Content uploaded by Mark Fellowes
Author content
All content in this area was uploaded by Mark Fellowes on Mar 12, 2024
Content may be subject to copyright.
Molecular Ecology. 2024;00:e17311.
|
1 of 13
https://doi.org/10.1111/mec.17311
wileyonlinelibrary.com/journal/mec
Received:15August2023
|
Revised:8December2023
|
Accepted:30January2024
DOI:10.1111/mec.17311
ORIGINAL ARTICLE
Does urbanisation lead to parallel demographic shifts across
the world in a cosmopolitan plant?
Aude E. Caizergues1,2 | James S. Santangelo3 | Rob W. Ness1,2 | Fabio Angeoletto4 |
Daniel N. Anstett5 | Julia Anstett6,7 | Fernanda Baena- Diaz8 | Elizabeth J. Carlen9 |
Jaime A. Chaves10,11 | Mattheau S. Comerford12 | Karen Dyson13 |
Mohsen Falahati- Anbaran14 | Mark D. E. Fellowes15 | Kathryn A. Hodgins16 |
Glen Ray Hood17 | Carlos Iñiguez- Armijos18 | Nicholas J. Kooyers19 |
Adrián Lázaro- Lobo20 | Angela T. Moles21 | Jason Munshi- South22 | Juraj Paule23 |
Ilga M. Porth24 | Luis Y. Santiago- Rosario25 | Kaitlin Stack Whitney26 |
Ayko J. M. Tack27 | Marc T. J. Johnson1,2
This is an op en access arti cle under the ter ms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction
in any medium, provided the original work is properl y cited an d is not use d for comm ercial purposes.
© 2024 The Aut hors. Molecular EcologypublishedbyJohnWiley&SonsLtd.
AudeE.C aizerg ues,JamesS.S antan gelo,RobW.NessandMarcT.J.Johnsonaremembersofthel eadtea m.
For Aff iliation refer p age on 10
Correspondence
Aude E. C aizergues, Ce ntre for Urban
Environments, University of Toronto
Mississauga,Mississauga,Ontario,
Canada.
Email: audeemiliecaizergues@gmail.com
Funding information
Canadian Network for Research and
InnovationinMachiningTechnology,
NaturalSciencesandEngineering
Research Council of Cana da; Canada
ResearchChairs;SchoolofCities
Handling Editor:RichardJAbbott
Abstract
Urbanisation is occurring globally, leading to dramatic environmental changes that are
alteringtheecologyandevolutionofspecies.Inparticular,theexpansionofhumanin-
frastructure and the loss and fragmentation of natural habitats in cities is predicted to
increase genetic drift and reduce gene flow by reducing the size and connectivity of
populations. Alternatively, the ‘urban facilitation model’ suggests that some species will
have greater gene flow into and within cities leading to higher diversity and lower differ-
entiation in urban populations. These alternative hypotheses have not been contrasted
across multiple cities. Here, we used the genomic data from the GLobal Urban Evolution
project (GLUE), to study the effects of urbanisation on non- adaptive evolutionary pro-
cesses of white clover (Trifolium repens)ata globalscale.Wefound thatwhite clover
populations presented high genetic diversity and no evidence of reduced Ne linked to
urbanisation.Onthecontrary,wefoundthaturbanpopulationswerelesslikelytoex-
periencearecentdecreaseineffectivepopulationsizethanruralones.Inaddition,we
found little genetic structure among populations both globally and between urban and
ruralpopulations,whichshowedextensivegeneflowbetweenhabitats.Interestingly,
white clover displayed overall higher gene flow within urban areas than within rural hab-
itats.Ourstudyprovidesthelargestcomprehensivetestofthedemographiceffectsof
urbanisation.Ourresultscontrastwiththecommonperceptionthatheavilyalteredand
fragmented urban environments will reduce the effective population size and genetic
diversity of populations and contribute to their isolation.
2 of 13
|
CAIZERGUES et al.
1 | INTRODUC TION
In the pas t century, urbanis ation has rapidl y modified land scapes
throughout the world. Urbanisation replaces natural and rural hab-
itats with highly disturbed, human- modified habitats, characterised
by a high density of humans, buildings and roads, and consequently
more impervious surfaces, increased pollution, and elevated habi-
tat loss and fragmentation (Grimm et al., 2008; Liu et al., 2020).
These landscape alterations influence multiple evolutionary pro-
cesses, including natural selection, mutation, gene flow and genetic
drift ( Johnson & Munshi-South, 2017; Mil es et al., 2019; Schmidt
et al., 2020; Somers et al.,2004; Yauk et al., 2008). For example,
we know that cities influence natural selection, yet not all species in
urban habitats have adapted to these new environments (Lamber t
et al., 2021). The potential of a population to adapt to a new envi-
ronment is heavily influenced by the availability of genetic diversity,
which is in part driven by demographic changes in population size
and the connections among populations. Hence, the demographic
history of a population can play a major role in its past and present
evolutionary trajectory, which raises concerns about the future of
urban dwelling species and their conservation. How urbanisation
influences the evolutionary demography of populations remains
poorly understood and is the focus of our study.
Habitat fragmentation and degradation can influence population
demographic processes, which in turn can affect genetic drift within
populations and gene flow between populations (Hanski, 1998).
Fragmentation occurs when habitats are divided into patches with
limited corridors for dispersal among remaining habitat fragments.
For instance, when non- urban habitats are converted into build-
ings and roads, it splits and potentially isolates natural patches. This
phenomenon can reduce the size, increase the isolation (i.e. by lim-
iting dispersal), and influence the persistence of populations. One
predicted evolutionary outcome to urban habitat fragmentation
is increased genetic drift , which can reduce fitness when there is
a loss of genetic diversity, which is often associated with elevated
inbreeding, accumulation of deleterious mutations, and reduced
ability to adapt to environmental changes. The reduced gene flow
thatensuesfromtheisolationofpopulationsisalsoexpectedtolead
to increased genetic differentiation between populations (Beninde
et al., 2018; Lourenço et al., 2017;Munshi-South et al.,2016). The
hypothesis of increased drif t and divergence of urban populations
is referred to as the ‘urban fragmentation model’, and it is widely
thought to be the most prevalent outcome of urbanisation on the
non-adaptiveevolutionofpopulations(Milesetal.,20 19).
While the u rban fragment ation model has t he most empiric al
support(Milesetal.,2019), urbanisation can also facilitate the eco-
logical success andevolutionarypotentialofsome species.Inpar-
ticular, human commensals maythrive inurban habitats (Carlen &
Munshi-South, 2021; Medina et al., 2018; Rochat et al., 2017). In
such cases, urban features and human behaviour can facilitate in-
dividual movement between populations and create corridors of
dispersal and gene flow, leading to higher genetic diversity within
urban populations and decreased divergence between urban pop-
ulations(Milesetal.,2018).Suchobservationshaveledtoanalter-
nativehypothesis–the‘urbanfacilitationmodel’(Milesetal.,20 19).
Understanding which of these two scenarios is likely to occur for
any given species is of major impor tance to understanding how ur-
banisation shapes evolution, because demographic processes such
as changes in population size and dispersal influence the amount of
genetic variation within and between populations, and thus the evo-
lutionary potential of populations.
Urbanisation is expected to cause similar evolutionary pro-
cesses and patterns across cities (Santangelo,Rivkin, et al.,2020),
since urbanisation frequently leads to similar environmental changes
(McKinney, 2006; Sant angelo et al., 2022). Whi le there has been
a focus on studying whether urbanisation causes parallel adap-
tive evolution to different cities (Caizergues et al., 2022; Salm ón
et al., 2021), there has not been the same focus on whether there
could be parallel non- adaptive processes. This leads to the question:
Doesurbanisationconsistentlyleadtoincreasedgeneticdriftwithin
populationsand divergence between populations?Somehavepre-
dicted that urbanisation can drive parallel non- adaptive processes
(Lambert et al., 2021;Santangelo, Miles, et al., 2020), butexisting
tests of this prediction are inconclusive and limited in spatial scale
(Beninde et al., 2018; Combs et al., 2018; Mueller et al., 2018;
Theodorou et al., 2018). Hence, understanding if urbanisation
causes parallel evolutionary and demographic patterns among cities
throughout the globe remains unresolved.
Wehave beenusingwhite clover(Trifolium repens L., Fabaceae)
as a model to understand how global urbanisation affects evolu-
tion as par t of the GLobal Urban Evolution (GLUE) project (www.
globa lurba nevol ution. com). Here we combine new data with the
large-scale d ataset gener ated as part of GLU E. We previously re -
portedthatwhite cloverfrequently exhibited urban–ruralclines in
the production of an antiherbivore defence trait, and this repeated
evolution was attributed to adaptive evolution. Previous results also
revealed that populations of clover consistently showed high genetic
diversityandlowworldwidegeneticstructure(Johnsonetal.,2018;
Santangelo et al.,2022;Wuet al.,2021),butwe did notexplicitly
investigate how urbanisation impacts genetic drift and gene flow, or
the potential for parallel non- adaptive evolution across the world.
Here, we seek to understand how urbanisation impact s non-
adaptive evolutionary processes, including demography and gene
flow,of whitecloverincities throughouttheworld. Weaddressed
the following questions: (1) Does urbanisation repeatedly cause
changes in effective population size across cities, thus influencing
KEYWORDS
effective population size, gene flow, genetic diversity, neutral evolution, urbanisation
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17311 by Test, Wiley Online Library on [12/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
|
3 of 13
CAIZERGUES et al.
genetic diversity and inbreeding? (2) Does urbanisation influence
gene flow, differentiation and structure among populations? To an-
swer these questions, we sampled and performed whole genome
sequencing of over 2000 plants from urban and rural populations of
white clover in 24 cities across the world, and used both population
genomic analyses and demographic modelling to reconstruct urban
and rural population histories. Given that the 24 cities sampled are
geographically distant from each other, and since the urbanisation
process occurred independently in each city, we can expect the
evolutionary trajectories of white clover populations to be suffi-
ciently independent among cities to treat them as independent tests
of evolutionary changes, and thus suitable for the study of parallel
evolution. Based on white clovers' strong positive association with
human modified habitats, we predicted demographic processes and
gene flow would better reflec t the urban facilitation model than the
urbanfragmentationmodel.Specifically,wepredictedthatT. repens
populations thrive more in urban areas than rural areas, which would
be seen as higher effective population size, higher genetic diversity,
lower inbreeding, and lower population struc ture due to high gene
flow among populations compared to rural areas.
2 | METHODS
2.1 | Study system and data sampling
Whiteclover(Trifolium repens L., Fabaceae) is an herbaceous peren-
nialplantnativetoEurasia.Itcanreproduceclonallyviastolonsthat
spreadhorizontally onthesoilsurfaceorthroughsexualreproduc-
tion via outcrossing (Burdon, 1983). Each plant produces inflores-
cences with numerous hermaphroditic flowers that are pollinated
bya diversity of bee species (Kakes, 19 97 ). White cloveris anal-
lotetraploid (Griffiths et al., 2 019)withdisomicinheritance(Williams
et al., 1998). Because of its ability to fix atmospheric nitrogen,
white clover has been introduced to all inhabited continents in the
past several hundred years as livestock fodder and as a cover crop
(Kjærgaard,2003).Itsglobaldistributioncoversawiderangeofcli-
mates, and the fact that it grows in anthropogenically modified habi-
tats (e.g. mowed grass, pastures), makes it an ideal model to study
urban evolutionary biology at a global scale.
OurstudyexpandsupontheGLUEprojectthataimedtoexam-
inehowglobalurbanisationaffectsparalleladaptation(Santangelo
et al., 2022). As a brief back ground, scientists from around the world
sampled 20–50 populations from each city along urban–rural tran-
sects.Intotal,110,019plantsfrom6169populationswerecollected
in 160 cities from 27 countries. Among these, a subset of 24 cities,
chosen to capture variation in geography and climate, were subject
to whole genome sequencing. For each of these selected cities, ~42
individuals from the five subpopulations closest to the city centre
and an equivalent number of individuals from the five furthest rural
subpopulations were selected for subsequent genomic analyses
(mean ± SD = 83.8 ± 14.6individuals/city,seeTable S1).
Whiletheinitialanalysesfocusedonadaptiveparallelevolution
of a single trait (i.e. ability to produce hydrogen cyanide) and the loci
controlling it (CY P79D15 and Li),herewespecificallyexaminedhow
urbanisationaffects non-adaptive genomicevolution.Weused an
improved and larger dataset including 19 cities from the initial GLUE
project for which good qualit y genomic data were available (remov-
ing cities where more than 50% of the sampled individuals had a
coverage <0.5×) and five newly sequenced cities (Palmerston North,
New Zealand; Punta Arenas, Chile; Sapporo, Japan; Vancouver,
Canada and Warsaw, Poland), for a worldwide sample across all
inhabited continent s (Figure 1). The final whole genome sequence
dataset included six European, three Asian, one African, seven
NorthAmerican,fourSouthAmericanandthreeOceaniancities,for
a total of 2013 individuals.
2.2 | Molecular biology and genome sequencing
To obtain our genomic dataset, we used the sequencing protocol
fromGLUE(Santangeloetal.,2022) when sequencing the five cit-
ies newly added to theoriginal genomic dataset.Briefly,weex-
tractedgenomicDNAfromfreeze-driedsamplesusingamodified
phenol–chloroform extraction protocol (detailed in Santangelo
et al., 2022),andqua nti fiedDNAaf terextra ctionu sin gth edsDNA
HSAssay Kit(FisherScientific, Mississauga, Canada). We gener-
ated dual-indexed genomic DNA libraries following Santangelo
et al. (2022) and Glenn et al. (2019), and sequenced genomic librar-
ieswithaconcentrationofDNA≥0.8 ng/μL. The genomes of 1922
FIGURE 1 Citiessampledforrural
and urban populations of white clover
Trifolium repens. Each circle represents a
city in which we sequenced the genomes
of80plantsonaveragefromurbanand
rural habitats. The colour scale represents
nucleotide diversity, measured as pi.
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17311 by Test, Wiley Online Library on [12/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
4 of 13
|
CAIZERGUES et al.
individualsfrom23cities(allexceptToronto)weresequencedon
aNovaseq6000S4platformusing150 bppaired-endreadsatlow
coverage (1.05× on average). Ninety plants from Toronto were
sequenced at ~13× as a part of another project, and downsam-
pledusingSAMtools(v1.10)to~2.5× to be included in this study.
Seven citie s from Santangelo et al. (2022) were removed from
the dataset because of low sequence quality and/or low sample
size (Bogotaand Medellin, Colombia; Canberra and Melbourne,
Australia;HiroshimaandKyoto,Japan;andParis,France),leaving
us with data from 24 cities (Figure 1).
2.3 | Sequence alignment and genotype likelihoods
Beginning with fastq sequence data, we first used fastpV0.20.1
to trim raw reads with the - trim_poly_g argument to remove polyG
tails commonly generated by the Novaseq platform and performed
quality checks on both raw and trimmed reads of every sample
using FastQCv0.11.9.WethenmappedtrimmedreadstotheT. re-
pensreferencegenome(NCBIBioProjectnumberPRJNA523044,
Grif fiths et al., 2019) with B WA MEM v0.7.17. Alt hough white clover
isanallotetraploid,i te xhibi tsdis omicinh eritance ,alllocia refunc-
tionally diploid. For this reason, reads could be mapped indepen-
dently toeach of the twoancestralgenomes.We usedSAMtools
to sort , index and mar k duplicates i n BAM files. We per formed
a quality check of mapped reads with Qualimap v2.2.2, Bamtools
v2. 5.1, BamUtil v1.0.14 and multiQCv1.11.Sincecallingindividual
genotypes from low coverage data can lead to bias in variant de-
tection (Han et al., 2014; Nielsen et al., 2012), we computed geno-
type likelihoods with ANGSDv0.933(Korneliussenetal.,2014) on
all four- fold degenerate sites (see below) using the SAMtools geno-
type likelihood model. Previous evidence demonstrated that allele
frequencies and population genetic estimates can be robustly es-
timated with sample sizes and sequence coverage levels similar to
those implemented in our study (Lou et al., 2021). As demographic
inference can be biased by sites under selection, we focused our
analyses on four- fold degenerate sites that are largely considered
toevolveneutrally.Wefilteredforaminimumphred-scaledbase
quality score of 20 and minimum mapping quality of 30, keep-
ing both variant and invariant sites present in a minimum of 50%
of sequen ced individu als (405,118 total sites pas sed the quali ty
filteringcriteriaandwere commonto all populations).While the
framework of our bioinformatic pipeline is built upon the ear-
lier Santangelo et al. (2022) pipeline (https://github.com/James
- S - S a n t a n g e l o / g l u e _ p c ), all analyses and results are new, using a
larger and improved dataset to specific ally focus on the evolution-
ary signatures of population demographic change in response to
urbanisation. A diagram summarising all the bioinformatic steps is
available in Figure S1 and all analyses were integrated into a repro-
ducible Snakemake (Mölder et al.,2021) pipeline (https:// github.
c o m / A u d e C a i z e r g u e s / g l u e _ d e m o g r a p h y / ).
Thesite frequencyspectra (SFS)was the basisfor mostof the
analyses performed in ANGSD and other s oftware. We es timated
theSFSatfour-folddegeneratesitesintheT. repens reference ge-
nome using the ‘degeneracy’ pipeline (g i t h u b . c o m / t v k e n t / D e g e n
eracy ). We retainedan average of 2,110,100four-fold degenerate
sites that were present in at least 60% of individuals per popula-
tion per city per habitat af ter filtering (see b elow). Since related
individuals can bias population genomic analyses, and clover has
the ability to grow clonally, we identified closely related individu-
als using NgsRelate (Hanghøj et al., 2 019) with default parameters,
and removed individuals with a pairwise relatedness rxy > .5,where
rxy = .5correspondstoaparent–offspringorfull-siblingrelationship;
atotal of28individuals wereremovedfrom subsequent analyses.
Afterremovingrelatedindividuals,weestimatedthefoldedSFSfor
eachurbanandruralpopulation.Wealsoestimatedthefoldedtwo-
dimensio nal (2D) SFS for each ur ban–rural pair p er city using th e
realSFS ANGSD function.
2.4 | Data analyses
2.4.1 | Geneticdiversityestimatesand
demographic inference
To investigate the effects of urbanisation on neutral evolution-
ary processes within populations, our first step was to estimate
genetic diversity, effective population size, gene flow and lev-
els of relatedness within each habitat (urban and rural) of a cit y.
To characterise population genomic parameters of diversit y, we
estimated nucleotide diversity based on π (pairwise nucleotide
diversity), Watterson's theta (θw) (Nei, 1975; Tajima, 198 3) and
Tajima'sD (Tajima,1989),fromthe SFS usingthe thetaStat func-
tion of ANGSD. These three estimators were estimated both at the
population level (all sub- populations of a habitat pooled together)
andatthesub-populationlevel.Inbreedinglevelswithinsubpopu-
lations were estimated as mean rxy (i.e. the pairwise relatedness,
Hedrick & L acy, 2015), computed with NgsRelate using default
parameters. To understand whether these diversity parameters
differedbetweenenvironmentsweincludedtheminlinearmixed
modelswithhabitat(urbanvs.non-urban)asanexplanatoryvari-
ableandcityasarandomeffect.Wethenestimatedtheeffective
population size (Ne)ineachhabitat.Sinceurbanpopulationsare
unlikely to be at mutation–drift equilibrium, statistics like π and θw
will not capture recent changes in Ne.Wethereforeusedamodel-
based approach to estimate the dynamics of contemporar y Ne.
Ineachcity,wereconstructedvariationofNe through time using
EPOS (Lynch et al., 2020), with 100 0 bootstrap iterations and a
mutationrateof1.8 × 10−8 (Griffiths et al., 2019). Raw outputs of
EPOS were converted to a plottable format using the epos2plot
function. These analyses gave us a detailed understanding of how
urbanisation affects genetic diversit y and effective population
size within populations.
In additi on, to explore th e potential role of ci ty age on clo-
ver's evolution, we investigated if population genetic estimates (π,
θwand Tajima's D)and Ne were correlated with the age of each
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17311 by Test, Wiley Online Library on [12/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
|
5 of 13
CAIZERGUES et al.
city.FollowingSantangelo,Rivkin,etal.(2020), we defined a city's
age as the number of years before 2020 that each city reached a
populationof150,000.Welog-transformedcityagetonormalise
the data and performed Pearson correlation tests with each of the
predictors.
2.4.2 | Geneticstructure
We next investigated how urbanisation influenced population
genetic structure and gene flow of T. repens within and between
urban and rural habitats. First, since linkage disequilibrium can
bias genetic structure analyses, we identified linked positions
among the four-fold degenerate SNPs with a minor allele fre-
quency ≥ .05usingngsLD(Foxetal.,2 019), and pruned the linked
SNPswithin20 kbusingar2 cut- off of 0.2, resulting in a dataset
of 41,543 sites. The pruned dataset was used for subsequent prin-
cipal component analysis (PCA) and admixture analysesonly. To
describe the genetic structure both between habitats and cities,
we performed a PC A with PCAngsd(Meisner&Albrechtsen,2018)
thatestimatesa variance–covariancematrixofallelefrequencies
directlyfromgenotypelikelihoods.Wethenestimatedpopulation
differentiation using Hudson's FST separately for each city (Hudson
et al., 1992).WecomputedFST with the realSFS fst index function
fromthetwo-dimensionalSFS.First,tounderstandgeneticstruc-
ture between habitats, we computed FST between each urban–
rural pair of populations, and similarly with diversity estimators,
investigatedifitwas correlated with cityage.Second,tofurther
analyse patterns of differentiation within habitats, we computed
FST between all pairs of subpopulations per city, and obtained an
average FST for urban–urban, urban–rural or rural–rural compari-
sons separately. As a complementary analysis of urban–rural dif-
ferentiation,weestimatedadmixtureproportionsforeachpairof
urban–rural populations with NGSadmix(Skotte etal., 2013). We
ran the analysis for cluster numbers (K) from 2 to 10, with 10 itera-
tions per K and selec ted the best K for each city using the method
of Evanno et al. (2005) implemented in CLUMPAK (Kopelman
et al., 2015). Finally, we used GADMA (with dadi engine) to estimate
migration rates between urban and rural populations (Gutenkunst
et al., 2010; Noskova et al., 2020).GADMAimplementsmethods
for automatic inference of the joint demographic history of multi-
plepopulations.Itthereforeeliminatestheneedtotestformulti-
ple demographic models in parallel as its algorithm automatically
models a multitude of scenarios and parameter values and selects
thebest supportedones. We modelled eachcity separately,and
set the initial structure to [1,1] (indicating one epoch before the
population split), final structure to [2,1] (allowing two epochs af ter
the split to model any change in population size), mutation rate to
1.8 × 10−8,generationtime to 1 year,sequencelengthtothesize
oftheSFS(number of sites) per city,andotherparameters were
set to their default mode. To have a more comprehensive under-
standing of gene flow we also estimated migration rate (m) using
Wright'sequationNem = (1/FST-1)/4(Wright,1984).
3 | RESULTS
3.1 | Does urbanisation repeatedly affect genetic
diversity, relatedness within populations and effective
population size?
Demographic processes frequently differed among cities, yet ge-
netic diversity and effective population size were, on average,
similarbetweenurbanandruralpopulations.Whilesomecitiesdis-
played higher nucleotide diversity than others within cities (e.g. see
Linkoping , Sweden, Figure 2a), both urban and rural clover popu-
lations had high nucleotide diversity (mean ± SD π = 0.025 ± 0.004,
Figure 2a; mean ± SD θw = 0.028 ± 0.006, Figure S2), which did
not differ between urban and rural habitats (habitat effect for π:
F1,23 = 0.505, p = .485; habitat effectfor θw: F1,23 = 1.303, p = .265).
Taj i ma' s Dvariedamongcitiesfrom−0.682to0.318(Figure 2b), and
was negative on average, which is consistent with recent population
expansion.Tajima'sD also varied within cities, but there was no con-
sistent dif ference between urban and rural habitats (F1,23 = 2.935,
p = .100). Such hig h levels of genetic di versity are co nsistent with
large effective population sizes. None of the π, θworTajima'sDper
sub- population (within habitat) estimates revealed any clear small-
scale pattern (Figures S3–S5).
Overall,Ne was high but varied substantially bet ween cities and
was more likely to decline in rural than urban habitats. Ne varied
among cities by four orders of magnitude, ranging from 1750 to
40,800,000(Table S2; Figure S6). Comparing urban and rural hab-
itats, Ne was twice as likely to be higher in the urban habitat; there
were seven occurrences of Ne urban > Ne rural(e.g.Munich,Germany,
Figure 3a), three occurrences of Ne urban < Ne rural (e.g. Toronto,
Figure 3b) and 14 occurrences of Ne urba n = Ne rural (e.g. Kunming ,
Figure 3c). When Ne was compared between habitats with para-
metric(LMER)and non-parametric(Wilcoxonranktest)analyses,
we found no consistent ef fect of urban/rural habitat on Ne(LMER:
𝜒
2
1 = 1.181, p = .27 7;W ilcoxon rank tes t: p = .107). Wef ound low
levels of relatedness between individuals, and relatedness did not
differ between urban and rural habitats (Figure 2c,Wilcoxonrank
test: p = .303),suggesting urbanisation did notaffect the propen-
sity of inbreeding. Using the whole genome dataset, we modelled
recent changes in Ne through time in both habitats. Although Ne
did not consistently differ between habitat s, urban populations
were less likely to show a recent decrease in Neinthelast500 years
compared tor uralp opulations.Specifically, Ne decreased in four
urban populations, whereas it decreased in 11 rural habitats, and
the remaining populations were stable between habitats (Figure S6;
𝜒
2 = 4.751,p = .029).
Finally, we found no link bet ween the age of a city and any of the
diversit y estimators or Ne (π: r = .040, p = .874;θw: r = .051,p = .840;
Tajima'sD:r = −.103,p = .690;Ne: r = .022,p = .932).
Taken together, our results suggest that while clover shows over-
all high Ne that maintains substantial genetic diversity within popu-
lations, urban habitats are more likely to maintain large and stable
populations than rural areas.
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17311 by Test, Wiley Online Library on [12/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
6 of 13
|
CAIZERGUES et al.
FIGURE 2 Summaryofpopulationgeneticparametersaveragedoverallcities(left,mean ± 95%CI)anddetailedpercity(right,
mean ± 95%CI)for:(a)pairwisenucleotidediversity,(b)Tajima'sD,(c)averagedrelatednessbetweenindividualsand(d)FST averaged
between pairs of sub populations. Lef t graphs represent averaged values per habit at for FST and rxy and genome wide averages for π and
Tajima'sD.In(a,b),greenrepresentsruralhabitatsandpurplerepresentsurbanhabitats.In(c,d),green,blackandpurplerespectively
represent rural–rural, rural–urban and urban–urban for FST and rxy.
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17311 by Test, Wiley Online Library on [12/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
|
7 of 13
CAIZERGUES et al.
3.2 | Does urbanisation influence gene flow,
differentiation and genetic structure?
Analysis of population structure shows that while cities can be ge-
netically differentiated from one another on a global scale, urban
and rural populations show limited structure within a given city. At
a global scale, PC A analyses revealed that many cities cluster close
to each other if they are geographically closer to one another. For
instance,NorthandSouthAmericancitiesweretypicallymorege-
netically similar to one another than they were to European cities
(Figure 4a,b). By contrast, several cities were genetically distinct,
suchasCapeTown,SouthAfrica,theonlyAfricancitysampled,and
Tehran,Iran,andThessaloniki,Greece,whicharegeographicallyiso-
lated from other sampled points (Figure 4a,b).Withinacity,PCAre-
vealed no strong pat tern of differentiation between urban and rural
populations (Figure S7).
WhenpopulationdifferentiationwasquantifiedwithFST, we ob-
served that overall between- population FST was low (Figure 2d) and
not correlated with city age (r = .11,p = .678),buttherewasevidence
of fine- scale differences when moving from between- to within-
habitat comparisons. Specifically, while the average FST between
urban and rural populations was similar to FST within rural habi-
tats (me an ± SD FSTurb-rur = 0.029 ± 0.028, FST rur-rur = 0.026 ± 0.016,
Wilcoxon rank test: p = .931), FST within urban habitats was 19%
lower than FST within rural habitat and 27% lower than FST between
habitats (mean ± SD FS Turb-urb = 0.021 ± 0.013; Wilcoxon rank test
FSTurb-urb- FS Trur-rur: p = .031,Wilcoxon ranktest FSTur b-urb- FSTurb-rur:
p = <.001). These differences in population differentiation between
habitatsareconsistentwithmoreextensivegeneflowamongsites
within urban habit ats compared to the gene flow among sites within
rural habitats, or between urban and rural habit ats.
Estimates of historical and recent demographic processes re-
veal extensive gene flow between urban and rural populations.
Admixture analyses identified admixturebetween urban andrural
populations,withtwotosixpopulations(bestsupported‘K’) found
within cities that were broadly shared between urban and rural in-
dividua ls. While we detec ted small scale s tructure be tween pop-
ulations, the proportion of each assumed ancestral cluster did not
differ between urban and rural habitats (Figure 5; per cit y G- test on
average adm ixture prop ortions p er habitat, a ll p > .05). More than
halfofthedemographicmodelsestimatedwithGADMAshowedbi-
orunidirectionalgeneflowbetweenhabitats(13of24).Wecaution
thatinnumerouscitiesmanydemographicmodelsexplainedsimilar
amounts of variation and these models often included substantial
gene flow. There was no dif ference in urban- to- rural and rural- to-
urban mig ration rates bas ed on estimate s from the best G ADMA
models (Figure S8). Similarly, estimates of Nem (Wright, 1931),
showed no clear variation between habitats (Figure S9), however for
low levels of FST such as observed here, Nem can vary widely (see
Figure 2ofWhitlock&McCauley,1999), so these results should be
interpretedwithcaution.Inanycase,whencombiningthesemigra-
tion rates results with PCA, FSTandadmix t urere sults,wefin dacle ar
pattern of high ongoing gene flow between habitat s.
4 | DISCUSSION
We tested whether global urbanisation leads to parallel non-
adaptive evolutionary processes and patterns in the cosmopolitan
plant white clover (T. repens). Specifically,wesoughttounderstand
whether urban and rural habitats consistently differed in genetic di-
versity, inbreeding, effective population size, genetic structure and
geneflow.Ourgenomewideanalysesconsistingof1922individuals
from 24 locations around the world revealed high levels of diversit y
in both urban and rural populations of white clover. Almost all urban
FIGURE 3 Effectivepopulationsize(Ne)inthepast1000 years
for three representative pairs of urban (purple lines) and rural
(greenlines)populationsinMunich(Germany),Toronto(Canada)
andKunming(China),showingthediversityofpatternsofrecentNe
variations across the dataset, where we find no consistent effect
of urbanisation on Ne.Solidlinesrepresentthemediananddashed
line the 5% and 95% quantiles.
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17311 by Test, Wiley Online Library on [12/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
8 of 13
|
CAIZERGUES et al.
populations had high stable Ne, which was not the case for some
rural populations, which disproportionately showed decreasing ef-
fective population sizes. We also detected low genetic structure
among urbanandruralpopulations and highadmixture, consistent
with frequent gene flow both within and among habitats. Gene flow
appears to be more prevalent among urban than rural populations,
leadingtogreatergenetichomogenisationinurbanhabitats.Ourre-
sults suggest that white clover demography aligns with the urban fa-
cilitation model more closely than the urban fragmentation model, in
that white clover thrives in urban environments and maintains high
evolutionary potential of populations.
4.1 | Effects of urbanisation on non- adaptive
evolutionary processes within populations
Urbanisation frequently leads to habitat fragmentation, which often
causes a decrease in effec tive population size and genetic diver-
sity within populations (i.e.the Urban Fragmentation Model). This
prediction has been supported by a broad diversity of studies. For
example,fragmentedforesthabitat sleadtoadecreaseingeneticdi-
versity in red- backed salamanders (Plethodon cinereus) in the city of
Montreal(Noëletal.,2007).Similarly,whitefootedmice(Peromyscus
leucopus) showed reduced diversity along an urbanisation gradient in
NewYork(Munshi-Southetal.,2016).Ourresultsdonotmatchthis
expectedpatternoftheurbanfragmentationmodel.
Consistent with the urban facilitation model, we found that
urban white clover was able to maintain high levels of genetic di-
versity and high ef fective population size throughout the sampling
rangeofourstudy.Ourresultsofhighgeneticdiversityinurbanand
rural populations are consistent with previous smaller scale stud-
ies of white clover that used microsatellite markers, which showed
that genetic diversity did not vary along urbanisation gradients in
multiplecitiesinOntario,Canada(Johnsonetal.,2018). These find-
ings may alsohelp explain how whiteclover hasbeen abletorap-
idly adapt to repeated urbanisation gradients throughout the world
(Santangeloetal.,2022).Adaptiveevolutionisexpectedtobemore
efficient when Neishighandselectionisstrong.Weconfirmedlarge
Ne in most cities, which is consistent with the obser vation that white
clover is frequently abundant and widespread throughout cities.
A novel result of the current study is that urban habitats are more
likely to maintain high and stable population sizes than rural areas.
Given that urban habitats frequently undergo rapid environmental
change, we may e xpect urb an populatio ns to experie nce stronger
selection than rural habitats.
There is mounting evidence that the urban facilitation model,
supported by our study, may apply to many species. For instance,
brown rats display high genetic diversit y in urban areas that can be
attributed to large population sizes (Combs et al., 2018). Similarly,
urban great tits (Parus major) showed higher genetic diversity in cit-
iesthanintheassociatedforests(Björklundetal.,2009).Inabroad
meta-analysis,Schmidtetal.(2020) suggested that human commen-
sal species could be more likely to display higher diversity in urban
compared to rural areas, especially in birds that are not affected by
habitatfragmentation(butnotnecessarilytrueformammals).White
clover has many traits (e.g. clonal growth, prostrate growth habit ,
ability to regrow following cut ting, polyploidy) that make it well-
adapted to human- dominated landscapes like lawns and well- grazed
pastures (Burdon, 1983;VanDrunen&Johnson,2022).Inparticular,
clover is a low- growing plant making it a poor competitor for sun-
light, but it has the ability to quickly regrow following damage, and
might thus benefit from mowing. Urban areas display a multitude of
well- mowed green spaces (parks, lawns, private backyards) where
clovercangrow.Inaddition,cloverissensitivetodroughtandbene-
fits from irrigation provided to many urban spaces. By contrast, rural
areas are frequently unmowed and unwatered, which can lead to
FIGURE 4 GeneticstructureamongT. repens populations revealed by a principal component analysis (PCA) on genotype likelihoods of
41,543 four- fold positions, (a) on PC1 and PC2, and (b) on PC1 and PC3. Each point represents the centroid of rural (green round) and urban
(purple triangle) populations for each city sampled.
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17311 by Test, Wiley Online Library on [12/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
|
9 of 13
CAIZERGUES et al.
cloverbeingoutcompetedandexperiencelowfitness.Finally,urban
white clover's abilit y to maintain high population sizes compared to
rural ones might also benefit from frequent introductions of clover
to cities (via seeds in sod and turfgrass) from diverse seed stocks,
leading to elevated diversity and low genetic structure. Hence, both
the life- history trait s of clover and its close link with intense human
activities make it a species well adapted to performing best in an-
thropogenic habitats.
FIGURE 5 AdmixtureproportionsbetweenurbanandruralhabitatsforeachcityestimatedwithNGSadmix. The best supported K value
for each cit y was selected with the method of Evanno et al. (2005). Each vertical bar represents an individual and the height of the colour
indicates the proportion of an individual's genome derived from each of the Kancestralpopulationswithinagivencity.Individualsare
grouped by sub- population.
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17311 by Test, Wiley Online Library on [12/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
10 of 13
|
CAIZERGUES et al.
4.2 | Effects of urbanisation on genetic
structure and gene flow between populations
Consistent with the urban facilitation model, we found that white
clover displays high gene flow between urban and rural areas, as
suggested by low levels of differentiation, limited genetic structure,
high admi xture propor tions and high es timates of disper sal. High
levels of gene flow between populations are consistent with high
genetic diversity and large population sizes, as connectivity between
populations enhances allele movement and reduces genetic drift. A
major element of white clover history is that it was often intention-
ally planted by humans as a fodder crop, to enrich soil nitrogen in
rotation farming, unintentionally introduced via sod or turfgrass, or
intentionallyplantedascloverlawns(Kjærgaard,2003). Thus, white
clover was potentially introduced multiple times in each area and
moved among locations by people. Such repeated introductions
have likely played a major role in the low genetic structure obser ved
betweenhabitatsandevidenceofsubstantialadmixture.
While urbanandruralcloverdisplay similarlow levelsof relat-
edness, we found the lowest dif ferentiation among urban popula-
tions (i.e. lower within urban FST than within rural FST), suggesting
the urban landscape facilitates gene flow. As an obligate outcrosser,
white clover depends on pollinators for it s reproduction. Cities
can harbour more pollinator diversity than rural areas (Wenzel
et al., 2020), and since movement of pollinators is not necessarily
constrained by the urban landscape (Theodorou et al., 2018), higher
gene flow within cities might be driven by pollinators. For instance,
in Toronto, Canada, previous research showed that pollinator vis-
itation rate to white clover was higher in urban sites than in rural
sites (Santangelo, Rivkin,etal., 2020). Additionally, clover is often
moved around by humans in urban areas, for instance via turfgrass,
which can in turn increase gene flow and reduce genetic differenti-
ation between urban subpopulations. This novel result of high gene
flow within urban versus within rural habitat s, also suggest s that
any adaptations that do arise (e.g. HCN production, Santangelo
et al., 2022) are likely to spread more quickly among urban than rural
populations.
Recent literature suggests that several species mayexhibitex-
tensive gene flow in urban areas. For instance, highly mobile spe-
cies such as the feral pigeons (Columba livia) can display high gene
flowin urbanareas (Carlen & Munshi-South, 2021).Insmallmam-
mals, Combs et al. (2018) found high genetic diversit y in brown rats
(Rattus novegicus) in four cities, likely due to high gene flow and large
population sizes. However, high gene flow in urban habitats is not
limited to hi ghly mobile spec ies. In fact, h uman commensal s that
are not necessarily good dispersers, like the western black widows
(Latrodectus hesperus), can directly benefit from human movement
andtransportationnetworksfortheir dispersal (Milesetal.,2018).
As a result, species that are intentionally or unintentionally moved
byhumansarelikelytodisplayhigherlevelsofgeneflow.Ourresults
on white clover contribute to the growing evidence that urbanisa-
tion can facilitate gene flow and reveal that such processes can be
repeated across cities worldwide.
4.3 | Future directions for the study of parallel
genomic evolution in cities
Cities offer a great opportunity to study how anthropogenic dis-
turban ces affect evo lutionar y processes a t a global scal e. In line
with our study, there has been a recent focus studying whether
similar environmental changes caused by urbanisation lead to in-
dependent, parallel evolutionary trajectories. While evidence of
repeated phenotypic shifts in urban areas is growing, they do not
necessarily arise from parallel independent adaptive evolution. For
example,acommongardenexperimentusingtheVirginiapepper-
weed (Lepidium virginicum) showed genetically based convergence
in the urban phenot ype of earlier bolting, larger size, producing
fewer leaves and more seeds. However, population genetic analy-
ses revealed that urban populations were likely derived from the
same inbred haplotype, demonstrating that such parallelism in
urbanphenotypewasfacilitatedbyextensivegeneflowamongcit-
ies combined with lineage sor ting within habitats, as opposed to
multipleindependent evolutionary events(Yakub & Tiffin,2017 ).
Similarly, in the Gulf killifish (Fundulus grandis), adaptive resist-
ance to industrial pollutants was found in four populations, which
was the result of de novo mutation in only one of the populations,
whereas the other three populations likely evolved from stand-
inggeneticvariationorintrogression(Ozioloretal.,2019).Inboth
cases, the adaptive parallelism was facilitated by demographic
events, highlightingthe need for combined analyses thatexplore
both adaptive and non- adaptive evolution to investigate how ur-
banisation shapes evolutionary trajectories. The parallelism of non-
adaptive processes has been particularly ignored in studies, even
though theory suggests that genetic drift and gene flow can lead to
parallelevolutionarypatterns(e.g.Santangeloetal.,2018).Ourre-
sults suggest that contrary to the prevailing hypothesis, cities can
facilitate the demographic spread, growth and adaptation of spe-
cies, especially for those with a cosmopolitan distribution.
AUTHOR CONTRIBUTIONS
AEC, JSS,RWNand MTJJconceptualisedthe study.JSSand MTJJ
developedlab protocols. RWN,JSS developed analytical methods.
AEC carried out the analysis. All authors collected samples and con-
tributed to the writing of the manuscript.
AFFILIATIONS
1CentreforUrbanEnvironments,UniversityofTorontoMississauga,
Mississauga,Ontario,Canada
2DepartmentofBiology,UniversityofTorontoMississauga,Mississauga,
Ontario,Canada
3DepartmentofIntegrativeBiology,UniversityofCaliforniaBerkeley,
Berkeley,California,USA
4ProgramadePós-GraduaçãoemGestãoeTecnologiaAmbientalda
UniversidadeFederaldeRondonópolis,Rondonópolis,Brasil
5DepartmentofPlantBiology,DepartmentofEntomology,PlantResilience
Institute,MichiganStateUniversity,EastLansing,Michigan,USA
6GenomicSciencesandTechnologyProgram,UniversityofBritishColumbia,
Vancouver,BritishColumbia,Canada
7DepartmentofMicrobiologyandImmunology,UniversityofBritish
Columbia,Vancouver,BritishColumbia,Canada
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17311 by Test, Wiley Online Library on [12/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
|
11 of 13
CAIZERGUES et al.
8ReddeEcoetología,InstitutodeEcologíaA.C,Xalapa,Mexico
9LivingEarthCollaborative,Washing tonUniversit yinSt.Louis,St.Louis,
Missouri,USA
10UniversidadSanFranciscodeQuito,Ecuador,Quito
11SanFranciscoStateUniversity,SanFrancisco,California,USA
12DepartmentofBiology,UniversityofMassachusettsBoston,Boston,
Massachusetts,USA
13Dendrolytics,Seattle,Washington,USA
14NTNUUniversityMuseum,NorwegianUniversityofScienceand
Technology, Trondheim, Norway
15SchoolofBiologicalSciences,UniversityofReading,Reading,UK
16SchoolofBiologicalSciences,MonashUniversity,Melbourne,Victoria,
Australia
17Depar tmentofBiologicalSciences,WayneStateUniversity,Detroit,
Michigan,USA
18LaboratoriodeEcologíaTropicalyServiciosEcosistémicos(EcoSs-Lab),
UniversidadTécnicaParticulardeLoja,Loja,Ecuador
19UniversityofLouisiana,Lafayette,Louisiana,USA
20BiodiversityResearchInstitute(IMIB),CSIC-UniversityofOviedo-
PrincipalityofAsturias,Mieres,Spain
21Evolution&EcologyResearchCentre,UNSW-Universit yofNewSouth
Wales,Sydney,NewSouthWales,Australia
22DepartmentofBiologyandLouisCalderCenter,FordhamUniversit y,New
YorkCity,NewYork,USA
23BotanischerGartenundBotanischesMuseumBerlin,FreieUniversität
Berlin, Berlin, Germany
24Institutdebiologieintégrativeetdessystèmes,UniversitéL aval,Quebec
City,Quebec,Canada
25DepartmentofEcology,EvolutionandBehavior,UniversityofMinnesota,
St.Paul,Minnesota,USA
26Science,Technology&SocietyDepartment,RochesterInstituteof
Technology,Rochester,NewYork,USA
27Depar tmentofEcology,EnvironmentandPlantSciences,Stockholm
University,Stockholm,Sweden
ACKNOWLEDGEMENTS
WewouldliketothankSimonInnes,SophieKoch,InderSheoranand
BeataCohanforperforminglabwork,DNAextractionsandmaking
the genomic libraries, all the GLUE collaborators for collecting clover
around the world and the two anonymous reviewers for their useful
comments.ThisresearchwasfundedbyCRC,SchoolofCitiesgrant
to AEC, NSERC D iscovery grant s to RWN and MTJJ, and NSERC
EWRSteacieAwardtoMTJJ.
DATA AVAIL AB ILI T Y S TAT EME NT
All code is available on h t t p s : / / g i t h u b . c o m / A u d e C a i z e r g u e s / g l u e _
demog raphy .BA M files are avai lable on the Euro pean Nucleot ide
Archive(ENABioProjectsPR JEB48967&PRJEB72257).
ORCID
Aude E. Caizergues https://orcid.org/0000-0003-4467-3912
James S. Santangelo https://orcid.org/0000-0002-5921-2548
Fernanda Baena- Diaz https://orcid.org/0000-0002-0660-0627
Carlos Iñiguez- Armijos https://orcid.org/0000-0002-9787-3451
Jason Munshi- South https://orcid.org/0000-0002-8067-4341
REFERENCES
Beninde,J.,Feldmeier,S., Veith, M., &Hochkirch, A.(2018).Admixture
of hybrid swarms of native and introduced lizards in cities is deter-
mined by the cityscape structure and invasion history. Proceedings
of the Royal Society B, 285(1883), 20180143. ht tps:// doi. org/ 10.
1 0 9 8 / r s p b . 2 0 1 8 . 0 1 4 3
Björklund,M.,Ruiz,I.,&Senar,J.C.(2009).Geneticdifferentiationinthe
urban habitat: The great tits (Parus major) of the parks of Barcelona
cit y. Biological Journal of the Linnean Society, 99(1), 9–19. ht t p s :// d o i .
o r g / 1 0 . 1 1 1 1 / j . 1 0 9 5 - 8 3 1 2 . 2 0 0 9 . 0 1 3 3 5 . x
Burdon,J.J.(1983).Trifolium Repens L. Journal of Ecology, 71(1), 307–330.
htt ps:// doi. org/ 10. 2307/ 2259979
Caizergues, A . E., Le Luyer, J., Grégoire, A., Szulkin , M., Senar, J.-C.,
Charmantier,A.,&Perrier,C.(2022).Epigeneticsandthecity:Non-
parallelDNAmethylationmodificationsacrosspairsofurban-forest
great tit populations. Evolutionary Applications, 15( 1 ), 1 4 9–1 65.
Carlen,E.,&Munshi-South, J.(2021).Widespread geneticconnectivity
of feral pigeons across the northeastern megacity. Evolutionary
Applications, 14(1), 150–162. h t t p s : / /d o i . o r g / 1 0 . 1 1 1 1 / E V A . 1 2 9 7 2
Combs, M.,Puckett,E.E., Richardson,J.,Mims, D., & Munshi-South, J.
(2018). Spatial populationgenomics of the brownrat (Rattus nor-
vegicus) in New York City. Molecular Ecology, 27(1),83–98.h tt p s : //
doi . org/ 10. 1111/ mec. 14 437
Evanno, G.,Regnaut, S., & Goudet,J.(2005).Detecting thenumberof
clusters of individuals using the software structure: A simulation
stud y. Molecular Ecology, 14 (8), 2611–2620. https:// doi. org/ 10.
1 1 1 1 / j . 1 3 6 5 - 2 9 4 X . 2 0 0 5 . 0 2 5 5 3 . x
Fox, E. A ., Wright, A . E., Fumagalli, M ., & Vieira, F. G. (20 19) . ngsLD:
Evaluating linkage disequilibrium using genotype likelihoods.
Bioinformatics, 35(19), 3855–3856. https:// doi. org/ 10. 1093/ bioin
forma tics/ btz200
Glenn,T.C.,Nilsen,R.A .,Kieran,T.J.,Sanders, J.G., Bayona-Vásquez,
N. J., Finger, J. W., Pierson, T. W., Bentley, K. E., Hoffberg, S. L.,
Louha, S ., Garcia-De Le on, F.J. , Del Rio Por tilla, M. A ., Reed, K .
D.,Anderson,J.L.,Meece,J.K.,Aggrey,S.E.,Rekaya,R.,Alabady,
M.,Belanger,M.,…Faircloth,B.C.(2019).AdapteramaI:Universal
stubs and primers for 384 unique dual-indexed or 147,456
combinatorially-indexed Illumina libraries (iTru &iNex t).P ee rJ, 7,
e7755. https:// doi. org/ 10. 7717/ peerj. 7755
Griffiths,A.G.,Moraga,R.,Tausen,M.,Gupta,V.,Bilton,T.P.,Campbell,
M. A., Ashby, R., Nagy, I., Khan, A., Larking, A., Anderson, C.,
Franzmayr,B.,Hancock,K.,Scott,A.,Ellison,N.W.,Cox,M.P.,Asp,
T.,Mailund,T.,Schierup,M.H.,&Andersen,S.U.(2019).Breaking
free:Thegenomicsofallopolyploidy-facilitatednicheexpansionin
white clover. The Plant Cell, 31(7 ), 14 66–1487. https:// doi. org/ 10.
1 1 0 5 / T P C . 1 8 . 0 0 6 0 6
Grimm,N.B.,Faeth,S.H.,Golubiewski,N.E.,Redman,C.L.,Wu,J.,Bai,
X., & Bri ggs, J. M. (2 008). Gl obal ghange a nd the ecolo gy of cit-
ies. Science, 319(5864), 756–760.https:// doi. org/ 10. 1126/ scien ce.
1150195
Gutenkunst,R.,Hernandez,R.,Williamson,S.,&Bustamante,C.(2010).
Diffusionapproximationsfordemographicinference:DaDi.Nature
Precedings. h t t p s : / / d o i . o r g / 1 0 . 1 0 3 8 / n p r e . 2 0 1 0 . 4 5 9 4 . 1
Han, E., Sinsheimer,J. S., & Novembre, J.(2014).Characterizingbias in
population genetic inferences from low- coverage sequencing data.
Molecular Biolog y and Evolution, 31(3), 723–735. https:// doi. org/ 10.
1 0 9 3 / M O L B E V / M S T 2 2 9
Hanghøj, K.,Moltke,I., Andersen, P.A., Manica,A.,& Korneliussen, T.S.
(2019). Fast and accurate relatedness estimation from high- t hroughput
sequencing data in the presence of inbreeding. GigaScience, 8(5),
giz034. https:// doi. org/ 10. 1093/ gigas cience/ giz034
Hanski,I.(1998).Metapopulation dynamics. Nature, 396(6706), 41–49.
h t t p s : / / d o i . o r g / 1 0 . 1 0 3 8 / 2 3 8 7 6
Hedrick,P.W.,&Lacy,R.C.(2015).Measuringrelatednessbetweenin-
bred individuals. Journal of Heredity, 10 6(1), 20–25. https:// doi. org/
10. 1093/ jhered/ esu 072
Hudson,R.R.,Slatkin,M.,&Maddison,W.P.(1992).Estimationoflevels
ofgeneflowfromDNAsequencedata.Genetics, 132(2),583–589.
h t t p s : / / d o i . o r g / 1 0 . 1 0 9 3 / g e n e t i c s / 1 3 2 . 2 . 5 8 3
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17311 by Test, Wiley Online Library on [12/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
12 of 13
|
CAIZERGUES et al.
Johnson,M. T.J., & Munshi-South, J. (2017).Evolutionoflife in urban
environments. Science, 358(6363), eaam8327. https:// doi. org/ 10.
1 1 2 6 / s c i e n c e . a a m 8 3 2 7
Johnson, M. T.J., Prashad, C. M., Lavoignat, M., & Saini, H. S. (2018).
Contrasting the effects of natural selection, genetic drift and
gene flow on urban evolution in white clover (Trifolium repens).
Proceedi ngs of the Royal Society B: Biological Sciences, 285(1883),
20181019.h t t p s : / / d o i . o r g / 1 0 . 1 0 9 8 / r s p b . 2 0 1 8 . 1 0 1 9
Kakes, P.(1997).Differencebetweenthemaleandfemale components
of fitness associated with the gene Ac in Trifolium repens. Acta
Botanica Neerlandica, 46(2), 219–223. ht tps:// doi . org/ 10. 1111 / plb.
1997. 46. 2. 219
Kjærgaard, T.(2003).Aplant thatchangedthe world:The riseandfall
of clover 1000- 2000. Landscape Research, 28(1), 41–49. ht t ps : // d o i .
o r g / 1 0 . 1 0 8 0 / 0 1 4 2 6 3 9 0 3 0 6 5 3 1
Kopelman,N.M.,Mayzel,J.,Jakobsson,M.,Rosenberg,N.A.,&Mayrose,
I.(2015).Clumpak:Aprogramforidentifyingclusteringmodesand
packaging population structure inferences across K. Molecular
Ecology Resources, 15(5), 1179–1191. h ttps :// doi. or g/ 10. 1111/
1 7 5 5 - 0 9 9 8 . 1 2 3 8 7
Korneliussen, T. S., Albrechtsen, A ., & Nielsen, R. (2014). ANGSD:
Analysisof nextgeneration sequencing data. BMC Bioinformatics,
15(1), 356. h t t p s : / / d o i . o r g / 1 0 . 1 1 8 6 / s 1 2 8 5 9 - 0 1 4 - 0 3 5 6 - 4
Lambert,M.R.,Brans,K.I.,DesRoches,S.,Donihue,C.M.,&Diamond,
S.E.(2021).Adaptiveevolutionincities:Progressandmisconcep-
tions. Trends in Ecolog y & Evolution, 36(3), 239–257. https:// doi. org/
10. 1016/j. tree. 2020. 11. 002
Liu, X.,Huang,Y., Xu, X.,Li, X.,Li, X.,Ciais, P.,Lin, P.,Gong, K ., Ziegler,
A. D., Chen,A., Gong, P.,Chen, J., Hu, G.,Chen,Y.,Wang,S.,Wu,
Q., Huang , K., Estes, L ., & Zeng, Z. (2020 ). High-spat iotemporal-
resolution mapping of global urban change from 1985 to 2015.
Nature Sustainability, 3(7), 564–570. h t t p s : // d o i . o r g / 1 0 . 1 0 3 8 / s 4 1 8 9
3 - 0 2 0 - 0 5 2 1 - x
Lou,R.N.,Jacobs,A.,Wilder,A.P.,&Therkildsen,N.O.(2021).Abegin-
ner's guide to low- coverage whole genome sequencing for popula-
tion genomics. Molecular Ecology, 30(23), 5966–5993 . h t t p s : // d o i .
org / 10. 1111/ me c. 1607 7
Lourenço,A.,Álvarez,D.,Wang,I.J., &Velo-Antón,G.(2017).Trapped
within t he city: Inte grating de mography, time s ince isolati on and
population- specific traits to assess the genetic ef fects of urbaniza-
tion. Molecular Ecology, 26(6),1498–1514.htt ps:// doi. org/ 10 . 1111/
mec. 14 019
Lynch,M.,Haubold,B.,Pfaffelhuber,P.,&Maruki,T.(2020).Inferenceof
historical population- size changes with allele- frequency data. G3:
Genes, Genomes, Genetics, 10(1), 211–223 . https:// doi. org/ 10. 1534/
g 3 . 1 1 9 . 4 0 0 8 5 4
McKinney,M.L.(2006).Urbanizationasamajorcauseofbiotichomog-
enization. Biological Conservation, 127(3), 247–260. https:// doi. org/
10. 1016/j. biocon. 2005. 09. 005
Medina, I., Cooke, G. M., & Ord, T. J. (2018). Walk, swim or fly?
Locomotor mode predicts genetic differentiation in vertebrates.
Ecology Letters, 21(5), 638–645. https:// doi . org/ 10. 1111/ ele .
12930
Meisner,J.,&Albrechtsen,A.(2018).Inferringpopulationstructureand
admixture proport ions in low-depth NGS data. Genetics, 210(2),
71 9–7 31. h t t p s : / / d o i . o r g / 1 0 . 1 5 3 4 / g e n e t i c s . 1 1 8 . 3 0 1 3 3 6
Miles,L.S.,Dyer,R.J.,&Verrelli,B.C.(2018).Urbanhubsofconnectiv-
ity: Contrasting pat terns of gene flow within and among cities in
the western black widow spider. Proceedings of the Royal Society B,
285(1884),20181224.h t t p s : / / d o i . o r g / 1 0 . 1 0 9 8 / R S P B . 2 0 1 8 . 1 2 2 4
Miles, L.S.,Rivkin,L.R.,Johnson,M.T.J.,Munshi-South,J.,&Verrelli,
B. C. (2019). Gene flow and genetic drift in urban environments.
Molecular Ecology, 28 (18), 4138–4151. http s:// doi. o rg / 10. 1111/
mec . 15221
Mölder,F.,Jablonski,K.P.,Letcher,B.,Hall,M.B.,Tomkins-Tinch,C.H.,
Sochat , V.,For ster, J., Lee, S ., Twardziok, S. O., K anitz, A ., Wilm,
A., Holtgrewe, M.,Rahmann, S., Nahnsen, S., & Köster,J. (2021).
SustainabledataanalysiswithSnakemake.F1000Research, 10, 33.
h t t p s : / / d o i . o r g / 1 0 . 1 2 6 8 8 / f 1 0 0 0 r e s e a r c h . 2 9 0 3 2 . 2
Mueller,J.C.,Kuhl,H.,Boerno,S.,Tella,J.L.,Carrete,M.,&Kempenaers,
B.(2018). Evolution of genomic variation inthe burrowing owl in
response to recent colonization of urban areas. Proceedings of the
Royal Societ y B, 285 (1878), 20180206. h t t p s : / / d o i . o r g / 1 0 . 1 0 9 8 /
rspb.2018.0206
Munshi-South,J.,Zolnik,C.P.,&Harris,S.E.(2016).Populationgenom-
ics of the Anthropocene: Urbanization is negatively associated
with genome- wide variation in white- footed mouse populations.
Evolutionary Applications, 9(4), 546–564. ht tp s:// doi . o rg / 10. 1111 /
eva. 12357
Nei,M.(1975).Molecularpopulationgeneticsandevolution.Frontiers of
Biology, 40,I–288.
Nielsen,R., Korneliussen, T., Albrechtsen, A ., Li,Y.,&Wang, J. (2012).
SNPcalling,genotypecalling,andsampleallelefrequencyestima-
tion from new- generation sequencing data. PLoS One, 7(7),e37558.
h t t p s : / /d o i . o r g / 1 0 . 1 3 7 1 / j o u r n a l . p o n e . 0 0 3 7 5 5 8
Noël,S.,Ouellet,M.,Galois,P.,&Lapointe,F.-J.(20 07).Impactofurban
fragmentation on the genetic structure of the eastern red- backed
salamander. Conservation Genetics, 8(3), 599–606. https:// doi. org/
10. 1007/ s1059 2- 006- 9202- 1
Noskova, E.,Ulyantsev, V., Koepfli, K.-P.,O'Brien, S. J.,& Dobrynin, P.
(2020).GADMA:Geneticalgorithmforinferringdemographichis-
tory of multiple populations from allele frequency spectrum data.
GigaScience, 9(3), giaa005. https:// doi. org/ 10. 1093/ gigas cience/
giaa005
Oziolor, E. M.,Reid, N. M.,Yair,S., Lee, K. M.,Guberman VerPloeg, S.,
Bruns, P.C.,Shaw, J. R ., Whitehead, A., & Matson, C. W. (2019).
Adaptiveintrogressionenablesevolutionaryrescuefromextreme
environmental pollution. Science, 364(6439), 455–457.
Rochat, E., Manel, S., Deschamps-Cottin, M., Widmer, I., & Joost, S.
(2017). Persistence of butterfly populations in fragmented habi-
tatsalongurbandensitygradients:Motilityhelps.Heredity, 119(5),
328–338.h t t p s : / /d o i . o r g / 1 0 . 1 0 3 8 / h d y . 2 0 1 7 . 4 0
Salmón,P.,Jacobs,A.,Ahrén,D.,Biard,C.,Dingemanse,N.J.,Dominoni,
D.M.,Helm,B.,Lundberg,M.,Senar,J.C.,Sprau,P.,Visser,M.E.,&
Isaksson,C .(2021).Repeatedgenomic signaturesofadaptationto
urbanis ation in a songbird across Europe. Nature Communications,
12(1), 1–14. h t t p s : // d o i . o r g / 1 0 . 1 0 3 8 / s 4 1 4 6 7 - 0 2 1 - 2 3 0 2 7 - w
Santangelo, J.S., Miles,L. S., Breitbart,S.T.,Murray-Stoker,D., Rivkin,
L.R.,Johnson,M. T.J.,&Ness,R.W.(2020).Urbanenvironments
asaframeworktostudyparallelevolution.InM.Szulkin,J.Munshi-
South,&A.Charmantier(Eds.),Urban evolutionary biology (pp. 36–
53).OxfordUniversityPress.
Santangelo, J.S., Ness,R.W.,Cohan, B., Fitzpatrick, C.R.,Innes,S.G.,
Koch, S., M iles, L. S., Mun im, S., Peres-Neto, P. R., Pras had, C.,
Tong,A.T.,A guirre,W.E.,Akinwole,P.O.,Alberti,M.,Álvarez,J.,
Anderson,J.T.,Anderson,J.J.,Ando,Y.,Andrew,N.R.,…Johnson,
M. T. J. (2022). Gl obal urban e nvironment al change dri ves adap-
tation in white clover. Science, 375(6586),1275–1281.h t t p s : // d o i .
o r g / 1 0 . 11 2 6 / s c i e n c e . a b k 0 9 8 9
Santangelo,J.S.,Rivkin,L.R.,Advenard,C.,&Thompson,K.A.(2020).
Multivariate phenotypicdivergence along an urbanization gradi-
ent. Biology Letters, 16(9), 20200511. h t t p s : / /d o i . o r g / 1 0 . 1 0 9 8 / r s b l .
2020. 0511
Santangelo,J.S.,Rivkin,L.R.,&Johnson,M.T.J.(2018).Theevolution
of city life. Proceedings of the Royal Society B: Biological Sciences,
285(1884),20181529.h t t p s : // d o i . o r g / 1 0 . 1 0 9 8 / r s p b . 2 0 1 8 . 1 5 2 9
Schmidt,C.,Domaratzki,M.,Kinnunen,R.P.,Bowman,J.,&Garroway,C.
J.(2020).Continent-wideeffectsofurbanizationonbirdandmam-
mal genetic diversity. Proceeding s of the Royal Society B, 287(1920) ,
20192497. h t t p s : / / d o i . o r g / 1 0 . 1 0 9 8 / r s p b . 2 0 1 9 . 2 4 9 7
Skotte, L., Korneliussen, T. S., &Albrechtsen,A. (2013).Estimating in-
dividualadmixture proportionsfromnextgenerationsequencing
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17311 by Test, Wiley Online Library on [12/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
|
13 of 13
CAIZERGUES et al.
data. Genetics, 195 (3), 693–702. https:// doi. org/ 10. 1534/ genet ics.
113.154138
Somers,C.M.,McCarr y,B.E.,Malek,F.,&Quinn,J.S.(2004).Reduction
of particulate air pollution lowers the risk of heritable mut ations
in mice. Science, 304(5673) , 1008 –1010. https:// doi. org/ 10. 1126/
science.1095815
Tajima,F.(1983). Evolutionary relationship of DNA sequencesin finite
populations. Genetics, 105(2), 437–460. ht tp s:// doi. org / 10. 1093/
genet ics/ 105.2. 437111/ eva. 13334
Tajima,F.(1989).TheeffectofchangeinpopulationsizeonDNApoly-
morphism. Genetics, 123(3), 597–601. https:// doi. org/ 10 . 1093/
genet ics/ 123.3. 597
Theodorou,P.,Radzevičiūtė,R.,Kahnt,B.,Soro,A.,Grosse,I.,&Paxton,
R. J. (2018). G enome-w ide single nucl eotide polym orphism scan
suggests adaptation to urbanization in an important pollinator, the
red- tailed bumblebee (Bombus lapidarius L.). Pro ceedings of the Royal
Society B: Biological Sciences, 285(1877),20172806.https:// doi. org/
1 0 . 1 0 9 8 / r s p b . 2 0 1 7 . 2 8 0 6
VanDrunen,W.E.,&Johnson,M.T.J.(2022).Polyploidyinurbanenvi-
ronments. Trends in Ecology & Evolution, 37(6), 507–516. ht t p s : // d o i .
org/ 10. 1016/j. tree. 2022. 02. 005
Wenzel,A.,Grass,I.,Belavadi,V.V.,&Tscharntke,T.(2020).Howurban-
ization is driving pollinator diversity and pollination – A systematic
revi ew. Biological Conservation, 241, 108321. https:// doi. org/ 10.
1 0 1 6 / j . b i o c o n . 2 0 1 9 . 1 0 8 3 2 1
Whitlo ck, M. C., & M cCauley, D. E. (1999). Indir ect measu res of gene
flowandmigration:FSTnotequalto 1/(4Nm+ 1). Heredity, 82(Pt
2), 117–125. h t t p s : // d o i . o r g / 1 0 . 1 0 3 8 / s j . h d y . 6 8 8 4 9 6 0
Williams,W.M.,Mason,K.M.,&Williamson,M.L.(1998).Geneticanal-
ysis of shikimate dehydrogenase allozymes in Trifolium repens L.
Theoretical and Applied Genetics, 96,859–868.
Wright, S . (1931). Evo lution in mend elian populat ions. Genetics, 16(2),
97–1 59. https:// doi. org/ 10. 1093/ genet ics/ 16.2. 97
Wright,S.(1984).Evolutionandthegeneticsofpopulations.InVariability
Within and Among Natural Populations(Vol. 4). Universit y of Chicago
Press.
Wu,F.,Ma, S., Zhou, J., Han, C.,Hu, R., Yang, X., Nie,G., & Zhang, X .
(2021). Genetic diversity and population structure analysis in a
large collection of white clover (Trifolium repens L.) germplasm
worldwide. PeerJ, 9, e11325. https:// doi. org/ 10. 7717/ peerj. 11325
Yakub, M., & Ti ffin, P. (2017). Living i n the city: Ur ban environme nts
shape the evolution of a native annual plant. Global Change Biology,
23(5),2082–2089.h t t p s : / / d o i . o r g / 1 0 . 1 1 1 1 / g c b . 1 3 5 2 8
Yauk, C., Po lyzos, A ., Rowan-Carr oll, A., S omers, C . M., Godsc halk, R.
W.,Van Schooten, F.J.,Berndt, M. L., Pogribny,I. P., Koturbash,
I.,Williams, A.,Douglas,G.R., & Kovalchuk, O. (2008). Germ-line
mutations,DNAdamage,andglobalhypermethylationinmiceex-
posed to particulate air pollution in an urban/industrial location.
Proceedi ngs of the National Academy of Sciences, 105(2), 605–610.
h t t p s : / /d o i . o r g / 1 0 . 1 0 7 3 / p n a s . 0 7 0 5 8 9 6 1 0 5
SUPPORTING INFORMATION
Additional supporting information can be found online in the
SupportingInformationsectionattheendofthisarticle.
How to cite this article: Caizergues,A.E.,Santangelo,J.S.,
Ness,R.W.,Angeoletto,F.,Anstett,D.N.,Anstett,J.,
Baena-Diaz,F.,Carlen,E.J.,Chaves,J.A.,Comerford,M.S.,
Dyson,K.,Falahati-Anbaran,M.,Fellowes,M.D.E.,Hodgins,
K.A.,Hood,G.R.,Iñiguez-Armijos,C.,Kooyers,N.J.,
Lázaro-Lobo,A.,Moles,A .T.…Johnson,M.T.J.(2024).
Doesurbanisationleadtoparalleldemographicshiftsacross
the world in a cosmopolitan plant? Molecular Ecology, 00,
e17311. ht tp s://doi.org/10 .1111/m ec .17311
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17311 by Test, Wiley Online Library on [12/03/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
