Using culturally significant birds to guide the timing of prescribed fires in the Klamath Siskiyou Bioregion

Abstract

Abstract Historically, wildfire and tribal burning practices played important roles in shaping ecosystems throughout the Klamath Siskiyou Bioregion of northern California and southern Oregon. Over the past several decades, there has been increased interest in the application of fire for forest management through the implementation of prescribed fires within habitats that are used by a diversity of migrant and resident land birds. While many bird species may benefit from habitat enhancements associated with wildfires, cultural burning, and prescribed fire, individuals may face direct or indirect harm. In this study, we analyzed the timing of breeding and molting in 11 species of culturally significant land birds across five ecologically distinct regions of northern California and southern Oregon to explore the potential timeframes that these bird species may be vulnerable to wildland fires (wildfire, prescribed fire, or cultural burning). We estimated that these selected species adhered to a breeding season from April 21 to August 23 and a molting season from June 30 to October 7 based on bird capture data collected between 1992 and 2014. Within these date ranges, we found that breeding and molting seasons of resident and migratory bird species varied temporally and spatially throughout our study region. Given this variability, spring fires that occur prior to April 21 and fall fires that occur after October 7 may reduce the potential for direct and indirect negative impacts on these culturally significant birds across the region. This timing corresponds with some Indigenous ecocultural burning practices that are aligned with traditionally observed environmental cues relating to patterns of biological phenology, weather, and astronomy. We detail the timing of breeding and molting seasons more specific to regions and species, and estimate 75%, 50%, and 25% quartiles for each season to allow for greater flexibility in planning the timing of prescribed fires and cultural burning, or regarding the potential implications of wildfires. The results of our study may serve as an additional resource for tribal members and cultural practitioners (when examined within the context of Indigenous Traditional Ecological Knowledge) and forest and wildland fire managers to promote stable populations of culturally significant bird species within fire‐dependent forest systems.

Full text

ARTICLE
Socio-Ecological Systems
Using culturally significant birds to guide the timing of
prescribed fires in the Klamath Siskiyou Bioregion
Linda L. Long
1
| Frank L. Lake
1
| Jaime L. Stephens
2
|
John D. Alexander
2
| C. John Ralph
1
| Jared D. Wolfe
3
1
USDA Forest Service, Pacific Southwest
Research Station, Arcata, California, USA
2
Klamath Bird Observatory, Ashland,
Oregon, USA
3
College of Forest Resources and
Environmental Science, Michigan
Technological University, Houghton,
Michigan, USA
Correspondence
Linda L. Long
Email: linda.l.long@usda.gov
Funding information
National Park Service; Oregon State
University’s Office of International
Programs; U.S. Bureau of Land
Management; U.S. Forest Service; Wildlife
Images; Klamath Bird Observatory
Handling Editor: Alisa Coffin
Abstract
Historically, wildfire and tribal burning practices played important roles in
shaping ecosystems throughout the Klamath Siskiyou Bioregion of northern
California and southern Oregon. Over the past several decades, there has been
increased interest in the application of fire for forest management through the
implementation of prescribed fires within habitats that are used by a diversity
of migrant and resident land birds. While many bird species may benefit
from habitat enhancements associated with wildfires, cultural burning, and
prescribed fire, individuals may face direct or indirect harm. In this study, we
analyzed the timing of breeding and molting in 11 species of culturally
significant land birds across five ecologically distinct regions of northern
California and southern Oregon to explore the potential timeframes that these
bird species may be vulnerable to wildland fires (wildfire, prescribed fire, or
cultural burning). We estimated that these selected species adhered to a breed-
ing season from April 21 to August 23 and a molting season from June 30 to
October 7 based on bird capture data collected between 1992 and 2014. Within
these date ranges, we found that breeding and molting seasons of resident and
migratory bird species varied temporally and spatially throughout our study
region. Given this variability, spring fires that occur prior to April 21 and fall
fires that occur after October 7 may reduce the potential for direct and indirect
negative impacts on these culturally significant birds across the region. This
timing corresponds with some Indigenous ecocultural burning practices that
are aligned with traditionally observed environmental cues relating to patterns
of biological phenology, weather, and astronomy. We detail the timing of
breeding and molting seasons more specific to regions and species, and
estimate 75%, 50%, and 25% quartiles for each season to allow for greater
flexibility in planning the timing of prescribed fires and cultural burning, or
regarding the potential implications of wildfires. The results of our study may
serve as an additional resource for tribal members and cultural practitioners
(when examined within the context of Indigenous Traditional Ecological
Received: 6 October 2020 Revised: 13 July 2021 Accepted: 6 August 2021
DOI: 10.1002/ecs2.4541
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided
the original work is properly cited.
© 2023 The Authors. Ecosphere published by Wiley Periodicals LLC on behalf of The Ecological Society of America.
Ecosphere. 2023;14:e4541. https://onlinelibrary.wiley.com/r/ecs2 1of27
https://doi.org/10.1002/ecs2.4541

Knowledge) and forest and wildland fire managers to promote stable
populations of culturally significant bird species within fire-dependent forest
systems.
KEYWORDS
bird, breed, Indigenous people, Klamath Siskiyou Bioregion, molt, wildland fire
INTRODUCTION
Indigenous peoples have lived along the northern Pacific
coast for millennia where they integrate burning as an
ecological process into cultural stewardship practices
(Boyd, 2022; Matson & Coupland, 1995). Historical
Indigenous fire use ultimately shaped forest physiog-
nomy while contemporary use of low-intensity fire in cul-
tural burns remains an important stewardship practice
among coastal tribes in Oregon and California, as exem-
plified by the Klamath River Prescribed Fire Training
Exchange, Indigenous Peoples’Burning Network,
Cultural Fire Management Council, and others (Buono,
2020;Crawfordetal.,2015;Longetal.,2018; Walsh et al.,
2015). These tribal burning practices are important
because they promote food security while maintaining
resilient ecosystems (Lake & Long, 2014;Longetal.,
2018). For example, the maintenance of open-canopy oaks
(Quercus spp.) through cultural burning provides tradi-
tional sources of food, thereby advancing cultural restora-
tion while preserving traditional fire knowledge
(Anderson, 2007;Huffman,2013). Also, low-intensity fires
diminish conifer recruitment while promoting oak
savanna landscapes, one of the most imperiled habitats
throughout the Klamath Siskiyou Bioregion of southern
Oregon and northern California (Altman & Stephens,
2012). Forest landscapes in the Klamath Siskiyou
Bioregion have been shaped by both wildfire (Taylor &
Skinner, 1998) and tribal burning practices (Lake, 2013;
Pullen, 1996) and host numerous culturally significant
bird species. Some bird species have recently experienced
population declines attributed to habitat degradation asso-
ciated with fire suppression and industrial scale-timber
management, among other factors (Table 1;Altman,2011;
Sauer et al., 2017). The absence of naturally occurring fire
and Indigenous burning has altered the habitat and com-
position of bird communities in comparison to similar
areas where fire has not been excluded (Marshall, 1963).
Similar to other regions globally (Whitehead et al.,
2003), birds in the Klamath Siskiyou Bioregion have
long been an integral part of tribal lifeways, Indigenous
traditional ecological knowledge (ITEK), and cultural
values, serving as indicators for ecological phenomena,
habitat quality, and environmental and seasonal changes
(Appendix S1: Table S1; Anderson, 2005; Kroeber &
Gifford, 1980). For example, patterns of avian lifecycle
phenology have been captured in traditional stories that
link seasonal bird behavior with tribal use of different
habitats and species. Birds are also commonly used in
spiritual regalia and ceremonies (Gleeson et al., 2012;
Riedler et al., 2012), and are revered, highly sought after,
and considered sacred by tribes throughout northern
California and southern Oregon (Long et al., 2018).
Further, avian species of cultural importance represent a
set of beliefs or “character”traits that form avian tradi-
tional knowledge, where some species are the focus of
story-teaching lessons and ethics of stewardship or used
for regalia and food (Anderson, 2005).
One aspect of tribal values for birds is as an indicator
to assure human responsibility is at the forefront in
limiting the impact of human fire use and stewardship
practices on the reproductive rights in nature. The impor-
tance of birds to regional tribal philosophies led to a
belief system that among some tribes includes the prac-
tice of prohibiting larger landscape burning that would
negatively affect birds in the spring and early summer
during mating, reproduction, or nesting (Long et al.,
2020, see also Mistry et al., 2016 for South American
tropical savannas). Because many bird species have
co-evolved with and benefit from fire and cultural burn-
ing, the time of their nesting can be used by cultural
practitioners as an indicator for when to cease burning at
larger spatial scales (see Karuk Tribe, 2019:77, fig. 3.8).
Among the belief systems of some Karuk tribal commu-
nity members and as the current policy of the Karuk
Tribe (residing in the mid-Klamath River and western
Klamath Mountain regions of northern California),
cultural burning is not conducted when the Pleiades star
cluster disappears and should stop when certain birds
indicate it is time to do so, only to start when other indi-
cators present themselves. During this time, fire is only
used in the context of heating and cooking (William
Tripp citing Karuk ITEK, personal communication).
“The appearance of Pleiades in the night sky denotes
the time for cultural burning …This knowledge gained
from attending to the land over generations is inscribed
in ceremonies and prayers”(Karuk Tribe, 2019:58).
Additionally, birds may be used as indicator species to
2of27 LONG ET AL.

TABLE 1 Bird species of cultural significance to tribes captured in mist nets in northern California and southern Oregon, 1992–2014.
Common name Scientific name BBS trend BBS cred.
a
Total captures
Residents
Mountain Quail Oreortyx pictus −1.62 1 15
California Quail Callipepla californica 0.56 1 159
Cooper’s Hawk Accipiter cooperii 2.06 3 2
Western Screech-Owl Otus kennicottii 2.02 3 32
Northern Pygmy-Owl Glaucidium gnoma 1.85 3 6
Northern Saw-whet Owl Aegolius acadicus
b
…7
Belted Kingfisher Ceryle alcyon −0.57 2 11
Acorn Woodpecker Melanerpes formicivorus 0.65 1 8
Red-breasted Sapsucker Sphyrapicus ruber 3.05 1 474
Downy Woodpecker Dryobates pubescens −1.20 1 392
Hairy Woodpecker Dryobates villosus 0.72 1 123
White-headed Woodpecker Dryobates albolarvatus −1.44 2 7
Northern Flicker Colaptes auratus −0.65 1 133
Pileated Woodpecker Dryocopus pileatus 1.06 2 3
Steller’s Jay Cyanocitta stelleri −0.10 1 318
California Scrub-Jay Aphelocoma californica 1.16 1 179
Pacific Wren Troglodytes pacificus −0.72 1 361
American Dipper Cinclus mexicanus 1.26 3 11
Golden-crowned Kinglet Regulus satrapa −1.64 1 197
Western Bluebird Sialia mexicana −3.51 1 13
Mountain Bluebird Sialia currucoides −5.35 3 2
American Robin Turdus migratorius −0.52 1 1636
Varied Thrush Ixoreus naevius −2.76 1 195
Evening Grosbeak Coccothraustes vespertinus 2.56 1 1
Pine Siskin Spinus pinus −1.22 1 518
Lesser Goldfinch Spinus psaltria −1.51 1 405
American Goldfinch Spinus tristis −2.26 1 1758
Spotted Towhee Pipilo maculatus −0.23 1 2745
Dark-eyed Junco Junco hyemalis −1.29 1 3915
Red-winged Blackbird Agelaius phoeniceus −1.69 1 31
Yellow-rumped Warbler Setophaga coronata −0.42
c
1 1559
Migrants
Black-chinned Hummingbird Archilocus alexandri
b
…1
Anna’s Hummingbird Calypte anna 8.17 2 182
Rufous Hummingbird Selasphorus rufus −2.56 1 560
Allen’s Hummingbird Selasphorus sasin −3.79 1 869
Calliope Hummingbird Stellula calliope 7.39 3 27
Tree Swallow Tachycineta bicolor −2.19 1 206
Violet-green Swallow Tachycineta thalassina −0.64 1 110
Barn Swallow Hirundo rustica −2.84 1 645
Ruby-crowned Kinglet Regulus calendula 0.04 2 633
Northern Mockingbird Mimus polyglottos 0.10 3 1
(Continues)
ECOSPHERE 3of27

guide burns at microscales, within specific vegetation
communities, such as savannah oak stands (Altman &
Stephens, 2012: fig. 2). Tribal fires during certain cultur-
ally determined timeframes are applied in relatively small
areas for vegetation management, such as clearing brush,
maintaining meadows, and enhancing the production of
basketry materials, heating, and cooking (Anderson &
Moratto, 1996; Lake, 2013), whereas at other times, fire is
used across larger spatial scales. Timing of cultural burns
is often guided by specific environmental cues that sug-
gest optimal conditions for fire initiation (Anderson &
Moratto, 1996; Lake et al., 2010). Such cues relate to
tribal belief systems that birds are omens or messengers
from the Creator, and as such are teachers to humans
regarding culturally appropriate conduct for stewardship
activities, such as the timing (within seasons) and speci-
ficity (habitat type/vegetation) that guides cultural burn-
ing considerations. However, birds are not the only
indicator for when fire is or is not used. For example, the
Karuk Tribe traditionally sets the World Renewal
Ceremony fires on certain mountains during certain
phases of the lunar cycle in late summer and early fall
(Karuk Tribe, 2010; William Tripp citing Karuk ITEK,
personal communication).
By contrast, Euro-American colonization and subse-
quent US fire policies in the early 1900s led to forest
management characterized by fire suppression as “the
first measure necessary for the successful practice of
forestry”(Graves, 1910:7), often leading to vegetation and
habitat changes as the result of excluding all fires (includ-
ing cultural burning). In drier western forests where fre-
quent fires previously occurred from both natural and
Indigenous sources, fire suppression led to increased
shrub and tree densities contributing to excessive fuel
accumulations (Knight et al., 2020,2022; Ryan et al.,
2013) or changes in successional patterns and increased
levels of surface fuels (Parsons & DeBenedetti, 1979).
Where ecotones occurred, such as in open oak savannas
and woodlands near coniferous forests (Altman &
Stephens, 2012) or juniper/prairie ecotones in the mid-
western United States (DeSantis et al., 2011), reduction in
fires allowed woody vegetation, generally fire-intolerant
species such as Douglas fir or junipers, to encroach
and create closed forests, thereby diminishing open
TABLE 1 (Continued)
Common name Scientific name BBS trend BBS cred.
a
Total captures
Cedar Waxwing Bombycilla cedrorum −1.91 1 460
Yellow-breasted Chat Icteria virens 0.41 1 2814
Bullock’s Oriole Icterus bullockii −1.43 2 469
Orange-crowned Warbler Oreothlypis celata −1.09 1 1763
Nashville Warbler Oreothlypis ruficapilla −1.90 1 692
MacGillivray’s Warbler Geothlypis tolmiei −2.04 1 3957
Common Yellowthroat Geothlypis trichas 0.24 1 93
Yellow Warbler Setophaga petechia −1.67 1 2806
Townsend’s Warbler Setophaga townsendi 0.37 1 18
Hermit Warbler Setophaga occidentalis −0.37 1 461
Wilson’s Warbler Cardellina pusilla −1.36 1 3250
Western Tanager Piranga ludoviciana 1.99 1 827
Black-headed Grosbeak Pheucticus melanocephalus 0.70 1 2080
Lazuli Bunting Passerina amoena −0.20 1 426
Grand total 38,566
Note: Breeding Bird Survey population trends (BBS trend) between 2005 and 2015 for North American Bird Conservation Region 5 (Northern Pacific Rainforest
[NABCI, 2020]) are shown as yearly percent change with BBS credibility category (BBS cred.; Sauer et al., 2017). BBS credibility categories incorporate the
potential for problems with population change estimates due to small sample sizes, low relative abundances on survey routes, imprecise trends, and missing
data. Species are grouped by migratory status: residents (i.e., nonmigrants) and migrants. Total captures is the number of adults captured in mist nets. Species
in boldface were selected for further analyses.
a
BBS credibility category: (1) highest credibility: reflects data with moderate abundance on routes, at least 14 samples in the long term, and of moderate
precision; (2) moderate credibility: reflects data with a deficiency with regional abundance, small sample size, or imprecise estimate; (3) lowest credibility:
reflects data with an important deficiency with very low abundance, very small samples, or very imprecise.
b
No data.
c
Both subspecies.
4of27 LONG ET AL.

grasslands (see Knight et al., 2020,2021,2022). Skinner
(1995), documenting forest mosaic changes after 40 years
of fire suppression between 1944 and 1985, reported that
forest openings in northern California mixed-conifer for-
ests were fewer and smaller. Similarly, the western
Klamath Mountains in northern California have more
contiguous landscape biomass–vegetation now as a result
of fire exclusion (Knight et al., 2021).
Early in the 1930s, forest managers in the southeast-
ern United States recognized the ecological benefits of
prescribed fire for the maintenance of upland game habi-
tat. Other regions of the United States were slower to pro-
mote this land management technique (Ryan et al., 2013;
Stephens & Ruth, 2005), but over the past two decades, it
has quickly become a tool used by forest managers, with
the number of prescribed fires set in the United States by
federal and state agencies increasing from 14,000 fires in
1998 totaling 355,000 hectares to over 450,000 fires total-
ing over 2.4 million hectares in 2018 (National
Interagency Fire Center, 2019). The greatest proportion
of the reported increase in prescribed fire was on tribal
lands, as tribes seek to re-introduce intentional fire on
the landscape (Kolden, 2019). Recognizing that fire
played an integral role in shaping ecosystems along the
northern Pacific coast (Huff et al., 2005), state and federal
agencies are integrating prescribed fire into forest man-
agement plans in the western United States to reduce fuel
loads, restore ecosystems, and enhance forest structure to
a desired condition (Agee, 2007; Huff et al., 2005;Long
et al., 2017; Ryan et al., 2013). At the same time, tribes
are both conducting prescribed fires with partners and
seeking to reinstitute cultural burns, working with agen-
cies and other fire use entities to reduce barriers to imple-
mentation (Clark et al., 2022; Karuk Tribe, 2010; Long
et al., 2018; Senos et al., 2006).
It has been asserted that cultural burning should not
be categorized with western Colonial prescribed fire
(Clark et al., 2022) and a prescribed fire may or may not
have cultural objectives. It is important to note and
distinguish the difference between cultural burning and
prescribed fire. Clark et al. (2022:3) state “Both [involve]
the act of setting fire to a specific landscape to achieve
a desired outcome, including fuel reduction and wildlife
habitat improvement. However, cultural burning and
prescribed fire are distinct concepts and are often
conducted by different groups for different purposes.
Prescribed fire is implemented based on a ‘prescription’
derived from models to determine conditions for burning.
Especially when state [or federal] agencies are involved,
prescribed fire typically includes the production of a burn
plan, smoke management plan, and completion of envi-
ronmental impact analysis. Cultural burning is typically
less formal and is integrative of holistic knowledge of
place to guide the timing and implementation of burning
activities. Cultural burning implies the purposeful use of
fire by a cultural group …for a variety of purposes and
outcomes.”
In the Klamath Siskiyou Bioregion, as well as across
the United States and other regions of the world, there is
heightened interest in evaluating the use of prescribed
fire by government agencies and local organizations
(e.g., USDA Forest Service and Fire Safe Councils)
for fuels and fire risk reduction. Wildland fires are poten-
tially detrimental to birds, several of which are culturally
significant indicators in the tribal belief systems about
the ethical uses of fire (Long et al., 2020). Conversely,
wildland fire may also provide benefits to birds (Bagne &
Purcell, 2011; Saab & Powell, 2005; Stephens et al., 2019).
By contrast, our objective in this study—which is
a collaborative partnership among tribal/Indigenous,
agency, academics, and nongovernmental organizations
(NGOs)—is to provide sound science to inform manage-
ment and address emergent challenges and to formulate
a better understanding of the potential effects of wildland
fires on land birds. The intent is not to “validate”tribal
knowledge or belief systems, but rather to explore the
implications of those concerns which have been raised by
tribes and forest and wildland fire managers in our study
region, or other Indigenous peoples globally (Mistry
et al., 2016; Whitehead et al., 2003). Cultural concerns
and ITEK relating to birds and other indicators may help
to inform prescribed fire and cultural burning practi-
tioners and may result in limiting impacts to birds of
both cultural importance and conservation concern.
Currently, state and federal agencies, tribes, and other
cooperative burn entities (i.e., NGOs) determine the best
times for setting these fires based primarily on air quality
regulations, authorized versus potential burn days, fire
personnel and resource availability, as well as environ-
mental, biophysical, or ecological variables such as fuels
treatment history, fuel loading, temperature, humidity,
wind, time of day, and seasonal restrictions (e.g., limited
operating periods) for sensitive wildlife species (Knapp
et al., 2009; Quinn-Davidson & Varner, 2012; Ryan et al.,
2013). Thus, prescribed fires are increasingly being
conducted in the spring to early summer and mid-to-late
fall in the western United States when “controlled”burns
(prescribed fires) are less likely to exceed intended sever-
ities and extents and fire personnel are more available
when not engaged in fire suppression or wildfire manage-
ment. As a result, many prescribed fires in northern
California and southern Oregon are implemented during
times when fire would have historically been likely
excluded by some tribes (e.g., late spring/early summer),
such as those tribal community members who hold such
beliefs among the Karuk Tribe (Karuk Tribe, 2010).
ECOSPHERE 5of27

However, there are also some examples where federal
agencies are aligning their collaborative burning efforts
with Indigenous indicators (see USDA Forest Service,
2018). Recent studies of low-to-mid elevation forests
show a low historical presence of fire scars in earlywood,
which would also suggest that Indigenous burning prac-
tices limited the extent of spring to mid-summer burns,
especially those of higher severity that would cause tree
scarring (Knight et al., 2022).
The timing of contemporary prescribed fires may coin-
cide with breeding or molting in land birds, which are
energetically taxing and vulnerable phases of the avian
lifecycle, and not often considered in fire planning (Huff
et al., 2005; Knapp et al., 2009, but see Ryan et al., 2013).
Fires during the nesting season may reduce populations
more than burning in other seasons (Lyon et al., 2000).
Direct effects of fire on birds during the breeding season
include destruction of active nests and mortality of young
or adults (though adults can generally escape fires;
Bagne & Purcell, 2009; Knapp et al., 2009). Besides direct
effects, in the short-term food resources and cover for
some species may become scarce depending on the scale
and severity of the burn (Lyon et al., 2000), while
long-term consequences include the displacement of some
species while other species may take advantage of new
post-fire resources (Huff & Smith, 2000; Knapp et al.,
2009). Bird nest site selection, territory establishment, and
nesting success can also be directly and negatively affected
by fire (Lyon et al., 2000). Ground-dwelling birds may be
affected by fires of any severity while canopy-dwelling
birds may not be as affected by understory, lower intensity
burns (Lyon et al., 2000).
Fires occurring during the prebasic (fall) molt—an
energetically taxing period when adult birds completely
replace their feathers—could directly endanger individ-
uals, particularly during periods of heavy molt when a
bird’s capacity to fly is diminished and they become less
capable of escaping fire (Swaddle & Witter, 1997). Fire can
also indirectly affect birds by reducing arthropod and fruit
availability during the prebasic molt when birds rely on
abundant food resources. Yet, the magnitude of the impact
to a bird’s diet may vary by habitat quality. For example,
while some food resources may decrease post-fire, others
may not be impacted or may increase, including
fire-adapted shrubs such as manzanita (e.g., sticky leaf
manzanita, Arctostaphylos viscida)(Fryer,2015)andselect
arthropod species, which may be attracted to fire (Huff &
Smith, 2000).
The energetic demands during molt are substantially
greater compared to daily maintenance needs at a time
when disruption in food availability during burns could
put additional metabolic stress on birds. Murphy and
Taruscio (1995) reported that the daily increase in
whole-body protein synthesis in molting White-crowned
Sparrows (Zonotrichia leucophrys) equaled at least a
3.5-fold increase over daily synthesis by non-molting
sparrows. Similarly, Murphy and King (1992) calculated
a daily energy cost for peak molt equal to 58% of basal
metabolic rate in addition to daily energy costs. Heise
and Rimmer (2000) reported that during late molt stages,
Gray Catbirds (Dumetella carolinensis) increased their
foraging, which coincided with significant increases in
fat stores.
To balance the ecological benefits, sociocultural
values, and management objectives of prescribed fires
and cultural burning with the inevitable complexities of
implementation, we synthesized data from long-term
scientific bird monitoring to inform the planning and
timing of fire use that better informs the potential of
fire-related impacts to those land bird species that are
an important part of the cultural heritage of local
tribes (Appendix S1: Table S1). According to the Karuk
Tribe (2019:60), “burn timing follows a gradient that
tracks the reproductive lifecycles of season and elevational
migrant species …[and] the nesting of birds. William
Tripp describes how the Karuk practice of careful observa-
tion is critical to this process: ‘When the birds come back
and nest it is time to move upriver or upslope with your
burning.’Fire management occurs working uphill in the
Spring along this gradient of reproductive timing [for resi-
dent and migratory land birds].”
There are reasons to consider, from an Indigenous
tribal perspective, that our study species may serve as
bioindicators for other bird species which may be more
susceptible to fires and potentially vulnerable to environ-
mental or climatic impacts (Long et al., 2020). Examples
of bioindicator species correlations are well established
in existing western scientific research and monitoring
(Chase & Geupel, 2005; Stephens et al., 2019), but not
well addressed by Indigenous peoples (Long et al., 2020,
see also Karuk Tribe, 2019). For example, Saab and
Powell (2005), in compiling results from multiple studies
of the effects of fire (wildfire and prescribed fires) on
birds, reported similarities within avian foraging and
nesting guilds. Aerial, ground, and bark insectivores were
positively influenced by fire, whereas foliage gleaners
were negatively influenced. Additionally, they reported
that species with closed nests were more positively
influenced by fire than those with open-cup nests, and
ground and canopy nesters more positively influenced
than shrub nesters.
In this study, we summarize the timing of breeding
and molting for 11 culturally significant land bird species
that regularly occur in five ecologically distinct regions in
the Klamath Siskiyou Bioregion of northern California
and southern Oregon. Our estimates of breeding and
6of27 LONG ET AL.

molting seasonality were derived from a long-term bird
banding dataset from the Klamath Bird Monitoring
Network (Alexander, 2011), at the center of which is the
long-standing investment in avian research by the US
Forest Service Pacific Southwest Research Station in
cooperation with Klamath Bird Observatory. The aim of
our analysis is to provide tribes and forest and wildland
fire managers with the best available science to support
efforts to better understand the potential impacts of the
timing of fire use on culturally significant bird species.
MATERIALS AND METHODS
We conducted our study in the Klamath Siskiyou
Bioregion of southern Oregon and northern California
(Figure 1) as described by Alexander et al. (2017). We
operated 96 bird banding stations between 1992 and
2014, each station operating from 2 to 22 years (Table 2).
Data were collected by multiple cooperators as part of the
Klamath Bird Monitoring Network (Alexander, 2011).
Stations were operated from May through October. From
May through August (breeding season), stations were
scheduled once every 10-day period and from September
through October (fall migration season), stations were
scheduled at least once per week (Ralph et al., 1993;
Stephens et al., 2010). Each station had 8–15 net sites that
were opened 15 min prior to sunrise and operated for
5–6 h during each sampling day. Banding stations were
typically placed in a water-associated or meadow riparian
zone to maximize bird capture rate. We grouped stations
into five “regions,”defined by elevation and proximity to
one another, thus reflecting similar habitats (Figure 1,
Table 2). Captured individuals were aged and sexed fol-
lowing Pyle (1997). Banding methods followed Ralph et al.
(1993). We followed US Federal Regulations as outlined by
the USGS Bird Banding Laboratory (BBL) (2019b) and its
attached documentation for obtaining and maintaining
ethical use of Federal Bird Banding and Marking Permits
(permits 09082 and 22834). We also adhered to the “Ethics
and Responsibilities of Bird Banders”(BBL, 2019a). Our
methods for capturing and processing land birds were
approved by Humboldt State University’s Institute for
Animal Care and Use Committee. The preliminary results
and framing of culturally significant birds were presented
to the Karuk Resource Advisory Board (KRAB), with addi-
tional review of content from William Tripp and Colleen
Rossier of the Karuk Tribe’s Department of Natural
Resources.
We compiled a list of culturally significant species
(n=55) derived from tribal ITEK holders (reviewed and
amended by KRAB, November 2019) and ethnographic
information of northwestern California and southwestern
Oregon tribal uses of birds for food, regalia, and cultural
teachings (see Table 1for scientific names). From this
list, we selected study species that had more than 1000
individuals captured (n=11) for subsequent analyses
which resulted in an adequate number of captures in at
least two season-region data subsets for an individual
species analyses (see details below on data subset require-
ments for analyses). The 11 species selected were 5 resi-
dents (American Robin, American Goldfinch, Spotted
Towhee, Dark-eyed Junco, and Yellow-rumped Warbler)
and 6 migrants (Yellow-breasted Chat, Orange-crowned
Warbler, MacGillivray’s Warbler, Yellow Warbler,
Wilson’s Warbler, and Black-headed Grosbeak), which
occur in a variety of habitats across the study area.
!(!(
"
"
!(
!(
!(
!(
!(
"
!(
"
"
!(
!(
!(
"
")")
!(
!(
!(
"
!(
")
")
!(
")
!(
!(
")
"
")
!(
!(
!(
"
!(
")")
")
")
!(
!(
!(
")
")
!(
"
")
")
!(
!(
!(
!(
")
!(
"
!(
!(
!(
!(
"
!(
")
!(
!(
")
!(!(!(
!(
!(
!(
!(
!(
!(
"
")
!(
!(!(
!(
!(
!(!(
!(
!(
!(
"
")
!(
"
")
")
")
Banding Station
Regions
!(
COAST
!(
KLMTR
!(
MENDO
")
MOUNT
")
REDWD
"
SISKI
Oregon
California
Pacific Ocean
±
05010025
Kilomet ers
FIGURE 1 Map of the Klamath Siskiyou Bioregion study area
in northern California and southern Oregon showing locations and
regional designations of 92 banding stations where adult birds were
captured. Region abbreviations are: COAST, Coast; KLMTR,
Klamath-Trinity; MENDO, Mendocino; MOUNT, Mountain;
REDWD, Redwood; SISKI, Siskiyou.
ECOSPHERE 7of27

For each species, we calculated the total number of
adult birds and total numbers of breeding or molting
adults per 10-day period, beginning with the start of sam-
pling Period 1 on May 1 (first period of data collection for
many of our stations) and ending with Period 19 on
November 7. Breeding birds were defined as adults with
a smooth, vascularized, or wrinkled brood patch (indicat-
ing egg incubation in females), or a medium, large, or
bulbous cloacal protuberance (indicating breeding in
males). Molting birds were defined as adults undergoing
the annual prebasic molt (when worn flight feathers are
replaced with new ones) as characterized by the observa-
tion of symmetrical flight feather molt. We then calcu-
lated the percent of total breeding or molting birds per
period, for all species combined and individual species,
by all regions combined and individual region. For our
calculations, we assumed that during the breeding season
both breeding and nonbreeding adults generally had
a similar likelihood of being captured. Similarly, we
assumed that during the molting season both molting
and non-molting adults had a similar likelihood of being
captured.
From each data subset, we calculated two second-
order polynomial equations in program R using the linear
model function (R Core Team, 2018), one for breeding
birds and the second for molting birds, which represented
the distribution of breeding or molting individuals over
time (Figure 2; Appendix S1: Table S2). Response variables
were percent breeding and molting birds, respectively,
while the explanatory variable was a 10-day sampling
period. We used only those data subsets with at least
100 individuals and 10-day periods where >5% of adults
showed breeding or molting condition in order to smooth
subsequent curves. This method allowed us to estimate
breeding season start dates and molting season end dates
that occurred outside our sampling period.
We calculated the area under the curve (AUC) for
each of the two polynomial equations (calculated above)
using SAS software’s Proc Expand (SAS, 2012). Since the
calculated equation represents the estimated total distri-
bution of breeding or molting individuals over time,
including estimates for time periods outside our data col-
lection, the AUC value represents an estimate of 100% of
the individuals in the distribution and can be used to cal-
culate the percentage of a subset of the population within
a selected time interval. To generate the breeding season
data for this calculation, we estimated the percent of
adults (y) exhibiting breeding characteristics by 5-day
periods (x) from each equation; the shorter time spans
(compared to the 10-day period for data collection)
allowed for greater precision in estimating the percentage
of breeding population during a selected time interval.
Proc Expand used these incremental (x,y) pairs to com-
pute the approximate AUC using cubic spline interpola-
tion and the trapezoid rule. The resulting AUC value
represented 100% of the estimated breeding season dura-
tion (e.g., Figure 3). We then estimated the AUC for 25%,
50%, and 75% of the breeding season duration (Figure 3).
To do this, we trimmed the calculated equation curve
from the edges in a symmetric fashion from both ends of
the curve in increments of 5-day periods and calculated a
new AUC with Proc Expand for an estimate of the per-
centage of the full season’s AUC for that time span. This
was repeated until we had estimated time spans for 25%,
50%, and 75% of the breeding season duration. This pro-
cess was repeated for molting birds to estimate the AUC
for 25%, 50%, 75%, and 100% of the molting season dura-
tion (Figure 3).
We considered 50% of the seasonal distribution to be
the “core”of the season and use this quartile value in our
results and discussion. In a few instances, we found
the estimates for the beginning of breeding season
TABLE 2 List of six “regions”with code, name, and number of banding stations resulting from grouping similar banding stations based
on proximity to one another, similar elevations, and distance inland from the Pacific Ocean, thus reflecting similar environmental
conditions.
Region code Region name No. stations
Elevation (m) Distance inland (km)
Min. Max. Mean Min. Max. Mean
COAST Coast 11 2 79 11 0.4 8.3 3.4
KLMTR Klamath-Trinity 42 79 869 347 34.0 151.5 77.2
MENDO Mendocino 5 291 1954 1249 24.8 81.3 62.4
MOUNT Mountain 15 513 1364 1176 45.7 80.0 61.8
REDWD Redwood 8 10 315 62 4.5 112.1 29.1
SISKI Siskiyou 15 249 1495 750 7.0 141.0 62.3
Total 96
Note: Minimum, maximum, and mean values of elevation and distance inland for stations are shown for each region.
8of27 LONG ET AL.

(100% of range) to be outside the reported ranges for certain
species’arrival dates to the study area. These early estimates
for breeding season dates were heavily influenced by the
lack of data prior to May 1, particularly for those species
which were already well into breeding seasonality at the
start of the monitoring season, such as American Robin,
Black-headed Grosbeak, and Spotted Towhee (Figure 4).
To test the assumptions of goodness of fit for using a
second-order polynomial as a model, we calculated the
adjusted R
2
value for each calculated model
(Appendix S1: Table S2). About 80% of the models had
adjusted R
2
values >0.70, suggesting good fit. To test
additional assumptions of the data, we plotted histograms
of the residuals for each model, most of which showed a
normal distribution and evaluated plots of fitted versus
residuals values for each model to assess assumptions of
homogeneity (Appendix S1: Table S2).
Despite low adjusted R
2
values (<0.70) for three of
our study species, we decided to report these species to
demonstrate the wide variation in breeding and molting
seasons between species and regions in this study.
RESULTS
A total of 38,566 adults of 55 culturally significant bird
species were captured in northern California and southern
Oregon from 1992 to 2014 (Table 1). We recorded a higher
percentage of adults with signs of breeding (85% on June 10)
than those undergoing molt (53% on August 9) (Figure 2).
Apr. 21 July 30
July 10
June 20May 31 Aug. 19May 11
20
40
60
80
100
0
Percent of adults captured
A) Breeding
Percent of adults captured
B) Molting
20
40
60
80
100
0
June 30 Oct. 8
Sept. 18
Aug. 29Aug. 9July 20
Captured
Calculated
y = 0.09 + 0.24x – 0.02x²
Captured
Calculated
y = –1.90 + 0.40x – 0.02x²
FIGURE 2 Percent of adult birds captured in (A) breeding and (B) molting condition over time in the Klamath Siskiyou Bioregion by
10-day sampling periods, all regions and all culturally significant bird species combined. Dots designate the actual percentage of captured
birds by 10-day period showing signs of breeding or molting and the estimated polynomial equations (line and equation in upper right).
Dates are the first day of the 10-day sampling period.
ECOSPHERE 9of27

We combined all captures of culturally significant species
to assess general patterns of breeding and molting
seasonality and determine timeframes with the greatest
potential for negative impacts across the 11 most
abundant culturally significant bird species.
Breeding season
We estimated that the breeding season of the most abun-
dant culturally significant bird species occurred from
April 21 to August 23 (Figure 3). We found 75% of
the estimated breeding season occurred from May 16 to
July 29, the core 50% from May 26 to July 14, and 25%
from June 10 to July 4.
Regional estimates for the beginning of the full breed-
ing season ranged from April 6 in REDWD to April 26 in
COAST and KLMTR (a difference of 20 days; Figure 3).
The end of the breeding season ranged from August 13 in
MENDO to September 7 in COAST (a difference of
25 days). The shortest breeding duration was 120 days for
KLMTR while the longest duration was 140 days for
REDWD. The beginning of the core breeding season
ranged from May 16 (MENDO and REDWD) to June 5
(COAST) (a difference of 20 days), with the end of the
season ranging from July 9 (MENDO, MOUNT, and
REDWD) to July 24 (COAST) (a difference of 15 days).
Individual species showed a wider range in estimated
breeding seasons. For example, beginning of breeding
season ranged from March 17 for Black-headed Grosbeak
to May 1 for Wilson’s Warbler (a difference of 45 days;
Figure 5A). The end of the breeding season ranged from
August 8 for Wilson’s Warbler to September 7 for
American Goldfinch and American Robin (a difference
of 30 days). Yellow-breasted Chat had the shortest breed-
ing duration of 60 days, American Robin and Spotted
Towhee had the longest breeding duration of 95 days
each (Table 3). The difference in breeding season start
Breeding
N4 14 246 16 26 5 15 25 5 15 25 3
April May JulyJune August September Octobe
r
6 16 26 3 13 23
Region
Molting
COAST
MOUNT
MENDO
KLMTR
REDWD
SISKI
ALL 31,585
5822
10,355
1281
4626
3040
6669
COAST
MOUNT
MENDO
KLMTR
REDWD
SISKI
ALL 31,585
5822
10,355
1281
4626
3040
6669
MoltBreedPercent
100
25
50
75
Values extrapolated from equation
FIGURE 3 Breeding and molting season begin and end date estimates by 5-day periods for all regions combined and by each region,
using all culturally significant birds combined. Gradations of fill indicate the percentage of the calculated area under the curve. Dates are the
first day of the 5-day period. Nis the number of individuals captured and used to estimate the polynomial equation. Region abbreviations
are: ALL, all regions combined; COAST, Coast; KLMTR, Klamath-Trinity; MENDO, Mendocino; MOUNT, Mountain; REDWD, Redwood;
SISKI, Siskiyou.
10 of 27 LONG ET AL.

dates between species was less variable when considering
the season core (Figure 5A). Core breeding start
dates ranged from May 6 (Black-headed Grosbeak) to
June 10 (American Goldfinch) (a difference of 35 days
compared to 45 days for full breeding season beginning
dates). Similarly, for end dates, the core breeding season
Percent of adults capturedPercent of adults captured Percent of adults captured
20
40
60
80
100
0
20
40
60
80
100
0
20
40
60
80
100
0
Captured
Calculated
y = 0.57 + 0.16x – 0.01x²
Captured
Calculated
y = 0.55 + 0.10x – 0.01x²
Captured
Calculated
y = 0.53 + 0.16x – 0.02x²
Mar. 22 July 20June 30June 10May 21 Aug. 9May 1Apr. 11 Aug. 29
Mar. 22 July 20June 30June 10May 21 Aug. 9May 1Apr. 11 Aug. 29
Mar. 22 July 20June 30June 10May 21 Aug. 9May 1Apr. 11 Aug. 29
A) American Robin
B) Black-headed Grosbeak
C) Spotted Towhee
FIGURE 4 Percent of adult birds breeding throughout the season in the Klamath Siskiyou Bioregion by 10-day periods, all regions
combined, for three species with early estimated breeding start dates. Dots designate the actual percentage of captured birds by 10-day
period showing signs of breeding and the estimated polynomial equations (line and equation in upper right). Dates are the first day of the
10-day sampling period.
ECOSPHERE 11 of 27

end ranged from July 4 (Orange-crowned Warbler) to
July 29 (American Goldfinch) (a difference of 25 days
compared to 30 days for the full breeding season end
dates).
Breeding start dates for species by regions ranged
from the earliest date of March 7 for Dark-eyed Junco
in MENDO and Black-headed Grosbeak at KLMTR to
May 6 for Yellow-breasted Chat in SISKI (a difference
of 60 days). End dates ranged from July 19 for
Yellow-breasted Chat in SISKI to September 27 for
American Robin in COAST (a difference of 70 days).
Core breeding start dates ranged from May 1 for
Black-headed Grosbeak in KLMTR to June 5 for Wilson’s
Warbler in KLMTR (a difference of 35 days).
Duration of breeding seasons for most species
varied between regions (Figure 6). The most variable
was Black-headed Grosbeak, with breeding seasons
ranging from March 7–August 28 at KLMTR and from
April 11 to August 18 at REDWD, with a difference
of 35 days between start dates and durations of
175 and 125 days, respectively. Conversely, Yellow-
rumped Warbler exhibited no variation in breeding
XXX
Breeding
Species N
March
6 16 26
April May
1209
1722
A.
Species
B.
American Robin
American
Goldfinch
Spotted Towhee
Dark-eyed Junco
Yellow-breasted
Chat
MacGillivray’s
Warbler
Orange-crowned
Warbler
Yellow Warbler
Wilson’s Warbler
Yellow-rumped
Warbler
Black-headed
Grosbeak
American Robin
American
Goldfinch
Spotted Towhee
Dark-eyed Junco
Yellow-breasted
Chat
MacGillivray’s
Warbler
Orange-crowned
Warbler
Yellow Warbler
Wilson’s Warbler
Yellow-rumped
Warbler
Black-headed
Grosbeak
Molting
3 17 27 6 16 26 5 15 25 5 15 25
JulyJune Septembe
r
August
4 14 24
1723
2689
2705
3784
1447
2003
807
2833
2073
N
June
20 30
AugustJuly OctoberSeptember
10 20 30 919 29 818 28 8
541
247
1642
1984
717
1483
981
828
367
1198
--
MoltBreedPercent
100
25
50
75
Values extrapolated from equation
FIGURE 5 Breeding (A) and molting (B) season begin and end date estimates for the 11 most abundant culturally significant birds in
the Klamath Siskiyou Bioregion by 5-day periods. Gradations of fill indicate the percentage of the calculated area under the curve. Dates are
the first day of the 5-day period. Nis the number of individuals captured and used to estimate polynomial equation. XXX indicates the
species was not captured in high enough numbers during the season for this analysis.
12 of 27 LONG ET AL.

season across region, running from April 21 to August 23
at both MENDO and MOUNT, a duration of 125 days.
Yellow-breasted Chat had the shortest estimated breed-
ing duration of 75 days in SISKI, while the American
Robin had the longest of 195 days in COAST (Table 3).
On average, breeding season for residents began
slightly earlier and lasted longer than migrants
(Figure 5A). Breeding season start dates for residents
averaged April 11 (March 27–April 22) and end dates
averaged August 31 (August 22–September 7), for an
average breeding duration of 143 days. Breeding start
dates for migrants averaged April 16 (March 17–May 1)
and end dates averaged August 15 (August 8–August 28)
for a breeding duration of 121 days. Breeding season for
residents had greater variability among species compared
with migrants. Resident start dates differed by a month
(March 27–April 26) while migrants differed by 15 days
(April 16–May 1), if the Black-headed Grosbeak’s early
outlier is removed (March 17).
Molting season
We estimated that the molting season of the most abun-
dant culturally significant bird species occurred from
June 30 to October 7, with 75% of the season occurring from
July 20 to September 17, 50% from July 30 to September 7,
and 25% from August 14 to August 28 (Figure 3).
Regional estimates for the beginning of the full molting
season ranged from June 24 in KLMTR, MOUNT, and
SISKItoJuly9inCOAST(adifferenceof15days;
Figure 3).Theendoftheseasonformostregionsoccurred
on October 2, ending in MENDO and COAST on
October 7 (a difference of 5 days). The shortest molting
time duration was 85 days in COAST, while the longest
was 100 days in KLMTR, MENDO, MOUNT, and SISKI
regions. The beginning of the core season ranged from
July 25 (MOUNT and REDWD) to August 4 (COAST)
(a difference of 10 days). The end of the core season ranged
from September 2 (MOUNT and REDWD) to September
12 for the remaining regions (a difference of 5 days).
Beginning of the molting season for individual species
ranged from June 20 for Orange-crowned and Yellow war-
blers to July 20 for American Goldfinch (a difference of
30 days; Figure 5B). The end of the season ranged from
September 2 for Yellow-rumped Warbler to October 12 for
American Robin (a difference of 40 days). Yellow-breasted
Chat had the shortest molting season duration of 60 days,
while American Robin and Spotted Towhee had the lon-
gest molting season duration of 95 days. We did not cap-
ture enough individuals of Black-headed Grosbeak in molt
to estimate seasonality and duration.
Difference in dates for the core molting season was
somewhat less than for the full season (Figure 5B). Core
molting seasons for species ranged from July 15 for
Orange-crowned and Yellow warblers to August 9 for
American Robin and American Goldfinch (a difference
of 25 days). Core molting season end dates ranged from
August 13 for Yellow Warbler to September 12 for
American Robin (a difference of 30 days).
We estimated molting start dates for individual spe-
cies by region (Figure 7). The earliest molting start date
was June 5 for Orange-crowned Warbler at MOUNT,
while the latest was July 15 for American Goldfinch at
COAST (a difference of 40 days). End dates ranged from
September 2 for Orange-crowned Warbler at KLMTR to
November 6 for American Goldfinch at COAST
(a difference of 65 days). Start dates for core molting sea-
son ranged from July 5 for MacGillivray’s Warbler at
KLMTR and Orange-crowned Warbler at MOUNT to
TABLE 3 Summary of breeding and molting season time spans in number of days for culturally significant bird species, listing the
species, and/or region with the shortest and longest time spans for the division (species, region, or species–region) and season.
Division and season
Shortest time span Longest time span
Species and/or region No. days Species and/or region No. days
Species
Breeding Wilson’s Warbler 100 American Robin, Black-headed Grosbeak 165
Molting Yellow-breasted Chat 60 American Robin, Spotted Towhee 95
Region
Breeding KLMTR 120 REDWD 140
Molting COAST 85 KLMTR, MENDO, MOUNT, SISKI 100
Species–Region
Breeding Yellow-breasted Chat-SISKI 75 American Robin-COAST 195
Molting Yellow-breasted Chat-KLMTR 75 American Goldfinch-COAST 115
Note: Region abbreviations are: COAST, Coast; KLMTR, Klamath-Trinity; MENDO, Mendocino; MOUNT, Mountain; REDWD, Redwood; SISKI, Siskiyou.
ECOSPHERE 13 of 27

August 19 for American Goldfinch at COAST
(a difference of 45 days). End dates for core molting sea-
son ranged from August 8 for Orange-crowned Warbler
at KLMTR to October 2 for American Goldfinch at
COAST (a difference of 55 days).
Duration of molting seasons was less variable between
regions when compared to breeding seasons for most spe-
cies (Figure 7). The most variable species were Dark-eyed
Junco and MacGillivray’s Warbler; start dates for each had
a 20-day difference between regions. Dark-eyed Junco
start dates ranged from June 25 in MOUNT to July 15 in
SISKI, with durations of 110 and 85 days, respectively.
MacGillivray’s Warbler start dates ranged from June 10 in
KLMTR to June 30 in SISKI, with durations of 95 and
75 days, respectively. Spotted Towhee was the least
variable, beginning on July 5 for all three regions in which
it occurred and ending on either October 7 or October 17
(a difference of 10 days). The shortest molting season
duration was 75 days for Yellow-breasted Chat at KLMTR;
the longest duration was 115 days for American Goldfinch
at COAST. We did not capture enough molting American
Robin or Yellow-rumped Warbler in any one region for
this analysis.
On average, residents started and ended molting
later than migrants, while molting duration was similar
(Figure 5B). Start of molt season for residents averaged
July 10 (July 5–July 20) and end averaged October 2
(September 27–October 12) for an average molting
American
Robin
American
Goldfinch
Spotted
Towhee
Dark-eyed
Junco
Yellow-breast.
Chat
MacGillivray’s
Warbler
Orange-cr.
Warbler
Yellow
Warbler
Wilson’s
Warbler
Yellow-rump.
Warbler
Black-headed
Grosbeak
Species NRegion
March JuneMayApril SeptemberAugustJuly
COAST
SISKI
MOUNT
COAST
KLMTR
MOUNT
SISKI
MOUNT
SISKI
MENDO
KLMTR
SISKI
SISKI
KLMTR
MOUNT
COAST
KLMTR
REDWD
SISKI
KLMTR
MOUNT
MOUNT
MENDO
COAST
KLMTR
REDWD
KLMTR
REDWD
SISKI
365
189
162
17
1572
1066
292
367
705
1110
956
286
2198
1541
1280
687
293
339
222
422
410
1232
185
363
442
829
979
931
173
543
62772616 6 2616 5 2515 5 2515 4 2414 3 2313
Percent
100
25
50
75
Extrapolated from equation
Breed
FIGURE 6 Breeding season begin and end date estimates for the 11 most abundant culturally significant birds in the Klamath Siskiyou
Bioregion by region and 5-day periods. Gradations of fill indicate the percentage of the calculated area under the curve. Dates are the first
day of the 5-day period. Nis the number of individuals captured and used to estimate polynomial equation. Region abbreviations are:
COAST, Coast; KLMTR, Klamath-Trinity; MENDO, Mendocino; MOUNT, Mountain; REDWD, Redwood; SISKI, Siskiyou.
14 of 27 LONG ET AL.

duration of 84 days, while migrants averaged June 27
(June 20–July 10) and end of season averaged
September 12 (September 2–September 22), an average
molting duration of 77 days.
Seasonal differences in use of regions by
species
We found that many species used different regions for
breeding and molting. One of the most distinct was the
Black-headed Grosbeak, which is characterized by molt
migration: it breeds in several regions of our study area
but leaves the area to molt in the monsoonal regions of
the Sonoran Desert (Figures 5B and 7; Pyle et al., 2009;
Siegel et al., 2016). We did not capture enough molting
American Robin and Yellow-rumped Warbler for analy-
sis by region, though we did capture enough to analyze
the molting season for the entire study area (Figures 5B
and 7).
Regions KLMTR and SISKI were used by the most
species for breeding, with seven species for each region
A
merican
Robin
A
merican
Goldfinch
Spotted
Towhee
Dark-eyed
Junco
Yellow-breast.
Chat
MacGillivray’s
Warbler
Orange-crown.
Warbler
Yellow Warbler
Yellow-rumped
Warbler
Wilson’s
Warbler
Black-headed
Grosbeak
381
515
224
983
653
1125
571
799
991
199
589
178
330
210
542
246
157
383
239
149
(none)
(none)
(none)
COAST
SISKI
MOUNT
KLMTR
SISKI
MOUNT
MENDO
KLMTR
SISKI
MOUNT
KLMTR
SISKI
MOUNT
KLMTR
MOUNT
KLMTR
REDWD
COAST
MOUNT
KLMTR
June SeptemberAugustJuly October Novembe
r
52515 5 2515 4 2414 3 2313 3 2313 2
Percent
25
100
75
50
Molt
XXX
XXX
XXX
Species N
Region
FIGURE 7 Molting season begin and end date estimates for the 11 most abundant culturally significant birds in the Klamath Siskiyou
Bioregion by region and 5-day periods. Gradations of fill indicate the percentage of the calculated area under the curve. Dates are the first
day of the 5-day period. Nis the number of individuals captured and used to estimate polynomial equation. XXX indicates a species was not
captured in high enough numbers during the season for this analysis. Region abbreviations are: ALL, all regions combined; COAST, Coast;
KLMTR, Klamath-Trinity; MENDO, Mendocino; MOUNT, Mountain; REDWD, Redwood; SISKI, Siskiyou.
ECOSPHERE 15 of 27

(Figure 6). By comparison, MENDO was used by only
two species for breeding: Dark-eyed Junco and
Yellow-rumped Warbler.
KLMTR and MOUNT were each used by six species
for molting, the most for the season by a region
(Figure 7). By comparison, COAST, MENDO, and
REDWD were the least used regions for molting, with
1–2 species each.
Five species used the same regions for both breeding
and molting, at least in part: Yellow Warbler used
KLMTR and MOUNT; Dark-eyed Junco, MacGillivray’s
Warbler, and Spotted Towhee used KLMTR, MOUNT,
and SISKI; and American Goldfinch used COAST
(Figures 6and 7). Other species added or subtracted one
or more regions where they underwent molt. Two
migrants were found upslope for molting: Wilson’s
Warbler used COAST, KLMTR, and REDWD for breed-
ing and molting, and added MOUNT for the molting sea-
son; Orange-crowned Warbler used COAST, KLMTR,
REDWD, and SISKI for breeding, which included two
coastal regions, but were found more exclusively inland
and higher in elevation for molting to KLMTR, MOUNT,
and SISKI. One migrant was found downslope during
molting: Yellow-breasted Chat used KLMTR and SISKI
for breeding but was found only in KLMTR for molting.
DISCUSSION
Precipitous declines in bird populations across North
America have renewed interest in the effects of manage-
ment actions, such as prescribed fires and cultural burns,
on avian communities and how the timing of such
actions may influence survival and reproductive success.
Overall, US bird populations have declined by an esti-
mated 30% over the last 50 years, with forest birds
decreasing by 22% (NABCI, 2019). Rosenberg et al. (2019)
estimated that over 50% of western forest birds are suffer-
ing population declines. Across the Northern Pacific
Rainforest Conservation Region (NABCI, 2020), which
includes the Klamath Siskiyou Bioregion, over 60% of
culturally significant species have shown a decline from
2005 to 2015, including 9 of the 11 species in our analysis,
with the steepest declines for American Goldfinch and
MacGillivray’s Warbler (Table 1; Sauer et al., 2017). Only
Yellow-breasted Chat and Black-headed Grosbeak
showed increasing population trends. A regional study
showed similar trends within the Klamath Siskiyou
Bioregion, including Yellow-rumpled Warbler population
declines and Yellow-breasted Chat and Black-headed
Grosbeak population increases (Rockwell et al., 2017).
Our study demonstrated that the energetically taxing
breeding and molting seasons of 11 culturally significant
resident and migratory adult land bird species varied
temporally and spatially in the Klamath Siskiyou
Bioregion as supported by the wide variation in timing of
lifecycle events. Nuanced differences in the timing of these
avian lifecycle phases can present challenges for tribes and
forest and wildland fire managers aiming to balance
potential negative and positive effects of cultural burning
and prescribed fires on bird communities (Huff et al.,
2005,seealsoKarukTribe,2019:73, fig. 3.8). Specifically,
our results indicate that tribes and land managers could
consider scheduling burns to avoid periods of physiologi-
cal stress (breeding and molting), which varied by region
and species across a variety of vegetation types.
Globally, paleoecological research has identified close
relationships between fire ignitions and Indigenous peo-
ples, as shown in Australia (Trauernicht et al., 2015),
whereby the local avifauna is dependent on the resulting
fire regime to such an extent that changes associated with
the arrival of European settlers have endangered several
fire-dependent endemic species (Olsen & Weston, 2005).
Similarly, local tribal knowledge among the different tribes
of northern California and southern Oregon have histori-
cally guided the application of fire use at specific times of
year, with seasonal variation for different habitats in
response to culturally determined cues, which naturally
encompass and protect stressful phenological periods of the
local fauna (Anderson, 1996,2007;Anderson&Moratto,
1996;Knightetal.,2022). For example, guidance from
ITEK among some tribal community members of the
Karuk traditional belief systems, and current Karuk Tribal
policy, leads practitioners to refrain from using fire for cul-
tural burning when the Pleiades star cluster is absent from
the sky, or beginning mid-April by the western calendar
(Karuk Tribe, 2019). Reintroducing fire long absent from
many of the habitats has ethical as well as sociocultural
considerations that have not necessarily been considered
by western-minded academically trained fire managers and
ecologists.
Our research was conducted in response to the Karuk
and other local tribes initially expressing concerns about
federal and state agencies and organizations (e.g., USDA
Forest Service, local Fire Safe Councils) conducting pre-
scribed fires during the spring and early summer (when
some tribal traditions generally end cultural burning as
noted above) as well as other times of the year under con-
ditions considered permissible to burn by local tribal tra-
ditions (Fry & Stephens, 2006; Karuk Tribe, 2019) for fire
risk reduction, hazardous fuels abatement, and public
safety in the Wildland Urban Interface and other areas of
the landscape. As described above, agencies often sched-
ule these burns when fire personnel and resources (such
as engines and tankers) are more available to implement
such management burns; that is, when they are not
16 of 27 LONG ET AL.

assigned to wildfires which typically occur during the
summer and fall in this region. Adding to the complica-
tions of scheduling prescribed fires, available burn win-
dows can be short (2–3 days) and few in number (as few
as two windows per month), as seen in California’s Lake
Tahoe Basin (Striplin et al., 2020). Additionally, with
expected temperature increases from climate change, the
number, length, and timing of potential burn windows
may change as observed in climate modeling for the
southeastern United States (Kupfer et al., 2020) and
Australia (Di Virgilio et al., 2020), resulting in reduced
opportunities for prescribed fire use.
Breeding seasonality
Based on previous research, our expectation was that
breeding would occur earliest at the lowest elevations
(Bears et al., 2009; LaBarbera & Lacey, 2018; Perfito
et al., 2004). In our study, we estimated on average
REDWD (mean elevation 62 m) had the earliest breeding
season start for all species combined (April 6), and SISKI,
MOUNT, and MENDO (mean elevations 750–1249 m)
had slightly later season starts (April 11) as expected.
Unexpectedly, we found that the other two low-elevation
regions, COAST (11 m) and KLMTR (347 m), had the lat-
est starts to the breeding season (April 26). LaBarbera
and Lacey (2018) reported the earliest hatching dates for
Dark-eyed Juncos were at low elevations while they were
later at high- and mid-elevation stations. Bears et al.
(2009) reported that with increasing elevation (stations
at 1000 and 2000 m), Dark-eyed Juncos delayed the
development of reproductive structures (such as cloacal
protuberances) and reduced the duration of their
reproductive period to less than half of the time used by
low-elevation birds. Similarly, Perfito et al. (2004) reported
that the testicular development of Song Sparrows
(Melospiza melodia) began earliest at coastal sites com-
pared to higher elevation sites in the mountain foothills in
Washington State. Perfito et al. (2004) also reported that
testis development was more strongly associated with
maximum temperature and emergence of new green
shoots rather than elevation and that plant phenology
tended to be more advanced on the coast than in the
mountains early in the breeding season. This suggests that
some habitat component other than elevation may have a
greater effect on timing of breeding in COAST and
KLMTR as compared to REDWD, contributing to later
breeding season starts. Relative to elevation, the Karuk
Tribe (2019:96) states “The [Yellow-breasted] chat is tied
to the responsibility of humans to realize that something
has to be done about fire. The chat is a migratory bird that
nests in the spring. When we are burning [at] low
[elevations], the return of the chat and other birds who
have come back to nest, signals that we are to stop burn-
ing. Other avian elevational migrants and birds who nest
at different elevations and times should be developed as
cultural indicators for fire applications. Humans have
burned as they moved up and down [seasonally] with
birds for thousands of years.”
In our study, we were able to compare breeding sea-
sonality between multiple species in one region. We
found the start of breeding seasonality differed by almost
a month between five warbler species in one of the most
used regions for breeding, KLMTR. By contrast,
Flockhart (2010) reported that hatch dates were not sig-
nificantly different between five species of warblers in
their study area in Alberta, Canada, with mean hatch
dates spanning only 7 days.
On average, we found breeding seasons for residents
started slightly earlier (5 days) than migrants. If the early
outlier Black-headed Grosbeak is removed from the
migrants estimate, the difference increases to 10 days. By
contrast, in England, Goodenough et al. (2010) reported a
difference of almost a month in mean lay dates between
residents and migrants. They reported that the mean lay
date for four resident species, Blue Tit (Cyanistes
caeruleus), Great Tit (Parus major), Coal Tit (Periparus
ater), and Nuthatch (Sitta europaea), ranged from
April 30 to May 7; conversely, they reported the mean lay
dates for two migrant species, European Pied Flycatcher
(Ficedula hypoleuca) and Redstart (Phoenicurus
phoenicurus), were on June 5 and June 6, respectively.
Some of the greatest variation in breeding seasonality
was within individual species between regions. Thus,
minimizing the impact of burning on a single species can
be influenced by localized seasonality as exhibited by the
Black-headed Grosbeak, with a difference of over a
month between breeding start dates by region, or by a
more general seasonality—irrespective of location—as
exhibited by the MacGillivray’s Warbler, with a 5-day dif-
ference between regional start dates. Given variability in
breeding seasonality exhibited by individual species,
spring burns prior to April 21 in any region would avoid
the majority of study species breeding throughout our
study area, or prior to May 26 to avoid the average esti-
mated core breeding season (Figure 3). By contrast, a
spring burn that is regionally specific could be scheduled
with more precision, such as burning prior to April 6 in
REDWD to avoid most breeding birds in the region, or
prior to May 16 to avoid the estimated core breeding
season in that area. Among other indicators, some Karuk
cultural fire practitioners understand the nesting of
certain bird species to indicate when to stop burning in
the late winter/early spring months (such as the Yellow-
breasted Chat discussed earlier), while other indicators
ECOSPHERE 17 of 27

will present themselves to signify that certain types of
cultural burning can resume again in mid-to-late June
(William Tripp citing Karuk ITEK, personal communica-
tion). Thus, the breeding and molting timeframes
outlined here are not fully aligned with the indicators
that some Indigenous practitioners may use to indicate
burn timing.
Molting seasonality
In our study, migrant species generally began and ended
molt earlier than resident species. De La Hera et al.
(2009) compared flight feather growth in migratory and
sedentary populations of Blackcaps (Sylvia atricapilla)in
southern Spain and reported that individuals that
migrated grew their feathers faster, which produced ligh-
ter feathers than those in sedentary individuals, thus
demonstrating a trade-off between molt speed and plum-
age quality in migratory birds. By comparison, Rimmer
(1988) reported that migrant Yellow Warblers in James
Bay, Ontario, regularly overlapped care of fledglings with
the beginning of prebasic molt, then after a rapid, intense
molt of 35–45 days, they began migration during the final
stages of remigial molt in early September.
Over half of the species (6 of 11) in our study did not
remain solely on their breeding grounds to molt; they
may have dispersed within our study area or completely
left the area to molt. Pyle et al. (2018) reported the
probability of a breeding bird being recorded molting at
the same station was about 0.47 (95% credibility
interval =0.38–0.57) in the western United States. Gilbert
et al. (2020) reported that Orange-crowned Warblers in the
Sierra Nevada foothills migrated to moister, higher eleva-
tion habitats to molt and that breeding Orange-crowned
Warblers in central inner-coastal California were rarely
observed molting on their breeding grounds, suggesting
they may molt elsewhere. Cassin’s Vireo in Washington
State moved from lower elevation dry pine or Douglas-fir
forests upslope at least 300 m, to molt in wetter,
high-elevation Douglas-fir forests (Rohwer et al., 2008).
Similarly, two species in our study (Wilson’sand
Orange-crowned warblers) were more abundant at higher
elevations during molt or added higher elevation regions
to molt. Two previous studies in our study region found
support for this as well (Wiegardt, Barton, et al., 2017;
Wiegardt, Wolfe, et al., 2017). Wiegardt, Barton, et al.
(2017) showed that Wilson’s Warblers were more likely to
breed at lower elevations and molt at higher elevations.
Wiegardt, Wolfe, et al. (2017) reported that long-distance
migrants such as Orange-crowned Warblers were found at
higher elevations during molt and Audubon’sWarbler
moved farther inland for molting compared to their
breeding habitats. Both studies suggest that some
individuals move altitudinally after breeding to com-
plete the definitive prebasic molt.
In our study, all species completed molt in all regions
by October 17 and after September 12 for the core molt
season (Figure 6). Beginning prescribed fires after these
dates would avoid the majority of molting birds across all
regions.
Fire effects and planning
Researchers have investigated the indirect effects of fire
on bird populations and communities, especially as
related to changes in forest structure or food resources
after wildfires and prescribed fires (e.g., Bagne & Purcell,
2009,2011; Fontaine et al., 2009; Huff et al., 2005;
Murphy et al., 2017; Russell et al., 2009; Seavy &
Alexander, 2006,2014; Stephens & Alexander, 2011;
Stephens et al.,