Blinded Manuscript

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

Click here to access/download;Blinded
Manuscript;Swift_connectivity_Blinded.docx

ABSTRACT
Like many aerial insectivores, Vaux’s swift (Chaetura vauxi) populations are rapidly declining
and therefore determining when and where populations are limited across the annual cycle is
important for their conservation. Establishing the linkages between wintering and breeding sites
and the strength of the connections between them is a necessary first step. In this study, we
analyzed three stable isotopes (δ13C, δ15N, δ2H) from feathers collected during spring migration
from Vaux’s swifts that perished during a stopover on Vancouver Island, British Columbia,
Canada. We previously analyzed claw tissue (grown during winter) from the same individuals,
revealing that the swifts likely wintered in two to three locations/habitats. Here, we used stable
isotope analysis of flight feathers presumed to have been grown on, or near the breeding grounds
to determine the likely previous breeding locations and presumed destinations for the swifts.
Stable isotope values (δ13C, δ15N, δ2H) showed no meaningful variation between age classes,
sexes or with body size. Surprisingly, ~26% of the birds sampled had feather isotope values that
were not consistent with growth on their breeding grounds. For the remaining birds, assigned
breeding origins appeared most consistent with molt origins on Vancouver Island. Overall,
migratory connectivity of this population was relatively weak (rM = 0.07). However, the degree
of connectivity depended on how many winter clusters were analyzed; the two cluster solution
suggested no significant connectivity, but the three cluster solution suggested weak connectivity.
It is still unclear whether low migratory connectivity as observed for Vaux swift and other aerial
insectivores may make their populations more or less vulnerable to habitat loss, therefore further
efforts should be directed to assessing whether aerial insectivores may be habitat limited
throughout the annual cycle.

1

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

INTRODUCTION
Migratory birds move hundreds to thousands of kilometres each year on their annual
journey between breeding and wintering sites. While we have some understanding of species
range-wide distribution on both the breeding and overwintering grounds for most migratory
birds, connecting specific breeding and wintering populations (i.e., migratory connectivity)
remains a major challenge. Strong connectivity occurs when individuals from specific breeding
populations winter together in one or a few distinct locations; weak connectivity occurs when
individuals from a specific breeding population disperse to multiple over-winter locations
(Webster et al. 2002). If breeding and wintering populations are strongly linked, stochastic
events (e.g., storms, disease) can be a serious conservation concern (Webster and Marra 2005).
Migratory connectivity can also affect evolutionary processes, influencing the ability of species
and populations to adapt to changing environmental conditions (Webster et al. 2002).
Delineating patterns of migratory connectivity has been constrained in large part by the
inherent difficulty of tracking small birds throughout their annual cycle. Despite this challenge,
several methods have been employed to establish patterns of migratory connectivity, most
notably large-scale banding efforts for mark-recapture studies. Unfortunately, recapture rates are
notoriously low, especially in tropical regions and for non-harvested species, and this method has
thus been largely ineffective for connecting breeding and wintering populations of small bodied
species, including many passerines (Webster et al. 2002, Van Wilgenburg and Hobson 2011).
The advent of light-level geolocators and miniaturized GPS devices is revolutionizing our ability
to track smaller birds throughout the annual cycle (Stutchbury et al. 2009). However, this
approach is limited by high cost and the necessity to recapture individuals to retrieve data, which
often limits samples sizes. In addition, extrinsic markers such as geolocators and GPS tags are

2

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

often not as useful as intrinsic markers for determining diet and habitat use during different
phases of the annual cycle.
Stable isotope analyses provide an alternative to the use of extrinsic markers in the study
of avian migratory connectivity. While the spatial resolution of inferred origins using stable
isotope analyses tends to be lower than that obtained by using extrinsic markers, they have been
used extensively to link breeding and wintering populations (Hobson and Norris 2008). In
addition, stable isotope analyses have also been used to infer diet and habitat use (Bearhop et al.
2004, Hobson and Norris 2008). Stable isotopes are incorporated into growing tissues from the
environment and by understanding the distribution of stable isotopes in the environment, one can
infer information about the location of tissue growth (Wassenaar 2008). A further advantage is
that this technique requires only a single capture event and can be applied broadly across many
individuals relatively inexpensively, thus allowing for the large sample sizes necessary to
ascertain population-level migratory connectivity. Furthermore, multiple intrinsic markers can be
assayed from the same tissue sample, potentially allowing for improved precision and accuracy
of inference (Haché et al. 2014).
Here, we employ a multiple stable isotope approach, using hydrogen (δ2H), carbon
(δ13C), and nitrogen (δ15N) to examine patterns of migratory connectivity in a population of
Vaux’s swifts at a northern migratory roost site. Much like other aerial insectivores (Nebel et al.
2010), Vaux’s swifts have experienced long-term (1970-2017) declines (-35%, 95% CI: -60.8 –
7.6%; Smith A.C. et al. unpublished, an update of Environment and Climate Change Canada
2017) and it is unclear which stage(s) of the annual cycle are most limiting for this species (Bull
and Collins 2007). The severity of declines in aerial insectivores has been linked to migration
distance, land-use change (Nebel et al. 2010) and may be related to a decline in food resources

3

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

(Nocera et al. 2012); declines appear to be more pronounced in populations that inhabit more
northern areas of their range (Fraser et al. 2012). An improved understanding of migratory
connectivity for this species is clearly needed (Reudink et al. 2015) and would facilitate where
and how to allocate conservation actions (Hobson et al. 2014).
Here we analyze the stable isotope composition of flight feathers grown post breedingseason the previous year (Pyle 1997) from Vaux’s swifts for which stable-isotope values of claw
material grown on the wintering grounds were analyzed (Reudink et al. 2015). Our objectives
were to estimate the approximate breeding distribution of this population, determine whether
birds from the previously delineated wintering areas were linked to distinct breeding populations,
and estimate the strength of migratory connectivity between breeding and wintering areas. Given
the suggestion of large movements on the wintering grounds for species such as Black Swift
(Cypseloides niger borealis; Beason et al. 2012) and other species of swift (Åkesson et al. 2012,
Liechti et al. 2013), we hypothesized that Vaux’s swifts would show relatively low migratory
connectivity and that breeding origins would not likely be tightly linked to previously defined
wintering clusters. In addition, we hypothesized that wing length would be negatively related to
δ2Hf (i.e., the individuals that had the lowest δ2Hf values would have the longest wings) since
migration distance should exert selective pressure on wing morphology and favor longer wings
for populations migrating farther (Nowakowski et al. 2014).

METHODS
Field Methods
On 9 May 2012, approximately 1,350 Vaux’s swifts became trapped and subsequently
suffocated in a migratory roost in Cumberland, British Columbia. The swifts used this roost
4

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

annually during northward spring migration and this single event resulted in a loss of
approximately 1.5-2.7% of the British Columbia population (Reudink et al. 2015). We analyzed
feather tissue from 97 individuals previously sampled by Reudink et al. (2015) for stable isotope
analysis of claw tips (used to infer winter habitat/location).
We sampled approximately 0.5 mg of the distal tip of the first primary (P1) feather from
each individual, taking care to avoid the rachis. We sampled the P1 feather as it is generally the
first flight feather replaced during molt in approximately late June (Bull and Collins 1993) and
therefore is likely to reflect conditions on the breeding grounds.

Stable Isotope Analysis
To determine the approximate breeding locations from this sample of Vaux’s swifts, we
examined stable isotope values (δ2H, δ13C, and δ15N) in primary flight feathers, which should be
grown on, or close to the breeding grounds. δ2H has been used frequently in studies of avian
migration and connectivity because values vary with both latitude and elevation. At more
northern latitudes and higher elevations, 2H is relatively less abundant (Hobson 2011), resulting
in more negative isotope values in the growing feathers. δ13C varies in plant tissues based on the
photosynthetic system and water stress in the environment; individuals occupying habitats with
higher proportions of C4 plants and plants that are under high water stress tend to exhibit less
negative δ13C values (Lajtha and Marshall 1994, Cerling et al. 1997; Still and Powell 2010).
Nitrogen isotope ratios (δ15N) offer a marker of trophic position because the heavier isotope
(15N) is preferentially incorporated into consumer tissues relative to the lighter isotope (14N) and
as a result, δ15N values tend to increase with trophic level (Post 2002, Poupin et al. 2011). δ15N

5

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

can also reflect the relative humidity of a biome, with δ15N being negatively correlated with
rainfall and positively correlated with temperature (Sealy et al. 1987, Craine et al. 2009).
Stable isotope analyses were conducted at the Smithsonian Institution OUSS/MCI Stable
Isotope Mass Spectrometry Facility in Suitland, Maryland, USA. Feathers were first washed in a
2:1 chloroform:methanol solution, and allowed to acclimate to lab conditions for 72 hours in a
fume hood prior to sample preparation. The samples were pyrolyzed in a Thermo TC/EA
elemental analyzer (Thermo Scientific, Waltham, Massachusetts, USA) at 1,350 °C and analyzed
using a Thermo Delta V Advantage isotope ratio mass spectrometer. We ran calibrated standards
for every 10 samples, which included the hydrogen standard International Atomic Energy
Agency CH-7 and three additional standards (Kudu Horn Standard (KHS): -54.1 ± 2.3‰,
Caribou Hoof Standard (CBS): -197.3 ± 2.0‰, Spectrum keratin: -121.6 ± 2.5‰). Using
methods from Wassenaar and Hobson (2003), any non-exchangeable δ2H values were corrected
to the keratin standard; these were then repeatable to within 3 ± 2‰ (mean ± SD, n = 10). Two
in-house standards (acetanalide and urea) were run for every 10 samples of both stable carbon
and nitrogen analysis. Stable isotope values are expressed in parts per thousand (‰) deviation
from international standards Vienna standard mean ocean water (VSMOW; hydrogen), Pee Dee
Belemnite (PDB; carbon), and air (nitrogen) using the following equation:
δX = {[(Rsample-Rstandard)-1] x 1000}
where X is the isotope ratio of interest (δ2H, δ13C, or δ15N) and R is the corresponding ratio
(2H:1H, 13C:12C or 15N:14N). Carbon and nitrogen samples were repeatable to within ± 0.2 ‰
based on repeated measurements of standards.

6

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

Statistical Analysis
We examined putative breeding origins of Vaux’s swifts using previously described
spatially explicit assignments to molt origin based on analysis of δ2H in feathers (Wunder 2010,
Van Wilgenburg and Hobson 2011, Vander Zanden et al. 2014). While previous work suggests
Vaux’s Swifts molt P1 on their breeding grounds during the boreal summer (Bull and Collins
1993), preliminary examination of our data suggested that some individuals in our sample may
not be isotopically consistent with the breeding season isoscape, even after accounting for
measurement and assignment error. Some individuals molt on migration or display interrupted
molt, which is completed on the non-breeding grounds (Howell 2010), may have had
adventitious feather replacement, or may have originated from regions in which the food web has
been altered by irrigation (see Hobson et al. 2012) and thus the isoscape may not accurately
predict origins. Therefore, prior to assigning birds to their breeding grounds, we first used
previously described, spatially-explicit approaches to assign birds to one of three potential molt
origins (wintering grounds, migratory route or breeding grounds) using a Bayesian framework to
compare observed feather δ2H values against predictions from calibrated feather isoscapes
(Wunder 2010, Van Wilgenburg and Hobson 2011, Vander Zanden et al. 2014). To begin, we
obtained δ2H in feathers from a subset of known-origin birds from data in Hobson et al. (2012).
Specifically, we used a subset of known origin birds that were either aerial insectivores (Tree
Swallow (Tachycineta bicolor; n=30) and Least Flycatcher (Empidonax minimus, n=4), and
American Redstarts (Setophaga ruticilla, n=55), which also forage on the wing. For each capture
location of the known-origin birds, we queried the predicted amount-weighted mean δ2H value
growing-season precipitation (δ2Hp) isoscape of Bowen et al. (2005). We then used 1000 nonparametric bootstrap regressions in which we randomly selected 25 known-source birds with

7

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

replacement to derive mean slope and intercept estimates relating δ2Hf to δ2Hp. This resulted in a
mean rescaling function of δ2Hf = -41.24 (SD = 9.28) + 0.81 (SD = 0.10) * δ2Hp. We derived a
predicted mean feather isoscape via rescaling of the Bowen et al. (2005) isoscape using the
aforementioned algorithm. The resulting feather isoscape had predicted δ2Hf values between 147.0 – -55.5 ‰ (Figure 1a). In order to derive a raster (pixel based) map depicting the variance
2
(𝜎𝑟𝑒𝑠𝑐𝑎𝑙𝑒
) of the rescaled δ2Hf isoscapes, we also saved separate calibrated isoscapes from each

iteration of the bootstrap (separately applied each regression equation) and calculated the
variance between the 1000 isoscapes on a pixel-by-pixel basis using the “cellStats” function of
the raster package (Hijmans 2015) in the R statistical computing environment, version 3.4.3 (R
Core Team 2017). We also calculated the mean variance of the residuals across the 1000
2
regressions and used this value to represent between individual variance (𝜎𝑖𝑛𝑑𝑖𝑣𝑖𝑑𝑢𝑎𝑙
). Finally,

following Vander Zanden et al. (2014) we derived a spatially explicit prediction error surface by
2
2
taking the square root of the sum of 𝜎𝑟𝑒𝑠𝑐𝑎𝑙𝑒
+ 𝜎𝑖𝑛𝑑𝑖𝑣𝑖𝑑𝑢𝑎𝑙
. The resulting prediction error surface

ranged in values from 13.5 – 15.3 ‰ (Figure 1b) which we used to propagate assignment error
(below).
We assigned the birds to origins by applying normal probability density functions to
provide a spatially-explicit assessment of the likelihood that a given feather was grown at a given
location within the predicted mean δ2Hf isoscape derived above. We parameterized the normal
probability density function by treating isoscape-predicted δ2Hf values (within every pixel) as the
mean values against which the δ2Hf was compared and by treating our spatially explicit
prediction error surface (Figure 1b) as the variance estimate (Vander Zanden et al. 2014). This
yielded one posterior probability density map per individual. To determine whether each
individual had likely grown its feather on the breeding grounds versus on wintering range or on

8

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

migration, we summed the posterior probability densities of all pixels falling within polygons
defining the breeding, migration and wintering ranges, respectively. To accomplish this, we used
the ‘zonal’ function in the raster package (Hijmans 2015) to calculate the sum of the posterior
probabilities of falling within polygons of the breeding, migratory and wintering ranges from a
digital species distribution map (BirdLife International and Handbook of the Birds of the World
2016). Each bird was then assigned to molt origins within one of breeding, migratory or
wintering ranges, based upon the region that had the highest cumulative likelihood of
representing a molt origin.
To determine location of breeding grounds (in the year prior to collection) for each
individual that was assessed as likely having molted on the breeding grounds (n = 72), we
selected the upper 67% of the cumulative probability distribution (consistent with 2:1 odds of
being correct versus incorrect) and recoded these as “likely” origins (1) and treated the lower
33% of the distribution as unlikely (0). We then summed the results of each of the individual
assignments by addition of the predicted origin surfaces (Hobson et al. 2009, Van Wilgenburg
and Hobson 2011).
We examined whether isotopic variation in breeding origin Vaux’s swifts was associated
with age, sex, wing length, and wintering ground cluster using a series of competing regression
models. Preliminary modeling of variation in δ13C, δ15N, and δ2H using linear models suggested
heterogeneity of variance and convergence issues, and therefore we modeled the data using
Bayesian robust regression to fit competing models to the data using the “brms” package
(Bürkner 2017) by specifying the student family and an identity link function in the R statistical
computing environment, version 3.4.3 (R Core Team 2017). Prior to analysis, we removed one
individual from the sample due to a missing value for δ15N and another for which the sex was

9

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

undetermined, resulting in 70 individuals for which we had complete data to model. We treated
each isotope ratio (δ13C, δ15N, δ2H) as the response variable in a series of analyses in which we
fit 11 candidate models to our data, with the least parameterized model containing only an
intercept (null model) and the most heavily parameterized model containing factors for age, sex,
wintering ground cluster and a linear covariate for wing length. Owing to sample size, we only
considered models with additive effects. We ran four chains, each with 30,000 samples, a burn-in
of 20,000, and thinning rate of 10, and assessed model convergence by inspecting trace plots and
posterior predictive checks, and the Gelman-Rubin statistic (Gelman and Rubin 1992). Finally,
we used the Widely Applicable Information Criterion (WAIC; Watanabe 2010) to rank our
competing models, and considered models within 2 WAIC units as potentially competitive.
We assessed the strength of migratory connectivity by calculating the Mantel correlation
coefficient (rM, range -1 to 1; sensu Ambrosini et al. 2009). Unlike previous applications which
have used the Mantel test between pairwise distance matrices of breeding and wintering sites, we
constructed pairwise distance matrices based solely on isotopic variation in δ13C, δ15N and δ2H
of tissues grown on the wintering (claws) versus breeding (feathers) grounds. Prior to analyses,
data were centered and standardized to zero mean and unit variance. Positive correlations
between the distance matrices indicate that individuals that breed closely together also winter
closely together (Ambrosini et al. 2009). We repeated this analysis for all individuals assessed as
having molted on the breeding grounds, and on subsets based on their wintering ground clusters
derived from Reudink et al. (2015).

10

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

RESULTS
Based on spatially explicit assignments to the δ2Hf isoscapes (Figure 2–4), 72 swifts were
classified as having grown their feathers on the breeding ground. The cumulative likelihoods
within the breeding ground portion of the isoscape were greater than the cumulative likelihoods
in either the migratory portion of the range or on the wintering grounds for these individuals. The
remaining 25 individuals were classified as molt migrants based on feather δ2H being most
consistent (highest likelihoods) with portions of the isoscape associated with areas of the species
range used on migration. None of the 97 Vaux’s swifts were assigned molt origins within the
species wintering range.
We detected substantial variation for the three stable isotopes (Table 1). Among the three
isotopes, only δ13Cf and δ2Hf were correlated with each other (Table 2); however, there was no
relationship between the feather and claw isotope values for the subset of 71 birds with complete
isotope data that molted on the breeding grounds (Supplemental Material Table S1). Despite
substantial variation in all three isotopes within feathers, there was little support for any of our
covariates driving variation δ2Hf, δ15Nf, or δ13Cf as in all three cases the Null model (intercept
only) received greater support than all other candidate models based on WAIC (Table 3). There
was weak evidence (ΔWAIC = 0.06) for there being a possible median difference of almost 5‰
in δ2Hf between males and females; however, both sexes had broad 95% credible intervals
(Female median = -116.1‰, 95% CI= -123.4 − -108.9 ‰; Male median = -111.1‰, 95% CI= 118.0 − -105.7‰). Based on WAIC, we did not find strong support for an effect of wintering
ground cluster (ΔWAIC relative to the null model of 1.01 and 1.37 respectively for the two and
three cluster solutions), age, or wing length on δ2Hf (Table 3). Similarly, WAIC model selection
suggested weak support for sex and wing length contributing to variation in δ15N (ΔWAIC for

11

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

the Sex + Wing model = 1.16 and 1.64 for the model including only Sex; Table 3). However,
95% credible intervals for the effect of sex (β = -0.01 and 95% CI = -0.49 − 0.46 for the Sex +
Wing length model; β = 0.00 and 95% CI = -0.46 − 0.50 for the model including only Sex) and
wing length (β = -0.02 and 95% CI = -0.11 − 0.07) suggested no biologically meaningful
variation was associated with these parameters. Finally, models exploring variation in δ13Cf
within 2 WAIC units of the Null model included parameters for sex, wing length and age (Table
3); however, parameters from these models showed no biologically meaningful variation (Sex β
= -0.25 and 95% CI = -0.69 − 0.20; Wing length β = 0.04 and 95% CI = -0.04 − 0.12; and Age β
= -0.34 and 95% CI = -1.14 − 0.45).
Regardless of which wintering cluster individuals were associated with (as per Reudink et
al. 2015), the vast majority of Vaux’s swifts in our sample were assigned to breeding origins that
overlapped the roost location where the birds were sampled (Figures 2–4). While assignment
uncertainty precludes us from excluding other portions of the breeding range as possible origins
(Figures 2–4), a local (Vancouver Island) origin is the most parsimonious assignment for the vast
majority of birds in our sample.
Overall, Vaux’s swifts displayed weak migratory connectivity (rM = 0.07, P = 0.11). The
putative strength of migratory connectivity, however, differed between wintering ground clusters
defined based on isotopic analysis of claws. In contrast with cluster 3, which displayed positive
migratory connectivity (rM = 0.20, P = 0.02), neither cluster 1 (rM = -0.11, P = 0.92) nor cluster 2
(rM = -0.02, P = 0.53) showed any significant correlation between wintering and breeding ground
distance matrices constructed using isotopic variation (δ13C, δ15N and δ2H) between claws
(wintering) versus feathers (breeding).

12

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

DISCUSSION
We examined the approximate breeding origins and patterns of migratory connectivity in
Vaux’s swifts recovered in spring from a migratory roost on Vancouver Island, Canada. Stable
isotope analysis (δ13C, δ15N and δ2H) revealed that the swifts were primarily migrating to local
breeding locations, likely on Vancouver Island. Consistent with recent work by Pyle et al.
(2018), we also determined that a substantial proportion of birds (27 out of 99 sampled) were
isotopically consistent with being molt-migrants, exhibiting δ2Hf values that were not consistent
with the breeding season isoscape, or may have grown their feathers in otherwise isotopically
enriched habitats. All three stable isotopes exhibited substantial isotopic variation, which likely
reflects the diverse array of habitats (e.g., coastal, inland, mountain) in coastal British Columbia.
While new technologies have revolutionized the study of migratory connectivity, strength
of migratory connectivity has been quantified for relatively few species (Table 4), mostly during
stationary periods of the annual cycle. Although Vaux’s swifts displayed some evidence of using
isotopically similar sites between breeding and wintering grounds, the inferred strength of
connectivity was weak (rM = 0.07; range: -0.11 − 0.20) when compared with most other species
(see Table 4). For example, Swainson’s thrush (Catharus ustulatus) exhibit high connectivity (rM
= 0.72; Cormier et al. 2013), with individuals from the same breeding population wintering
within the same general area of Mexico. On the other end of the spectrum, the barn swallow
(Hirundo rustica) exhibits extremely weak migratory connectivity (rM = 0.025 – 0.22; Ambrosini
et al. 2009); barn swallows in that population neither wintered nor bred together. The low
connectivity we observed for Vaux’s swift fit our a priori hypothesis, and may be generally true
for aerial insectivores given the weak migratory connectivity estimated for other species
(Ambrosini et al. 2009, Fraser et al. 2017), the generally low connectivity of long-distance

13

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

migrants in general (Finch et al. 2017), and the large-scale movements within the non-breeding
season by some aerial insectivores (Åkesson et al. 2012, Beason et al. 2012, Liechti et al. 2013).
The relatively weak connectivity we found for Vaux’s swifts roosting together during spring
migration suggests that factors (e.g., habitat loss) influencing populations on the wintering
grounds should have relatively weak impacts spread over multiple breeding populations.
Alternatively, the weaker connectivity we observed may in part owe to mixing of individuals on
migration. Future efforts examining strength of connectivity of birds on stopover would provide
useful insight as to whether population mixing is greater along migratory routes than during
stationary periods of the annual cycle.
Although we found evidence for a connection between wintering and breeding grounds,
the strength of migratory connectivity depended upon which wintering cluster the individuals
were associated with. Our initial analysis (Reudink et al. 2015) revealed that the swifts likely
originated from two or three clusters. When we examined connectivity using the two cluster
solution, there was no evidence of connectivity between winter clusters and breeding origins.
However, when we examined the three cluster solution, we did identify weak positive migratory
connectivity. While it is possible the finding of connectivity in the three cluster solution is
spurious, the fact that we found evidence of connectivity in the three cluster, but not two cluster
solution, may lend support to the idea that these swifts originated from three, rather than two,
distinct wintering locations/habitats. We cannot completely preclude the possibility that birds
may have been panmictic and simply selecting isotopically distinct habitats within the same
region; however, a lack of sex-based differences (see below) suggests this may be less probable.
Calibration of wintering ground multi-isotope claw isoscapes would further increase our
confidence in the strength of connectivity estimates for the analysis based on our three cluster

14

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

solution. Alternatively, future work could deploy miniaturized tracking devices such as archival
GPS tags (Hallworth and Marra 2015) on Vaux’s swift on the breeding grounds and multiisotope assays of claw material from recaptured individuals could be used to assess whether birds
from differing wintering ground isotopic clusters were indeed spatially segregated.
Segregation of migratory birds during the non-breeding season by age, sex and breeding
origins has previously been described for several species (Marra 2000; Valdez-Juárez et al.
2018). In some instances, competition amongst individuals may result in varying age, sex or
body size ratios along gradients of habitat quality, some of which may be isotopically distinct
(Marra et al. 1998). For example, older American redstart (Setophaga ruticilla) males hold
territories in high quality mangrove habitats and competitively exclude females and young
males, relegating them to neighboring, lower quality scrub habitat (Marra et al. 1993; Marra
2000). The Vaux’s swift does not appear to exhibit sex-based segregation during winter
(Reudink et al. 2015) and our results here also indicate that the sexes do not use isotopically
distinct habitats or locations during the post-breeding molt. Along with a lack of sex-based
segregation, we found no relationship between body size (wing or tail length) and all three stable
isotopes. The lack of relationship between body size and δ2Hf in particular may indicate that
body size does not affect migration distance; however, this relationship is unlikely to be apparent
at such a fine spatial scale (localized) and may instead only be visible at a range-wide scale.
Assignment of individuals to a feather isoscape revealed that a large portion (25.8%) of
the sample may have molted during migration, south of the breeding region, or were otherwise
isotopically inconsistent with this portion of the isoscape, while the remaining individuals
(74.2%) likely molted on or near the breeding grounds. While this result appears in stark contrast
to work by Bull and Collins (1993) that suggests P1 feathers should reflect the breeding grounds,

15

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

our results do appear generally consistent with recent analyses by Pyle et al. (2018) and Tonra
and Reudink (2018) that point to molt migration being more commonplace than previously
described in the literature. Indeed, stable isotope analysis has previously revealed the apparent
ubiquity of molt-migration and is proving indispensable for determining the location of feather
growth for both small and large-bodied birds (Rohwer et al. 2015) and can be effectively used on
species that migrate thousands of kilometers (Hobson et al. 2014). While it is possible that some
individuals may have grown P1 in isotopically enriched food-webs within the breeding range
(e.g. areas influenced by irrigation), it seems improbable that such a high proportion would
derive from enriched food-webs. Here, we used a likelihood-based method to link the isotopic
signatures of P1 feathers to the period of the annual cycle during which they were most likely
grown. We suggest future work should employ similar methods to ours to assess whether molt
may have occurred on versus off the breeding grounds given the apparent prevalence of molt
migration (Pyle et al. 2018), and to filter out individuals for which it may be more conservative
to not make isotopic assignments to breeding season origin using isoscapes.
Many studies have used stable isotope analysis as a method to determine either breeding
location (from individuals caught on their wintering grounds; Hobson et al 2014) or wintering
location or habitat (from individuals caught on breeding grounds; Bearhop et al. 2004, Marra et
al. 1998). We took a novel approach, as we sampled birds traveling en route north to their
breeding grounds and used stable isotopes to determine both the wintering clusters (Reudink et
al. 2015) and approximate breeding locations. While stable isotope analysis has many benefits,
one limitation is the uncertainty of the hydrogen isoclines (Hobson et al. 2014). With more
complete sampling, especially on the wintering grounds, further studies will improve the
resolution of feather and claw isoscapes and allow a better picture of connectivity. In particular,

16

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

improvement of the isoscapes that have been developed for central and South America would
allow for better assignments of individuals to specific winter and breeding regions.
It has previously been established that spatial or habitat based segregation of birds from
differing populations, ages or sexes has the potential to explain population trajectories and
fluctuations on the breeding grounds (Marra et al. 2015; Rushing et al. 2016, Taylor and
Stutchbury 2016). In addition, recent theoretical (Taylor and Norris 2010) and experimental
(Betini et al. 2015) evidence suggests that the strength of migratory connectivity has the potential
to moderate the effect of environmental perturbations on populations. The apparent weak
migratory connectivity observed for Vaux’s swift and other aerial insectivores might buffer
aerial populations against climate driven shifts in habitat because more individuals will manage
to reach suitable habitat if it shifts when the population disperses widely (Finch et al. 2017).
Given the pervasive and persistent declines of aerial insectivores in North America (Nebel et al.
2010), this does not appear to be the case. Alternatively, Finch et al. (2017) suggest that weak
connectivity might simultaneously increase susceptibility to habitat loss. In this light, the
continued loss and degradation of habitat from deforestation, agricultural intensification,
urbanization, and land use change (Paquette et al. 2014, Curtis et al. 2018) should be the focus of
further research and conservation efforts for aerial insectivores. Clearly there is a need for
continued work to determine factors limiting their populations across the annual cycle and to
devise optimal strategies for their conservation (Martin et al. 2007).

ACKNOWLEDGEMENTS
We thank S. Joly for assistance in data collection. We would also like to thank the Smithsonian
Mass Spectrometry Facilities and Christine France for providing analytical support. We thank J.

17

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

Nocera, C. T. Collins and an anonymous reviewer for comments that greatly improved the
manuscript.
Funding statement: Funding was provided by a Natural Sciences and Engineering Research
Council of Canada (NSERC) Discovery Grant and a TRU Internal Research Fund Award to
M.W.R. and a National Science Foundation (NSF) grant to P.P.M.
Ethics statement: Birds were collected under a Canadian salvage permit (BC-SA-0049-12).
Author contributions: E.L.S., M.W.R, S.V.W. and A.E.M. formulated the study; S.V.W. and
E.L.S. analyzed the data; and all authors substantially contributed to writing the paper.

18

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

LITERATURE CITED
Åkesson, S., R. Klaassen, J. Holmgren, J. W. Fox, and A. Hedenström. (2012). Migration routes
and strategies in a highly aerial migrant, the Common Swift Apus apus, revealed by lightlevel geolocators. PLoS ONE 7: e41195.
Ambrosini, R., A., P. Moller, and N. Saino. (2009). A quantitative measure of migratory
connectivity. Journal of Theoretical Biology 257: 203–211.
Bearhop, S., G. M. Hilton, S. C. Votier, and S. Waldron. (2004). Stable isotope ratios indicate
that body condition in migrating passerines is influenced by winter habitat. Proceedings of
the Royal Society B 271: S215–S218.
Beason, J. P., C. Gunn, K. M. Potter, R. A. Sparks, and J. W. Fox. (2012). The Northern Black
Swift: migration path and wintering area revealed. The Wilson Journal of Ornithology 124:
1-8.
Betini, G. S., M. J. Fitzpatrick, and D. R. Norris. (2015). Experimental evidence for the effect of
habitat loss on the dynamics of migratory networks. Ecology Letters 18:526-534.
Bowen G. J., L. I. Wassenaar, and K. A. Hobson. (2005). Global application of stable hydrogen
and oxygen isotopes to wildlife forensics. Oecologia 143: 337-348.
Bracey, A., S. Lisovski, D. Moore, A. McKellar, E. Craig, S. Matteson, F. Strand, J. Costa, C.
Pekarik, P. Curtis, and G. Niemi. (2018). Migratory routes and wintering locations of
declining inland North American Common Terns. The Auk 135: 385-399.
Brown, D., and G. Miller. (2016). Band recoveries reveal alternative migration strategies in
American Robins. Animal Migration 3: 35-47.

19

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

Bull, E. L., and C. T. Collins. (1993). Nesting chronology, molt, and ectoparasites of Vaux's
Swifts in northeastern Oregon. Avocetta 17: 203–207.
Bull, E. L., and C. T. Collins. (2007). Vaux’s Swift (Chaetura vauxi). In Birds of North America
Online (A. Poole, Editor). Cornell Lab of Ornithology, Ithaca, NY, USA.
http://bna.birds.cornell.edu/bna/species/077
Bürkner, P.−C. (2017). brms: An R package for Bayesian multilevel models using Stan. Journal
of Statistical Software 80: 1–28. doi:10.18637/jss.v080.i01
Cerling, T. E., J. M. Harris, B. J. MacFadden, M. G. Leakey, J. Quade, V. Eisenmann, and J. R.
Ehleringer. (1997). Global vegetation change through the Miocene/Pliocene boundary.
Nature 389: 153–158.
Chabot, A. A., K. A. Hobson, S. L. Van Wilgenburg, G. E. Pérez, and S. C. Lougheed. (2018).
Migratory connectivity in the Loggerhead Shrike (Lanius ludovicianus). Ecology and
Evolution 8: 10662-10672. https://doi.org/10.1002/ece3.4415
Cooper, N. W., D. N. Ewert, K. R. Hall, S. M. Rockwell, D. Currie, J. M. Wunderle Jr, J. D.
White, and P. P. Marra. (2018). Resighting data reveal weak connectivity from wintering
to breeding grounds in a range-restricted and endangered long-distance migratory
passerine. Avian Conservation and Ecology 13(1): 9. https://doi.org/10.5751/ACE-01159130109
Cormier, R. L., D. L. Humple, T. Gardali, and N. E. Seavy. (2013). Light-level geolocators
reveal strong migratory connectivity and within-winter movements for a coastal California
Swainson's Thrush (Catharus ustulatus) population. The Auk 130: 283-290.

20

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

Cormier, R. L., D. L. Humple, T. Gardali, and N. E. Seavy. (2016). Migratory connectivity of
Golden-crowned Sparrows from two wintering regions in California. Animal Migration 3:
48-56.
Craine J. M., A. J. Elmore, M. P. M. Aidar, M. Bustamante, T. E. Dawson, E. A. Hobbie, A.
Kahmen, M. C. Mack, K. K. Mclauchlan, A. Michelsen, G. B. Nardoto, L. H. Pardo, J.
Peñuelas, P. B. Riech, E. A. G. Schuur, W. D Stock, P. H. Templer, and R. Virginia.
(2009). Global patterns of foliar nitrogen isotopes and their relationships with climate,
mycorrhizal fungi, foliar nutrient concentrations, and nitrogen availability. New
Phytologist 183: 980–992.
Curtis, P. G., C. M. Slay, H. L. Harris, A. Tyukavina, and M. C. Hansen. (2018). Classifying
drivers of global forest loss. Science 361: 1108-1111.
Environment and Climate Change Canada (2017). North American Breeding Bird Survey Canadian Trends Website, Data-version 2015. Environment and Climate Change Canada,
Gatineau, Quebec, K1A 0H3.
Finch, T., S. J. Butler, A. M. A Franco, and W. Cresswell. (2017). Low migratory connectivity is
common in long‐distance migrant birds. Journal of Animal Ecology 86: 662–673.
Finch, T., P. Saunders, J. M. Avilés, A. Bermejo, I. Catry, J. de la Puente, T. Emmenegger, I.
Mardega, P. Mayet, D. Parejo, and E. Račinskis. (2015). A pan-European, multipopulation
assessment of migratory connectivity in a near-threatened migrant bird. Diversity and
Distributions 21: 1051–1062.
Fraser, K. C., B. J. M. Stutchbury, C. Silverio, P. M. Kramer, J. Barrow, D. Newstead, N.
Mickle, B. F. Cousens, J. C. Lee, D. M. Morrison, T. Shaheen, P. Mammenga, K.

21

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

Appelgate, and J. Tautin. (2012). Continent-wide tracking to determine migratory
connectivity and tropical habitat associations of a declining aerial insectivore. Proceedings
of the Royal Society B 279: 4901–4906.
Fraser, K. C., A. Shave, A. Savage, A. Ritchie, K. Bell, J. Siegrist, J. D. Ray, K. Applegate, and
M. Pearman. (2017). Determining fine‐scale migratory connectivity and habitat selection
for a migratory songbird by using new GPS technology. Journal of Avian Biology 48: 339345.
Gelman, A., and D. B. Rubin. (1992). Inference from iterative simulation using multiple
sequences. Statistical science 7: 457−472.
Haché, S, K. A. Hobson, E. Bayne, S. L. Van Wilgenburg, and M.-A. Villard. (2014). Tracking
natal dispersal in a coastal population of a migratory songbird using feather stable isotope
(δ2H, δ34S) tracers. PLoS ONE 9: e94437. https://doi:10.1371/journal.pone.0094437
Hallworth, M.T., and P. P. Marra. (2015). Miniaturized GPS tags identify non-breeding
territories of a small breeding migratory songbird. Scientific Reports 5: 11069.
Handbook of the Birds of the World and BirdLife International. (2016). Handbook of the Birds
of the World and BirdLife International digital checklist of the birds of the world. Version
9.1. Available at: http://datazone.birdlife.org/species/taxonomy
Hijmans, R. (2015). Raster: Geographic data analysis and modeling. R package version 2.3-24.
http://CRAN.R-project.org/package=raster
Hobson, K. A. (2011). Isotopic ornithology: A perspective. Journal of Ornithology 152(1
SUPPL): S49-S66.

22

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

Hobson, K. A., and D. R. Norris. (2008). Animal migration: A context for using new techniques
and approaches. In Tracking Animal Migration with Stable Isotopes (K. A. Hobson and L.
I. Wassenaar, Editors) Academic Press, London, UK. pp. 1-19.
Hobson, K. A., H. Lormée, S. L. Van Wilgenburg, L. I. Wassenaar, and J. M. Boutin. (2009).
Stable isotopes (δD) delineate the origins and migratory connectivity of harvested animals:
the case of European woodpigeons. Journal of Applied Ecology 46: 572-581.
Hobson, K. A., S. L. Van Wilgenburg, L. I. Wassenaar, and K. Larson. (2012). Linking hydrogen
(δ2H) isotopes in feathers and precipitation: sources of variance and consequences for
assignment to isoscapes. PLoS ONE 7: e35137.
Hobson, K. A., S. L. Van Wilgenburg, J. Faaborg, J. D. Toms, C. Rengifo, A. Llanes Sosa, Y.
Aubry, and R. Brito Aguilar. (2014). Connecting breeding and wintering grounds of
Neotropical migrant songbirds using stable hydrogen isotopes: A call for an isotopic atlas
of migratory connectivity. Journal of Field Ornithology 85: 237–257.
Howell, S. N. G. (2010). Molt in North American Birds. Houghton Mifflin Harcourt Publishing
Company, New York, USA.
Knight,S. M., D. W. Bradley, R. G. Clark, E. A. Gow, M. Bélisle, L. L. Berzins, T. Blake, E. S.
Bridge, L. Burke, R. D. Dawson, P. O. Dunn, D. Garant, G. L. Holroyd, D. J. T. Hussell,
O. Lansdorp, A. J. Laughlin, M. L. Leonard, F. Pelletier, D. Shutler, L. Siefferman, C. M.
Taylor, H. E. Trefry, C. M. Vleck, D. Vleck, D. W. Winkler, L. A. Whittingham, and D. R.
Norris. (2018). Constructing and evaluating a continent‐wide migratory songbird network
across the annual cycle. Ecological Monographs 88: 445−460.
https://doi.org/10.1002/ecm.1298

23

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

Lajtha, K., and J. D. Marshall. (1994). Sources of variation in the stable isotopic composition of
plants. In Stable Isotopes in Ecology and Environmental Science (R. H. Michener and K.
Lajtha, Editors), Blackwell Scientific Publications, Oxford. UK. pp. 1–21.
Liechti, F., W. Witvliet, R. Weber, and E. Bächler. (2013). First evidence of a 200-day non-stop
flight in a bird. Nature Communications 4: 1–7.
Koleček, J., P. Procházka, N. El‐Arabany, M. Tarka, M. Ilieva, S. Hahn, M. Honza, J. de la
Puente, A. Bermejo, A. Gürsoy, and S. Bensch. (2016). Cross‐continental migratory
connectivity and spatiotemporal migratory patterns in the great reed warbler. Journal of
Avian Biology 47: 756-767.
Marra, P. P. 2000. The role of behavioral dominance in structuring patterns of habitat occupancy
in migrant bird during the nonbreeding season. Behavioural Ecology 11: 299-308.
Marra, P. P., K.A. Hobson, and R. T. Holmes. (1998). Linking winter and summer events in a
migratory bird using stable-carbon isotopes. Science 282: 1884:1886.
Marra, P. P., T. W. Sherry, and R. T. Holmes (1993). Territorial exclusion by a long-distance
migrant Warbler in Jamaica: a removal experiment with American Redstarts (Setophaga
ruticilla). The Auk 110: 565–573.
Marra, P. P., C. E. Studds, S. Wilson, T. S. Sillett, T. W. Sherry, and R. T. Holmes. (2015). Nonbreeding season habitat quality mediates the strength of density-dependence for a
migratory bird. Proceedings of the Royal Society B 282: 20150624.
Martin, T. G., I. Chadès, P. Arcese, P. P. Marra, H. P. Possingham, and D. R. Norris. (2007).
Optimal conservation of migratory species. PLoS ONE 2: e751.

24

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

Nebel, S., A. Mills, J. D. McCracken, and P. D. Taylor. (2010). Declines of aerial insectivores in
North America follow a geographic gradient. Avian Conservation and Ecology 5(2): 1.
http://dx.doi.org/10.5751/ACE-00391-050201
Nocera, J. J., J. M. Blais, D. V. Beresford, L. K. Finity, C. Grooms, L. E. Kimpe, K. Kyser, N.
Michelutti, M. W. Reudink, and J. P. Smol. (2012). Historical pesticide applications
coincided with an altered diet of aerially foraging insectivorous chimney swifts.
Proceedings of the Royal Society B 1740: 3114-3120.
Nowakowski, J. K., J. Szulc, and M. Remisiewicz. (2014). The further the flight, the longer the
wing: relationship between wing length and migratory distance in Old World reed and bush
Warblers (Acrocephalidae and Locustellidae). Ornis Fennica 91: 178-186.
Paquette, S. R., F. Pelletier, D. Garant, and M. Bélisle. (2014). Severe recent decrease of adult
body mass in a declining insectivorous bird population. Proceedings of the Royal Society
B 281: 20140649.
Post, D. M. (2002). Using stable isotopes to estimate trophic position: models, methods, and
assumptions. Ecology 83: 703–718.
Poupin, N., C. Bos, F. Mariotti, J. F. Huneau, D. Tomé, and H. Fouillet. (2011). The nature of
the dietary protein impacts the tissue-to-diet 15N discrimination factors in laboratory rats.
PLoS ONE 6: e28046.
Pyle, P. (1997). Identification guide to North American birds, Part 1, Columbidae to Ploceidae.
Slate Creek Press, California, USA.

25

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

Pyle, P., J. F. Saracco, and D. F. DeSante. (2018). Evidence of widespread movements from
breeding to molting grounds by North American landbirds. The Auk 135: 506-520.
Reudink, M. W., S. L. Van Wilgenburg, L. S. Steele, A. G. Pillar, P. P. Marra, and A. E.
McKellar. (2015). Patterns of migratory connectivity in Vaux’s Swifts at a northern
migratory roost: A multi-isotope approach. The Condor 117: 670–682.
R Core Team. (2017). R: A language and environment for statistical computing. R Foundation
for Statistical Computing, Vienna, Austria. https://www.R-project.org/.
Rohwer, S., A. D. Fox, T. Daniel, and J. F. Kelly. (2015). Chronologically sampled flight feather
permits recognition of individual molt-migrants due to varying protein sources. Peer J 3:
e743. https://doi.org/10.7717/peerj.743
Rushing, C. S., T. B. Ryder, and P. P. Marra. (2016). Quantifying drivers of population dynamics
for a migratory bird throughout the annual cycle. Proceedings of the Royal Society B 283:
20152846.
Sealy, J. C., N. J. van der Merwe, J. A. Lee-Thorp, and J. L. Lanham. (1987). Nitrogen isotope
ecology in Southern Africa: implications for environmental and dietary tracing.
Geochimica et Cosmochimica Acta 51: 2707-2717.
Stutchbury, B. G. M., S. A. Tarof, T. Done, E. Gow, P. M. Kramer, J. Tautin, J. W. Fox, and V.
Afanasyev. (2009). Tracking long-distance songbird migration by using geolocators.
Science 323: 896–896.
Still, C. J., and R. L. Powell. (2010). Continental-scale distributions of vegetation stable carbon
isotope ratios. In Isoscapes: Understanding Movements, Pattern and Process on Earth

26

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

through Isotope Mapping (J. B. West, G. J. Bowen, T. E. Dawson, and K. P. Tu, Editors).
Springer, New York, NY, USA. pp. 179–194.
Taylor, C. M., and D. R. Norris. (2010). Population dynamics in migratory networks. Theoretical
Ecology 3: 65-73.
Taylor, C. M., and B. J. M. Stutchbury. (2016). Effects of breeding versus winter habitat loss and
fragmentation on the population dynamics of a migratory songbird. Ecological
Applications 26: 424-437.
Tonra, C. M. and M. W. Reudink. (2018). Expanding the traditional definition of molt-migration.
The Auk 135: 1123-1132.
Trierweiler, C., R. H. G. Klaassen, R. H. Drent, K.-M. Exo, J. Komdeur, F. Bairlein, and B. J.
Koks. (2014). Migratory connectivity and population-specific migration routes in a longdistance migratory bird. Proceedings of the Royal Society B 281: 20132897.
Valdez-Juárez, S. O., A. Drake, K. J. Kardynal, K. A. Hobson, E. A. Krebs, and D. J. Green.
(2018). Use of natural and anthropogenic land cover by wintering Yellow Warblers: The
influence of sex and breeding origin. The Condor 120: 427-438.
Van Wilgenburg, S. L. and K. A. Hobson. (2011). Combining stable-isotope (δD) and band
recovery data to improve probabilistic assignment of migratory birds to origin. Ecological
Applications 21: 1340–1351.
Vander Zanden, H. B., M. B. Wunder, K. A. Hobson, S. L. Van Wilgenburg, L. I. Wassenaar, J.
M. Welker, and G. J. Bowen. (2014). Contrasting assignment of migratory organisms to

27

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

geographic origin using long term versus year-specific precipitation isotope maps. Methods
in Ecology and Evolution 5: 891−900.
van Wijk, R. E., M. Schaub, S. Hahn, N. Juárez-García-Pelayo, B. Schäfer, L. Viktora, M.
Martín-Vivaldi, M. Zischewski, and S. Bauer. (2018). Diverse migration strategies in
hoopoes (Upupa epops) lead to weak spatial but strong temporal connectivity. The Science
of Nature 105: 42−42.
Wassenaar, L. I. (2008). An introduction to light stable isotopes for use in terrestrial animal
migration studies. In Tracking animal migration with stable isotopes (K. A. Hobson and L.
I. Wassenaar, Editors). Elsevier, London, UK. pp 21–44.
Wassenaar, L. I., and K. A. Hobson. (2003). Comparative equilibration and online technique for
determination of non-exchangeable hydrogen of keratins for use in animal migration
studies. Isotopes in Environmental and Health Studies 39: 211–217.
Watanabe, S. (2010). Asymptotic equivalence of Bayes cross validation and Widely Applicable
information criterion in singular learning theory. Journal of Machine Learning Research,
11: 3571–3594.
Webster, M. S., and P. P. Marra. (2005). The Importance of understanding migratory
connectivity and seasonal interactions. In Birds of Two Worlds: The Ecology and
Evolution of Migration (R. Greenberg and P. P. Marra, Editors). Johns Hopkins University,
Baltimore, Maryland, USA. pp. 199–209.
Webster, M. S., P. P. Marra, S. M. Haig, S. Bensch, and R. T. Holmes. (2002). Links between
worlds: Unraveling migratory connectivity. Trends in Ecology and Evolution 17: 76-83.

28

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

Wunder, M. B. (2010). Using isoscapes to model probability surfaces for determining geographic
origins. In Isoscapes: understanding movement, pattern, and process on Earth through
isotope mapping (J. B. West, G.J. Bowen, T.E. Dawson and D.P. Tu, Editors). Springer,
New York, New York, USA. pp. 251-270.

29

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

TABLE 1. Mean, SD, Min and Max of the δ13Cf, δ15Nf and δ2Hf feather isotope analysis of
Vaux’s swifts from a northern migratory roost.

Isotope

Mean

δ2Hf
δ13Cf
δ15Nf

-107.6
-24.2
8.2

SD
23.0
1.0
1.0

Min
-176.1
-26.3
6.25

Max
-41.1
-19.4
10.56

30

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

TABLE 2. Correlation among feather isotopes δ13Cf, δ15Nf and δ2Hf of Vaux’s swifts from a
northern migratory roost.

Isotope

δ2Hf

δ13Cf
0.29

δ15Nf
0.06

δ2Hf

-

δ13Cf

P= 0.003

-

0.01

δ15Nf

P=0.56

P=0.94

-

31

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

TABLE 3. WAIC - based model selection for Bayesian robust regression models examining
variation in δ2Hf, δ15N, and δ13C respectively of Vaux’s swifts from a northern migratory roost.
Model
WAIC
SE
ΔWAIC
2
δ Hf
Null
610.88
17.41
0
Sex
610.94
16.85
0.06
Cluster (2)1
611.89
16.43
1.01
1
Cluster (3)
612.25
18.58
1.37
Age
612.45
17.37
1.57
Wing
612.59
16.83
1.71
Sex + Wing
613.02
16.19
2.14
Age + Sex + Wing
614.46
16.08
3.58
Age + Wing
614.60
16.69
3.72
Age + Sex + Wing + Cluster (2)
616.04
15.12
5.16
Age + Sex + Wing + Cluster (3)
618.61
15.94
7.73
δ15N
Null
Sex + Wing
Sex
Cluster (2)
Age
Wing
Cluster (3)
Age + Wing
Age + Sex + Wing
Age + Sex + Wing + Cluster (2)
Age + Sex + Wing + Cluster (3)

201.95
203.11
203.89
203.99
204.01
204.23
204.25
206.31
208.51
210.61
210.86

9.94
10.06
9.93
9.93
10.32
10.08
9.81
10.48
10.37
10.46
10.40

0
1.16
1.94
2.04
2.06
2.28
2.3
4.36
6.56
8.66
8.91

δ13C
Null
186.92
10.03
0
Sex
187.84
10.26
0.92
Wing
188.09
10.59
1.17
Age
188.62
9.70
1.70
Sex + Wing
189.02
10.64
2.10
Cluster (2)
189.1
10.17
2.18
Age + Wing
189.75
10.27
2.83
Cluster (3)
190.62
10.8
3.7
Age + Sex + Wing
190.81
10.19
3.89
Age + Sex + Wing + Cluster (2)
193.01
10.24
6.09
Age + Sex + Wing + Cluster (3)
194.63
10.59
7.71
1
Cluster (2) and Cluster (3) are 2 level and 3 level factors (respectively) representing the
alternative wintering ground cluster solutions from Reudink et al. (2015)

32

12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

Species
Ovenbird (Seiurus aurocapilla)
American Robin (Turdus migratorious)
Swainson's Thrush (Catharus ustulatus)
Golden-crowned Sparrow (Zonotrichia atricapilla)
Great Reed Warbler (Acrocephalus arundinaceus)
Tree swallow (Tachycineta bicolor)
Montagu's Harrier (Circus pygargus)
European Roller (Coracias garrulus)
European Robin (Erithacus rubecula)
American woodcock (Scolopax minor)
European Roller (Coracias garrulus) - eastern populations
Wood Thrush (Hylocichla mustelina)
European Roller (Coracias garrulus) - western populations
Purple Martin (Progne subis)
Loggerhead Shrike (Lanius ludovicianus)
Vaux’s swift
Common Tern (Sterna hirundo)
Barn Swallow (Hirundo rustica)
Hoopoe (Upupa epops)
Kirtland’s Warbler (Setophaga kirtlandii)

Paper
Hallworth and Marra 2015
Brown and Miller 2016
Cormier et al. 2013
Cormier et al. 2016
Koleček et al. 2016
Knight et al. 2018
Trierweiler et al. 2014
Finch et al. 2015
Ambrosini et al. 2016
Moore and Krementz 2017
Finch et al. 2015
Stanley et al. 2015
Finch et al. 2015
Fraser et al. 2017
Chabot et al. 2018
This study
Bracey et al. 2018
Ambrosini et al. 2009
van Wijk et al. 2018
Cooper et al. 2018

TABLE 4. Strength of migratory connectivity between breeding and non-breeding populations of migratory birds.
Mantel
Correlation
Coefficient (rM)
0.84
0.78
0.72
0.66
0.53 − 0.56
0.51 − 0.53
0.50
0.50
0.48
0.42
0.36
0.33
-0.30
-0.11 − 0.22
-0.12 − 0.22
-0.11 – 0.20
-0.03 − 0.19
0.025 – 0.22
0.06
-0.02 − 0.02

33

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

FIGURE LEGENDS

FIGURE 1. Predicted δ2H of feathers (δ2Hf) of Vaux’s swifts from a northern migratory roost a)
derived from 1000 bootstrap regressions relating δ2Hf to predicted δ2Hp from the isoscape of
Bowen et al. (2005) and b) spatially explicit estimates of prediction error (SD); see Methods.

FIGURE 2. Predicted origins of 35 Vaux’s swifts from a northern migratory roost that were
isotopically consistent with a common wintering environment (cluster 1 in Reudink et al. 2015)
based on spatially explicit assignment to a δ2Hf isoscape. Red dot depicts the sampling location
on Vancouver Island, Canada.

FIGURE 3. Predicted origins of 11 Vaux’s swifts from a northern migratory roost that were
isotopically consistent with a common wintering environment (cluster 2 in Reudink et al. 2015)
based on spatially explicit assignment to a δ2Hf isoscape. Red dot depicts the sampling location
on Vancouver Island, Canada.

FIGURE 4. Predicted origins of 26 Vaux’s swifts from a northern migratory roost that were
isotopically consistent with a common wintering environment (cluster 3 in Reudink et al. 2015)
based on spatially explicit assignment to a δ2Hf isoscape. Red dot depicts the sampling location
on Vancouver Island, Canada.

34

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

35

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

FIGURE 1

36

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

FIGURE 2

37

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

FIGURE 3

38

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

FIGURE 4

39

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65

SUPPLEMENTAL MATERIAL
Supplemental Material Table S1. Relationship between claw and feather isotopes (δ13C, δ15N
and δ2H), using only feathers identified as molted on the breeding ground (n = 71) of Vaux’s
swifts from a northern migratory roost.

Claw

δ2H
δ13C
δ15N

δ2H
r P
-0.08 0.44
-0.01 0.89
0.01 0.91

Feather
δ13C
r P
-0.03 0.80
-0.11 0.31
-0.16 0.11

δ15N
r P
0.00 0.40
0.09 1.00
-0.08 0.41

40