pika genetics and adaptation

1
The American pika (Fig. 2; Ochotona princeps) is a small lagomorph that inhabits talus slopes and rocky habitats throughout mountainous regions of western North America. • In the northern most part of their range, the central Coast Mountains of BC (Fig. 1), pikas are present along altitudinal gradients ranging from sea level to above 1500m, making them an ideal system to investigate the genetic basis of local adaptation to changing environmental conditions. Investigating the genetic basis of local adaptation in American pika using genomic scans Philippe Henry and Michael A. Russello Department of Biology and Centre for Species at Risk and Habitat Studies (SARAHS), University of British Columbia Okanagan, Kelowna, BC, Canada. Correspondence: [email protected] Results Context Objective Discussion References • Alpine species are starting to show responses to global environmental changes. One obvious example is the upslope movement of species as they seek to retain their optimal niches. • Besides the ability to migrate, species with special habitat requirements may be constrained to stay put. These species will have to respond to the environmental changes by adapting to the changing conditions. • Their responses to environmental changes will largely depend on the underlying adaptive genetics variability present in its populations. • To date the task of identifying genomic regions under selection from environmental changes has been a difficult one, mostly owing to the lack of genetic markers for natural populations. • Technical advances that have lead to the identification of a large number of genetic markers along with statistical developments have allowed researchers to tease apart adaptive from neutral genetic variation by comparing levels of genetic differentiation between loci, where outlier loci with high differentiation are interpreted as been subject to natural selection. Study Species The advent of molecular markers such as AFLP have enabled researcher to study the adaptive genetic variation in wild, non‐model species. In the present study we show that a small fraction of loci found in pikas are potentially under natural selection, while most loci are selectively neutral. These loci were only present in samples from high (A, C, E) and low (B, D) elevations but absent from the mid elevation site. This point suggests that the outlier loci are responsible for the local adaptation to environmental conditions that vary with altitude. Yet it is important to bear in mind that these results are preliminary as they were obtained from only 32 samples. Besides all advantages of using AFLPs, one drawback is that this type of markers are anonymous and thus it is difficult to pinpoint the genes or genomic regions under selection. Other complementary approaches (SAM) can also be used to shed light on the association between loci and eco‐climatic variables. This latter approach will be implemented in the future on a larger dataset, and will hopefully identify the main driver of selection in American pikas from the Bella Coola valley. Figure 2: An American pika emerging from the rocks in the Bella Coola valley, British Columbia (Philippe Henry Photo) Methods Figure 1: Map of the location of the study site: the Bella Coola Valley, British Columbia, Canada. This site is of particular interest since pikas are distributed along an altitudinal gradient ranging from the bottom of the valley at 300m, up to 1500m over very short geographical distances. The highway descending from the Chilcotin plateau into the Bella Coola valley offers ideal access to populations at different altitudes. • Use a genomic scan approach to identify candidate loci under natural selection that may be responsible for local adaptation in the American pika. A. Samples • Pika populations were sampled from five populations at three different elevations: 300m (N=3), 800m (N=1) and 1500m (N=1) above sea level in the Bella Coola Valley, British Columbia, Canada. • Non‐invasive snares 1 were used to collect hair from 32 individuals. B. Lab work • DNA was extracted from 25 hairs using the Promega DNA IQ system. • A genomic scan was undertaken using AFLPs. The protocol included digestion of 50ng of DNA with EcoRI and TaqI followed by ligation of adapters, pre‐selective amplification and selective amplification with a total of 64 primer combinations 2 . • 44 primer combinations produced bands that were run onto an ABI 3130XL genotyper and visualized using Genemapper 3.7. C. Analyzes • The software scanAFLP 3 was used to select the most repeatable and informative primer combinations. 15 primer combinations with high signal to noise ratios and low error rates were thus retained. Furthermore, loci with less than 5% and more than 95% presence were removed from the dataset. • The program Bayescan 4 which uses a Bayesian approach to directly calculate the probability of been under selection for each locus was used to detect candidate loci potentially under natural selection. Figure 3: Bayescan analysis graphic with, on the x‐axis, the posterior probability for a locus to be under natural selection. The vertical line indicates the threshold beyond which there is strong evidence for selection to act. 5 out of 1238 loci show strong evidence for selection: A ‐ E32T35_115, B – E43T44_123, C – E43T35_83, D – E31T37_96, E – E43T35_73. 1. Henry & Russello (Submitted) 2. Bonin et al., 2005 3. Herrmann et al., 2010 4. Foll & Gagiotti, 2008 Acknowledgements We would like to thank the members of the ECGL for help in the field and lab, Kurt Galbreath and Mary Peacock for kindly providing samples. Sarka Jahodova is thanked for providing guidance with AFLP. A total of 1238 loci were obtained with 15 selective primer combinations that proved to be informative and repeatable. Bayescan analysis detected that 5 of the 1238 loci (0.4%) showed strong evidence for natural selection in our populations sampled along an altitudinal gradient (Fig. 3). Funding

Upload: philippe-henry

Post on 14-Apr-2017

100 views

Category:

Environment


0 download

TRANSCRIPT

Page 1: Pika genetics and adaptation

•  The American pika (Fig. 2; Ochotona princeps) is a smalllagomorphthatinhabitstalusslopesandrockyhabitatsthroughoutmountainousregionsofwesternNorthAmerica.•  In the northern most part of their range, the central CoastMountains of BC (Fig. 1), pikas are present along altitudinalgradientsrangingfromsealeveltoabove1500m,makingthemanidealsystemtoinvestigatethegeneticbasisoflocaladaptationtochangingenvironmentalconditions.

Investigating the genetic basis of local adaptation in American pika using genomic scans

Philippe Henry and Michael A. Russello Department of Biology and Centre for Species at Risk and Habitat Studies (SARAHS), University of British Columbia Okanagan, Kelowna, BC, Canada.

Correspondence: [email protected] Results Context

Objective

Discussion

References

•  Alpine species are starting to show responses to globalenvironmental changes. One obvious example is the upslopemovementofspeciesastheyseektoretaintheiroptimalniches.•  Besides the ability to migrate, species with special habitatrequirements maybeconstrainedto stay put. Thesespecies willhavetorespondtotheenvironmentalchangesbyadaptingtothechangingconditions.•  Theirresponsestoenvironmentalchangeswilllargelydependonthe underlying adaptive genetics variability present in itspopulations.•  Todatethetaskofidentifyinggenomicregionsunderselectionfromenvironmentalchangeshasbeenadifficultone,mostlyowingtothelackofgeneticmarkersfornaturalpopulations.•  Technicaladvancesthathaveleadtotheidentificationofalargenumberofgeneticmarkersalongwithstatisticaldevelopmentshaveallowedresearcherstoteaseapart adaptivefromneutral geneticvariation by comparing levels of genetic differentiation betweenloci,whereoutlierlociwithhighdifferentiationareinterpretedasbeensubjecttonaturalselection.

Study Species

•  The advent of molecular markers such as AFLP have enabledresearchertostudytheadaptivegeneticvariationinwild,non‐modelspecies.• Inthepresentstudyweshowthatasmallfractionoflocifoundinpikas are potentially under natural selection, while most loci areselectivelyneutral.• Theselociwereonlypresentinsamplesfromhigh(A,C,E)andlow(B,D)elevationsbutabsentfromthemidelevationsite.• Thispointsuggeststhattheoutlierlociareresponsibleforthelocaladaptationtoenvironmentalconditionsthatvarywithaltitude.Yetitis importanttobear inmindthattheseresults arepreliminaryastheywereobtainedfromonly32samples.• Besidesall advantagesofusingAFLPs, onedrawbackis thatthistypeofmarkersareanonymousandthusitisdifficulttopinpointthegenesorgenomicregionsunderselection.• Othercomplementaryapproaches(SAM)canalsobeusedtoshedlightontheassociationbetweenlociandeco‐climaticvariables.• Thislatterapproachwillbeimplementedinthefutureonalargerdataset, andwill hopefully identifythemaindriverof selection inAmericanpikasfromtheBellaCoolavalley.

Figure2:AnAmericanpikaemergingfromtherocksintheBellaCoolavalley,BritishColumbia(PhilippeHenryPhoto)

Methods

Figure 1: Map of thelocationofthestudysite:the Bella Coola Valley,BritishColumbia,Canada.This site is of particularinterest since pikas aredistributed along ana l t i tud ina l grad ientranging fromthebottomofthevalleyat300m,upto1500moververyshortgeographical distances.The highway descendingfromtheChilcotinplateauintotheBellaCoolavalleyoffers ideal access topopulations at differentaltitudes.

• Use a genomic scanapproach to identify candidate loci undernaturalselectionthatmayberesponsibleforlocaladaptationintheAmericanpika.

A.Samples• Pika populations were sampled from five populations at threedifferent elevations: 300m (N=3), 800m (N=1) and 1500m (N=1)abovesealevelintheBellaCoolaValley,BritishColumbia,Canada.• Non‐invasivesnares1wereusedtocollecthairfrom32individuals.B.Labwork• DNA was extracted from25 hairs using the Promega DNA IQsystem.•  A genomic scan was undertaken using AFLPs. The protocolincludeddigestionof50ngofDNAwithEcoRIandTaqIfollowedbyligation of adapters, pre‐selective amplification and selectiveamplificationwithatotalof64primercombinations2.• 44primercombinationsproducedbandsthatwererunontoanABI3130XLgenotyperandvisualizedusingGenemapper3.7.C.Analyzes• ThesoftwarescanAFLP3wasusedtoselectthemostrepeatableandinformativeprimercombinations.15primercombinationswithhighsignaltonoiseratiosandlowerrorrateswerethusretained.Furthermore,lociwithlessthan5%andmorethan95%presencewereremovedfromthedataset.•  The program Bayescan4 which uses a Bayesian approach todirectlycalculatetheprobabilityofbeenunderselectionforeachlocuswasusedtodetectcandidatelocipotentially undernaturalselection.

Figure 3: Bayescananalysis graphic with,on the x‐axis, theposteriorprobabilityfora locus to be undernatural selection. Thevertical line indicatesthe threshold beyondwhich there is strongevidence for selectiontoact.5outof1238locishow strong evidenceforselection:A‐E32T35_115,B–E43T44_123,C–E43T35_83,D–E31T37_96,E–E43T35_73.

1.  Henry&Russello(Submitted)2.  Boninetal.,20053.  Herrmannetal.,20104.  Foll&Gagiotti,2008

AcknowledgementsWewouldliketothankthemembersoftheECGLforhelp

inthefieldandlab,KurtGalbreathandMaryPeacockfor kindly providing samples. Sarka Jahodova isthanked for providing guidance with AFLP.

•  A total of 1238 loci were obtained with 15 selective primercombinationsthatprovedtobeinformativeandrepeatable.• Bayescananalysisdetectedthat5ofthe1238loci(0.4%)showedstrong evidence for natural selection in our populations sampledalonganaltitudinalgradient(Fig.3).

Funding