Diversification across the New World within the ‘blue’ cardinalids (Aves: Cardinalidae)

Description
Diversification across the New World within the ‘blue’ cardinalids (Aves: Cardinalidae)

Please download to get full document.

View again

of 13
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information
Category:

Environment

Publish on:

Views: 0 | Pages: 13

Extension: PDF | Download: 0

Share
Tags
Transcript
  ORIGINALARTICLE Diversification across the New Worldwithin the ‘blue’ cardinalids(Aves: Cardinalidae) Robert W. Bryson Jr 1 *, Jaime Chaves 2,3 , Brian Tilston Smith 4 ,Matthew J. Miller 5,6 , Kevin Winker 6 , Jorge L. P  erez-Em  an 7,8 and John Klicka 1 1 Department of Biology and Burke Museumof Natural History and Culture, University of Washington, Seattle, WA 98195-1800, USA,  2 Department of Biology, University of Miami,Coral Gables, FL 33146, USA,  3 Universidad San Francisco de Quito, USFQ, Colegio deCiencias Biol   ogicas y Ambientales, y Extensio´n Gala´ pagos, Campus Cumbay   a,Casilla Postal 17-1200-841, Quito, Ecuador, 4 Louisiana Museum of Natural Science,Louisiana State University, Baton Rouge, LA70803, USA,  5 Smithsonian Tropical ResearchInstitute, Panam  a, Rep  ublica de Panam  a, 6  University of Alaska Museum, Fairbanks, AK 99775, USA,  7  Instituto de Zoolog   ıa y Ecolog   ıaTropical, Universidad Central de Venezuela,Caracas 1041-A, Venezuela,  8 Colecci  onOrnitol   ogica Phelps, Caracas 1010-A,Venezuela *Correspondence: Robert W. Bryson Jr,Department of Biology and Burke Museumof Natural History and Culture, University of Washington, Box 351800, Seattle,WA 98195-1800, USA.E-mail: brysonjr@uw.edu ABSTRACT Aim  To examine the history of diversification of ‘blue’ cardinalids (Cardinali-dae) across North and South America. Location  North America (including Middle America) and South America. Methods  We collected 163 individuals of the 14 species of blue cardinalidsand generated multilocus sequence data (3193 base pairs from one mitochon-drial and three nuclear genes) to infer phylogeographical structure and recon-struct time-calibrated species trees. We then estimated the ancestral range ateach divergence event and tested for temporal shifts in diversification rate. Results  Twenty-five lineages of blue cardinalids clustered into two majorclades: one confined to North America, and a second concentrated in SouthAmerica. Blue cardinalids probably srcinated in North America, but recon-structions were influenced by how migrant taxa were assigned to biogeographi-cal regions. Most of the pre-Pleistocene divergences between extant taxaoccurred in the North American clade, whereas most divergences in SouthAmerica and adjacent Middle America occurred during the Pleistocene. Despitethese differences, the rate of diversification for both clades has been similarand relatively constant over the past 10 million years, with little geographicalexchange between North and South America outside the Panamanian isthmusregion. Main conclusions  Our reconstruction of the diversification history of bluecardinalids indicates a role of both Neogene and Quaternary events in generat-ing biotic diversity across North and South America. Although ancestral areareconstruction suggests a possible North American srcin for blue cardinalids,the occurrence of seasonal migration in this group and their relatives limitsinference. Our study highlights the importance of considering ecological andbehavioural characteristics together with palaeogeological events in order togain an understanding of the diversification history of widespread, mobile tax-onomic groups. Keywords Ancestral area, birds,  Cyanocompsa , dispersal, diversification rates, migration,Neotropics,  Passerina , phylogeography. INTRODUCTION Patterns of diversification within widespread lineages provideinsight into the broad-scale processes that generate bioticdiversity (Avise, 2000). Studies on the origins of taxa anddiversification across the New World provide examples of how geological processes (Bryson  et al. , 2012), climatechange (Weir & Schluter, 2004), niche conservatism (Smith et al. , 2012a) and the colonization of novel regions (Simp-son, 1980) influence lineage accumulation. The two conti-nents of the New World, North and South America, have anumber of important differences that impacted diversifica-tions, including connections with other landmasses (Webb,1991), geographical size (Vrba, 1992), climatic stability  ª  2013 John Wiley & Sons Ltd http://wileyonlinelibrary.com/journal/jbi  587 doi:10.1111/jbi.12218  Journal of Biogeography   (  J. Biogeogr. ) (2014)  41 , 587–599  (Hawkins  et al. , 2006) and niche availability (Ricklefs, 2002).Consequently, modes and rates of evolutionary radiations onthese continents may have differed (Rabosky & Lovette,2008; Derryberry   et al. , 2011).Much research on avian diversification in North Americahas focused on the impact of Quaternary climate change(Klicka & Zink, 1997; Avise & Walker, 1998; Johnson & Cicero, 2004). Although divergences between bird speciesare generally much older than the late Pleistocene (Klicka& Zink, 1997), many species exhibit phylogeographicalstructure that was probably generated throughout the Pleis-tocene (e.g. Klicka  et al. , 2011; Smith  et al. , 2011; van Els et al. , 2012). The overall rate of diversification in NorthAmerican birds during the Pleistocene is less well under-stood. Parts of northern North America were periodically covered with glacial ice during the Pleistocene, causing dra-matically fluctuating biotic communities (Webb & Bartlein,1992). Although habitat fragmentation promoted diversifi-cation in some birds (Weir & Schluter, 2004), decreasinghabitat availability limited ecological opportunities acrossmuch of temperate North America, potentially causing adecreasing rate of diversification (Rabosky & Lovette, 2008)or an absence of a rate increase (in part owing to higherextinction rates; Zink & Slowinski, 1995; Zink   et al. , 2004;Zink & Klicka, 2006). This absence of an increasing diver-sification rate during the Pleistocene has also been observedin several co-distributed, wide-ranging, non-avian taxa(snakes: Burbrink & Pyron, 2010; lizards: Harmon  et al. ,2003; and bats: Barber & Jensen, 2012). In Middle America(herein considered a part of the North American continent;Winker, 2011), the impact of glaciers displacing entire bio-tas was less extreme (Metcalfe  et al. , 2000). In this region,cyclical habitat fragmentation driven by Pleistocene glacial  –  interglacial cycles across a more topographically complex area may have triggered diversification in many birds(Miller  et al. , 2011; Barrera-Guzm  an  et al. , 2012; Smith et al. , 2012b).In South America, avian species diversity reaches its pin-nacle (Haffer, 1990), and the processes producing this diver-sity probably span a temporal continuum (Rull, 2011), fromthe early Cenozoic (Hawkins  et al. , 2006) to the end of thePleistocene (Ribas  et al. , 2012). A variety of historical pro-cesses and the formation of biogeographical barriers, forexample Andean uplift, Pleistocene forest fragmentation andriverine barriers, have been linked to this diversification(Hoorn  et al. , 2010; Ribas  et al. , 2012). However, the effectof Quaternary processes on the generation of avian speciesdiversity across South America is still debated (Bates  et al. ,2003; Hoorn  et al. , 2010; Smith  et al. , 2012a). Clearly, diver-sification in many South American birds occurred during thePleistocene (e.g. Patel  et al. , 2011; Guti  errez-Pinto  et al. ,2012; d’Horta  et al. , 2012; Smith  et al. , 2013), but whatremains unclear is the relative importance of processesoccurring during the past 2.6 Myr compared to eventsoccurring earlier, during the Neogene (the 20 Myr precedingthe Pleistocene).We examined diversification within a widespread group of songbirds in the ‘blue’ clade of Cardinalidae ( sensu  Klicka et al. , 2007; hereafter referred to as the blue cardinalids).The 14 species in this group collectively range across theentire New World, from Canada south to Argentina (Figs 1& 2; Orenstein & Brewer, 2011), making this an attractivespecies complex for examining diversification patterns atcontinental scales. Three species of   Passerina  ( P. amoena , P. ciris  and  P. cyanea ) breed in temperate North America(Canada south to northern Mexico), and three additional Passerina  species ( P. leclancherii ,  P. rositae ,  P. versicolor)  and Cyanocompsa parellina  are restricted to Middle America(Mexico to Nicaragua).  Amaurospiza carrizalensis ,  Amaurosp-iza moesta ,  Cyanocompsa brissonii  and  Cyanoloxia glaucocaer-ulea  are residents of South America.  Passerina caerulea  isdistributed widely across both temperate North America andMiddle America, and  Amaurospiza concolor   and  Cya-nocompsa cyanoides  range across Middle America into SouthAmerica. Seasonal migration is a life history strategy amongmany species in this group, simultaneously providing anopportunity to determine whether it might affect diversifica-tion rates and a potential problem in determining ancestralareas.To address the relative impact of Quaternary and pre-Quaternary processes on the tempo and mode of diversifica-tion in blue cardinalids, and to assess possible effects of sea-sonal migration, we combined intraspecific range-widesampling with multilocus data. We used mitochondrial DNA(mtDNA) and three nuclear loci to infer phylogeographicalstructure and estimate time-calibrated species trees. We thenreconstructed the ancestral range at each divergence eventand tested for temporal shifts in diversification rates. MATERIALS AND METHODSTaxon sampling and laboratory methods We collected 163 individuals of blue cardinalids (Klicka et al. , 2007) from throughout their distributions (Figs 1 & 2,and see Appendix S1 in the Supporting Information). Oursampling spanned the geographical distributions of all 14currently recognized species and of 20 of the 30 putativesubspecies within this group (Clements, 2007). We includedthe sister species  Spiza americana , and used  Pheucticus ludo-vicianus  and  Habia fuscicauda  as more distant outgroups(Klicka  et al. , 2007; Barker  et al. , 2012).Total genomic DNA was extracted using a DNeasy tissueextraction kit (Qiagen, Valencia, CA, USA). We sequencedthe mtDNA gene  ND2  (1038 base pairs, bp) for all 163 indi-viduals and outgroups, and for a subset of specimens( n  =  37) representing the main lineages inferred from ourcomplete mtDNA data (see below). We also sequenced threenuclear introns, including 951 bp of aconitase 1 intron 9(  ACO1 ), 685 bp of myoglobin intron 2 (  MYC  ), and 519 bpof beta-fibrinogen intron 5 ( FGB-I5 ). Laboratory protocolsand primers follow Smith & Klicka (2013).  Journal of Biogeography   41 , 587–599 ª  2013 John Wiley & Sons Ltd 588 R. W. Bryson Jr  et al.  We edited and manually aligned forward and reversesequences using  Sequencher  4.2 (Gene Codes Corporation,Ann Arbor, MI, USA).  InDelligent  1.2 (Dmitriev & Rakitov,2008) was used to resolve indels between homologous nuclearalleles. The gametic phase of heterozygous variants was deter-mined using  phase  2.1.1 (Stephens & Donnelly, 2003). Foreach nuclear dataset, separate runs of 400 iterations each wereperformed, accepting results with probability   ≥  0.7. All poly-morphic sites with probability   <  0.7 were coded in both alleleswith the appropriate IUPAC ambiguity code. The  ACO1 intron is sex-linked on the  Z  -chromosome, and so only malescontain two alleles. We therefore cross-checked sequencesagainst the sexes of our vouchered specimens (Appendix S1).Male specimens were retained with biallelic states, whereasphased  ACO1  sequences from females and specimens of unknown sex were reduced to only one allele. Sequences weredeposited in Dryad (see Data Accessibility below). Phylogeographical estimation To assess range-wide genetic structure and delineate geo-graphical lineages, we generated a mtDNA phylogeny of allindividuals ( n  =  166 including outgroups) using  MrBayes 3.2.1 (Ronquist  et al. , 2012). Lineages were defined as geneti-cally distinct geographical clusters with strong support values( ≥  0.95 Bayesian posterior probability; Huelsenbeck & Rann-ala, 2004), consistent with the term ‘phylogroup’ (Avise & Walker, 1998; Avise  et al. , 1998; Rissler & Apodaca, 2007).Single divergent samples from unique geographical areaswere also referred to as lineages for convenience. We imple-mented two separate models for the first plus second and forthe third codon positions.  jModelTest  0.1.1 (Posada, 2008)was used to select the best-fit model of evolution, based onthe Akaike information criterion (AIC), for each partition(1st + 2nd: HKY  +  I  +  G; 3rd: GTR   +  I). Four runs were con-ducted using the ‘nruns  =  4’ command, each with threeheated and one cold Markov chain sampling every 100 gen-erations for 10 million generations. Changing the tempera-ture of the heated chain from the default value of 0.1 to 0.03improved the efficiency of the Metropolis coupling. Outputparameters were visualized using  Tracer  1.4 (Rambaut & Drummond, 2007) to ascertain stationarity and convergence.We further assessed convergence between runs using  awty (Nylander  et al. , 2008). The first 25% of generations werediscarded as burn-in. Species trees and divergence times We used * beast  (Heled & Drummond, 2010), part of the beast  1.7.4 package (Drummond & Rambaut, 2007), to amoena 3 NV caerulea 5 HON amoena 2 ID caerulea 4 MS amoena 1 OR amoena 6 MX amoena 5 MX caerulea 1 NV caerulea 3 LA caerulea 2 OK amoena 4 NV leclancherii   3 MEX versicolor   5 MEX versicolor   2 MEX rositae 2 MEX leclancherii   4 MEX leclancherii   2 MEX ciris  5 TX versicolor 7 MEX cyanea  6 NC ciris  1 LA ciris  7 FL ciris  4 NC cyanea  2 MN cyanea  3 KS rositae 1 MEX leclancherii   6 MEX versicolor   4 MEX cyanea  1 WI versicolor   3 MEX versicolor   6 MEX cyanea  4 WV ciris  6 GA versicolor   1 MEX leclancherii   5 MEX leclancherii   1 MEX ciris  2 LA ciris  3 OK cyanea  5 AR  parellina  17 MEX  parellina  2 MEX  parellina  1 MEX  parellina  6 MEX  parellina  7 MEX  parellina  23 MEX  parellina  15 MEX  parellina  18 MEX  parellina  16 MEX  parellina  8 MEX  parellina 11 MEX  parellina  5 HON  parellina  14 MEX  parellina  4 SAL  parellina  12 MEX  parellina  13 MEX  parellina  20 MEX  parellina  9 MEX  parellina  21 MEX  parellina  3 SAL  parellina  19 MEX Passerinarositae  AB ABCC lechlancherii ciriscaeruleaamoenaPasserina cyaneaCyanocompsa parellinaPasserinaCyanocompsa parellina  ACCCBB AC A A    A    ACCCCCCCCBB   CC P. caerulea PP ≥ 0.95PP ≥ 0.70PP ≥ 0.50 Passerina cyaneaPasserina lechlancherii Cyanocompsa parellinaPasserina versicolor Passerina amoenaPasserina rositae Passerina cirisversicolor Passerina Figure 1  The blue cardinalids of North America. The map on the right shows the breeding distributions of the eight currently recognized species in North America (from Ridgely   et al. , 2007) and the localities of the specimens sampled. The mitochondrialphylogeny on the left was used to infer phylogeographical structure, and major geographical lineages are noted. Bayesian posteriorprobability (PP) support for nodes is indicated by coded dots. Multilocus data were generated from specimens marked with an arrow.Additional locality data can be found in Appendix S1.  Journal of Biogeography   41 , 587–599 ª  2013 John Wiley & Sons Ltd 589 Diversification of the ‘blue’ cardinalids  reconstruct a time-calibrated multilocus species tree using 1  –  4individuals ( n  =  37) from each currently recognized speciesor phylogeographical lineage inferred from our completemtDNA data and the three outgroups. We used a subset of individuals in order to capture genetic diversity but reducecomputational burden and sequencing costs. Best-fit modelsof evolution were estimated using  jModelTest  (  ND2 :GTR   +  I  +  G;  ACO1 ,  FGB-I5 : GTR   +  G;  MYC  : HKY  +  G).We used a Yule speciation prior and relaxed uncorrelatedlognormal clock for each gene tree. To calibrate our speciestree, we used the  ND2  substitution rate of 1.25  9  10  2 sub-stitutions/site/Myr (2.5% change between lineages per Myr)from Smith & Klicka (2010), and of 1.35  9  10  3 substitu-tions/site/Myr for autosomal (  MYC  ,  FGB-I5 ) and1.45  9  10  3 substitutions/site/Myr for sex-linked (  ACO1 )intron rates (Ellegren, 2007). We specified a lognormal concolor 3 MEX concolor 2 GUA concolor   1 SAL carrizalensis  3 VEN moesta 2 BRA carrizalensis  1 VEN moesta 1 BRA concolor 4 ECU carrizalensis 2 VEN cyanoides 46 BEL cyanoides 65 PAN cyanoides 56 MEX cyanoides 45 BEL cyanoides  59 MEX cyanoides  48 BEL cyanoides  41 MEX cyanoides  44 MEX cyanoides 77 CR cyanoides  11 HON cyanoides 40 MEX cyanoides  22 MEX cyanoides  42 MEX cyanoides 58 MEX cyanoides  18 PAN cyanoides 28 MEX cyanoides  29 MEX cyanoides 49 BEL cyanoides  66 PAN cyanoides  57 MEX cyanoides 30 BEL cyanoides  75 VEN cyanoides  43 MEX cyanoides 47 BEL cyanoides  15 VEN cyanoides 16 BEL cyanoides  99 VEN cyanoides  2 HON cyanoides 98 VEN cyanoides  8 NIC cyanoides 17 BEL cyanoides  5 MEX cyanoides  74 VEN cyanoides  76 CR cyanoides  39 PAN cyanoides 31 BEL cyanoides 7  8 CR cyanoides 20 PAN cyanoides 33 PAN cyanoides 21 PAN cyanoides 19 PAN cyanoides 68 PAN cyanoides 50 PAN cyanoides 32 PAN cyanoides 38 PAN cyanoides 97 VEN cyanoides 67 PAN cyanoides 52 PAN cyanoides 36 PAN cyanoides 34 PAN cyanoides 69 PAN cyanoides 35 PAN cyanoides 54 PAN cyanoides  55 PAN cyanoides 71 PAN cyanoides 73 VEN cyanoides 51 PAN cyanoides 70 PAN cyanoides 53 PAN cyanoides 37 PAN cyanoides 25 ECU brissonii 3 BRA brissonii   1 BOL brissonii 4 BRA brissonii 5 PAR brissonii 10 VEN brissonii 7 ARG brissonii 9 BOL brissonii 6 ARG brissonii 8 ARG brissonii 2 PAR Cyanoloxia 1 URU Cyanoloxia 2 PAR cyanoides 72 BOL cyanoides 3 PER cyanoides 9 VEN cyanoides 7 PER cyanoides 6 PER cyanoides 62 GUY cyanoides 60 ECU cyanoides 24 BRA cyanoides 1 BOL cyanoides 10 GUY cyanoides 14 VEN cyanoides 64 ECU cyanoides 12 GUY cyanoides 22 PER cyanoides 63 GUY cyanoides 4 PER cyanoides 13 VEN cyanoides 23 BRA cyanoides 61 GUY cyanoides 26 ECU carrizalensis  4 VEN  Amaurospiza  ABDEF ABC concolor  A concolor B moestaCyanocompsa cyanoidesCyanoloxiaCyanocompsa brissonii  AmaurospizaCyanocompsa cyanoidesCyanocompsa brissonii Cyanoloxia PP ≥ 0.95PP ≥ 0.70 carrizalensisCyanocompsa cyanoidesCyanocompsa brissonii Cyanoloxia glaucocaerulea Amaurospiza concolor  Amaurospiza moesta Amaurospiza carrizalensis  A A AB   B A A A A A ABBBBBBBBBCC   CC   B ACBDDEEEEDD   DFDEEB A A ACDEE Cyanocompsa cyanoides Figure 2  The blue cardinalids of South America and adjacent countries. The map on the right shows the breeding distributions of thesix currently recognized species in this region (from Ridgely   et al. , 2007) and the localities of the specimens sampled. The mitochondrialphylogeny on the left was used to infer phylogeographical structure, and major geographical lineages are noted. Bayesian posteriorprobability (PP) support for nodes is indicated by coded dots. Multilocus data were generated from specimens marked with an arrow.Additional locality data can be found in Appendix S1.  Journal of Biogeography   41 , 587–599 ª  2013 John Wiley & Sons Ltd 590 R. W. Bryson Jr  et al.  distribution and a relatively wide logarithmic standard devia-tion of 0.2 for each gene, thus encompassing alternative sub-stitution rates (e.g. Lerner  et al. , 2011). Analyses were runfor 100 million generations and sampled every 1000 genera-tions.  Tracer  was used to confirm acceptable mixing andlikelihood stationarity, appropriate burn-in, and adequateeffective sample sizes ( >  200). After discarding the first 10million generations (10%) as burn-in, the parameter valuesof the samples from the posterior distribution were summa-rized on the maximum clade credibility tree using TreeAnnotator  1.7.4 (Drummond & Rambaut, 2007).We reconstructed a second species tree from the fullmtDNA dataset to estimate divergences based on all 163specimens of blue cardinalids and the three outgroups.Although mtDNA provides only a single-gene estimate of aspecies tree, * beast  can estimate divergence times frommtDNA data that account for gene divergences that may pre-date species divergences (Heled & Drummond, 2010).This approach allowed us to run biogeographical analysesusing time-calibrated species trees estimated from two dis-tinct datasets. Following the same approach as outlinedabove, each specimen was assigned to a currently recognizedspecies or geographical lineage inferred from our mtDNAgene tree. We calibrated the tree using the same  ND2  substitu-tion rate of 1.25  9  10  2 substitutions/site/Myr, and analyseswere run for 100 million generations.  Tracer  was used toconfirm acceptable mixing and stationarity, appropriateburn-in, and adequate effective sample sizes. The first 10million generations were discarded as burn-in, and the pos-terior parameter values were summarized on the maximumclade credibility tree using  TreeAnnotator . Ancestral area reconstruction The ancestral range at each divergence event was recon-structed using Bayesian binary Markov chain Monte Carlo(MCMC) analysis (BBM) as implemented in  rasp  2.0b (Yu et al. , 2011). We ran independent analyses on 40,000 post-burn-in trees produced from our multilocus species tree andmtDNA species tree reconstructions to account for phyloge-netic uncertainty and topological differences between thesetwo phylogenies. Each sample from our species trees wasassigned to one or more of three broad biogeographicalregions (following Winker, 2011): (A) temperate NorthAmerica (including the northern Chihuahuan Desert), (B)Middle America, and (C) South America (including centraland eastern Panama). The probabilities for nodes in eachphylogeny with  >  0.90 posterior probability (PP) were esti-mated to incorporate information from most nodes (17/25)of the tree but minimize phylogenetic ‘noise’ from poorly supported relationships. The number of areas was set to 3, aF81  +  G model (the most complex model available in  rasp )was used, and analyses were conducted for 1 million genera-tions using 10 chains, sampling every 100 generations.  Spizaamericana  was set as the outgroup. The first 25% of genera-tions were discarded as burn-in.Attempts to infer the srcin of migratory bird species havebeen contentious (Zink, 2002; Joseph, 2005). Consistent withthe hypothesis that Cardinalidae is derived from a NorthAmerican ancestor (Barker  et al. , 2004; Klicka  et al. , 2007),we chose to assign migratory species to biogeographical areasbased on their breeding distributions. However, this assumesthat the distributions of blue cardinalids have been staticthrough time, and that current breeding distributions reflectareas of species origins. To test these assumptions, we rantwo additional analyses using the multilocus dataset in  rasp (Appendix S2). In the first of these, we assigned winteringareas rather than breeding areas as points of origin. In thesecond, we redefined areas of srcin to include both breedingand wintering areas. Finally, because outgroup choice andtheir distributions can influence ancestral area reconstruc-tions, we ran one more analysis using the multilocus datasetto assess the effect of including four more outgroups onancestral area reconstructions (Appendix S2). Diversification rates We analysed temporal shifts in diversification rates withinblue cardinalids using maximum likelihood-based diversifica-tion-rate analysis (Rabosky, 2006a) and divergence dates esti-mated from both the multilocus and the mtDNA datasets.The fits of different birth  –  death models implementing twoconstant rates (pure birth and birth  –  death) and three vari-able rates (exponential and logistic density-dependent andtwo-rate pure birth) were computed with  laser  2.3 (Rabo-sky, 2006b). Model fit was measured using AIC scores, andthe significance of the change in AIC scores between the bestrate-constant and best rate-variable model was determinedthrough simulations implemented in  laser . Log-likelihoodand AIC values were calculated for three models (SPVAR,EXVAR and BOTHVAR; Rabosky & Lovette, 2008) that per-mit differential extinction and speciation rates. We repeatedthe analyses and generated lineage-through-time (LTT) plotsto visualize the tempo of diversification within the two majorgeographical clades (North American and predominantly South American) inferred from our multilocus species treereconstructions. RESULTSPhylogeographical estimation Within blue cardinalids, we identified two major mtDNAclades with 25 evolutionary lineages (Figs 1 & 2). A northern‘North American’ clade contained all  Passerina  species and asingle member of   Cyanocompsa  ( C. parellina ; Fig. 1), render-ing  Cyanocompsa  polyphyletic. Support for this clade wasweak (PP  =  0.56). All  Passerina  species lacked strong phylo-geographical structure, with the exception of   P. versicolor  ,which formed three geographical lineages. The Middle Amer-ican endemic  Cyanocompsa parellina  was also structured intothree geographically isolated lineages. Considerably more  Journal of Biogeography   41 , 587–599 ª  2013 John Wiley & Sons Ltd 591 Diversification of the ‘blue’ cardinalids
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks