, Corvallis. 63 pp.
Otvos, I.S. 1965. Studies on avian predators of Dendroctonus brevicomis le Conte (Coleoptera: Scolytidae) with special reference to Picidae. Canadian Entomologist 97: 1184-1199.
Raphael, M.G. and M. White. 1984. Use of snags by cavity-nesting birds in the Sierra Nevada. Wildlife Monograph No. 86. 66 pp.
Schreiber, B. and D.S. deCalesta. 1992. The relationship between cavity-nesting birds and snags on clearcuts in wetern Oregon. Forest Ecology and Management 50: 299-3 16.
Van Horne, B. 1983. Density as a misleading indicator of habitat quality. Journal of Wildlife Management 47:893-901. 12
Weikel, J. 1997. Habitat use by cavity-nesting birds in young thinned and unthinned Douglas-fir forests of western Oregon. M.S. Thesis, Oregon State University, Corvallis. 102 pp.
Wiens, J.A. 1976. Population responses to patchy environments. Annual Review of Ecological Systematics 7: 81-120.
Wiens, J.A. 1989. Spatial Scaling in Ecology. Functional Ecology 3: 385-397. 13
Habitat Associations of Woodpeckers at Multiple Scales inManaged Forests of the Southern Oregon Cascades
CHAPTER 2 Nest-Site Characteristics of Woodpeckers in Managed Forests of the Southern Oregon Cascades
INTRODUCTION
Over the past 40 years, the forests of the Pacific Northwest have been managed for efficient extraction of wood products, using even-aged regeneration methods. Until the last decade, snags were typically removed for economic, safety or logistic concerns (Neitro et al. 1985). Removal of snags during logging operations generally has negative effects on populations of cavity-nesting species (Balda 1975, Thomas Ct al. 1975, Franzreb 1978, Evans and Conner 1979, Raphael and White 1984, Zarnowitz and Manuwal 1985). Primary cavity-excavators and secondary cavity-users depend on dead trees, or snags for survival; snags are essential for nesting, foraging, androosting by many species (Bull 1980, Madsen 1985, Nelson 1989, Schreiber and deCalesta 1992, Bunnell et a!. 1997). Thirty-nine species of birds and 14 species of small mammals in the Pacific Northwest are dependent on cavities (Neitro et al. 1985, Bull 1986a). Furthermore, the absence of suitable snags can be a limiting factor for wildlife (Haapanen 1965, Balda 1975, Mannan et al. 1980, Raphael and White 1984). Woodpeckers are primary cavity- excavators, annually excavating fresh cavities for nesting and providing habitat for secondary cavity-users in the process. Secondary cavity-nesters, including the northern pygmy owl (Glaucidum gnoma) and the western bluebird (Sialia mexicana), cannot excavate cavities themselves, but depend upon abandoned woodpecker nest cavities for nesting and roosting habitat. Woodpeckers are an integral part of a healthy forest ecosystem, suppressing insect populations and promoting decay and nutrient cycling by excavating in dead trees (Otvos 14
1965, 1979, 1985; Bull 1977, Mannan et a!. 1980). Woodpeckershave been shown to be important predators on harmful forest insects such as the western pine beetle (Dendroctonus brevicomus), the Engelmann spruce beetle (Dendroctonusengelmanni), and the Douglas-fir tussock moth (Orgyiapseudotsugata) (Knight 1958,Otvos 1965). Because of their integral role in forest ecosystems, woodpeckers aresometimes considered to be a keystone species, and a prime indicator of overall foresthealth (Bunnell et a!. 1997). Numerous studies have shown that woodpeckers select snags for nesting and foraging (Raphael and White 1984, Madsen 1985, Nelson 1989). However, only in the past 15-20 years have land management agencies considered snags as animportant limiting resource for many wildlife species. Guidelines have been developed and implemented to ensure that some snags are retained after harvest. Forest Service regulations require that two hard and one soft snag per acre are left during timber removal (USDA Forest Service Resource Management Plan 1990). Forest plans recommend that, whenever possible, snags> 35 cm be left standing following harvest. Neitro et a!. (1985) recommend leaving 5 snags per hectare, 4.8 of which are soft snags in decay class 4-5, and 0.2 hard snags, in decay class 2-3. These numbers may be inadequate, however, to sustain viable populations of cavity-nesters. Hope and McComb (1994) showed that a majority of biologists in the Pacific Northwest believe that current guidelines for snag management are inadequate to maintain target populations of woodpeckers. Important questions remain to be answered regarding specific snag characteristics preferred by woodpeckers in order to develop management plans that maintain adequate numbers of suitable snags to sustain populations of woodpeckers. These questions must not only examine qualities of the snag itself, but also the environment, or immediate surroundings in which the snag is found. Woodpeckers most likely select snags for nesting based upon a combination of characteristics including tree-top condition, degree of hardness, tree species, and presence or absence of various fungi or insect species. However, woodpeckers probably also consider the characteristics of the surrounding environment when selecting a snag for nesting. For example, does the snag occur in a mature forest stand, or is it in the middle of an open meadow? Are there high densities of 15 snags and large amounts of down woody material in the surrounding area or not? If it occurs in a forested stand, what is the species composition of the trees in the stand? Additionally, studies are needed to examine the differences between co-existing species of woodpeckers, to determine the unique characteristics selected for by each species. Providing habitat to maintain a single species of cavity-nesters may not necessarily meet the needs of other cavity-nesters. With these questions in mind, the goals of my study were: •To estimate characteristics of nest trees selected by hairy woodpeckers, red-breasted sapsuckers and northern flickers •To determine whether microsites surrounding nest trees were associated with nest-site selection by woodpeckers
STUDY AREA The study was conducted in the Central Oregon Cascade Range, on the Diamond Lake Ranger District of the Umpqua National Forest. Study sites are located approximately 70 miles east of Roseburg, Oregon, in Douglas County between 47° 54' and 47° 56' latitude (Fig. 1). Study sites were within the Umpqua National Forest on sites varying in elevation from 1066 m-1463 m in the Umpqua High Cascade sub- province. The climate is characterized by moderate temperatures, wet, mild winters and relatively cool, dry summers (Franklin and Dyrness 1988). The region receives an average of 152 cm of precipitation annually; of which 75-85% occurs between October and April, primarily as rain, with snow at high elevations (Azet and McCrimmon 1990). The study area is located within the western hemlock (Tsuga heterophylla) zone, the most extensive vegetation zone found in the Pacific Northwest (Franklin and Dyrness 1988). In this zone western hemlock is considered the climax species, but this state is rarely reached due to forest management activities (Harris 1984). The overstory is typically dominated by Douglas-fir (Pseudotsuga menziesii), though white fir (Abies concolor), western hemlock, mountain hemlock (Tsuga mertensiana), Pacific silver fir (Abies amabilis), Pacific yew (Taxus and Shasta red fir (Abies commonly occur. 16
Figure 1. Location of study sites for woodpecker habitat association study on the Diamond Lake Ranger District, Umpqua National Forest Southern Oregon Cascades. STUDY SITE LOCATION Sites
Figure 1. 0 Uometei ,col. 30 18
Other species present include: incense-cedar (Calocedrus decurrens), westernredcedar (Thuja plicata), ponderosa pine (Pinus ponderosa) and sugar pine (Pinus lambertiana). Lodgepole pine (Pinus contorta) is found in sites located on the eastern end of the study area. Understory vegetation varies depending onmicrosite conditions, but generally is dominated by golden chinquapin (Castanopsis chrysophylla), Sitka alder (Alnus sinuata), Pacific rhododendron (Rhododendron macrophyllum) and vine maple (Acer circinatum). Buckrush (Ceanothus integerrimus), and snowbrush (Ceanothus velutinus) are found in young, regenerating stands. Ground cover is variable andincludes sword fern (Polystichum bracken fern (Pteridium aquilinum), vanillaleaf (Achlys triphylla), Cascade Oregon-grape (Berberis nervosa), huckleberries (Vacciniumspp.), salal (Gaultheria shallon), and prince's pine (Chimophila umbellata). The study sites were dominated by mixed-conifer forest composed of at least 70%, but not more than 85% Douglas-fir trees. The remaining trees were a mixture of western hemlock, mountain hemlock, true fir, and pine trees. Three sites had a small component of lodgepole pine, not exceeding more than 5% of the total area. The sites were at approximately the same elevation. Volcanic activity has influenced landscape formation found in the study area, depositing a layer of pumice across the landscape. Prior to European settlement in the middle 19th century, the natural disturbance regime of the area was dominated by wild fire events.Since its designation as National Forest land in 1905, the resource planning strategy has emphasized efficient timber production, using even-aged management methods. Due to management practices over the past 40 years, the condition of the landscape is highly variable, consisting of a mosaic of habitat patches of varying age and structure. Five species of woodpeckers are known to breed within the study area; the black- backed woodpecker, hairy woodpecker, northern flicker, pileated woodpecker, and red- breasted sapsucker. Other species of woodpecker, including red-naped sapsuckers (Sphyrapicus nuchalis) and white-headed woodpeckers (Picoides albolarvatus) are residents on the east side of the Cascade Range, but occur only occasionally in the study area. 19
METHODS Study Site Selection Eleven sites within the study area were selected for the study. Site selection criteria were based on the proportion of the area comprised of mature forest, defined as closed-canopy forest with> 40% canopy cover and trees averaging 53 cm in diameter. The proportion of total forested area within study sites varied from 10% to 94%. The remaining sites had intermediate amounts of forest, with the intent to select sites that represented a continuum of amount of forest cover. Six sites were surveyed in 1995, and seven in 1996; two of the sites sampled in 1995 were resampledin 1996. Study sites chosen for woodpecker sampling were 200-ha in size. Home range estimates of resident woodpeckers range from 12 ha for the red-breasted sapsucker to 200 ha for the pileated woodpecker (Bull 1977, Mannan et al. 1980, Brown et al. 1985, Winkler et a!. 1995). I chose 200-ha study sites because they were sufficiently large to include multiple woodpecker home ranges, yet small enough to meet logistical constraints.
Bird Sampling and Nest Searches A 1600- x 1200-rn grid was established at each study site. Five, transect lines were located parallel to one another within each grid, with stations established every 200 meters along transect lines. Station points were used as references for recording locations of nests. I located woodpecker nests by walking along transect lines. Approximately every 200 meters I stopped to listen for woodpeckers. I attempted to visually locate each woodpecker heard. Once located, I observed bird behavior, looking for evidence of a nest. Indicators of nest presence included wood chips at the base of a tree, drumming between pair members, territorial calling, and begging calls of nestlings. I followed woodpeckers for as long as possible or until a nest was discovered. Woodpeckers were not followed if they left the study site. Woodpeckers were easier to locate in open- canopy than in closed-canopy forest conditions, where they are more visible. I used the 20 relative amount of closed-canopy forest present in a transect tostandardize search time. To minimize potential sampling biases, I allotted approximatelytwice the amount of time to conduct searches in closed-canopy than Idid in open-canopy forest. All sites were visited a minimum of five times between 10 Apriland 14 July in 1995 and 1996, with 6 sites sampled in 1995 and 7 in 1996. Surveys wereconducted between 0500 and 1500. Potential nests were marked and laterrevisited to confirm nesting status. Nests were considered to be confirmed when membersof a pair were observed excavating a cavity, exchanging incubation duties, or feeding fledglings at a cavity entrance, or when fledglings were seen or heard in the cavity.
Habitat Sampling Nest Trees--For each nest tree I recorded species, height, diameter at breast height (dbh), condition (broken or unbroken top), decay class from 1-5 based on Cline et al. 1980, and height and orientation of nest (Table 1). Tree and nest height were measured with a clinometer, dbh with a diameter tape, and cavity orientation by compass. I visually estimated bark cover as percentage of bark remaining. Canopy cover was measured using a moosehorn (Garrison 1949). I recorded readings at 6 bearings 6 meters from the base of the tree and calculated the average to obtain an estimate of percentage overstory canopy. In addition, cover was estimated visually to the nearest 5% atfive different canopy heights; 0-5, 5-10, 10-20, 20-30, and >30 m above the ground. I collected data on ground cover composition, densities of shrubs and trees, and canopy cover in a 0.04-ha plot surrounding each nest tree. Treesand shrubs were counted and species and dbh of all stems >10 cm dbh were measured. Snags greater than 10 cm in diameter were recorded; species was recorded if it could be determined. I visually estimated percent cover of shrub, herbaceous material, bare ground, and litter. Random plots--To sample random snags, I selected a random compass bearing and distance within the home-range of the nesting bird. Using the nest tree as the starting point, the random bearing and distance were followed until arriving at the designated point. The nearest snag to that point was selected for habitat analysis. Data collected for 21
Table 1. Definition of habitat characteristics measured in plots centered around woodpecker nest trees and random trees.
Habitat Variable Description Tree diameter Diameter at breast height (cm) Tree height Height of tree (m) Nest height Height of nest (m) Nest cavity orientation Direction of cavity opening (degrees) Tree top condition Top intact or broken Decay stage Category from 1-5 (Cline et al. 1980) Bark Percentage of bark remaining on bole Cavity Number of visible nest cavities on tree Plot stems Measure of diameter and species for live tree stems> 10 cm within 0.04-ha plot
Stand density index (SDI) Reineke's index of stand density accounting for average tree size and density in the 0.04-ha plot (Mctague and Patton 1989)
Mean diameter of plot stems Calculation of mean diameter of all trees in 0.04-ha plot
Vertical structure Percent foliage in 5verticallayers from 0- 70 m
Canopy cover Mean percent of overstory canopy cover in 0.04-ha plot
Ground cover Percent cover of ground vegetation (all species) 22
nest trees and vegetation characteristics of random plots wereidentical to those recorded for nest trees.
Data Analysis I performed analyses that compared habitat characteristics of nest sites tothose of random sites for hairy woodpeckers, red-breasted sapsuckers and northernflickers. Sample sizes for pileated woodpeckers (n=8), red-naped sapsuckers (n1), andblack- backed woodpeckers (n=1) were inadequate to analyze statistically. Variables were analyzed using PROC UN! VARIATE (SAS Institute, Inc. 1990) to obtain means, ranges, and standard deviations. Univariate tests showed that many of the variables were not normally distributed. In addition, many of the variables were categorical, with 2 to 4 levels of response. Categorical data are common among habitat association studies, and are best analyzed by using multivariate techniques (Ramsey eta!. 1994). Multiple logistic regression was chosen for analysis. Logistic regression allows habitat variables to be measured on continuous scales, and, unlike discriminant function andrank-based tests, is not sensitive to non-normality (Ramsey et al. 1994). I used PROC GENMOD (SAS Institute, Inc. 1997) to perform logistic regression. The procedure fits a generalized linear model to the data using maximum likelihood estimations of parameters. To begin the process of variable selection, variables were considered based on likely biological relevancy to cavity-nesters. Correlation analysis revealed that diameter and tree height were highly correlated (r =0.62);and height was dropped from the analysis because dbh has been shown to be a stronger indicator of suitable habitat for woodpeckers (Nelson 1989). I pooled tree species into four categories: Douglas-fir, white fir, hemlock, and other. The "hemlock" category included mountain and western hemlock, and "other" included pines (ponderosa, lodgepole, and western white) and true firs other than white fir (Shasta red, Pacific silver fir). This categorization was done to give the model greater power in detecting differences between nests and random trees. I performed logistic regression analysis for each species of woodpecker independently. I constructed single-variable logistic models using each independent variable and used the 23 drop-in-deviance test (Hosmer and Lemeshow 1989) to determine the amountof variance accounted for by each variable. The model that had the greatestsignificant drop in deviance was chosen. This process was repeated iteratively, addingadditional variables one at a time to the model. Variablesresulting in the greatest drop-in-deviance were added to the model if they were significant at the P <0.05 level. In this manner, amodel was developed for each woodpeckerspecies predicting the probability of use of a particular tree for nesting. Although models for different species of woodpeckers are not directly comparable to one another, the results identify the suite of variables most closely associated with each species in nest-site selection. Selection of random trees was 'matched' to nest trees. This design was used to eliminate some of the extraneous variation inherent in a study of this magnitude. Because logistic regression assumes samples are not paired, I also used an alternative approach that is commonly used in matched case-control studies, in which the dependent variable is stratified based on the variables believed to be associated with the outcome (Hosmer and Lemeshow 1989). My study could be interpreted as a matched-case study, in which the dependent variable NEST, is stratified based on a definable area, in this case home range. The design was developed for more practical reasons, however,it made sense biologically, and therefore is not truly a paired design. Therefore, I performed the final analysis using logistic regression. The results of this analysis and those of the logistic regression analysis are essentially identical, and only the results of logistic regression are presented here.
RESULTS I observed seven species of primary cavity-excavators in the study area during the 1995-1996 breeding seasons. I found 91 nests of woodpeckers in 1995, and 82 nests of woodpeckers in 1996 (Table 2). 24
Table 2. Number of nests of woodpeckers found during 1995-96 breeding seasons in 11 study-sites located in theSouthern Oregon Cascades.
Woodpecker Species Number Number Total nests-1995 nests-1996
Red-naped sapsucker 0 1 1
Red-breasted sapsucker 36 32 68
Hairy woodpecker 14 18 32
Black-backed woodpecker 1 0 1
Northern flicker 35 28 63
White-headedwoodpeckerb 0 0
Pileated woodpecker 5 2 7
Totals 91 82 173
bTwoindividuals were observed at Bear Site, no nests were located
Characteristics of Nest Trees Tree diameter and height- Woodpeckers nested primarily in large-diameter, dead trees (Fig. 2). The average diameter for nest trees was 79 cm, varying only slightly among species (Table 3). Mean height of nest trees was 24.6 m, with a rangeof 4-80 m. Tree diameter and tree height were highly correlated (r= 0.62). Mean nest height was 18 m and although this number varied among species, the variation was notstatistically significant (Table 3). Tree condition- Eighty-six percent of all nests were found in dead trees, and all of the live trees with nests showed visible signs of decay. Usually a portion of the tree was dead or diseased, suggesting the presence of some type of fungal rot, or previous injury to the tree. 25
60
50
• West
U Random
0 0 0 0 0 0 0 (9 V 0 0 • (0 0It0 A (0 0
Diameter(cm) Figure 2. Diameter of woodpecker nest trees versus random trees.
100
90
80
• Broken
Ulniact © 40
30
20
10
0 HAWO RBSA NOFL RAND Nesttree and random trees
Figure 3. Tree-top condition for hairy woodpeckers, red-breasted sapsuckers, northern flickers and random trees. 26
All species most frequently used trees with broken tops:81% of nest trees of red- breasted sapsuckers and hairy woodpeckers and 94% of nest treesof northern flickers had broken tops. Flickers used broken-top snags the majority of thetime, with 94% of all nests found in trees with broken tops. Only 4 outof 63 northern flicker nest trees had intact tops(Fig. 3). Decay- Of 163 nest trees, 14% were in decay stage 1,45% indecay stage 2 and 30% in decay stage 3; and 10% in decay stage 4 trees. There weredifferences in pattern of use of snags in different decay stages among species of woodpeckers (Fig.4). Twenty- two percent of nests of red-breasted sapsuckers were in decay stage1 trees, and only 9- 10% of nests of hairy woodpeckers and northern flickers were in decay stage 1 trees. Nests of red-breasted sapsuckers were almost exclusively in hard trees,98% of nests of this species were found in decay stage 1, 2, or 3 trees. Nest trees of hairywoodpeckers were typically of intermediate hardness; 82% of nestsof this species were found in trees in decay stages 2 or 3, 9% were in decay stage 1 trees, and 9% were in trees of decay stage 4. Northern flickers used soft snags for nesting; 32% of nests werelocated in trees in decay stage 3 and 17% were found in decay stage 4 trees. Of the remaining nests, 10% were in decay stage 1 trees, and 41% in snags in decay stage 2. Tree species-Douglas-fir was most often used for nesting by cavity-nesters; 47% of all nest cavities were located in Douglas-fir, 25% of the nests were in white fir trees, and 15% were in hemlock trees (Fig. 5). The remaining nests occurred in pine and other species of true fir besides white fir. There were not any notable differences among species of woodpeckers in pattern of use of tree species.
Multivariate analysis Red-breasted sapsucker- Trees used by red-breasted sapsuckers for nesting were larger in diameter than random trees. Nest trees had a mean dbh of 80.5 cm whereas the mean dbh of random trees was 42.1 cm. The logistic regression modelpredicted that red- breasted sapsuckers were 10.5 times more likely to use a tree for nesting for each 10 cm increase in diameter (P <0.0001) (Table 4). Sapsuckers nested more frequently in trees 27
Table 3. Mean values of nest tree characteristics for hairy woodpeckers, northern flickers and red-breasted sapsuckers. Standard deviations are reported (below the mean values).
VARIABLE Hairy Northern Red-breasted Random woodpecker flicker sapsucker Trees (n=32) Diameter (cm) 80.4 77.7 80.5 42.1
. (31.4) (31.3) (28.7) (31.9) Height (m) 28.0 19.7 27.5 15.2 (16.4) (12.8) (12.3) (12.9)
Height of nest 18.6 14.3 20.4 N/A . (m) (12.6) (10.8) (10.3) Bark 68.5 66.7 90.3 69.0 remaining (38.0) (36.6) (21.9) (38.6) Number of 5.5 5.9 8.1 0.85 cavities (5.2) (7.5) (8.1) (2.7) 28
IHAWO • ABSA •NOFL •RAND
Decay 1 Decay 2 Decay 3 Decay 4
DecayClass Figure 4. Decay class for hairy woodpeckers, red-breasted sapsuckers, northern flickers and random trees.
50
I HAWO •RBSA BNOFL •RAND
Ui 0 C.) C.) Ui Cl) 0 I
Treespecies
Figure 5.Treespecies for nests of hairy woodpeckers, red-breasted sapsuckers, northern flickers and random trees. 29 with broken tops than in those with intact tops (P <0.0108). Trees withbroken tops were 2.6 times more likely to be used for nesting by red-breasted sapsuckersthan were those with intact tops. All characteristics measured at the plot level wereinsignificant. After accounting for dbh and tree-top condition, sapsuckers did not seem to be using treesfor nesting based on vegetation composition and structure within a 0.04-ha areaaround the nest.
Table 4. Results of multiple logistic regression comparing tree characteristics for nests of red-breasted sapsuckers compared to random trees. Variables are presented in the order in which they appeared in the model, and probability of use is based upon specified parameters. n68.
Coefficient SE Prob> Condition Probability Chi of use
Intercept -2.60 0.5599 0.0001
Dbh 0.052 0.0092 0.0001 10 cm increase in 10.5-fold dbh increase
Tree-top -1.38 0.5403 0.01 Broken-top 2.6-fold condition increase
Hairy woodDecker- Diameter and decay class significantly distinguished nest trees of hairy woodpeckers from random trees. Mean diameter of nest trees was 80.4 cm versus 42.1 cm for randomly selected trees. Hairy woodpeckers were 10.4 times more likely to select a nest tree for every 10 cm increase in diameter (P <0.0001)(Table 5). Trees in decay stage three were 16.4 times more likely to be selected for use by nesting hairy woodpeckers over trees in decay stage 4 (P <0.0077). After accounting for dbh and decay class, I could not detect any significant habitat relationships at the plot level. 30
Statistical analysis did not reveal any other habitat characteristics that were significantly associated with nest-site selection for hairy woodpeckers.
Table 5. Results of multiple logistic regression comparing tree characteristics for nests of hairy woodpeckers compared to random trees. Variables are presented in the order in which they appeared in the model, and probability of use is based upon specified parameters. n=32.
Coefficient SE Prob> Condition Probability Chi of use
Intercept -3.97 -1.18 0.0008
Dbh 0.04 0.01 0.0003 10 cm increase in 10.4-fold dbh increase
Decay 2.84 1.07 0.008 Decay stage 3 tree 16.4-fold increase 31
Northern flicker- As with the red-breasted sapsuckers, northern flickersselected nest trees based on tree diameter and tree top condition. Themodel predicted a 10.2-fold increase in probability of use for each 10 cm increase in diameter (P <0.0001)(Table 6). Northern flickers are 4.3 times more likely to use a tree with a broken top (P <0.001). Logistic regression did not detect any other nest level characteristics significant in predicting use by flickers.
Table 6. Results of multiple logistic regression comparing tree characteristics for nests of northern flickers compared to random trees. Variables are presented in the order in which they appeared in the model, and probability of use is based upon specified parameters. n=63.
Coefficient SE Prob> Condition Probability Chi of use
Intercept -1.11 0.4980 0.025
Dbh 0.02 0.0070 0.0004 10 cm increasein 10.2-fold
. dbh increase
Tree-top -2.06 0.6304 0.001 Broken-top 4.3-fold condition increase
Tree species I did not detect any significant pattern of use of tree species by individual woodpeckers. The sample size may have been insufficient to detect any differences if they did exist. 32
DISCUSSION
Woodpeckers select nest trees based on a combination of characteristics thatmake a tree suitable for nesting. The 'ideal' treeis not characterized solely by one component, but by the presence and interaction of several factors. Qualities thatdescribe tree suitability from the perspective of a woodpecker are tree species, diameter andheight, wood hardness, tree-top condition, and the amount and integrity of the bark cover.
Diameter Woodpeckers selected large diameter trees for nesting. Less than 3% of all nest trees were smaller than 30 cm in diameter. My results are consistentwith other studies conducted in the Pacific Northwest that have reported woodpeckers using large-diameter trees for nesting (Madsen 1985, Mannan et al. 1980, Raphael and White 1984,Nelson 1989, Lundquist and Mariani 1991, Schreiber and deCalesta 1992). There are several hypotheses that may explain selection of large-diameter trees by woodpeckers. Woodpeckers require a tree large enough to allow excavation of a cavity suitable for the nesting bird. Cavity dimensions of nests of the red-breasted sapsucker range from 15-25 cm in depth, and 10-13 cm in width; northern flicker nestcavities average 46 cm deep by 20 cm wide, but have been recorded as large as 60 by 15 cm;and nest cavities of hairy woodpeckers range from 25 to 30.4 cm in depth (Bent 1964). Woodpeckers are most likely restricted to nesting in larger trees that can accommodate cavities of the appropriate dimensions. Large trees also provide increased protection from severe weather conditions and enhance thermal insulating capabilities (Beebe 1974, Jackman 1974, Karisson and Nilsson 1977, Conner 1979). In March and April, when woodpeckers are beginning to excavate nests in the Oregon Cascades, snow is not uncommon, the weather conditions are often cold and wet for extended periods, and the greater insulation provided by larger trees may increase the chance of nest success. In my study, diameter was confounded with height, suggesting that woodpeckers are using trees that are taller than average and have large diameters. Tall trees allow woodpeckers to place nest cavities further from the ground, offering better protection 33 from potential predators such as pine martens (Martes americana) andweasels (Mustela); (Nilsson 1984, Gutzwiller and Anderson 1987). In addition to their importance as potential nest trees, large snags support greater densities of insects, a primary food source for woodpeckers. Cline (1977) reportedhigher concentrations of beetles and carpenter ants (Componotus spp) in larger snags.Mariani and Manuwal (1990) reported an increase in abundance of arthropods toincreased tree diameter and bark furrow depth. Weikel (1997) emphasized the importance of maintaining large snags as foraging substrate for hairy woodpeckers and other cavity- nesters in the Oregon Coast Range. Based on my results, I hypothesizethat nest-site selection is strongly influenced by tree diameter, not only as a nesting substrate, but also because of enhanced foraging opportunities and the increased protection from predation. Currently, land management agencies base much of their snag management on the habitat needs of the pileated woodpecker. The pileated woodpecker is considered to be a forest indicator species, selecting older forest with greater canopy cover for nesting, and utilizing large snags for nesting, roosting, and foraging (Bull 1977, Bull 1980, Madsen 1985, Mellen 1987, Nelson 1989. It is assumed that by providing habitat for the pileated woodpecker, the habitat needs of other woodpeckers and secondary cavity-users will subsequently be met. Snag management guidelines presented in the Northwest Forest Plan (USDA Forest Service and USD1 Bureau of Land Management 1994) require that adequate numbers of snags are retained to support 40% of the potential population of pileated woodpeckers. The Northwest Forest Plan requires that 5 hard snags (decay stage 1-2) and 0.3 soft snags (decay stage 3-4) per hectare greater than 35 cm in diameter are left after harvest (USDA Forest Service and USD1 Bureau of Land Management 1994). There is strong evidence, however, that all woodpeckers select snags significantly larger than 35 cm in diameter for nesting (Bull 1977, Bull 1980, Raphael and White 1984, Zarnowitz and Manuwal 1985, Madsen 1985, Nelson 1989, Lundquist and Mariani 1991, Schreiber and deCalesta 1992, Bunnell et al. 1997, this study), and that large snags additionally provide important resources for foraging (Weikel 1997). Emphasis should be placed on managing forests to provide and maintain larger snags, not only because they are more desirable for nesting and foraging, but because they remain standing 34 longer, and are able to provide habitat over a longer period of time (Cline et a!. 1980, Neitro et al. 1985, Raphael and Morrison 1987).
Tree Condition and Decay Considerable emphasis has been placed on size of snag and its importance to nesting woodpeckers. The state of deterioration of a snag is also critical; unless sufficient rot is present, a tree is unusable by woodpeckers. Mannan et al. (1980), reported that woodpeckers in western Oregon nested most frequently in trees of decay stage 3. Lundquist and Mariani (1991) found that cavity-nesters used harder snags in the Washington Cascades, but mentioned that most nest holes were located in the tops of snags, where decay and rot were present. I found that red-breasted sapsuckers used harder snags in decay classes 1-2 for nesting, hairy woodpeckers used trees in the intermediate stages of decay, and northern flickers used trees that were well-decayed. These results suggest that it is important to maintain snags representing a range of decay stages to provide nesting habitat for woodpeckers. The heartwood of a tree must be sufficiently decayed for woodpeckers to excavate nest cavities (Conner et al. 1976, Bent 1964, Mannan et al. 1980). Location of the original source of decay, along with weather and other environmental factors influence the dynamics driving tree deterioration (Cline et al. 1980, Neitro et al. 1985). Snag deterioration is initiated by invasion of insects, fungi and bacteria, promoted by weather conditions and accentuated by age (Kimmey and Fumiss 1943). This process differs among individual tree species; for Douglas-fir, the decay generally proceeds from top to bottom, and outer to inner bole (Wright and Harvey 1967). In my study, 86% of all nest trees and snags had broken tops. Broken tops serve as an entrance portal for disease and fungal growth (McClelland and Frissell 1975) and can accelerate deterioration of heartwood to the point where it is suitable for excavation.
Tree Species I found woodpeckers nesting primarily in Douglas-fir and white fir trees; however, I could not detect any disproportionate use of tree species by any of the three 35
speciesof woodpecker studied. In Oregon, several studies have indicated thatcavity- nesters use Douglas-fir trees for nesting more thanwould be expected based on availability (Nelson 1989, Mannan Ct al. 1980). In the southern Cascades ofWashington, Lundquist and Mariani (1991) and Madsen (1985) reported disproportionate useby woodpeckers of western white pine. It may be the case in my study that sample size limited the ability of the statistical test to detect any differences between treespecies for nest trees and random trees. Selection of tree species for nestingis probably influenced by local factors, such as presence of absence of fungal disease, insects, and regional climatic conditions.
CONCLUSIONS Tree characteristics influence nest-site selection by woodpeckers in the Southern Oregon Cascades. Tree diameter, tree-top condition, and decay stage all appeared to influence use of trees by woodpeckers for nesting. Large diameter trees were associated with nests of all three species; the hairy woodpecker, northern flicker, and red-breasted sapsucker. Tree-top condition was associated with northern flickers and red-breasted sapsuckers. These species nested more often in trees with broken tops than in trees with intact tops. Nests of hairy woodpeckers were associated with soft snags in decay class 3. Although I found that differences in structure and condition of snags influenced use by woodpeckers, I concluded from my results that variation within thesurrounding microsite did not appear to affect nest-site-selection by woodpeckers. I then modified my scale of resolution, to examine nest-site selection at a larger spatial scale, and determine whether characteristics at the landscape level had any association with nest-site selection by woodpeckers. The results of the landscape level study are presented in the Chapter 3: Habitat Associations of Woodpeckers at the Landscape level in Managed Forests of the Southern Oregon Cascades 36
LiTERATURECiTED
Azet, T. L.A. McCrimmon. 1990. Preliminaiy Plant Association of the Southern Oregon Cascade Mountain Province. USDA Forest Service Publication. Siskiyou National Forest. 113 pp.
Balda, R.P. 1975. The relationships of secondary cavity-nesters to snag densities in western coniferous forests. USDA Forest Service Wildlife Technical Bulletin 1, Albuquerque, N.M. 37 pp.
Beebe, S.B. 1974. Relationships between insectivorous hole-nesting birds and forest management. Yale University School of Forestry and Environmental Studies. 49 pp.
Bent, A.C. 1964. Life Histories of North American woodpeckers. Dover Publications. New York, N.Y. 334 pp.
Brown, E.R. tech. ed. 1985. Management of wildlife and fish habitats in forests of western Oregon and Washington. Part 2 Appendices. USDA Forest Service, General Technical Report PNW-R8-F&WL-192-1985. Portland, OR. 302 pp.
Bull, E.L. 1977. Habitat utilization of the pileated woodpecker, Blue Mountains, Oregon. M. S. Thesis. Oregon State University, Corvallis. 58 pp.
Bull, E.L. 1980. Resource partitioning among woodpeckers in northeastern Oregon Ph.D. Dissertation. University of Idaho, Moscow. 109 pp.
Bull, E.L. 1986a. Ecological value of dead trees to cavity-nesting birds in northeastern Oregon. Oregon Birds: 12: 9 1-99.
Bunnell, F. L., L.L. Kremsater, R.W. Wells. 1997. Likely consequences of forest management on terrestrial, forest-dwelling vertebrates in Oregon. Report M-7 of the Centre for Applied Conservation Biology, University of British Columbia. 129 pp.
Cline, S.P. 1977. The characteristics and dynamics of snags in Douglas-fir forests of the Oregon Coast Range. M.S. thesis. Oregon State University, Corvallis. 107 pp.
Cline, S.P., A.B. Berg, and H.M. Wight. 1980. Snag characteristics and dynamics in Douglas-fir forests, western Oregon. Journal of Wildlife Management 44: 773- 786. 37
Conner, RN., O.K. Miller, C.S. Adkisson. 1976. Woodpecker dependence on trees infected by fungal heart rots. Wilson Bulletin 88: 575-581.
Conner, R.N. 1979. Seasonal changes in woodpecker foraging methods: strategies for winter survival. Pages 95-105 In: J.G. Dickson, R.N. Conner, R.R. Fleet, eds. The role of insectivorous birds in forest ecosystems: Proc. of a symposium. Nacogdoches,Texas. Academic Press, Inc. New York, N.Y.
Evans, K., and R.N. Conner. 1979. Snag management. pages 2 14-225 R.M.DeGraaf and K.E. Evans, eds. Management of north central and northeastern forests for nongame birds. USDA Forest Service General Technical Report NC-5 1, St. Paul, M.N.
Franklin, J.F., C.T. Dryness. 1988. Natural vegetation of Oregon and Washington. Oregon State University Press, Corvallis. 452 pp.
Franzreb, K.E. and R.D. Ohmart. 1978. The effects of timber harvesting on breeding birds in a mixed coniferous forest. Condor 80: 431 -441.
Garrison, G.A. 1949. Uses and modifications for moosehorn crown closure estimation. Journal of Forestry 47: 733-735.
Gutzwiller, K.J. and S.H. Anderson. 1987. Multiscale associations between cavity- nesting birds and features of Wyoming streamside woodlands. Condor 89: 534- 548.
Haapanen, A. 1965. Bird fauna of the finnish forests in relation to forest succession. I. Am. Zool. Fenn. 2:153-196
Harris, L.D. 1984. The fragmented forest. University of Chicago Press, Chicago, I.L. 211 pp.
Hope, S. and W.C. McComb. 1994. Perceptions of implementing and monitoring wildlife tree prescriptions on national forests in western Washington and Oregon. Wildlife Society Bulletin 22: 383-392.
Hosmer, D.W. Jr, and S. Lemeshow. 1989. Applied logistic regression. John Wiley and Sons, New York, N.Y. 307 pp.
Jackman, S.M. 1974. Woodpeckers of the Pacific Northwest; their characteristics and their role in the forest. M.S. Thesis, Oregon State University, Corvallis. 147 pp.
Karlsson, J. and S.G. Nilsson. 1977. The influence of nest-box area on clutch size in some hole-nesting birds. Ibis 119: 207-211. 38
Kimmey, J.W. and R.L. Furniss. 1943. Deterioration of fire-killed Douglas-fir.USDA Technical Bulletin 351. U.S. Governmen Printing Office. Washington, D.C. 61 pp.
Knight, F.B. 1958. The effects of woodpeckers on populations of the Engelmann spruce beetle. Journal of Economic Entomology 51: 603-607.
Lundquist R.W. and J.M. Mariani. 1991. Nesting habitat and abundance of snag- dependent birds in the southern Washington Cascade Range. Pages 22 1-239 jjjL. F. Ruggiero, K.B. Aubry, A.B. Carey, and M.H. Huff, tech. coords., Wildlife and vegetation of umnanaged Douglas-fir forests. USDA Forest Service General Technical Report PNW-GTR-285.
Madsen, S.J. 1985. Habitat use by cavity-nesting birds in the Okanogan National Forest, Washington. M.S. Thesis. University of Washington, Seattle. 113 pp.
Mannan, R.W., E.C. Meslow, H.M. Wight. 1980. Use of snags by birds in Douglas-fir forests, western Oregon. Journal of Wildlife Management 44: 787- 797.
Mariani, J.M. and D.A. Manuwal. 1990. Factors influencing brown creeper (Certhia americana) abundance patterns in the southern Washington Cascade Range. Studies in Avian Biology No. 13:53-57.
McClelland, B.R. and S.S. Frissell. 1975. Identifying forest snags useful for hole- nesting birds. Journal of Forestry 73: 414-417.
Mellen, T.K. 1987. Home range and habitat use by pileated woodpeckers. M.S. Thesis. Oregon State University, Corvallis. 96 pp.
Nelson, S.K. 1989. Habitat use and densities of cavity-nesting birds in the Oregon Coast Ranges. M.S. Thesis. Oregon State University, Corvallis. 157 pp.
Neitro, W.A., V.W. Binkley, S.P. Cline, R.W. Mannan, B.G. Marcot, D. Taylor, F.F. Wagner. 1985. Snags. Pages 129-168 In Brown, E.R. tech. ed. Management of wildlife and fish habitats in forests of western Oregon and Washington. Part 2 - Appendices.USDA Forest Service, Pacific Northwest Region, Portland, OR.
Nilsson, S.G. 1984. The evolution of nest-site selection among hole-nesting birds: the importance of nest predation and competition. Ornis Scand. 15: 167-175
Otvos, I.S. 1965. Studies on avian predators of Dendroctonus brevicomis le Conte (Coleoptera: Scolytidae) with special reference to Picidae. Canadian Entomologist 97: 1184-1199. ______
39
1979.The effects of insectivorous bird activities in forest ecosystems: an evaluation. Pages 341-374 J.G. Dickson et aL, eds. The role of insectivorous birds in forest ecosystems. Academic Press. New York, N.Y.
Otvos, I.S. and R.W. Stark. 1985. Arthropod food of some forest-inhabitingbirds. The Cannadian Entomologist 117:971-990
Ramsey, F.L., Marti McCracken, J.A. Crawford, M.S. Drut and W.J. Ripple.1994. Habitat association studies of the northern spotted owl, sage grouse, and flammulated owl. Pages 189-209 In N. Lange, L. Ryan, L. Billard, D. Brillinger, Conquest and J. Greenhouse (eds.). Case Studies in Biometry. John Wiley and Sons, New York, N.Y. 496 pp.
Raphael, M.G. and M. White. 1984. Use of snags by cavity-nesting birds in the Sierra Nevada. Wildlife Monograph No. 86. 66 pp.
Raphael, M.G. and M.L. Morrison. 1987. Decay and dynamics of snags in the Sierra Nevada, California. Forest Science 33: 774-783.
SAS Institute, Inc. 1990. SAS/STAT Procedures Guide, Version 6, Third Edition, Cary, NC:. '705 pp.
1997.SAS/STAT software: changes and enhancements through release 6.12. Cary,N.C. :ll67pp.
Schreiber, B. and D.S. deCalesta. 1992. The relationship between cavity-nesting birds and snags on clearcuts in wetern Oregon. Forest Ecology and Management 50: 299-316.
Thomas, J.W., G.L. Crouch, R.S. Bumstead, and L.D. Bryant. 1975. Silvicultural options and habitat values in coniferous forests. Proc. symposium management of forest and range habitats for nongame birds. USDA Forest Service General Technical Report W0-1.
USDA Forest Service. 1990. Land and Resource Management Plan -UmpquaNational Forest. Final Environment Impact Statement: Appendices -VolumeII. USDA Forest Service, Pacifice Northwest Region. 315 pp.
USDA Forest Service and USD1 Bureau of Land Management. 1994. Record of Decision for amendments to Forest Service and Bureau of Land Management Planning Documents within the range of the northern spotted owl. Standards and Guidelines for management of habitat for late successional and old-growth related species within the range of the northern spotted owl. USDA Forest Service and USD1 Bureau of Land Management. 74 pp. 40
Weikel, J.1997. Habitat use by cavity-nesting birds in young thinned andunthinned Douglas-fir forests of western Oregon. M.S. Thesis, Oregon StateUniversity, Corvallis. 102 pp.
Winkler, H., D.A. Christie, and D. Nurney. 1995. Woodpeckers: Anidentification guide to the woodpeckers of the world. Houghton Mifflin Co. NewYork, N.Y. 406 pp.
Wright, K.H. and G.M. Harvey. 1967. The deterioration of beetle-killedDouglas-fir in Oregon and Washington. USDA Forest Service Research Paper. Pacific Northwest Region, Portland, OR. 22 pp.
Zarnowitz, J.E. and D.A. Manuwal. 1985. The effects of forest management on cavity- nesting birds in northwestern Washington. Journal of Wildlife Management 49: 255-263. 41
Habitat Associations of Woodpeckers at Multiple Scales in Managed Forests of the Southern Oregon Cascades
CHAPTER 3 Habitat Associations of Woodpeckers at the Landscape Level in Managed Forests of the Southern Oregon Cascades
INTRODUCTION "The term 'patch' typically implies a discrete and internally homogeneous entity, yet such patches are rarely observed in nature. Instead hierarchical mosaics of patches within patches occur over a broad range of scales. Progress in understanding how organisms respond to patchiness requires more realistic conceptualizations that explicitly recognize these hierarchical layerings of patch structure" Wiens 1989
Wiens (1989) captured the crux of the dilemmas we face today in landscape ecology and ecosystem management. There have been many recent studies investigating processes and systems at large scales, with varying degrees of success. Researchers are discovering that the study of systems at larger scales is a challenge, requiring that the methodology and the very maimer in which we formulate questions be re-examined to address problems at larger scales. Likewise, land managers are discovering that management of ecosystems is a complex undertaking, and are floundering with few previous examples from the past to use as guides. Studies have shown that woodpeckers select sites for nesting based on characteristics of individual trees that allow excavation of nest cavities (Madsen 1985, Bull 1986c, Swallow and Gutierrez 1986, Schreiber 1987, Nelson 1989, Milne and Hejl 1989, Sedgwick and Knopf 1990, Chapter 2 this study). Selection for nest sites, however, may begin at a broader scale, with woodpeckers selecting areas containing ample opportunities for foraging and shelter. It has been proposed that habitat selection is hierarchical in nature, with organisms choosing sites based on characteristics at multiple levels (Block and Brennan 1993, Johnson 1980). Few studies that I am aware of have 42 examined habitat associations of woodpeckers beyond the nest tree orstand level. One study examined associations between cavity-nesters and habitatfeatures in Wyoming bottomlands, (Gutzwiller and Anderson 1987) and found that cavity-nesters were responding to characteristics at different spatial scales. Because of the recent shift in forest management to lookbeyond the stand level and plan for desired conditions over broader spatial andtemporal scales to meet diverse objectives, there is a need for information on habitat needs for nestingwoodpeckers over landscapes. The difficulty, however, in defining habitat suitablity increasesmarkedly when considering that selection is not based on one or two habitatcharacteristics, but rather the combined effects produced by characteristics across different spatialscales that elicit the observed species' response. In Chapter two, I presented the resultsof one part of my study, examining the characteristics of nest trees of woodpeckers, and characteristics of microsites immediately around the nest tree. In the second part of my study, I expanded the scope of resolution to encompass the landscape context in which the nests were located. I considered habitat features at the landscape scale that may be influencing habitat selection by woodpeckers to determine whether associations existed between habitat features and nest densities of woodpeckers. My overall goals were two-fold; to determine whether habitat characteristics at the landscape level were associated with woodpecker nest densities, and furthermore to integrate the results of the study at the larger scale of resolution with the results of the nest-level study to provide a more complete picture of habitat needs for nesting woodpeckers.
OBJECTIVES Specifically, my objectives were to:
• Quantify selected habitat characteristics at the landscape level of resolution • Determine whether nest densities of northern flickers, hairy woodpeckers and red -breasted sapsuckers were associated with measured habitat characteristics 43
STUDY AREA
Thestudy was conducted in the Central Oregon Cascade Range, on the Diamond Lake Ranger District of the Umpqua National Forest. Study sites are located approximately 70 miles east of Roseburg, Oregon, in Douglas County between 47° 54' and 47° 56'latitude(Fig. 1). Study sites were within the Umpqua National Forest on sites varying in elevation from 1066 m-1463 m in the Umpqua High Cascade sub- province. The climate is characterized by moderate temperatures, wet, mild winters and relatively cool, dry summers (Franklin and Dymess 1988). The region receives an average of 152 cm of precipitation annually; of which 75-85% occurs between October and April, primarily as rain, with snow at high elevations (Azet and McCrimmon 1990). The study area is located within the western hemlock (Tsuga heterophylla) zone, the most extensive vegetation zone found in the Pacific Northwest (Franklin and Dyrness 1988). In this zone western hemlock is considered the climax species, but this state is rarely reached due to forest management activities (Harris 1984). The overstory is typically dominated by Douglas-fir (Pseudotsuga menziesii), though white fir (Abies concolor), western hemlock, mountain hemlock (Tsuga mertensiana), Pacific silver fir (Abies amabilis), Pacific yew (Taxus and Shasta red fir (Abies commonly occur. Other species present include: incense-cedar (Calocedrus decurrens), western redcedar (Thujaplicata), ponderosa pine (Pinusponderosa) and sugar pine (Pinus lambertiana). Lodgepole pine (Pinus contorta) is found in sites located on the eastern end of the study area. Understory vegetation varies depending on microsite conditions, but generally is dominated by golden chinquapin (Castanopsis chrysophylla)4, Sitka alder (Alnus sinuata), Pacific rhododendron (Rhododendron macrophyllum) vine maple (Acer circinatum). Buckrush (Ceanothus integerrimus), and snowbrush (Ceanothus velutinus) are found in young, regenerating stands. Ground cover is variable and includes sword fern (Polystichum bracken fern (Pteridium aquilinum), vanillaleaf (Achlys triphylla), 44
Figure 6. Location of study sites for woodpecker habitat association study on the Diamond Lake Ranger District, Umpqua National Forest Southern Oregon Cascades. STUDY SITE LOCATION idy Sites
0 kaomete.i 'Co's 30 46
Cascade Oregon-grape (Berberis nervosa), huckleberries (Vacciniumspp.), salal (Gaultheria shallon), and prince's pine (Chimophila umbellafa) The study sites were dominated by mixed-conifer forest composed of atleast 70%, but not more than 85% Douglas-fir trees. The remaining trees were amixture of western hemlock, mountain hemlock, true fir, and pine trees.Three sites had a small component of lodgepole pine, not exceeding more than 5% of thetotal area. The sites were at approximately the same elevation. Volcanic activity has influenced landscape formation found in the study area, depositing a layer of pumice across the landscape. Prior to European settlement in the middle 19th century, the natural disturbance regime of the area was dominated by wild fire events.Since its designation as National Forest land in 1905, the resource planning strategy has emphasized efficient timber production, using even-aged management methods. Due to management practices over the past 40 years, the condition of the landscape is highly variable, consisting of a mosaic of habitat patches of varying age and structure. Five species of woodpeckers are known to breed within the study area; the black- backed woodpecker, hairy woodpecker, northern flicker, pileated woodpecker, and red- breasted sapsucker. Other species of woodpecker, including red-naped sapsuckers (Sphyrapicus nuchalis) and white-headed woodpeckers (Picoides albolarvatus) are residents on the east side of the Cascade Range, but occur only occasionally in the study area.
METHODS Study Site Selection Eleven sites within the study area were selected for the study. Site selection criteria were based on the proportion of the area comprised of mature forest, defined as closed-canopy forest with> 40% canopy cover and trees averaging53 cm in diameter. The proportion of total forested area within study sites varied from 10% to 94%. The remaining sites had intermediate amounts of forest, with the intent to select sites that 47 represented a continuum of amount of forest cover. Six sites were surveyed in 1995,and seven in 1996; two of the sites sampled in 1995 wereresampled in 1996. Study sites chosen for woodpecker sampling were 200-ha in size. Home range estimates of resident woodpeckers range from 12 ha for the red-breastedsapsucker to 200 ha for the pileated woodpecker (Bull 1977, Mannan et al. 1980, Brown et al. 1985, Winider et a!. 1995). I chose 200-ha study sites because they were sufficiently large to include multiple woodpecker home ranges, yet small enough to meet logistical constraints.
Bird Sampling and Nest Searches A 1600- x 1200-rn grid was established at each study site. Five transect lines were located parallel to one another within each grid, with stationsestablished every 200 meters along transect lines. Station points were used as references for recording locations of nests. I located woodpecker nests by walking along transect lines. Approximately every 200 meters I stopped to listen for woodpeckers. I attempted to visually locate each woodpecker heard. Once located, I observed bird behavior, looking for evidence of a nest. Indicators of nest presence included wood chips at the base of a tree, drumming between pair members, territorial calling, and begging calls of nestlings. I followed woodpeckers for as long as possible or until a nest was discovered. Woodpeckers were not followed if they left the study site. Woodpeckers were easier to locate in open- canopy than in closed-canopy forest conditions, where they are more visible. I used the relative amount of closed-canopy forest present in a transect to standardize search time. To eliminate potential sampling biases, I allotted approximately twice the amount of time to conduct searches in closed-canopy than I did in open-canopy forest. All sites were visited a minimum of five times between 10 April and 14 July in 1995 and 1996, with 6 sites sampled in 1995 and 7 in 1996. Surveys were conducted between 0500 and 1500. Potential nests were marked and later revisited to confirm nesting status. Nests were considered to be confirmed when members of a pair were 48 observed excavating a cavity, exchanging incubation duties, or feedingfledglings at a cavity entrance, or when fledglings were seen or heard in the cavity.
Creating a GIS Layer Universal transverse mercator (UTM) coordinates for all nest trees wereobtained using a hand held GPS data logging system (TrimbleNavigation©). I generated UTM coordinates of each nest tree to the nearest 10 meters. Nest locations weredownloaded to a computer and correction factors fromthe U.S. Department of Defense-Portland base were used to correct UTM coordinates.Corrected coordinates were brought into AutoCAD (AutoCAD 1990). Maps of each study site were generated,integrating nest locations, forest road systems, habitat patches, and site boundaries. Thecompleted coverages were imported into Arc/Info, a GeographicalInformation System (GIS) software package that facilitates interactive analysis of spatial data (INFO 1994). All patches within coverages were attributed with codes representing vegetation composition and seral stage. I used definitions for seven different patch types to attribute individual coverages (Table 7).
Statistical Analysis FRAGSTATS-Background I used FRAGSTATS (Marks and McGarigal 1995) to describe the structural characteristics of the landscape. FRAGSTATS uses spatial coverages to calculate 50 metrices representing landscape and patch-type characteristics. I analyzed images complete with landscape boundary and border. The boundary defined the outside perimeter of the area and the border consisted of a 200-m buffer strip surrounding the site. A border provides information about patches adjacent to the landscape edge, and allows computation of meaningful edge metrices. FRAGSTATS generates indices characterizing a specified area at the patch, class, and landscape level. I focused my analysis at the class level and restricted the analysis to the pre-defined target habitat patch type, mature, mixed-conifer forest. Class-level indices described the characteristics of each patch type, or class within each study site. 49
Table 7. Patch type classification system used for FRAGSTATS analysis to classify vegetation in 200-ha study sites. Mixed-conifer defined as forest composed of at least 70%. but not more than 85% Douglas-fir trees; with the remaining trees a mixture of western/mountain hemlock, true fir, and pine trees. Lodgepole is defined as forest composed of >80% lodgepole pine.
PATCH TYPE DEFINITION
Mature mixed-conifer >90%mixed-conifer; average tree diameter> 53 cm, canopy closure >40%.
Mature lodgepole >90%lodgepole; average tree diameter >38cm, stand density equal to or greater than 250 trees/ha.
Mixed-conifer pole >90%mixed-conifer; tree diameter between 10-28 cm, stand density equal to or greater than 1000 trees/ha.
Mixed-conifer shelterwood >90%mixed-conifer; 30-45 trees/ha, snags present.
Mixed-conifer regeneration >90%mixed-conifer; average tree diameter <10cm, with snags open canopy with <40% canopy closure. <60% shrub cover composed of Ceanothus spp, snags present.
Mixed-conifer regeneration, >90%mixed-conifer; average tree diameter <10cm, shrub understory, with snags open canopy with <40% canopy closure, >60%shrub cover composed of Ceanothus spp., snags present.
Mixed-conifer regeneration, >90%mixed-conifer; open canopy with <40% canopy shrub understory, no snags closure,> 60% shrub cover composed of Ceanothus spp., snags absent. 50
Class indices are useful when examining habitat fragmentation becausethey allow the researcher to separate a landscape into distinct patches associated with uniquehabitat characteristics (MarksandMcGarigal 1995).
Class Level Analysis A rank-correlation test performed on the data set indicated that several of the descriptive variables were highly correlated (SAS Institute, Inc. 1990). In the case of highly intercorrelated variables (p> 0.80), one of the variables was dropped from analysis. FRAGSTATS generates indices that are redundant, expressing the same information in a slightly different format. In cases where two indices measured the same characteristic, but one was density-dependent, I included the density-dependent index in the analysis. This allowed for comparisons between study sites. An attempt was made to include variables that were relevant both biologically and within the context of the questions asked. In some cases this decision was subjective, based upon interpretation of the variables given my experience in the field. Through this process, several variables were eliminated from the dataset, either because they were redundant, orbecause they had no biological relevancy to the questions of interest. The resultant set of 6 variables was modeled using multiple regression, to determine whether measured characteristics at the landscape level were associated with nest densities of red-breasted sapsuckers, northern flickers, or hairy woodpeckers (Table 8). Univariate analysis revealed that several of the variables were not normally distributed. Where necessary, log transformations were used to normalize distributions. To examine the association between landscape characteristics and nest abundance for each woodpecker species, I used multiple regression (SAS Institute, Inc. SAS/STAT 1997). I used a forward stepwise procedure, setting the significance level at p <0.05for entry into the model. I ran the analysis separately for each woodpecker species to determine if differences could be detected. 51
Table 8. Definitions of class level habitat variables generated by FRAGSTATS to describe landscape characteristics in 200-ha sites; Mixed-conifer defined as forest composed of at least 70%, but not more than 85%Douglas-firtrees; with the remaining trees a mixture of western/mountain hemlock, true fir, and pine trees. Edge contrastis based on a weighting scheme to represent the degree of contrast between patch types, based on McGarigal and McComb (1995) and Brooks (1997) (See Appendix C.).
FRAGSTATS DEFINITION VARIABLE Percent mature forestThe proportion of area representing mature, mixed-conifer forest with average tree diameter> 53cm.
Patch density Number of patches of mature, mixed-conifer forest per unit area.
Mean patch size Average size in hectares of mature, mixed-conifer forest.
Edge density The amount of edge habitat for mature, mixed-conifer forest per unit area.
Area-weighted mean Index that estimates the degree of contrast between a patch of edge-contrast index mature, mixed-conifer forest and its adjacent environment, weighted by area of the patch.
Core area density The density of core areas of mature, mixed-conifer forest within landscape. Core areas are defined as patches large enough to contain interior forest, with at least 100 meters from the core center to an edge. 52
Edge Analysis To assess relationships between nest-site location and habitat edges, I generated 159 random coordinates and plotted these random locations on each study site coverage. I measured the distance from the random points and from each nest to the nearest habitat edge, points that fell directly on an edge were assigned a value of zero. I did not include the unfragmented site in this analysis as it did not posess edge habitat, and therefore was not relevant to the question of interest. For each species of woodpecker, I performed a Chi-square goodness-of-fit test to determine whether woodpeckers disproportionately used trees along habitat edges for nesting. I also tested for differences in distances of nests and random locations to the nearest habitat edge using a t-test. This was done for each species of woodpecker, for all nests not located directly on an edge. Data were log-transformed for this analysis. The statistical power of the t-test was calculated based on Cohen (1988).
RESULTS FRAGSTATS Analysis of Landscape Indices Results generated by FRAGSTATS analysis revealed several relationships between nest density and habitat features. There was a significant negative trend between the amount of mature forest and the density of woodpecker nests, suggesting that woodpeckers select more open habitats for nesting. Regression analysis on FRAGSTATS indices revealed slight differences among species. Nest densities for red-breasted sapsuckers were negatively associated with mean patch size of mature forest habitat (P <0.036)(Fig. 7). Nests of northern flickers were negatively associated with the proportion of mature forest in a landscape (P <0.033) (Fig. 8). This association was also observed with the hairy woodpecker (P <
0.009)(Fig. 9). Regression did not detect any other variables among the habitat characteristics calculated that were significant in explaining the variability in nest densities between sites. Figure 7. Association red-breasted sapsucker nests over landscapes between the log of mean patch size and density of 10 Y=9.75p EdgeAnalysis Placement of nests along habitat edges was significantly different between nests of all woodpecker species and random points (Table 9). This association was highly 2 significant for red-breasted sapsuckers (X= 34.12,df =1,P <5x1 Out of a 61 red-breasted sapsucker nests sampled, 23 occurred on habitat edges. Northern flickers 2 (X2= 8.23,df= 1, P <0.0041) and hairy woodpeckers(X 4.14,df= 1, P <0.042) also nested in trees located along an edge more frequently than would be expected. Of the 31 hairy woodpecker nests included in analysis, 4 were located along forest edges, defined as the boundary between a patch of mature forest and regenerating stand, and one was at the edge between a shelterwood and a regenerating stand. Table 9. Location of nests of red-breasted sapsuckers, hairy woodpeckers and northern flickers with respect to edge within landscapes. Edge was defined as representing a distinct boundary between a stand of mature forest and a regenerating stand, or as the boundary between a shelterwood and a regenerating stand. OBSERVED On Edge Not on Edge Red-breasted Sapsucker 23 38 Northern flicker 12 51 Hairy woodpecker 5 26 Random 10 149 57 Comparisons using t-tests T-tests performed on nests not located directly on an edge did not indicate a significant difference between nest tree placement and random locations in proximity to edge (Table 12). The statistical power of the t-test was high (>0.99 to detect a difference between nest sites and random sites). Therefore I am confident that the conclusion can be made that once nests located directly on habitat edges are accounted for, there is no difference in distance to an edge between nest trees and random trees. Table 10. Comparison of distance from an edge between woodpecker nests and random locations, after all points located directly on an edge were removed. Distances were log- transformed for analysis, but non-transformed values for means and standard deviations are presented here. n X StdDev P<{t) (meters) Random 149 63.4 2.43 Hairy woodpecker 24 75.9 2.75 0.3655 Northern flicker 50 69.4 2.54 0.5583 Red-breasted sapsucker 38 54.6 2.54 0.4323 58 DISCUSSION Nest densities of woodpeckers were associated with habitat variation within landscapes. Woodpecker nest-site selection appears to be influenced by overall landscape pattern and location with respect to edges. Northern flickers and hairy woodpeckers exhibited negative associations with the proportion of mature, mixed-conifer forest. This is consistent with results of similar studies of woodpeckers throughout the Pacific Northwest that have found greater abundances of woodpecker nests in openings and open-canopy forests(Hagar 1960, Madsen 1985, Hagar et al. 1996, Weikel 1997). Hairy woodpeckers and northern flickers will nest in young regenerating forests, if adequate numbers of snags are present (Hagar 1960, Bull 1980, Mannan et al. 1980, Raphael and White 1984, Schreiber and deCalesta 1992). Madsen (1985) reported that woodpecker nest-sites had significantly less canopy cover than randomly located sites in the southern cascades of Washington. Additionally, some cavity-nesters in the Oregon Coast Range respond positively to thinning (Hagar et al. 1996, Weikel 1997). Hagar et al. (1996) attributed this behavioral response to the enhancement of foraging habitat created by the residual slash and trees left behind after harvest activities, but Weikel (1997) questioned this conclusion. Species-specific patterns I detected statistically significant associations with one habitat component and nest-site location for northern flickers at the landscape level. Densities of northern flicker nests decreased as the amount of mixed-conifer forest increased within a landscape. This species has been described as a habitat generalist, and is often found nesting in forest openings (Hagar 1960, Bent 1964, Raphael and White 1984, Nelson 1989, Lundquist and Mariani 1991). Raphael and White (1984) reported that northern flickers nest in a wide variety of habitats in northern California, primarily in burned areas and lodgepole/meadow conditions. Hagar (1960) found that densities of northern flickers increased in response to logging, provided that ample snags for nesting and foraging were left following harvest. The diet of the northern flicker consists primarily of ants (Hymenoptera), but is supplemented by termites (Isoptera), caterpillars (Lepidoptera), 59 beetles and beetle larvae (Coleoptera), crickets (Orthoptera), aphids (Homoptera) spiders (Araneida) and berries (Beal 1911, Winkler et al. 1995). Northern flickers were observed foraging in regenerating stands, often with little ground cover, where preferred forage was available. The foraging habits of the northern flicker may partially explain their affinity for open-canopy forest. Northern flickers are weak excavators, and use soft, well-decayed snags for excavating nest cavities (Raphael and White 1984, Lundquist and Mariani 1991, Winkler et al. 1995, this study). Well-decayed, soft snags suitable for excavation by northern flickers are often found in regenerating stands and forest openings, where exposure to the wind and weather accelerates the rate of decay (Cline et al. 1980). Density of red-breasted sapsucker nests was negatively associated with the size of mature forest patches throughout a landscape. Though this statistic is related to the overall amount of mature forest, it relates more closely to distribution of that patch type across the landscape. Spatial structure of habitat patches over a landscape may have an effect on species' distribution, influencing the amount of edge, amount of interior forest, and the availability of required resources. Red-breasted sapsuckers have been described as forest specialists (Raphael and White 1984), whereas other studies have associated them with older forests (Nelson 1989, Lundquist and Mariani 1991). I often observed red-breasted sapsuckers foraging in stands of mature forest, where they foraged mainly on hemlock and yew trees. Sapsuckers that had nested in a regenerating stand were seen making frequent forays to adjacent forested stands when feeding nestlings. Relatively little is known about the foraging habits of the red-breasted sapsucker. They feed primarily on sap, as do other members of Sphyrapicus, but it is thought that ants may be an important food source for nestlings (Winkler et al. 1995). Dietary needs may be specialized enough to restrict sapsuckers to certain habitat types for foraging. The negative association with mature forest patch size may be due in part to the sapsucker's need for diverse patch types within their home range to meet habitat requirements. It may be advantageous for sapsuckers to nest in trees that offer access to several different patch types. 60 Density of hairy woodpecker nests were positively associated with open canopy forest with snags present, and negatively associated with the amount of mature forestin landscapes. This observation is consistent with those of other studies. Hagar et at. (1996) and Weikel (1997) both reported positive responses of hairy woodpeckers to thinning in forests of the Oregon Coast Range. Weikel also detected more hairy woodpeckers in heavily thinned stands than in moderately thinned stands. She hypothesized that this was due to increased abundance of prey in thinned stands. The diet of the hairy woodpecker consists mainly of species from the order Coleoptera; supplemented by those in Hymenoptera, Orthoptera, Araneida, and Diptera (Otvos and Stark 1965, Winkler et at. 1995). However, hairy woodpeckers are quite versatile and will supplement their diet with seeds and fruit (Bent 1964). This versatility may allow them to take advantage of foraging opportunities available in different stand types. Weikel (1997) reported that hairy woodpeckers foraged heavily on large, well-decayed snags and logs. This suggests that the presence of large dead wood is animportant habitat attribute for hairy woodpeckers, and may be a stronger driving force in habitat selection than stand type or seral stage. Edge Effect Woodpeckers chose trees along edge habitatsmore frequently than would be expected due to chance alone This affinity for edge habitat by woodpeckers has been described by others (McGarigal and McComb 1995). For red-breasted sapsuckers, the association with edge habitat was very strong, roughly one-third of all red-breasted sapsucker nests were located on a habitat edge. Aldo Leopold, one of the first wildlife biologists to recognize the importance of edge habitat for wildlife, stated that edge habitat is desirable because it "offers simultaneous access to more than one environmental type, or the greater richness of border vegetation, or both" (Leopold 1933). Hunter (1990) defined edges as transition zones between two habitat patches with a high degree of contrast in vegetative and/or structural characteristics. Unique conditions found in edges create suitable habitat for a variety of plants and animals (Chen et at. 1993). Helle and Muona (1985) reported that 61 several types of invertebrates were relatively more abundant near forest-clearcut edges. I hypothesize that edge habitats support higher densities of insects, providing enhanced foraging opportunities for woodpeckers. Conditions found along edges also may promote fungal growth on trees that accelerate tree decay and encourage the invasion of woodpecker prey species. It probably is not due to one single factor but the combination and integration of several factors that make edges preferred locations for nesting by these three woodpecker species. While it is true that edge habitats support a greater diversity of wildlife species, they can also be inhospitable to some species. Edges may be detrimental to species who are considered interior forest-dwellers. Studies, primarily conducted in the eastern United States, have shown increased rates of nest predation on songbirds nesting near edges than for those nesting in forest interior (Hunter 1990). When developing long-term management plans, careful consideration should be given to overall landscape diversity with respect to edges. The goal of enhancing wildlife diversity by the creation of edges through management activities must be balanced with the need to maintain the integrity of the forest matrix for forest-dwelling species. CONCLUSIONS Nesting habitat for woodpeckers can be provided within managed landscapes as long as adequate numbers of suitable nest snags are made available. In some cases, the enhancement of foraging opportunities created by forest openings may allow woodpeckers to nest at higher densities in areas with little overstory canopy, once again provided that appropriate numbers of suitable snags are retained for nesting. Woodpeckers exhibited a positive association to increased landscape diversity. Results of landscape analysis revealed that woodpeckers nested at higher densities in areas characterized by greater habitat complexity. The results of this study suggest that habitat complexity, specifically with regard to patch heterogeneity and the presence of edge habitat, may be important factors in nest-site selection for woodpeckers. 62 L1TERAT(JRECITED Azet,1. L.A. McCrimmon. 1990. Preliminary Plant Association of the Southern Oregon Cascade Mountain Province. USDA Forest Service Publication. Siskiyou National Forest. 113 pp. Autodesk, Inc. 1990. AutoCAD Release 11: Reference manual, publication AC1 1RM. 640 pp. Beal, F.E.L. 1911. Food of the woodpeckers of the United States. Biological Survey Bulletinv. 37. Bent, A.C. 1964. Life Histories of North American woodpeckers. Dover Publications. New York, N.Y. 334 pp. Block and Brennan. 1993. The habitat concept in ornithology: Theory and applications. Pages 35-83 D.M. Power, ed. Current ornithology, Volume 11. Plenum Press. New York, N.Y. Brooks, J.P. 1997. Bird-habitat relationships at multiple spatial resolutions in the Oregon Coast Range. M.S. Thesis, Oregon State University, Corvaffis. 46 pp. Brown, E.R. tech. ed. 1985. Management of wildlife and fish habitats in forests of western Oregon and Washington. Part 2 Appendices. USDA Forest Service, General Technical Report PNW-R8-F&WL-192-1985. Portland, OR. 302 pp. Bull, E.L. 1977. Habitat utilization of the pileated woodpecker, Blue Mountains, Oregon. M. S. Thesis. Oregon State University, Corvallis. 58 pp. Bull, E.L. 1980. Resource partitioning among woodpeckers in northeastern Oregon Ph.D. Thesis. University of Idaho, Moscow. 109 pp. Bull, E.L., S. R. Peterson, J.W. Thomas. 1986c. Resource partitioning among woodpeckers in northeastern Oregon. USDA Forest Service Research Note, PNW-444. 19 pp. Chen, J., J.F. Franklin, and T.A. Spies. 1993. Contrasting microcimates among clearcut, edge, and interior of old-growth Douglas-fir forest. Agricultural and Forest Meteorology 63: 219-237. 63 Cline, S.P., A.B. Berg, and H.M. Wight. 1980. Snag characteristics anddynamics in Douglas-fir forests, western Oregon. Journal of Wildlife Management. 44: 773- 786. Cohen, J. 1988. Statistical power analysis for the behavioral sciences.Second edition. Lawrence Eribaum Associations, Inc. Hillsdale, N.J. Franidin, J.F., C.T. Dryness. 1988. Natural vegetation of Oregon and Washington. Oregon State University Press, Corvallis. 452 pp. Gutzwiller, K.J., and S.H. Anderson. 1987. Multiscale associations between cavity- nesting birds and features of Wyoming streamside woodlands. Condor 89: 534- 548. Hagar, D. C. 1960. The interrelationships of logging, birds and timber regeneration in the Douglas-fir region of northwest California. Ecology 41: 116-125. Hagar, J.C., W.C. McComb, and W.H. Emmingham. 1996. Bird communities in commercially thinned and unthinned Douglas-fir stands of western Oregon. Wildlife Society Bulletin 24: 353-366. Harris, L.D. 1984. The fragmented forest. University of Chicago Press, Chicago, I.L. 2llpp. Helle P., and J. Muona. 1985. Invertebrate numbers in edges between clear-fellings and mature forests in northern Finland. Silva Fenn. 19: 28 1-294. Found In: Hunter, Malcom. 1990. Wildlife, forests and forestry: Principles of managing forests for biological diversity. Prentice-Hall, Inc. Englewood Cliffs, N.J. 370 pp. Hunter, M.L. Jr. 1990. Wildlife, forests and forestry: Principles of managing forests for biological diversity. Prentice-Hall, Englewood Cliffs, N.J. 370 pp. INFO. Arc/Info Version 7. 1994. Environmental Systems Research Institute, Inc. Redlands, C.A. 337 pp. Johnson, D.H. 1980. The comparison of usage and availability measurements for evaluating resource preference. Ecology 61: 65-70. Leopold, A. 1933. Game Management. Charles Scribner Sons. New York, N.Y. 481 pp. Lundquist R.W. and J.M. Mariani. 1991. Nesting habitat and abundance of snag- dependent birds in the southern Washington Cascade Range. Pages 22 1-239 L. F. Ruggiero, K.B. Aubry, A.B. Carey, and M.H. Huff, technical coordinators., Wildlife and vegetation of unmanaged Douglas-fir forests. USDA Forest Service General Technical Report PNW-GTR-285. ______ 64 Madsen, S.J. 1985. Habitat use by cavity-nesting birds in theOkanoganNational Forest, Washington. M.S. Thesis. University of Washington, Seattle. 113 pp. Mannan, R.W., E.C. Meslow, H.M. Wight. 1980. Use of snags bybirds in Douglas-fir forests, western Oregon. Journal of Wildlife Management. 44: 787- 797. Marks, B. and K. McGarigal. 1995. Fragstats: Spatial pattern analysis program for quantifying landscape structure. USDA Forest Service General Technical Report, PNW-GTR-351. 122 pp. McGarigal, K. and W.C. McComb. 1995. Relations between landscape structure and breeding birds in the Oregon Coast Range. Ecological Monographs. No. 65: 235- 260. Mime, K.A., and S.J. Heji. 1989. Nest-site characteristics of white-headed woodpeckers. 1989. Journal of Wildlife Management 53: 50-54. Nelson, S.K. 1989. Habitat use and densities of cavity-nesting birds in the Oregon Coast Ranges. M.S. Thesis. Oregon State University, Corvallis. 157 pp. Otvos, I.S. and R.W. Stark. 1985. Arthropod food of some forest-inhabiting birds. The Cannadian Entomologist 117:971-990 Raphael, M.G. and M. White. 1984. Use of snags by cavity-nesting birds in the Sierra Nevada. Wildlife Monograph No. 86. 66 pp. 1990.SAS/S TAT Procedures Guide, Version 6, Third Edition, Cary, N.C: 705 pp. 1997.SAS/STAT software: changes and enhancements through release 6.12. Cary,N.C. :1167 pp. Schreiber, B. 1987. Bird use of snags on clearcuts in central coastal Oregon. M.S. Thesis, Oregon State University, Corvallis. 63 pp. Schreiber, B. and D.S. deCalesta. 1992. The relationship between cavity-nesting birds and snags on clearcuts in wetern Oregon. Forest Ecology and Management 50: 299-3 16. Sedgwick, J.A., and F.L. Knopf. 1990. Habitat relationships and nest site characteristics of cavity-nesting birds in cottonwood floodplains. Journal of Wildlife Management 54: 112-124. 65 Swallow, S.K. and R.J. Gutierrez. 1986. Primary cavity-site selection by birds.Journal of Wildlife Management 50: 576-583. USDA Forest Service and USD1 Bureau of Land Management. 1994.Record of Decision for amendments to Forest Service and Bureau of Land Management Planning Documents within the range of the northern spotted owl. Standardsand Guidelines for management of habitat for late successional and old-growth related species within the range of the northern spotted owl. USDA Forest Service and USD1 Bureau of Land Management. 74 pp. Weikel, J.1997. Habitat use by cavity-nesting birds in young thinned and unthinned Douglas-fir forests of western Oregon. M.S. Thesis, Oregon State University, Corvallis. 102 pp. Wiens, J.A. 1989. Spatial Scaling in Ecology. Functional Ecology 3: 385-397. Winkler. H., D.A. Christie, and D. Nurney. 1995. Woodpeckers: An identification guide to the woodpeckers of the world. Houghton Mifflin Co. New York, N.Y. 406 pp. 66 Habitat Associations of Woodpeckers at Multiple Scales in Managed Forests of the Southern Oregon Cascades CHAPTER FOUR Conclusions and Implications for Management for Woodpeckers in the Oregon Cascades Associations between woodpeckers and habitat characteristics occurred at more than one spatial scale, and observations made at one scale were not always detected at another. Two important conclusions can be drawn from these results. First, it is imperative that spatial scale is considered in wildlife-habitat association studies. If scale is disregarded, associations may go undetected, and in some cases, results may be incorrectly interpreted. Second, management plans designed to provide and maintain suitable habitat for woodpeckers must consider the larger landscape context in which they are found. It is important to retain adequate numbers of suitable nesting snags but consideration also should be given to snag placement over landscapes. LARGE DIAMETER SNAGS Current guidelines for snag management outlined in the Northwest Forest Plan (NFP) set minimal requirements for retention of snag habitat. These guidelines are based on nesting habitat needs of the pileated woodpecker, and require that a sufficient number of snags be retained to support 40% of the potential populations level for pileated woodpeckers. The plan recommends that snag management occur within areas of green tree retention, and that green trees can be left as 'snag recruitment' trees if snag density is not adequate (USDA-Forest Service and USD1-Bureau of Land Management 1994). The Umpqua National Forest Land and Resource Management Plan (USDA Forest Service 1990) requires that 5trees/hagreater than 30 cm in diameter and 0.8 trees/ha greater than 50 cm in diameter be retained, and that 15 green trees/ha be left to provide future snags. 67 When snag densities are insufficient, guidelines call for the recruitment of snags by artificial means. Guidelines require that snags be maintained that are 35 cm or greater in dbh. I found woodpeckers nesting in trees substantially larger than 35 cm; average diameter for woodpecker nest trees varied from 78 cm (northern flicker) to 81 cm (red- breasted sapsucker). My results, supported by numerous other studies throughout the Pacific Northwest, indicate that woodpeckers use larger diameter trees for nesting (Mannan et al. 1980, Raphael and White 1984, Madsen 1985, Nelson 1989, Lundquist and Mariani 1991, this study). There is evidence that large diameter snags provide foraging habitat in addition to nesting habitat for woodpeckers. Weikel (1997) reported that large diameter well-decayed snags were an important foraging habitat component for hairy woodpeckers and other cavity-nesters in the Oregon Coast Range. Providing large snags will help meet many wildlife objectives, and will aid in the overall maintenance of healthy forest ecosystems. For example, flying squirrels and long-legged myotis (Myotis volans) use large diameter snags for roosting (Feen 1997, Ormsbee 1996). Over time, snags decay and are recruited to the stand as down woody material, functioning as an important foraging substrate for many species of cavity- nesters (Weikel 1997) and creating travel corridors for small mammals such as the red- backed vole (Thompson 1996). Based on my results, I recommend that snags 80 cm or greater be retained following harvest to provide nesting habitat for woodpeckers. This number represents the average diameter of nest tree used by all three species of woodpeckers. Providing snags equal to or greater than the average size reported is preferable to leaving snags that meet only the minimal requirements, and will provide more suitable habitat over the long run. EDGE Edgehad a dramatic effect on density and distribution of woodpecker nests in my study area. Although an association with edge has been described for several taxa by other researchers (MacGarigal and McComb 1995), to my knowledge no other study has documented the importance of edge habitat for woodpeckers. Of 163 total nests, 40 were located directly on an edge. The mechanism driving this association remains unclear. 68 The importance of edge for some species has long been recognized, particularly for big game species such as deer and elk (Hunter 1990). Recently, the term 'edge effect' has come into use to describe the overall increase in wildlife species richness along edges. For some species, especially those that require interior forest for nesting and foraging, edges can be detrimental. Reports of large declines in songbird populations in the eastern United States have been attributed to increased rates of nest parasitism along habitat edges (Hunter 1990). Fragmentation resulting from logging has increased the amount of edge in forest lands. Consideration must be given to the development and adoption of a management strategy that provides some edge habitat to promote wildlife diversity, while minimizing the detrimental effects that occur when too much edge is created. At the landscape level, it may be important to provide habitat that represents a diversity of patch types for nesting woodpeckers. Based on the results of my research, I suggest implementing a plan that incorporates the use of alternative regeneration techniques over a broad landscape, such as partial cut and shelterwood prescriptions. Prescriptions could be used to treat small 20-3 0 acre patches over a landscape to create habitat diversity while maintaining the integrity of the forest matrix--in effect emulating natural forest openings resulting from tree mortality, blow down events, or low-intensity wildfires. Because my results indicate that woodpecker nest density increases with overall habitat complexity, using harvest strategies that create complexity would meet the goals of providing timber resources while at the same time enhancing habitat for woodpeckers. NATURAL AGENTS OF DIVERSITY In attempting to provide suitable habitat for woodpeckers in other areas, regional conditions need to be considered, to tailor management plans to meet the needs of resident cavity-nesters. Lundquist and Mariani (1991) reported that woodpeckers in the southern Washington Cascades disproportionately used western white pine trees for nesting. These results may be driven more by unique site conditions of the region than by any other factor. In my study area, white fir snags were found at high densities in small pockets of forest infected by a root rot disease, armallaria (Armallaria obscura). Stands 69 infected with the disease provide an abundance of snags that serve as nesting and foraging substrate for woodpeckers. These snags eventually fall and are recruited to the stand as part of the down wood component, continuing to serve as foraging substrate for woodpeckers. Viewed from a landscape perspective, there are two concerns that need to be considered. The first is provision of long-term nesting and foraging habitat for woodpeckers. The second concern is a silvicultural objective, to minimize the loss of timber due to disease, and also minimize the rate of spread to uninfected stands. Van der Kamp (1991) suggested taking a different approach to management of pathogens including armallaria, in ecosystems. He asserted that root diseases can be influential in creating unique habitats which favor plant and animal species not well-suited for living in dense, even-aged, coniferous forests, and that pathogens are agents of diversity that can be utilized to create desired habitat conditions. If small pockets of armallaria root rot are allowed to persist within the landscape, spreading at an average rate of 1 meter annually (Goheen et a!. 1994), they will provide a continous supply of nesting habitat for woodpeckers. To minimize spread of disease, adjacent stands can be planted with resistant tree species, thus acting as a natural barrier to further spreading and maintaining a localized patch of infected trees. Forest gaps created by tree death will add to overall habitat complexity, which will also benefit woodpecker populations. These ideas can be incorporated into effective management plans to provide suitable nesting habitat for woodpeckers over landscapes. Wildlife objectives and timber production can both be achieved if a scheme is used that balances forest diversity, overall forest health, and commercial productivity. RESEARCH RECOMMENDATIONS Because I did not measure the reproductive fitness of nesting woodpeckers, I can make oniy limited inferences about habitat suitability. Species abundance does not necessarily equate to suitable habitat (Van Home 1983). Further study needs to be done to measure reproductive fitness of individual woodpecker species in response to forest 70 management activities. Studies that examine nest success of woodpeckers in different habitat patch types over landscapes will provide information on habitat suitability. Furthermore, I examined habitat associations of nesting woodpeckers, and did not attempt to foraging behavior of woodpeckers. Very little information exists on foraging needs of woodpeckers, much emphasis has been placed instead upon nesting habitat needs of woodpeckers. To develop management plans that provide both nesting and foraging habitat components, more information is needed on foraging resource requirements of individual woodpecker species. The observed importance of edge requires further examination. Currently, the mechanism responsible for this association remain unclear. Though I hypothesize that this response is the result of enhanced foraging opportunities near edge, further researchis needed to test this.Specifically, insect population density along edges needs to be determined, and compared with populations in adjacent stands. Studies that examine the interrelationship between woodpeckers and food resources may provide answers to explain this phenomena, and will be beneficial in creating suitable plans for management of woodpeckers. LITERATURECiTED Feen, J.S. 1997. Winter den sites of northern flying squirrels in Douglas-fir forests of the south central Oregon Cascades. M.S. thesis. Oregon State University, Corvallis. 45pp. Goheen, E.M., D.J. Goheen, and K. Marshall. 1994. Insect and disease assessment of the demonstration of ecosystem management options site: Diamond Lake Ranger District, Umpqua National Forest. Southwest Oregon Forest Insect and Disease Technical Center publication. 17 pp. Hunter, M.L. Jr.1990. Wildlife, forests and forestry: Principles of managing forests for biological diversity. Prentice-Hall, Englewood Cliffs, N.J. 370 pp. 71 Lundquist R.W. and J.M. Mariani. 1991. Nesting habitat and abundance of snag- dependent birds in the southern Washington Cascade Range. Pages 22 1-239 L. F. Ruggiero, K.B. Aubry, A.B. Carey, and M.H. Huff, technical coordinators., Wildlife and vegetation of unmanaged Douglas-fir forests. USDA Forest Service General Technical Report PNW-GTR-285. Madsen, S.J. 1985. Habitat use by cavity-nesting birds in the Okanogan National Forest, Washington. M.S. Thesis. University of Washington, Seattle. 113 pp. Mannan, R.W., E.C. Meslow, H.M. Wight. 1980. Use of snags by birds in Douglas-fir forests, western Oregon. Journal of Wildlife Management 44: 787- 797. McGarigal, K. 1993. Relationship between landscape structure and avian abundance patterns in the Oregon Coast Range. Ph.D. Dissertation. Oregon State University, Corvallis. 273 pp. Nelson, S.K. 1989. Habitat use and densities of cavity-nesting birds in the Oregon Coast Ranges. M.S. Thesis. Oregon State University, Corvallis. 157 pp. Ormsbee, P. 1996. Selection of day roosts by female long-legged myotis (Myotis volans) in forests of the central Oregon Cascades. M.S. Thesis. Oregon State University, Corvallis. 63 pp. Raphael, M.G. and M. White. 1984. Use of snags by cavity-nesting birds in the Sierra Nevada. Wildlife Monograph No. 86. 66 pp. Thompson, R.L. 1997. Home range and habitat use of western red-backed voles in mature coniferous forests in the southern Oregon Cascades. M.S. thesis. Oregon State University, Corvallis. pp. USDA Forest Service. 1990. Land and Resource Management Plan -UmpquaNational Forest. Final Environment Impact Statement: Appendices -VolumeII. USDA Forest Service, Pacific Northwest Region. 315 pp. USDA Forest Service and USD1 Bureau of Land Management. 1994. Record of Decision for amendments to Forest Service and Bureau of Land Management Planning Documents within the range of the northern spotted owl. Standards and Guidelines for management of habitat for late successional and old-growth related species within the range of the northern spotted owl. USDA Forest Service and USD1 Bureau of Land Management. 74 pp. Van Home, B. 1983. Density as a misleading indicator of habitat quality. Journal of Wildlife Management 47:893-901. 72 Van der Kamp, B.J. 1991. Pathogens as agents of diversity in forest landscapes. The Forestry Chronicle 67: 353-354. Weikel,J.1997. Habitat use by cavity-nesting birds in young thinned and unthinned Douglas-fir forests of western Oregon. M.S. Thesis, Oregon State University, Corvallis. 102 pp. 73 BIBLIOGRAPHY Autodesk, Inc. 1990. AutoCAD Release 11: Reference manual, publication AC! 1RM. 640 pp. Azet, T., and L.A. McCrimmon. 1990. Preliminary Plant Association of the Southern Oregon CascadeMountain Province. USDAForestService Publication.Siskiyou National Forest. 113 pp. Balda, R.P. 1975. The relationships of secondary cavity-nesters to snag densities in western coniferous forests. USDA Forest Service Wildlife Technical Bulletin 1, Albuquerque, N.M. 37 pp. Beal, F.E.L. 191!. Food of the woodpeckers of the United States. Biological Survey Bulletinv. 37. Beebe, S.B. 1974. Relationships between insectivorous hole-nesting birds and forest management. Yale University School of Forestry and Environmental Studies. 49 pp. Bent, A.C. 1964. Life Histories of North American woodpeckers. Dover Publications. New York, N.Y. 334 pp. Block, W.M., and L.A. Brennan. 1993. The habitat concept in ornithology: Theory applications. Pages 35-83 D.M. Power, ed. Current ornithology, Volume 11. Plenum Press. New York, N.Y. Brooks, J.P. 1997. Bird-habitat relationships at multiple spatial resolutions in the Oregon Coast Range. M.S. Thesis, Oregon State University, Corvallis. 46 pp. Brown, E.R. tech. ed. 1985. Management of wildlife and fish habitats in forests of western Oregon and Washington. Part 2 Appendices. USDA Forest Service, General Technical Report PNW-R8-F&WL-192-1985. Portland, OR. 302 pp. Bull, E.L. 1977. Habitat utilization of the pileated woodpecker, Blue Mountains, Oregon. M. S. Thesis. Oregon State University, Corvallis. 58 pp. Bull, E.L. 1980. Resource partitioning among woodpeckers in northeastern Oregon Ph.D. Dissertation. University of Idaho, Moscow. 109 pp. Bull, E.L. 1986a. Ecological value of dead trees to cavity-nesting birds in northeastern Oregon. Oregon Birds 12: 9 1-99. 74 Bull, E.L.I 986b. Methods of killing trees for use by cavity-nesters. Wildlife Society Bulletin 14:142-146. Bull, E.L., S.R. Peterson, and J.W. Thomas.1 986c. Resource partitioning among woodpeckers in northeastern Oregon. USDA Forest Service Research Note, PNW-444. 19 pp. Bunnell, F.L., L.L. Kremsater, and R.W. Wells. 1997. Likely consequences of forest management on terrestrial, forest-dwelling vertebrates in Oregon. Report M-7 of the Centre for Applied Conservation Biology, University of British Columbia. 129 pp. Capra, F. 1982. The turning point: science, society and the rising culture. Bantam. New York, N.Y. 464 pp. Chambers, C.L., T. Carrigan, T.E. Sabin, J.Tappeiner, and W.C. McComb. 1997. Use of artificially created Douglas-fir snags by cavity-nesting birds. Western Journal of Applied Forestry 12:93-97. Chen, J., J.F. Franklin, and T.A. Spies. 1993. Contrasting microclimates among clearcut, edge, and interior of old-growth Douglas-fir forest. Agricultural and Forest Meteorology 63: 219-237. Cline, S.P. 1977. The characteristics and dynamics of snags in Douglas-fir forests of the Oregon Coast Range. M.S. thesis. Oregon State University, Corvallis. 107 pp. Cline, S.P., A.B. Berg, and H.M. Wight. 1980. Snag characteristics and dynamics in Douglas-fir forests, western Oregon. Journal of Wildlife Management 44: 773- 786. Cohen, J.1988. Statistical power analysis for the behavioral sciences. Second edition. Lawrence Erlbaum Associations, Inc. Hilisdale, N.J. Conner, R.N., O.K. Miller, and C.S. Adkisson. 1976. Woodpecker dependence on trees infected by fungal heart rots. Wilson Bulletin 88: 575-58 1. Conner, R.N. 1979. Seasonal changes in woodpecker foraging methods: strategies for winter survival. Pages 95-105 In: J.G. Dickson, R.N. Conner, R.R. Fleet, eds. The role of insectivorous birds in forest ecosystems: Proc. of a symposium. Nacogdoches,Texas. Academic Press, Inc. New York, N.Y. den Boer, P.J. 1981. On the survival of populations in a heterogeneous and variable environment. Oecologia 50: 39-53. 75 Edgerton, P.J., and J.W. Thomas. 1977. Silvicultural options and habitat values in coniferous forests. Paper presented at the Nongame Bird Habitat Management in Coniferous Forests of the Western United States, Portland, Oregon. Feb. 7-9, 1977. pp. 56-65. Evans, K., and R.N. Conner. 1979. Snag management. pages 214-225 III: R.M. DeGraaf and K.E. eds. Management of north central and northeastern forests for nongame birds. USDA Forest Service General Technical ReportNC-5 1, St. Paul, M.N. Feen, J.S. 1997. Winter den sites of northern flying squirrels in Douglas-fir forests of the south central Oregon Cascades. M.S. thesis. Oregon State University, Corvallis. 45 pp. Filip, G. Personal Communication. October 15, 1995. Forest pathology expert. Oregon State University, Corvallis. Forman, R.T., and M. Godron. 1981. Patches and structural components for a landscape ecology. Bioscience 31: 733-740. Forman, R.T., and M. Godron. 1986. Landscape ecology. John Wiley and Sons, New York,N.Y. 619 pp. Franklin, J.F., and R.T. Forman. 1987. Creating landscape patterns by forest cutting: ecological consequences and principles. Landscape Ecology 1: 5-18 Franklin, J.F., and C.T. Dryness. 1988. Natural vegetation of Oregon and Washington. Oregon State University Press, Corvallis. 452 pp. Franzreb, K.E., and R.D. Ohmart. 1978. The effects of timber harvesting on breeding birds in a mixed coniferous forest. Condor 80: 43 1-441. Garrison, G.A. 1949. Uses and modifications for moosehorn crown closure estimation. Journal of Forestry 47: 733-735. Goheen, E.M., D.J. Goheen, and K. Marshall. 1994. Insect and disease assessment of the demonstration of ecosystem management options site: Diamond Lake Ranger District, Umpqua National Forest. Southwest Oregon Forest Insect and Disease Technical Center publication. 17 pp. Graves, K. 1995. Personal Commnication. March 15, 1997. USDA Forest Service Natural Resource Manager, Diamond Lake Ranger District, Umpqua National Forest. 76 Gutzwiller, K.J., and S.H. Anderson. 1987. Multiscale associations between cavity- nesting birds and features of Wyoming streamside woodlands. Condor 89: 534- 548. Haapanen, A. 1965. Bird fauna of the finnish forests in relation to forest succession. I. Am. Zool. Fenn. 2:153-196 Hagar, D.C. 1960. The interrelationships of logging, birds and timber regeneration in the Douglas-fir region of northwest California. Ecology 41: 116-125. Hagar, J.C., W.C. McComb, and W.H. Emmingham. 1996. Bird communities in commercially thinned and unthinned Douglas-fir stands of western Oregon. Wildlife Society Bulletin 24: 353-366. Harris, L.D. 1984. The fragmented forest. University of Chicago Press, Chicago, I.L. 2llpp. Helle P., and J. Muona. 1985. Invertebrate numbers in edges between clear-fellings and mature forests in northern Finland. Silva Fenn. 19: 28 1-294. Found In: Hunter, Malcom. 1990. Wildlife, forests and forestry: Principles of managing forests for biological diversity. Prentice-Hall, Inc. Englewood Cliffs, N.J. 370 pp. Hope, S., and W.C. McComb. 1994. Perceptions of implementing and monitoring wildlife tree prescriptions on national forests in western Washington and Oregon. Wildlife Society Bulletin 22: 383-392. Hosmer, D.W. Jr., and S. Lemeshow. 1989. Applied logistic regression. John Wiley and Sons, New York, N.Y. 307 pp. Hunter, M.L. Jr. 1990. Wildlife, forests and forestry: Principles of managing forests for biological diversity. Prentice-Hall, Englewood Cliffs, N.J. 370 pp. INFO. Arc/Info Version 7. 1994. Environmental Systems Research Institute, Inc. Redlands, C.A. 337 pp. Jackman, S.M. 1974. Woodpeckers of the Pacific Northwest; their characteristics and their role in the forest. M.S. Thesis, Oregon State University, Corvallis. 147 pp. Johnson, D.H. 1980. The comparison of usage and availability measurements for evaluating resource preference. Ecology 61: 65-70. Karlsson, J., and S.G. Nilsson. 1977. The influence of nest-box area on clutch size in some hole-nesting birds. Ibis 119: 207-211. 77 Kimmey, J.W., and R.L. Furniss. 1943. Deterioration of fire-killed Douglas-fir. USDA Technical Bulletin 351. U.S. Government Printing Office. Washington, D.C. 61 pp. Knight, F.B. 1958. The effects of woodpeckers on populations of the Engelmann spruce beetle. Journal of Economic Entomology 51: 603-60 7. Kotliar, N.B., and J.A. Wiens. 1990. Multiple scales of patchiness and patch structure: a hierarchical framework for the study of heterogeneity. Oikos 59: 253-260. Leopold, A. 1933. Game Management. Charles Scribner Sons. New York, N.Y. 481 pp. Lundquist R.W., and J.M. Mariani. 1991. Nesting habitat and abundance of snag- dependent birds in the southern Washington Cascade Range. Pages 221-239 L. F. Ruggiero, K.B. Aubry, A.B. Carey, and M.H. Huff, technical coordinators., Wildlife and vegetation of unmanaged Douglas-fir forests. USDA Forest Service General Technical Report PNW-GTR-285. Madsen, S.J. 1985. Habitat use by cavity-nesting birds in the Okanogan National Forest, Washington. M.S. Thesis. University of Washington, Seattle. 113 pp. Mannan, R.W. 1977. Use of snags by birds, Douglas-fir region, western Oregon. M.S. Thesis. Oregon State University, Corvallis. 113 pp. Mannan, R.W., E.C. Meslow, and H.M. Wight. 1980. Use of snags by birds in Douglas-fir forests, western Oregon. Journal of Wildlife Management 44: 787- 797. Mariani, J.M., and D.A. Manuwal. 1990. Factors influencing brown creeper (Certhia americana) abundance patterns in the southern Washington Cascade Range. Studies in Avian Biology 13:53-57. Marks, B., and K. McGarigal. 1995. FRAGSTATS: Spatial pattern analysis program forquantifying landscape structure. USDA Forest Service General Technical Report, PNW-GTR-351. 122 pp. McClelland, B.R., and S.S. Frissell. 1975. Identifying forest snags useful for hole- nesting birds. Journal of Forestry 73: 414-417. McClelland, B.R. 1977. Relationships between hole-nesting birds, forest snags and decay in western larch, Douglas-fir forests of the northern Rocky Mountains. Ph.D. Dissertation. University of Montana, Missoula. 495 pp. 78 McGarigal, K. 1993. Relationship between landscape structure and avian abundance patterns in the Oregon Coast Range. Ph.D. Dissertation. Oregon State University, Corvallis. 273 pp. McGarigal, K., and W.C. McComb. 1995. Relations between landscape structure and breeding birds in the Oregon Coast Range. Ecological Monographs 65: 23 5-260. Mellen, T.K. 1987. Home range and habitat use by pileated woodpeckers. M.S. Thesis. Oregon State University, Corvallis. 96 pp. Mime, K.A., and S.J. Heji. 1989. Nest-site characteristics of white-headed woodpeckers. 1989. Journal of Wildlife Management 53: 50-54. Naveh, Z., and A. S. Liebermann. 1984. Landscape ecology, theory and application. Springer-Verlag, New York, N.Y. 356 pp. Nelson, S.K. 1989. Habitat use and densities of cavity-nesting birds in the Oregon Coast Ranges. M.S. Thesis. Oregon State University, Corvallis. 157 pp. Neitro, W.A., V.W. Binkley, S.P. Cline, R.W. Mannan, B.G. Marcot, D. Taylor, and F.F. Wagner. 1985. Snags. Pages 129-168 In: Brown, E.R. tech. ed. Management of wildlife and fish habitats in forests of western Oregon and Washington. Part 2 - Appendices.USDA Forest Service, Pacific Northwest Region, Portland, O.R. Nilsson, S.G. 1984. The evolution of nest-site selection among hole-nesting birds: the importance of nest predation and competition. Ornis Scand. 15: 167-175 Ormsbee, P. 1996. Selection of day roosts by female long-legged myotis (Myotis volans) in forests of the central Oregon Cascades. M.S. Thesis. Oregon State University, Corvallis. 63 pp. Otvos, I.S. 1965. Studies on avian predators of Dendroctonus brevicomis le Conte (Coleoptera: Scolytidae) with special reference to Picidae. Canadian Entomologist 97: 1184-1199. 1979.The effects of insectivorous bird activities in forest ecosystems: an evaluation. Pages 341-374 In: J.G. Dickson et al., eds. The role of insectivorous birds in forest ecosystems. Academic Press. New York, N.Y. Otvos, I.S. and R.W. Stark. 1985. Arthropod food of some forest-inhabiting birds. The Cannadian Entomologist 117:971-990 ______• 79 Ramsey, F.L., M. McCracken, J.A. Crawford, M.S. Drut, and W.J. Ripple. 1994. Habitat association studies of the northern spotted owl, sage grouse, and flammulated owl. Pages 189-209 N.Lange, L. Ryan, L. Billard, D. Brillinger, Conquest and J. Greenhouse (eds.). Case Studies in Biometry. John Wiley and Sons, New York, N.Y. 496 pp. Raphael, M.G., and M. White. 1984. Use of snags by cavity-nesting birds in the Sierra Nevada. Wildlife Monograph No. 86. 66 pp. Raphael, M.G., and M.L. Morrison. 1987. Decay and dynamics of snags in the Sierra Nevada, California. Forest Science 33: 774-783. 1990.SAS/STAT Procedures Guide, Version 6, Third Edition, Cary, N.C: 705 pp. 1997.SAS/STAT software: changes and enhancements through release 6.12. Cary,N.C. :ll67pp. Schreiber, B. 1987. Bird use of snags on clearcuts in central coastal Oregon. M.S. Thesis, Oregon State University, Corvallis. 63 pp. Schreiber, B., and D.S. deCalesta. 1992. The relationship between cavity-nesting birds and snags on clearcuts in wetern Oregon. Forest Ecology and Management 50: 299-3 16. Sedgwick, J.A., and F.L. Knopf 1990. Habitat relationships and nest site characteristics of cavity-nesting birds in cottonwood floodplains. Journal of Wildlife Management 54: 112-124. Swallow, S.K., and R.J. Gutierrez. 1986. Primary cavity-site selection by birds. Journal of Wildlife Management 50: 516-583. Thomas, J.W., G.L. Crouch, R.S. Bumstead, and L.D. Bryant. 1975. Silvicultural options and habitat values in coniferous forests. In: Proc. symposium management of forest and range habitats for nongame birds. USDA Forest Service General Technical Report WO- 1. Thompson, R.L. 1997. Home range and habitat use of western red-backed voles in mature coniferous forests in the southern Oregon Cascades. M.S. thesis. Oregon State University, Corvallis. pp. USDA Forest Service. 1990. Land and Resource Management Plan -UmpquaNational Forest. Final Environment Impact Statement: Appendices -VolumeII. USDA Forest Service, Pacific Northwest Region. 315 pp. 80 USDA Forest Service and USD1 Bureau of Land Management. 1994. Record of Decision for amendments to Forest Service and Bureau of Land Management Planning Documents within the range of the northern spotted owl. Standards and Guidelines for management of habitat for late successional and old-growth related species within the range of the northern spotted owl. USDA Forest Service and USD1 Bureau of Land Management. 74 pp. Van der Kamp, B.J. 1991. Pathogens as agents of diversity in forest landscapes. The Forestry Chronicle 67: 353-354. Van Home, B. 1983. Density as a misleading indicator of habitat quality. Journal of Wildlife Management 47:893-901. Weikel, J. 1997. Habitat use by cavity-nesting birds in young thinned and unthinned Douglas-fir forests of western Oregon. M.S. Thesis, Oregon State University, Corvallis. 102 pp. Wiens, J.A. 1976. Population responses to patchy environments. Annual Review of Ecological Systematics 7: 81-120. Wiens, J.A. 1989. Spatial Scaling in Ecology. Functional Ecology 3: 385-397. Winkler, H., D.A. Christie, and D. Nurney. 1995. Woodpeckers: An identification guide to the woodpeckers of the world. Houghton Mifflin Co. New York, N.Y. 406 pp. Wright, K.H., and G.M. Harvey. 1967. The deterioration of beetle-killed Douglas-fir in Oregon and Washington. USDA Forest Service Research Paper. Pacific Northwest Region, Portland, OR. 22 pp. Zamowitz, J.E., and D.A. Manuwal. 1985. The effects of forest management on cavity- nesting birds in northwestern Washington. Journal of Wildlife Management 49: 255-263. 81 APPENDICES 82 APPENDIX A- Maps of eleven study sites showing habitat patch type and spatial arrangement, and location of woodpecker nests at each site. STUMP LAKE SITE N Neat, LEGEND •A Hafry AS-breasted wood peck.. espauckec PII.Sedwooc*.ecker •*Patth 2 Mitt,.1 Mthve lodgepolePOWSITUb FareS ian 6 ClearcuI7 Clearcut with wIIh etmjbe-no snag WoodpecSouthernIn the Dia mondker nestOregon Lake locations Ranger over District, 200-Ha Umpqua lands 1995-1996 capesNational Forest 2 00 0 200 400 600 800 100 0 Meters FOREBAY SITE a a a aa a aaaam a\.7a Nests LEGEND a a a a a aaaaaaaaa a a a a a a aaa4a a a a a *•£ HairyNorthernPileatedRed-breasted woodpecker flicker sapsucker woodpecker a a a a a Patch type ShaiterwoodMatLts Forest a a cleercutClsarcutClewout withwith shrub.-snags shrubs snag a a kaa.. Wood pecker nest locations over 200-Ha lands capes 00 0 200 400 600 800 100 0 Meters Southin the emDiamond Oregon Cascades,Lake Ranger 1995-1996 District, Umpqua National Forest TRAP CREEK SITE N #81 111111 S 'S A LEGEND Neste *•A HairyPfleated Red-breasted woodpecker woodpecker sapsuc Nofthern flicker ker •Patch type 53 PoIe(shrubClearcut1 Mature with Forest snags WoodSouthIn the empeckerDiamond Oregon nest Cascades, Lakelocations Ranger over1995-1 District, 200-Ha 996 Umpqua lands capesNational Forest 2 00 0 200 400 600 800 100 6 Clearcut with shrubs 0 Meters 00 00 kor t National Forest z capes it mu C C LEGEND wE o a'C 0) o m —xoMs. o 0 C sEQS Woodpec in the Dia Southern o Meters 0 •/ 0 ID 0 0 c'J 0 SPRUCE SITE 0 LITTLE BEAR SITE N Nest. £ HafryRed-breasted woodpecker LEGEND sapsLIclcer *• NaqthecnPatol, type flicks, 2 Mature1 Mature lodgepol. Forest wo*eoker 7S Clearcut CrewcutClearciitShellerwood with snags vithwith siTub, 0 SouthernInWood the DIapec mondkerOregon nestLake locations RangerCascades, District,over 1995-1996 200-Ha Umpqua lands capesNational Forest 2 00 0 200 400 600 800 100 0 Meters 00 MOWICH SITE N I SstA HairyRed-StasiS woodpecker LEGEND sapsucker Patc.* htypePileeledNorthern flicker woodpecker woodpecker C15* CIOaTCUtClearcutSheMature ito iwood Forest with with snags ,hmbs SouthIntheWood empeckerDiamond Oregon nest Cascades,Lakelocations Ranger over 1995-1996 District, 200-Ha Umpqua lands capesNational Forest 2 00 0 200 400 600 800 100 0 Meters 89 ker S z U 0 0 Meters inUI nI..chr 0• LEGEND 0 i:;:!— 0 0 CD as 0 0 a c'J / (4 0 A 4 VW( 4 .• 2 <4 4 4 4 4 4 •• a % . . :4•. 4 4. %A X • . X tS a . . — V I . A I — • .. . V / / I<((<<(V<<< ¼ A . . .1 A flr 4 t . A A 4 4 C 5 C. .7/ m E rn_iC • 0 .xEO0 Woodpec In the Dia Southern VELVET CREEK SITE ROUGH CREEK SITE N Nests A Red-breasted LEGEND sepsuckec • HakywoodpeckPatch Northern tUckerI Mature Foresttype .. 65 ClearcutCloarcut with shrubssnagsPole/shrub SoutherninWoodpec the Dia mondkerOregon nestLake locations RangerCascades, District,over 1995-1 200-Ha Umpqua 996 lands capesNational Forest 2 00 0 200 400 600 I 800 100 0 Meters C MAPLE CREEK SITE N Men £ Red-breasled sapsuc LEGEND ker *• HairyPetch woodpecker type Mawr. rorest woodpecker CleartutCleamutCIeawutShfle,woodPol&eIwLt wiltiwith with shrubssnaga sirute -no snags SouthInWood the empeckerDiamond Oregon nest Cascades, Lakelocations Ranger over 1995-1996 District, 200-Ha Umpqua lands capesNational Forest 2 00 0 200 400 600 800 100 0 Meters — WATSON CREEK SITE N LEGEND Nests *•A NorthernHairyPileated woodpecker woodpecker flicker sapsuc ker + Red-napedPatchE21 5 Clearcut with snags 1 Mature Foresttype sapsucker 2 00 0 200 400 600 800 Clearcut with shrubs 100 0 Meters SouthInWood the empeckerDiamond Oregon nest Cascades, Lakelocations Ranger over1995-1 DistrIct, 200-Ha 996 Umpqua lands capesNational Forest ____LEGEND PYGMY KNOLL SITE N A A Nests •S NorthernHefrywoodpecker Red-breasted flicker saps Patth type S1 ClearcutMsttn Forest with SnPole/skit age SouthInWood the empeckerDiamond Oregon nest Cascades, Lakelocations Ranger over 1995-1996 District, 200-Ha Umpqua lands capesNational Forest 2 00 0 200 400 600 800 100 0 Meters 94 Appendix B. Weight file used in FRAGSTATS analysis to characterize vegetation characteristics of 200-ha landscapes in the Diamond Lake Ranger District, Umpqua National Forest, Southern Oregon Cascades, 1995-1996 BOUNDARY TYPE WEIGHT 1-2 0.05 1-3 0.20 1-4 0.05 1-5 0.25 1-6 0.45 1-7 0.45 2-3 0.25 2-4 0.10 2-5 0.30 2-6 0.50 2-7 0.50 3-4 0.25 3-5 0.05 3-6 0.25 3-7 0.25 4-5 0.20 4-6 0.40 4-7 0.40 5-6 0.20 5-7 0.20 6-7 0.00 woodpeckerAppendix1996. C. nests Class were level located indices on generated the Diamond in FRAGSTATS Lake Ranger District, to quantify habitat characteristics for 11 200-ha landscapes where Umpqua National Forest, Southern Oregon Cascades. 1995- SITE Numberof Nests % Mature forest Forestpatch Mean patch size (ha) density Edge Edge % Core Density of ForebayStump Lake 2119 43.710.1 density 2.21.1 20.39.4 21.835.1(m) contrast 24.133.0 area forest 0.09.3 core area 0.12.2 LittleSpruceTrapCreek Bear 18146 58.994.841.9 0.540.55 3.8 109.5173.311.2 31.231.76.5 0.9637.413.3 64.716,69.9 0.552.21.6 RoughVelvetMowich Creek 13147 41.776.889.2 0.570.530.52 145.4171.573.6 34.521.915.3 39.215.312.6 34.151.46.9 0.530.521.13 PygmyWatsonMaple CreekKnoll Creek 191312 35.460.948.9 2.11.61.7 21.735.423.1 55.240.314.5 22.337.629.5 15.19.89.6 0.542.92.7