Content uploaded by Yuanzhen Liu
Author content
All content in this area was uploaded by Yuanzhen Liu on Dec 21, 2023
Content may be subject to copyright.
Molecular Ecology. 2023;00:e17251.
|
1 of 18
https://doi.org/10.1111/mec.17251
wileyonlinelibrary.com/journal/mec
Received:4October2023
|
Revised:24November2023
|
Accepted:11December2023
DOI :10.1111/mec.17251
ORIGINAL ARTICLE
Genomic variation in montane bumblebees in Scandinavia:
High levels of intraspecific diversity despite population
vulnerability
Yuanzhen Liu1 | Anna Olsson1 | Tuuli Larva1 | Aoife Cantwell- Jones2 |
Richard J. Gill2 | Björn Cederberg3 | Matthew T. Webster1
This is an open access article under the terms of the CreativeCommonsAttribution-NonCommercial License, which permits use, distribution and reproduction
in any medium, provided the original work is properly cited and is not used for commercial purposes.
©2023TheAuthors.Molecular Ecology published by John Wiley & Sons Ltd.
1DepartmentofMedicalBiochemistryand
Microbiology,ScienceforLifeLaboratory,
Uppsala University, Uppsala, Sweden
2Depar tmentofLifeSciences,Georgina
MaceCentreforTheLivingPlanet,
ImperialCollegeLondon,Ascot,UK
3SwedishUniversityofAgricultural
Sciences,SwedishSpeciesInformation
Centre, Uppsala, Sweden
Correspondence
MatthewT.Webster,IMBIM,BMCD11:3,
Box582,UppsalaUniversity,75123
Uppsala, Sweden.
Email: matthew.webster@imbim.uu.se
Funding information
Vetenskapsrådet,Grant/AwardNumber:
2018-05973;Naturvårdsverket,Grant/
AwardNumber:225-21-003
Handling Editor:SeanD.Schoville
Abstract
Populations of many bumblebee species are declining, with distributions shift-
ing northwards to track suitable climates. Climate change is considered a major
contributingfactor.Arcticspeciesareparticularlyvulnerableastheycannot shift
furthernorth,makingassessmentoftheirpopulationviabilityimportant.Analysis
oflevelsofwhole-genome variation is a powerful way to analyse population de-
clines and fragmentation. Here, we use genome sequencing to analyse genetic vari-
ation in seven species of bumblebee from the Scandinavian mountains, including
two classified as vulnerable. We sequenced 333 samples from across the ranges
of these species in Sweden. Estimates of effective population size (NE) vary from
~55,000 for species with restricted high alpine distributions to 220,000 for more
widespreadspecies.Populationfragmentationisgenerallyveryloworundetect-
able over large distances in the mountains, suggesting an absence of barriers to
gene flow. The relatively high NE and low population structure indicate that none of
the species are at immediate risk of negative genetic effects caused by high levels
of genetic drift. However, reconstruction of historical fluctuations in NE indicates
that the arctic specialist species Bombus hyperboreushasexperiencedpopulation
declines since the last ice age and we detected one highly inbred diploid male of
this species close to the southern limit of its range, potentially indicating elevated
geneticload.Althoughthelevelsofgeneticvariationinmontanebumblebeepopu-
lations are currently relatively high, their ranges are predicted to shrink drastically
due to the effects of climate change and monitoring is essential to detect future
population declines.
KEYWORDS
bumblebees, climate change, conservation genetics, genomics/proteomics, population
dynamics, population genetics – empirical
2 of 18
|
LIU et al.
1 | INTRODUCTION
Bumblebees are important pollinators, particularly in temperate
regions of the northern hemisphere. However, declines in species
abundance and distribution have been reported globally in recent
decades (Cameron et al., 2011; Cameron & Sadd, 2020; Williams
et al., 2009).AccordingtotheIUCNRedList,nearlyone-quarterof
the bumblebee species that have been assessed are declining(De
Meulemeesteretal.,2016).Asbumblebeesaremainlycoldadapted,
they are particularly vulnerable to climate change, which is predicted
to have a major negative impact on their populations (Cameron &
Sadd, 2020). Climatic niche modelling of future suitable habitats on
the basis of multiple future climate change scenarios has shown that
the majority of species are predicted to lose a significant proportion
of their current ranges within this century. Up to 36% of species are
predic ted to lose more t han 80% of thei r current ran ge (Ghisbain
et al., 2023; Rasmont et al., 2015).
ThedistributionsofmanyNorthAmericanand European bum-
blebee species have shifted northwards in recent decades (Kerr
et al., 2015; Mar tinet et al., 2015; Soroye et al., 2020). However,
although the southern edges of their distributions have retreated
rapidly,expansionsofthenorthernedgesarereportedtobeoccur-
ring muchmore slowly.Climate change is the mostlikely explana-
tion for these changes, and in particular the increased occurrence
of extreme weather events that limit sur vival. Bumblebee popu-
lations in the arctic and alpine regions are particularly threatened
by climate change because it is impossible for them to shift their
distributions furthernorth (Kerretal.,2015; Rasmont et al., 2015;
Soroye et al., 2020; Williams & Osborne, 2009). The assessment of
the viability of arctic and mountain specialist bumblebee species is,
therefore, high priority.
Analysi s of genetic diver sity is a power ful way to assess p op-
ulation viability and inform conservation strategies (Allendorf
et al., 2010; Hohenlohe et al., 2021). Rarer species of bumblebees
inEuropeandNorthAmericahavebeenfoundtoexhibitthelower
levels of genetic diversity and the elevated levels of genetic dif-
ferentiation (Cameron et al., 2011; Charman et al., 2010; Darvill
et al., 2006; Ellis et al., 2006;Kentetal., 2018; Lozier et al., 2011;
Lozier & Cameron, 2009), which suggests that an analysis of genetic
variation could be an important way to assess conservation status of
bumblebees.Analysisofgeneticdiversityonawhole-genomescale
enables a range of more powerful analyses than traditional conser-
vation genetic analyses based on limited numbers of loci (Webster
et al., 2023). These include reconstructing population history and
quantifying levels of inbreeding and genetic load in natural pop-
ulations ( Allendor f et al., 2010; Hohenlohe et al., 2021; Supple &
Shapiro, 2018; Webster et al., 2023).
The main focus of this study is to provide an assessment of
populationhistoryandvulnerabilitybasedongenome-widevaria-
tion in bumblebee species found in mountain habitats in Sweden,
which are among the most threatened by climate change (Rasmont
et al., 2015).Wefocuson sixspeciesthatarerestrictedtoarctic
and mountain environments, and one species (Bombus pascuo-
rum) that is present in the Swedish mountains but is also wide-
spread in multiple habitats throughout Europe (Table 1; Figure 1).
Of the montane species, Bombus balteatus, Bombus hyperboreus,
Bombus pyrrhopygus and Bombus alpinus belong to the subgenus
Alpinobombus(Cameronetal.,2007; Williams et al., 2008). B. bal-
teatus is relatively common in the Swedish mountains, whereas the
other three have a more restricted range, inhabiting higher lati-
tudes and elevations (Rasmont et al., 2015; Williams et al., 2019).
Bombus lapponicus and Bombus monticola belong to the subgenus
Pyrobombus. Both of these species are found throughout the
Swedish mou ntains (Rasmont e t al., 2015). I n a d dit i o n t o t h e irran g e
in Scandinavia, two of the species are also present in mountain en-
vironments further south in Europe. B. monticola is found in high
elevationsat manylocations across Europe including the British
Isles,Pyrenees,AlpsandBalkans.B. alpinus occurs at the highest
elevationsin the Alps inadditiontothe Scandinavianmountains.
The other montane species B. lapponicus, B. pyrrhopygus, B. baltea-
tus and B. hyperboreusarenotpresentsouthofScandinavia.Allof
these species apart from B. alpinus and B. monticola are also found
further east in Siberia (Rasmont et al., 2015; Williams et al., 2019).
Incontrasttotheotherspecies,B. hyperboreus is a parasitic spe-
cies, which lays eggs in the nests of other bumblebee species
(Ødegaard et al., 2015).
TAB LE 1 Speciesofbumblebeesanalysedinthisstudy.
Bombus sp. Subgenus English Swedish
IUCN red list
(3.1)
Climate
riskaDistribution in Europe
alpinus Alpinobombus Alpinebumblebee Alphumla Vulnerable HHR EuropeanhighlandsandArctic
pyrrhopygus Alpinobombus Polarbumblebee Polarhumla Least concern HHHR Arctic
balteatus Alpinobombus Golden-beltedbumblebee Fjällhumla Least concern HHR ScandinavianhighlandsandArctic
hyperboreus Alpinobombus High-Arcticbumblebee Tundrahumla Vulnerable HHHR Arctic
lapponicus Pyrobombus Lapland bumblebee Lapphumla Least concern HHR ScandinavianhighlandsandArctic
monticola Pyrobombus Mountainbumblebee Berghumla Least concern HHR EuropeanhighlandsandArctic
pascuorum Thoracobombus Common carder bee Åkerhumla Least concern RCommon
aClimateriskcategories:HHHR,extremelyhighclimatechangerisk,lossof>95%ofrangeby2100;HHR,veryhighclimatechangerisk,lossof>85%
of range by 2100, HR, high climate change risk, loss of >70% of range by 2100, R, climate change risk, loss of >50% of range by 2100, LR, lower
climate change risk, <50% loss of range by 2100 (Rasmont et al., 2015).
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17251 by Yuanzhen Liu - New Aarhus University , Wiley Online Library on [21/12/2023]. 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 18
LIU et al.
Allofthese montane bumblebeespecies arethreatenedbycli-
mate change. Those that are judged to be most at risk are the ones
restricted to the highest elevations and latitudes: B. pyrrhopygus, B.
alpinus and B. hyperboreus. Of these, B. alpinus and B. hyperboreus
are classified as vulnerable in the species red list whereas all the
other species studied here are considered least concern (Table 1)
(IUCN,2022). However, these risk assessments have been shown
to be unreflective of projected population trends that model future
scenarios of climate change (Ghisbain et al., 2023). Species have
been classified into climate risk categories based on species distri-
bution modelling (Rasmont et al., 2015). Two of the species under
study here, B. pyrrhopygus and B. hyperboreus, are assigned the most
extremeriskclass,HHHR,indicatingthat>95%oftheirinhabitable
range is predicted to be lost by 2100 under a realistic future climate
change scenario. The other four montane species, B. alpinus, B.
balteatus, B. lapponicus and B. monticola, are assigned the category
HHR, indicating a loss of >85%.
We also included the samples of Bombus pascuorum in the study,
whichisamemberofthesubgenusThoracobombus.Incontrasttothe
other spe cies, it is common all ove r Europe, present in a r ange of habitats
fromtheArctictotheMediterranean.Itisthemostabundantbumble-
bee species in Europe and is relatively unthreatened by climate change
compared to other species (Rasmont et al., 2015).Itthereforeserves
asauseful comparisontoother species as it is expectedtohaverela-
tively high levels of genetic variation. Bombus pascuorum is a highly poly-
typic species with 24 described subspecies found throughout Europe
(Lecocq et al., 2015). Four of these subspecies are found in Sweden:
B. p. smithianus north of the Arctic circle, B. p. sparreanus in northern
SwedensouthoftheArcticcircle,B. p. pallidofacies in central and south-
ern Sweden, B. p. gotlandicusonGotland.B. pascuorum is the most wide-
spread and abundant bumblebee species in Europe and can survive in
urban and suburban areas where other bumblebee species are scarce
(Rasmont et al., 2015). B. pascuorum is assigned the lower climate risk
category R, indicating >50% predicted loss in a similar time.
FIGURE 1 Distributionofstudy
speciesinSweden.Pointsonthe
map represent observations since
1997reportedinSwedishSpecies
Observation System (N is the number of
observations). B. lapponicus, B. monticola
and B. balteatus are found widely in the
SwedishmountainsandintheArctic.B.
alpinus, B. pyrrhopygus and B. hyperboreus
are restricted to higher elevations and
latitudes in these areas. B. pascuorum is
a common bumblebee in the whole of
Europe,distributedfromtheArcticto
theMediterranean.MapsfromSwedish
Species Observation System (artportalen.
se).
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17251 by Yuanzhen Liu - New Aarhus University , Wiley Online Library on [21/12/2023]. 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 18
|
LIU et al.
Here,weperformedwhole-genomesequencingon333samples
of the seven bumblebee species collected throughout their ranges
in Sweden. The data were used to estimate a range of population
genetic parameters for each species to inform about their conserva-
tion status and population viability. We first generated phylogenetic
trees of all the samples in order to confirm evolutionary relationship
between species and search for the presence of any previously un-
knowncrypticspecies.Wenextperformedananalysisofpopulation
structure in each species to detect the presence of subpopulations
and barriers to gene flow across their ranges. We estimated the lev-
els of genetic variation across the genome and used these values
to estimate effective population size (NE) for each species. We also
performed a reconstruction of historical variation in NE using the
programmeSMC++, which analyses variation in coalescent time and
the site frequency spectrum of genetic variation across the genome
(Terhorst et al., 2017). Finally, we analysed the length distribution of
runs of homozygosity (ROH) in samples of each species to infer the
prevalence of inbreeding (Ceballos et al., 2018).
2 | METHODS
2.1 | Sampling and DNA extraction
Specimens of seven bumblebee species were collected during the
summer of 2021 from several locations in Sweden. This involved
field trips to high elevation locations in Lapland, Härjedalen and
Jämtland to collect montane species and additional collections of the
widespread species B. pascuorum in locations throughout Sweden.
Sampling locations are displayed in Figure 2, a summary of collec-
tions is found in Table 2 and a full list of samples and associated
data is presented in Table S1. Samples were captured with sweeping
hand nets and kept temporarily in Falcon tubes accompanied with
ice packs. Species and sex/caste of each sample were identified
using a guide (Ødegaard et al., 2015). The identity of each sample
was confirmed by the genetic analysis, which also enabled a subset
of samples for which field identification was ambiguous or incorrect
tobecorrectlyidentified (see below).Sampleswere storedin95%
ethanolat−20°C.TheDNAwasextractedusingtheQiagenDNeasy
Blood&TissueKitfromamiddlelegandthoraxmuscletakenfrom
thecoxalcavityofthemiddleleg.
2.2 | Sequencing and variant calling
All species from the Pyrobombus and Alpinobombus subgen-
era studied here are closely related to a single species from their
subgenus for which a high-quality genome assembly is available.
It is, ther efore, possible to use one genome assembl y from each
subgenus to map reads from all of the other species of that sub-
genus. We used a genome assembly of B. lapponicus sylvicola
(ASM1967717v1;PRJNA646847) to map Pyrobombusspeciesand
a genome assembly of B. kirbiellus(ASM1920181v1;PRJNA704506)
to map Alpinobombus species. Bothof these genome assemblies
FIGURE 2 Mapofbumblebeesample
collection for this study. The solid dots
in different colours denote the sampling
sites for each species across Sweden.
Numberofmales(M)andfemales(F;
workers and queens) collected are given.
MapfromStamendesign(CCBY4.0).
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17251 by Yuanzhen Liu - New Aarhus University , Wiley Online Library on [21/12/2023]. 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 18
LIU et al.
arederivedfromsamplesfromtheRockyMountains,UnitedStates
(Christmas et al., 2021, 2022). B. lapponicus sylvicola in United States
isan extremelycloserelative ofB. lapponicus in Sweden. The most
recent ta xonomic revis ion considers t hem to be the sa me species
(B. lapponicus) although they were previously considered to be two
species, B. sylvicola (North America) and B. lapponicus (Eurasia)
(Mart inet et al., 2019). A simil ar situation ex ists for B. kirbiellus in
NorthAmerica and B. balteatus in Eurasia. In this case, these two
species were previously both considered to be B. balteatus, but the
mostrecenttaxonomicrevisionhassplitB. balteatus into two species
B. kirbiellus in North A merica and B. balteatus in Eurasia (Williams
et al., 2019). The genome assemblies that are available are therefore
from species currently known as B. lapponicus sylvicola (previously B.
sylvicola) and B. kirbiellus (previously B. balteatus), which are closely
related to the study species B. lapponicus and B. balteatus found in
Sweden. Bombus pascuorum is distantly related to the other study
species and weused apubliclyavailable genomeassembly (iyBom-
Pasc1.1;PRJEB43481)tomapreadsfromthisspecies.
Paired-end sequencinglibrarieswerepreparedwith the Nextera
DNAFlexLibraryPreparationKitandsamplesweresequencedonan
IlluminaNovaSeq6000with2 × 150 bpreads.Thereaddataforeach
species were mapped to their corresponding reference genomes. B.
monticola and B. lapponicus were mapped to B. lapponicus sylvicola
(Pyrobombus).B. balteatus, B. alpinus, B. hyperboreas and B. pyrrhopygus
were mapped to B. kirbiellus (previously B. balteatus,Alpinobombus),
and B. pascuorum was mapped to B. pascuorum (Thoracobombus).
Mappingwasperformed using the BWA-MEMalgorithm (Li, 2013).
Mappingsweresortedandwrittentobamfilesusingsamtools.Picard
was used to add read groups and mark duplicate reads in the bam files
thatwerethenindexedinsamtools.Thevariantswerecalledfollowing
theGATKbestpractices fornon-modelspecies (Poplinetal.,2018).
HaplotypeCaller was run with each bam fileswith flag ‘-ploidy’ to
generate individual-specific gVCF files. All gVCF files were merged
on a per contig basis byGenomicsDBImport and called variants by
GenotypeGVCFs.The resulting variant data sets were hard-filtered
followingtheGATKrecommendedthresholds.
Samples were initially all called as diploids, which allowed us to
confidentlyconfirmthesex ofeach sample bythe absenceofhet-
erozygoussitesinhaploidmales.Samplesforwhichthesexassigned
fromidentificationinthefieldwasincongruentwithploidywereex-
aminedmorphologicallyinmoredetailtoconfirmsex.Heterozygous
variants that were detected when HaplotypeCaller was run with
‘-ploidy 2’ for haploid sampleswereconsideredunreliableandfil-
teredoutforeachcohortdataset.Allsampleswereassigned their
correct ploidy to infer genotypes for the final data set.
AdditionalSNP filters wereappliedwith bcftools(Li, 2011): (1)
only including biallelic SNPswith minimum minorallele countof2
orgreaterandexcluding SNPs withreaddepthandgenotypequal-
itypersiteineachsampleunder3 and20,respectively(-vsnps-m
2 -M 2 --min-ac 2:minor -e ‘FMT/DP < 3 | FMT/GQ < 20’); (2) ex-
cluding sites with missing data exceeding 0.5, that is, ‘F_MISSING
> 0.5’; (3) excluding sites with highmean depth across all samples
andexcessiveheterozygosity(INFO/DP > X|INFO/ExcessHet> 25),
whereX = 1000forbothB. pascuorumandAlpinobombusdatasets
andX = 3134forPyrobombustoexcludesiteswithdepthinthetop
5% of the distr ibution. Th e additional f ilters INFO/DP and IN FO/
ExcessHetallowedustoremovesitesduetoalignmentstomultiple
regionsandvariant callingerrors.ThreeSNPdatasetsweregener-
ated, corresponding to sets of samples mapping to each of the three
reference genomes, ready for the downstream analyses.
In order to directly compare the levels of genetic variation
in B. lapponicus and B. balteatus with their sis ter species in N orth
America,the gVCFfilesgeneratedbyHaplotypeCallerwerecom-
bined with a similar data set of samples collected previously in the
UnitedStatesRockyMountains(Christmasetal.,2021, 2022)(NCBI
PRJNA646847and PRJNA704506).The variantswere thengeno-
typed and filtered for the two combined sets of gVCF files follow-
ing the same steps and filters as above. For the combined data sets,
‘INFO/DP> 25000’wasused.Thiscut-offwaschosenbyexamining
theINFO/DPdistributionsineachdataset.
2.3 | Population structure
Reconstructionofneighbour-joining(NJ)trees(Saitou&Nei,1987), prin-
cipalcomponentanalyses(PCA)andadmixture(Alexanderetal.,2009)
analyses were performed to reveal the genetic diversity of each geo-
graphicalpopulation using the SNPdatasets. Fortheseanalyses, only
TAB LE 2 Numberandlocationofbumblebeesamplesusedinthisstudy.
Bombus sp.
Province
Tot alLappland Jämtland Härjedalen Västerbotten Uppland Dalarna Östergötland Västergötland Gotland
alpinus 12 0 0 0 0 0 0 0 0 12
pyrrhopygus 5 1 0 0 0 0 0 0 0 6
balteatus 42 0 1 0 0 0 0 0 0 43
hyperboreus 91 0 0 0 0 0 0 0 10
lapponicus 109 16 3 0 0 0 0 0 0 128
monticola 63 115 0 0 0 0 0 0 79
pascuorum 23 0 4 82 2 4 4 855
Tot al 263 19 23 82 2 4 4 8333
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17251 by Yuanzhen Liu - New Aarhus University , Wiley Online Library on [21/12/2023]. 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 18
|
LIU et al.
SNPswithminorallelefrequencygreaterthan5%wereincluded.Scripts
providedbySimonMartinwereusedtocalculatedistancematrixforNJ
treestaking intoaccountmixed ploidyofsamples (https:// github. com/
simonhmartin/genomics_general). The trees were made and visualized
withFastMe(Guindon&Gascuel,2003) and FigTree v1.4.4 (http:// tree.
bio. ed. ac. uk/ softw are/ figtr ee/ ), respectively.
PCA analyses for each species were implemented in PLINK
v1.9(Chang etal., 2015). To better present the geographical clus-
tering, one sample(DA_2)of B. pascuorum overlapped with sample
DA_03 wasremoveddue totheir closerelatedness, probably from
thesame nest. TheAdmixtureprogramwas usedforeight runsof
population structure by specifying the level of ploidy for each sam-
ple (Alexander et al., 2009). The best-supported values ofK were
inferred byselectingtheonewiththelowestcross-validationerror
byadding‘--cv’flaginthisprogram.
2.4 | Confirmation of species identification
TheNJtreeswereusedinconjunctionwithdetailedmorphological
analysis to correctly identify samples where species identification in
the field was incongruent with genetic analysis. This was particularly
important for species with similar colouration patterns, which is the
case for B. lapponicus and B. monticola, and for B. alpinus and B. pyr-
rhopygus.AWILDM5-47147stereomicroscopeforpreparationwas
used for this purpose.
AllspecimensofB. lapponicus in this study had yellow hairs on
the last two tergites in contrast to samples of B. monticola which had
orange red hairsonthelasttergites.Detailedanalysisofthischar-
acter was particular important for populations of B. lapponicus in the
southern part of the mountain chain, which are darker than northern
populations and difficult to distinguish from B. monticola in the same
area (Ødegaard et al., 2015).
Agood diagnostic field character to separate B. alpinus and B.
pyrrhopygus proved to be the colour of hairs on the two first metaso-
malsegments.InB. alpinus, there is a distinct limit between the black
hairsonthefirsttergiteandtheredhairsonthesecondtergite.In
B. pyrrhopygus, the colour transition is much more diffuse and does
notnecessarilyfollow the tergitelimitation. Inaddition,the cuticle
in B. pyrrhopygus is densely shagreened and dull while in B. alpinus,
it has a smoother, shinier surface. This can preferably be seen on
the outer s urface of the h ind tibiae. Af ter detaile d morphologi cal
analysis of all samples for which genetic analysis was incongruent
with field identification, we were able to confidently determine the
species for each sample.
2.5 | Estimation of effective population size
Effective population size (NE) for each species was estimated using an
estimate of population mutation rate, Watterson's estimator (θW), in-
volving two equations:
𝜃
W=
K
a
n
, where K is the number of segregating
sites in the species and an is the (n–1)th harmonic number, n denoting
the number of chromosome sets. NE can be obtained using an estimate
of θW per base and an estimate of the mutation rate per base μ based
on θ = 2pNEμ, where p is the ploidy of the species (p = 1.5forhaplodip-
loidbumblebees)(Ferretti&Ramos-Onsins,2015; Fu, 1994). For these
analyses, we used a direct estimate of mutation rate from B. terrestris
(μ = 3.6 × 10−9) (Liu et al., 2017).
2.6 | Detection of runs of homozygosity (ROH)
The segments of ROHs for females from each species were assayed
usingthe‘--homozyg’flaginPLINKv1.9(Changetal.,2015), which
uses a sliding window approach to scan each individual's genotype at
each marker position across the genome. The following criteria were
defined to present the length of ROHs: (1) a minimum ROH length of
10 kbcontainingatleast50SNPs;(2)amaximumgapbetweencon-
secutiveSNPsof1 Mb;(3)aminimumdensityofoneSNPin50 kb;(4)
aslidingwindowof50SNPsacrossthegenomewithoneSNPstep;
(5)notoleranceofheterozygoussitesandamaximumthreemissing
genotypes; (6) a default window threshold of 0.05. The lengths of
ROHweredividedintothreeclasses: 0.01–0.5 Mb, 0.5–1 Mb,1–1.5
(detectedmaximumlength)Mb.
2.7 | Inference of population demographic history
SMC++ (Terhorst et al., 2017) was used to infer past fluctuations in
NE. This method uses the inferred distribution of coalescence times at
lociacrossthegenomeandthesitefrequencydistributionofSNPsto
infer temporal variation in NE (Terhorst et al., 2017).To ensure the reli-
abilityofSNPcalling,onlyfemalesampleswithreadcoveragegreater
than 13×were selected forrunningSMC++, in line with other stud-
ies (Wang et al., 2021).None of the samples included in this analy-
sisshowedhighrelatednessaccordingtothePCAanalysis.Themean
depthperindividualwasestimatedusingthe‘--depth’flaginVCFtools
(Daneceket al.,2011). The previously filtered VCF files were used as
inputs by keeping only these samples. Eventually, 49 samples were
used for the demographic inference, including seven B. pascuorum, 10
B. monticola and 11 B. lapponicus, as well as 12 B. balteatus and three for
B. alpinus, B. hyperboreus and B. pyrrhopygus, respectively. We used the
‘-d’optiontospecifybetween3and12differentsamplesperspeciesas
‘distinguishedindividuals’acrossrunsinordertogenerateacomposite
likelihood, which should result in improved estimates of population size
history (Terhorst et al., 2017). EstimationsfromSMC++ were scaled
using a generation time (g)of1 yearandamutation rate(μ)3.6 × 10−9
substitutions per site per year (see above).
3 | RESULTS
3.1 | Field sampling of seven bumblebee species
The samples of the study species were collected during the summer
of2021.Samplecollection in Abisko and thevicinity yielded sam-
plesofallsevenspecies(263samplesintotal).Additionalcollections
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17251 by Yuanzhen Liu - New Aarhus University , Wiley Online Library on [21/12/2023]. 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 18
LIU et al.
were also made in Jämtland (19samples) and Härjedalen (23 sam-
ples). Samples of the widespread species B. pascuorum were col-
lected from throughout its range in Sweden, including the island of
Gotland.Thisincorporatedsixmaingeographicareasinwhichfour
different subspecies and hybrids between them were located.
The final data sets consisted of 333 samples (Webster, 2022).
Three of the species—B. alpinus, B. pyrrhopygus and B. hyperboreus—
have a more res tricted distribution at higher elevations and latitudes.
These species were found in lower numbers than the other, more
widespread, montane species and were only present in a subset of
localities. The sampling locations for all seven species are displayed
in Figure 2.Asummaryofsamplescollectedineachlocationispre-
sented in Table 2. A full list of sa mples that wer e sequenced an d
associated metadata is presented in Table S1.
3.2 | Whole- genome sequencing and
bioinformatic analysis
All samples were subjected to Illumina whole-genome sequenc-
ing. Raw sequence reads from each sample were aligned to the
appropriate reference genome to generate three separate multi-
speciesalignments:(1)The207Pyrobombussamples(B. lapponicus
and B. monticola) were aligned to B. lapponicus sylvicola (Christmas
et al., 2021). (2) The 71 Alp inobombus s amples (B. alpinus, B. pyr-
rhopygus, B. balteatus, B. hyperboreus) were aligned to the B. kirbiellus
reference assembly (Christmas et al., 2022). (3) The 55 B. pascuorum
samples aligned to the B. pascuorum reference genome assembly.
Meansequencingdepth was 10.5xperindividualanddidnotvary
substantially between species (Table S1).
We first determined ploidy of each sample by detecting the pres-
ence of heterozygous sites. This analysis confirmed whether samples
were male (haploid) or female (diploid). One sample of B. hyperboreus
was identified morphologically as a male, but was found to be diploid
on the basis of the genetic analysis, indicating that it is a diploid male
(see below for additional analysis). We updated sample classifications
in cases of i ncorrect classif ication in the fi eld. Species ide ntity was then
confirmedbyconstructingNJtreesforeachofthethreemultispecies
alignments. The seven different species were clearly distinguishable
from NJ trees. A small number of samples thathad beenincorrectly
classified were identified from NJ trees and correctly classified. In
addition, the identity of samples of the two species B. alpinus and B.
pyrrhopygus, which are difficult to identify due to their convergence in
morphology, was confirmed by a combination of genetic clustering in
NJtreesanddetailedmorphologicalanalysis.The numbersinTable 2
refer to correctly classified samples.
3.3 | Levels of genetic variation
We estimated the levels of genetic variation in all species using
Watterson's theta (θW) and nucleotide diversity (average number of
pairwise differences, π) (Watterson, 1975) (Table 3). Estimates of θW
TAB LE 3 GeneticvariationinsevenSwedishbumblebeespecies.
Bombus sp. Male Female
Haploid genomes
sampledaNo. of SNPs
Watterson's theta per
base (θW) (%)
Nucleotide diversity
(π) (%)
Effective population size
(NE)
Observed
heterozygosity (HOBS)
alpinus 12 24 570,149 0.061 0.08 56,534 0.333
pyrrhopygus 612 1,390,927 0.184 0.236 170,545 0.371
balteatus 17 26 69 2,370,054 0.197 0.24 182,673 0.283
hyperboreus 10 20 660,899 0.074 0.099 68,978 0.306
lapponicus 62 66 194 3,095,046 0.21 0.227 194,583 0.280
monticola 28 48 124 1, 085,476 0.08 0.091 73,814 0.298
pascuorum 14 41 96 3,80 0,079 0. 241 0.273 222,742 0.218
aSum of haploid genomes found in both haploid and diploid individuals.
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17251 by Yuanzhen Liu - New Aarhus University , Wiley Online Library on [21/12/2023]. 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 18
|
LIU et al.
are 0.06%–0.25% (Wallberg et al., 2014), which we converted into
estimates of long-term NE based on a direct estimate of mutation
rate in the bumblebee Bombus terrestris (see Section 2). Levels of NE
vary between species. The highest estimates of NEareexhibitedby
B. pascuorum (~220,000; Table 3). This is consis tent with the fact t hat
it is a widespread species in Europe, found in many habitats, which
maybeexpandinginnumbers(Rasmontetal.,2015). Two other spe-
cies that are relatively widespread in the Swedish mountains, B. lap-
ponicus and B. balteatus,alsoexhibitrelativelyhighNE (~195,000and
~180,000,respectively).B. pyrrhopygusalsoexhibitedrelativelyhigh
NE (~170,000), which is surprising considering that it restricted to
high elevations and was the species observed the fewest number
of times during collection. Conversely, those species with the most
restricted distributions had significantly lower levels of genetic di-
versity. Estimates of NE for B. alpinus and B. hyperboreus were both
in the vicinity of 60,000 (Table 3). Surprisingly, B. monticola also has
an NE in this range, despite being one of the more abundant montane
species. In general, however,differences in levels of long-term NE
betweenspecieswereconsistentwithexpectationsbasedon spe-
cies abundance and range size.
3.4 | Population structure
We estimated population structure and species relationships of all
species using three methods: Neighbour-joining(NJ)trees (Saitou &
Nei,1987),principlecomponentanalysis(PCA)andadmixtureanalyses
(Alexanderetal.,2009).TheNJ treeofB. lapponicus and B. monticola
reveals that the samples cluster by species with no additional struc-
turing apparent (Figure 3a). Samples of B. lapponicus from Lappland,
Jämtland and Härjedalen showed no evidence of population substruc-
tureorcluste ringbyge og raphyinthePCA(Figure S1). This was further
supportedbytheadmixtureanalysis,inwhichasinglepopulationwas
thebest supportedmodelbycross-validation (Figure 3b). Samples of
B. monticolafromAbiskoformedclustersaccordingtoPCA(Figure S1).
However, the three samples that form a separate cluster were col-
lected inexactlythesamelocationand themost likely reasonforthe
clusteringis thereforethattheyarerelated.Admixtureanalysisfor B.
monticola also supports a single panmictic population (Figure 3b).
NJ trees f or the species of sub genus Alpinobu mbus—B. bal-
teatus, B. alpinus, B. pyrrhopygus and B. hyperboreus—reflect the
evolutionary relationships between these species (Figure 4a).
Tree topologies are the same as those published based on anal-
yses of phenotypic characters and molecular barcodes (Williams
et al., 2019), which place B. alpinus and B. pyrrhopygus as sis-
ter taxa. No significant evidence of clustering either within or
among geographical regions was observed for any of the four
AlpinobombusspeciesusingPCA(Figure S1).Asinglepopulation
was the be st-suppor ted model by cross-valid ation using admix-
ture for all species (Figure 4b). However, it should be noted that
the sample sizes for B. alpinus, B. pyrrhopygus and B. hyperboreus
were all relatively low (6–12 individuals per species), which results
in low power to detect population structure. B. alpinus was only
sample d in Abisko, which me ans it is impossibl e to identify ge-
netic structuring among geographic regions for this species. For
B. hyperboreus, a single sample collected in Jämtland did not clus-
terwiththeothersamplesfromAbisko,whichcouldindicatethe
presence of geographical structure in the species although more
samples would be required to confirm or refute this.
In contrast to the species already discussed, B. pascuorum is a
polytypic species found in a wide range of habitats across Europe
(Lecocq et al., 2015). The sampling strategy for this species incorpo-
rated the four subspecies present in Sweden, in addition to samples
identified as hybrids between pairs of subspecies collected in hy-
bridzones. TheNJtreeofthese samples revealedlimited levelsof
clustering (Figure 5a).ThesamplescollectedfromGotland,fromthe
subspecies, B. p. gotlandicus, were clearly separated from the other
samples.Alimitedamountofseparationwasalsoobservedforsam-
plesfromAbisko,consisting ofthesubspecies B. p. smithianus. The
remaining samples, consisting of B. p. sparreanus, B. p. pallidofacies
and hybrids between subspecies, collected over a wide geographical
area,donotshowevidenceofsubstructure.ThePCAplotrevealsa
similarpattern,withsamplesfromGotlandandAbiskoformingsep-
arate clusters from the remaining samples from mainland Sweden.
These remaining samples do not show evidence for geographic
structuring (Figure S1). The a dmixture anal ysis also revealed t hat
samplesfromAbiskoandGotlandformedseparateclustersfromthe
remaining samples (Figure 5b). Cross-valida tion analysis indic ated
that K = 2hadbestsupport,implyingtwomajorpopulations,oneon
GotlandandoneonmainlandSweden,althoughtheadmixtureplots
alsoshowsomedegreeofdistinctnessoftheAbiskopopulation.
Taken together, these analyses indicate that the six montane
bumblebee species studied here do not exhibit obvious popula-
tion subs tructure a cross their ra nges in Sweden. By co ntrast, th e
widespread polytypic species B. pascuorumexhibitsadegreeofsub-
structure.ApopulationonGotlandappearstobeisolatedfromthe
mainlandpopulation.Furthermore,samplescollectedinAbiskoap-
pear to have a degree of isolation from other mainland populations
further south, which display different colouration patterns and are
described as belonging to different subspecies (Lecocq et al., 2015).
3.5 | Inference of historical fluctuations in NE using
SMC++
Weusedthe program SMC++ to infer historical fluctuations in NE
in all seven of the study species (Figure 6). We infer that popula-
tions of B. hyperboreus have declined since ~104 years ago (the end
of the last ice age). The analysis suggests that current NE could be
less than 5000, although historically it has been more than 10 times
higher (Table 3, ~70,000). The species B. pyrrhopygus also shows
signs of decline in this period, which could indicate a contemporary
NE around 25,000 compared to a value of ~75,000 estimated from
θW. The other montane species under investigation do not show the
signals of majorpopulation expansions or declines and have been
relativelystableduring theHolocene.Incontrast to the arctic and
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17251 by Yuanzhen Liu - New Aarhus University , Wiley Online Library on [21/12/2023]. 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 18
LIU et al.
montane species, there are signs that B. pascuorum has undergone a
populationexpansionduringthisperiod.
3.6 | Runs of homozygosity
We scanned the genomes of all diploid samples (typically females) for
runs of homozygosity (ROH) (Table 4; Figures S2–S5). Figure 7 shows
the proportion of the genome of each sample that falls into ROH
>0.1 Mb, grouped byspecies. The most striking finding was a single
individual of B. hyperboreus(WO_001)with>55% of the genome lying
inROHlongerthan0.1 Mb(Figure 7). This pattern is predicted to result
from a sib-mating.This sample was also classified as a diploid male,
whichindicatesthatitishomozygousatthesex-determininglocus.The
sample is the only B. hyperboreuscollected from Jämtland.All of the
othersamplesofthisspecieswerefoundinthevicinityofAbisko.
FIGURE 3 (a)Neighbour-joiningtreeforsamplesofB. lapponicus and B. monticolageneratedusingwhole-genomesequencedata.Both
speciesareinthesubgenusPyrobombus.Eachexternalbranchrepresentsanindividualsample.(b)Admixtureanalysesforallgeographical
populations of B. lapponicus (top) and B. monticola(bottom).NosignificantstructurewasdetectedforeitherspeciesandK = 1wasthebest
supported parameter value for both species.
K =2
K =3
Abisko/Kiruna
Jämtland
Härjedalen
Populations of B. lapponicus
K =2
K =3
Abisko/Kiruna
Jämtland
Härjedal
en
Populations of B. monticola
0.05
AC-028
M_L018
3
9
0
-
C
A
AC-012
M_L005
M_L223
M_L077
M_L121
M_L085
M_L012
BH_02
M_L144
M_L215
M_L004
1
8
0
-
C
A
M_L007
AC-017
M_L081
M_L150
M_L205
WO_111
4 5 1 L _ M
M_L021
AC-141
M_L199
M_L056
M_L130
M_L122
BH_06
AC-034
M_L185
M_L139
8 4 1 L _ M
M_L143
M_L140
M_L184
M_L054
WO_358
M_L030
BH_19
M_L003
M_L138
190
-C
A
M_L027
M_L169
M_L220
M_L166
500_
O
W
WO_006
M_L187
M_L171
BH_09
M_L180
AC-111
WO_815
M_L216
AC-021
M_L055
AC-061
M_L115
AC-043
M_L110
M_L174
AC-118
BH_14
AC-129
M_L076
M_L206
M_L124
M_L114
M_L218
21
_
H
B
AC-038
M_L167
M_L203
M_L060
M_L153
WO_007
AC-073
WO_448
BH_01
M_L211
M_L192
M_L176
M_L175
M_L109
M_L146
AC-082
M_L113
M_L172
M_L147
M_L071
M_L194
AC-135
AC-014
M_L009
AC-115
M_L075
4
4
0
-
CA
AC-049
M_L183
M_L089
M_L200
WO_449
M_L152
M_L083
M_L043
M_L112
WO_002
BH_08
M_L136
M_L125
AC-138
M_L160
M_L157
AC-133
M_L017
M_L090
M_L086
AC-080
M_L091
M_L226
M_L099
M_L092
AC-006
BH_18
M_L100
M_L186
M_L133
M_L097
M_L159
WO_067
M_L189
M_L028
M_L074
M_L013
M_L219
BH_05
M_L008
BH_15
WO_269
721L_ M
M_L178
BH_11
M_L120
M_L224
AC-136
M_L117
AC-105
M_L082
M_L072
AC-137
5
6
8
_
O
W
M_L111
M_L029
M_L123
AC-071
M_L093
M_L191
AC-087
AC-069
M_L145
WO_703
M_L025
M_L095
BH_03
M_L135
M_L014
AC-046
AC-023
M_L096
M_L214
M_L193
BH_07
BH_10
AC-067
WO_004
1
4 1
L _
M
M_L087
M_L221
WO_383
AC-106
WO_331
M_L057
BH_04
M_L210
M_L213
WO_003
M_L119
M_L002
M_L088
BH_16
M_L131
M_L046
M_L084
M_L116
M_L173
AC-072
BH_17
AC-150
M_L061
AC-047
M_L197
M_L018
3
9
0
-
C
A
AC-012
M_L005
M_L077
M_L121
M_L085
M_L012
M_L144
M_L215
1
8
0
-
C
A
AC-017
M_L081
WO_111
4
5 1
M_L021
AC-141
M_L199
M_L122
M_L185
M_L139
8
4
1
L
_
M
M_L184
WO_358
M_L030
M_L003
M_L138
1
9
0
-
C
A
5
0
0
_
O
W
WO_006
M_L187
M_L171
M_L180
WO_815
M_L216
M_L115
AC-043
M_L110
M_L174
M_L076
M_L124
M_L114
2
1
_
H
B
A
C-038
M_L060
WO_007
AC-073
WO_448
M_L192
M_L176
M_L175
M_L109
AC-082
M_L113
M_L071
M_L194
AC-135
M_L009
AC-115
4
4
0
-
C
A
M_L183
M_L089
M_L200
WO_449
M_L083
M_L043
M_L112
M_L125
AC-138
M_L157
M_L090
M_L086
A
C-080
M_L091
M_L226
M_L099
M_L092
AC-006
M_L100
M_L186
M_L097
WO_067
M_L189
M_L028
M_L074
M_L008
WO_269
7 2
1
L
_
M
M_L178
BH_11
M_L120
A
C-136
M_L117
AC-105
M_L082
M_L072
A
C-137
5
6
8
_
O
W
M_L111
M_L123
AC-071
AC-087
AC-069
M_L145
WO_703
M_L095
M_L135
M_L014
M_L096
M_L193
BH_10
AC-067
WO_004
1
4
1
L
_
M
M_L087
WO_383
WO_331
M_L057
WO_003
M_L002
M_L088
M_L046
M_L084
M_L116
M_L173
AC-072
AC-150
M_L061
AC-028
M_L223
BH_02
M_L004
M_L007
M_L150
M_L205
M_L056
M_L130
BH_06
AC-034
M_L143
M_L140
M_L054
BH_19
M_L027
M_L169
M_L220
M_L166
BH_09
AC-111
AC-021
M_L055
AC-061
AC-118
BH_14
AC-129
M_L206
M_L218
M_L167
M_L203
M_L153
BH_01
9
M_L211
ML22
ML22
M_L146
M_L172
M_L147
AC-014
M_L075
AC-049
M_L152
WO_002
BH_08
BH_09
BH_09
M_L136
M_L160
AC-133
M_L017
B
H_18
M_L133
M_L159
L166
M_L013
M_L219
BH_05
B
H_15
M_L224
M_L029
M_L093
M_L191
M_L025
BH_03
AC-046
AC-023
M_L214
M_L224
M_L224
BH_07
M_L221
AC-106
BH_04
ML210
ML211
ML211
M_L213
ML075
M_L119
BH_16
M_L131
BH_17
5
5
AC-047
M_L197
Bombus lapponicusBombus monticola
M_L168
AC-020
M_L202
M_L202
(a)
(b)
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17251 by Yuanzhen Liu - New Aarhus University , Wiley Online Library on [21/12/2023]. 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 18
|
LIU et al.
Populations of B. balteatus
Populations of B. alpinus
Populations of B. pyrrhopygus
Populations of B. hyperboreus
K =2
Abisko/Kiruna_ML
Abisko/Kiruna_AC
K =2
Abisko/Kiruna
Jämtland
K =2
Abisko/Kiruna
Härjedalen
Abisko/Kiru
na
Jämtland
0.05
AC-152
AC-029
0
1
0-
C
A
WO_456
AC-022
M_L179
M_L155
M_L137
AC-090
M_L022
M_L020
M_L129
AC-027
M_L033
M_L188
AC-045
M_L019
M_L011
M_L126
M_L102
M_L106
AC-121
M_L195
AC-007
M_L149
M_L026
M_L062
M_L053
M_L032
AC-016
M_L098
AC-041
M_L151
WO_001
AC-003
M_L177
M_L198
AC-153
M_L024
AC-025
AC-019
AC-004
M_L196
AC-008
M_L103
M_L181
M_L006
AC-126
M_L058
M_L010
AC-156
AC-002
M_L142
AC-078
AC-015
AC-033
M_L015
BH_13
M_L105
M_L049
M_L107
M_L108
AC-113
M_L101
AC-018
M_L190
AC-112
AC-011
AC-036
M_L182
AC_026
M_L020
AC-027
M_L062
M_L053
62
AC-041
WO_001
M_L058
AC-015
3
AC-018
AC-036
Bombus hyperboreus
AC-152
AC
-029
AC
-
022
M_L179
M_L155
M_L137
AC
-
090
M_L129
M_L033
M_L188
M_L126
M_L102
M_L106
AC-121
M_L195
M_L149
M_L032
AC-016
M_L098
M_L151
AC-003
M_L177
M_L198
M_L024
AC-004
M_L196
M_L103
M_L181
AC-126
AC-156
AC-002
M_L142
AC-078
BH_13
M_L105
M_L049
M_L107
M_L108
AC-113
M_L101
M_L190
AC-011
M_L182
Bombus balteatus
WO_456
M_L026
AC
-
025
M_L015
A
C-112
AC_026
Bombus pyrrhopygus
0
1
0
-
C
A
M_L022
AC-045
M_L019
M_L022
M_L
M_L011
AC-007
AC-153
AC-019
AC
A
AC-008
M_L006
M_L010
M_L
M_
AC
A
AC-033
M_L
Bombus alpinus
K =2
(a)
(b)
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17251 by Yuanzhen Liu - New Aarhus University , Wiley Online Library on [21/12/2023]. 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 18
LIU et al.
AsmallnumberofindividualsofB. pyrrhopygus, B. pascuorum
and B. monticola were also identified with 5%–20% of the genome
covered by ROH bl ocks longer than 0.1 Mb indicating that in-
breeding occurs sporadically in these species as well (Figure 7).
Outliers with long ROH tracts were not detected in the other spe-
cies, but may still occur at low frequencies. Only a small number
of sample s contained any ROH s longer than 1 Mb, an d for the
majority of species, none of samples had any such long ROHs
(Table 4). The mean coverage of the genome by ROH is <5% for all
species apart from B. hyperboreus, where the presence of the one
highly inbred diploid male individual strongly influences this mea-
sure. Average levelsofROH in B. pascuorum are high compared
FIGURE 4 (a)Neighbour-joiningtreeforB. alpinus, B. pyrrhopygus, B. balteatus and B. hyperboreus.Allspeciesareinthesubgenus
Alpinobombus.Eachexternalbranchrepresentsanindividualsample.(b)AdmixtureanalysesforallgeographicalpopulationsofB. alpinus, B.
pyrrhopygus, B. balteatus and B. hyperboreus.NosignificantstructurewasdetectedforanyofthespeciesandK = 1wasthebestsupported
parameter value for all species.
FIGURE 5 (a)Neighbour-joiningtreeforB. pascuorum. Several subspecies of B. pascuorumarepresentinSweden.Eachexternalbranch
representsanindividualsample.Samplinglocationsandsubspeciesareindicated,includingseveralfromhybridzones.(b)Admixtureruns
for all geographical populations of B. pascuorum. The best supported parameter value is K = 2,whichindicatesthepresenceofadistinct
subpopulationonGotland(B. pascuorum gotlandicus).
K = 2
K = 3
K = 4
Abisko
Härjedalen/Ume
å
Ammarnäs
Västergötland
Uppsala/Dalar
na
Gotland
0.02
NS_02
M_L068
SS_07
SL_04
M_L204
GT_01
SS_02
GT_05
M_L201
GT_02
M_L064
NS_01
GT_04
M_L209
SS_08
M_L132
M_L065
M_L134
GT_08
M_L208
M_L041
M_L066
DA_04
SL_01
DA_02
NS_07
SS_05
M_L207
M_L070
M_L067
SS_01
GT_03
M_L039
M_L170
DA_03
M_L051
M_L063
SS_03
GT_06
SS_04
NS_08
NS_05
M_L050
NS_03
SS_06
M_L069
GT_07
NS_04
M_L016
M_L001
M_L217
NS_06
SL_02
SL_03
DA_01
B. pascuorum smitheanus
Abisko
B. pascuorum gotlandicus
Gotland
B. pascuorum sparreanus/smitheanus
Ammarnäs
B. pascuorum sparreanus
Härjedalen/Umeå
B. pascuorum pallidofacies
Västergötland
B. pascuorum pallidofacies/sparreanus
UppsalaDalarna
(a)
(b) Populations of B. pascuorum
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17251 by Yuanzhen Liu - New Aarhus University , Wiley Online Library on [21/12/2023]. 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 18
|
LIU et al.
to the other species, with ROH covering ~5% of the genome on
average (Table 4).
3.7 | Comparison of levels of genetic variation with
North American montane bumblebee species
Sister species of B. lapponicus and B. balteatus are found in mountain
andarcticlocationsinNorthAmerica.Thesespeciesarehighlysimi-
lar in morphology and recently evolutionarily diverged from their
European counterparts. The sister species of B. lapponicusinNorth
AmericaisalsoknownasB. lapponicus (subspecies sylvicola). The sis-
ter species of B. balteatusinNorthAmericaisB. kirbiellus. We com-
pared our results to a data set of 214 samples of B. lapponicus and
299 sampl es of B. kirbiellusf rom the Rock y Mountains, C olorado,
United States (Christmas et al., 2021, 2022). The data set generated
here also permitted comparisons of genetic diversity in this popula-
tion (Table 5). To do this, we jointly called genetic variation in both
data sets simultaneously in order to directly compare the levels of
genetic variation. For the two populations of B. lapponicus(USAand
Sweden), we find that the levels of nucleotide diversity and NE are
extremelysimilar(~160,000). We find that B. kirbiellus(USA)hassig-
nificantly higher NE compared to B. balteatus (Sweden) (NE ≈ 310,000
compared to 170,000).
4 | DISCUSSION
Weusedwhole-genomesequencingtosurveygeneticvariationand
analyse population history and structure of seven bumblebee spe-
ciesfoundinSweden:sixspeciesrestrictedtoarcticandmountain
environmentsandonewidespreadspecies.Ingeneral,wefindalack
of population structure in all of the montane species, and compara-
ble levels of NE in both montane and widespread species. We iden-
tify some indications of declines and elevated genetic load in the
rarer montane species. The results are useful to evaluate the conser-
vation status of each species and to inform management decisions.
4.1 | Relatively high NE and low population
structure in montane bumblebees
Forall of the species under study here, we estimate long-term NE
from levels of genetic variation to be greater than 50,000. Variation
in estimates of NE among species reflects differences in their range
and abundance. B. pyrrhopygus, B. balteatus and B. lapponicus have
NE close to that estimated for the widespread species B. pascuorum
(~200,000), which indicates their range and abundance are also
relatively large. These levels of NE are similar to other widespread
social bees such as honeybees (Wallberg et al., 2014).Bycontrast,
FIGURE 6 EstimationofhistoricalvariationinNEusingtheSMC++ method. The x-axisdenotesyearsbeforepresent.Strongsignalsof
population decline are inferred for B. hyperboreus, whereas population size increases are inferred for B. pascuorum.
Bombus pascuorum
Bombus monticola
Bombus lapponicus
Bombus balteatus
Bombus alpinus
Bombus hyperboreus
Bombus pyrrhopygus
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17251 by Yuanzhen Liu - New Aarhus University , Wiley Online Library on [21/12/2023]. 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 18
LIU et al.
B. alpinus, B. monticola and B. hyperboreus have lower NE than the
other montane bumblebee species. B. alpinus and B. hyperboreus
have a more restricted distribution in at higher elevations and lati-
tudes (Figure 1) (Ødegaard et al., 2015; Rasmont et al., 2015) and are
thereforeexpectedtohavesmallerpopulationsizes. Avalue of NE
for a population less than 500 is commonly used to suggest it is vul-
nerableandlikelytoexperiencethenegativeeffectsofgeneticdrift
and inbreeding (Allendorf etal., 2022). Levels of NE in the species
studied here are substantially higher than this.
The six mon tane species of bu mblebees stu died here did not
show evidence for population structure among the montane loca-
tions whe re they were collec ted in Lapland , Jämtland and Här jedalen.
This indicates that there are no substantial barriers to gene flow and
that the populations have historically been continuous across the
Swedish mountain range. Studies of genetic structure B. kirbiellus
and B. lapponicus sylvicolafromNorthAmer icathataresiste rs pecies
oftwo of the species studied here (Martinet etal., 2019; Williams
et al., 2019) do not show evidence for geographical substructure
within the Rocky Mountains, Colorado, United States (Christmas
et al., 2022). Other studies have, however, found evidence for ge-
netic differentiation of these populations on different mountain
rangesinNorthAmerica(Kochetal.,2017; Whitley, 2018). We find
that B. lapponicus samples collected from populations in Sweden and
Colorado have similar levels of genetic variation, whereas popula-
tions of B. kirbiellus from Colorado have significantly higher levels
of genetic variation compared to its sister species B. balteatus in
Sweden. This indicates that B. kirbiellus may be relatively more abun-
dantinNorthAmerica,whichisconsistentwithobservationsfrom
field collections (Christmas et al., 2022).
Ingeneral,studiesinbothEuropeandNorthAmericahavefound
that relatively abundant bumblebee species such as B. terrestris do
notexhibitpopulationstructureoverthousandsofkilometres(Colgan
et al., 2022; Ghisbain et al., 2020; Heraghty et al., 2023; Lozier
et al., 2011), whereas very rare and declining species (e.g. B. musco-
rum)mayshowstructureonascaleofonly10 km(Darvilletal.,2006).
Geneticdifferentiationofpopulationscanoccurinmountainrangesif
species are restricted to high elevations, as with B. vancouverensis in
the southern part of its range in the western United States (Jackson
et al., 2018; Lozier et al., 2023). The results presented here are there-
fore consistent with observations from other studies indicating that
bumblebeesdonotexhibitpopulationstructureintheabsenceofgeo-
graphical barriers unless they have particularly low abundance.
We analysed population structure in four subspecies of B. pas-
cuorum, which are geographically separated and differ in coloura-
tion (Lecocq et al., 2015). The subspecies B. p. gotlandicus, which is
restricted to the island of Gotland,was clearly genetically distinct
fromtheother subspecies,consistentwithisolationin allopatry.In
addition, the subspecies B. p. smithianus, present in the far north of
Sweden, was genetically separable from other subspecies (Figure 5).
Incontrast, however,the twosubspecies B. p. sparreanus and B. p.
pallidofacies were indistinguishable genetically, indicating that there
are unlikely to be any barriers to gene flow between these two
subspecies.
TAB LE 4 Summarystatisticsofrunsofhomozygosity(ROH).
Bombus sp. Female no.
Total no. of ROH of each length
Mean total length of ROH per sample (Mb) Mean genome coverage by ROH per sample (%)0.01–0.1 Mb 0.1–1 Mb >1 Mb
alpinus 12 301 48 01.72 0.69
pyrrhopygus 6290 87 05.17 2.07
balteatus 26 4797 55 04. 51 1.80
hyperboreus 10 1844 804 125.47 10.18
lapponicus 53 7511 771 05.45 2.16
monticola 49 3799 971 26.99 2.77
pascuorum 41 12,329 1325 314.42 4.69
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17251 by Yuanzhen Liu - New Aarhus University , Wiley Online Library on [21/12/2023]. 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
14 of 18
|
LIU et al.
4.2 | Evidence of inbreeding and
population declines
Haplodiploid organisms such as bumblebees and other Hymenoptera
are particularly vulnerable to inbreeding because it can lead to ho-
mozygosityat the sex locus, whichproduces diploid males, which
are unable to successfully fertilize females. The presence of diploid
males can therefore severely reduce the fitness of a population.
When populations become small and inbreeding is common, this ef-
fect can leadto a particularly extremeextinctionvortex (Lozier &
Zayed, 2017; Zayed & Packer,2005). Diploid male production has
been observed in populations of rare and declining bumblebee spe-
ciesandisanindicatorofhighvulnerability(Darvilletal.,2006; Ellis
et al., 2006).
Inouranalysis,weidentifiedoneexampleofadiploidmale—aB.
hyperboreus collected in Jämtland. This sample also had the highest
degree of inbreeding detected in any sample: >50% of its genome
consistedof ROH greater than 100 kbpin length.It wasthe only
B. hyperboreus collect ed in Jämtland. It is n oteworthy that thi s is
the only observation of B. hyperboreus this far south reported in
Swedish Species Observation System, which has records over the
last25 years,withalloftheother143reportsrestrictedtoLapland
(themajorityinthevicinityofAbisko).Thisindicatesthatthissample
was found at or close to the southern limit of the species distribu-
tion in the Swedish mountains. This finding indicates that inbreeding
is occurring in the population of B. hyperboreus, particularly on the
edges of its range where it is most likely to be influenced by the
effects of climate change, although it is not possible to determine
theextentofinbreedinginapopulationbasedontheobservationof
a single diploid male. We do not find evidence for inbreeding in any
other species. Relatively high average levels of ROH are found in B.
pascuorum.AsB. pascuorumisanextremelywidespreadbumblebee,
itisunlikelythatitisexperiencinginbreeding.Itisthereforeunlikely
that the average levels of ROH observed in any of the species here
FIGURE 7 Proportionofgenomeof
eachsamplecomprisingROH > 0.1 Mbp.
Outliers are labelled.
AC−029
M_L198
WO_001
WO_456
AC−028
AC−111
BH_09
M_L169
GT_02
GT_03
GT_05
GT_06
GT_07
B. alpinus
B. balteatus
B. hyperboreus
B. lapponicus
B. monticola
B. pascuorum
B. pyrrhopygus
020 40
Proportion of genomewith ROH greater than 0.1 Mb (%)
Species
TABLE 5 ComparisonofgeneticvariationinrelatedspeciesfromNorthernEuropeandNorthAmerica.
Species
Haploid genomes
sampledcNo. of SNPs
Watterson's theta per
base (θw) (%)
Nucleotide diversity
(π) (%)
Effective
population size (NE)
lapponicus cluster
B. lapponicus (SWE) 194 2,514,341 0.171 0.161 158,075
B. lapponicusa(USA) 428 3,010,424 0.18 0.219 166,652
balteatus cluster
B. balteatus (SWE) 69 2, 250,496 0.187 0.207 173,458
B. kirbiellusb(USA) 598 5,824,609 0.334 0.271 3 0 9,42 7
aThespeciesinNorthAmericacurrentlynamedBombus lapponicus (subspecies sylvicola) was previously considered to be a separate species named
Bombus sylvicola(Martinetetal.,2019).
bThe species Bombus kirbiellusinNorthAmericawaspreviouslyknownasBombus balteatus (Williams et al., 2019).
cSum of haploid genomes found in both haploid and diploid individuals.
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17251 by Yuanzhen Liu - New Aarhus University , Wiley Online Library on [21/12/2023]. 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
|
15 of 18
LIU et al.
have been influenced by inbreeding as they do not differ markedly
from those observed in B. pascuorum.
AlthoughweestimatethatNE is relatively high in all of the study
species, it is important to note that our estimate of NE from θW re-
flects the harmonic mean over evolutionary time since the coales-
cence of the sequences in the sample, rather than being an estimate
of contemporary NE(Nadachowska-Brzyskaet al., 2022). We also
performedanSMC++ analysis to estimate fluctuations in NE over
the evolutionary history of the populations (Terhorst et al., 2017).
Asignificantfindingof thisanalysis was thatB. hyperboreus and to
alesserextent B. pyrrhopygus showed signals of population decline
during the current postglacial period in the last ~10,000 years.Both
of these species have restricted arctic distributions, but may have
been more widespread earlier in the current post-glacial period
when conditions were more favourable in a greater part of Europe.
Itis therefore possiblethat their populations have declined in the
current post-glacial period. This could suggest that the range of
these species has shrunk and that current NE is much smaller than
predicted from overall levels of genetic variation. The observation of
inbred individuals in these species from the ROH analysis also sup-
portsthissuggestion.Bycontrast,B. pascuorum is currently the mos t
widespread and abundant bumblebee in Europe and is able to sur-
vive in a range of modern habitats (Lecocq et al., 2015).TheSMC++
analysisindicatesthatthisspeciesmayhaveexpandedsincethelast
ice age, although it is possible that the weak population structure
observed in this species may have influenced this result.
Acontrastingpatternofinferred population history usingSMC
approachesisexhibitedbythemontanebumblebeeB. vancouveren-
sis, which shows greater degree of population fragmentation along
anorth–south gradientin western North America. In this species,
more isolated southern montane populations show lower genetic
variation and population declines (Lozier et al., 2023). This could
reflect a greater degree of habitat fragmentation, whereby sub-
populations of this species have become geographically isolated
athighelevationmountainpeaks.By contrast, thecommon North
AmericanbumblebeeB. vosnesenskii shows low population structure
and no evidence of decline (Heraghty et al., 2023), similar to our ob-
servations in B. pascuorum.
4.3 | Absence of cryptic species
Asmentionedabove,themontanebumblebeespeciesstudiedhere
havesister speciesinNorthAmericaandtheir subgenera have cir-
cumpolardistributions(Martinetetal.,2019; Williams et al., 2019).
Previous studies have used whole-genome sequencing to assess
genetic variation and population structure in B. lapponicus sylvicola
(sister species of B. lapponicus) and B. kirbiellus (sister species of B.
balteatus) in several high-elevation locations in Colorado, United
States (Christmas et al., 2021, 2022). One significant finding was the
detection of the presence of a previously unknown cryptic species,
which was indistinguishable from B. lapponicus sylvicola on the basis
ofstandardmorphological characters:67of281samplesoriginally
identif ied as B. l. sylvicola formed a distin ct cluster of samp les, termed
‘incognitus’,thatwerehighlydivergentfromtheremainingsamples
(Christmas et al., 2021). However, we did not observe any distinct
clusters related to B . lapponicusin our dat ase tof129s a mpl es.Thisis
consistentwiththehypothesisthat‘incognitus’isrestrictedtoNorth
America,andwhetheritoccursoutsideofColoradoisunknown.
4.4 | Assessment of species vulnerability
ArcticandmountainousregionsofScandinaviaharbourbumblebee
species that are most threatened by climate change and are also
projected to become refugia for more European bumblebee species
infuture(Ghisbainetal.,2023). Habitat loss due to agriculture or
urbanization is not currently prevalent in these regions, although
could potentially increase in future, which could have negative ef-
fects on bumblebees (Rasmont et al., 2015). Three of the montane
species studied here, B. lapponicus, B. monticola and B. balteatus, are
found relatively commonly in the Swedish mountains. Of these, B.
monticolais alsofound at higherelevationsinthe UnitedKingdom
and in mountain ranges in mainland Europe, whereas the other spe-
ciesarerestrictedtotheScandinavianmountains.Accordingtothe
IUCN RedList, all these speciesare categorized as ‘least concern’
(IUCN,2022). However, they also fall in the climate risk category
HHR (very high risk), indicating they are predicted to lose >85%of
their current range by 2100 (Rasmont et al., 2015). This implies that
all of these species would occupy a significantly reduced range in the
Scandinavian mountains, and B. monticola would additionally lose
muchofitsrangeelsewhereinEuropeexceptthehighAlps.
Our analysis of genome variation in these three species did not
uncover any evidence for depleted levels of genetic variation that
wouldsuggestthey lie in the immediate riskzoneforexperiencing
negative genetic effects due to inbreeding or genetic drift. We find
the levels of v ariation to be lower tha n expected in B. monticola
based on its relative abundance, but no indications of population de-
cline or inbreeding. This suggests that the populations are currently
healthy, despite the prediction that they will be strongly threatened
by climate change in future.
The other three montane species, B. alpinus, B. pyrrhopy gus and B.
hyperboreus, have a much more restricted dis tribution in the Swedish
mountains, occurring at higher elevations and latitudes (Rasmont
et al., 2015). Of these, B. alpinus is also present at higher altitudes
intheAlps,whereastheothersarerestricted toScandinavia. B. al-
pinus and B. hyperboreus are catego rized as ‘vulne rable’ accord ing
to the red list, whereas B. pyrrhopygus isconsidered‘leastconcern’
(IUCN,2022). B. hyperboreus and B. pyrrhopygus are placed in the
HHHRclimateriskcategory(extremelyhighrisk),indicatingtheyare
predicted to lose >95%oftheirrangeby2100whereasB. alpinus is
placed in category HHR (very high risk) (Rasmont et al., 2015).
Amongthesethreespecies, wefindevidenceofpopulationde-
clines or inbreeding over evolutionary time. B. hyperboreus is a less
abundant species with a very restricted range and, in contrast to the
other species studied here, it has a specialized parasitic lifestyle. Our
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17251 by Yuanzhen Liu - New Aarhus University , Wiley Online Library on [21/12/2023]. 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
16 of 18
|
LIU et al.
analysis indicated that this species has become much less abundant
in the current postglacial period, and we uncovered one highly in-
bred diploid male individual among the 10 samples we collected.
This speciesis considered to be oneofthe mostvulnerable to ex-
tinction in Europe (Rasmont et al., 2015) and our analysis suggests
it may be suffering negative genetic effects of population declines,
particularly in the southern limit of its range. We also find evidence
that B. alpinus and B. pyrrhopygus exhibitlowerordecliningNE rel-
ative to the other species, consistent with their relatively restricted
abundance and distribution.
Climatechangeisthemainthreattothesixmontanebumblebee
species studied here (Ghisbainet al., 2023). Our analysis suggests
their current ranges in the Swedish mountains would be sufficient
to support healthy populations in future. However, these ranges
are predicted to shrink drastically under future scenarios of climate
change (Rasmont et al., 2015), which could lead to population bottle-
necks and fragmentation and eventually loss of all suitable habitat.
Monitoring of these p opulations using both genetic and conven-
tional methods would be advisable in order to understand how their
populations are affected by changing climate in future.
AUTHOR CONTRIBUTIONS
MatthewT.Websterdesignedtheresearch.AoifeCantwell-Jones,
Richard J. G ill, Björn Cederberg and M atthew Webster collected
samplesand performedspeciesidentification.Yuanzhen Liu, Anna
Olsson andTuuli Larva performedtheDNA extractions. Yuanzhen
Liu analy sed the data . Matthew T. Webster w rote the manusc ript
withfeedbackfromallco-authors.
ACKNO WLE DGE MENTS
This project was mainly funded by the Swedish Environmental
Protect ion Agenc y (agreement nr. 225-21-003). The pr oject made
use of facilitiesat theAbisko Scientific Research Station, which is
connectedtoSITES(SwedishInfrastructureforEcosystemScience),
INTERACT (International Network for Terrestrial Research and
MonitoringintheArctic)andSLU(SwedishUniversityofAgricultural
Sciences). Aoife Cantwell-Jones is funded by the NERC Science
andSolutionsforaChangingPlanet Doctoral Training Programme,
ImperialCollegeLondon(NE/S007415/1).Collectionswerealsosup-
portedbygrantsfromINTERACT(H2020—agreementno.730938)
an d Q u e k e t t M icroscop i c a l C l u b awardedt o R i c h a r dGill.T h e c o m p u -
tational analysis was enabled by resources provided by the Swedish
National InfrastructureforComputing(SNIC) at UPPMAX partially
funded by the Swedish Research Council through grant agreement
no.2018-05973. We alsoacknowledge supportfrom the National
Genomi cs Infrastr ucture in Stockh olm funded by Sci ence for Life
Laboratory, the Knut and Alice Wallenberg Foundation and the
Swedish Research Council. We thank the following people for help
with sample collection: Johan Lindell (Lappland), Willhelm Osterman
(Jämtland), Peter Nolbrant (Västergötland), Niklas Johansson
(Östergötland), Sven Hellqvist (Västerbotten). Turid Everitt and
MatthewChristmasassistedwithdataanalysis.
CONFLICT OF INTEREST STATEMENT
There are no conflicts of interest.
DATA AVAIL AB ILI T Y STATEMEN T
All sequence data generated by the study have been deposited at
theNationalCenterforBiotechnologyInformation(NCBI)together
withassociatedmetadata underBioProject PRJNA890771.Scripts
for bioinformatic analysis are available at GitHub: https:// github.
c o m / y z l i u 0 1 / S w e d i s h _ m o u n t a i n _ b u m b l e b e e .
ORCID
Yuanzhen Liu https://orcid.org/0000-0003-0212-0674
Matthew T. Webster https://orcid.org/0000-0003-1141-2863
REFERENCES
Alexan der, D. H., Novembr e, J., & Lange , K. (200 9).Fas t model-ba sed
estimation of ancestry in unrelated individuals. Genome Research,
19, 1655–1664. h t t p s : / /d o i . o r g / 1 0 . 1 1 0 1 / g r . 0 9 4 0 5 2 . 1 0 9
Allendorf,F.W.,Funk,W.C.,Aitken,S.N.,Byrne,M.,&Luikart,G.(2022).
Conservation and the genomics of populations. Oxford University
Press.
Allendorf, F.W.,Hohenlohe, P.A., & Luikar t, G.(2010). Genomicsand
the future of conservation genetics. Nature Reviews. Genetics, 11,
69 7–7 0 9. h t t p s : / /d o i . o r g / 1 0 . 1 0 3 8 / n r g 2 8 4 4
Cameron,S.A.,Hines,H.M.,&Williams,P.H.(2007).Acomprehensive
phyloge ny of the bumble b ees (Bomb us). Biological Journal of the
Linnean Society, 91,161–188.h t 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 7. 0 0 7 8 4 . x
Cameron,S.A.,Lozier,J.D.,Strange,J.P.,Koch,J.B.,Cordes,N.,Solter,
L. F., & Griswo ld, T.L . (2011). Patter ns of widespre ad decline in
NorthAmericanbumblebees.Proceedings of the National Academy
of Science s of the United States of Ame rica, 108, 662–667. https:// d o i .
o r g / 1 0 . 1 0 7 3 / p n a s . 1 0 1 4 7 4 3 1 0 8
Cameron,S.A.,&Sadd,B.M.(2020).Globaltrendsinbumblebeehealth.
Annual Review of Entomology, 65, 209–232. https:// doi. org/ 10.
1 1 4 6 / a n n u r e v - e n t o - 0 1 1 1 1 8 - 1 1 1 8 4 7
Ceballos,F.C.,Joshi,P.K.,Clark,D.W.,Ramsay,M.,&Wilson,J.F.(2018).
Runs of homozygosity: Windows into population history and trait
architecture. Nature Reviews. Genetics, 19, 220–234. ht tps: // doi.
o r g / 1 0 . 1 0 3 8 / n r g . 2 0 1 7 . 1 0 9
Chang,C.C., Chow,C.C.,Tellier,L.C.A.M.,Vattikuti,S.,Purcell,S.M.,&
Lee,J.J.(2015).Second-generationPLINK:Risingtothechallengeof
larger and richer datasets. GigaScience4:s13742-015-0047–8,4, 7.
Charman, T. G., Sears, J., Green, R. E., & Bourke, A. F. G. (2010).
Conservation genetics, foraging distance and nest density of the
scarce GreatYellowBumblebee (Bombus distinguendus). Molecular
Ecology, 19, 2661–2674. h t t p s : / / d o i . o r g / 1 0 . 1 1 1 1 / j . 1 3 6 5 - 2 9 4 X .
2 0 1 0 . 0 4 6 9 7 . x
Christmas, M. J., Jones, J. C., Olsson, A ., Wallerman, O., Bunikis, I.,
Kierczak, M., Peona, V., Whitley, K. M., Larva, T., Suh, A., Miller-
Struttmann,N.E.,Geib,J.C.,&Webster,M.T.(2021).Geneticbar-
riers to historical gene flow between cryptic species of alpine bum-
blebees revealed by comparative population genomics. Molecular
Biology and Evolution, 38, 3126–3143. h t t p s : / /d o i . o r g / 1 0 . 1 0 9 3 /
molbev/msab086
Christmas, M. J., Jones, J. C., Olsson, A ., Wallerman, O., Bunikis,
I., Kierczak, M., Whitley, K. M., Sullivan, I., Geib, J. C., Miller-
Strut tmann, N. E ., & Webster, M. T. (2022). A ge nomic and mo r-
phometric analysis of alpine bumblebees: Ongoing reductions in
tongue length but no clear genetic component. Molecular Ecology,
31, 1111–1127. h t t p s : / / d o i . o r g / 1 0 . 1 1 1 1 / m e c . 1 6 2 9 1
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17251 by Yuanzhen Liu - New Aarhus University , Wiley Online Library on [21/12/2023]. 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
|
17 of 18
LIU et al.
Colgan , T.J., A rce, A. N., G ill, R. J., Ram os Rodrigue s, A., Kante h, A.,
Duncan,E.J.,Li,L.,Chittka,L.,&Wurm,Y.(2022).Genomicsigna-
tures of recent adaptation in a wild bumblebee. Molecular Biology
and Evolution, 39, msab366. h t t p s : / /d o i . o r g / 1 0 . 1 0 9 3 / m o l b e v /
msab366
Danecek,P.,Auton,A.,Abecasis,G.,Albers,C.A.,Banks,E.,DePristo,M.
A.,Handsaker,R.E.,Lunter,G.,Marth,G.T.,Sherry,S.T.,McVean,
G., Dur bin, R., & 10 00 Genom es Project A nalysis Gro up. (2011).
The variant call format and VCFtools. Bioinformatics, 27, 2156–
2158.h t t p s : / /d o i . o r g / 1 0 . 1 0 9 3 / b i o i n f o r m a t i c s / b t r 3 3 0
Darvill,B.,Ellis,J. S.,Lye,G. C., &Goulson,D.(2006). Populationstruc-
ture and inbreeding in a rare and declining bumblebee, Bombus
muscorum(Hymenoptera:Apidae).Molecular Ecology, 15, 601–611.
h t t p s : // d o i . o r g / 1 0 . 1 1 1 1 / j . 1 3 6 5 - 2 9 4 X . 2 0 0 6 . 0 2 7 9 7 . x
De Meulemeester, T., Window, J., Dewulf, A., Nieto, A., Tomozii, B.,
Lhomme , P., García C riado, M., Ku hlmann, M., R adchenko, V. G.,
Pauly,A .,DeLaRúa, P.,Praz,C.,Kemp,J.,Quaranta,M.,Roberts,
S.P.M.,Dathe,H.H.,Scheuchl,E.,&Smit,J.(2016).European red
list of bees.PublicationsOf ficeoftheEuropeanUnion.h t tps:// doi.
o r g / 1 0 . 2 7 7 9 / 5 1 1 8 1
Ellis,J.S.,Knight,M.E.,Darvill,B.,&Goulson,D.(2006).Extremelylow
effective population sizes, genetic structuring and reduced genetic
diversity in a threatened bumblebee species, Bombus sylvarum
(Hymenoptera:Apidae).Molecular Ecology, 15,4375–4386.h ttps://
d o i . o r g / 1 0 . 1 1 1 1 / j . 1 3 6 5 - 2 9 4 X . 2 0 0 6 . 0 3 1 2 1 . x
Ferretti,L., & Ramos-Onsins,S. E. (2015). A generalizedWattersones-
timator for next-generation sequencing: From trios to autopoly-
ploids. Theoretical Population Biology, 10 0, 79–87.https:// doi. org/
10. 1016/j. tpb. 2015. 01. 001
Fu, Y. X. (1994). Estimati ng effect ive populat ion size or mutat ion rate
using the frequencies of mutations of various classes in a sample
ofDNA sequences. Genetics, 138,1375–1386.https:// doi. org/ 10.
1 0 9 3 / g e n e t i c s / 1 3 8 . 4 . 1 3 7 5
Ghisba in, G., Lozier, J. D., Ra hman, S. R ., Ezray, B. D., Tian , L., Ulmer,
J. M., Heraghty, S. D., Strange, J. P.,Rasmont, P.,& Hines, H. M.
(2020). Substantial genetic divergence and lack of recent gene
flow support cr yptic speciation in a colour polymorphic bumble
bee (Bombus bifarius)species complex. Systematic Entomolog y, 45,
635–652. h t t p s : / /d o i . o r g / 1 0 . 1 1 1 1 / s y e n . 1 2 4 1 9
Ghisbain,G., Thiery,W.,Massonnet, F.,Erazo,D.,Rasmont, P.,Michez,
D.,&Dellicour,S.(2023).ProjecteddeclineinEuropeanbumblebee
populationsinthetwenty-firstcentury.Nature. https:// doi. org/ 10.
1 0 3 8 / s 4 1 5 8 6 - 0 2 3 - 0 6 4 7 1 - 0
Guindon,S.,&Gascuel,O.(2003).Asimple,fast,andaccuratealgorithm
to estim ate large phylo genies by maxi mum likeliho od. Systematic
Biology, 52,696–704.
Heragh ty, S. D., Jackso n, J. M., & Lozier, J. D. (2023). W hole genome
analyses reveal weak signatures of population structure and en-
vironmentally associatedlocal adaptation in an important North
American pollinator, the bumble bee Bombus vosnesenskii.
Molecular Ecology, 32, 5479–5497. https:// doi. org/ 10. 1111/ mec.
17125
Hohenlohe,P.A.,Funk,W.C.,&Rajora,O.P.(2021).Populationgenom-
ics for wildlife conservation and management. Molecular Ecology,
30,62–82.https:// doi. org/ 10. 1111/ mec. 15720
IUCN. (2022).The IUCN Red List of Threatened Species.Version 2022-2.
https:// www. iucnr edlist. org/
Jackson,J.M., Pimsler,M.L., Oyen,K. J., Koch-Uhuad,J.B., Herndon,
J.D., Strange, J. P.,Dillon, M.E., & Lozier, J. D. (2018).Distance,
elevation and environment as drivers of diversity and divergence
in bumble bees across latitude and altitude. Molecular Ecology, 27,
2926–2942.https:// doi. org/ 10. 1111/ mec. 14735
Kent, C. F., Dey, A ., Patel, H., Tsvetkov, N., T iwari, T., MacPh ail, V. J.,
Gobeil,Y.,Harpur,B. A.,Gurtowski, J.,Schatz, M. C.,Colla,S. R.,
&Zayed,A.(2018).Conservationgenomics ofthedecliningNorth
American bumblebee Bombus terricola reveals inbreeding and
selection on immune genes. Frontiers in Gene tics, 9, 316. h t tps:// doi.
o r g / 1 0 . 3 3 8 9 / f g e n e . 2 0 1 8 . 0 0 3 1 6
Kerr,J.T.,Pindar,A.,Galpern, P.,Packer,L., Pott s,S.G.,Roberts,S.M.,
Rasmont,P.,Schweiger,O.,Colla,S.R.,Richardson,L.L.,Wagner,D.
L.,Gall,L.F.,Sikes,D.S.,&Pantoja,A.(2015).CLIMATECHANGE.
Climate change impacts on bumblebees converge across conti-
nents. Science, 349, 177–180. https:// doi. org/ 10. 1126/ scien ce.
aaa7031
Koch,J.B.,Looney,C.,Sheppard,W.S.,&Strange,J.P.(2017).Patterns
of population genetic structure and diversity across bumble bee
communi ties in the Pa cific Nor thwest. Conservation Genetics, 18,
50 7–520. h t t p s : // d o i . o r g / 1 0 . 1 0 0 7 / s 1 0 5 9 2 - 0 1 7 - 0 9 4 4 - 8
Lecocq,T.,Brasero,N.,Martinet, B.,Valterovà,I.,&Rasmont,P.(2015).
Highly polytypic taxon complex: Interspecific and intraspecific
integrative taxonomic assessment of the widespread pollina-
tor Bombus pascuorum Scopoli 1763 (Hymenoptera: Apidae).
Systematic Entomolog y, 40,881–890.https:// doi. org/ 10. 1111/ syen .
12137
Li,H.(2011).AstatisticalframeworkforSNPcalling,mutationdiscovery,
association mapping and population genetical parameter estima-
tion from sequencing data. Bioinformatics, 27,2987–2993.
Li, H. (2013). A ligning sequence reads, clone sequenc es and assembly co ntigs
with BWA- M EM.arXivpreprintarXiv:1303.3997.
Liu, H., Jia, Y.,Sun, X., Tian, D., Hurst, L. D., & Yang, S. (2017). Direct
determination of the mutation rate in the bumblebee reveals evi-
denceforweakrecombination-associatedmutationandanapprox-
imate rate constancy in insects. Molecular Biology and Evolution, 34,
119–1 3 0 . h t t p s : / / d o i . o r g / 1 0 . 1 0 9 3 / m o l b e v / m s w 2 2 6
Lozier,J.D., & Cameron, S.A .(2009).Comparativegeneticanalysesof
historical and contemporary collections highlight contrasting de-
mographic histories for the bumble bees Bombus pensylvanicus and
B. impatiensinIllinois. Molecular Ecology, 18 , 1875–1886.ht t p s://
d o i . o r g / 1 0 . 1 1 1 1 / j . 1 3 6 5 - 2 9 4 X . 2 0 0 9 . 0 4 1 6 0 . x
Lozier, J. D., Stra nge, J. P., & Herag hty, S. D. (2023). Whole gen ome
demographic models indicate divergent effective population size
histories shape contemporary genetic diversity gradients in a mon-
tane bumble bee. Ecology and Evolution., 13,e9778.https:// doi. org/
1 0 . 1 0 0 2 / e c e 3 . 9 7 7 8
Lozier,J.D.,Strange,J.P.,Stewart,I.J.,&Cameron,S.A.(2011).Patterns
of range-wide genetic variation in six North American bumble
bee(Apidae:Bombus)species.Molecular Ecology, 20, 4870–4888.
h t t p s : // d o i . o r g / 1 0 . 1 1 1 1 / j . 1 3 6 5 - 2 9 4 X . 2 0 1 1 . 0 5 3 1 4 . x
Lozier,J.D.,&Zayed,A.(2017).Beeconservationintheageofgenomics.
Conservation Genetics, 18,713–729.h t t p s : / /d o i . o r g / 1 0 . 1 0 0 7 / s 1 0 5 9
2 - 0 1 6 - 0 8 9 3 -7
Martinet, B., Lecocq, T., Brasero, N., Gerard, M., Urbanová, K.,
Valterová, I.,Gjershaug, J. O.,Michez,D.,& Rasmont,P.(2019).
Integrativet axonomy of an arctic bumblebee speciescomplex
highlights a new cr yptic species (Apidae: Bombus). Zoological
Journal of the Linnean Society, 187, 599–621. https:// doi. org/ 10.
1 0 9 3 / z o o l i n n e a n / z l z 0 4 1
Martinet, B., Rasmont, P., Cederberg, B., Evrard, D., Ødegaard, F.,
Paukkunen,J.,&Lecocq,T.(2015).Forwardtothenorth:TwoEuro-
Mediterranean bumblebee species now cross the Arctic Circle.
Annales de la Société Entomologique de France (N.S.), 51, 303–309.
h t t p s : // d o i . o r g / 1 0 . 1 0 8 0 / 0 0 3 7 9 2 7 1 . 2 0 1 5 . 1 1 1 8 3 5 7
Nadachowska-Brzyska, K., Konczal,M.,&Babik, W.(2022). Navigating
the temporal continuum of ef fective population size. Methods in
Ecology and Evolution, 13, 22–41. h t t p s : / / d o i . o r g / 1 0 . 1 1 1 1 / 2 0 4 1 -
2 1 0 X . 1 3 7 4 0
Ødegaard,F.,Staverløkk,A.,Gjershaug,J.O.,Bengtson,R.,&Mjelde,A .
(2015). Humler i Norge.Norskinstitut tfornaturforskning.
Poplin,R.,Ruano-Rubio,V.,DePristo,M.A.,Fennell,T.J.,Carneiro,M.O.,
VanderAuwera,G.A.,Kling,D.E.,Gauthier,L.D.,Levy-Moonshine,
A.,Roazen,D.,Shakir,K.,Thibault,J.,Chandran,S.,Whelan,C.,Lek,
M.,Gabriel,S.,Daly,M.J.,Neale,B.,MacArthur,D.G.,&Banks,E.
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17251 by Yuanzhen Liu - New Aarhus University , Wiley Online Library on [21/12/2023]. 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
18 of 18
|
LIU et al.
(2018).Scaling accurategeneticvariantdiscoveryto tens ofthou-
sands of samples. bioRxiv. h t t p s : // d o i . o r g / 1 0 . 1 1 0 1 / 2 0 1 1 7 8
Rasmont, P., Franzen, M., Lecocq, T., Harpke, A., Rober ts, S.,
Biesmeijer,K.,Castro,L.,Cederberg,B.,Dvorak,L.,Fitzpatrick,
U., Gonse th, Y., Haubruge, E. , Mahe, G., Mani no, A., Miche z,
D.,Neumayer,J., Odegaard,F.,Paukkunen, J., Pawlikowski,T.,
… Schweiger, O. (2015). Climatic risk and distribution atlas of
European bumblebees. Biorisk, 10, 1–236. https:// doi. org/ 10.
3 8 9 7 / b i o r i s k . 1 0 . 4 7 4 9
Saitou,N.,&Nei,M.(1987).Theneighbor-joiningmethod:Anewmethod
for reconstructing phylogenetic trees. Molecular Biology and Evolution,
4, 406–425.
Soroye, P., Newbo ld, T., & Kerr, J. (2020). Clima te change contri butes
to widespread declines among bumble bees across continents.
Science, 367,685–688.h t t p s : / / d o i . o r g / 1 0 . 1 1 2 6 / s c i e n c e . a a x 8 5 9 1
Supple,M.A .,&Shapiro,B.(2018).Conservation of biodiversityinthe
genomics era. Genome Biology, 19, 131. h t t p s : / / d o i . o r g / 1 0 . 1 1 8 6 /
s 1 3 0 5 9 - 0 1 8 - 1 5 2 0 - 3
Terhorst,J.,Kamm,J.A.,&Song,Y.S.(2017).Robustandscalableinference
of population history from hundreds of unphased whole genomes.
Nature Genetics, 49,303–309.https://doi.org/10.1038/ng.3748
Wallberg,A.,Han,F.,Wellhagen, G., Dahle,B.,Kawata,M.,Haddad,N.,
Simões,Z.L.P.,Allsopp,M.H.,Kandemir,I.,delaRúa,P.,Pirk,C.W.,
&Webster,M.T.(2014).Aworldwidesur veyofgenomesequence
variation provides insight into the evolutionary history of the hon-
eybee Apis mellifera. Nature Genetics, 46, 1081–1088 . ht t p s : // doi .
o r g / 1 0 . 1 0 3 8 / n g . 3 0 7 7
Wang, M.-S., Zhang, J.J., Guo,X., Li, M., Meyer,R., Ashari,H., Zheng,
Z.Q.,Wang, S.,Peng, M. S.,Jiang, Y., Thakur,M.,Suwannapoom,
C.,Esmailizadeh,A.,Hirimuthugoda,N.Y.,Zein,M.S.A.,Kusza,S.,
Kharrati-Koopaee,H.,Zeng,L.,Wang,Y.M.,…Zhang,Y.P.(2021).
Large- s cale genomi c analysis reve als the genet ic cost of chicken
domestication. BMC Biology, 19, 118. h t t p s : // d o i . o r g / 1 0 . 1 1 8 6 /
s 1 2 9 1 5 - 0 2 1 - 0 1 0 5 2 - x
Watterson, G. A.(1975).Onthe numberofsegregating sitesingeneti-
cal models without recombination. Theoretical Population Biology,
7, 256–276.
Webster,M.T.(2022).Assessment of genetic variation in Swedish bumble-
bees.NCBIBioProject;PRJNA890771[dataset].
Webster, M. T., Beaurepaire, A., Neumann, P., & Stolle, E. (2023).
Population genomics for insect conservation. Annual Review of
Animal Biosciences, 11, 115–140.
Whitley,K.M.(2018).Genetic structure and gene flow barriers among pop-
ulations of an alpine bumble bee (Bombus balteatus) in the central
Rocky Mountains.MSc,AppalachianStateUniversity.
Williams,P.,Colla,S., &Xie,Z.(2009).Bumblebeevulnerability:Common cor-
relates of winners and losers across three continents. Conservation Biology,
23,931–940.https://doi.org/10.1111/j.1523-1739.2009.01176.x
Williams,P.H.,Berezin,M.V.,Cannings,S.G.,Cederberg,B.,Ødegaard,
F., Rasmussen, C., Richardson, L. L., Rykken, J., Sheffield, C. S.,
Thanoo sing, C., & By valtse v,A . M. (2019). The arct ic and alpine
bumblebees of the subgenus Alpinobombus revised from inte-
grative assessment of species' gene coalescents and morphology
(Hymenoptera,Apidae,Bombus).Zootaxa, 4625, 1–68.h t t ps : // doi.
o r g / 1 0 . 1 1 6 4 6 / z o o t a x a . 4 6 2 5 . 1 . 1
Williams,P.H.,Cameron,S.A.,Hines,H.M.,Cederberg,B.,&Rasmont,
P.(2008).Asimplifiedsubgenericclassificationofthebumblebees
(genus Bombus). Apidologie, 39, 46–74. https:// doi. org/ 10. 1051/
apido: 2007052
Will ia ms,P.H.,&Osb orn e, J. L.(20 0 9).Bum ble beevu lnerabil it yan dc on-
servationworld-wide.Apidologie, 40,367–387.https:// doi. org/ 10.
1 0 5 1 /a p i d o / 2 0 0 9 0 2 5
Zayed,A.,&Packer,L.(2005).Complementary sexdeterminationsub-
stantially increases extinction proneness of haplodiploid popula-
tions. Proceedings of the National Academy of Sciences of the United
States of America, 102, 10742–10746 . https:// doi. org/ 10. 1073/
pnas. 05022 71102
SUPPORTING INFORMATION
Additional supporting information can be found online in the
SupportingInformationsectionattheendofthisarticle.
How to cite this article: Liu,Y.,Olsson,A.,Larva,T.,
Cantwell-Jones,A.,Gill,R.J.,Cederberg,B.,&Webster,M.T.
(2023).Genomicvariationinmontanebumblebeesin
Scandinavia: High levels of intraspecific diversity despite
population vulnerability. Molecular Ecology, 00, e17251.
https://doi.org/10.1111/mec.17251
1365294x, 0, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/mec.17251 by Yuanzhen Liu - New Aarhus University , Wiley Online Library on [21/12/2023]. 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