Received: 12 April 2021

| Revised: 15 July 2021 | Accepted: 5 August 2021

DOI: 10.1002/ece3.8047

ORIGINAL RESEARCH

Evolution of winter molting strategies in European and North
American migratory passerines
Claudie Pageau1
| Jared Sonnleitner1 | Christopher M. Tonra2
Matthew W. Reudink1
1
Department of Biological Sciences,
Thompson Rivers University, Kamloops, BC,
Canada
2

| Mateen Shaikh3 |

Abstract
Molt is critical for birds as it replaces damaged feathers and worn plumage, enhancing

School of Environment and Natural
Resources, The Ohio State University,
Columbus, Ohio, USA

flight performance, thermoregulation, and communication. In passerines, molt gen-

3

However, some species of migrant passerines that breed in the Nearctic and Western

Department of Mathematics & Statistics,
Thompson Rivers University, Kamloops, BC,
Canada
Correspondence
Matthew W. Reudink, 805 TRU Way,
Kamloops, BC V2C 0C8, Canada.
Email: mreudink@tru.ca
Funding information
This research was funded by a Natural
Sciences and Engineering Research Council
of Canada Discovery Grant (MWR), a British
Columbia Graduate Scholarship (CP), and a
Master's research scholarship from the Fond
de Recherche Nature et technologies (CP).

erally occurs on the breeding grounds during the postbreeding period once a year.
Palearctic regions have evolved different molting strategies that involve molting on
the overwintering grounds. Some species forego molt on the breeding grounds and
instead complete their prebasic molt on the overwintering grounds. Other species
molt some or all feathers a second time (prealternate molt) during the overwintering
period. Using phylogenetic analyses, we explored the potential drivers of the evolution of winter molts in Nearctic and Western Palearctic breeding passerines. Our
results indicate an association between longer photoperiods and the presence of
prebasic and prealternate molts on the overwintering grounds for both Nearctic and
Western Palearctic species. We also found a relationship between prealternate molt
and generalist and water habitats for Western Palearctic species. Finally, the complete prealternate molt in Western Palearctic passerines was linked to longer days
on the overwintering grounds and longer migration distance. Longer days may favor
the evolution of winter prebasic molt by increasing the time window when birds can
absorb essential nutrients for molt. Alternatively, for birds undertaking a prealternate
molt at the end of the overwintering period, longer days may increase exposure to
feather-­degrading ultra-­violet radiation, necessitating the replacement of feathers.
Our study underlines the importance of the overwintering grounds in the critical process of molt for many passerines that breed in the Nearctic and Western Palearctic
regions.
KEYWORDS

molt, Nearctic, Passeriformes, phylogenetic analysis, Western Palearctic

This is an open access article under the terms of the Creat​ive Commo​ns Attri​bution License, which permits use, distribution and reproduction in any medium,
provided the original work is properly cited.
© 2021 The Authors. Ecology and Evolution published by John Wiley & Sons Ltd.
Ecology and Evolution. 2021;11:13247–13258.	

www.ecolevol.org

| 13247

13248

|

1 | I NTRO D U C TI O N

PAGEAU et al.

distance as potential drivers of the evolution of molt-­migration.
Phylogenetic analyses are important for the study of evolutionary

Molt, the systematic replacement of old feathers by new ones, is

processes to correct for the nonindependence among species (Ives

critical for birds; fresh, high-­quality plumage enhances flight perfor-

& Zhu, 2006), but they have not been used to look at the evolu-

mance, thermoregulation, visual communication, and attractiveness

tion of winter molt strategies in passerines. In a recent study of

(Gill, 2006). Hence, the timing, location, and number of molts are

the Nearctic–­Neotropical Family Parulidae, Terrill et al. (2020) con-

critical for birds to grow feathers of sufficient quality and maintain

cluded that structural needs driven by feather damage during the

relatively fresh plumage throughout the annual cycle. Most migra-

annual cycle drive the evolution of prealternate molts. However, the

tory passerines (Order Passeriformes) that breed in the Nearctic

extent to which this is the case across the diverse molt strategies of

and Western Palearctic regions undergo only one complete molt

passerines as a whole remains unknown.

(i.e., all feathers are replaced) during the annual cycle (Humphrey &

Although phylogenetic analyses have not been conducted yet to

Parkes, 1959; Jenni & Winkler, 2020; Pyle, 1997). This molt generally

study winter molts in all Passeriformes, potential drivers of the evolu-

occurs after breeding on the breeding grounds and is termed the

tion of the prebasic molt on the overwintering grounds have been hy-

prebasic molt for adult birds (Humphrey & Parkes, 1959). However,

pothesized (but see Svensson & Hedenström, 1999; Terrill et al., 2020

when and where the prebasic molt occurs can be quite variable. For

for studies on particular families). Barta et al. (2008) created models

example, some species or populations employ a strategy in which

that linked winter molt in migratory birds with food seasonality; lack

there is a temporal overlap between molt and migration (molt-­

of resources at the end of the summer combined with an abundance of

migration; Tonra & Reudink, 2018). This is especially common in

resources on the overwintering grounds during winter could have led

western North America where many passerine species depart their

to the evolution of winter molt (see also Remisiewicz et al., 2019). This

arid breeding grounds to molt in the highly productive Mexican mon-

would particularly be the case for Western Palearctic species migrat-

soon region prior to completing their southward migration (Pageau

ing to sub-­Saharan overwintering grounds that are productive during

et al., 2020; Pyle et al., 2009; Rohwer et al., 2005). Other species

the northern fall/beginning of winter because of the rainy season at

migrate prior to molt and undergo the prebasic molt on the over-

these locations (Jenni & Winkler, 2020; Kiat et al., 2019). Molting on

wintering grounds (Barta et al., 2008; De la Hera et al., 2012; Kiat

the overwintering grounds after migration would also occur most often

et al., 2019). Molt on the breeding grounds has been reported to be

in longer-­distance migrants; longer migration distance would impose a

the ancestral state of the prebasic molt, and other strategies evolved

time constraint between breeding and migration and, thus, could favor

later (Svensson & Hedenström, 1999), likely in response to environ-

molt somewhere other than breeding grounds (Kjellén, 1994; Leu &

mental and life-­history trade-­offs (Pageau, Tonra, et al., 2020).

Thompson, 2002). De la Hera et al. (2012) recorded molt duration of

In addition to prebasic molt, some species undergo a second molt,

98 Nearctic passerines and found that migrant species molting on the

termed prealternate and resulting in alternate plumage (Humphrey

overwintering grounds have a molt duration as long as resident species

& Parkes, 1959), at some point during the annual cycle, usually in

molting on breeding grounds. However, migrant species molting on

late winter/early spring prior to breeding, although prealternate

the breeding grounds have a shorter molt duration. Previous studies

molt-­migration has also been documented (e.g., Wright et al., 2018).

have indicated that a longer molt duration results in feathers of higher

Generally, this molt is partial (i.e., not all feathers are replaced) and

quality (e.g., Dawson et al., 2000; Griggio et al., 2009; Serra, 2001);

often only involves body feathers (Jenni & Winkler, 2020). However,

however, de la Hera et al. (2012) observed the opposite. Thus, it could

some migratory passerines undergo a complete prealternate molt,

be more advantageous for long-­distance migrants to molt on the

replacing all their feathers on the overwintering grounds prior

overwintering grounds where there are fewer time constraints, which

to fall migration (e.g., Dolichonyx oryzivorus Renfrew et al., 2011,

would allow them to grow high-­quality feathers.

Phylloscopus trochilus Underhill et al., 1992). The presence of a

The presence of a second molt during the annual cycle may be

complete prealternate molt during the annual cycle is rare among

favored due to rapid degradation of the feathers from UV exposure

Nearctic migrant passerines, but more common among Western

(Bergman, 1982; Jenni & Winkler, 2020, see studies by Svensson &

Palearctic passerines (Jenni & Winkler, 2020; Renfrew et al., 2011).

Hedenström, 1999; Jiguet et al., 2019). Thus, species living in open

Generally, species undergoing a complete prealternate molt in the

habitats would be more affected by UV degradation and more likely

Palearctic have an incomplete prebasic molt. Rarely will passerines,

to undergo a prealternate molt (Guallar et al., 2021; Pyle, 1998;

both in the Nearctic and in the Palearctic, have two complete molts

Pyle & Kayhart, 2010). This would particularly be the case for some

during the annual cycle (Renfrew et al., 2011). This extensive varia-

Western Palearctic migrants who winter in more sun-­exposed en-

tion in molt strategies among migratory passerines begs the ques-

vironments such as savannahs (Jones, 1995; Rohwer et al., 2005).

tion: Which life-­history characteristics and/or environmental factors

For Nearctic migrant passerines, Froehlich et al. (2005) observed

have driven the evolution of these different molt strategies?

a link between a plant-­based diet and the presence of a prealter-

Pageau, Tonra, et al. (2020) studied the evolution of molt-­

nate molt. Complete prealternate molt is more common in Western

migration in North America using phylogenetic analysis to correct

Palearctic migrant passerines than Nearctic migrant passerines; only

for the nonindependence among species. The results indicated the

Dolichonyx oryzivorus (Renfrew et al., 2011) undergoes a complete

importance of the aridity of the breeding grounds and migration

prealternate molt among Nearctic migrant passerines which results

| 13249

PAGEAU et al.

in two complete molts during the annual cycle. Rohwer et al. (2005)
proposed that a complete prealternate molt is rare among Nearctic
migrant passerines because most species winter in moister, denser,
and heavily shaded habitats with softer foliage which would result

TA B L E 1 Number of species for Nearctic (total of 183 species)
and Western Palearctic (total of 116 species) passerines performing
different molting strategies: winter prebasic, prealternate, and
complete prealternate molt
Nearctic

in relatively less damage to feathers compared with open habitats.
Here, we explore potential drivers behind the evolution of win-

Palearctic

Winter prebasic

13/183 (7.1%)

9/116 (7.8%)

ter molts, specifically the prebasic and prealternate winter molts in

Prealternate

87/183 (47.5%)

56/116 (48.3%)

Nearctic and Western Palearctic migratory passerines. We tested

Complete prealternate

2/183 (1.1%)

11/116 (9.5%)

whether migration distance, aridity of the breeding and overwintering grounds, average photoperiod (i.e., average day length over
a certain period at a particular location), diet, and habitat of the

2.2 | Classification molt strategies

overwintering grounds were associated with the presence of winter molts. Based on previous studies, we predicted that the pre-

We followed Humphrey and Parkes (1959) system to classify two

basic molt on the overwintering grounds will be associated with

types of molt: prebasic molt, which results in the basic plumage and

longer-­distance migrants for both Nearctic and Western Palearctic

generally occurs after breeding, and prealternate molt, which results

(Kjellén, 1994; Leu & Thompson, 2002). We also predicted that

in the alternate plumage and generally occurs before spring migra-

productive overwintering grounds would be an additional driver

tion. For the prebasic molt, we classified if the species molted on

of the winter prebasic molt for Western Palearctic species (Barta

their overwintering grounds or not as a binary variable (overwinter-

et al., 2008; Remisiewicz et al., 2019). For the prealternate molt, we

ing grounds = 1, not = 0; Table 1). Species undergoing molt-­migration

predicted that overwintering in open habitats and locations where

or conducting a suspended prebasic molt were not considered

the average photoperiod is longer are important factors associated

as molting on the overwintering grounds (classified as 0). For pre-

with a second molt for both Nearctic and Western Palearctic mi-

alternate molt, we recorded whether the molt was complete, par-

grant passerines because these species would be more exposed

tial, or absent. Thus, we created two binary response variables for

to UV radiation resulting in feather degradation (Bergman, 1982;

prealternate molt: (A) presence or absence of prealternate molt and

Jenni & Winkler, 2020; Jiguet et al., 2019; Pyle, 1998; Svensson &

(B) complete prealternate molt or absence of a complete prealter-

Hedenström, 1999; Terrill et al., 2020). We also predicted that her-

nate molt (Table 1). To categorize these molting strategies, we ex-

bivorous Nearctic species would be associated with the presence of

tracted molt information from species description in The Birds of the

a prealternate molt (Froehlich et al., 2005).

World (2020) for Nearctic passerines and the Handbook of Western
Palearctic Birds (Shirihai & Svensson, 2018) for Western Palearctic

2 | M ATE R I A L S A N D M E TH O DS
2.1 | Species selection

passerines. When the molting accounts were unclear, vague, or lacking, we searched other sources to find or verify the information and
used the most recent information (The Identification Guide to North
American Birds, Pyle, 1997; and peer-­reviewed journal articles;
Voelker & Rohwer, 1998; Butler et al., 2002; Rohwer et al., 2005;

We collected data for all species of Nearctic and Western Palearctic

Butler et al., 2006; Pyle et al., 2009; Jahn et al., 2013). When varia-

migratory passerines according to Birds of North America (now The

tion in molting strategy among individuals of the same species was

Birds of the World, 2020) and Handbook of Western Palearctic Birds

encountered, we used the predominant strategy to classify the spe-

(Shirihai & Svensson, 2018) lists. For the Nearctic species, we only

cies. For the case of Tyrannus vociferans and Delichon urbicum, we

kept species with a breeding distribution located in Canada or USA

classified the species as having a prebasic molt on the overwintering

(excluding Hawaii) and overwintering grounds in the Americas. For

grounds even though the species have been reported to molt both

the Western Palearctic, we kept species with a breeding distribu-

on the breeding and overwintering grounds equally. We recorded

tion located in Europe, northern Africa, or western Asia (we used

the reference used for each species in the data file.

61°46′35.2″ as an approximate eastern boundary and the Sahara
Desert as southern boundary) and with overwintering grounds in
Africa, Europe, or western Asia. We excluded species that were not

2.3 | Data collection of predictor variables

considered full migrants by IUCN (2020) in addition to Pinicola enucleator, Galerida cristata, Corvus corone, and Cinclus cinclus as they

To classify the amount of primary productivity on the breeding

were not considered full migrant by Handbook of the Birds of the

and overwintering grounds, we calculated the normalized differ-

World (now The Birds of the World, 2020). We deleted any species

ence vegetation index (NDVI) using distribution maps of the breed-

from our dataset that lacks important data for our analyses. Our final

ing and overwintering grounds from BirdLife International Data

data consisted of 183 species (including 6 subspecies) of Nearctic

Zone after filling out a request (2018). NDVI is a measure of live

migratory passerines and for 116 species (including 2 subspecies) of

green vegetation and was used to indicate the aridity of the breed-

Western Palearctic migratory passerines.

ing grounds during the postbreeding period (1 July to 31 August)

13250

|

PAGEAU et al.

and the overwintering grounds during the nonbreeding period (15

we then created the maximum clade credibility trees for Nearctic

September to 15 April). Data were available from the Application

and Palearctic with our 1,000 trees using 1% burn-­in (as states) and

for Extracting and Exploring Analysis Ready Samples (AppEEARS

mean heights for node heights. We added the subspecies (6 Nearctic,

Team, 2019). We extracted the NDVI values from 2000 to 2019

2 Palearctic) in R (R Core Team, 2020) using the package phytools

using the “Area Sample” function, chose the product MOD13A3.006

(Revell, 2012) to obtain a maximum clade credibility tree of 183

(Didan, 2015), selected “Native projection,” and calculated the mean

species and subspecies for the Nearctic and 116 for the Western

value for each species. One hundred ninety-­eight distribution maps

Palearctic. We used trees including the species and subspecies for all

were not available from BirdLife International (2018) or unusable in

our analyses. The visual representations of the phylogenies (Figures 1

AppEEARS. Instead, we created distribution maps as polygons using

and 2) were created using the phytools package (Revell, 2012).

template maps from IUCN (2020). Methods for NDVI data collection
followed those employed by Pageau, Tonra, et al. (2020).
Migration distance was approximated as the distance (megam-

2.5 | Statistical analysis

eter; Mm) from the centroid of the breeding distribution to the
centroid of the wintering distribution using the package Geosphere

For each response variable (prebasic winter, prealternate, and com-

(Hijmans, 2019) in R (R Core Team, 2020). Centroid values were cal-

plete prealternate molt), we separated the analysis between Western

culated in decimal degrees using polygon maps provided by BirdLife

Palearctic and Nearctic passerines as we hypothesized that both

International (2018) in ArcMap. One hundred seventy maps were

systems are influenced differently by the environmental conditions.

unavailable from BirdLife International, and the centroid values were

Note that we could not analyze the presence of a complete prealter-

calculated by creating polygons in Google Earth pro using template

nate molt in Nearctic passerines because only one species in North

maps from IUCN (2020) and processing the polygons in Earth Point

America exhibits a complete prealternate molt. Prior to our analyses,

(2020).

we tested for collinearity; migration distance, and average photoper-

Average photoperiods at the overwintering grounds were mea-

iod were highly correlated (r > 0.5) and were never included in the

sured using each species centroid on the National Research Council

same model. Next, we used phylogenetic logistic linear models to test

of Canada (2020) sunset/sunrise calculator. Photoperiods on the

the predictors and created all the possible models (total of 23 models

overwintering grounds were retrieved from 15 September to 15

for prebasic molt and 47 for prealternate). We selected the best mod-

April 2018/2019 and averaged for each species.

els using Akaike's information criterion (AIC) and determined that

We retrieved the list of preferred breeding and nonbreeding

models were similar if they differed by <4 ΔAIC. The Akaike weights

habitats for each species from BirdLife Data Zone online database

were obtained using the qpcR package (Spiess, 2018) and the R 2

(November 2018). We used data mining on the webpage http://dataz​

using the function R2.lik from the rr2 package (Ives & Daijiang, 2018).

one.birdl​ife.org/speci​es/facts​heet/common_name-­scien​tific_name/

Finally, we examined the 95% confidence intervals of the parameter

details for this task, and spaces were replaced by the character “-­”

estimates of every predictor included in the top models to assess

on common and scientific names. All entries were checked for no-

which variables were informative. Analyses were conducted in R (R

menclature inconsistencies. For our analysis, we categorized and

Core Team, 2020; version 1.2.5001) using the package phyloglm (Ho

simplified the type of habitat for the overwintering grounds by pri-

& Ane, 2014). The “logistic_MPLE” method was applied with a btol of

oritizing major habitat and then suitable. We also categorized the

35, a log.alpha.bound of 10, and 100 bootstraps.

habitat into four major categories: dense (forest, shrubland), open
(grassland, savanna, open woodland, rocky areas), water (wetland,
marine), and generalist. We classified species as generalist when two
or more major habitats were encountered. Habitat classification followed methods employed by Pageau, Vale, et al. (2020).

3 | R E S U LT S
3.1 | Prebasic winter molt

Diet data were retrieved from Willman et al. (2014), and we classified
the species into three categories: herbivore, omnivore, and invertivore.

For the Western Palearctic species, the top model predicting the

We could not include diet in the winter prebasic models as every winter

presence of a prebasic winter molt included NDVI of the breeding

molt species except one had a diet that mostly consists of invertebrates.

grounds and average photoperiod on the overwintering grounds
(Table 2). For the Nearctic species, the same two variables were also

2.4 | Phylogeny

present in the top models, but with the addition of migration distance and NDVI of the overwintering grounds (Table 2). For both
Western Palearctic and Nearctic passerines, only photoperiod had

Using BirdTree.org (Jetz et al., 2012), we downloaded 1,000 possi-

a 95% confidence interval that did not overlap zero (Table 3). The

ble trees from “Ericson All Species: a set of 10,000 trees with 9,993

parameter estimates were both positive (Western Palearctic: 2.19,

OTUs each” (Ericson et al., 2006) for a phylogeny subset of the 177

Nearctic: 1.48) which indicate that longer photoperiods were asso-

Nearctic species and a subset of the 114 Western Palearctic spe-

ciated with a prebasic molt on the overwintering grounds in both

cies. Using TreeAnnotator V.1.10.4 (Rambaut & Drummond, 2018),

Palearctic and Nearctic passerines (Figures 1–­3).

| 13251

PAGEAU et al.

F I G U R E 1 Phylogeny of the 116 species and subspecies of Western Palearctic passerines. The color of the branches represents the
average photoperiod where red is longer photoperiod. The red dots indicate the species doing a prealternate molt while the black dots
indicate a prebasic molt on the overwintering grounds. We labeled the passerine families with more than 5 members. See supporting
information for phylogeny with tips labeled with species name

3.2 | Presence of a prealternate molt

association with prealternate molt; longer photoperiod, species
wintering in water habitat or being generalist in their selection of

The top models explaining the presence of a prealternate molt in

habitat were associated with the presence of a prealternate molt

Western Palearctic passerines included the following variables:

(Figures 1 and 3). For the Nearctic passerines, the top models con-

NDVI of the breeding and overwintering grounds, photoperiod,

tained all the variables (NDVI breeding and overwintering grounds,

habitat of the overwintering grounds, and diet (Table 2). Of these

migration distance, photoperiod, habitat of the overwintering

five variables, the confidence intervals of average photoperiod,

grounds, and diet), but only the confidence intervals of photoper-

generalist habitat, and water habitat did not overlap zero (Table 3).

iod did not overlap zero (Tables 2 and 3). Longer photoperiod (0.50)

The parameter estimates of average photoperiod (1.29), general-

was associated with the presence of a prealternate molt in Nearctic

ist habitat (1.37), and water habitat (1.97) all indicated a positive

passerines (Figures 2 and 3).

13252

|

PAGEAU et al.

F I G U R E 2 Phylogeny of the 183 species and subspecies of Nearctic passerines. The color of the branches represents the average
photoperiod where red is longer photoperiod. The red dots indicate the species doing a prealternate molt while the black dots indicate a
prebasic molt on the overwintering grounds. We labeled the passerine families with more than 5 members. See supporting information for
phylogeny with tips labeled with species name

3.3 | Presence of a complete prealternate molt

are associated with the presence of a complete prealternate molt
(Figure 3).

Western Palearctic passerines that underwent a complete prealternate molt had top models that included the following variables: NDVI
of the breeding and overwintering grounds, migration distance, pho-

4 | D I S CU S S I O N

toperiod, and habitat of the overwintering grounds (Table 2). Only
migration distance and photoperiod had a 95% confidence interval

The aim of this study was to explore the potential drivers of the

that did not overlap zero (Table 3). The parameter estimate of mi-

evolution of winter molts in Nearctic and Western Palearctic pas-

gration distance and photoperiod was both positive (0.21 and 3.12),

serines. We examined the prebasic molt on the overwintering

suggesting that longer migration distance and longer day length

grounds, the presence of a prealternate molt (second molt) on the

| 13253

PAGEAU et al.

TA B L E 2 Top-­ranked models (<4 AIC units from top model) explaining a prebasic molt on the overwintering grounds and the presence of
a second molt (prealternate), which can be completed, in Western Palearctic and Nearctic passerines
Molt

Region

Top models

AIC

Prebasic winter

Western
Palearctic

NDVI breed + photoperiod

52.65

Nearctic

Prealternate

Western
Palearctic

Nearctic

Complete
prealternate

Western
Palearctic

ΔAIC
0

R2

w
1

0.35

Photoperiod

84.12

0

0.19

0.26

NDVI breed

85.24

1.12

0.11

0.24

NDVI w

85.25

1.13

0.11

0.24

Migration distance

85.35

1.23

0.10

0.24

NDVI w + photoperiod

86.11

1.99

0.07

0.26

NDVI w + migration dist.

87.43

3.31

0.04

0.24

NDVI breed + migration dist.

87.54

3.42

0.03

0.24

NDVI breed + NDVI w

87.55

3.43

0.03

0.24

NDVI breed + NDVI w + photoperiod

88.04

3.92

0.03

0.26

Photoperiod + habitat w

137.84

0

0.34

0.35

NDVI w + Photoperiod + habitat w

139.51

1.67

0.15

0.35

NDVI breed + Photoperiod + habitat w

139.67

1.83

0.14

0.35

Photoperiod

139.79

1.95

0.13

0.28

NDVI w + photoperiod

139.91

2.07

0.12

0.29

NDVI w + photoperiod + diet

141.25

3.41

0.06

0.30

Photoperiod + diet

141.78

3.94

0.05

0.29

Photoperiod

221.8

0

0.16

0.25

Migration distance

221.8

0

0.16

0.25

NDVI w

222

0.2

0.14

0.24

NDVI breed + migration dist.

222.1

0.3

0.14

0.26

NDVI breed + NDVI w + migration

223.2

1.4

0.08

0.26

NDVI breed

223.5

1.7

0.07

0.24

NDVI w + migration dist.

224

2.2

0.05

0.24

NDVI w + diet

224.3

2.5

0.04

0.26

NDVI w + photoperiod

224.4

2.6

0.03

0.24

Migration dist. + habitat w

224.9

3.1

0.03

0.26

Diet

225.1

3.3

0.03

0.24

NDVI breed + NDVI w

225.2

3.4

0.03

0.24

Habitat w

225.4

3.6

0.03

0.25

NDVI breed + NDVI w + diet

225.6

3.8

0.02

0.26

NDVI breed + migration dist. + diet

225.8

4

0.02

0.26

NDVI breed + photoperiod

45.39

0

0.31

0.52

NDVI breed + migration dist.

46.49

1.1

0.18

0.51

Migration distance

47.25

1.86

0.12

0.46

NDVI breed + NDVI w + photoperiod

47.47

2.08

0.11

0.52

Photoperiod

47.64

2.25

0.10

0.46

NDVI breed + photoperiod + habitat w

48.48

3.09

0.07

0.57

NDVI w + migration dist.

48.62

3.23

0.06

0.48

NDVI w + photoperiod

48.70

3.31

0.06

0.47

Note: NDVI breed = NDVI breeding grounds, NDVI w = NDVI overwintering grounds, migration distance = migration distance between the centroid
of the breeding and overwintering grounds, photoperiod = average photoperiod at the centroid of the wintering ground between 15 September and
15 April, habitat w = main habitat used on the overwintering grounds, diet = principal diet of the species (omnivore, invertivore, herbivore).

13254

|

PAGEAU et al.

TA B L E 3 Model-­averaged parameter estimates and 95% confidence intervals for variables included in the top-­ranked models (<4 AICc
units of best model) explaining a prebasic molt on the overwintering grounds and the presence of a second molt (prealternate), which can be
completed, in Western Palearctic and Nearctic passerines
Prebasic winter

Prealternate
completed

Prealternate

Western Palearctic

Nearctic

Western Palearctic

Nearctic

Western
Palearctic

−3.39 (−6.61, 1.10)

0.001 (−2.21, 2.88)

−0.94 (−2.74, 0.028)

−0.32 (−1.66, 1.18)

4.63 (−0.002, 10.1)

NDVI w

0.44 (−2.86, 3.35)

0.80 (−1.40, 3.30)

Migration distance

0.021 (−0.30, 0.26)

NDVI breed

1.48 (1.44, 1.64)

4.98 (−0.65, 9.30)
0.21, (0.10, 0.90)
3.12 (3.02, 3.24)

1.29 (1.21, 1.36)

0.50, (0.44, 0.59)

Diet—­Invertivore

−0.059 (−1.00, 0.74)

0.21 (−0.61, 0.98)

Diet—­Omnivore

−0.28 (−1.21, 0.58)

0.051 (−0.89, 0.93)

Habitat Generalist

1.37 (0.64, 2.60)

0.0066 (−0.55, 0.64)

0.22 (−0.48, 1.45)

Habitat—­Open

0.51 (−0.62, 1.71)

−0.15 (−1.08, 0.61)

0.083 (−1.05, 0.96)

Habitat—­Water

1.97 (0.28, 3.74)

0.75 (−0.30, 1.88)

0.13 (−13.55, 1.16)

Photoperiod

2.19 (2.11, 2.30)

1.83 (−0.14, 4.26)
0.14 (−0.070, 0.33)

Note: Values in bold indicate that the 95% CI did not overlap zero. See Table 2 for variable's definition.

overwintering grounds, and the presence of a complete prealternate

grounds. We could not add diet in our models for the prebasic molt

molt. We found that a prebasic molt on the overwintering grounds

as most bird (78%) undertaking a winter prebasic molt has a diet that

is associated with longer average photoperiods on the overwinter-

mostly consists of invertebrates (e.g., Empidonax spp., Hippolais spp.).

ing grounds for Nearctic as well as Western Palearctic passerines.

Despite being unable to include this factor in our analysis, we argue

Our results also indicate, for Western Palearctic passerines, that

that this relationship likely indicates a link between diet and win-

longer photoperiods on the overwintering grounds and species liv-

ter prebasic molt. A link between diet and molt strategy has already

ing in water (wetlands or marine) habitats or being generalist in their

been observed by Froehlich et al. (2005), but for the prealternate

habitat choice exhibit an association with the presence of prealter-

molt, long-­distance migrants with an extensive plant diet are more

nate molt. For Nearctic passerines, only longer photoperiods were

likely to undergo a prealternate molt for Nearctic passerines.

associated with the prealternate molt. Finally, the complete prealter-

The evolution of the prealternate molt in Nearctic and Western

nate molt in Western Palearctic passerines is associated with longer

Palearctic migrants is proposed to result from the amount of feather

migration distance and longer photoperiod on the overwintering

degradation caused by the habitat where they overwinter (Pyle &

grounds. Collectively, these results indicate the powerful selective

Kayhart, 2010; Rohwer et al., 2005; Terrill et al., 2020). In over-

force of overwintering conditions in the evolution of molt strategies.

wintering grounds with longer days throughout the winter, both

The prebasic molt on the overwintering grounds for both

Nearctic and Western Palearctic passerines may have evolved

Nearctic and Western Palearctic migrants is influenced by longer

prealternate molts to cope with increased feather wear from UV

photoperiod on the overwintering grounds. This result was unex-

light exposure (Barta et al. 2008; Bergman, 1982; Pyle, 1998). For

pected because we predicted that longer migration distance (Kjellén,

Western Palearctic migrant passerines, species who inhabited water

1994; Leu & Thompson, 2002) and primary productivity (Barta

habitats, or were generalists, were more likely to undergo a prealter-

et al., 2008; Remisiewicz et al., 2019) would be associated with win-

nate molt. This result was counter to our hypothesis that species

ter prebasic molt. However, photoperiod and migration distance

living in open habitats were more likely to undergo a prealternate

were correlated (see Methods), making these two factors difficult

molt because the harsher conditions and exposure to UV light would

to disentangle. Additionally, we were expecting longer photoperiods

increase feather wear (Rohwer et al., 2005). Feather replacement is

to be a potential driver of the prealternate molt and not the preba-

energetically costly (Murphy & King, 1992; Lindström et al., 1993)

sic molt because we hypothesized that longer photoperiods would

and requires sufficient resources, particularly proteins (Froehlich

expose the feathers to UV radiation and, as a consequence, faster

et al., 2005). Thus, species living in habitats with no or little food

feather degradation which would be counterproductive to a preba-

limitation might be able to afford the replacement of some or all

sic molt on the overwintering grounds. However, longer photoperi-

feathers before spring migration. Extended UV light exposure from

ods could have some benefits by reducing the costs associated with

longer photoperiods coupled with resources availability could have

molt by increasing the time window when birds can absorb essential

led to Western Palearctic passerines evolving a second molt during

nutrients for molt (Murphy & King, 1991; Renfrew et al., 2011). Thus,

their annual cycle. Habitat was not a significant variable for Nearctic

some species might have evolved a winter prebasic molt strategy

passerines; these passerines generally overwinter in tropical hab-

to take advantage of the longer photoperiods on the overwinter-

itats with softer foliage which does not damage feathers (Rohwer

ing grounds compared with shorter daylights in fall on the summer

et al., 2005). In summary, the prealternate molt, which is often

| 13255

PAGEAU et al.

(a)

(b)

(c)

(d)

(e)

(f)

(g)

F I G U R E 3 Boxplots and bar graph of the significant predictors (parameter estimates did not overlap 0) associated winter prebasic (a, b),
prealternate (c, d, e), and complete prealternate molts (f, g). Plots to the left (a, c) correspond to the results for the Nearctic passerines, and
plots to the right (b, d, e, g, f) are the results for the Western Palearctic species
incomplete, seems to be associated with species affected by strong

The presence of a complete prealternate molt in Western

feather wear and need to replace those specific feathers to maximize

Palearctic passerines appears to be associated with species with

fitness. This result supports the findings of Terrill et al. (2020) which

longer migration distance and longer photoperiod on the overwin-

identified feather wear as a driver of the evolution of prealternate

tering grounds. This result fits with our predictions that a complete

molt in Parulidae, but not of Guallar and Figuerola (2016) that identi-

prealternate molt would have evolved in species wintering in harsher

fied migration distance as a predominant factor and sexual selection

and sun-­exposed environments (longer photoperiod and open hab-

as a limited factor in Motacillidae.

itat types) which damage the feathers faster, hence the need to re-

Interestingly, it appears that in most cases where winter prebasic

place them in a second molt (Jenni & Winkler, 2020; Jones, 1995;

molt is present (13 species in the Nearctic and 9 species in the Western

Rohwer et al., 2005). Even though we did not find an association

Palearctic), a prealternate molt is uncommon (36% in Nearctic, 33% in

with habitat type, longer photoperiod on the overwintering grounds

Palearctic). Having a later prebasic molt on the overwintering grounds

could cause the need to molt before spring migration due to rapid

could result in fresher feathers prior to the spring migration and

degradation of the feathers from UV exposure (Bergman, 1982;

breeding season, making it unnecessary to molt feathers a second

Jenni & Winkler, 2020, see studies by Svensson & Hedenström,

time in the spring. The effect of a prebasic molt on the presence or

1999; Jiguet et al., 2019). However, it is difficult to disentangle the

absence of a prealternate molt should be explored further.

effects of migration distance and photoperiod as these variables are

13256

|

PAGEAU et al.

highly correlated. Still, having to migrate a long distance could re-

One challenge with this study was the classification of molt strat-

quire a fresh molt prior to it to ensure feathers of high quality for the

egies. Many species (e.g., Tyrannus vociferans) have interindividual

migration or longer migration distances could cause more feather

variation or plasticity in molt. In addition, there is variation in which

wear and create the need to replace the feathers a second time.

types of feathers and how many are molted, especially for the pre-

Figuerola and Jovani (2001) also identified longer migration distance

alternate molt. As a consequence, a simple binary classification does

as an important factor associated with prealternate molt in Western

not capture the complexity of molt even though this approach was

Palearctic species while Guallar and Figuerola (2016) found this as-

necessary due to methodological limitations. To fully understand the

sociation in North American Motacillidae. We could not repeat the

biological complexity of winter molts, it will be necessary for future

complete prealternate molt analysis with Nearctic passerines be-

studies to categorize the extent of molt as a continuous variable (e.g.,

cause only one species (Dolichonyx oryzivorus; Renfrew et al., 2011)

Table 3, Jenni & Winkler, 2020).

that breeds in North America undergoes a complete prealternate

In conclusion, our results suggest that the evolution of winter

molt. Future analyses on the full range of prealternate molt extents

molt strategies in Nearctic and Western Palearctic migratory pas-

(not just presence/absence of complete molt), as recently under-

serines was likely driven by multiple factors, but photoperiod, migra-

taken on preformative molts by several recent studies (e.g., Guallar

tion distance, and overwintering ground conditions are particularly

et al., 2021), could be informative.

important. What remains to be seen is how the availability of re-

We had hypothesized that NDVI could be an important pre-

sources, and their influence on the costs of feather production and

dictor of molting strategies on the overwintering grounds, but

plumage quality, has played a role in these systems. It is highly im-

NDVI was not a significant predictor of any molt strategy. This lack

portant to understand the drivers behind the different molting strat-

of relationship could be explained by our choice of averaging the

egies because the quality of the molt will impact fitness throughout

NDVI of the overwintering grounds over a period of 7 months (15

the annual cycle (Dawson et al., 2000; Harrison et al., 2011; Nilsson

September to 15 April). This decision could have reduced the im-

& Svensson, 1996). There is a need for a full annual cycle focus in an-

portance of certain environmental events and reduced the impor-

imal ecology to effectively conserve populations (Marra et al., 2015).

tance of NDVI as a potential driver in our results. It is difficult to

In that context, our findings indicate that rapid changes in condi-

select a precise time window that suits every species as molt timing

tions on the overwintering grounds (e.g., through climate change

can be variable and timing is often unknown. However, it could be

and/or habitat loss) could have substantial impacts on the selective

interesting for future studies to separate the NDVI during the over-

forces shaping molt strategies, potentially requiring populations to

wintering months in multiple time periods that cover the earlier pre-

have sufficient plasticity or adaptive capacity to overcome impacts

basic molt (September–­December) and the later prealternate molt

on survival and reproduction. Alternatively, the strong role of static

(January–­April).

components of the abiotic environment, such as photoperiod, in

Molt is a complex process that affects birds for multiple seasons

the evolution of molt strategies may preclude some species from

and impacts many aspects of their life such as flight, thermoregu-

responding to changing biotic conditions, with unknown conse-

lation, communication, and sexual selection (Gill, 2006). Therefore,

quences for fitness.

the evolution of molting strategies was likely influenced by multiple variables impacting at least one function of the plumage. In this

AC K N OW L E D G M E N T

study, we focused on the importance of having a fresh plumage of

We would like to thank B. Turner for his help with the distribution

high quality, but we did not examine the role of sexual selection on

maps and A. Veale for his revisions.

molt strategies, particularly prealternate molt. For future work, it
would be important to examine feather coloration and degree of

C O N FL I C T O F I N T E R E S T

sexual dichromatism, especially as it may play an important role in

We declare we have no conflict of interests.

the evolution of prealternate molt which generally happens before
spring migration and breeding. Terrill et al. (2020) found that, in

AU T H O R C O N T R I B U T I O N S

Parulidae, seasonal dichromatism can only evolve when a prealter-

Claudie Pageau: Conceptualization (equal); data curation (lead); for-

nate molt already exists. Froehlich et al. (2005) hypothesized that, if

mal analysis (lead); funding acquisition (equal); methodology (equal);

the cost of a second molt is too high, birds might keep their conspic-

visualization (equal); writing-­original draft (lead); writing-­review

uous plumage throughout the winter even though this strategy may

& editing (equal). Jared Sonnleitner: Conceptualization (support-

be maladaptive. Thus, it would be interesting to examine the relation

ing); data curation (equal); formal analysis (supporting); methodol-

between dichromatism and prealternate molt across all passerines.

ogy (supporting); writing-­original draft (supporting); writing-­review

Another factor that would have been interesting to analyze is the

& editing (supporting). Christopher M. Tonra: Conceptualization

impact of molt duration on molt strategies (De la Hera et al., 2012);

(equal); data curation (equal); investigation (equal); methodology

however, current available information on this subject is limited. It

(equal); supervision (equal); validation (equal); writing-­original draft

would also be interesting to extend the Tökölyi et al. (2008) study to

(equal); writing-­review & editing (equal). Mateen Shaikh: Formal

all passerines and test breeding onset in relation to the presence of

analysis (equal); methodology (equal); software (equal); supervi-

the prealternate molt.

sion (equal); validation (equal); visualization (equal); writing-­original

| 13257

PAGEAU et al.

draft (supporting); writing-­review & editing (supporting). Matthew
W. Reudink: Conceptualization (equal); data curation (equal); formal
analysis (equal); funding acquisition (equal); investigation (equal);
methodology (equal); resources (equal); supervision (equal); validation (equal); visualization (equal); writing-­original draft (equal);
writing-­review & editing (equal).
DATA AVA I L A B I L I T Y S TAT E M E N T
Data are accessible on Dryad (https://doi.org/10.5061/dryad.5dv41​
ns6b).
ORCID
Claudie Pageau

https://orcid.org/0000-0003-0371-5602

Christopher M. Tonra

https://orcid.org/0000-0002-3499-2576

Matthew W. Reudink

https://orcid.org/0000-0001-8956-5849

REFERENCES
AppEEARS Team (2019). Application for Extracting and Exploring Analysis
Ready Samples (AppEEARS). Ver. 2.30. NASA EOSDIS Land Processes
Distributed Active Archive Center (LP DAAC), USGS/Earth Resources
Observation and Science (EROS) Center. https://lpdaa​c svc.cr.usgs.
gov/appeears
Barta, Z., McNamara, J. M., Houston, A. I., Weber, T. P., Hedenström,
A., & Feró, O. (2008). Optimal moult strategies in migratory birds.
Philosophical Transactions of the Royal Society B: Biological Sciences,
363, 221–­229.
Bergman, G. (1982). Why are the wings of larus fuscus fuscus so dark?
Ornis Fennica, 59, 77–­83.
BirdLife International (2018). Data zone. http://dataz​one.birdl​ife.org/
home
BirdLife International and Handbook of the Birds of the World (2018).
Bird species distribution maps of the world. Version 2018.1. http://dataz​
one.birdl​ife.org/speci​es/reque​stdis
Birds of the World (Billerman S. M. et al Editors) (2020). Cornell
Laboratory of Ornithology. https://birds​ofthe​world.org/bow/home
Butler, L. K., Donahue, M. G., & Rohwer, S. (2002). Molt-­migration in
western tanagers (Piranga ludoviciana): Age effects, aerodynamics,
and conservation implications. The Auk, 119, 1010–­1023.
Butler, L. K., Rohwer, S., & Rogers, M. (2006). Prebasic molt and molt-­
related movements in ash-­throated flycatchers. Condor, 108, 647–­660.
Dawson, A., Hinsley, S. A., Ferns, P. N., Bonser, R. H. C., & Eccleston,
L. (2000). Rate of moult affects feather quality: A mechanism linking current reproductive effort to future survival. Proceedings of the
Royal Society of London, 267, 2093–­2098.
De la Hera, I., DeSante, D. F., & Mllá, B. (2012). Feather growth rate and
mass in Nearctic passerines with variable migratory behavior and
molt pattern. The Auk, 129, 222–­230.
Didan, K. (2015). MOD13A3 MODIS/Terra vegetation indices monthly L3
global 1km SIN grid V006. NASA EOSDIS Land Processes DAAC.
https://doi.org/10.5067/MODIS/​MOD13​A3.006
Earth Point (2020). KML shapes -­Polygon area, linestring length, placemark
point [cited April 12, 2020]. http://www.earth​point.us/Shapes.aspx
Ericson, P. G., Zuccon, D., Ohlson, J. I., Johansson, U. S., Alvarenga, H., &
Prum, R. O. (2006). Higher-­level phylogeny and morphological evolution of tyrant flycatchers, cotingas, manakins, and their allies (Aves:
Tyrannida). Molecular Phylogenetics and Evolution, 40, 471–­483.
Figuerola, J., & Jovani, R. (2001). Ecological correlates in the evolution
of moult strategies in Western Palearctic passerines. Evolutionary
Ecology, 15, 183–­192.
Froehlich, D. R., Rohwer, S., & Stutchbury, B. J. (2005). Spring molt constraints versus winter territoriality. In R. Greenberg, & P. P. Marra

(Eds.), Birds of two worlds (pp. 321–­335). John Hopkins University
Press.
Gill, F. B. (2006). Ornithology (3rd ed.). W.H. Freeman and Company.
Griggio, M., Serra, L., Licheri, D., Campomori, C., & Pilastro, A. (2009).
Moult speed affects structural feather ornaments in the Blue Tit.
Journal of Evolutionary Biology, 22, 782–­792.
Guallar, S., & Figuerola, J. (2016). Factors influencing the evolution
of moult in the non-­breeding season: Insights from the family
Motacillidae. Biological Journal of the Linnean Society, 118, 774–­785.
Guallar, S., Rueda-­Hernández, R., & Pyle, P. (2021). Evolution of the
preformative molt in Cardinalidae correlates with transitions from
forest to open habitats. Ornithology, 138, 1–­14.
Harrison, X. A., Blount, J. D., Inger, R., Norris, D. R., & Bearhop, S. (2011).
Carry-­over effects as drivers of fitness differences in animals. Journal
of Animal Ecology, 80, 4–­18.
Hijmans, R. J. (2019). Package “geosphere”. https://CRAN.R-­proje​c t.org/
packa​ge=geosp​here
Ho, L. S. T., & Ane, C. (2014). A linear-­time algorithm for Gaussian and non-­
Gaussian trait evolution models. Systematic Biology, 63, 397–­4 08.
Humphrey, P. S., & Parkes, K. C. (1959). An approach to the study of molts
and plumages. The Auk, 76, 1–­31.
IUCN (2020). The IUCN red list of threatened species. Version 2020-­1.
https://www.iucnr​edlist.org
Ives, A. R., & Daijiang, L. (2018). rr2: An R package to calculate R 2s for
regression models. The Journal of Open Source Software, 3, 1028.
Ives, A. R., & Zhu, J. (2006). Statistics for correlated data: Phylogenies,
space, and time. Ecological Applications, 16, 20–­32. https://doi.
org/10.1890/04-­0702
Jahn, A. E., Cueto, V. R., Fox, J. W., Husak, M. S., Kim, D. H., Landoll, D. V.,
Ledezma, J. P., Lepage, H. K., Levey, D. J., Murphy, M. T., & Renfrew,
R. B. (2013). Migration timing and wintering areas of three species of
flycatchers (Tyrannus) breeding in the Great Plains of North America.
The Auk, 130, 247–­257.
Jenni, L., & Winkler, R. (2020). Moult and ageing of European passerines
(2nd ed.). HELM.
Jetz, W., Thomas, G. H., Joy, J. B., Hartmann, K., & Mooers, A. O.
(2012). The global diversity of birds in space and time. Nature, 491,
444–­4 48.
Jiguet, F., Kardynal, K. J., Piha, M., Seimola, T., Copete, J. L., Czajkowski,
M. A., Dombrovski, V., Efrat, R., Minkevicius, S., Raković, M.,
Skierczyǹski, M., & Hobson, K. A. (2019). Stable isotopes reveal the
common winter moult of central rectrices in a long-­distance migrant
songbird. Journal of Ornithology, 160, 1077–­1085.
Jones, P. J. (1995). Migration strategies of Palearctic passerines in Africa.
Israel Journal of Zoology, 41, 393–­4 06.
Kiat, Y., Izhaki, I., & Sapir, N. (2019). The effects of long-­distance migration on the evolution of moult strategies in Western-­Palearctic passerines. Biological Reviews, 94, 700–­720.
Kjellén, N. (1994). Moult in relation to migration –­A review. Ornis Svecica,
4, 1–­24.
Leu, M., & Thompson, C. W. (2002). The potential importance of migratory stopover sites as flight feather molt staging areas: A review for
neotropical migrants. Biological Conservation, 106, 45–­56.
Lindström, Â., Visser, G. H., & Daan, S. (1993). The energetic cost of
feather synthesis is proportional to basal metabolic rate. Physiological
Zoology, 66(4), 490–­510.
Marra, P. P., Cohen, E. B., Loss, S. R., Rutter, J. E., & Tonra, C. M. (2015). A
call for full annual cycle research in animal ecology. Biological Letters,
11, 20150552.
Murphy, M. E., & King, J. R. (1991). Energy and nutrient use during moult
by white-­crowned sparrows Zonotrichia leucophrys gambelii. Ornis
Scandinavica, 23, 304–­313.
Murphy, M. E., & King, J. R. (1992). Energy and nutrient use during moult
by white-­crowned sparrows Zonotrichia leucophrys gambelii. Ornis
Scandinavica, 23, 304–­313.

13258

|

National Research Council Canada (2020). Advanced options and sun angles.
https://www.nrc-­cnrc.gc.ca/eng/servi​ces/sunri​se/advan​ced.
html
Nilsson, J.-­A ., & Svensson, E. (1996). The cost of reproduction: A new
link between current reproductive effort and future reproductive
success. Proceedings of the Royal Society of London B, 263, 711–­714.
Pageau, C., Tonra, C. M., Shaikh, M., Flood, N. J., & Reudink, M. W.
(2020). Evolution of moult-­migration is directly linked to aridity
of the breeding grounds in North American passerines. Biological
Letters, 16, 20200155.
Pageau, C., Vale, M. M., de Menezes, M. A., Barçante, L., Shaikh, M.,
Alves, M. A. S., & Reudink, M. W. (2020). Evolution of altitudinal
migration in passerines is linked to diet. Ecology and Evolution, 10,
3338–­3345.
Pyle, P. (1997). Identification guide to North American birds Part I. Slate
Creek Press.
Pyle, P. (1998). Eccentric first-­year molts in certain Tyrannid flycatchers.
Western Birds, 29, 29–­35.
Pyle, P., & Kayhart, R. (2010). Replacement of primaries during the prealternate molt of a Yellow Warbler. North American Bird Bander, 35,
178–­181.
Pyle, P., Leitner, W. A., Lozano-­Angulo, L., Avilez-­Teran, F., Swanson, H.,
Limón, E. G., & Chambers, M. K. (2009). Temporal, spatial, and annual variation in the occurrence of molt-­migrant passerines in the
Mexican Monsoon region. Condor, 111, 583–­590.
R Core Team (2020). R: A language and environment for statistical computing. R Foundation for Statistical Computing.
Rambaut, A., & Drummond, A. J. (2018). TreeAnnotator v1.10.4: MCMC
output analysis. http://beast.commu​nity/
Remisiewicz, M., Bernitz, Z., Bernitz, H., Burman, M. S., Raijmakers, J. M.
H., Raijmakers, J. H. F. A., Underhill, L. G., Rostkowska, A., Barshep,
Y., Soloviev, S. A., & Ilona Siwek, I. (2019). Contrasting strategies for
wing-­moult and pre-­migratory fuelling in western and eastern populations of Common Whitethroat Sylvia communis. IBIS, 161, 824–­838.
Renfrew, R. B., Frey, S. J. K., & Klavins, J. (2011). Phenology and sequence
of the complete prealternate molt of Bobolinks in South America.
Journal of Field Ornithology, 82, 101–­113.
Revell, L. J. (2012). phytools 0.6.99: A R package for phylogenetic comparative biology (and other things). Methods in Ecology and Evolution,
3, 217–­223.
Rohwer, S., Butler, L. K., & Froehlich, D. (2005). Ecology and demography
of east-­west differences in molt scheduling of neotropical migrant
passerines. In R. Greenberg, & P. P. Marra (Eds.), Birds of two worlds:
The ecology and evolution of migration (pp. 87–­105). Johns Hopkins
University Press.
Serra, L. (2001). Duration of primary moult affects primary quality in Grey
Plovers Pluvialis squatarola. Journal of Avian Biology, 32, 377–­382.

PAGEAU et al.

Shirihai, H., & Svensson, L. (2018). The handbook of western Palearctic
birds: Passerines. Bloomsbury Publishing.
Spiess, A.-­N. (2018). Package “qpcR”. https://CRAN.R-­proje​c t.org/packa​
ge=qpcR
Svensson, E., & Hedenström, A. (1999). A phylogenetic analysis of the
evolution of moult strategies in Western Palearctic warblers (Aves:
Sylviidae). Biological Journal of the Linnean Society, 67, 263–­276.
Terrill, R. S., Seeholzer, G. F., & Wolfe, J. D. (2020). Evolution of breeding
plumages in birds: A multiple-­step pathway to seasonal dichromatism
in New World Warblers (Aves: Parulidae). Ecology and Evolution, 10,
9223–­9239.
Tökölyi, J., Bókony, V., & Barta, Z. (2008). Seasonal colour change by
moult or by the abrasion of feather tipes: A comparative study.
Biological Journal of the Linnean Society, 94, 711–­721.
Tonra, C., & Reudink, M. W. (2018). Expanding the traditional definition
of molt-­migration. The Auk, 135, 1123–­1132.
Underhill, L. G., Prys-­Jones, R. P., Dowsett, R. J., Herroelen, P., Johnson,
D. N., Lawn, M. R., Norman, S. C., Pearson, D. J., & Tree, A. J. (1992).
The biannual primary moult of Willow Warblers Phylloscopus trochilus in Europe and Africa. IBIS, 134, 286–­297.
Voelker, G., & Rohwer, S. (1998). Contrasts in scheduling of molt and
migration in eastern and western warbling-­vireos. The Auk, 115,
142–­155.
Willman, H., Belmaker, J., Simpson, J., de la Rosa, C., Rivadeneira, M. M.,
& Jetz, W. (2014). Eltontraits 1.0: Species-­level foraging attributes
of the world's birds and mammals. Ecology, 95(7), 2027. https://doi.
org/10.1890/13-­1917.1
Wright, J. R., Tonra, C. M., & Powell, L. L. (2018). Prealternate molt-­
migration in Rusty Blackbirds and its implications for stopover biology. Condor, 120, 507–­516.

S U P P O R T I N G I N FO R M AT I O N
Additional supporting information may be found online in the
Supporting Information section.

How to cite this article: Pageau, C., Sonnleitner, J., Tonra, C.
M., Shaikh, M., & Reudink, M. W. (2021). Evolution of winter
molting strategies in European and North American
migratory passerines. Ecology and Evolution, 11, 13247–­
13258. https://doi.org/10.1002/ece3.8047