Revised: 20 July 2022

Accepted: 11 August 2022

DOI: 10.1002/ecs2.4316

ARTICLE

Rapid shifts in migration routes and breeding latitude in
North American bluebirds
Jared Sonnleitner 1 | Stefanie E. LaZerte 2
Nancy J. Flood 1 | Matthew W. Reudink 1

| Ann E. McKellar 3 |

1

Department of Biological Sciences,
Thompson Rivers University, Kamloops,
British Columbia, Canada

2

Department of Biology, Brandon
University, Brandon, Manitoba, Canada

3

Wildlife Research Division, Environment
and Climate Change Canada, Saskatoon,
Saskatchewan, Canada
Correspondence
Matthew W. Reudink
Email: mreudink@tru.ca

Funding information
Natural Sciences and Engineering
Research Council of Canada; Thompson
Rivers University Undergraduate
Research Experience Award Program
(UREAP); NSERC Discovery Grant
Handling Editor: Jeffrey F. Kelly

Abstract
Spatial and temporal shifts in the migratory patterns of birds have become
more frequent as climate change and habitat alteration continue to impact
ecosystems and the species dependent on them. In this study, we used eBird
community science data collected over ten years to examine potential changes
in the migratory patterns of three North American bluebird species: eastern
(Sialia sialis), western (Sialia mexicana), and mountain (Sialia currucoides)
bluebirds. Community science datasets such as those provided through eBird
are a valuable tool for examining population-level processes, as such data are
often costly and time-consuming to collect through other approaches
(e.g., directly tracking individuals). Using generalized additive models, we produced smoothed migration paths for all three species over each season from
2009 to 2018. We asked whether there were changes over this 10-year period
in the timing of spring and fall migration and migration speed, and the population centroids during breeding and migration. In contrast to many species
that are experiencing poleward shifts in their distributions, the population centroids during the breeding period of all three bluebird species appear to have
shifted southward over the past decade. Perhaps most surprisingly, we also
detected strong longitudinal shifts in the population centroids during migration in eastern and western bluebirds, with both species shifting toward the
center of the continent. Despite these changes in migratory routes and breeding distributions, we detected no change in the migratory timing or speed of
any of the species. Our analysis indicates that bluebirds are rapidly altering
the pattern of their migration, likely in response to changing environmental
conditions, but not always in the direction predicted.
KEYWORDS
bluebirds, breeding, conservation, latitude, longitude, migration, routes, shift, speed, timing

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.
© 2022 The Authors. Ecosphere published by Wiley Periodicals LLC on behalf of The Ecological Society of America.
Ecosphere. 2022;13:e4316.
https://doi.org/10.1002/ecs2.4316

https://onlinelibrary.wiley.com/r/ecs2

1 of 12

21508925, 2022, 12, Downloaded from https://esajournals.onlinelibrary.wiley.com/doi/10.1002/ecs2.4316 by Cochrane Canada Provision, Wiley Online Library on [29/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

Received: 29 April 2022

INTRODUCTION
Since the mid-1900s, global temperatures have risen by
approximately 1 C (Allen et al., 2018), and the Earth has
experienced unprecedented anthropogenic changes that
have altered ecosystems and disrupted biological processes (Walther et al., 2002). Spring is advancing in the
Northern Hemisphere, both in terms of first-leaf and
first-bloom dates (Buermann et al., 2013). These advances
are forcing insects—and the wildlife dependent on
them—to alter their life cycles accordingly (Schwartz
et al., 2006). The rapidity of the change in spring phenology can result in a mismatch between peak resource
availability and the peak resource demands of many
animals, including migratory birds, which may be
constrained in their ability to alter the timing of life
history events such as migration and reproduction.
With advancing springs, migratory birds may need to
arrive on their breeding grounds earlier to take advantage
of earlier spring productivity (Mayor et al., 2017; Saino
et al., 2011). Individuals may also increase the speed of
migration; many bird species already migrate faster during spring compared with fall migration due to the time
constraints on arrival to obtain high-quality territories
and mates (Nilsson et al., 2013), so changes to migration
speed may be physiologically possible in some cases.
Speed of migration may also be influenced by environmental conditions experienced en route (Bridge
et al., 2016), such as higher wind speeds resulting in earlier arrival to migratory roosts in Vaux’s swifts (Chaetura
vauxi) (Prytula et al., 2021).
In addition to temporal shifts, species may also be
undergoing spatial shifts in their distribution as a result of
climate change (Huang et al., 2017; La Sorte &
Graham, 2020; La Sorte & Thompson, 2007). For instance,
in response to increasing temperatures, many northern
avian species’ ranges are shifting to higher latitudes
(Curley et al., 2020; Thomas & Lennon, 1999) and elevations (Tingley et al., 2012). There is also evidence for shifts
in the longitudinal distributions of many species, but
these studies often focus on species presence or absence at
the range margins (Huang et al., 2017), rather than
changes in the population centroid of the species across
their total distribution. Analyzing a species’ entire range
may be more informative about changes to their entire
longitudinal distribution (Virkkala & Lehikoinen, 2014)
than examining the occurrence of small populations,
which may be heavily influenced by local events such as
land use changes or irregular annual variation in snowpacks or temperature.
Historically, many studies of migratory birds have
focused on events occurring during the breeding
periods, but to fully understand the annual cycle of a

SONNLEITNER ET AL.

species, we need to improve our understanding of events
occurring outside of its relatively short breeding period
(Faaborg et al., 2010). Increasing attention is now being
directed toward carryover effects and full annual cycle
dynamics (Marra et al., 2015). Recent advances in tracking technology (including satellite tracking, GPS tags,
and geolocators) have shed light on individual-level patterns of migration, while large-scale datasets, such as
those obtained through community science or weather
radar, can be used to examine changes in entire
populations (Curley et al., 2020; Kelly et al., 2012;
Rushing et al., 2020). Here, we make use of community
science data available from eBird (Sullivan et al., 2009)
to examine shifts in migration patterns among three
closely related migratory bird species.
Specifically, we ask whether eastern bluebirds (Sialia
sialis), western bluebirds (Sialia mexicana), and mountain bluebirds (Sialia currucoides) have shifted their
migration patterns over the past decade. Each species
exhibits different migration life histories: eastern and
western bluebirds are considered partial migrants as
populations in parts of the range remain as year-round
residents (Gowaty & Plissner, 2020; Guinan et al., 2020),
while mountain bluebirds are considered fully migratory
(Johnson & Dawson, 2020). The species also exhibit
varying short-term (2009–2019) population trends based
on data from the North American Breeding Bird Survey
(BBS) (A. C. Smith et al., unpublished, an update of
Environment Canada, 2017). Eastern bluebird populations
have declined by 1.14% per year (95% credible interval:
1.61, 0.65), and mountain bluebird populations have
also declined, although credible intervals overlap zero
( 0.92% per year 2.21, 0.39) (A. C. Smith et al.,
unpublished, an update of Environment Canada, 2017).
By contrast, western bluebirds have shown an increase of
0.77% per year, although again credible intervals overlap
zero ( 1.02, 2.47). Thus, while closely related, the three
species have distinct ecologies and different population
trends, and therefore may differ in the degree to which
they are impacted by climate change and in the strategies
they use to respond.
To examine potential changes in the migratory patterns
of North American bluebirds, we used 10 years of community science data from eBird (Sullivan et al., 2009).
Large-scale community science datasets such as those available from eBird have allowed researchers to harness millions of bird observations, which can allow for comparisons
of year-to-year shifts in migratory patterns (Sullivan
et al., 2014; Supp et al., 2015). In response to environmental
changes, we expected that bluebirds may have altered
aspects of their migratory behavior such as the speed,
timing, and routes used, in order to remain in synchrony
with advancing spring phenology (Mayor et al., 2017;

21508925, 2022, 12, Downloaded from https://esajournals.onlinelibrary.wiley.com/doi/10.1002/ecs2.4316 by Cochrane Canada Provision, Wiley Online Library on [29/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

2 of 12

Visser & Both, 2005). Based on recent trends observed
across a broad range of taxa, we also expected that the three
species may be shifting their breeding distributions in
response to rapidly changing environmental conditions in
North America (Parmesan, 2006; Parmesan & Yohe, 2003).
Specifically, we predicted that western bluebirds would
demonstrate an eastward shift in population centroids during breeding and possibly also during migration as a result
of factors such as changes in fire suppression regimes and
increased nesting opportunities through nest box programs,
both of which have been linked to the rapid range expansion of the species (Duckworth, 2009; Duckworth &
Badyaev, 2007). By contrast, we did not have an ecological
reason to expect that eastern or mountain bluebirds would
demonstrate a consistent, species-level longitudinal shift
during breeding or migration. While migratory timing is
often associated with large-scale environmental change,
longitudinal shifts may be much more reliant on local conditions en route to or from the breeding and nonbreeding
grounds (Stanley et al., 2012). Consistent with studies
conducted on other migratory birds, we predicted that as a
result of increasing temperatures, all three species of bluebirds would arrive earlier on their breeding grounds
(Jonzen, 2006; Parmesan & Yohe, 2003) and/or increase
spring migratory speed (Nilsson et al., 2013), and shift the
population centroid during breeding toward a higher latitude over time (Rushing et al., 2020).

METHODS
We obtained presence data for eastern, western, and mountain bluebirds for each day of the year from 2009 through
2018 from eBird (eBird, 2019; Sullivan et al., 2009)
and processed these data using the auk package (v0.4.0,
Strimas-Mackey et al., 2018) for R (v3.6.2, R Core
Team, 2020). Following recommended best practices, this
processing involved filtering the eBird checklists to include
only those that were “Stationary” or “Traveling” (leaving
out those that were “Incidental” or “Historical”), had a duration of 0–5 h, a distance of between 0 and 5 km, and that
were “complete” (i.e., all species observed were recorded)
(Strimas-Mackey et al., 2020). Finally, we “zero-filled” the
data, a process that adds counts of 0 for each checklist that
did not include any bluebird observations.
We then followed the methodology of Supp et al.
(2015) to summarize presence by geographic location
using equal-area icosahedron hex grids. Hex grids were
created at a resolution of 23,323 km2 using the dggridR R
package (v2.0.3, Barnes, 2018). Presence was summarized
into daily measures by binning checklists into hexes by
date and then calculating the proportion of checklists

3 of 12

that included a bluebird of a given species (number of
checklists with a bluebird observed/total number of
checklists). For each species we calculated daily weighted
mean longitude and latitude using coordinates of the hex
centroids. These means were weighted by the proportion
of checklists that included observations of a bluebird species. This resulted in a measure of mean species presence
per day, per hex. Hexes without checklists were omitted
from the analysis.

General additive models of migration
Generalized additive models (GAMs) were used following
methodologies from Supp et al. (2015) to model the relationship between daily latitude/longitude and time for each species. GAMs were created with the mgcv R package (v1.8-31,
Wood, 2011) using a penalized regression spline-based
smoothing parameter with a basis dimension of 40 (i.e., k = 40) and a gamma of 1.5 (the degree of smoothing).
These models predicted smooth paths from the weighted
daily mean locations. These smoothed paths were then used
to predict the daily latitude and longitude of the mean population centroid for each species (Figure 1a–c).

Migration timing
Migration timing was calculated in two steps. First, a
coarse migration timing for each species in each year
was defined as the date the daily latitudes predicted
from the GAMs crossed southern or northern latitudinal
thresholds calculated for each season using specific ordinal date ranges. These latitudinal thresholds were
defined as the most northerly latitudinal extent during
the nonbreeding period and the most southerly latitudinal extent during the breeding period for each species.
Specifically, migrations that started or ended in the
south (i.e., the start of Northern Hemisphere spring
and end of fall migration) used the minimum latitude
of the upper limit of the 99% confidence band of
predicted daily locations calculated over ordinal dates
1–80 (spring) and 285–345 (fall) (as in Supp et al., 2015).
Migrations that started or ended in the north (i.e., the
end of Northern Hemisphere spring and the start of fall
migration) used the maximum latitude of the lower limit
of the 99% confidence band of predicted daily locations
calculated over ordinal dates 80–175 (spring) and
225–285 (fall).
Migration timing was then fine-tuned with segmented
regressions (segmented R package v1.0-0, Vito &
Muggeo, 2008). The date calculated using the thresholds

21508925, 2022, 12, Downloaded from https://esajournals.onlinelibrary.wiley.com/doi/10.1002/ecs2.4316 by Cochrane Canada Provision, Wiley Online Library on [29/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

ECOSPHERE

SONNLEITNER ET AL.

F I G U R E 1 Mean location of bluebird populations overall (a–c) and by latitude (d–f) over 10 years for eastern (a, d), mountain (b, e), and
western (c, f) bluebirds. In the maps (a–c), each point represents the daily mean weighted location of the population, colored by month; lines
represent generalized additive model (GAM) smoothed migration paths, one for each year, colored by month. The latitude plots (d–f) show
GAM smoothed predicted latitude (of the population centroid; black) overlaid with the calculated starts and ends of spring and fall migration.

in the previous step was used as the starting point, to
determine the break point where there was no longer a
relationship between predicted latitude and date. This
resulted in a more precise calculation of the start and end
of migration (as in Supp et al., 2015). We omitted the fall
values during 2009 and 2012 for western bluebirds as the
migration was too short and gradual for the start and
ends to be detected in these years.

Migration speed
Maximum daily migration speed for each species in each
year was calculated for both spring and fall migration as
the median distance traveled (in kilometers) over the five
fastest days in each period and is expressed as kilometers
per day (spring ordinal dates 1–175; fall ordinal dates
225–340; as in Supp et al., 2015). Date ranges extending

21508925, 2022, 12, Downloaded from https://esajournals.onlinelibrary.wiley.com/doi/10.1002/ecs2.4316 by Cochrane Canada Provision, Wiley Online Library on [29/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

4 of 12

into the nonbreeding period were used because there was
little movement in the nonbreeding period and to ensure
that start and end of migration were not excluded from
the calculations. As such, our measure of migration speed
should be interpreted as the maximum population-level
migration speed, rather than the flight speed of individual birds.

Maximum latitude and median longitude
The maximum latitude during the breeding season for
each species in each year was defined as the maximum
predicted daily latitude (of the population centroid)
between the end of spring migration and the start of fall
migration. The median longitude of both spring and fall
migrations for each species and each year was defined as
the median predicted daily longitude between the calculated start and end dates of each migration period. The
median breeding longitude was calculated as the median
predicted daily longitude between the end of spring
migration and the start of fall migration. As with migration timing, we omitted the fall values during 2009 and
2012 for western bluebirds as the migration was too short
and gradual for the start and ends to be detected in these
years.

Statistical analysis
We ran a series of linear models with explanatory parameters, species, year, and their interaction, with the following response variables: start/end of spring/fall migration,
spring/fall migration speed, maximum breeding latitude,
median breeding longitude, and median spring/fall
migration longitude. If we detected a significant interaction, we ran post hoc analyses as separate linear
regressions for each of the three species to examine the
relationship between year and the response variable
independently. If the interaction was not significant,
we removed it from the model. Significant differences
among species were explored using post hoc analyses
with the emmeans R package (v1.7.5, Lenth, 2018),
controlling for multiple testing using the false discovery rate method. Significance was evaluated using an
alpha value of 0.05 and analyses were completed using
R (v3.6.2, R Core Team, 2020). Model residuals were
assessed for normality and heteroscedasticity using the
DHARMa R package (v0.4.3, Hartig, 2021). Results are
presented as Type III ANOVAs (using the car R package v3.0-8, Fox & Weisberg, 2019), or, where interpretation of the slopes was required, as parameter
estimates.

5 of 12

Maps, data, and scripts
Maps were created using the R packages sf (v0.9-6,
Pebesma, 2018) and ggplot2 (v3.3.2, Wickham, 2016), with
data obtained from Natural Earth (https://naturalearthdata.
com) via the rnaturalearth R package (v0.1.0, South, 2017).
Raw data are available from eBird (eBird, 2019; Sullivan
et al., 2009). Scripts for data summarization, analysis, and
figures are available from Zenodo: https://doi.org/10.5281/
zenodo.6885688.

RESULTS
Migration timing
When examining migratory timing (Figure 1d–f), we
found no significant year  species interaction for the
start (F2,24 = 0.24, p = 0.79) and end (F2,24 = 2.24,
p = 0.13) of spring migration or the start (F2,22 = 0.66,
p = 0.53) and end (F2,22 = 0.91, p = 0.42) of fall migration. Therefore, the interaction terms were removed
from subsequent models (Appendix S1: Table S1). We
found no effect of year on the start (F2,26 = 0.08,
p = 0.78) or end (F2,26 = 1.02, p = 0.32) of spring migration or the start (F2,24 = 0.38, p = 0.54) and end
(F2,24 = 2.75, p = 0.11) of fall migration. We found no
significant effect of species on the start (F2,26 = 0.25,
p = 0.78) of spring migration, but did find a significant
effect on the end (F2,26 = 108.88, p < 0.0001) of spring
migration as well as on the start (F2,24 = 11.77, p = 0.0003)
and end (F2,24 = 5.68, p = 0.010) of fall migration.

Migration speed
When examining migration speed (Figure 2), the
interaction between year and species was not significant in spring (F2,24 = 0.75, p = 0.48) or fall
(F2,24 = 1.02, p = 0.38) and was subsequently removed
from the models (Appendix S1: Table S1). After
removing the interaction term from the model, there
remained no effect of year for spring (F2,26 = 0.94,
p = 0.34) or fall (F2,26 = 2.98, p = 0.10). There was
however an effect of species during both the spring
(F2,26 = 108.44, p < 0.0001) and fall (F2,26 = 58.27,
p < 0.0001).

Maximum latitude during breeding
When we examined whether the maximum breeding latitude of the population centroids of the three species

21508925, 2022, 12, Downloaded from https://esajournals.onlinelibrary.wiley.com/doi/10.1002/ecs2.4316 by Cochrane Canada Provision, Wiley Online Library on [29/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

ECOSPHERE

SONNLEITNER ET AL.

F I G U R E 2 Differences in the spring (a) and fall (b) migratory speed of eastern, western, and mountain bluebirds. Speed of migration
was calculated as the mean distance traveled (in kilometers) over the fastest 5 days. Boxplots represent the median value and interquartile
range for each species (Q1 below as 25th percentile of data and Q3 above as 75th percentile of data).

changed over time, the interaction between year and species was not significant (F2,24 = 0.68, p = 0.52) and was
subsequently removed from the model (Appendix S1:
Table S1). We found an overall significant effect of species (F2,26 = 865.25, p < 0.0001) and year (F2,26 = 4.69,
p = 0.0396) on maximum breeding latitude. Specifically,
maximum breeding latitudes shifted southward by 0.053
degrees per year.

Median longitude during migration
When we examined median annual longitude during
migration, we found a significant year by species interaction for both spring (F2,24 = 11.05, p = 0.0004; Figure 3a)
and fall migration (F2,22 = 14.46, p < 0.0001; Figure 3b).
We then examined the relationship between median longitude and year separately for each species (Appendix S1:
Table S2). The median longitude during spring migration
for both eastern (r 2 = 0.66, p = 0.009) and western bluebirds (r 2 = 0.71, p = 0.005) changed from 2009 to 2018,
with eastern bluebirds shifting their distribution westward by 0.22 degrees of longitude per year and western
bluebirds shifting eastward by 0.26 degrees of longitude

per year. By contrast, for mountain bluebirds, the median
longitude during spring migration did not change over
that time period (p = 0.21). Similar to spring migration,
the median longitude during fall migration in eastern
bluebirds shifted westward by 0.12 degrees of longitude
per year (r 2 = 0.59, p = 0.038), while the median longitude during fall migration in western bluebirds shifted
eastward by 0.31 degrees of longitude per year from 2009
to 2018 (r 2 = 0.68, p = 0.0002). As with spring migration,
for mountain bluebirds, the median longitude during fall
migration (p = 0.09) did not change over time.

Median longitude during breeding
When examining the median longitude of the population
centroid during breeding (Figure 3c), the interaction
between year and species was significant (F2,22 = 17.19,
p < 0.0001). We then examined the relationship between
median longitude and year separately for each species
(Appendix S1: Table S2). The median longitude during
the breeding period for eastern (r 2 = 0.69, p = 0.003),
western (r 2 = 0.52, p = 0.0007), and mountain bluebirds
(r 2 = 0.48, p = 0.045) changed from 2009 to 2018.

21508925, 2022, 12, Downloaded from https://esajournals.onlinelibrary.wiley.com/doi/10.1002/ecs2.4316 by Cochrane Canada Provision, Wiley Online Library on [29/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

6 of 12

7 of 12

(a)
Spring median longitude

−86.0
−86.5
−87.0
−87.5
−88.0
−110
−111
−112
−113
−114
−115
−116
−117

2010

2012

2014

2016

2018

2016

2018

2016

2018

Year

(b)

Fall median longitude

−86.5
−87.0
−87.5
−110
−111
−112
−113
−114
−115
−116
2010

2012

2014

Year

Breeding median longitude

(c) −86.0
−86.5
−87.0
−87.5
−111
−112
−113
−114
−115
−116
2010

2012

2014

Year
F I G U R E 3 Changes in the spring (a), fall (b), and breeding (c) median longitude during migration of eastern (black points), western
(yellow points), and mountain bluebirds (blue points) from daily population centroid longitudes and latitudes predicted using generalized
additive models. Bottom trend line shows western bluebirds shifting eastward whereas top trend line shows eastern bluebirds shifting
westward. The horizontal line denotes a split on the y-axis, in order to show all three species efficiently.

21508925, 2022, 12, Downloaded from https://esajournals.onlinelibrary.wiley.com/doi/10.1002/ecs2.4316 by Cochrane Canada Provision, Wiley Online Library on [29/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

ECOSPHERE

We found that eastern bluebirds shifted their distribution
westward by 0.16 degrees of longitude per year, western
bluebirds shifted eastward by 0.24 degrees of longitude
per year, and mountain bluebirds shifted eastward by
0.09 degrees of longitude per year.

DISC USS I ON
Using a 10-year community science dataset, we demonstrated consistent and rapid species-level changes in
migration patterns of North American bluebirds,
including shifts in migratory routes and breeding latitude and longitude. Somewhat surprisingly, bluebirds
do not appear to be advancing the timing of spring
migration and, in contrast to northward shifts seen in
many other species, the population centroids during
breeding appear to be shifting slightly southward. The
strongest pattern we detected was a longitudinal shift
during spring and fall migration, with both eastern and
western (but not mountain) bluebirds migrating closer
to the center of the continent. By examining three
North American bluebird species concurrently, we were
able to examine the differences in how these recently
divergent lineages are responding to a rapidly changing
planet.
There are some important considerations when
interpreting our results, given that we were examining
shifts in bluebird population centroids over time based
on eBird checklists. While any shifts in population centroids over time could be caused by overall shifts in the
population distribution, they could also be the result of
changes in abundance or spatial variation in abundance
over time. As discussed further below, potential northerly declining population trends could result in a contraction toward the core of the species range, resulting
in an apparent southward shift in maximum breeding
latitude in our case. Similarly, increases in abundance
in certain parts of a species’ range would tend to pull
the species’ centroid toward those areas, assuming
increased abundance leads to a higher proportion of
checklists including that species. Future studies could
incorporate spatial variation in abundance based on
BBS data (e.g., Bled et al., 2013) to control for this
possibility. Another caveat has to do with the
semi-structured nature of eBird data, especially when
compared with structured programs such as the BBS
that use randomly selected survey routes. While we
used standard methods to control for effort level
(e.g., only including complete checklists, weighting by
proportion of checklists that included observations of a
bluebird species, etc.), other sources of variation could
include observer effects and imperfect detection of

SONNLEITNER ET AL.

birds. However, we do not have a strong reason to
believe that these factors would change over time to
produce a bias in our results.
Arriving early to the breeding grounds has been
attributed to competition for mates and territories
(Kokko, 1999), and advancements in spring arrival
timing have been correlated to changing ecological conditions such as first-leaf and first-bloom dates (Buermann
et al., 2013). Youngflesh et al. (2021) examined 56 bird
species and found that fluctuations in spring phenology
at the breeding grounds were broadly associated with earlier spring arrival dates. By contrast, and contrary to our
predictions, we found no changes to the timing of spring
migration or the spring arrival of any bluebird species. In
addition to changes in breeding ground conditions,
nonbreeding ground conditions may also affect the
migratory timing of birds (Paxton et al., 2014;
Robson & Barriocanal, 2011). Although we did not
examine changes to breeding ground conditions, western and mountain bluebirds tend to arrive at their
breeding grounds earlier in the season when compared
with eastern bluebirds and other passerine species. We
found that both species had mean arrival dates as early
as March in some years and thus may be able to
endure current climate-induced advancements in early
spring phenology without yet needing to advance their
arrival. Alternatively, interannual variation in environmental conditions, or variation in the strength of
phenological adjustments throughout a species range
(Youngflesh et al., 2021), could have masked any
directional change in range-wide population-level
migration timing over the relatively short time period
that we examined.
While there were interspecific differences in migration speed (defined as the maximum population-level
migration speed), we found that there was no change
in the migratory speed of bluebirds over ten years.
These interspecific differences in migration speed may
be partially the result of divergent migratory strategies
resulting from differences in the routes, stopover sites,
and stopover durations the three species utilize along
their migration paths. Another possibility is that the
mountain bluebird population as a whole appeared to
be moving more quickly because mountain bluebirds
are the only species considered fully migratory. Similar
to the issue of spatial patterns in abundance described
above, the slower pace of eastern and western bluebirds could be the result of fewer members of these
partially migratory populations moving as a proportion
of the total population.
Poleward shifts have been observed across many taxa
(Parmesan, 2006; Parmesan & Yohe, 2003) and are
thought to be an ecological response to warmer

21508925, 2022, 12, Downloaded from https://esajournals.onlinelibrary.wiley.com/doi/10.1002/ecs2.4316 by Cochrane Canada Provision, Wiley Online Library on [29/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

8 of 12

conditions in both winter and early spring at higher latitudes (Buermann et al., 2013; Rushing et al., 2020). A
recent analysis of breeding distributions of 73 species of
North American migratory birds based on the BBS
showed that 55% of shifts were toward the north and
44% were toward the south from 1994 to 2017
(McCaslin & Heath, 2020). In this BBS study, bluebird
species either showed no change, or shifted toward the
north, depending on the species and region (the analysis
examined eastern, central, and western regions of North
America separately). Although our analysis is not
directly comparable due to different data sources
(eBird vs. BBS), time periods, and regional breakdown,
our results are contradictory in that they suggest that the
population centroids during breeding for the three bluebird species have shifted toward the south. As indicated
above, an apparent shift in breeding distributions toward
the south could potentially be explained by recently
(2009–2019) declining population trajectories, at least for
eastern and mountain bluebirds. However, McCaslin
and Heath (2020) found that across the suite of species
in their study, breeding distribution shifts could not be
explained by trends in abundance, although results were
variable across species and regions. An additional consideration is that we were unable to examine potential
altitudinal shifts in our analysis, and bluebirds could also
be responding to changing environmental conditions by
shifting their ranges to higher elevations (Tingley
et al., 2012). Taken together, our results and those of the
above studies suggest that bluebird range edges may be
shifting north, but that the center of abundance of their
populations may have shifted southward. Another interesting pattern we observed was that population centroids
during breeding tended to decline immediately following
the end of spring migration in most years (Figure 1d–f).
One likely explanation for this pattern is that young will
generally fledge earlier at lower latitudes, and thus more
bluebirds will be recorded at more southerly latitudes
earlier in the breeding season, prior to the onset of fall
migration.
The longitudinal shifts we detected during migration,
with western and eastern bluebirds shifting toward the
center of the continent, could potentially be explained by
the Pacific and Atlantic Oceans limiting coastward range
expansion, and the movement of human populations
toward coastal areas (Neumann et al., 2015; Seto
et al., 2011), reducing the available habitat in these
regions (Isaksson, 2018). Thus, range expansion during
migration could only occur toward the east for western
bluebirds and the west for eastern bluebirds. Species’
ranges may also shift in response to different population
trends occurring in different parts of the range
(Virkkala & Lehikoinen, 2017). For example, Hejl (1994)

9 of 12

and Duckworth (2009) proposed that changes in fire suppression regimes, along with nest box programs, have
led to improved nesting opportunities for western bluebirds, facilitating an increase in population size—also
supported by the BBS (see Introduction)—and an eastward range expansion. Potential range expansion of all
three species through increased nesting opportunities
may be supported by our finding that all three species
have experienced shifts to their breeding ground longitude. In contrast to migration longitude where mountain
bluebirds experienced no shift, we found that during
breeding, mountain bluebird population centroids did in
fact shift westward over the ten years we examined. This
indicates that there may be shifts in a species’ breeding
longitude without a concurrent shift in their migration
longitude.
With a changing environment due to climatic and
other anthropogenic factors, conservation of migratory
birds is increasingly critical, yet there is a lack of data
on migratory systems, particularly outside of the breeding period (Marra et al., 2015). While standardized surveys such as the BBS focus on breeding bird populations
and have been invaluable in estimating population
trends, among many other uses (Rosenberg et al., 2019;
U.S. Geological Survey and Canadian Wildlife Service,
2020), eBird presents the opportunity to track and monitor the movements and distributions of species throughout their entire annual cycle. Our findings demonstrate
that shifting patterns of migration and changes in the
mean population centroid (which could be the result of
shifts in species ranges) are occurring within bluebirds
and potentially other passerine species. For conservation
efforts to be effective, we will need to understand how
changes to population centroids reflect range-wide patterns of species’ distributions and why these changes are
occurring—a task that can be difficult using community
science data such as eBird. However, data from eBird
present an opportunity to address broadscale issues at a
low cost, which is beneficial as directed funding can be
limited and large-scale, long-term datasets are needed to
answer some of the most pressing large-scale conservation issues. In addition to monitoring shifting migrations and breeding population centroids, eBird can
identify high priority areas for designation and protection (Cañizares & Reed, 2020) or retroactively determine
the effectiveness of conservation efforts (Cazalis
et al., 2020).
ACKNOWLEDGMENTS
We are grateful to S. Supp and colleagues for making
their well-annotated code and raw data available from
their 2015 publication, without which this project
would not have been possible. A special thank you to

21508925, 2022, 12, Downloaded from https://esajournals.onlinelibrary.wiley.com/doi/10.1002/ecs2.4316 by Cochrane Canada Provision, Wiley Online Library on [29/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

ECOSPHERE

SONNLEITNER ET AL.

Dr. Sean Mahoney for guidance and assistance. Funding
for this work was provided through the Thompson
Rivers University Undergraduate Research Experience
Award Program (UREAP) to Jared Sonnleitner and an
NSERC Discovery Grant to Matthew W. Reudink.
CONFLICT OF INTEREST
The authors declare no conflict of interest.
DATA AVAILABILITY STATEMENT
Data and code (Sonnleitner & LaZerte, 2022) are available from Zenodo: https://doi.org/10.5281/zenodo.
6885688.
ORCID
Stefanie E. LaZerte
8360

https://orcid.org/0000-0002-7690-

R EF E RE N C E S
Allen, M. R., K. P. Shine, J. S. Fuglestvedt, R. J. Millar, M. Cain,
D. J. Frame, and A. H. Macey. 2018. “A Solution to the
Misrepresentations of CO2-Equivalent Emissions of
Short-Lived Climate Pollutants under Ambitious Mitigation.”
Climate and Atmospheric Science 1(16): 1–8.
Barnes R. 2018. “dggridR: Discrete Global Grids.” R Package
Version 2.0.3. https://CRAN.R-project.org/package=dggridR.
Bled, F., J. Sauer, K. Pardieck, P. Doherty, and J. A. Royle.
2013. “Modeling Trends from North American Breeding
Bird Survey Data: A Spatially Explicit Approach.” PLoS
One 8(12): e81867. https://doi.org/10.1371/journal.pone.
0081867.
Bridge, E. S., J. D. Ross, A. J. Contina, and J. F. Kelly. 2016. “Do
Molt-Migrant Songbirds Optimize Migration Routes Based on
Primary Productivity?” Behavioral Ecology 27(3): 784–92.
Buermann, W., P. Bikash, M. Jung, D. Burn, and M. Reichstein.
2013. “Earlier Springs Decrease Peak Summer Productivity in
North American Boreal Forests.” Environmental Research
Letters 8(2): 024027.
Cañizares, J. R., and J. M. Reed. 2020. “Identification of Priority
Shorebird Conservation Areas in the Caribbean.” PeerJ 8:
e9831.
Cazalis, V., K. Princé, J.-B. Mihoub, J. Kelly, S. Butchart, and A.
Rodrigues. 2020. “Effectiveness of Protected Areas in
Conserving Tropical Forest Birds.” Nature Communications
11(1): 4461.
Curley, S., L. Manne, and R. Veit. 2020. “Differential Winter and
Breeding Range Shifts: Implications for Avian Migration
Distances.” Diversity and Distributions 26(4): 415–25.
Duckworth, R. A. 2009. “Maternal Effects and Range Expansion:
A Key Factor in a Dynamic Process?” Philosophical
Transactions: Biological Sciences 364(1520): 1075–86.
Duckworth, R. A., and A. V. Badyaev. 2007. “Coupling of
Dispersal and Aggression Facilitates the Rapid Range
Expansion of a Passerine Bird.” Proceedings of the National
Academy of Sciences of the United States of America 104(38):
15017–22.

eBird. 2019. “eBird Basic Dataset.” Versions: ebd_easblu_200501_
201901_relNov-2019, ebd_moublu_200501_201901_relNov-2019,
ebd_wesblu_200501_201901_relNov-2019. Ithaca, NY: Cornell
Lab of Ornithology.
Environment Canada. 2017. North American Breeding Bird
Survey—Canadian Trends Website, Data-Version 2015.
Gatineau: Environment Canada.
Faaborg, J., R. Holmes, A. Anders, K. Bildstein, K. Dugger,
S. Gauthreaux, P. Heglund, et al. 2010. “Recent Advances in
Understanding Migration Systems of New World Land Birds.”
Ecological Monographs 80(1): 3–48.
Fox, J., and S. Weisberg. 2019. “car: Companion to Applied
Regression.” https://CRAN.R-project.org/package=car.
Gowaty, P. A., and J. H. Plissner. 2020. “Eastern Bluebird (Sialia
sialis), Version 1.0.” In Birds of the World, edited by A. F.
Poole. Ithaca, NY: Cornell Lab of Ornithology. https://doi.org/
10.2173/bow.easblu.01.
Guinan, J. A., P. A. Gowaty, and E. K. Eltzroth. 2020. “Western
Bluebird (Sialia mexicana), Version 1.0.” In Birds of the World,
edited by A. F. Poole. Ithaca, NY: Cornell Lab of Ornithology.
https://doi.org/10.2173/bow.wesblu.01.
Hartig, F. 2021. “DHARMa: Residual Diagnostics for Hierarchical
(Multi-Level/Mixed) Regression Models.” https://CRAN.Rproject.org/package=DHARMa.
Hejl, S. J. 1994. “Human-Induced Changes in Bird Populations in
Coniferous Forests in Western North America During the Past
100 Years.” Studies of Avian Biology 15: 232–46.
Huang, Q., J. R. Sauer, and R. O. Dubayah. 2017. “Multidirectional
Abundance Shifts among North American Birds and the
Relative Influence of Multifaceted Climate Factors.” Global
Change Biology 23(9): 3610–22.
Isaksson, C. 2018. “Impact of Urbanization on Birds.” In
Fascinating Life Sciences, edited by D. Tietze, 235–57. Cham:
Springer.
Johnson, L. S., and R. D. Dawson. 2020. “Mountain Bluebird (Sialia
currucoides), Version 1.0.” In Birds of the World, edited by
P. G. Rodewald. Ithaca, NY: Cornell Lab of Ornithology.
https://doi.org/10.2173/bow.moublu.01.
Jonzen, N. 2006. “Rapid Advance of Spring Arrival Dates in
Long-Distance Migratory Birds.” Science 312: 1959–61.
Kelly, J. F., J. R. Shipley, P. B. Chilson, K. W. Howard, W. F. Frick,
and T. H. Kunz. 2012. “Quantifying Animal Phenology in the
Aerosphere at a Continental Scale Using NEXRAD Weather
Radars.” Ecosphere 3(2): 16.
Kokko, H. 1999. “Competition for Early Arrival in Migratory
Birds.” Journal of Animal Ecology 68: 940–50.
La Sorte, F. A., and C. H. Graham. 2020. “Phenological Synchronization
of Seasonal Bird Migration with Vegetation Greenness across
Dietary Guilds.” Journal of Animal Ecology 90: 343–55.
La Sorte, F. A., and F. R. Thompson. 2007. “Poleward Shifts in
Winter Ranges of North American Birds.” Ecology 88(7):
1803–12.
Lenth, R. 2018. “lsmeans: Least Squared Means.” https://CRAN.Rproject.org/package=lsmeans.
Marra, P. P., E. B. Cohen, S. R. Loss, J. E. Rutter, and C. M. Tonra.
2015. “A Call for Full Annual Cycle Research in Animal
Ecology.” Biology Letters 11(8): 20150552.
Mayor, S. J., R. P. Guralnick, M. W. Tingley, J. Otegui, J. C. Withey,
S. C. Elmendorf, M. E. Andrew, S. Leyk, I. S. Pearse, and

21508925, 2022, 12, Downloaded from https://esajournals.onlinelibrary.wiley.com/doi/10.1002/ecs2.4316 by Cochrane Canada Provision, Wiley Online Library on [29/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

10 of 12

D. C. Schneider. 2017. “Increasing Phenological Asynchrony
between Spring Green-Up and Arrival of Migratory Birds.”
Scientific Reports 7: 1902.
McCaslin, H. M., and J. A. Heath. 2020. “Patterns and Mechanisms
of Heterogeneous Breeding Distribution Shifts of North
American Migratory Birds.” Journal of Avian Biology 51:
e02237.
Neumann, B., A. T. Vafeidis, J. Zimmermann, and R. J. Nicholls.
2015. “Future Coastal Population Growth and Exposure to
Sea-Level Rise and Coastal Flooding—A Global Assessment.”
PLoS One 10(3): e0131375.
Nilsson, C., R. Klaassen, and T. Alerstam. 2013. “Differences in
Speed and Duration of Bird Migration between Spring and
Autumn.” The American Naturalist 181(6): 837–45.
Parmesan, C. 2006. “Ecological and Evolutionary Responses to
Recent Climate Change.” Annual Review of Ecology, Evolution,
and Systematics 37: 637–69.
Parmesan, C., and G. Yohe. 2003. “A Globally Coherent Fingerprint
of Climate Change Impacts across Natural Systems.” Nature
421: 37–42.
Paxton, K. L., E. B. Cohen, E. H. Paxton, Z. Németh, and F. R.
Moore. 2014. “El Niño-Southern Oscillation Is Linked to
Decreased Energetic Condition in Long-Distance Migrants.”
PLoS One 9: e95383.
Pebesma, E. 2018. “sf: Simple Features for R.” https://CRAN.Rproject.org/package=sf.
Prytula, E. D., A. E. McKellar, L. Schwitters, and M. W. Reudink.
2021. “Rapid Advancement of Spring Migration and en Route
Adjustment of Migration Timing in Response to Weather during Fall Migration in Vaux’s Swifts (Chaetura vauxi).”
Canadian Journal of Zoology 100: 56–63.
R Core Team. 2020. R: A Language and Environment for Statistical
Computing. Vienna: R Foundation for Statistical Computing
https://www.R-project.org/.
Robson, D., and C. Barriocanal. 2011. “Ecological Conditions in
Wintering and Passage Areas as Determinants of Timing of
Spring Migration in Trans-Saharan Migratory Birds.” Journal
of Animal Ecology 80: 320–31.
Rosenberg, K. V., A. M. Dokter, P. J. Blancher, J. R. Sauer, A. C.
Smith, P. A. Smith, J. C. Stanton, et al. 2019. “Decline of the
North American Avifauna.” Science 366(6461): 120–4.
Rushing, C. S., J. A. Royle, D. J. Ziolkowski, and K. L. Pardieck.
2020. “Migratory Behavior and Winter Geography Drive
Differential Range Shifts of Eastern Birds in Response to
Recent Climate Change.” Proceedings of the National Academy
of Sciences of the United States of America 117(23): 12897–903.
Saino, N., R. Ambrosini, D. Rubolini, V. H. Jost, A. Provenzale,
K. Hüppop, O. Hüppop, et al. 2011. “Climate Warming, Ecological
Mismatch at Arrival and Population Decline in Migratory Birds.”
Proceedings: Biological Sciences 278(1707): 835–42.
Schwartz, M. D., R. Ahas, and A. Aasa. 2006. “Onset of Spring
Starting earlier across the Northern Hemisphere.” Global
Change Biology 12(2): 343–51.
Seto, K. C., M. Fragkias, B. Güneralp, and M. K. Reilly. 2011. “A
Meta-Analysis of Global Urban Land Expansion.” PLoS One
6(8): e23777.
Sonnleitner, J., and S. LaZerte. 2022. “steffilazerte/bluebird_
migration.” Zenodo. Data and Code. https://doi.org/10.5281/
zenodo.6885688.

11 of 12

South, A. 2017. “rnaturalearth: World Map Data from Natural
Earth.” https://CRAN.R-project.org/package=rnaturalearth.
Stanley, C. Q., M. Macpherson, K. C. Fraser, E. A. Mckinnon, and
B. J. M. Stutchbury. 2012. “Repeat Tracking of Individual
Songbirds Reveals Consistent Migration Timing but Flexibility
in Route.” PLoS One 7(7): e40688.
Strimas-Mackey, M., W. M. Hochachka, V. Ruiz-Gutierrez, O. J.
Robinson, E. T. Miller, T. Auer, S. Kelling, D. Fink, and
A. Johnston. 2020. Best Practices for Using eBird Data. Version
1.0. 41. Ithaca, NY: Cornell Lab of Ornithology. https://doi.
org/10.5281/zenodo.3620739, https://cornelllabofornithology.
github.io/ebird-best-practices/.
Strimas-Mackey, M., E. Miller, and W. Hochachka. 2018. “auk:
eBird Data Extraction and Processing With AWK.” https://
cornelllabofornithology.github.io/auk/.
Sullivan, B. L., J. L. Aycrigg, and J. H. Barry. 2014. “The eBird
Enterprise: An Integrated Approach to Development and
Application of Citizen Science.” Biology Conservation 169: 31–40.
Sullivan, B. L., C. L. Wood, M. J. Iliff, R. E. Bonney, D. Fink, and
S. Kelling. 2009. “eBird: A Citizen-Based Bird Observation
Network in the Biological Sciences.” Biological Conservation
142: 2282–92.
Supp, S. R., F. A. La Sorte, T. A. Cormier, M. C. W. Lim, D. R.
Powers, S. M. Wethington, S. Goetz, and C. H. Graham. 2015.
“Citizen-Science Data Provides New Insight into Annual and
Seasonal Variation in Migration Patterns.” Ecosphere 6(1): 15–9.
Thomas, C. D., and J. J. Lennon. 1999. “Birds Extend their Ranges
Northwards.” Nature 399(6733): 213.
Tingley, M. W., M. S. Koo, C. Moritz, A. C. Rush, and S. R.
Beissinger. 2012. “The Push and Pull of Climate Change
Causes Heterogeneous Shifts in Avian Elevational Ranges.”
Global Change Biology 18(11): 3279–90.
U.S. Geological Survey and Canadian Wildlife Service. 2020.
“Strategic Plan for the North American Breeding Bird Survey,
2020–30.” U.S. Geological Survey Circular 1466, 10 pp. https://
doi.org/10.3133/cir1466. [Supersedes USGS Circular 1307].
Virkkala, R., and A. Lehikoinen. 2014. “Patterns of
Climate-Induced Density Shifts of Species: Poleward Shifts
faster in Northern Boreal Birds than in Southern Birds.”
Global Change Biology 20: 2995–3003.
Virkkala, R., and A. Lehikoinen. 2017. “Birds on the Move in the
Face of Climate Change: High Species Turnover in Northern
Europe.” Ecology and Evolution 7: 8201–9.
Visser, M. E., and C. Both. 2005. “Shifts in Phenology Due to Global
Climate Change: The Need for a Yardstick.” Proceedings of the
Royal Society B: Biological Sciences 272: 2561–9.
Vito, M., and R. Muggeo. 2008. “Segmented: An R Package to Fit
Regression Models with Broken-Line Relationships.” R News
8/1: 20–5. https://cran.r-project.org/doc/Rnews/.
Walther, G., E. Post, P. Convey, A. Menzel, C. Parmesan, T. J. C.
Beebee, J. Fromentin, O. Hoegh-Guldberg, and F. Bairlein.
2002. “Ecological Responses to Recent Climate Change.”
Nature 416(6879): 389–95.
Wickham, H. 2016. ggplot2: Elegant Graphics for Data Analysis.
New York: Springer-Verlag https://ggplot2.tidyverse.org.
Wood, S. N. 2011. “Fast Stable Restricted Maximum Likelihood and
Marginal Likelihood Estimation of Semiparametric Generalized
Linear Models.” Journal of the Royal Statistical Society Series B
73(1): 3–36.

21508925, 2022, 12, Downloaded from https://esajournals.onlinelibrary.wiley.com/doi/10.1002/ecs2.4316 by Cochrane Canada Provision, Wiley Online Library on [29/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

ECOSPHERE

Youngflesh, C., J. Socolar, B. R. Amaral, A. Arab, R. P. Guralnick,
A. H. Hurlbert, R. LaFrance, S. J. Mayor, D. A. W. Miller, and
M. W. Tingley. 2021. “Migratory Strategy Drives Species-Level
Variation in Bird Sensitivity to Vegetation Green-Up.” Nature
Ecology and Evolution 5: 987–94.

S UP PO RT ING IN FOR MAT ION
Additional supporting information can be found online
in the Supporting Information section at the end of this
article.

SONNLEITNER ET AL.

How to cite this article: Sonnleitner, Jared,
Stefanie E. LaZerte, Ann E. McKellar, Nancy
J. Flood, and Matthew W. Reudink. 2022. “Rapid
Shifts in Migration Routes and Breeding Latitude
in North American Bluebirds.” Ecosphere 13(12):
e4316. https://doi.org/10.1002/ecs2.4316

21508925, 2022, 12, Downloaded from https://esajournals.onlinelibrary.wiley.com/doi/10.1002/ecs2.4316 by Cochrane Canada Provision, Wiley Online Library on [29/12/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License

12 of 12