University of Montana ScholarWorks at University of Montana
Graduate Student Theses, Dissertations, & Professional Papers Graduate School
1996
Effects of fire and salvage logging on the cavity-nesting bird community in northwestern Montana
Elaine L. Caton The University of Montana
Follow this and additional works at: https://scholarworks.umt.edu/etd Let us know how access to this document benefits ou.y
Recommended Citation Caton, Elaine L., "Effects of fire and salvage logging on the cavity-nesting bird community in northwestern Montana" (1996). Graduate Student Theses, Dissertations, & Professional Papers. 10470. https://scholarworks.umt.edu/etd/10470
This Dissertation is brought to you for free and open access by the Graduate School at ScholarWorks at University of Montana. It has been accepted for inclusion in Graduate Student Theses, Dissertations, & Professional Papers by an authorized administrator of ScholarWorks at University of Montana. For more information, please contact [email protected]. INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.
The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely afreet reproduction.
In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.
Photographs included in the original manuscript have been reproduced xerographically in this copy. Ifigher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. UMI A Bell & Howell Infonnaüon Company 300 North Zed) Road, Ann Arbor MI 48106-1346 USA 313^761-4700 800/521-0600
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. EFFECTS OF FIRE AND SALVAGE LOGGING ON THE CAVITY-NESTING BIRD COMMUNITY IN NORTHWESTERN MONTANA
by Elaine L . Caton B.A., The University of Montana, 1984
Presented in partial fulfillment of the requirements
for the degree of Doctor of Philosophy The University of Montana 1996
Approved by:
Chairperson
Dean, Graduate School
______Date
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 9718524
UMI Microform 9718524 Copyright 1997, by UMI Company. All rights reserved.
This microform edition is protected against unauthorized copying under Title 17, United States Code.
UMI 300 North Zeeb Road Ann Arbor, MI 48103
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. i
Maureen and Mike MANSFIELD LIBRARY
The UniversityoflVIONTANA
Pemiission is granted by tlie autlior to reproduce this material in its entirety, provided tliat tliis material is used for scholarly purposes and is properly cited in published works and reports.
** ** Please check "Yes" or "No" and provide signature
Yes, I grant permission No, I do not grant permission ______
Author's Signature ^
Date ______
Any copying for commercial purposes or financial gain may be undertaken only with tlie author's explicit consent.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT Caton, Elaine L., Ph.D., December 1996 Biological Sciences
Effects of fire and salvage-logging on a cavity-nesting bird community in northwestern Montana (115 pp.) Advisor: Richard L. Hutto
I studied the effects of a large, stand-replacement fire on a cavity-nesting bird community in northwestern Montana for five years, starting two years post fire. I compared bird abundances in burned and unbumed mid-elevation coniferous forests, and investigated the importance of nest-site and foraging-site availability in explaining differences in abundance. Differences in abundance appeared to be due largely to differences in foraging opportunities, rather than to nest-site availability. Timber-drilling woodpeckers, aerial insectivores, and ground foragers were more abundant in post-fire habitat, whereas gleaners were less abundant. Pairs of species drawn from the same foraging guild were significantly more similar in abundance patterns than pairs drawn from different guilds, but species with similar nest sites did not show similar patterns of abundance. I recorded characteristics of sites used for foraging and nesting in the post-fire forest. Use of sites appeared to be determined by presence of foraging sites as well as nest-tree availability. Woodpeckers foraged on large larch, Douglas- fir, and lodgepole pines in stands with basal areas greater than expected based on availability. Correspondingly, nests were placed more often than expected in stands of high basal area and large-diameter trees. Nests were predominantly in large, broken-top aspen. I compared habitat and bird abundances between salvage- logged and unlogged post-fire habitat. Bird abundances and nest densities were lower in post-fire forest stands that were logged, and fewer sites in logged stands were suitable for nesting. Within logged areas, birds nested in sites most similar to nest sites in unlogged habitat. Tree-foraging cavity nesters were less abundant in clearcut stands than in partially logged stands. Variation in forest stcinds around nests rather than in characteristics of nest trees seemed to best explain this difference in abundance. Fire maintains cavity nester diversity by providing plentiful foraging and nesting sites. Management actions that address only nest tree provisions will fall short of providing for all community members. Fire-created habitats that undergo little or no post-fire logging may become réfugia for species facing rapidly dwindling sources of standing dead wood.
11
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGEMENTS My interest in and curiosity about the natural world
were stimulated at an early age by the enthusiasm and
interest of my parents, especially my mother. I thank all of my family for their support of and pride in my endeavors.
Riley McClelland has been a source of inspiration and
encouragement for many years. I am indebted to Riley and Pat
McClelland for all their professional help, and for treating
me as a member of the family when ny own was so far away. Rob Bennetts helped catalyze my interests in research biology, and our many thought-provoking discussions enriched my education. Many friends and colleagues at Glacier
National Park helped me in countless ways during my research,
with logistics, meals, friendships, housing, and by looking for me when I was late in returning from my solitary fieldwork. Although I can't name them all here, I especially
thank Susan Bechtel Allen, Kathy Dimont, Mac Donofrio, Carl
and Lindy Key, Richard Menicke, Gary Vodehnal, Rick Yates, and Jack Polzin and the North Fork Trailcrew. My field assistants Lynn Bacon and Rosalind Yanishevsky provided not just outstanding field work under sometimes miserable conditions, but also many useful ideas and
thoughtful criticisms. Carl Key and Richard Menicke of Glacier National Park gave generously of their time and expertise to supply me with GIS maps and data. Steve Gniadek and Jim Tilmant provided helpful input and provided logistic
iii
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. support along the way. Shari Gregory of the Flathead
National Forest supplied information on the study area. I benefitted greatly from input provided by my committee members Sallie Hejl, Dick Hutto, Penny Kukuk, Riley
McClelland, and David Patterson. I especially thank my
advisor, Dick Hutto, for compelling me to clarify my ideas and put them into context. This project and my own education profited significantly from his enthusiastic involvement and stimulating conversations.
My fellow graduate students helped me maintain my
equilibrium during trying times; their friendships are much
appreciated. Susan Hitchcox, Vita Wright, and Jock Young
gave thoughtful criticisms and helpful suggestions through the many stages of the project.
I am particularly grateful to Tim Swanberg for his
enthusiastic support and complete confidence in my abilities, which made this accomplishment seem possible and worthwhile. His humor and unending patience sustained me countless times when all else failed.
Funding for this research was provided by Glacier National Park and a Bertha Morton Award.
IV
Reproduced with permission ot the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS ABSTRACT ...... ii
ACKNOWLEDGEMENTS ...... iii
LIST OF TABLES ...... vii
LIST OF FIGURES ...... ix Chapter I INTRODUCTION ...... 1
Literature Cited ...... 3
II THE EFFECTS OF FIRE ON A CAVITY-NESTING BIRD COMMUNITY IN NORTHWESTERN M O N T A N A ...... 7 Study A r e a ...... 12
Methods ...... 15
Results ...... 23
D i s c u s s i o n ...... 29
Literature Cited ...... 3 8
III USE OF AN EARLY POST-FIRE FOREST BY CAVITY-
NESTING BIRDS: IMPORTANCE OF NESTING AND FORAGING OPPORTUNITIES ...... 50 Study Site ...... 52
Methods ...... 53
Foraging habitat ...... 53
Nesting habitat ...... 54 Data Analyses ...... 56
Results ...... 57
Foraging habitat ...... 57
Nesting Habitat ...... 51
V
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Literature Cited ...... 83 IV THE EFFECTS OF POST-FIRE SALVAGE LOGGING ON CAVITY-NESTING BIRDS ...... 90 Study Site ...... 93
Methods ...... 94
Nest and bird abundances ...... 94
Habitat measurements ...... 95
Results _ ...... 97
Nest and bird abundances ...... 97
Habitat characteristics ...... 102
Unlogged and logged ...... 102
Partially cut and clearcut . 104 Discussion ...... 107 Literature Cited ...... Ill
VI
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES
TABLE Page 1 Variables recorded at nest trees and random trees in post-fire habitat...... 19
2 Bird species and foraging group assignments. . . 21
3 Common name, scientific name, and number of nests found within the Red Bench Fire study area for each of the 21 cavity-nesting species encountered. . . 24
4 Change in relative nest abundance from unbumed to burned forest in the current study, number of studies that reported increases and decreases in abundances, and p-value of chi-square tests of response distributions...... 26 5 Variables recorded at nest and random trees and sites in post-fire habitat...... 55
6 Common name, scientific name, and number of nests found within the Red Bench Fire study area of Glacier National Park for 20 cavity-nesting species. 62
7 Means (±SE) of nest and random tree variables in post-fire habitat...... 63
8 Tree species availability and use by cavity-nesting birds in early post-fire habitat. Glacier National Park, Montana...... 65
9 Tree condition availability and use by cavity- nesting birds in early post-fire habitat. Glacier National Park, Montana...... 68
10 Means of variables recorded around nest trees and randomly selected trees in post-fire habitat. . . 71
11 Nest-site characteristics of cavity-nesting bird species grouped by foraging habits...... 75
12 Results of stepwise multiple logistic regression of nest and random trees in post-fire habitat, based on characteristics of trees and the surrounding 50-m- radius forest stand...... 76
13 Variables recorded at nest and random trees and sites in post-fire habitat...... 96
VIX
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 14 Number of nests of each cavity-nesting bird species found in burned forests of Glacier National Park and Flathead National Forest from 1990- 1994. . . . 98 15 Means ±SE of habitat variables recorded at randomly located trees and nest trees in burned, unlogged forests of Glacier National Park and in burned, logged forests of Flathead National Forest .... 103
16 Means ±SE of habitat variables recorded at randomly located trees and nest trees in partially cut and clearcut post-fire stands of Flathead National Forest...... 106
Vlll
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF FIGURES FIGURE Page
1 Location of study area...... 14
2 Relative nest abundances of cavity-nesting bird species inburned and unburned forests. 24 3 Cavity-nesting bird abundances at point counts in burned and unbumed forests...... 28 4 Cavity-nest abundances in burned, clearcut and bumed, partially cut stands. 30
5 Tree diameter and % bark disturbed by woodpeckers in post-fire habitat. .... 59
6 Tree species and category of % bark disturbed by woodpeckers in post-fire habitat...... 60 7 Use and availability of tree species for nesting in post-fire habitat. .... 67
8 Tree condition use and availability for nesting in post-fire habitat. .... 69
9 Use and availability of sites of different b u m severity for nesting in post-fire habitat...... 73
10 Pattem of similarity of nest sites of cavity-nesting bird species and randomly selected trees in post-fire habitat...... 74
11 Nest abundances in bumed, clearcut and bumed, partially logged stands. . . . 100 12 Bird abundances at point counts in burned, clearcut and bumed, partially logged stands. 101 13 Pattern of similarity of nest sites and randomly selected sites in bumed, unlogged and bumed, logged forests. . . . 105
IX
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter I INTRODUCTION Relationships between cavity-nesting birds and forest
habitat have received considerable attention from ecologists,
forest managers, and conservationists. Cavity-nesting birds
comprise approximately 25% of forest bird species in
temperate zones (Gibbs et al. 1993). They are a diverse group of species from several orders, and they fill many
ecological roles. Most cavity nesters are primarily
insectivorous (Erlich et al. 1988), and may affect the populations and activities of their prey species (Hutchinson
1951, Otvos 1965, Beebe 1974). The feeding activities of tree-foraging cavity nesters may introduce or increase
pathways of decay into trees and make food resources
accessible for other species (Bent 1939, Foster and Tate
1966, Erlich et al. 1988) . Nest cavities provide denning and
resting sites for small mammals and réfugia from weather and predators for other bird species (McLaren 1963, Kilham 1971, McClelland 1979).
Changes in forest characteristics affect use by cavity nesters, because of their restrictive nesting requirements
and dependence on forest insects (McClelland 1980, Scott and Oldemeyer 1983, Renken and Wiggers 1989). Habitat of cavity nesters has been altered by timber harvest and fire
suppression, two major anthropogenic factors of change in the
Rocky Mountains. Both activities tend to reduce the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. parameters important to the group. Logging removes potential
foraging and nesting substrates and alters forest age structure and species composition (Weliner 1984). Several cavity-nesting species are among those most negatively
affected by widespread alterations of forests from logging
(Scott and Oldemeyer 1983, Raphael and White 1984, Hejl et al. 1995) .
Fire has been a major influence on forests in western North America for millennia (Gruell 1983, Agee 1993). It has maintained conditions that allow for the persistence of some
species, while limiting or excluding others (Habeck and Mutch
1973) . Stand-replacement fires typical of the pre-European
fire regime in mid-elevation coniferous forests generated tremendous expanses of standing dead trees (Amo 1980, Agee
1993) . These snag forests provided habitat for numerous
insect species that flourish in freshly dead wood (Evans 1971). Some fire-killed trees also must have served as
nesting sites for birds. Although most cavity-nesting birds are abundant in post-fire habitat (Hutto 1995), little is
known about the mechanisms by which fire affects cavity
nesters, the use of post-fire forests by cavity-nesting
species, and how logging of bumed forests affects bird abundances.
In the following chapters, I present the results of a 5- year study on cavity-nesting bird ecology in a recently
bumed forest in northwestern Montana. In chapter II, I
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. discuss how fire affects bird abundances, the mechanisms by
which those changes occur, and subsequent influences on community structure and diversity. Chapter III describes foraging and nesting habitat use by cavity nesters within the
post-fire environment. Chapter IV assesses the impacts of
salvage logging on cavity-nesting bird habitat and
abundances. Because each chapter was written to be submitted
as a separate journal publication, there is some repetition
of study area, methods, and citations among chapters.
Literature Cited
Agee, J. K. 1993. Fire ecology of Pacific Northwest forests. Island Press, Washington, D. C.
Arno, S. F. 1980. Forest fire history in the Northern Rockies. J. For. 78:460-465.
Beebe, S. B. 1974. Relationships between insectivorous hole- nesting birds and forest management. Yale Univ. School of
Forestry and Environmental Studies. New Haven, CT. 49 pp.
Bent, A. C. 1939. Life histories of North American
woodpeckers. U.S. Nat. Mus. Bull. 174. Washington, D.C.
Ehrlich, P. R., D. S. Dobkin, and D. Wheye. 1988. The
birder's handbook: A field guide to the natural history of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. North American birds. Simon and Schuster, New York.
Evans, W. G. 1971. The attraction of insects to forest fires.
In Tall Timbers Conf. on Ecol. Animal Control by Habitat Manage. Proc. 3:115-127.
Foster, W. L., and J. Tate, Jr. 1966. The activities and
coactions of animals at sapsucker trees. Living Bird 5:87- 113 .
Gibbs, J. P., M. L. Hunter, and S. M. Melvin. 1993. Snag
availability and communities of cavity-nesting birds in
tropical versus temperate forests. Biotropica 25(2):236- 241.
Gruell, G. E. 1983. Fire and vegetative trends in the
Northern Rockies: interpretations from 1871-1982
photographs. General technical report INT-150. U.S. Forest Service, Ogden, Utah
Habeck, J. R., and R. W. Mutch. 1973. Fire-dependent forests in the Northern Rocky Mountains. J. Quaternary Res. 3(3):408-424.
Hejl, S. J., R.L. Hutto, C.R. Preston, and D.M. Finch. 1995.
The effects of silvicultural treatments on forest birds in
the Rocky Mountains. Pages 220-244 in Population ecology
and conservation of neotropical migratory birds (T. Martin,
and D.M. Finch, eds.). Oxford University Press, New York.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Hutchinson, F. T. 1951. The effect of woodpeckers on the Engelitiann spruce beetle, Dendroctonus engelmannii Hopkins. M.S. Thesis, Colo. State Univ., 73 pp.
Hutto, R. L. 1995. Composition of bird communities following
stand-replacement fires in Northern Rock/ Mountain (U.S.A.)
conifer forests. Cons. Biol. 9(5):1041-1058.
Kilham, L. 1971. Reproductive behavior of Yellow-bellied
Sapsuckers I. Preference for nesting in Fames infected aspens and nest hole interrelations with flying squirrels,
raccoons, and other animals. Wilson Bull. 83:159-171.
McClelland, B. R. 1979. The Pileated Woodpecker in forests of
the northern Rocky Mountains. In The role of insectivorous birds in forest ecosystems (J.G. Dickson, R.N. Conner, R.R. Fleet, J.A. Jackson, and J.C. Kroll, eds.). Academic Press, N.Y.
McClelland, B. R. 1980. Influences of harvesting and residue
management on cavity-nesting birds. In Proceedings of a symposium on environmental consequences of timber harvesting in Rocky Mountain coniferous forests. USDA
Forest Service General Technical Report INT-90, Ogden, Utah.
McLaren, W. D. 1963. A preliminary study of nest-site
competition in a group of hole-nesting birds. M.S. Thesis,
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Univ. of British Columbia, Vancouver.
Otvos, I. M. 1965. Studies on avian predators of Dendroc tonus
brevicomis Lecone (Coleoptera: Scolytidae) with special reference to Picidae. Canadian Entomology 97:1184-1199.
Raphael, M. G . , and M. White. 1984. Use of snags by cavity- nesting birds in the Sierra Nevada. Wildl. Monog. 86:1-66.
Renken, R. B., and E. P. Wiggers. 1989. Forest
characteristics related to Pileated Woodpecker territory size in Missouri. Condor 91:642-652.
Scott, V. E., and J. L. Oldemeyer. 1983. Cavity-nesting bird
requirements and response to snag cutting in ponderosa
pine. Pp. 19-23 in Proceedings of a symposium on snag
habitat management (J. W. Davis, G. A. Goodwin, and R. A.
Ockenfels, tech. coords.). USDA Forest Service, General Technical Report RM-99.
We liner, C . A. 1984. History and status of silvicultural
management in the interior Douglas-fir and grand fir forest
types. Pp. 3-10 in Proceedings of a symposium on silvicultural management strategies for pests of the interior Douglas-fir and grand fir forest types (D. M.
Baumgartner and R. Mitchell, comps.). Washington State Univ., Pullman.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter II THE EFFECTS OF FIRE ON A CAVITY-NESTING BIRD
COMMUNITY IN NORTHWESTERN MONTANA
Introduction
The effects of disturbance on the structure of
biological communities have received considerable attention
from ecologists. In early views, disturbance was considered a rare event that destroyed the normal equilibrium of ecosystems (Clements 1916). More recently, ecologists
recognize natural disturbance as a process important in
maintaining species diversity (May 1973, Thiery 1982, Pickett and White 1985, Denslow 1985).
According to this more recent view, ecosystems and communities are dynamic, with populations and resources often
in a state of temporal or spatial flux. This dynamism may
mediate community interactions and influence species composition. Disturbance may increase environmental heterogeneity by creating a patchy landscape, due to spatial variation in disturbance frequency, size, and intensity
(Grubb 1977, Denslow 1982, Wiens 1985). Disturbance may
increase the availability of some resources used by organisms
and decrease other resources. Because species use unique sets of resources within the same environment, a single disturbance event may effect complex changes in a community.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8
The consequences of disturbance on communities may vary
greatly at different spatial and temporal scales. For
example, disturbance may decrease species diversity locally
by eliminating some species, but increase regional diversity by providing habitat for species usually absent in undisturbed habitat (Wiens 1985). Disturbance also may
decrease diversity initially, but allow a variety of species
to occupy a site through different successional stages,
creating a temporally dynamic community (Taylor and Barmore 1980) .
Natural fire has played a role in the evolution of many
species and in the composition of the communities in which
they occur. In the northern Rocky Mountains, fire has been a vital force in shaping ecosystems (Heinselman 1971, Habeck and Mutch 1973, A m o 1980) . Fire influences forest
composition by allowing species such as western larch (Larix occidentalis) and ponderosa pine (Pinus ponderosa) to
persist, while excluding or limiting others (Agee 1993). Fire also shapes forest structure by creating snags and new age classes. Cavity-nesting birds, which comprise approximately 25% of the area's forest bird species, are
greatly influenced by forest composition and structure
(McClelland et al. 1979, Raphael and White 1984, Zarnowitz and Manuwal 1985).
Cavity-nesting birds depend on specific types of trees because of restrictive needs for nest construction. Cavity
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. nesters require trees of adequate minimal diameter, usually with soft heartwood (McClelland 1977, Raphael and White 1984), and may even require particular tree species and
degrees of decay (Shigo and Kilham 1968, Conner et al. 1976).
Cavity nesters are primarily insectivorous, and many of their
prey species are associated with specific forest conditions.
For example, many beetle larvae (on which woodpeckers feed) develop only in certain species and sizes of trees, and often
are restricted to freshly dead wood (Fumiss and Carolin
1977, Fellin 1980).
Because of these specialized nesting and foraging
habits, many cavity nesters are probably sensitive to changes in forest structure. Before Euro-American fire suppression
altered fire regimes, fires maintained forests in mosaics of
various successional stages and compositions (Habeck and
Mutch 1973, Arno 1980). This provided diverse habitats, supporting the full range of cavity-nesting species.
Fire in the northern Rocky Mountains may benefit cavity
nesters in various ways. The immediate effects of fire on
cavity nesters are probably minimal (i.e., temporary physical displacement during the actual event). Long-term effects may
be substantial due to extreme alteration of the environment.
Because nesting and foraging requirements vary among species,
different cavity-nesting species may respond differently to
fire's effects. Variation in responses would change
community structure and species' interactions.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 10
Some authors have documented the effects of fire on bird abundances in Rocky Mountain forests (see Hutto 1995a for a
review). Cavity nesters generally respond positively to post-fire conditions (Bock et al. 1978, Neimi 1978, Pfister 1980, Taylor and Barmore 1980, Harris 1982). However, it is
unclear by which mechanisms fire affects cavity nesters, and
why species in this group differ from one another in their response to fire.
Fire may act on at least two important resources: nest sites and food. Fire-created snags may serve as nesting
substrates because most cavity nests are in dead trees or
dead portions of trees (e.g., Conner et al. 1976, McClelland 1977). However, an advanced stage of heartrot decay also is
an important component of nest substrate suitability (Shigo and Kilham 1968, Conner et al. 1976, McClelland 1977, Mannan et al. 1980, Raphael and White 1984), and fire-killed
conifers generally do not reach such a stage of decay until at least 5 years post-fire (Kimmey 1955, Etheridge 1961, Cline et al. 1980). Moreover, pre-existing snags tend to burn completely (Horton and Mannan 1988). Aspen are used in both
decayed and undecayed states by some species (Harris 1982,
Wintemitz and Cahn 1983), and decay tends to be prevalent in both live and dead aspen (Riley 1952, Basham 1958). In sum,
the effects of fire on nest-site availability in the first
few years after fire are unclear, but potentially important.
Availability of nest sites often is suggested as the
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 11
limiting factor for cavity-nesting bird populations (Von Haartman 1957, Hilden 1965, Conner 1978, Slagsvold 1978,
Gustafsson 1988; but see Waters 1990, Welsh and Capen 1992).
Nevertheless, factors such as food availability may be
limiting, particularly in unmanaged forests where nest sites may not be scarce (Wesolowski 1989). Events that affect food resources, then, may influence bird abundances in some
habitats. Fires attract numerous insects, especially
wood-boring beetles (Evans 1971) that are fed on by several cavity-nesting species (Beal 1911). Wood-borer larvae are generally eibundant in b u m s for the first few years after
fire, and concentrations of woodpeckers feeding on wood-borer
larvae have been observed in post-fire habitat (Blackford 1955, Koplin 1969).
The objectives of this study were to document how cavity-nesting birds were affected by a stand-replacement fire in northwestern Montana, and to investigate mechanisms
by which changes in bird abundances occurred. Specifically, I investigated whether differences in nest or food
availability most influenced differences in abundances of various cavity-nesting species between bumed and unburned
forests. Specifically, I tested predictions that follow from each of the two hypotheses, as outlined below:
H^: Cavity nesters respond primarily to nest-site
modifications.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12
P^. Differences in bird abundances between bumed and
unbumed habitat will be more pronounced in the breeding season than in the nonbreeding season, because cavity
nesters should be less constrained by nest-site
availability in the nonbreeding season.
P^. Species that use similar nest-site conditions will
show similar patterns of abundance between burned and unbumed forests.
P^. The most abundant cavity-nesting species in bumed
habitat will nest in trees most similar to those most available in burned habitat.
Hg: Birds respond primarily to changes in food availability.
P^. Abundance patterns between the two habitats will be
more similar for species with like foraging habits than for species with different foraging habits. Pg. Removal of most trees from post-fire habitat will
affect abundances of tree-foraging species only,
provided suitable trees remain for nesting.
These two hypotheses are not mutually exclusive, and failure to disprove either may indicate that both factors are important.
Study Area
In 1988, the Red Bench Fire burned approximately 15,000
hectares in northwestern Montana (Fig. 1). The fire started
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 13
in the Flathead National Forest (FNF) and burned 4000
hectares. It then crossed the North Fork of the Flathead
River into adjacent Glacier National Park (GNP) where 11,000 hectares burned. Much of the burned area consisted of
extensive stands of lodgepole pine {Pinus contorta var.
latifolia) that originated in 1910 and previously were killed
by mountain pine beetle {Dendroctonus ponderosae) . Interspersed with these single-aged lodgepole stands were
various-aged stands containing Douglas-fir {Pseudotsuga menziesii), Engelmann spruce (Picea engelmaimii), subalpine fir {Abies lasiocarpa), western larch, ponderosa pine, black
cottonwood {Populus trichocarpa), trembling aspen {Populus
tremuloides) , paper birch {Betula papyxifera) , and some riparian and grassland communities.
Past fire regimes greatly influenced vegetation in the
North Fork drainage. Fire chronologies beginning in 1650
identified a mixed-severity fire regime ranging from
underbums to stand-replacing crown fires (Barrett et al. 1991). Mean fire intervals were 25-75 years. This regime produced a mosaic of young and old stands as well as a range
of single- to multiple-aged stands. Fire frequency in this
area apparently has been altered by suppression efforts. The estimated mean number of fire years (years in which one or
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 14
Red Bench Fire Study Area
tu rn e d
U n b u m e d
Glacier National Park
Flathead National Forest
10 km
Montana
Figure 1. Red Bench Study Area in northwestern Montana.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 15
more fires occurred) declined from 20 per century between 1650 and 1935 to 2 per century between 1935 and present (Barrett et al. 1991). The Red Bench Fire severity ranged from unbumed
inclusions to severe crown b u m (Key 1988) . Low severity
areas were left with many live standing trees, especially of
more fire-resistant, thick-bar)ced conifers (Douglas-fir,
western larch, and ponderosa pine). In areas of moderate to
high severity, most trees were killed and left as snags
except for some large, thick-barked individuals (Dutton and
Cooper 1988) . Approximately 75% of the area experienced crown fire; in over half of that area, the fire was intense enough to consume needles and smaller branches, killing all trees (Key 1988).
In the GNP portion of the bum, most forest stands were
not altered by human activity. In contrast, extensive and
intensive timber harvest occurred in the adjacent FNF fire area. Timber was removed in pre- and post-fire commercial
harvesting and as a perceived safety hazard during and after
the fire. This left a mosaic of burned and uncut, burned and partially cut, and burned and clearcut stands. Methods
Nest surveys
I searched for active cavity nests along random
transects throughout the GNP part of the Red Bench Burn
during the nesting seasons of 1990-1994. I established
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 16
transects throughout the study area by choosing random starting points within the perimeter of the bum. Randomly
selected compass bearings were extended from those points to
the edge of the bum. I surveyed 12 variable-length transects, totalling 42,000 meters (the maximum coverage that time allowed) . Transects were walked slowly, with brief
stops at 150-m intervals. I walked each transect once
between early June and mid-July, when most cavity nesters
were in nestling stage and nests were consistently easiest to locate. Nests were located by listening for young calling from nests and by following adults to nests.
Random transects were established in the FNF portion of
the b u m in 1992. I selected an area that was homogeneous in
aspect, elevation, and slope to control for factors other
than forest-stand conditions. FNF transects totalled 16,000 meters in length. Nests were located in the breeding seasons
of 1992-1994, using the same methods described for the GNP
transects. Different observers searched for nests each year. I attempted to reduce observer bias by conducting a training period at the beginning of each nesting season.
Nest occurrences in this study were compared with data
collected in unbumed areas in FNF and GNP by McClelland
(1977). In both studies, active cavity nests were located
using the same methods (visual and auditory clues) and equal
efforts made to locate nests of all species. I estimated the
relative abundance (percentage of the total nests) of each
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 17
species in each habitat. I calculated the ratio of the relative abundance of a species in burned forest to that in unbumed forest, as an index of a species response to the
effects of fire. Values less than 1 (indicating a negative
response) were standardized by dividing them into -1. Final
values potentially ranged between zero and positive infinity
for species that increased, and zero and negative infinity for species that declined in relative abundance. I used data
from only those species for which 5 or more nests were found
to calculate this fire response index.
Differences between the two communities could have been due to factors other than fire. The study areas overlapped
but probably differed in overall forest composition, and the
surveys occurred 15 years apart. Therefore, I reviewed data
from 12 other studies that compared bird abundances in burned
and unbumed habitats, and conducted a meta-analysis to test whether my results were consistent with others. For each species, I recorded the number of studies (out of 4 or more) in which the species showed the same response to fire
(positive or negative) as I had found. I used chi-square to
test whether the percent of studies that reported a species
to be more abundant in post-fire habitat was significantly different from 50% (the expected value if a species was unaffected by fire).
Microhabitat characteristics
At each nest, I recorded characteristics of the nest
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 18
cavity and tree (Table 1) . I used standardized mean values
of nest variables to calculate squared Euclidean distances between each pair of bird species. This resulted in a matrix
of distances between each pairwise combination of species. Pairs of species whose nest site characteristics were similar
would be "closer" than pairs whose nest sites were more
different. I calculated an analogous distance matrix using the fire response index, as described above. A correlation
coefficient between these two distance matrices was computed. Because data points were not independent (i.e., data from each species is used 12 times in constructing each distance
matrix), I used the program RT (Manly 1992) to perform a
Mantel randomization test on the correlation. This test
determines the significance of a correlation between matrices by randomly reallocating the order of the data points of one
of the matrices. It compares correlations between reordered
matrices to the correlation of the original matrices (Manly 1991).
To randomly select trees for comparison with nest trees, I used a random starting point on each nest-search transect.
At 250-m intervals, I found the tree nearest to the transect,
until 150 trees were sampled. The same variables were
recorded for random trees as were recorded for nest trees. I
used mean values of tree variables to calculate the Euclidean
distance between nest trees and random trees for each bird
species. I used Pearson correlation analysis to test for a
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 19
Table 1. Variables recorded at nest trees and random trees in post-fire habitat.
Variable Method Diameter at breast height (dbh) Diameter tape Height relascope % bark estimated ° lean compass Height of cavity relascope Condition (live/dead, top intact/broken) Tree species
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 20
linear relationship between the numerical fire response index
and use of trees similar to random b u m trees. Species were assigned to 1 of 4 foraging categories
based on life-history information and foraging techniques
(Bent 1939, Ehrlich et al. 1988, pers. obs., Table 2). Each
pairwise combination of species was then categorized as a pair whose members were in either the same or different foraging groups. I used Mann-Whitney U test to determine
whether the mean distance based on response to fire was less
for species pairs in the same foraging categories than for
species pairs in different categories. Bird censuses
I set up survey points in 1992 in areas on both sides of
the b u m perimeter in GNP. Ten random points at least 250-m
apart were established in each of the following forest types:
1) unbumed, 80-yr-old lodgepole pine; 2) burned, 80-yr-old lodgepole pine; 3) unbumed mature mixed-species, and 4)
burned mature mixed-species. Mature, mixed-species stands
originated before 1844 (Key, pers. comm.) and were dominated
by Douglas-fir and western larch. Paired burned and unburned
areas were similar in elevation, slope, aspect, and pre-bum
forest stand type and age. Burned areas all had undergone crown fire with no surviving trees.
I censused cavity nesters from 16 July to 10 October
1992-1994, using the fixed-radius point count method (Hutto
et al. 1986) . Counts were 10 minutes long, and all
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 21
Table 2. Bird species and foraging group assignments.
Primary foraging Method Bird Species Timber Drill Downy Woodpecker Picoides puhescens Hairy Woodpecker Picoides villosus Three-toed Woodpecker Picoides tridactylus Black-backed Woodpecker Picoides arcticus Pileated Woodpecker Dryocopus pileatus
Bark/Foliage Glean Red-naped Sapsucker Sphyrapicus nuchalis Black-capped Chickadee Parus atricapillus Mountain Chickadee Parus gambeli Red-breasted Nuthatch Sit ta canadensis Ground Capture American Kestrel Falco sparverius Northern Flicker Colaptes auratus House Wren Troglodytes aedon Aerial Capture Tree Swallow Tachycineta bicolor Mountain Bluebird Sialia currucoides
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 22
cavity-nesting birds detected within 50 m were recorded (this
limit was set to reduce potential biases due to differences
in detectability) . Paired areas were sampled on consecutive
days, 5 times each year. Multivariate ANOVA was used to test
whether species differed in abundances between the two habitats, with year and forest type as covariates. I used
these data to compute a percent similarity index for burned
and unbumed forest communities in the nonbreeding season.
To calculate the similarity index, I computed the relative abundance of each bird species in each of the two habitats.
The lower of these two values for each species were summed to
give the index. In a concurrent project, breeding-season point counts were conducted in 58 burned and 39 nearby
unbumed plots (Dumin 1992) . With these data I computed a similarity index for the breeding season.
To determine the effects of tree removal in post-fire habitat, point counts were conducted in the FNF portion of the bum, in forest stands subjected to different cutting
intensities. In some stands, individual tree selection left
most trees standing (partial cuts), whereas other stands were
clearcut, with only a few snags left to serve as potential
nest trees. Random points were established in both types of cutting units; points served as the centers of 10-minute,
50-m-radius counts. Forty-five counts were conducted in
clearcut stands and 120 in partially cut stands. I used
Mann-Whitney U tests to compare mean abundances of
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 23
tree-foraging and non-tree-foraging species in partial cuts
to abundances in clearcuts.
Results Five hundred and fifty-one cavity nests of 21 bird
species were located in the study area during 1990-1994
(Table 3). Relative nest abundances of species differed
significantly between this study and McClelland's (1977)study
of u n b u m e d habitat (%^ = 265, df = 12, p < 0.001; Fig. 2) .
Some species showed dramatically different abundances in the
two habitats. For example. Mountain Bluebirds made up only
1% of nests in unbumed forest, compared to 12% of the nests
in the post-fire forest. Conversely, while 43% of nests in
the unbumed forest were Red-naped Sapsucker nests, only 12%
of nests in the unbumed forest belonged to this species.
Differences between the habitats were consistent with those
detected in other studies. For each of 13 species, I found
between 4 and 11 studies that compared abundances in burned and unburned forests (Table 4). For all species, the results from a statistically significant proportion of other
published studies were consistent with results from my study.
For 8 species, all studies showed the same response to fire,
and for the other 5 species, only a single study reported a trend opposite that of all other studies.
The postbreeding point count data also showed
significantly different species abundances between burned and
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 24
Table 3. Common name, scientific name, and number of
nests found within the Red Bench Fire study area for
each of the 21 cavity-nesting species encountered.
Common Name Scientific Name # Nests Tree Swallow Tachycineta bicolor 129 Northern Flicker Colaptes auratus 95 Red-naped Sapsucker Sphyrapicus nuchalis 67 Mountain Bluebird Sialia currucoides 62 Three-toed Woodpecker Picoides tridactylus 48 Hairy Woodpecker Picoides villosus 38 Red-breasted Nuthatch Sit ta canadensis 25 Downy Woodpecker Picoides pubescens 22 American Kestrel Falco spazverius 16 Black-backed Woodpecker Picoides arcticus 12 Mountain chickadee Parus gambeli 7 House Wren Troglodytes aedon 6 European Starling Stumus vulgaris 5 Pileated Woodpecker Dryocopus pileatus 5 Black-capped Chickadee Parus atricapillus 3 Bufflehead Bucephala albeola 4 Lewis' Woodpecker Melanerpes lewis 2 Brown Creeper Certhia americana 2 Common Merganser Mergus merganser 1 Northern Saw-whet Owl Aegolius acadicus 1 Chestnut-backed Parus rufescens 1 Chickadee Total 551
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 25
20 30 40 SO - 1 Mountain Bluebird
Black-backed Woodpecker Unburned forest Burned forest (n=303) (n=551)
Three-toed Woodpecker
Downy Woodpecker
Tree SwallowH
Bouse Wren
Hairy Woodpecker- r Northern FlickerH
American Kestrel
Red-breasted NuthatchH
Pileated Woodpecker
Red-naped Sapsucker-
Chickadee spp. *i
1 I I I I 10 20 30 40 SO % of nests Figure 2. Relative abundances of cavity nests in burned and unbumed forests in and around Glacier National Park, Montana. Species are arranged in descending order based on the ratio of nests in burned to unbumed forest. Nest distributions of species differed significantly (chi-square test, p < 0.05) between the two habitats.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 26
Table 4. Change in relative nest abundance from unburned to burned forest in the current study, number
of studies that reported increases and decreases in
abundances, and p-value of chi-square tests of
response distributions.
Bird Species Response* + /-b P American Kestrel + 4/0 0.046
Red-naped Sapsucker^ - 0/6 0.014 Downy Woodpecker + 6/1 0.059 Hairy Woodpecker + 11/0 0.0009 Three-toed Woodpecker + 8/0 0.0047 Black-backed Woodpecker + 9/1 0.011 Northern Flicker + 11/0 0.0009 Tree Swallow + 7/1 0.034
Mountain Chickadee - 1/7 0.019
Black-capped Chickadee - 0/6 0.014
Red-breasted Nuthatch - 1/7 0.034 House Wren + 7/0 0.0082 Mountain Bluebird + 11/0 0.0009 a Response to fire (in this study) as indicated by change in relative abundance of nests. b # of studies reporting positive response to fire/# of studies reporting negative response to fire, as indicated by differences in abundances. Data from Koplin (1969), Roppe (1974), Davis (1976), Pfister (1980), Taylor and Barmore (1980), Harris (1982), Raphael and White (1984), Raphael et al. (1987), Breininger and Smith (1992), Skinner (1989), Dumin (1992), Sallabanks (1995), and postbreeding censuses in this study. c Includes Yellow-bellied and Red-breasted Sapsuckers, which hybridize with Red-naped Sapsuckers and have in the past been considered conspecific.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 27
unburned habitat = 395, df = 12, p < 0.001; Fig. 3) .
Trends were identical with trends shown by nest abundance
data. Cavity nesters as a whole were more abundant in the
burned forest (mean number per point = 1.88 ± 0.222) than in
the unbumed forest (mean = 1.32 ± 0.112; p = 0.025). Four species were significantly more abundant in the burned forest
and two were significantly more abundant in the unbumed forest (Fig. 3).
The bird communities in burned eind unbumed forests were more similar to one another in the breeding season
(similarity index = .4739) than in the postbreeding season (similarity index = .2635). Pairs of species that used
similar nest-site conditions did not show similar patterns of abundance (based on the fire response index) between the two
habitats (Mantel randomization test, r = 0.059, p = 0.289).
Additionally, species with positive numerical response to fire did not nest in sites similar to randomly available
sites in the burned forest (Pearson correlation, r = 0.001, p = 0.339).
Pairs of species drawn from the same foraging guild were significantly more similar in abundance pattems (mean
Euclidean distance = 23.89) than pairs drawn from different guilds (mean Euclidean distance = 53.69; p = 0.005). Point
count data indicated consistent responses within foraging
groups (Fig. 3). Timber drillers, aerial insect foragers
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 28
0.25 0.75 1.25
Mountain Bluebird
Black-backed Woodpecker
Q Unburned forest American Kestrel I Burned forest
House Wren
Hairy Woodpecker
Downy Woodpecker
Tree Swallow
Three-toed Woodpecker
Northern Flicker
Chickadee spp. -i * r - Red-breasted Nuthatchi *
Red-naped Sapsucker
0.25 0.5 0.75 1 1.25 Average # per point Figure 3. Bird abundances in 50-m-radius plots in recently burned and unburned forests of Glacier National Park, Montana. Species are arranged in descending order based on the ratio of abundance in burned to unbumed habitat. Data are from 180 point counts in each habitat. Species indicated by * differed significantly (ANOVA, p < 0.05) between the two habitats.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 29
(including those flycatching and hawking), and ground foragers were more abundant in burned habitat; however,
foliage- and bark-gleaning species were less abundant
therein.
Relative nest abundances differed significantly between
partially cut and clearcut stands with leave trees (%^ =19.8
df = 11, p < 0.05). Nests of most species, including most
timber drillers, were found only in areas where partial
cutting occurred (Fig. 4). A similar trend was revealed from
the point count data. Non-tree-foraging species as a group were equally abundant in the two types of treatments, whereas
tree-foraging species were less abundant in clearcuts (mean
per point = .244 ± .0722) than in partial cuts (mean per
point = .776 ± .073, p = 0.001). These differences were
probably minimized, since partially cut stands had areas within them that were effectively clearcut. Additionally,
all tree-foraging species were less abundant in clearcuts,
but abundance pattems of non-tree-foraging species were inconsistent.
Discussion
Previous studies have shown cavity-nesting birds to be more abundant in recently burned forests than in unbumed
forests (Hutto 1995a). These differences have been
attributed largely to the creation of nest sites (Taylor and
Barmore 1980, Raphael and White 1984, Sallabanks 1995). In
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30
0.25 0.5 0.75 I 1- __L. Three-toed Woodpecker 4
Red-naped Sapsucker-
S Partial Cut | Clearcut Red-breasted Nuthatch ss
American Kestrel a
Black-backed Woodpecker
Downy Woodpecker
Chestnut-backed Chickadee
Mountain Chickadee
Mountain Bluebird
Hairy Woodpecker 4
Northern Flicker A Tree Swallow F , r- T 0 0.25 0.5 0,75 No. nests/km transect Figure 4. Active cavity nest abundances in areas of different logging intensities in a post-fire forest. Species are arranged in descending order based on the ratio of nest abundances in partial cuts to clearcuts. The difference in species distributions between the two areas is significant (chi-square, p < 0.05). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 31 this study, fire affected cavity-nesting bird species in disparate ways, thereby altering the structure of the cavity- nesting bird community. Some species were very rare in one of the two habitats. Fire, therefore, appears to be important in maintaining diversity, at least on a regional scale. Black-backed Woodpeckers, Three-toed Woodpeckers, Downy Woodpeckers, Hairy Woodpeckers, and Mountain Bluebirds were much more abundant in the burned forest. These species all likely have declined in past decades (Hejl 1994); Black- backed, Three-toed, and Downy woodpeckers are particularly uncommon in this region (Hutto 1995b). Historic fire regimes produced mosaics of forest stands in various stages of post-fire succession. Mid-elevation forests in the northern Rockies frequently experienced high- intensity, stand-replacement fires (Amo 1980, Fischer and Bradley 1987) that left a large proportion of this region in stages of early post-fire succession (Gmell 1980) . Large, stand-rep lacement fires were not uncommon in the North Fork drainage (Barrett et al. 1991). The Red Bench Fire was within the range of normal intensity and size for naturally- occurring fires in this area and forest type. Alterations of natural regimes through suppression may have serious consequences for species such as Black-backed Woodpeckers, which are rare in areas without burned forest stands (Hutto 1995a, this study). Because woodpeckers provide nest cavities for species unable to excavate their own holes. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 32 factors that contribute to their decline on the landscape ultimately may affect other members of the community as well (Pfister 1980) . I was not able to directly measure fire's effects with pre- and post-fire data. Microhabitat differences (unrelated to the fire) between burned and unbumed forest sites may have influenced results. I attempted to reduce this potential problem by using sites in close proximity and in similar forest types. There is some indication, however, that trends I observed were typical across a broad geographical range (Table 4). The close correspondence of trends from nest abundance and point count data, and the agreement of trends among several studies support the conclusion that observed pattems were due largely to the effects of fire. Nonetheless, a planned comparison of pre- and post-fire data with controls is needed to corroborate the conclusions of this and other studies. Additionally, effects may vary in forest systems with different fire regime, different bird species, and different insect prey species. Results from this study should not be extrapolated too liberally beyond the mid-elevation forests of this region. Differences in abundances of most species were probably due to differences in foraging opportunities rather than differences in nest-site availability. Predictions based on nest-site changes were not supported. Abundance differences between the two habitats were less pronounced in the breeding Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 33 season, when nest sites should be most important, than in the non-breeding season. Species that used similar nesting sites did not respond similarly to fire; nor did species that used the types of nest conditions most abundant in the b u m respond most positively to fire. In support of foraging-site predictions, species with similar foraging habits showed similar pattems of abundances, and removal of fire-created foraging substrates affected only tree-foraging species. Nest-site availability is frequently cited as the primary factor determining cavity-nesting bird distributions (Von Haartman 1957, McLaren 1963, Hilden 1965, Taylor 1973, Mannan et al. 1980, Raphael 1983, Gustafsson 1988). Competition for nest sites has been documented in both natural and experimental settings (Power 1966, Taylor 1973, Troetschler 1970, van Balen 1982, Nilsson 1984, Ingold 1989), suggesting that the cavity-nesting community is shaped largely by the outcomes of nest-site competition. In heavily managed forests, suitable nest trees may be scarce due to removal of snags and other economically undesirable trees (McClelland 1977, McComb et al. 1986, Ingold 1989). In such conditions, nest sites likely are limiting for some species. Because of this information, land-use agencies often make management decisions based on the idea that providing appropriate nesting trees will satisfy the needs of cavity-nesting birds and maintain viable populations of cavity-nesting species in human-altered forests. Such Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 34 options probably are not sufficient for species that forage on trees. Results of this study suggest the cavity-nesting bird community is shaped as much or more by foraging opportunities, especially in forests with naturally occurring densities of snags and large trees (McClelland 1979). Most woodpeclcers are likely to be limited by food, and respond dramatically to increased prey abundance (Blackford 1955, Yeager 1955, Koplin 1969, Lester 1980). Timber-drilling woodpeckers, especially Black-backed and Three-toed woodpeckers, feed predominantly on wood-boring beetle larvae (Beal 1911, Stallcup 1962), which are plentiful in early post-fire forests (Evans 1971, Fellin 1980, Harris 1982). Even in managed forests, tree-foraging species may be affected by loss of foraging sites while potential nest trees are still available. In this study, most tree-foraging species did not nest in clearcuts, although other species nested in the snags that had been left in clearcuts (Chap. Ill). In partially cut or second-growth forests where decayed trees are scarce, this situation is less probable. Nest-site availability may be more critical for species unable to excavate their own holes. Nonexcavating species such as Mountain Bluebirds and Tree Swallows strongly compete for nest holes (Power 1966, Taylor 1973). Fire improves foraging habitat for these species that feed in open areas and are eliminated by canopy closure (Taylor 1973). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 35 Nonetheless, secondary cavity nesters use holes created both by species that decline in post-fire habitat (e.g., Red-naped Sapsuckers) and increase in post-fire habitat (e.g., Northern Flickers). In this study, nonexcavators also used holes that appeared to have been created by the fire (31 or 5.6% of nests were in such cavities). It is unclear whether nest-site availability for non-excavating species differs between unbumed and burned forests. In a closed canopy forest, species such as bluebirds and swallows may be limited by foraging-site rather than nesting-site availability. In meadow, savannah, and edge habitats, where foraging sites are plentiful, nest sites probably are limiting. Post-fire habitat is a unique environment, open enough to provide appropriate foraging habitat, yet with many more nest sites than typically occur in open areas. Because several species use a similar resource (dead or decaying trees) for both nesting and foraging, it is difficult to separate these two factors. Overall, there may be a tendency for excavator species to respond primarily to factors related to foraging, while nonexcavator species respond to nest-site availability. As excavators increase in post-fire habitat, they may augment cavities available for nonexcavators, leading to increases in nonexcavators over time after fire (Taylor and Barmore 1980) . In the Red Bench study area, nonexcavators became relatively more abundant over time, compared to excavating species. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 36 Management decisions that address only nest-site provisions for cavity-nesting birds may fall short of providing for some community members. Although within the same nesting guild and often managed similarly, species respond to different factors. For example, a Tree Swallow that uses trees only for nesting is not likely to represent the needs of a tree-foraging Red-breasted Nuthatch. A single event such as fire has opposite effects even on closely related species that use similar nest trees (e.g., Red-naped Sapsuckers and Downy Woodpeckers). Grouping species for research or management purposes may be inappropriate, given the wide range of resource use within the group (Hutto et al. 1987) . Generalizations from some species or environments may lead to erroneous assumptions about the group as a whole. For example, Raphael (1983) modeled cavity-nesting bird abundances in response to snag densities in post-fire habitat. His model assumed that bird numbers would track the densities of snags of species, sizes, and decay stages appropriate for nest sites. The lack of correspondence he found between his model predictions and actual bird abundances may have been due to birds ' response to other factors such as foraging opportunities, and nest sites were not limited. Many community-level studies emphasize comparisons of species diversity between habitats and measure biological response to disturbance by calculating diversity indices. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 37 However, such indices may conceal or divert attention from important and interesting trends of individual species, and between-patch (regional or beta) diversity. Rare species, influenced by atypical factors, may respond to circumstances very differently than do more common species. Research that addresses trends only above the species level, or that assumes all members of a guild or community respond similarly, may overlook factors important to the least abundant elements of the community. Habitat loss often is cited as the single greatest threat to the survival of species (e.g., Shaffer 1987, Soule and Kohm 1989) . Habitat destruction generally is viewed as removal of native vegetation for resource extraction or development. However, habitat loss can result from the disruption of natural processes (Brussard 1991). When these processes produce unique habitats to which species are adapted, the disruption of these processes may have significant ecological consequences. Public land managers concerned about maintaining natural populations of native species will want to consider the important functions that fire has performed in this ecosystem. Fire produces a unique set of environmental conditions. These conditions in turn produce a distinctive cavity-nesting bird community. Fire opens the forest and creates expansive areas of standing dead trees, favoring several species uncommon in unbumed forests. Manipulating Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 38 habitat to manage for targeted species or groups of species is likely to be ineffective in fire-dominated ecosystems that cover vast areas and have a diversity of microhabitats. Management actions such as providing styrofoam nest trees (Peterson and Grubb 1983) are often undesirable treatments of symptoms and may create more problems than they solve. Allowing natural processes like fire to occur may be the only way to sustain intact communities of cavity-nesting birds on the landscape. Literature Cited Agee, J. K. 1993. Fire ecology of Pacific Northwest forests. Island Press, Washington, D. C. Amo, S. F. 1980. Forest fire history in the Northern Rockies. J. For. 78:460-465. Barrett, S. W. , S. F. Amo, and C. H. Key. 1991. Fire regimes of western larch-lodgepole pine forests in Glacier National Park, Montana. Can. J. For. Res. 21:1711-1720. Basham, J. T. 1958. Decay of trembling aspen. Can. J. Botany 36:491-505. Beal, F. E. L. 1911. Food of the woodpeckers of the United States. U.S. Dept. Agri. Biol. Surv. Bull. 37. Bent, A. C. 1939. Life histories of North American Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 39 woodpeckers. U.S. Nat. Mus. Bull. 174. Washington, D.C. Blackford, J. L. 1955. Woodpecker concentration in a burned forest. Condor 57:28-30. Bock, C. E., M. Raphael, and J. H, Bock. 1978. Changing avian community structure during early post-fire succession in the Sierra Nevada. Wilson Bull. 90(1): 119-123. Breininger, D. R., and R. B. Smith. 1992. Relationships between fire and bird density in coastal scrub and slash pine flatwoods in Florida. Am. Midi. Nat. 127:233-240. Brussard, P. F. 1991. The role of ecology in biological conservation. Ecol. Appl. 1(1): 6-12. Clements, F. E. 1916. Plant succession: An analysis of the development of vegetation. Carnegie Inst. Washington Publ. No. 242. Cline, S. P., A. B. Berg, and H. M. Wight. 1980. Snag characteristics and dynamics in Douglas-fir forests. Western Oregon. J. Wildl. Manage. 44(4) :773-786. Conner, R. N. 1978. Snag management for cavity-nesting birds. Pages 120-128 in Management of Southern Forests for Nongame Birds, Workshop Proc. U.S.D.A. For. Serv. Gen. Tech. Rep.(SE-14) . Conner, R. N. , 0. K. Miller, and C. S. Adkisson. 1976. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 40 Woodpecker dependence on trees infected by fungal heart rots. Wilson Bull. 88(4): 575-581. Davis, P. R. 1976. Response of vertebrate fauna to forestfire and clear-cutting in south central Wyoming. Ph.D. diss., Univ. of Wyoming, Laramie. Denslow, J. S. 1985. Disturbance-mediated coexistence of species. Pages 307-323 in The Ecology of Natural Disturbance and Patch Dynamics (S.T.A. Pickett and P.S. White, eds.). Academic Press, San Diego. Dumin, A. M. 1992. Avian communities of the Red Bench Fire and adjacent areas in Glacier National Park. Unpub. rep., National Park Sevice, Glacier National Park, Mont. Dutton, B. L ., and S. B. Cooper. 1988. Red Bench Fire intensity mapping and permanent plot installations. Unpub. rep., Dutton Resource Consulting, Missoula, Mont. Ehrlich, P. R., D. S. Dobkin, and D. Wheye. 1988. The birder's handbook: A field guide to the natural history of North American birds. Simon and Schuster, New York. Etheridge, D. E. 1961. Factors affecting branch infection in aspen. Can. J. Botany 39:799-816. Evans, W. G. 1971. The attraction of insects to forest fires. Jn Tall Timbers Conf. on Ecol. Animal Control by Habitat Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 41 Manage. Proc. 3:115-127. Fellin, F . G. 1980. A review of some interactions between harvesting, residue management, fire, and forest insects and diseases. Pages 335-414 in Symposium Proceedings : Environmental consequences of timber harvesting in Rocky Mountain coniferous forests. Gen. Tech. Rep. INT-90, U.S. Forest Service, Ogden, Utah. Fischer, W. C., and A. F. Bradley. 1987. Fire ecology of western Montana forest habitat types. General technical report INT-223. U.S. Forest Service, Ogden, Utah. Fumiss, R. L., and V. M. Carolin. 1977. Western forest insects. Misc. Publ. 1339. U.S. Forest Service, Washington, D.C. Gruel 1, G. E. 1983. Fire and vegetative trends in the Northern Rockies: interpretations from 1871-1982 photographs. General technical report INT-150. U.S. Forest Service, Ogden, Utah. Grubb, P. J. 1977. The maintenance of species richness in plant communities : the importance of the regeneration niche. Biol. Rev. Cambridge Philos. Soc. 52:107-145. Gustafsson, L. 1988. Inter- and intraspecific competition for nest holes in a population of the collared flycatcher Ficedula albicollis. Ibis 130:11-16. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 42 Habeck, J. R., and R. W. Mutch. 1973. Fire-dependent forests in the Northern RoclQr Mountains. J. Quaternary Res. 3(3):408-424. Harris, M. A. 1982. Habitat use among woodpeckers in forest burns. M. S. Thesis, Univ. of Montana, Missoula. Heinselman, M. L. 1971. The natural role of fire in northern conifer forests. Pages 61-72 in Fire in the Northern Environment Symposium Proceedings. College, Alaska. Hejl, S. J. 1994. Human-induced changes in bird populations in coniferous forests in western North America during the past 100 years. Studies in Avian Biology No. 15:232-246. Hilden, 0. 1965. Habitat selection in birds. Ann. Zool. Fenn. 2:53-75. Horton, S. P., and R. W. Mannan. 1988. Effects of prescribed fire on snags and cavity-nesting birds in southeastern Arizona pine forests. Wildl. Soc. Bull. 16:37-44. Hutto, R. L. 1995. Composition of bird communities following stand-replacement fires in Northern Rocky Mountain (U.S.A.) conifer forests. Conservation Biology 9(5):1041-1058. Hutto, R. L. 1995. Distribution and habitat relationships. U.S. Forest Service Northern Region Songbird Monitoring Program Second Report. U.S. Forest Service, Missoula, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 43 Montana. Hutto, R. L., S. M. Pletschet, and P. Hendricks. 1986. A fixed-radius point count method for nonbreeding and breeding season use. Auk 103:593-602. Hutto, R. L . , S. Reel, and P. B. Landres. 1987. A critical evaluation of the species approach to biological conservation. Endangered Species Update 4(12):1-4. Univ. Of Michigan, Ann Arbor, Michigan. Ingold, D. J. 1989. Nesting phenology and competition for nest sites among Red-headed and Red-bellied Woodpeckers and European Starlings. Auk 106:209-217. Key, C. 1988. Red Bench Fire Intensity Map. Unpub. rep., Glacier National Park, Mont. Kimmey, J. W. 1955. Rate of deterioration of fire-killed timber in California. U.S. Dept. Agric. Circ. 962:1-22. Koplin, J. R. 1969. The numerical response of woodpeckers to insect prey in a subalpine forest in Colorado. Condor 71:436-438. Lester, A. N. 1980. Numerical response of woodpeckers and their effect on mortality of mountain pine beetles in lodgepole pine in northwestern Montana. M.S. Thesis, Univ. of Montana, Missoula, 103 pp. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 44 Mcinly, B. F. J. 1991. Randomization and Monte Carlo methods in biology. Chapman and Hall, London. Manly, B. F. J. 1992. RT, Version 1.02. Centre for Applications of Statistics and Mathematics, Univ. Of Otago. Mannan, R. W., E. C. Meslow, and H. M. Wight. 1980. Use of snags by birds in Douglas-fir forests, western Oregon. J. Wildl. Manage. 44(4):787-797. May, R. M. 1973. Stability and complexity in model ecosystems. Princeton Univ. Press, Princeton. 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. diss., Univ. of Montana, Missoula, 492 pp. McClelland, B. R. , S. S. Frissell, W. C. Fischer, and C. H. Halvorson. 1979. Habitat management for hole-nesting birds in forests of western larch and Douglas-fir. J. Forestry 77:480-483. McComb, W. C . , S.A. Bonney, R.M. Sheffield, and N.D. Cost. 1986. Snag resources in Florida— are they sufficient for average populations of primairy cavity-nesters? Wildl. Soc. Bull. 14:40-48. McLaren, W. D. 1963. A preliminary study of nest-site Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 45 competition in a group of hole-nesting birds. M.S. Thesis, Univ. of British Columbia, Vancouver. Neimi, G. J. 1978. Breeding birds of burned and unbumed areas in northern Minnesota. Loon 50:73-84. Nilsson, S. G. 1984. The evolution of nest-site selection among hole-nesting birds: the importance of nest predation and competition. Omis Scandinavica 15:167-175. Peterson, A. W. , and T. C. Grubb. 1983. Artificial trees as a cavity substrate for woodpeckers. J. Wildl. Manage. 47(3):790-798. Pfister, A. R. 1980. Postfire avian ecology in Yellowstone National Park. M. S. Thesis, Univ. of Washington, Pullman. Pickett, S. T. A., and P. S. White. 1985. Natural disturbance and patch dynamics: an introduction. Pages 3-13 in The Ecology of Natural Disturbance and Patch Dynamics (S.T.A. Pickett and P.S. White, eds.). Academic Press, San Diego. Power, H. W. 1966. Biology of the Mountain Bluebird in Montana. Condor 68:351-371. Raphael, M. G. 1983. Cavity-nesting bird response to declining snags on a burned forest: a simulation model. Pages 211-215 in Proceedings of Snag Habitat Management Symposium at Northern Arizona Univ., Flagstaff, AZ. (J.W. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 46 Davis, G.A. Goodwin, and R.A. Ockenfels, Technical coordinators) U.S. For. Serv. Rocl^y Mtn. For. Range Exp. Sta. Gen. Tech. Rep. RM-99. Raphael, M. G., and M. White. 1984. Use of snags by cavity- nesting birds in the Sierra Nevada. Wildl. Monog. 86:1-66. Raphael, M. G., M. L. Morrison, and M. P. Yoder-Williams. 1987. Breeding bird populations during twenty-five years of postfire succession in the Sierra Nevada. Condor 89:614- 626. Riley, C. G. 1952. Studies in forest pathology. IX. Fames igniarius decay of poplar. Can. J. Botany 30:710-734. Roppe, J. A. 1974. Effects on wildlife of a fire in a lodgepole pine forest. M.S. Thesis, Colorado State Univ., Ft. Collins. Sallabanks, R. 1995. Effects of wildfire on breeding bird communities in coniferous forests of northeastern Oregon. Annual report to Blue Mountains Natural Resources Institute. Shaffer, M. 1987. Minimum viable populations: coping with uncertainty. Pages 69-86 in Viable populations for conservation (M. Soule, ed.). Cambridge Univ. Press, Cambridge. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 47 Shigo, A. L., and L. Kilham. 1968. Sapsuckers and Fames igniarius var. Populinus. U. S. For. Serv. Res. Note NE-84. NE Forest Exp. Sta., Upper Darby, Pa. Skinner, N. G. 1989. Seasonal avifauna use of burned and unbumed lodgepole pine forest ecotones. M.S. Thesis, Univ. of Montana, Missoula. Slagsvold, T. 1978. Con^etition between the great tit (Parus majar) and the pied flycatcher (Ficedula hypoleuca) : an experiment. Omis. Scand. 9:46-50. Soule, M. E. , and K. A. Kohm. 1989. Research Priorities for Conservation Biology. Island Press, Washington, D.C. Stallcup, P. L. 1962. Contents of 103 stomachs of Northem Three-toed Woodpeckers. J. Co lor ado-Wyoming Acad. Sci. 5:39. Taylor, D. L. 1973. Competition between Mountain Bluebirds and Tree Swallows in post-fire areas of Yellowstone National Park. Natl. Geogr. Soc. Res. Rep. Taylor, D. L., and J. W. J. Barmore. 1980. Post-fire succession of avifauna in coniferous forests of Yellowstone and Grand Teton National Parks, Wyoming. Pages 130-145 in Management of westem forests and grasslands for nongame birds. U.S.D.A. For. Ser. Gen. Tech. Rep. INT-86. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 48 Thiery, R. G. 1982. Environmental instability and community diversity. Biol. Rev. Cambridge Philos. Soc. 91:691-710. Troetschler, R. G. 1970. The impact of the Starling (Stumus vulgaris) on the native hole nesting species of the San Francisco Bay region. M.S. Thesis, San Jose State Univ., San Jose, California. van Balen, J. H., C. J. H. Booy, J. A. van Franeker, and E. R Odieck. 1982. Studies on hole-nesting birds in natural nest sites: 1. availability and occupation of natural nest sites. Ardea 70:1-24. Von Haartman, L. 1957. Adaptation in hole-nesting birds. Evolution 11:339-347. Waters, J. R., B. R. Noon, and J. Verner. 1990. Lack of nest-site limitation in a cavity-nesting bird community. J. Wildl. Manage. 54:239-245. Welsh, C. J., and D. E. Capen. 1992. Availability of nesting sites as a limit to woodpecker populations. For. Ecol. and Mgmt. 48:31-41. Wesolowski, T. 1989. Nest-sites of hole-nesters in a primaeval temperate forest (Bialowieza National Park, Poland). Acta Omith. 25 (3 ): 321-351. Wiens, J. A. 1985. Vertebrate responses to environmental Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 49 patchiness in arid and semiarid ecosystems. Pages 169-193 in The Ecology of Natural Disturbance and Patch Dynamics (S.T.A. Pickett and P.S. White, eds.). Academic Press, San Diego. Wintemitz, B. L., and H. Cahn. 1983. Nestholes in live and dead aspen. Pages 102-106 in Proceedings of Snag Habitat Management Symposium at Northem Arizona Univ., Flagstaff, AZ. (J.W. Davis, G.A. Goodwin, and R.A. Ockenfels, Technical coordinators) U.S. For. Serv. RocJcy Mtn. For. Range Exp. Sta. Gen. Tech. Rep. RM-99. Yeager, L. E. 1955. Two woodpecker populations in relation to environmental change. Condor 57:148-153. Zamowitz, J. E., and D.A. Manuwal. 1985. The effects of forest management on cavity-nes ting birds in northwestem Washington. J. of Wildl. Manage. 49(1):255-263. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter III USE OF AN EARLY POST-FIRE FOREST BY CAVITY-NESTING BIRDS: IMPORTANCE OF NESTING AND FORAGING OPPORTUNITIES Introduction Determining effects of habitat variation on distributions of organisms has long engaged ecologists. For some species, presence of an appropriate nesting or denning site appears to influence distribution patterns (Brush 1983, Steele 1992). For others, there is strong empirical evidence for close tracking of food (e.g., Hutto 1985) . Which factors are deemed important may vary with the scale at which they are investigated. For example, Orians and Wittenberger (1991) found that at larger scales. Yellow-headed Blackbirds settled disproportionately in marshes that had more of their primary prey. However, within the marsh, nests were not located in areas of higher food availability. Vegetation density and other variables (presumably associated with higher nest success) appeared to determine nest locations. Nest-site selection has been the focus of studies of habitat use in cavity-nesting birds. Excavating species make cavities in trees of adequate diameters with relatively soft (often decayed) heartwood (Shigo and Kilham 1968, Conner et al. 1976, McClelland 1977, Mannan et al. 1980, Raphael and White 1984). Nonexcavating species depend on suitable holes, 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 51 usually excavated by primary cavity nesters (Raphael and White 1984) . Because of these relatively nsirrow criteria, nest-site availability is generally thought to determine distributions of cavity-nesting birds (Conner 1978, Scott et al. 1980, Raphael and White 1984, Zamowitz and Manuwal 1985, Gustafsson 1988, Li and Martin 1991, Gibbs et al. 1993; but see Waters et al. 1990, Welsh and Capen 1992). Studies of habitat used by cavity-nesting birds have found nest tree characteristics to be inç)ortant in predicting tree use (McLaren 1963, Mannan et al. 1980, van Balen et al. 1982, Wesolowski 1989, Schreiber and deCalesta 1992). Attributes of areas around nest trees were found to be meaningful as well, perhaps due to their influence on foraging opportunities (Conner et al. 1975, McClelland 1977, Brush 1983, Korol and Hutto 1984, Raphael and White 1984, Peterson and Gauthier 1985, Swallow et al. 1986, Bull 1987, Li and Martin 1991). In temperate regions, most cavity-nesting bird species are more abundant in recently bumed forests than in unburned forests (Bock and Lynch 1970, Pfister 1980, Taylor and Barmore 1980, Hutto 1995, Chap. II). However, little is known about microhabitat use by cavity-nesting birds within the post-fire environment (but see Harris 1982, Hitchcox in prep). Fire creates a distinct environment not reproduced by other events, including human activities such as logging (Hutto 1995). Differences in topography, fire intensity, and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 52 pre-fire forest conditions may influence suitability of post-fire habitat to cavity nesters. Fire probably affects both nesting and foraging factors by generating snags, eliminating foliage, altering plant species composition, and modifying tree age-class distributions. It is unclear, therefore, whether nest-site or food availability is of primary importance in determining cavity nester distributions in this unique habitat. In the past few decades, natural resource agencies have begun to recognize the natural role of fire. Policy changes that encourage increased natural and management-ignited fires reflect that recognition (Heinselman 1970, Laverty et al. 1993) . However, natural fire regimes are often difficult to determine, and numerous considerations act to constrain fire's return to public lands. Information on the importance of the context of fire can facilitate management decisions when natural fire regimes are impractical. Research on cavity nesters in burned areas can provide information on the use of a potentially significant but little-studied habitat. I studied foraging and nesting habitat used by cavity-nesting birds from 2 to 6 years after a 1988 fire in a northwestern Montana forest. The primary objective was to assess the influence of food and nest-site availability in bird distribution patterns within burned forests. Study Site The study area was within the 11,000 hectares burned by Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 53 the Red Bench Fire in 1988 in Glacier National Park (GNP) , Montana. Previous to the fire, this area primarily consisted of mid-elevation coniferous forests dominated by lodgepole pine (Pinus contorta var. latifolia) , westem larch (Larix occidentalis), and Douglas-fir (Pseudotsuga menziesii). Over 70% of the area bumed at stand-replacement intensity (Key 1988) . Large swaths of trees were removed near roads and buildings; however, most of the area was left in a natural post-fire condition (see chap. II for a map and complete description of the study area). Methods Foraging habitat In 1990 I sampled trees in the Red Bench B u m for evidence of woodpecker foraging. I examined woodpecker foraging at 2 sites, each of which encompassed all b u m severities. The sites differed in pre-bum stand ages and types. Site MC was a mature (pre-1844 origin) mixed-conifer forest dominated by westem larch. Site LP was a single-aged stand dating from a b u m in 1910. It primarily included lodgepole pines that were previously killed from beetle outbreaks. At each site, I randomly selected a transect starting point and set up sampling plots at 100-m intervals on the transect. At each point, I set up a lOO-mf circular plot centered on that point, and recorded for each tree within the plot: 1) species; 2) diameter at breast height (dbh); 3) condition Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 54 (intact snag, broken snag, dead-topped live, or totally live tree), 4) whether it was charred by fire, and; 5) % bark area disturbed by woodpeckers (0=0%, 1=1-25%, 2=26-50%, 3=51-75%, 4=>75%) . I also recorded b u m severity (defined by Key 1988) of each plot as one of the following: underbum (canopy >75% green) ; crown scorch (canopy >75% scorched tan or brown with probable canopy replacement) ; crown black (canopy >85% black with zero tree survivorship) ; and severe crown b u m (<15% snag remnants visible) . I sampled 262 trees within 64 plots in the MC site and 321 trees in 42 plots in the LP site. Nesting habitat I searched for nests along random transects in the Red Bench B u m during the nesting seasons of 1990-1994. The same 12 transects, totaling 42,000 meters, were surveyed each year (see chap. II for details of methods) . At each nest I recorded bird species, location, and characteristics of the nest cavity, the nest tree, and the 50-m-radius site around the nest tree (Table 5). I selected a series of random trees by using a random starting point on each transect and, at 250-m intervals, finding the tree nearest to the transect. The same characteristics were recorded for randomly located trees as for nest trees (except nest-cavity characteristics) . The same variables were recorded for any cavity nests found incidentally, e.g., while traveling in the study area to and from transects. These data were not used in comparisons with Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 55 Table 5. Variables recorded at nest and random trees and sites in post-fire habitat. Site variables were measured in a 50-m-radius plot around trees. Variable Method Tree Diameter at breast height (dbh) Diameter tape Height relascope Percent bark estimated Degree of lean compass Height of cavity relascope Condition (live/dead, top intact/broken) Tree species Site Percent canopy (pre-bum) estimated Percent live canopy estimated average tree size (dbh) diameter tape basal area (live and dead) relascope b u m severity (defined by Key 1988) visual Percent slope relascope Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 56 randomly located trees; however, they were used in descriptive summary statistics of nesting habitat. Data Analyses I compared average percent of bark disturbed on forage trees of different species, conditions, char status, and b u m intensities using Kruskal-Wallis analysis of variance (ANOVA). Differences among categories were tested for significance using the post-hoc test of least significant difference. Mean, dbh of trees in different bark disturbcince categories were compared with ANOVA. I used ANCOVA to determine the effects of several variables simultaneously on the percent of bark removed from trees. B u m intensity, tree condition, char status, and tree species were entered as main effects, and dbh was entered as a covariate. I compared characteristics of used and randomly selected nest trees for each bird species with 5 or more nests. I grouped species by foraging habits (Chap. II) and compared nest microhabitat use among the groups. I also combined nest data from all species for comparisons with randomly located sites. Univariate analyses were performed first. Following methods suggested by the resource selection function theory of Manly et al. (1993), selectivity indices were calculated for categorical variables to compare habitat at used (nest) and available sites. The selectivity indices give the relative probability that a habitat unit (tree or site) in a particular category will be used. These indices are ratios Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 57 of the percent of used units in a category to the percent of available units in that category (i.e., %used / %available) , standardized to sum to 1. I used a Chi-square analysis to test whether the distributions were significantly different between used and unused sites. Mann-Whitney U tests of mean ranks were used to compare values of continuous variables between used and available sites. I applied Bonferroni adjustments to final p-values to account for the large number of tests (one for each bird species) . To determine which species used nest sites most different from randomly located sites, I used mean values of numeric variables to calculate Euclidean distances between nest and random trees for each species. I used standardized 2-scores of variables to calculate distances. Backwards stepwise multiple logistic regression was used to assess the relative, simultaneous importance of variables in predicting use of sites. Variables were first screened with univariate analyses; any variables with significance values of 0.25 or less were entered into the regression model. Variables were examined for curvilinear trends to determine whether quadratic equations would be appropriate in the model. Results Foraging habitat Use of trees by foraging woodpeckers was nonrandom. The average percent bark disturbance category for trees at the MC site (1.43 ± 0.07) was significantly higher than the value Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 58 for trees at the LP site (0.39 ± 0.04; p < 0.0001) . At the MC site, tree dbh [mean = 36 cm ± 1.25) was significantly higher than at the LP site (mean = 24 cm ± 0.48; p < 0.0001). The number of trees per plot was higher at the LP site (mean = 7.64) than at the MC site (mean = 4.09). Because of these differences, I analyzed data separately for the two sites. At both sites, bark disturbance increased significantly (p < 0.0001) as tree diameter increased (Fig. 5). Average bark disturbance differed significantly among tree species at both sites (p < 0.0001), and was greatest for w e s t e m larch, lodgepole pine, and Douglas-fir (Fig. 6). Broken-topped snags were used less than live trees and unbroken snags at the MC site (p < 0.0001), whereas tree condition was unrelated to use at the LP site. Bark disturbance was lowest in areas of severe crown b u m at both sites (ANOVA, p < 0.0003; Fig. 6). Bark disturbance category of charred trees (mean = 1.58 ± 0.08) was significantly higher than value of non-charred trees (mean = 0.29 ± 0.08; p < 0.0001) at the MC site; the difference was not significant at the LP site (p = 0.06), although the trend was the same. Associations among independent variables may have influenced results. For example, tree species and condition were not independent (%^ = 402, df = 24, p < 0.0001); most larch were live trees, while most spruce, fir, and lodgepole were dead. Average dbh of snags (mean = 26 cm ± 0.54) was Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5 9 80-1 60 - su T g Q 2 0 - I T 0 4 % bark disturbance category S Mixed C o n ife r S it e O Lodgepole Pine Site (n=262) (n=335) Figure 5. Average diameter at breast height (dbh) of trees in different bark disturbance categories, based on amount of bark removed by woodpeckers: category 0=0%, 1=1-25%, 2=26-50%, 3=51-75%, 4=>75%. Differences among categories were significant at both sites (p < 0.0001). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 60 as I m &Î ffl < % iiP- o < % < CO itil5 ^ S .2? <« T3 C "= 2 O u. Z m < < -P“ XjoSaiBD aoaBqjiasfp aSiuaAy I § V ii g % m \ ffi 3 o V (0& u V ill s a m <, 3 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 61 smaller than the dbh of live trees (mean = 46 cm ± 2.19; p < 0.0001). When all factors were considered simultaneously using ANCOVA, only tree species and dbh were significantly related to amount of foraging use at both sites (p < 0.01). Nesting- Habitat I found 501 active cavity nests of 20 bird species in the study area; 392 were located from transects (Table 6). Nests were found in all b u m severities. Most nests (80%) were in aspen, and almost all nest trees (96%) were snags. Microhabitat characteristics of and around nest trees differed from those of randomly located trees in several ways. Nest trees tended to be of greater diameter and have less bark than random trees (Table 7) . Most species used aspen (Populus tremuloides) and cottonwood (Populus trichocaipa) more than expected based on availability (%^ = 353, df = 10, p < 0.0001; Table 8, Fig. 7) . Nest trees also displayed signs of decay much more often than did random trees. Broken-top snags were used proportionately more than trees of other conditions (%^ = 24, df = 3, p < 0.0001; Table 9, Fig. 8) , and more nest trees (62%) had conks than did random trees (2%; = 127, df = 1, p < 0.0001; Table 9). Site characteristics also differed between nest-tree- centered and random- tree - c entered sites. Average tree diameter in the stand, basal area (a measure of the area of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 52 Teüale 6. Common name, scientific name, and number of nests found within the Red Bench Fire study area of Glacier National Park for 20 cavity-nesting species. Coimon Name Scientific Name # Nests Tree Swallow Tachycineta bicolor 119 Northern Flicker Colap tes aura tus 81 Red-naped Sapsucker Sphyrapicus nuchalis 64 Mountain Bluebird Sialia currucoides 54 Three-toed Woodpecker Picoides tridactylus 45 Hairy Woodpecker Picoides villosus 34 Red-breasted Nuthatch Sit ta canadensis 23 Downy Woodpecker Picoides pubescens 21 American Kestrel Falco sparverius 14 Black-backed Woodpecker Picoides arcticus 11 Mountain chickadee Parus gambeli 6 House Wren Troglodytes aedon 6 European Starling Stumus vulgaris 5 Pileated Woodpecker Dryocopus pileatus 5 Bufflehead Bucephala albeola 4 Black-capped Chickadee Parus atricapillus 3 Lewis' Woodpecker Melanerpes lewis 2 Brown Creeper Certhia americana 2 Common Merganser Mergus merganser 1 Northem Saw-whet Owl Aegolius acadicus 1 Total 501 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 63 Table 7. Means (±SE) of nest and random tree variables in post-fire habitat. P-values of Mann- Whitney U tests (after Bonferroni adjustment) of the difference between nest and random trees are given below the summary statistics for species with 5 or more nests. Bird Species Dbh (cm) °Lean Nest Ht Tree Ht % Bark (n) (m) (m) Bufflehead 55 ±2.9 4 ± 1.2 12 ± 0 25.8 ± 1.6 100 ± 0 (4) Common M erganser 71 16 32.9 (1 ) American Kestrel 50 ±3.6 4 ± 2.0 13 ± 2.3 19.5 ± 2.0 71 ± 11 (14) .0075 1.0 1.0 1.0 Lewis Woodpecker 78 ± 2.5 4 ± 4 22 ± 5.0 25.3 ± 8.3 50 ± 50 (2 ) Red-naped 37 ± 1.1 2 ± 0.8 10 ± 0.6 22.6 ± 1.0 90 ± 2.7 Sapsucker .0000 .44 1.0 .027 (64) Downy Woodpecker 29 ±1.6 4 ± 1.4 9 ± 1.0 21.4 ± 1.1 77 ± 6.1 (21 ) 1.0 1.0 1.0 1.0 Hairy Woodpecker 35 ± 1.4 5 ± 1.0 10 ± 0.9 23.4 ± 0.8 67 ± 8.6 (34) .081 1.0 1.0 1.0 T h re e -to e d 31 ±1.5 5 ± 1.2 7 ± 0.7 22.7 ± 0.9 68 ± 7.0 Woodpecker 1.0 1.0 .71 1.0 (45) Black-backed 40 ±5.0 2 ± 0.6 11 ± 2.8 28 ± 1.6 99 ± 0.9 Woodpecker 1.0 1.0 .45 .90 (11) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 64 Northern Flicker 41 ±1.2 3 ± 0.7 12 ± 0.8 21.6 ± 1.0 43 ± 5.3 (81) .0000 .44 1.0 .015 pileated Woodpecker 54 ±3.3 1 ± 1.0 15 ±0.5 30.2 ± 0.3 83 ±17 (5) .063 1.0 .072 1.0 N o rth e rn 61 10 30.5 100 Saw-whet Owl (1 ) Tree Swallow 32 ±0.9 2 ± 0.6 10 ± 0.6 20.9 ± 0.8 50 ± 4.3 (119) .027 .0000 1.0 .123 Black-capped 38 ±9.0 4 ± 1.9 17 ± 5.4 21.2 ± 5.6 83 ± 16 C hickadee 1.0 1.0 1.0 1.0 (3) Mountain Chickadee 35 ± 5.2 2 ± 1.6 17 ± 2.422.7 ± 2.2 78 ± 3.3 (6 ) 1.0 1.0 1.0 1.0 Red-breas ted 34 ±1.6 3 ± 0.9 12 ± 1.3 20.0 ± 2.0 80 ± 9.8 N u th atch 1.0 1.0 1.0 1.0 (23) Brown Creeper 72 ± 0 5 ± 0 2 ± 0 32.6 ± 0 73 ± 23 (2 ) House Wren 30 ±2.3 4 ± 2.2 8 ± 0.8 14.7 ± 2.4 13±7.7 (6 ) 1.0 1.0 .56 .36 Mountain Bluebird 34 ±2.3 3 ± 0.6 8 ± 0.9 17.9 ±1.8 42 ±7.6 (54) .054 .18 1.0 .13 European Starling 31 ±1.4 3 ± 1.2 12 ± 1.0 20.2 ± 3.3 13 ± 13 (5) 1.0 1.0 1.0 1.0 A il Species 35 ± 0.6 3 ± .04 10 ± 0.3 22 ± 0.4 57 ± 2.1 (391) . 0000 .041 1.0 1.0 Random 31 ±1.3 3 ± 0.4 63 ± 3.7 (153) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 65 Table 8. Tree species availability and use by cavity-nesting birds in early post-fire habitat. Glacier National Park, Montana. Values indicate % of nests that were in each tree species. Bird species Tree species^ n QA BC PP WL SF EF SP LP PB Bufflehead 4 100 0 0 0 0 0 0 0 0 Common Merganser 1 0 100 0 0 0 0 0 0 0 American Kestrel 14 71 0 21 7 0 0 0 0 0 Northern Saw-whet 1 100 0 0 0 0 0 0 0 0 Owl Red-naped Sapsucker 64 93 3 0 0 0 0 0 0 4 Downy Woodpecker 21 100 0 0 0 0 0 0 0 0 Hairy Wooc^ecker 34 88 3 0 9 0 0 0 0 0 Three-toed- 45 44 2 0 29 4 0 18 2 0 Wood^)ecker Black-backed 11 0 0 082 018 0 0 0 WoocftÆcker Northern Flicker 81 76 6 3 8 0 0 0 0 0 Pileated Woodpecker 5 100 0 0 0 0 0 0 0 0 Tree Swallow 119 92 3 0 3 0 1 1 0 0 Black-capped 3 100 0 0 0 0 0 0 0 0 Chickadee Mountain Chickadee 6 83 0 0 17 0 0 0 0 0 Red-breasted 23 74 4 0 9 0 9 4 0 0 Nuthatch Brown Creeper 2 0 0 0 100 0 0 0 0 0 House Wren 6 100 0 0 0 0 0 0 0 0 Mountain Bluebird 54 61 9 0 17 0 6 4 4 0 European Stcirling 5 80 0 0 0 0 0 20 0 0 All Species 391 80 4 1 10 <1 2 2 <1 <1 Randomly selected 153 5 1 1 14 6 14 26 33 0 Selectivity Indexé .724 .181 .045 .032 .007 .006 .003 .001 n/a ^ QA = quaking aspen; BC = black cottonwood; PB = paper birch (Betula papyri fera ) ; WL = western larch; DF = Douglas-fir; SP = Engelmann spruce (Picea engelmanniij ; SF = subalpine fir (Abies lasiocarpa); LP = Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 66 lodgepole pine; PP = ponderosa pine (Pinus ponderosa) . ^ The selectivity index is the %used/%available of each tree species, standardized to sum to 1. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 67 80 n Nest Trees Random Trees (nc392) (n=153) 60 40“ Figure 7. Use and availability of tree species for cavity nests; nest data are for all bird species combined. Tree species are listed in descending order based on the ratio of use to availability. Differences between nest and random trees are significant (p < 0.0001). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 68 Table 9. Tree condition availability and use by cavity-nesting birds in early post-fire habitat. Glacier National Park, Montana. Values are % of each bird species' nest trees in each category. Bird Species Tree Condition n Broken Intact Broken Intact % with snag snag live live conks Bufflehead 4 0 0 100 0 100 Common M erganser 1 100 0 0 0 100 American Kestrel 14 50 50 0 0 40 Red-naped Sapsucker 54 22 56 0 22 95 Downy Woodpecker 21 17 83 0 0 87 Hairy Woodpecker 34 19 81 0 0 67 Three-toed Woodpecker 45 17 80 0 3 42 Black-backed Woodpecker 11 17 83 0 0 100 Northern Flicker 81 37 58 0 4 53 Pileated Woodpecker 5 100 0 0 0 100 Northern Saw-whet Owl 1 0 0 0 0 100 Tree Swallow 119 30 69 0 1 59 Black-capped Chickadee 3 33 67 0 0 100 Mountain Chickadee 6 40 40 20 0 50 Red-breasted Nuthatch 23 75 19 6 0 75 Brown Creeper 2 100 0 0 0 0 House Wren 6 33 67 0 0 0 Mountain Bluebird 54 34 66 0 0 44 European Starling 5 40 60 0 0 40 A ll Species 391 29 67 <1 4 62 Random 153 9 80 <1 11 2 Selectivity Index^ .620 .166 .1 4 1 .073 a The selectivity index is the %used/%availabie of each tree species , standardized to sum to 1. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 69 80n Nest Trees [] Random Trees (n=392) (n=153) 60- 2 o 40“ 20 - Intact Snag Broken Snag Broken Live Intact Live Tree Condition Figure 8. Conditions of nest and randomly-selected trees in post-fire habitat. Distributions of tree types differed significantly (Chi-square, p < 0.00 1) between the two groups. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 70 standing wood) , and b u m severity were greater around nest trees than around randomly selected trees, whereas percent slope was less at nest trees (Table 10) . Areas of crown fire, except for the most intense b u m category, were used more than expected (%^ = 21, df = 4, p = 0.0003; Fig. 9) . Pileated Woodpecker, European Starling, and Black-backed Woodpecker nests were least similar to nests of other species and to random sites (Fig. 10). Nest-site characteristics differed among categories of bird species grouped by foraging habits; timber-drilling woodpeckers and gleaners nested in sites of greater basal area, average tree dbh, and live canopy than did ground and air foragers (Table 11) . Similar to univariate analyses, multiple logistic regression indicated that both tree and surrounding site characteristics influenced use. For cavity-nesting species as a whole, tree species, basal area, b u m severity, tree condition, and amount of bark all contributed significantly to predicting use (Table 12). The model correctly classified 98% of nests, and probability plots (not shown) demonstrated that nests and random sites were clearly separated. Analyses for individual bird species showed that although tree species and tree condition were significant for all species, importance of surrounding site variables differed among species (Table 12). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 71 < u o tn es es o es tH es es o eo 44 o o <3 a 2 o 44 • 44 44 • 4 4 • 44 • 44 • 4 1 es tH tH tH tH tH 44'=: d P S tH < 1 o kO r n t d t tH « U •H I «W a a L f l 0 0 un I g A m o . Q tH tH tH Q a T > 0 O o o o O ^ e o O e z ) Q ca "H o 4 4 • 4 4 C 3 4 4 • 44 • 44 • 41 Q 9 a 4 1 «H 4 4 Q 4 4 ° t H t H t H e z i r ~ o kO kO e n m • a d p e s m e s g M ro m tH tH tH tH en tH a M o 0 « b— O o O O OO O Q H « w kn eo O tH lfl O i n ° 03 C3 d k 1 4 1 V 4 1 • 4 1 4 1 • 4 4 0 0 4 1 tH 4 4 o 0 t H tH tH 4 1 " ° o •H n 0 iH o •H a G 3 a û AS a O 4J •H 4J I—t •H jj W k -H I m a 73 u 73 .'S 4-1 a 5 73 O, •H a o k a a a a O a o ■3 A 4J m ► G o s 0 3 Q m Eh S m s 2 A3 4J a g u Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 72 cs CS Tf 03 cs cs o o CS tH o K +, o O o o m Q O tH 44 • 44 • 44 44 • tH 44 • tn 44 44 . tH tH tH 44 • t-t 44 • in o O CM TT t tH tH cs cs rH ro cs ro tH CS ro 00 o o o o o Q o o o o O O O o C3 Q Q O 44 O o Q o o o 44 • O 44 O 44 44 • 44 • 44 ro 44 O 44 • o tH O tH tH o o tH 44 o 41 in 03 O ro tjt 03 kO V kO cs ro cs cs ro ro ro ro ro ro kO kO CM C3 tH ro ro o 00 cs o tH o tH Q Q 41 cs 44 44 4 4° 4-1 44 O 44 o 41 o 44 (S O 44 o o kO tH kO O t> ^ 00 in 00 tH CM C3 O' ' tH o tH tH tH fH V cs cs 03 cs ro ro kO CS in cs m o cs CS m tH tH in tH CO 41 ko CN 4-1 Q 4 4 ^ 4-1 44 C3 44 C3 4 4 ° 41 CO C3 tH 41 + 1 ^ CO 44 . • tH tH CO o 03 in o ^ tH in 03 kO to ^ o in tHcs tH CS tH tH tH tH cs tH in kO ro tH n- tH CN O ro r4 CN in in ro O 41 tH kO 44 03 41 4 1 ° 44 m 44 - H ° 4 1 ° +'S? CO ro tH CO 41 kO • cs V o • 00 ko 03 " cs ro CN ro ro CN CN CN 03 ■ ro CN CN in 03 m kO fO CS U3 in tH m tH CM in O in tH in tH rH 3 73 73 a 3 0 a a 0, a k o iH Û. OJ a c a a rH Qt a a a a a •H 1 73 AS C 4-> a a a c a a XI k k C 73 G D) u >i73 a u k a 3 U 73 •H 73 a Ü u 2 •H k a G a rH a a a XI CO 1 a a a k OJ a -H a -H a E OJ a Q< XI 3 AS AS 0-1 AS jQ a H a OJ P Qt iH ca 0 u a 73 4J I a Ü u C 0 1 XI 3 a G a O k 73 a rH o k 3 a a -H 3 -H 73 oJ 0 0 3 3 k a rH G tH •H 0 0 a k tH jd 0 XI a 3 k o 0 tH 3 OJ tH a a & s 2 M E-t m u s u OS 2 ca K 2 ca M M < os a Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 73 80-1 Nest Sites Q Random Sites (n=392) (n=153) 60- 40 - 20 - Unburned ünderburn Crown scorch Crown black Severe crown Fire Severity Class Figure 9. Percentages of nests and random sites in each category of fire severity; nest data are for all bird species combined. The difference between nest and randomly-selected sites was significant (Chi-square, p = 0.0003). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 74 Hairy Woodpecker Three-toed Woodpecker Downy Woodpecker Red-breasted Nuthatch Red-naped Sapsucker Mountain Chickadee RANDOM American Kestrel Mountain Bluebird Tree Swallow Northern Flicker House Wren Black-backed Woodpecker European Starling Pileated Woodpecker Figure 10. Cavity nest sites and random sites, clustered by standardized mean values of microhabitat variables of the tree and surrounding 50-m-radius area. The dendrogram is based on squared Euclidean distances and average linkages between groups. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 75 Table 11. Nest-site characteristics of cavity- nesting bird species grouped by foraging habits. Differences among the bird groups are significant (ANOVA, p < 0.0001} for all variables. Variable Foraging group Timber- Air/ground dri He r s Gleaners foragers (109) (79) J202) Mean SE Mean SE Mean SE B a s a l area (m2/ha) 25 1.3 20 1.4 17 0.84 (live and dead) (Aa) (B) (B) Average dbh (cm) 30 0.88 31 0.98 26 0.59 (A) (A) (B) % live canopy 4 0.79 6 1.12 1 0.21 (A) (B) (C) a Groups that share a letter do not have significantly different mean values for the variable, using ad hoc tests of least significant differences. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 76 Table 12. Results of stepwise multiple logistic regression of nest and random trees in post-fire habitat, based on characteristics of trees and the surrounding 50-m-radius forest stand. Bird Species (n) Model P % nests Variables (d f) c o r r e c t ly in model classified A l l (392) 547.38 .0000 98 B a s a l a re a (19) Tree species Burn severity Tree condition % baric T h re e -to e d 97.40 .0000 74 B a s a l a re a Woodpeclcer (30) (12) Tree species Tree condition Average dbh Northern Fliclcer (69) 254.29 .0000 94 B a s a l a re a (19) Tree lean ( - ) % canopy (-) live canopy (-) Dbh Bum severity S lope Tree species Tree condition Mountain Bluebird (45) 181.76 .0000 93 Burn severity (16) Tree condition S lo p e ( - ) % canopy Tree species % baric Average dbh Tree Swallow (108) 320.13 .0000 99 Burn severity (11) B asal a re a % b a rk S lo p e ( - ) Tree species Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 77 Discussion Pre-fire forest conditions, fire severity, and topography all influenced cavity-nesting bird use of the post-fire forest. Woodpeckers foraged more in areas with larger larch, Douglas-fir, and lodgepole pine snags. Amman and Ryan (1991) found that beetle larvae were more abundant in larger trees in burned forests of Yellowstone National Park. Thus, an abundance of prey probably makes large trees valuable foraging sites for woodpeckers. Amman and Ryan (1991) also detected more wood-boring beetle larvae in subalpine fir than in lodgepole and Douglas-fir. However, when only dead trees were considered, more lodgepole and Douglas-fir had beetle larvae. Beetle larvae in firs were limited to the small areas of bark that did not slough from the tree following burning. Hutto (1995) found more evidence of woodpecker foraging on larger trees in post-fire habitat in Montana and Wyoming. He reported Douglas-fir and western larch were used heavily, but observed little foraging evidence on lodgepole pine. Because it was not possible to identify the species of woodpeckers from foraging evidence, observed differences in tree species used may have been due in part to differences in preferences of individual bird species. Local differences in insect species colonizing post-fire habitat may also influence use of trees by foraging birds. Raphael and White (1984) found that medium-sized trees were foraged on more Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 78 than expected; the largest trees may have been energetically unfavorable to use because they were few and far between. Although smaller trees were more abundant in my study area, large trees were used most for foraging. Clearly, further research is needed to clarify the importance of different tree species and sizes as foraging sites in burned forests. Such research should include observations of foraging woodpeckers, preferably in various seasons. In post-fire habitat, nest-site locations apparently depended on availability of both foraging and nesting substrate. Forest patches around nest trees had more and larger standing trees than randomly located sites. Nests were placed in sites with characteristics that corresponded to bird species' foraging habits. Swallow et al. (1986) suggested the greater snag diversity they found around nests offered more foraging substrates. Li and Martin (1991) reported cavity nests in sites of greater conifer density than random sites, and suggested those sites offered more foraging opportunities. Conner et al. (1975), McClelland (1977), and Harris (1982) all found that tree-foraging woodpecker species (Pileated, Three-toed. Black-backed, Hairy and Downy woodpeckers) nested in areas of greater tree densities than the predominately ground-foraging Northern Flicker. However, Bull et al. (1986) found Hairy Woodpeckers nesting in stands slightly less dense than stands used by flickers. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 79 Factors other than foraging opportunities may explain variation in sites surrounding nests. Predation pressure may result in use of sites with vegetation characteristics that differ from random sites and differ among bird species (Martin 1993). In my study it seems unlikely that variation is due solely to predation avoidance, given the relatively large size of plots around nests and the correspondence of site characteristics with foraging habits of different species. In post-fire habitat, higher tree density does not result in more nest concealment by foliage, since most trees are dead. The greater basal area around nests was due primarily to conifers rather than to the aspen used commonly for nesting. It is unlikely, therefore, that birds selected sites with more potential nest substrates for predators to search (Martin 1993). Li and Martin (1991) found cavity nest success was lower (due primarily to predation) in sites with more large conifers, indicating that conifer density increased predation. Another selection pressure, such as foraging potential, may act to balance selection against dense stands. Characteristics of nest trees also influenced use. Large aspen that showed signs of advanced decay (broken tops and conks) were particularly important. Use of aspen for nest cavities has varied among other studies. McClelland (1977), working in pre-fire conditions in an area overlapping my study area, found western larch was used much more than Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 80 any other species in proportion to availability. Raphael and White (1984) reported that use of aspen was less than expected in the Sierra Nevada, although others (McLaren 1963, Winternitz and CcOm 1983, and Li and Martin 1991) found aspen was used frequently for nesting. Use of aspen may depend on the presence and conditions of other tree species. Aspen may be favored in early post-fire habitat because existing conifer snags tend to b u m completely (Horton and Mannan 1988), and fire-killed conifers may not reach appropriate decay states for several years post-fire (Kimmey 1955). Hutto (1995) and Hitchcox (pers. comm.) found that nests in early post-fire habitat were predominantly in aspen showing pre-fire decay. Raphael and White (1984) found aspen were used little compared to availability, but since they were working 7+ years post fire, most aspen snags may have decayed to an unfavorable stage compared to the harder conifer snags. Nest and foraging trees differed substantially. Although decayed aspen (those with broken tops and conks) were used disproportionately often for nests, conifers that were not obviously decayed were used more than other tree types for foraging. Wood-boring beetle larvae on which woodpeckers feed usually infest freshly dead wood of conifers (Evans 1971). Trees in varying states of decay may fulfill different requirements, in early stages providing appropriate foraging sites and in later stages fumishing nest sites. Nest tree and forest stand characteristics were both Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 81 important in predicting use by most cavity nesters. Most species nested in broken-topped aspens surrounded by stands of high basal area and crown bum. However, nests of gleaner species were in areas with more live canopy than those of other species, while nests of air and ground foragers were in stands of lower basal area. Swallow et al. (1986) found forest characteristics (primarily snag basal area) better predicted cavity presence in a mature hardwood forest than did tree characteristics; they did not identify bird species. Site selection may occur in a hierarchical way (Hutto 1985), in which a bird selects habitat at increasingly smaller spatial scales until a particular position in a tree is chosen. A timber-drilling woodpecker, therefore, may select a forest stand with characteristics indicating high food availability and then select a suitable nest tree within that stand. Future research should include comparing nest abundances in aspens that vary in context of surrounding forest stands. Such research might be done in conjunction with carefully planned timber harvest of post-fire forests. Additionally, it would be instmotive to compare abundances in areas with suitable aspens to those in similar areas without aspen, to determine if absence of aspen precludes post-fire nesting for any species. Past forest management for cavity-nesting species has focused on nest-site availability as a limiting factor of populations. Although lack of nest trees and cavities likely Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 82 limits some species in some habitats, food availability also plays an important role in determining distributions (Brawn and Baida 1988, Waters et al. 1990, Sedgewick and Knopf 1992, Welsh and Capen 1992). Management often entails providing potential nest snags and boxes in areas where habitat has been manipulated, under the assumption that needs of cavity nesters will then be met. These practices probably have been inadequate for some cavity-nesting species. Foraging as well as nesting requirements need to be considered to maintain healthy populations of species like timber-drilling woodpeckers (McClelland 1979). Managing for large patches of naturally dense snags, rather than for single trees or small groups of trees, is more likely to sustain less common species like Black-backed and Three-toed woodpeckers. Public land managers will want to consider several factors within the context of fire-generated habitat. Fire management should include, whenever possible, allowing fires to b u m as they would in natural situations. This may include large fires that encompass the various microhabitats used by different bird species. Crown fires tend to create suitable conditions for most species, especially in areas with large aspen, larch, Douglas-fir, and lodgepole pine. Leaving sizable bumed areas of these kinds unaltered will allow opportunities for further research of post-fire habitat use. In view of the many uncertainties regarding the relationship of cavity nesters and fire, this may provide Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 83 habitat important to various species. Fire also has long term, positive effects on cavity-nesting bird habitat. Fire perpetuates tree species important for nesting and foraging (aspen, larch, and lodgepole pine). Without the influence of fire, these tree species and the birds that use them may become rare in the regional landscape. Literature Cited Amman, G. D., and K.D. Ryan. 1991. Insect infestation of fire-injured trees in the greater Yellowstone area. Research note INT-398, U.S. Forest Service, Ogden, Utah Bock, C. E., and J. F. Lynch. 1970. Breeding bird populations of bumed and unbumed conifer forest in the Sierra Nevada. Condor 72:182-189. Brawn, J. D., and R. P. Baida. 1988. Population biology of cavity nesters in nor th e m Arizona: do nest sites limit breeding densities? Condor 90:61-71. Brush, T. 1983. Cavity use by secondary cavity-nesting birds and response to manipulations. Condor 85 (4):461-466. Brussard, P. F . 1991. The role of ecology in biological conservation. Ecol. Appl. 1(1):6-12. Bull, E. L. 1987. Ecology of the Pileated Woodpecker in northeastem Oregon. Joumal of Wildlife Management Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 84 51(2): 472-481. Bull, E. L., S. R. Peterson, and J. W. Thomas. 1986. Resource partitioning among woodpeckers in northeastem Oregon. Res. Note PNW-444 LaGrande, OR. U.S. Dept, of Agriculture, Forest Service, Pacific Northwest Research Station 19 pp. Conner, R. N. 1978. Snag management for cavity-nesting birds. Pages 120-128 in Management of Southem Forests for Nongame Birds, Workshop Proc. U.S.D.A. For. Serv. Gen. Tech. Rep. SE-14. Conner, R. N., R. G. Hooper, H. S. Crawford, and H. S. Mosby. 1975. Woodpecker nesting habitat in cut and uncut woodlands in Virginia. J. Wildl. Manage. 39(1) : 144-150. Conner, R. N., 0. K. Miller, and C. S. Adkisson. 1976. Woodpecker dependence on trees infected by fungal heart rots, Wilson Bull. 88(4): 575-581. Evans, W. G. 1971. The attraction of insects to forest fires. In Tall Timbers Conf. on Ecol. Animal Control by Habitat Manage. Proc. 3:115-127. Gibbs, J. P., M. L. Hunter J.R., and S. M. Melvin. 1993. Snag availability and communities of cavity-nesting birds in tropical versus temperate forests. Biotropica 25(2):236-241. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 85 Gustafsson, L. 1988. Inter- and intraspecific competition for nest holes in a population of the collared flycatcher Ficedula alhicollis. Ibis 130:11-16. Harris, M. A. 1982. Habitat use among woodpeckers in forest bums. M. S. Thesis, Univ. of Montana, Missoula. Heinselman, M. L. 1971. The natural role of fire in northem conifer forests. Pages 61-72 in Fire in the Northem Environment Symposium Proceedings. College, Alaska. Horton, S. P., and R. W. Mannan. 1988. Effects of prescribed fire on snags and cavity-nesting birds in southeastern Arizona pine forests. Wildl. Soc. Bull. 16:37-44. Hutto, R. L. 1985. Habitat selection by nonbreeding, migratory land birds. Pages 455-476 in Habitat Selection in Birds (M. L. Cody, ed. ) . Academic Press, New York. Hutto, R. L. 1995. Composition of bird communities following stand-replacement fires in Northem Rocky Mountain (U.S.A.) conifer forests. Cons. Biol. 9 (5): 1041-1058. Key, C. 1988. Red Bench Fire Intensity Map. Unpub. Rep. , Glacier National Park, Mont. Kimmey, J. W. 1955. Rate of deterioration of fire-killed timber in Califomia. U.S. Dept. Agric. Circ. 962:1-22. Korol, J. J., and R.L. Hutto. 1984. Factors affecting nest site Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 86 location in Gila Woodpeckers. Condor 86:73-78. Laverty, L., J. Williams, B. Kilgore, S. Woodley, and K. Corrigall. 1993. Panel: Wilderness and Park Fire Policies and Programs. Fire in Wilderness and Park Management Symposium Intermountain Fire Sciences Laboratory, Missoula, Montana. Li, P., and T. E. Martin. 1991. Nest-site selection and nesting success of cavity-nesting birds in high elevation forest drainages. Auk 108:405-418 Manly, B. F., L. McDonald, and D. Thomas. 1993. Resource selection by animals. Chapman and Hall, London. Mannan, R. W . , E. C. Mes low, and H. M. Wight. 1980. Use of snags by birds in Douglas-fir forests, western Oregon. J. Wildl. Manage. 44 (4)-.787-797 . Martin, T. E. 1993. Nest predation and nest sites : new perspectives on old patterns. BioScience 43(8):523-532 . McClelland, B. R. 1977. Relationships between hole-nesting birds, forest snags, and decay in western larch - Douglas-fir forests of the northem Rocky Mountains. Ph.D. Diss., Univ. of Montana, Missoula. McClelland, B. R. 1979. The Pileated Woodpecker in forests of the northem Rocky Mountains. In The role of insectivorous birds in forest ecosystems (J.G. Dickson, R.N. Conner, R.R. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 87 Fleet, J.A. Jackson, and J.C. Kroll, eds.). Academic Press, N.Y. McLaren, W. D. 1963. A preliminary study of nest-site competition in a group of hole-nesting birds. M.S. Thesis, Univ. of British Columbia, Vancouver. Orians, G., and J. Wittenberger. 1991. Spatial and temporal scales and habitat selection. American Naturalist 137(S):S29-S49. Peterson, B., and G. Gauthier. 1985. Nest-site use by cavity-nesting birds of the Cariboo Parkland, British Columbia. Wilson Bulletin 97 (3) .-319-331. Pfister, A. R. 1980. Postfire avian ecology in Yellowstone National Park. M. S. Thesis, Univ. of Washington, Pullman . Raphael, M. G ., and M. White. 1984. Use of snags by cavity- nesting birds in the Sierra Nevada. Wildl. Monog. 86:1-66. Schreiber, B . , and U.S. deCalesta. 1992. The relationship between cavity-nesting birds and snags on clearcuts in western Oregon. Forest Ecology and Management 50:299-316. Scott, V. E., J.A. Whelan, and P.L. Svoboda. 1980. Cavity-nesting birds and forest management. Pages 311-324 in Symposium Proceedings of Management of western forests and grasslands for nongame birds. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 88 Sedgewick, J. A., and F.L. Knopf. 1992. Cavity turnover and equilibrium cavity densities in a cottonwood bottomland. J. Wildl. Manage. 56(3):477-484. Shi go, A. L. , and L. Kilham. 1968. Sapsuckers and Fames ignicLrius var. Populinus. U. S. For. Serv. Res. Note NE-84. NE Forest Exp. Sta., Upper Darby, Pa. Steele, B. B. 1992. Habitat selection by breeding Black-throated Blue Warblers at two spatial scales. Ornis Scandinavica 23 :33-42 . Swallow, S. K . , R.J. Gutierrez, and R.A. Howard. 1986. Primary cavity-site selection by birds. J. Wildl. Manage. 50 (4) :576-583. Taylor, D. L ., and J. W. J. Barmore. 1980. Post-fire succession of avifauna in coniferous forests of Yellowstone and Grand Teton National Parks, Wyoming. Pages 130-145 in Management of western forests and grasslands for nongame birds. U.S.D.A. For. Ser. Gen. Tech. Rep. INT-86. van Balen, J. H., C. J. H. Booy, J. A. van Franeker, and E. R Odieck. 1982. Studies on hole-nesting birds in natural nest sites:1. availability and occupation of natural nest sites. Ardea 70:1-24. Waters, J. R. , B. R. Noon, and J. Vemer. 1990. Lack of nest-site limitation in a cavity-nesting bird community. J. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 89 Wildl. Manage. 54:239-245. Welsh, C. J., and D. E. Capen. 1992. Availability of nesting sites as a limit to woodpecker populations. For. Ecol. and Mgmt. 48:31-41. Wesolowski, T. 1989. Nest-sites of hole-nesters in a primaeval temperate forest (Bialowieza National Park, Poland) . Acta Omith. 25(3) :321-351. Wintemitz, B. L., and H. Cahn. 1983. Nestholes in live and dead aspen. Snag Habitat Management Symposium, Flagstaff, AZ. Zamowitz, J. E., and D.A. Manuwal. 1985. The effects of forest management on cavity-nesting birds in northwestern Washington. J. of Wildl. Manage. 49(1):255-263. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter IV THE EFFECTS OF POST-FIRE SALVAGE LOGGING ON CAVITY-NESTING BIRDS Introduction Fire in northern Rocky Mountain forests dramatically alters vegetation and environmental conditions, thereby influencing bird communities (Davis 1976, Taylor and Barmore 1980, Harris 1982; see Hutto 1995 for a review). Cavity-nesting birds increase in abundance in post-fire conditions (Pfister 1980, Taylor and Barmore 1980, Harris 1982, Hutto 1995). Woodpecker populations have been shown to increase after fires and remain high for 1-3 or more years in burned areas (Blackford 1955, Bock and Lynch 1970, Pfister 1980, Taylor and Barmore 1980, Harris 1982). Cavity nesters may respond positively for a variety of reasons. Fires create extensive stands of snags, providing potential nesting substrates. These fire-killed trees also attract wood-boring beetles and other insects, which serve as an abundant food resource for many species (Blackford 1955, Evans 1966) . Before fire suppression altered the fire regime of this region, recently burned areas probably were widespread on the landscape at any given time (Habeck and Mutch 1973, A m o 1980) . Remnant snags from fires, left to decay and fall 90 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 91 naturally, likely were significant components of forest structure in the northern Rocky Mountains. Although burned areas are not the only habitat used by cavity nesters, they may be critical to the persistence of some species (Hutto 1995) . Fire may produce high quality habitat that sustains populations in sufficient numbers for long-term persistence. Fire also increases regional habitat heterogeneity, allowing species to exploit different habitats in fluctuating environments (Simberloff and Abele 1982, Boecklen 1986). Finally, the increase in habitat diversity resulting from fire may help maintain genetic diversity within species that occupy a broad range of conditions (Bell and Lechowicz 1991, Mitchell-Olds 1992). Logging is often referred to as a mimic of natural disturbances such as fire (e.g., Blake 1982, Greenberg et al. 1995), and some studies have used logged areas as sites to represent early post-fire habitat (e.g., Catt 1991). However, differences in these two environments are obvious with regard to the needs of cavity-nesting birds. Fires leave forests of standing snags which serve as nesting and foraging substrates, whereas logging removes standing trees and snags. Some cavity-nesting species are among those most negatively affected by the widespread alterations of forest structure and age resulting from logging (Hejl 1994). The effects on cavity-nesting birds of logging green forests in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 92 the Rocky Mountains have been well documented (Hejl et al. 1995); however, effects of "salvage" logging of post-fire habitat have not been well documented at all. Although most cavity-nesting species use burned forests to a much greater extent than predicted based on its availability (Blackford 1955, Bock and Lynch 1970, Pfister 1980, Taylor and Barmore 1980, Harris 1982, Hutto 1995), resource management decisions often are made under the assumption that logging dead trees has little environmental consequence. Pre- and post-fire management activities, however, may have a significant influence on the availability of suitable habitat for cavity-nesting birds. Foraging and nesting sites may be reduced or eliminated by timber removal, affecting bird population levels. However, there has been very little research on the effects of logging in post-fire habitat (but see Blake 1982, S. Hitchcox pers. comm., V. Saab pers.comm.). The objectives of this study were to determine how logging in post-fire forests affected cavity-nesting birds in northwestern Montana. I recorded cavity-nesting bird abundances and nest densities in unlogged and salvage-logged forests that burned in the same fire. I compared bird abundances in logged sites to those in unlogged areas. I also compared abundances between treatments of low and high cutting intensities. Habitat characteristics were measured to determine if differences between areas could explain differences in bird use. Specifically, I addressed the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 93 following questions: 1. Does logging decrease the suitability of post-fire habitat for cavity-nesting birds? If so, suitable nest sites should be rarer in logged sites. I predicted that nest and randomly located sites in burned, logged forests would be less similar to one another than nest and randomly located sites in burned, unlogged forests. 2. If logging reduces nest-site availability, do birds using logged areas modify their nest-site selection patterns and use different types of sites than they do in unaltered habitat? If so, nest sites in logged areas should differ substantially from nest sites in unlogged areas. For species that are unable to modify their nest-site selection, I predicted lower nest densities in logged areas. Study Site The Red Bench Fire in 1988 burned 15,000 hectares in the North Fork of the Flathead River valley in northwestern Montana. Approximately 4000 hectares burned in the Flathead National Forest (FNF), where the fire started. The remaining 11,000 hectares burned in adjacent Glacier National Park (GNP). Before the fire, extensive lodgepole pine (Pinus contorta var. latifolia) stands intermixed with mixed-species coniferous forests dominated by western larch {Larix occidentalis) and Douglas-fir {Psuedotsuga menziesii) . A range of fire intensities left a mosaic of unbumed, lightly- burned, and severely-burned areas (Key 1988; see Chap. II for Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 94 a map and complete description of the study area). Trees were removed in commercial harvesting and as a perceived safety hazard during and after the fire, primarily in FNF. Except along roads and other developments, most of the post-fire habitat in GNP was left unaltered by human activities. In FNF, almost all of the area within the fire perimeter was logged either before or after the fire. Some stands were clearcut, where all trees were removed except for a few snags left to serve as potential nest trees. Other stands were partially cut; individual trees or small groups of trees were logged. Limited information was available on stand prescriptions and histories; therefore, I report descriptive statistics of forest stand characteristics in both logging treatment types. Methods Nest and bird abundances I used random transects to search for nests throughout the GNP portion of the Red Bench B u m during the nesting seasons of 1990-1994 (see Chap. II) . The same 12 variable- length transects, totalling 42,000 meters, were surveyed once each year between early June and mid-July. To locate nests and to census cavity nesters, transects were established in the FNF portion of the b u m in 1992, in an area that was homogeneous in aspect, elevation, and slope. Transects totaled 16,000 meters in length, and were surveyed once each year from 1992-1994. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 95 Cavity nesters were censused at 200-m intervals along the FNF transects in June and early July, using the fixed- radius point count method (Hutto et al. 1986). Counts were 10 minutes long and all cavity-nesting birds detected within 50 meters were recorded. Sixty points in partially cut stands and 15 points in clearcuts were censused once each year from 1992-1994. A different observer searched for nests and conducted point counts the first two years at the FNF site; the third year I collected all data at both sites. All observers were adept at identifying the species of interest and experienced at finding nests. I reduced observer bias by conducting a training period at the beginning of each nesting season. Habitat measurements At each nest, I recorded bird species, location, distance from nest to transect, and microhabitat characteristics. I recorded characteristics of the nest tree and of the forest stand in a 50-m-radius plot around each nest tree (Table 13). Starting from a randomly selected point on each transect, I found the tree nearest the transect at 250-m intervals and recorded the same variables that I recorded at nest trees. These data were used to compare nest characteristics to characteristics of and around randomly located trees. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 96 Table 13. VariaüDles recorded at nest and random trees and sites in post-fire habitat. Site variables were measured in a 50-m-radius plot around trees. Variable Method Tree Diameter at breast height (dbh) Diameter tape Height relascope Percent bark estimated Degree lean compass Height of cavity relascope Condition (live/dead, top intact/broken) Tree species Site Canopy height estimated Percent live canopy estimated Average tree size (dbh) diameter tape Basal area (live and dead) relascope Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 97 Data analyses Nest densities were estimated with line transect distance sampling methods (Buckland et al. 1993) using the program DISTANCE (Laake et al. 1993). I calculated densities separately for each year, and estimated separate detectability functions and densities for GNP and FNF. I tested for significant differences in nest densities between the 2 areas using Student's t-test. I used chi-square to test whether relative nest abundances differed significantly between GNP and FNF, and between partially cut and clearcut stands. Mann-Whitney U tests of mean ranks were used to determine if bird abundances, estimated from the point count data, differed significantly between the two cutting treatments. Mann-Whitney U tests were used to compare values of continuous habitat variables between areas. Categorical variables were compared between areas using Chi-square tests. Using standardized habitat variables, I calculated Euclidean distances between nest and random sites in GNP and FNF. Results Nest and bird abundances Three hundred and ninety-two active cavity nests were found along transects in GNP and 50 in FNF (Table 14) . Sample sizes in FNF were too small to allow density estimates. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 98 Table 14. Number of nests of each cavity-nesting bird species found in burned forests of Glacier National Park (GNP) and Flathead National Forest (FNF) from 1990-1994. Forty-eight km of transect in FNF and 210 km of transect in GNP were surveyed. The percentage of the total nests in each location are given in parentheses for each bird species. Common name Scientific name # nests # nests (%) in (%) in GNP FNF Tree Swallow Tachycineta bicolor 109 (28) 10 (20) Northern Flicker Colaptes auratus 69 (18) 14 (28) Mountain Bluebird Sialia currucoides 45 (12) 8 (16) Red-naped Sapsucker Sphyrapicus nuchalis 46 (12) 3 (6) Three-toed Picoides tridactylus 30 (8) 3 (6) Woodpecker Hairy Woodpecker Picoides villosus 16 (4) 4 (8) Downy Woodpecker Picoides pubescens 18 (5) 1 (2) Red-breas ted Sitta canadensis 17 (4) 2 (4) Nuthatch American Kestrel Falco sparverius 10 (3) 2 (4) Black-backed Picoides arcticus 7 (2) 1 (2) Woodpecker Mountain chickadee Parus gambeli 5 (1) 1 (2) House Wren Troglodytes aedon 6 (1) 0 (0) European Starling Stumus vulgaris 5 (1) 0 (0) Pileated Woodpecker Dryocopus pileatus 4 (1) 0 (0) Black-capped Parus atricapillus 3 (1) 0 (0) Chickadee Brown Creeper Certhia americana 2 (<1) 0 (0) Chestnut-backed Parus rufescëns 0 (0) 1 (2) Chickadee Total 392 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 99 for individual species, but total nest density was two times greater in GNP (.123 ± .060 nests/ha) than in FNF (.063 ± .022 nests/ha); this difference was significant (t = 2.0, df = 129, p < 0.05) . Nest densities did not differ significantly among years (p > 0.05). Relative nest abundances among species did not differ significantly between GNP and FNF 18.9, df = 16, p = 0.2730) . Within FNF, species relative nest abundances differed between partially cut and clearcut stands =19.8, df = 11, p < 0.05; Fig. 11). Nests of most species, especially tree-foraging insectivores, were found only in partially cut stands; a few species were more abundant in clearcuts. The sample was not large enough to estimate densities separately for each cutting type; nevertheless, the number of nests found per kilometer of transect surveyed in clearcut stands (1.3 nests/km) was much less than in partially cut stands (3.7 nests/km). This estimate does not take into account distance from transect, which is necessary to calculate actual densities. However, detectability distance is likely to be greater in clearcuts because of fewer trees to obstruct sounds and views. Differences in actual nest densities between the two treatments, therefore, are probably greater than indicated by the non-adjusted abundance data. Results from the point count data were similar to the nest data (Fig. 12). Power tests indicated that samples were Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 100 0 0J5 0.5 0.75 t I -L. I Three-toed Woodpecker- Red-naped Sapsucker- B Partial Cut Clearcut Red-breasted Nuthatch American Kestrel Black-backed Woodpecker Downy Woodpecker Chestnut-backed Chickadee- Mountain Chickadee- Mountain Bluebird- Hairy Woodpecker-j L Northern Flicker-iCVXNXXVO r Tree Swallow- I I I 0 0.25 0.5 0.75 # nests/km transect Figure 11. Active cavity nest abundances in areas of different logging intensities in a post-fire forest. Nests were found along 38 kilometers of transects. Bird species are listed in descending order based on the ratio of nests in partial cuts to clearcuts ; the difference in species distributions between the two treatments is significant (chi-square, p < 0.05). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 101 0.2 0.4 I I 0.6I Three-toed Woodpecker B Partial Cut Clearcut (n=180) (n=45) Black-backed WoodpeckerJ Chickadee spp. J Red-naped Sapsucker -\ r TREE-FORAGERS Red-breasted Nuthatch H Hairy Woodpecker *i Tree Swallow NON-TREE-FORAGERS American Kestrel H Northern Flicker [ Mountain Bluebird H F I I r 0.2 0.4 0.6 Average # per point Figure 12. Cavity-nesting bird abundances at point counts in partially cut and clearcut sites in post-fire habitat. Bird species are listed in descending order based on the ratio of abundance in partial cuts to clearcuts. As a group, tree-foraging species were significantly more abundant in selection cuts than in clearcuts (p = 0.001). Non-tree-foraging species did not differ significantly in abundance between the sites (p = .0630). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 102 not sufficient to detect significant differences in abundances for most species (power < 0.3 for all but Mountain Bluebird) , so I tested for differences between tree-foraging and non-tree-foraging species as groups. Non-tree-foraging species did not differ significantly in abundance between the two types of sites (p = 0.063); however, tree-foraging species were less abundant in clearcuts (.244 ± .0722 per point) than in partially cut stands (.776 ± .073 per point; p = 0.001). All tree-foraging speciesindividually were less abundant in clearcuts; non-tree-foragers showed mixed results. Habitat characteristics Unloaaed and logged forests Conditions of and around randomly selected trees in GNP and FNF differed significantly for almost all variables I measured (Table 15) . Trees in FNF (clearcut and partially cut sites'combined) were smaller in diameter and height, had less bark and lean, and were surrounded by stands of lower basal area, smaller average tree diameter, and shorter canopies. Proportionally more random trees in GNP were Douglas-fir and lodgepole pine, whereas in FNF more were western larch and subalpine fir (;^ = 76, df=8, p< 0.0001). Aspen made up 5% of randomly selected trees in GNP and 0% in FNF. Tree condition was the only variable that did not differ significantly (%^ = 1.6, df = 3, p = 0.6515). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 103 Table 15. Means ±SE of habitat variables recorded at randomly located trees and nest trees in burned, unlogged forests of Glacier National Park (GNP) and in burned, logged forests of Flathead National Forest (FNF) . P-values are for Mann-Whitney U tests of mean ranks. Variable GNP FNF P* Random trees n 153 133 Tree dbh (cm) 30 ± 1.3 20 ±1.54 <0.0001 Tree height (m) 21 ± 0.56 14 ±0.74 <0.0001 % baric 63 ±3.66 56 ±5.01 0.0004 ° lean 3.1 ±0.42 1.4 ±0.45 <0.0001 Ave. dbh in 23 ±0.64 19 ± 5.34 <0.0001 stand (cm) Basal areaa 15 ± 0.88 10 ± 0.51 <0.0001 (m2/ha) Canopy height 26 ± 0.55 17 ±0.43 <0.0001 (m) Mest trees n 392 50 Tree dbh (cm) 35 ± 0.56 40 ± 2.80 0.73 Tree height (m) 22 ±0.45 19 ±0.97 0.0001 % bark 57 ± 2.39 45 ±4.62 0.0001 ° lean 2.7 ± 0.32 5.7 ±1.15 0.0020 Ave. dbh in 29 ±0.57 21 ± 1.27 0.0001 stand (cm) Basal area 20 ±0.82 15 ±1.47 0.0001 (m2/ha) Canopy height 26 ± 0.25 19 ± 0.56 0.0001 (m) % live canopy 1.9 ± 0.32 2.8 ± 0.96 0.32 ^ Basal area includes both live and dead t r e e s . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 104 Most of the same habitat variables also differed significantly between nests in GNP and nests in FNF (Table 15). However, differences were less pronounced than for randomly selected trees, and nest trees in FNF were of greater dbh and lean than GNP nest trees, opposite the trends for randomly selected trees. In FNF, more nests were broken- top snags (49%) than in GNP (27%; = 12, df = A, p = 0.0151. In GNP, most nests (78%) were in aspen; nests in FNF were in western larch (36%), black cottonwood (16%), and subalpine fir (16%; = 166, df = 10, p < 0.0001). Cluster analysis showed GNP nest and random sites were more similar than FNF nest and random sites (Fig. 13). Habitat at nests differed less between the two areas than did habitat at random trees; in fact, conditions at FNF nests were more similar to those at random sites in GNP than to those at randomly selected sites in FNF. Partially cut and clearcut stands Conditions around randomly selected trees differed significantly in several respects between partially cut and clearcut stands in FNF (Table 16) . Trees in partial cuts were taller and surrounded by stands of greater basal area, average diameter (by rank) , and canopy height. Most trees in partially cut stands (89%) were intact snags, while in clearcuts, 40% of the trees left were broken--top snags Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 105 GNP nests 2.41 GNP random 2.89 FNF nests 3.74 FNF random Figure 13. Dendrogram of similarity between nest and random sites in burned, unlogged forest stands in Glacier National Park (GNP) and burned, logged stands in Flathead National Forest (FNF) . Euclidean distances between groups are derived from microhabitat variables of trees and surrounding forest stands; clusters are based on average linkages between groups. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 106 Table 16. Means ±SE of habitat variables recorded at randomly located trees and nest trees in partially cut and clearcut post-fire stands of Flathead National Forest. P-values are for Mann-Whitney U tests of mean ranks. Variable Partial cut Clearcut P* Random trees n 112 21 Tree dbh (cm) 21 ± 1.88 16 ± 0.94 0.067 Tree height (m) 15 ± 0.88 11 ± 0.83 0.013 % bark 56 ± 5.84 47 ± 10.7 0.590 ° lean 1.4 ± 0.49 0.3 ± 0.3 0.360 Ave. dbh in 19 ± 0.82 12 ± 1.85 0.0042 stand (cm) Basal areaa 11 ± 0.54 3 ± 0.57 <0.0001 (m2/ha) Canopy height 17 ± 0.46 12 ± 0.84 0.001 (m) Nest trees n 37 13 Tree dbh (cm) 43 ± 3.52 32 ± 3.03 0.12 Tree height (m) 19 ± 1.18 18 ± 1.71 0.41 % bark 8 ± 3.61 10 ± 6.82 0.61 ° lean 6.6 ± 1.5 3.3 ± 0.94 0.63 Ave. dbh in 24 ± 2.28 17 ± 1.02 0.039 stand (cm) Basal area 17 ± 1.81 7 ± 1.63 0.0025 (m2/ha) Canopy height 19 ± 0.66 17 ± 0.81 0.026 (m) a Basal area includes both live and dead trees. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 107 = 13, df = 2, p = 0.0012) . The distribution of random tree species did not differ significantly between the two treatments. Nest-tree characteristics were not significantly different between the two types of treatments ; however, stands around nest trees in partial cuts had significantly greater average tree diameter, canopy height, and basal area than nests in clearcuts (Table 16). Discussion Harris (1982) found timber-drilling woodpeckers were less abundant in burned, logged stands than in burned, unlogged stands. Blake (1982) reported that bark and foliage insectivores were less abundant in burned, logged stands than in burned, unlogged stands in Arizona. Hitchcox (pers. comm.) studied logged and unlogged post-fire stands in western Montana, and found lower nest densities and fewer suitable potential nest sites in logged stands. In this study, cavity-nesting bird abundances and nest densities were significantly lower in logged post-fire forest stands than in unlogged, burned stands. These differences may have been due to differences in available habitat. Forest stand characteristics (basal area, average tree size and canopy height) differed greatly between the logged and unlogged areas. Differences between used and randomly selected trees (assumed to represent available habitat) were more pronounced in the logged forest, indicating that fewer Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 108 sites were suitable for nesting. Birds may have responded to habitat alterations somewhat by shifting their nesting habits. This is suggested by different nest microhabitat conditions between logged and unlogged areas. However, nest microhabitat did not differ nearly as much between logged and unlogged areas as did microhabitat at randomly selected trees. Within logged areas, they selected sites that were most similar to nest sites in unlogged habitat. Logging intensity also was a significant factor in site use by some species. Even though logging took place on a relatively small scale (2-35 ha) in the FNF, and potential nest snags were left in clearcuts, a negative effect on tree- foraging species was evident. Logging reduced the number of potential nest trees of cavity-nesting birds, which have fairly restrictive nesting requirements (Shigo and Kilham 1968, Conner et al. 1976, McClelland 1977, Raphael and White 1984, Schrieber and deCalesta 1992). Leaving snags in clearcuts to serve as potential nest trees has been shown to mitigate logging effects for some cavity-nesting bird species (Scott 1979, Dickson et al. 1983, Marcot 1983), and in this study some species nested in snags left in clearcuts. However, logging also removes foraging substrates used by many insectivorous cavity-nesting species. Even if a few trees are left to serve as nest substrates, some species may be limited by food. Changes in food availability may be the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 109 primary factor influencing use of post-fire habitat by many cavity-nesting bird species (Chap. II) . Lower nest abundances in clearcuts, and differences in relative species abundances between clearcut and partially cut stands appears to be due to variation in conditions around trees (forest stand variables) rather than to characteristics of the nest trees themselves. There were no significant differences in nest tree characteristics between the two treatments. However, average tree size, basal area, and canopy height were all significantly greater in partial cuts. Fewer appropriate trees may have been available for nesting in clearcuts. Even available (random) trees, however, differed only in height between the two treatments; trees in clearcuts were shorter, probably because more of them were broken-topped. Because broken-topped snags were used more often than expected for nests (Chap. Ill), it seems unlikely that this factor made trees in clearcuts less suitable. Thus, for cavity nesters that forage primarily on standing trees, logging practices that remove a large portion of standing fire-killed trees may have particularly detrimental effects. Such effects are not likely to be mitigated simply by leaving a few trees as nesting substrates. Most tree-foraging cavity nesters are excavators that create nest holes used by other species. Their low numbers may ultimately contribute to lower densities of nonexcavating species. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 110 Large portions of public lands sire salvage logged after fires, under the assumption that removal of burned trees has minimal effects on wildlife. A clear example that the assumption of "no effect" is invalid comes from the Black- backed Woodpecker, a species found primarily in forests affected by fire or insect infestations (Hutto 1995). This species is listed as a Sensitive Species by the Northern Region of the U.S. Forest Service, primarily because it is relatively restricted to burned forests. This species was less abundant in my intensely logged post-fire stands, and other timber-drilling woodpeckers showed similar trends. In fact, those species most rare in the cavity nester community tended to be those most positively affected by fire (Chap. II) and most negatively affected by salvage logging. Currently, logging of burned forests may legally proceed with no public comment, and "salvage" sales are promoted as environmentally harmless resource extraction. This policy does not reflect the importance of post-fire habitat to cavity-nesting birds and the impact of post-fire salvage on species like the Black-backed Woodpecker. Fire suppression has severely reduced the areal coverage of recently burned forests on Rocky Mountain landscapes (Gruell 1983, Agee 1993). Much post-fire habitat is unavailable for wildlife due to the prevalent "salvage" cutting of public lands. Salvage logging removes precisely the element that makes recently burned forests unique and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ill valuable to cavity-nesting birds— standing dead trees. Fire-created habitat where little or no logging takes place may become réfugia for species facing rapidly dwindling sources of standing dead wood. Literature Cited Agee, J. K. 1993. Fire ecology of Pacific Northwest forests. Island Press, Washington, D. C. Amo, S. F . 1980. Forest fire history in the Northern Rockies. J. For. 78:460-465. Bell, G. and M. J. Lechowicz. 1991. The ecology and genetics of fitness in forest plants. I. Environmental heterogeneity measured by explant trials. J. Ecol. 79:663-695. Blackford, J. L. 1955. Woodpecker concentration in a burned forest. Condor 57:28-30. Blake, J. G . 1982. Influence of fire and logging on nonbreeding bird communities of ponderosa pine forests. J. Wildl. Manage. 46(2): 404-415. Bock, C. E., and J. F . Lynch. 1970. Breeding bird populations of burned and unbumed conifer forest in the Sierra Nevada. Condor 72:182-189. Boecklen, W. J. 1986. Effects of habitat heterogeneity on the species-area relationships of forest birds. J. Biogeography Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 112 3:59-68. Buckland, S. T., D.R. Anderson, K.P. Bumham, and J.L. Laake. 1993. Distance Sampling. Chapman and Hall, London. Catt, D. J. 1991. Bird communities and forest succession in the subalpine zone of Kootenay National Park, British Columbia. M.S. Thesis, Simon Fraser Univ., British Columbia. Conner, R. N., 0. K. Miller, and C. S. Adkisson. 1976. Woodpecker dependence on trees infected by fungal heart rots. Wilson Bull. 88 (4) :575-581. Davis, P. R. 1976. Response of vertebrate fauna to forestfire and clear-cutting in south central Wyoming. Ph.D. Diss., Univ. of Wyoming, Laramie. Dickson, J. G., R. N. Conner, and J. H. Williamson. 1983. Snag retention increases bird use of a clear-cut. J. Wildl. Manage. 47(3):799-804. Evans, W. G . 1966. Perception of infra-red radiation from forest fires by Melanophila acuminata DeGeer (Coleoptera: Buprestidae). Ecoogy 47:1061-1065. Greenberg, C.H., D.G. Neary, L.D. Harris. 1994. Effects of high- intensity wildfire and silvicultural treatments on reptile communities in sand-pine scrub. Cons. Biol. 8:1047- Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 113 1057. Gruel1, G. E. 1983. Fire and vegetative trends in the Northern Rockies: interpretations from 1871-1982 photographs. General technical report INT-150. U.S. Forest Service, Ogden, Utah Habeck, J. R., and R. W. Mutch. 1973. Fire-dependent forests in the Northern Rocky Mountains. J . Quaternary Res. 3 (3) -.408-424. Harris, M. A. 1982. Habitat use among woodpeckers in forest bums. M. S. Thesis, Univ. of Montana, Missoula. Hejl, S. J. 1994. Human-induced changes in bird populations in coniferous forests in western North America during the past 100 years. Studies in Avian Biology 15:232-246. Hejl, S. J. , R.L. Hutto, C.R. Preston, and D.M. Finch. 1995. The effects of silvicultural treatments on forest birds in the Rocky Mountains. Pages 220-244 In Population ecology and conservation of neotropical migratory birds (T. Martin, and D.M. Finch, eds.). Oxford University Press, New York. Hutto, R. L. 1995. Composition of bird communities following stand-replacement fires in Northern Rocky Mountain (U.S.A.) conifer forests. Cons. Biol. 9(5):1041-1058. Hutto, R. L., S. M. Pletschet, and P. Hendricks. 1986. A Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 114 fixed-radius point count method for nonbreeding and breeding season use. Auk 103:593-602. Key, C. 1988. Red Bench Fire Intensity Map. Unpub. rep., Glacier National Park, Montana. Laake, J. L., S. T. Buckland, D. R. Anderson, and K. P. Burnham. 1993. DISTANCE. Colorado Cooperative Fish and Wildlife Research Unit, Colorado State Univ., Fort Collins. Marcot, B . G. 1983. Snag use by birds in Douglas-fir clearcuts. Pages 134-13 9 in Proceedings of Snag Habitat Management Symposium at Northern Arizona Univ., Flagstaff, AZ. (J.W. Davis, G.A. Goodwin, and R.A. Ockenfels, Technical coordinators) U.S. For. Serv. Rocky Mtn. For. Range Exp. Sta. Gen. Tech. Rep. RM-99. 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. diss., Univ. Of Montana, Missoula. Mitchell-Olds, T. 1992. Does environmental variation maintain genetic variation? A question of scale. Tree 7(12):397-398. Pfister, A. R. 1980. Postfire avian ecology in Yellowstone National Park. M. S. Thesis, Univ. of Washington, Pullman. Raphael, M. G., and M. White. 1984. Use of snags by Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 115 cavity-nesting birds in the Sierra Nevada. Wildl. Monogr. 86 :1- 66 . Schreiber, B., and D.S. deCalesta. 1992. The relationship between cavity-nesting birds and snags on clearcuts in western Oregon. Forest Ecology and Management 50:299-316. Scott, V. E. 1979. Bird response to snag removal in ponderosa pine. J. For. 77(l):26-28. Shigo, A. L., and L. Kilham. 1968. Sapsuckers and Pomes ignarius var. Populinus. U. S. For. Ser. Res. Note NE-84. NE Forest Exp. Sta., Upper Darby, Pa. Simberloff, D., and L. G. Abele. 1982. Refuge design and island biogeography theory : effects of fragmentation. Amer. Nat. 120:41-50. Taylor, D. L., and J. W. J. Barmore. 1980. Post-fire succession of avifauna in coniferous forests of Yellowstone and Grand Teton National Parks, Wyoming. Pages 130-145 in Management of western forests and grasslands for nongame birds. U.S.D.A. For. Ser. Gen. Tech. Rep. INT-86. Reproduced with permission ot the copyright owner. Further reproduction prohibited without permission.