Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Phylogenetics and introgression of Habronattus jumping spiders using transcriptomes (Araneae:Salticidae) Leduc-Robert, Geneviève 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


24-ubc_2016_february_leducrobert_genevieve.pdf [ 2.26MB ]
JSON: 24-1.0221354.json
JSON-LD: 24-1.0221354-ld.json
RDF/XML (Pretty): 24-1.0221354-rdf.xml
RDF/JSON: 24-1.0221354-rdf.json
Turtle: 24-1.0221354-turtle.txt
N-Triples: 24-1.0221354-rdf-ntriples.txt
Original Record: 24-1.0221354-source.json
Full Text

Full Text

PHYLOGENETICS AND INTROGRESSION OF  HABRONATTUS JUMPING SPIDERS USING TRANSCRIPTOMES (ARANEAE: SALTICIDAE)   by    Geneviève Leduc-Robert    B.Sc., The University of British Columbia, 2011   A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF   MASTER OF SCIENCE  in   The Faculty of Graduate and Postdoctoral Studies  (Zoology)      THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)      December 2015     ! Geneviève Leduc-Robert, 2015 ! ""!Abstract Habronattus is a diverse genus of jumping spiders with complex courtship displays and colourful ornaments in males. A well-resolved species phylogeny would provide an important framework to study these traits, but has not yet been achieved because of conflicting signals from the few genes available. While such discordant gene trees could be the result of deep coalescence in the recently diverged group, there are many indications that hybridization may have occurred and could be the source of conflict. To infer Habronattus phylogenetic relationships and to investigate the cause of gene tree discordance, we assembled transcriptomes for 34 Habronattus species and 2 outgroups. We conducted a concatenated phylogenetic analysis using Maximum Likelihood for 2.41 Mb of nuclear data and for 12.33 kb of mitochondrial data. The concatenated nuclear phylogeny was resolved with high bootstrap support (95-100%) at most nodes with some uncertainty surrounding the relationships of H. icenoglei, H. cambridgei, and H. oregonensis, and Pellenes cf. levii. There are several nodes of the mitochondrial phylogeny that are incongruent to the nuclear phylogeny and indicate possible mitochondrial introgression: the internal relationships of the americanus and the coecatus group, the relationship between the altanus, decorus, banksi, and americanus group, and between H. clypeatus and the coecatus group. To determine the extent of incomplete lineage sorting (ILS) and introgression, we analyzed gene tree discordance for loci longer than 1 kb using Bayesian Concordance Analysis (BCA) for the americanus group (679 loci) and the viridipes/clypeatus/coecatus (VCC) group (517 loci) and found high levels of genetic discordance, especially in VCC group. Finally, we tested specifically for nuclear introgression in the concatenated nuclear matrix with Patterson’s D statistics and DFOIL. We found nuclear introgression resulting in substantial admixture between americanus group species, and between H. sp. (ROBRT) and the clypeatus group, and more ! """!minimal nuclear introgression between the clypeatus group and the coecatus group, and between the americanus group and several distant species. Our results indicate that hybridization may have been historically common between phylogenetically distant species of Habronattus, and that reproductive isolation is yet to be complete across the Habronattus phylogeny.                  ! "#!Preface The collection of specimens was conducted as a team led by W.P. Maddison in Arizona and Mexico. Other specimens were already collected by David Maddison, Junxia Zhang, Damian Elias, and Marshal Hedin. I designed the study with the help of W.P. Maddison. I completed all the lab work and data analysis, except for the Bioanalyzer step completed by Joanne Denny, the cDNA library preparation completed by Debbie Adams and Anastasia Kuzmin and the Illumina sequencing step competed by Anastasia Kuzmin. There were steps in alignment and sequence processing completed with custom modules, designed by W.P. Maddison, in the Mesquite Gataga package. I prepared the manuscript with input from W.P. Maddison.             ! #!Table of contents Abstract .......................................................................................................................................... ii Preface .......................................................................................................................................... iv Table of contents ............................................................................................................................ v List of tables ..........................................................................................................................….. vii List of figures .............................................................................................................................. viii Acknowledgements ....................................................................................................................... ix 1. Introduction ................................................................................................................................ 1  2. Methods ...................................................................................................................................... 6 2.1 Taxon sampling ............................................................................................................ 6  2.2 Molecular extractions and sequencing ......................................................................... 7 2.3 Sequence read filtering and trimming .......................................................................... 8 2.4 Reference transcriptome .............................................................................................. 9 2.5 Reference-based assembly of transcriptomes ............................................................ 10 2.6 Sequence alignments .................................................................................................. 11 2.7 Phylogenetic analyses ................................................................................................ 13 2.7.1 Nuclear and mitochondrial matrices matrices ............................................. 13 2.7.2 16SND1 and Ef1-a phylogenies with broader species sample ................... 15 2.8 Introgression .............................................................................................................. 15  2.8.1 Bayesian concordance analysis ................................................................... 15 2.8.2 Patterson’s D statistic and DFOIL .................................................................. 17 3. Results ...................................................................................................................................... 21 3.1 Transcriptome assemblies .......................................................................................... 21 3.1.1 Reference transcriptome ............................................................................. 21  3.1.2 Reference-based transcriptomes assemblies ............................................... 22 3.2 Substitution model selection ...................................................................................... 22  3.3 Nuclear phylogenetic analyses ................................................................................... 23 3.4 Mitochondrial phylogenetic analyses ........................................................................ 26 3.5 16SND1 and Ef1-a phylogenies (broader species sample) ........................................ 27 3.6. Introgression ............................................................................................................. 28   3.6.1 Bayesian concordance analysis ................................................................... 28 3.6.2 Patterson’s D statistic and DFOIL .................................................................. 30 4. Discussion ................................................................................................................................ 33  4.1 Habronattus phylogeny ............................................................................................. 33 4.2 Introgression in Habronattus ..................................................................................... 35 4.2.1 Introgression within species groups ............................................................ 38  4.2.2 Distant introgression ................................................................................... 41  4.2.3 Potential drivers of hybridization ................................................................ 46 4.2.4 Evolutionary consequences of introgression .............................................. 49  4.2.5 Future research ............................................................................................ 51 5. Conclusions .............................................................................................................................. 53 Tables and figures ........................................................................................................................ 54 Bibliography ................................................................................................................................ 69 Appendices ................................................................................................................................... 76  1 List of specimens............................................................................................................76  ! #"!2 Summary of sequencing and read assemblies ............................................................... 81  3 Concordance factors and their credibility intervals for the americanus group ............. 86 4 Concordance factors and their credibility intervals for the VCC clade ........................ 87  5 Counts of biallelic patterns for DFOIL ............................................................................. 90 6 Counts of biallelic patterns for Patterson’s D statistics ................................................ 91 7 DFOIL results ................................................................................................................... 92  8 D-statistics results ......................................................................................................... 93                                         ! #""!List of tables  Table 1. Sets of species tested using D-statistics and DFOIL ......................................................... 54                    ! #"""!List of figures Figure 1. Flow chart of reference transcriptome filtering steps ................................................... 55 Figure 2. Flow chart of data partitioning scheme for phylogenetic analysis ............................... 56 Figure 3. Concatenated nuclear phylogeny................................................................................... 57 Figure 4. Variable nodes for Pellenes cf. levii, H. icenoglei, H. oregonensis, H. cambridgei .... 58 Figure 5. Variable nodes for H. virgulatus, H. ophrys, H. sp. (ROBRT) .................................... 59 Figure 6. Concatenated mitochondrial phylogeny ....................................................................... 60 Figure 7. 16SND1 phylogeny ...................................................................................................... 61 Figure 8. Ef1-a phylogeny ........................................................................................................... 62 Figure 9. Americanus group primary concordance tree ............................................................... 63 Figure 10. VCC clade primary concordance tree ......................................................................... 64 Figure 11. D-statistics and DFOIL results for americanus group ................................................... 65 Figure 12. DFOIL results for clypeatus and coecatus group ........................................................... 66 Figure 13. D-statistics results for H. sp. (ROBRT) group ........................................................... 67 Figure 14. D-statistics and DFOIL results for H. decorus ............................................................... 68              ! "$!Acknowledgements I’d like to thank Wayne Maddison for his insights, encouragement, and contagious enthusiasm. His expertise on Habronattus and phylogenetics were instrumental to this thesis. Beyond his help with this project, I look to him as a role model of what a scientist, mentor, and educator can be. Thank you for setting such a strong example.  A special thanks to Heather Proctor, Sam Evans, and Abraham Meza for help collecting specimens for this project in August 2013, and to Marshal Hedin and Megan Porter, who contributed sequencing data.   Thank you to Saemundur Sveinsson for his patient help as I started navigating bioinformatics and sequencing data for the first time. Thanks to Allyson MisCampbell and Carol Ritland from the Forestry Genomic Data Centre at UBC, and Jeff Richards and his students, for their help and the resources they shared in the lab. I am also grateful to Anastasia Kuzmin for being so helpful and collaborative throughout the sequencing process. Thanks to Westgrid (CanadaComputes) for access to their computer clusters.  Thank you to my committee Sally Otto, Loren Rieseberg, and Darren Irwin, and to Maddison lab members past and present, Edy Piascik, Junxia Zhang, Gwylim Blackburn, and Sam Evans, for their helpful feedback and ideas.  Thank you to Sean Naman for his daily positive presence, kindness, and inspiration and to the friends who have made these years so enjoyable. I’m especially thankful to my parents, Josée Leduc and André Robert, for their encouragement and support, and for instilling in me a love of learning and of the natural world.  This work was funded by an NSERC grant to W.P. Maddison,   ! %!1. Introduction A major challenge in phylogenetics is accurately inferring species relationships in recently and rapidly diverged clades (Degnan and Rosenberg 2009). Because of recent divergence, these clades are prone to incomplete lineage sorting (ILS) and are likely to have weak reproductive barriers that are susceptible to hybridization (Donnelly et al. 2004; Grant et al. 2005), two processes that contribute to signals discordant to the species tree.  A clade’s true evolutionary history is therefore reflected both in the species tree and in discordant signals (Maddison 1997; Baum 2007). Determining the relative contributions of introgression and ILS to genetic discordance can help us understand the relationship between hybridization and divergence during diversification, a key field of research in evolution (Mallet 2005; Abbott et al. 2013; Seehausen 2004; Baack and Rieseberg 2007). Hybridization can have several different effects: it can accelerate speciation through reinforcement (Turelli et al. 2001), reduce genetic divergence between hybridizing species (Anderson 1949), or it can introduce new genetic variation at a faster rate than would be possible by mutation alone, potentially sharing adaptive loci across lineages and facilitating diversification (Slatkin 1985; Twyford and Ennos 2012; Rieseberg et al. 2003). While adaptive introgression and hybrid speciation are documented as important and common events in plants, it is only recently that examples of adaptive introgression have been described in animals (Seehausen 2004; Mallet 2005). However, it is still unclear if introgression plays a major or minor role in the diversification of animals. A better characterization of nuclear introgression in different clades is needed to determine the relative contribution of introgression to adaptive genetic variation. Because discordant gene trees produced by introgression or ILS are difficult to distinguish (Joly et al. 2009), a large sample of nuclear genes is required to identify nuclear ! &!introgression. Until recently, phylogenetic datasets have included only a small sample of genes; thus, nuclear introgression is poorly described in most organisms relative to mitochondrial introgression, which may leave a different historical signature that does not correspond to nuclear introgression (Toews and Brelsford 2012). This issue can be addressed using phylogenomic datasets, which can successfully resolve species trees with high levels of discordance (Dunn et al. 2008; Lemmon and Lemmon 2012), and which can also identify patterns of historical introgression (Yu et al. 2011; Kronforst et al. 2006) because characteristics unique to introgression emerge in genome-wide data. For example, two gene trees discordant to the species tree (in a rooted four-taxon tree) are predicted to occur at equal frequencies when ILS is the only cause of discordance (Durand et al. 2011). Because ILS is a random process, it should not favour one particular discordant topology or another. This is not the case if introgression has occurred, because the discordant gene trees which link together species that introgressed as sister taxa should be more frequent due to a higher frequency of shared (via introgression) loci between those two species. Therefore, asymmetric frequencies of discordant signals favouring close relationships between potentially hybridizing species can be used as a diagnosis for introgression with a large enough set of gene trees (Baum 2007; Larget et al. 2010) or SNPs (Durand et al. 2011; Martin et al. 2015; Pease and Hahn 2015). Coalescence times can also differentiate introgression from ILS (Joly et al. 2009). Gene trees produced by ILS should coalesce prior to the time of speciation while introgression should result in coalescence following speciation (Joly et al. 2009). Habronattus is a young jumping spider (Salticidae) genus comprised of about 100 described species that diversified primarily in North America, with some species present as far south as Costa Rica. They possess an impressive diversity of colourful male ornaments and ! '!courtship behaviours, amongst the most complex found in arthropods (Peckham and Peckham 1889; Peckham and Peckham 1890; Griswold 1987; Maddison and McMahon 2000; Maddison and Hedin 2003; Elias et al. 2012). Multimodal courtship displays can involve up to 23 male ornaments, motion displays, and vibratory signals, arranged temporally into motifs  (Elias et al. 2012). As a result of this complex behaviour, Habronattus is emerging as a model to study the role of sexual selection in divergence (Hebets and Maddison 2005; Scheidemantel 1997; Elias 2006; Blackburn and Maddison 2014; Maddison and McMahon 2000; Masta and Maddison 2002), the evolution of sex chromosomes (Maddison and Leduc-Robert 2013; Maddison 1982), and arachnid visual systems (Zurek et al. 2015).  Several clades within Habronattus are morphologically distinctive and have been partially phylogenetically resolved using Ef1-a (nuclear gene) and 16SND1 (mitochondrial ribosomal subunit and adjacent ND1 gene) (Maddison and Hedin 2003) and morphological traits (Griswold 1987). The genus as a whole is defined by an elbowed tegular apophysis (part of the male palp), which is secondarily lost in the coecatus group (Maddison and Hedin 2003; Griswold 1987), and shorter first legs in most species compared to sister genus Pellenes. The americanus, agilis, amicus, tranquillus (together referred to as the AAT clade), and dorotheae groups were first inferred based on morphology (Griswold 1987) and later confirmed with molecular data (Maddison and Hedin 2003). The decorus, coecatus, and viridipes groups began to take form in the morphological phylogeny (Griswold 1987) but only fully emerged with molecular data (Maddison and Hedin 2003). In particular, the viridipes group was further divided into the clypeatus group (primarily southwestern species) and the viridipes group (primarily northern species) based on molecular data (Maddison and Hedin 2003). Despite this success, other aspects of Habronattus phylogeny have been difficult to resolve because of insufficient data and ! (!widespread discordance (Maddison & Hedin 2003; Griswold 1987). The internal relationships of most species groups and the deeper phylogenetic structure of the genus remain poorly resolved. As a result, the analysis of character traits, such as male ornaments and courtship behaviour, within a phylogenetic framework has been limited. There are many indications that introgression may be partially responsible for poor phylogenetic resolution due to gene tree discordance. There have been several molecular hints of mitochondrial introgression, both between distant species (Maddison and Hedin 2003), and between sympatric and closely related species that form mitochondrial clades based on geographic proximity rather than by morphospecies (Hedin and Lowder 2009; Maddison and Hedin, unpubl.). Introgression could also explain why some morphologically divergent species group together with strong support in the nuclear phylogeny (Maddison and Hedin 2003). The possibility of introgression is further supported by behavioural tests, which found that females from different populations of H. pugillis can have preferences for foreign males with divergent courtship displays, a possible result of antagonistic coevolution between the sexes (Hebets and Maddison 2005). If widespread hybridization is possible, in particular between more distant species that are expected to be reproductively isolated, traits novel to the recipient lineages could potentially introgress. For example, sexually selected ornaments in males from divergent and allopatric H. pugillis populations exhibit convergence that could be explained by introgression (Maddison and McMahon 2000). To determine if historical hybridization in Habronattus has been significant and if its evolutionary effects have been creative, neutral, or destructive, the extent of hybridization across the group and the amount of nuclear introgression resulting from this hybridization first needs to be quantified. ! )!We set out to further resolve the Habronattus phylogeny and to determine the extent of introgression in the clade. We collected transcriptome data for 34 Habronattus species and two outgroups, the first genomic dataset assembled for Salticidae spiders. With a well-resolved phylogenetic tree, we were then able to use a phylogenetic approach to determine the extent of nuclear and mitochondrial introgression in the group. We focused on the americanus group and the VCC group, two of the most diverse species groups within Habronattus. To investigate nuclear introgression, we first conducted a Bayesian Concordance Analysis (BCA) to investigate discordance (caused by either ILS or introgression) in gene trees, and we then applied Patterson’s D statistic and DFOIL tests to explicitly test allele patterns for introgression. We resolved most nodes of the phylogeny with high support and identified several instances of hybridization in the group.             ! *!2. Methods 2.1. Taxon sampling We sampled a total of 36 species, including representatives from all major clades within Habronattus, and 2 outgroups (Appendix 1). We sampled more deeply in species-rich groups (i.e., viridipes/clypeatus/coecatus group, and the americanus group), and prioritized the coecatus group because of possible introgression (Maddison and Hedin 2003). The outgroup Pellenes cf. levii is from the genus sister to Habronattus (subtribe Harmochirinae), and Evarcha proszjinski is more distantly related (subtribe Plexippinae). Both the genera Evarcha and Pellenes have been previously used in phylogenies to support Habronattus as monophyletic (Maddison and Hedin 2003).  The undescribed species H. sp. (ROBRT) and H. sp. (CHMLA) retain their 5 letter coded names from Maddison and Hedin (2003) and H. cf. dossenus (Silvercity) of Maddison and Hedin (2003) is renamed here as undescribed species H. sp. (SLCTY). This is the first phylogenetic analysis to include the undescribed species H. sp. (ESTU), H. sp. (BLNDI), and H. sp. (SUNGL). Because these species have yet to be fully described, they are given new 4-5 coded letter names until a complete species description is made in the future. H. sp. (ESTU) was found near water at low tide along the estuaries of Puerto Peñasco, Mexico and has an americanus group style courtship behaviour. H. sp. (BLNDI), which was caught near H. sp. (ESTU) but in a drier habitat, resembles a blond H. pyrrithrix (a species also found sympatrically) with a similar distinctive bright red face. H. sp. (SUNGL) was found in high desert grass clumps near Mount Hopkins, Arizona. It resembles a paler and heavier bodied H.  pugillis, and has a bright orange and blue face.  Specimens were collected from 2012 to 2014 in various locations in Canada (British Columbia, Ontario, and Alberta), the USA (Arizona, California, and New Mexico), Mexico ! +!(Sonora and Jalisco), and Panama (localities of all specimens are listed in Appendix 1). Habronattus are primarily ground-dwellers found on rocks, leaf litter or grasses; therefore, we collected specimens by catching them off the ground and we used a beat sheet method for the few species that were arboreal. Each specimen was photographed alive in the field. Adult male specimens were preferred because they are easier to distinguish morphologically between species. We resorted to adult females for H. sp. (ESTU), H. sp. (SUNGL), H. captiosus, and H. sp. (SLCTY) because males were not available. Both a male and a female specimen were included for H. ophrys and H. festus in an effort to assemble a more complete reference transcriptome. H. paratus DNA was preserved in 95% EtOH. All other specimens were killed by submersion in RNAlater for RNA preservation. To maximize tissue exposure, the cephalothorax and abdomen were opened immediately upon submersion. All specimens were stored at -20°C. Legs and the male palps or the female epigynum were preserved separately as vouchers (stored at the Beaty Biodiversity Museum at the University of British Columbia).   2.2. Molecular extractions and sequencing We chose RNA-seq because the assembled transcripts are long enough to permit a gene tree approach that is difficult to implement with other common phylogenomic methods (e.g., RADseq, targeted sequence capture). Transcriptomes also avoid uninformative sequences that may form a major part of the large Habronattus genome (haploid genome size = 5.73 pg, Gregory and Shorthouse 2003) and that would be assembled using whole-genome methods. Total RNA was extracted from whole specimens using a combination of TRIzol extraction (Life Technologies) and RNeasy Mini Kit (Qiagen) for RNA purification and DNAse digestion. DNA ! ,!was extracted from the legs and abdomen of H. paratus using a QIAamp DNA Investigator Kit (Qiagen). Samples were selected for sequencing based on three criteria: (1) quantity (minimum 2 ug) as measured by a spectrophotometer, (2) purity, (minimum 260/280 = 1.8 for RNA and 260/280 = 2 for DNA, minimum 230/260 = 1.5) and (3) RNA integrity, assessed on a 2100 Bioanalyzer. RNA samples were kept at -80°C and DNA samples were kept at -20°C until the cDNA library preparation step.  Libraries were constructed with BIO-O NEXTflex Library Prep Kits (Bioo Scientific, Inc.) with insert sizes averaging 220bp for RNA and 300 bp for DNA, and sequenced as 100bp paired-end reads on an Illumina HiSeq 2000 (Illumina, Inc.) at the Biodiversity Research Centre Next Generation Sequencing Facilities (University of British Columbia). We aimed for at least 20 million paired reads per species, with the exception of potential reference transcriptomes H. ophrys and H. festus, which were more deeply sequenced (approximately half a sequencing lane each, see Appendix 2 for sequencing summary). H. signatus and H. ustulatus were prepared at UC San Diego (M. Hedin, unpubl.) and sequenced as 50 bp paired end sequences, and H. sp. (SLCTY) was prepared at the University of South Dakota (M. Porter, unpubl.) and sequenced as 150 bp paired end sequences.   2.3 Sequence read filtering and trimming  Any sequence read with an average Phred score under Q=30 was discarded. All remaining reads were quality checked with FASTQC V0.10.1 (Andrews 2010).  Terminal nucleotides were trimmed using fastq-mcf from ea-utils (Aronesty 2011) if they had a score below Q=30 or if they were sequencing adaptors. Reads that were 95% or more homopolymer were discarded and any suspected contaminants, detected from the GC content ! -!curve of FASTQC, were trimmed using PRINSEQ-lite (Schmieder and Edwards 2011). Any read shorter than 33 bp after trimming was discarded. Contamination was ruled out as the source of overrepresented sequences when the top blastn hits returned for these sequences were Habronattus mitochondrial and ribosomal RNA sequences.  2.4 Reference transcriptome   Reference transcriptomes were assembled for H. festus and H. ophrys in Trinity RNA-seq_v20140717 (Haas et al. 2013) with the command: --seqType fq --left leftreads.fq –right rightreads.fq --JM 110G --CPU 12 --inchworm_cpu 12 --bflyCPU 12 --min_contig_length 200 –-kmer_cov 2  Both assemblies were similar in size and quality, so we selected H. ophrys as the reference transcriptome because of predicted ease of obtaining this species for future studies. We filtered transcripts from the H. ophrys prior to using the transcriptome as a reference for subsequent assemblies according to the steps outlined in Figure 1. We determined transcript abundance with RSEM v1.2.19 (Li and Dewey 2011) and kept only the most abundant transcript variant per gene. Any remaining redundant transcripts identified with CD-HIT-EST (Fu et al. 2012) with a similarity threshold of 95% were removed. To decrease the likelihood of paralogous genes rather than true orthologs assembling on a reference transcript during reference-based assemblies, we conducted an all-versus-all BLAST with all remaining H. ophrys contigs and removed any contig with a contig other than itself as a significant hit (blastx, evalue = 10-3). To set codon positions, the reference transcriptome was scanned for open reading frames (ORFs) using TransDecoder_r20131110 (Haas et al. 2013) and the longest ORF of a transcript ! %.!was chosen as its protein coding region. If multiple non-overlapping coding regions were found on a transcript, those transcripts were split between coding regions. We conducted a BLAST search of the entire H. ophrys transcriptome (evalue= 10-3, min. HSP length = 33, max_target_seqs=20) to proteins from the SWISS-PROT database (Boeckmann et al. 2003), the African social velvet spider (Stegodyphus mimosarum) (Sanggaard et al. 2014) and the Brazilian white-knee tarantula (Acanthoscurria geniculata) (Sanggaard et al. 2014) to assess the quality and completeness of the transcriptome. The transcriptome was also BLASTed to the H. oregonensis mitochondrial genome (Masta and Boore 2004) and any significant hits were removed from the final reference nuclear transcriptome. The annotated H. oregonensis mitochondrial genome (Masta and Boore 2004), was used instead as the reference for all mitochondrial assemblies.  2.5 Reference-based assembly of transcriptomes For each of the species other than H. ophrys, sequencing reads were first mapped to the H. oregonensis mitochondrial genome and the remaining unmapped reads were mapped to the H. ophrys reference transcriptome using CLC Genomics Workbench (CLC Bio) with assembly parameters: mismatch cost=2, insertion cost=3, deletion cost= 3, length fraction=0.5, similarity fraction=0.8. Polymorphisms were retained in the consensus sequence as ambiguous sites if the variant represented 30% or more of mapped alleles. Nuclear sequences with average coverage less than 5x were discarded, and contigs were split into fragments at any region where coverage was less than 5x. Only contigs longer than 200 bp following these steps were retained.   ! %%!2.6 Sequence alignments We converted species-based FASTA files into locus-based FASTA files and trimmed sequences of any remaining poly-A or poly-T tails using custom python scripts. If a loci was fragmented into multiple sequences for a species, and the lengths of those fragments when added together met our length cut-off, then those sequence fragments were scaffolded with “N”s to represent the missing data. Orthologous loci with a minimum average length of 200 bp  (excluding “N”s) were aligned using MAFFT v7.058b (Katoh and Standley 2013) using L-INS-I and parameters --localpair --maxiterate 100. Nucleotides trailing from either end of an alignment (present in 30% or less of species) were trimmed and a sequence was discarded if it was 30% or less than the average sequence length for that alignment. Matrices were partitioned as codon position 1,2,3, or non-coding based on Transdecoder results. To better understand the cause of gaps and ambiguous nucleotides in these alignments, we aligned a subset of transcripts with annotated sequences with known protein-coding regions matched using a BLAST search to the SWISS-PROT database, to Ef1-a (Maddison and Hedin 2003), and to the annotated mitochondrial genome (Masta and Boore 2004). All sites with gaps were caused by an insertion of a nucleotide in one or a few sequences. These insertions introduced gaps in the sequences that lacked the inserted nucleotide, causing highly unlikely frame-shifts in conserved genes. These insertions only occurred in the reference-based assemblies and never occurred in the reference sequence. Therefore, we inferred that these insertions of gaps and nucleotides were assembly errors rather than true insertions. As a result, columns with gaps were excluded if they were not present in the reference sequence whenever they were in a coding region (to avoid any frame-shift mutations) or if they were present in at least 50% of species in a non-coding region. If ambiguous sites constituted more than 3% of an ! %&!alignment, those alignments were excluded from phylogenetic analyses. High levels of ambiguous sites might not represent true heterozygosity, but rather multiple transcript variants or paralogs assembled on the same reference transcript. A 3% ambiguity threshold is very conservative and based on ambiguity levels in the sequences of three Ef1-a exons (exons described in Hedin and Maddison 2000). While exon 2 (ambiguity levels = 13%) and exon 1 (ambiguity levels =1.5%) have a top BLAST nr database hit to Habronattus ef1-a, ef1-a exon 3 (ambiguity levels = 21%) has more distant (not Habronattus) jumping spiders as its best BLAST nr database match. We therefore interpret high levels of ambiguity in a consensus sequence as a possible indication of multiple variants or paralogs contributing to the consensus sequence. All alignment trimming and partitioning steps were completed with the package Gataga (Maddison and Maddison, unpubl.) in Mesquite 3.02 (Maddison and Maddison 2015). Because reference-based assemblies generate sequences aligned to the same reference transcripts, sequences assembled on the same reference are assumed to be orthologous. Filtering for paralogs in the H. ophrys transcriptome, reducing reference transcriptome redundancy, and removing sequences with high levels of ambiguous sites are expected to have sufficiently reduced the possibility of paralogs assembled on a reference sequence. To verify that paralogs were not an issue following these steps, gene trees were constructed for genes of lengths over 1 kb with a single search replicate in RAxML 7.7.9 (Stamatakis 2006). None of the trees produced had unusually long branches or phylogenetic structures obviously indicative of paralogy, so we considered these loci to be reliable orthologs.    ! %'!2.7 Phylogenetic analyses To determine optimal models of substitution, we analyzed in PartitionFinder v.1.1.1 (Lanfear et al. 2012) partitions for gene, codon positions (pos1, pos2, pos3) , and non-coding sites (N) for a sample of 20 nuclear genes and all mitochondrial genes using a greedy algorithm search and AIC model selection. All maximum likelihood phylogenetic searches were run with substitution model GTR+G+I, 20 search replicates, and 1000 bootstraps in RAxML 7.7.9 (Stamatakis 2006).  2.7.1. Nuclear and mitochondrial matrices  We conducted several phylogenetic analyses using different concatenated matrices, each composed of different sets of loci (Figure 2). We had multiple goals when separating loci into different matrices for phylogenetic analysis. First, we wanted to concatenate loci deemed to be of highest quality for our primary nuclear phylogenetic analysis. The 1,884 nuclear loci meeting the benchmarks for high quality (i.e., present in 25 or more species, with coding region) were concatenated together using Mesquite 3.02 as the concatenated nuclear matrix (total length = 2.41 Mb). Because concatenated datasets can overinflate support values (Kubatko and Degnan 2007), we also separated our concatenated nuclear matrix into 8 even subsets (nuclear subsets 1-8; 302,200 bp each) to better assess the strength of the phylogenetic signal. Contig order was randomized prior to concatenation and subset division to ensure that each subset represents a random sample of loci. Coding sites and non-coding sites of the concatenated nuclear matrix were also analyzed as two separate matrices (coding matrix, 1.93 Mb, and non-coding matrix, 565.8 kb).  ! %(!We analyzed the remaining nuclear loci (those that do not meet all our “quality” requirements: present in less than 25 species, without coding regions), to see if these requirements truly mattered as indicators of quality and affect phylogenetic outcomes. There were 1,020 loci present in 15 to 24 species, and these were concatenated together as the missing species matrix (concatenated matrix length = 1.1 Mb). We did not assess loci present in less than 15 species because these were highly skewed towards species with much more sequencing data (H. ophrys, H. festus, H. signatus, H. ustulatus, H. sp. (SLCTY)). The 237 transcripts without coding regions present in 25 or more species were concatenated together as the unidentified transcripts matrix (118.1 kb).  The mitochondrial phylogeny includes an additional species, H. paratus, for which a DNA sample was sequenced rather than RNA. We only include the mitochondrial H. paratus sequences for phylogenetic analysis because these sequences are complete and have adequate coverage, while most nuclear sequences for H. paratus are not. Mitochondrial sequences for 16S RNA (1022 bp), 12S RNA (691 bp), ND1 (921 bp), ND2 (959 bp), ND3 (342 bp), ND4 (1289 bp), ND4L (268 bp), ND5 (1638 bp), ND6 (429 bp), ATP8 (158 bp), Cytochrome B (1111 bp), COX1 (1542 bp), COX2 (666 bp), COX3 (786 bp) were aligned, concatenated (concatenated mitochondrial matrix, 12.33 kb), and assigned codon positions based on annotations from the H. oregonensis mitochondrial genome (Masta and Boore 2004). To assess the strength of the mitochondrial phylogenetic signal with less data, we divided mitochondrial rRNA (1.72 kb) from protein-coding sites and then separated protein-coding alignment into 4 even subsets (mtDNA subsets 1-4; 2.53 kb each).   ! %)!2.7.2. 16SND1 and Ef1-a phylogenies with broader species sample  To place our dataset in the context of the broader Habronattus phylogeny and previously sequenced data, sequence alignments from genes Ef1-a (819 bp) and 16SND1 (1047 bp) were integrated with pre-existing Sanger Sequencing data from Maddison and Hedin (2003). These datasets included a much broader sample of Habronattus species and populations. These combined Sanger and transcriptome alignments were reanalyzed using RAxML with 20 search replicates and a GTR+G+I substitution model. There were 160 Sanger sequences available for 16SND1 and 105 Sanger sequenced specimens for Ef1-a. The two Ef1-a introns (Intron 1 = 148 bp, Intron 2 = 91 bp) were coded as missing in transcriptome sequences. To see how additional species fell into the concatenated nuclear phylogeny while avoiding poor resolution of known clades due to low phylogenetic informativeness of the genes, any node recovered with minimum 90% bootstrap support in either the concatenated mitochondrial phylogeny (for the 16SND1 tree) or the concatenated nuclear phylogeny (for the Ef1-a analysis) was enforced as a phylogenetic constraint.  2.8 Introgression 2.8.1 Bayesian concordance analysis  Bayesian concordance analysis (BCA) uses a set of Bayesian gene trees to detect both the dominant phylogenetic signal arising from the genes, represented as a primary concordance tree (the estimated species tree), and secondary signals that are also substantially supported by some genes but discordant with the dominant species tree (Ané et al. 2007; Baum 2007). These discordant signals are considered to be part of the true evolutionary history of the group, and may be the result of introgression, incomplete lineage sorting (Larget et al. 2010), or noise due to ! %*!low phylogenetic informativeness of single genes (Smith et al. 2015). Genomic support for a particular clade is represented as a Concordance Factor (CF) and a 95% CF credibility interval. The CF represents an estimation of the proportion of sampled gene trees for which a clade is true (Baum 2007). The width of the credibility interval is indicative of the confidence of the CF. A wide credibility interval, and in particular an interval overlapping 0, indicates greater CF uncertainty due to lower posterior probabilities for that clade from Bayesian gene tree searches (Ané 2010) We conducted all Bayesian gene tree searches with MrBayes 3.2.2 (Ronquist and Huelsenbeck 2003) with 4 chains (3 cold, 1 hot), 2 runs, and 25% burn-in. To determine the number of MCMC generations required for convergence, 20 genes were run until convergence as a test. Convergence was measured as a standard deviation (SD) of split frequencies <0.01 (Ronquist et al. 2005). The convergence time ranged from 570,000 to 13,087,000 and averaged 4,025,500 generations, so we set the number MCMC generations conservatively at 20,000,000 generations per gene. Codon positions were used as partitions. BCA analyses were conducted using BUCKy 1.4.3 (Larget et al. 2010) with 2 runs and 4 chains per analysis. BUCKy does not assume that any specific biological process is causing gene tree incongruence (Larget et al. 2010) and so can be used to identify potential sources of introgression or incomplete lineage sorting. Due to the computational limitations of BUCKy, we only analyzed the americanus group and the VCC group, and only included genes longer than 1 kb that were present in all species being analyzed. There were 679 genes included for the americanus group analysis (with H. signatus outgroup) and 517 genes included for the VCC group analysis (with H. ophrys outgroup).  ! %+!The single adjustable parameter in BUCKy is !, which represents the expected level of gene tree discordance; a larger value of ! corresponds to greater gene tree discordance. To test !, we tried values of ! = 0.1, 1, 2, 5 and 10. We found no substantial difference in results using different ! for the americanus group, so kept ! set to 1. The VCC clade had difficulty converging at higher !, so ! was set to 0.1 for that analysis, although there was very little difference between CFs depending on ! for VCC. The americanus group analysis ran for 10,000,000 generations and the VCC clade analysis ran for 30,000,000 generations due to the longer time required for convergence.   2.8.2 Patterson’s D statistic and DFOIL  To distinguish between incomplete lineage sorting (ILS) and introgression patterns in SNP data, we conducted Patterson’s D statistic tests (Durand et al. 2011) and the related test DFOIL (Pease and Hahn 2015). These tests compare patterns of shared SNPs across sets of 4 and 5 taxa, respectively. Under ILS in a 4-taxa binary tree, it is expected that the number of shared alleles between non-sister taxa would be equal for each possible non-sister pairing (i.e., patterns ABBA and BABA) (Durand et al. 2011). Introgression between two species is detected when they have significantly more shared alleles than expected (i.e., more ABBA than BABA or vice versa). The same principle can be applied to a 5-taxa tree, with some added complexity (Pease and Hahn 2015). Because there are more possible ways to share alleles and have introgression between 5 taxa, DFOIL uses 4 different D statistics (one per species tested) to detect a broader range of possible events. Whether Patterson’s D statistic or DFOIL was used depended on the structure of the species tree of the taxa being tested. If the species tree did not agree with the ! %,!phylogenetic structure (((species 1, species 2),(species 3, species 4)), outgroup) required by DFOIL, Patterson’s D statistic was used instead.  Specimens tested (listed in Table 1) have some variability in sex chromosomes. We included female specimens (i.e., H. sp. (SLCTY), H. sp. (ESTU)), males with Y chromosomes (i.e., H. borealis, H.mexicanus, H. zapotecanus, and H. decorus), and specimens with have unknown karyotypes (i.e., H. sp. (BLNDI), H. sp. (ESTU), H. clypeatus, H. sp. (SLCTY)), because excluding all of these species from tests would impose too many limitations on testable hypotheses. An excess of shared alleles between species possessing Y chromosomes (or between species lacking a Y) resulting in a false positive for introgression remains a possibility, although the effect should not be substantial. Habronattus Y chromosomes are very young neo-Ys resulting from an X-autosome fusion that most likely retain some level of recombination with the neo-X (Maddison and Leduc-Robert 2013; Maddison 1982). As a result, they are assumed to be mostly, if not all, orthologous to autosomes of species lacking Ys. Any Y-specific gene should be excluded from the analysis, because only sites present in all 4 or 5 species were considered when counting SNPs.  We tested sets of species based on indications of introgression in previous studies (Maddison and Hedin 2003), in the mitochondrial phylogeny, or in the Bayesian concordance analysis. There were 4 principal hypotheses tested: (1) introgression among species of the americanus group, (2) introgression between the clypeatus group and coecatus group, (3) introgression between H. sp. (ROBRT) with the VCC group and (4) introgression between the decorus/altanus/banksi group and the americanus group.   For hypothesis 1, we tested all 5 americanus group species. For hypotheses 2, 3, and 4, we did not test all possible species because multiple large clades were involved. Instead, we selected ! %-!a few species as representatives of each clade. With a careful choice of species as a clade representative, we can still draw general conclusions about introgression across Habronattus clades. For hypothesis 2, we included 2 coecatus and 2 clypeatus species per DFOIL test. H. pyrrithrix (coecatus group) and H. clypeatus (clypeatus group) are our clade representatives that are always included. H. pyrrithrix was chosen because its phylogenetic position is consistent in both the concatenated nuclear and mitochondrial phylogenies, and H. clypeatus was chosen because of mitochondrial introgression detected in the mitochondrial phylogeny and in other members of its species group by Maddison and Hedin (2003). By using more or less divergent sister taxa in different tests for comparison, we can better approximate where in the coecatus and clypeatus group introgression has occurred. For example, H. pyrrithrix/H. mexicanus tests for introgression across the coecatus group because H. mexicanus is a distant coecatus group species, H. pyrrithrix/H. borealis tests for introgression at a shallower scale, and H. pyrrithrix/H. sp. (BLNDI) tests for very recent introgression with either sister species. Likewise, H. clypeatus/H. sp. (SLCTY) tests for recent introgression with H. clypeatus or H. sp. (SLCTY) while H. clypeatus/H. aztecanus tests for introgression across the clypeatus group. For hypothesis 3, we tested only the species with the most sequencing data per clade for the four clades involved (H. sp. (ROBRT), and viridipes, clypeatus, coecatus groups). For hypothesis 4, we used the species with the most sequencing data for each clade involved, and to identify the depth of introgression, we used species at incremental phylogenetic distances from H. decorus. D-statistics and DFOIL tests are sensitive to missing data. If the phylogeny is sparsely sampled, these tests may detect introgression from a lineage that is not sampled (i.e., a ghost lineage) and incorrectly attribute the introgression signal to the lineage most closely related to the ghost taxon (Eaton et al. 2015). We only sampled 2 of 10 known viridipes species, 3 of 11 clypeatus group ! &.!species, 7 of 17 coecatus group species, and 5 of 11 americanus group species. As a result, even if we tested all possible combinations of sampled species, it is impossible to attribute an introgression signal to a specific species because there is a high probability that ghost taxa have had an effect (Eaton et al. 2015).  DFOIL was run in mode dfoilalt to reduce noise from synapomorphic sites. We used a custom R script to count allele patterns. All sites that included ambiguous nucleotides, gaps, or missing data were excluded from the analysis. DFOIL estimated divergence T-values were verified against the assumption that T12<T34<T1234. We adjusted significance for 68 comparisons (Bonferroni correction) to a p-value lower than 0.00074 to indicate 95% significance.             ! &%!3. Results  3.1 Transcriptome assemblies Total raw sequencing reads, trimmed sequencing reads, and the number of reads assembled, and total transcripts, and coverage are summarized for every species in Appendix 2.  3.1.1 Reference transcriptome  The unfiltered H. ophrys reference transcriptome includes 117,859 transcripts (total 53,927,457 bases assembled), with an N50 (analogous to median contig length, Miller et al. (2010)) of 516, and an average coverage of 103x. Following filtering for redundancy, selection of a single variant per gene, removal of possible paralogs, and the separation of connected transcripts, there were 92,343 transcripts left. After reads were remapped, 51,143 transcripts had sufficient (5x) coverage (all filtering steps are summarized in Figure 1). There are 13,901 H. ophrys loci with ORFs longer than 200 bp and coverage greater than 5x. Coding regions are most frequently identified in loci longer than 1000 bp (95%) and this number decreases the shorter the loci (500 to 1000bp: 77%, 200-500 bp: 15%). Almost all transcripts with coding regions (93%) have significant SWISS-PROT hits, compared to only 66% of the 78,442 contigs that do not have any detected ORFs. There are 18,053 H. ophrys loci with significant hits to the African social velvet spider protein set (26,889 proteins, 67.1% overlap) and 16,796 H. ophrys loci have significant hits with the Tarantula protein set (76,238 proteins, 22% overlap). These reference protein sets serve as best approximations of a complete spider transcriptome.  ! &&!3.1.2. Reference-based transcriptomes assemblies Reference-based transcriptome assemblies mapped on average 77% of trimmed reads to either the nuclear reference transcriptome (average nuclear coverage = 67x) or mitochondrial reference genome (average mitochondrial coverage = 13,640x). There were an average of 10,164 transcripts assembled per species, although numbers ranged widely (depending on the number of reads) from 3,746 for H. sp. (ROBRT) to 28,846 for H. festus. Contigs without identified coding regions were much more likely to be discarded based on high levels of ambiguous sites (59% in the transcripts without coding regions matrix compared to 11% in concatenated nuclear matrix and 22% in missing species matrix), and sequences discarded based on ambiguity levels were overwhelmingly short transcripts under 500 bp.  3.2 Substitution model selection  GTR+G+I or GTR+G was chosen as the optimal substitution model using AIC for all mitochondrial partitions. For the 20 nuclear genes tested, AIC selected the General Time Reversible (GTR) substitution model for 33 partitions; transversion model (TVM) for 22 partitions; unequal-frequency Kimura 3-parameter (K81uf) for 10 partitions; transition model (TIM) for 5 partitions; Hasegawa-Kishino-Yano (HKY) for 4 partitions; and Tamura-Nei (TrN) for 3 partitions. These include additional variations of these models, like gamma-distribution of rates (+G) or proportion of invariable sites (+I). We were unable to set a different model for each partition in concatenated matrices due to computational limitations. Instead, we set GTR+G+I as the substitution model in all phylogenetic Maximum Likelihood (ML) and Bayesian analyses (nst = 6 rates = invgamma) because it was the most commonly chosen and most widely applicable model. ! &'!3.3 Nuclear phylogenetic analyses   We resolved most phylogenetic relationships with high bootstrap support (95-100%) in the concatenated nuclear phylogeny (Figure 3). This phylogeny replicates clades previously resolved in the molecular phylogeny of Maddison and Hedin (2003) and the morphological phylogeny of Griswold (1987) with high bootstrap support. The amicus, agilis, and tranquillus groups are well supported and form the monophyletic and most basal AAT group, a pattern consistent with previous molecular and morphological phylogenetic studies (Maddison and Hedin 2003; Griswold 1987). All Habronattus other than the AAT group (i.e., those at the bottom, or the right side of the phylogeny in Fig.3) also form a clade. Within this clade forming the right of the phylogeny, H. geronimoi is sister to the rest of the clade. The VCC group, and the americanus group are also resolved with high support. The concatenated nuclear phylogeny also provides clarifications for several previously ambiguous phylogenetic relationships. Two well-supported clades are nested in the coecatus group: H. festus, H. captiosus and H. borealis together will be referred to as the Northern clade (all specimens in this clade were collected in Canada), while H. sp. (BLNDI) and H. pyrrithrix are together and form the Southern clade (since they are both found in the Southern USA neighboring Mexico). H. virgulatus is also sister to the Northern coecatus clade with high support (nuclear bootstrap = 97%). H. altanus with H. decorus are sisters, and the banksi group (H. sp. (CHMLA) and H. zapotecanus) is their sister. The previously intractable H. hallani is strongly supported as sister to the pugillis group (H. sp. (SUNGL) and H. pugillis). The relationships between the subgroups of the VCC clade are resolved with high bootstrap support; H. sp. (ROBRT) is sister to the viridipes, clypeatus, and coecatus group, the viridipes group (nuclear bootstrap = 83%) is sister to the clypeatus and the coecatus group, and the clypeatus and ! &(!coecatus group are sisters. H. jucundus groups with H. calcaratus rather than with the oregonensis group (a poorly supported relationship found in Maddison and Hedin (2003)), confirming that the viridipes group is in fact monophyletic. The internal relationships of the americanus group and the VCC group are also well resolved. Most of the relationships found in the concatenated nuclear phylogeny are also replicated in phylogenies from different matrices (i.e., nuclear subsets 1-8, coding sites, non-coding sites, missing species, and transcripts without coding regions matrices). This suggests that high quality orthologs were assembled, even in matrices predicted to be of “lower” quality (i.e., the transcripts without coding regions matrix and missing species matrix), and that the phylogenetic signal for most branches is robust even with less data (i.e., nuclear subsets 1-8), and partitions that fit to different evolutionary models (i.e., coding sites, non-coding sites). The AAT, americanus, altanus/decorus/banksi, pugillis, and VCC groups are always monophyletic across analyses, as are the positions of H. geronimoi and H. hallani, and the internal relationships within the americanus group and the VCC clade are, for the most part (see below), consistent across analyses.  The most poorly resolved nodes of the concatenated nuclear phylogeny are also the most variable across the phylogenies from nuclear subsets 1-8, coding sites, non-coding sites, missing species, and transcript without coding regions (variable nodes are summarized in Figure 4). Pellenes cf. levii falls outside Habronattus (as expected) in the Maximum likelihood concatenated nuclear phylogeny, but falls within the Habronattus in most other nuclear phylogenies, including the bootstrapped consensus concatenated nuclear phylogeny (bootstrap support = 67%) (Fig. 4A-D). The positions for H. icenoglei, H. oregonensis and H. cambridgei also conflict across nuclear analyses (Fig. 4E-H). H. icenoglei and H. oregonensis  are sisters in ! &)!the Maximum likelihood tree as well as in phylogenies from 3/8 nuclear subsets, coding sites, and non-coding sites (Fig. 4E). However, H. cambridgei and H. oregonensis are sisters in the bootstrap consensus tree (bootstrap support = 66%) and in phylogenies from 2/8 nuclear subsets (Fig. 4F). H. icenoglei is the most volatile species, even grouping at times with the VCC clade in 3/8 nuclear subsets (Fig. 4H).  There are other relationships that, despite relatively high bootstrap support, show some variability across phylogenies from different nuclear matrices (Fig. 5). H. sp. (ROBRT), while strongly supported as sister to the rest of the VCC clade (Fig. 5A), also groups frequently with the clypeatus group (Fig. 5B). H. virgulatus (bootstrap support = 97%) departs from its dominant concatenated nuclear position (Fig. 5D) to group with either the Southern coecatus clade (Figure 5E) or as a basal branch of the coecatus group (Fig. 5F) in some nuclear subsets. H. ophrys is positioned as the sister to H. sansoni and H. americanus in half of all data subsets (Fig. 5H), despite being sister to H. tarsalis (bootstrap support = 100%) in the concatenated nuclear phylogeny (Fig. 5G).  High bootstrap support (95-100%) should be indicative of good resolution for a phylogenetic relationship in the concatenated nuclear phylogeny. Nodes with high bootstrap support are replicated in most or all additional phylogenetic analyses using different matrices. Therefore, based on the criterion of high bootstrap support and consistency across analyses, we consider the concatenated nuclear phylogeny (Fig. 3) to be a reflection of the true species phylogeny, except for some ambiguity that remains regarding the positions of Pellenes cf. levii and H. icenoglei/H. oregonensis/H. cambridgei (Fig. 4). Because of high bootstrap support, we accept the positions of H. sp. (ROBRT), H. virgulatus, and H. ophrys/H. tarsalis (Fig. 5) despite variability found across analyses for their phylogenetic positions. ! &*!3.4 Mitochondrial phylogenetic analyses The mitochondrial phylogenetic signal is strong for most clades and consistent across rRNA and protein-coding mitochondrial subsets (Figure 6). The americanus group and the VCC group remain monophyletic in all subsets. The most variability across subsets is found on the left side of the phylogeny (i.e., the outgroups, H. paratus, H. geronimoi and the AAT clade in Figure 3), although H. oregonensis, H. hallani, and the pugillis group also vary.  There are several notable differences between the mitochondrial and the nuclear phylogenies. Those most convincingly suggestive of introgressive hybridization involve distant, morphologically disparate species grouping together in the mitochondrial phylogeny. H. clypeatus is found with coecatus group species H. sp. (BLNDI), H. borealis, and H. captiosus with strong support (bootstrap support = 100%, replicated in each of rRNA and protein-coding subsets 1-4). This pattern of mitochondrial introgression repeats a pattern of clypeatus/coecatus discordance previously identified in Maddison and Hedin (2003), where H. velivolus (clypeatus group) and H. sp. (CHIH) (clypeatus group) specimens were polymorphic for coecatus and clypeatus group mitochondria. Unlike the nuclear phylogeny, H. altanus, H. decorus, and the banksi group, branch together at the base of the americanus group (bootstrap support = 100%, replicated in rRNA, protein-coding subsets 1-4). There are also mito-nuclear differences in the internal relationships of the americanus and coecatus groups. Because these species are very recently diverged, both ILS and hybridization are likely explanations for mito-nuclear discordance. Ongoing hybridization could explain why H. ophrys is the sister taxon to sympatric species H. americanus (bootstrap support = 100%), rather than to morphologically similar H. sansoni (sister to H. ophrys in the nuclear tree). Another difference between more closely related taxa is the relationship of H. icenoglei as sister to the pugillis group, H. hallani, and the VCC ! &+!group (bootstrap support = 86%) rather than H. oregonensis. There is also some support for the monophyly of H. hallani, pugillis group and VCC group (bootstrap support = 52%); however, the exact position of H. hallani within the clade is unresolved.   3.5 16SND1 and Ef1-a phylogenies (broader species sample) When transcriptome and Sanger sequenced 16SND1 (Figure 7) and Ef1-a (Figure 8) sequences are combined and analyzed using ML methods, gene trees resolve most established species groups. In the 16SND1 tree, the americanus group and the coecatus group are each monophyletic, although they are poorly resolved at the species level. With the exception of H. borealis, H. virgulatus, and H. festus, specimens of the same species do not form individual clades. Instead, they sometimes group together and sometimes group with other species that are closely related. The transcriptome (GLR218) and Sanger (d436) H. sp. (SUNGL) specimens are not sisters but they do form a clade with two H. pugillis Sanger specimens (M-105MX, M-29GL), confirming the place of H. sp. (SUNGL) within the pugillis species complex. Both clypeatus group specimens H. sp. (CHIH)-HA292 and H. velivolus-HA659 are found within the coecatus group, replicating the patterns of mitochondrial introgression from the coecatus to the clypeatus group found in Maddison and Hedin (2003). H. clypeatus-GLR227 (transcriptome) is sister to clypeatus group H. sp. (CHIH)-HA292 (Sanger) within the coecatus group.  The Ef1-a tree resolves only some clades supported by the transcriptome concatenated nuclear phylogeny and is poorly resolved at the species level across the entire phylogeny. Only the outgroups, AAT group, dorotheae group, americanus group, and VCC group (with the exception of a few specimens) are monophyletic. The poor resolution and variable branch lengths in the Ef1-a tree can be explained by two factors. First, a short and highly conserved ! &,!alignment is at least partially responsible for the low resolution of the phylogeny. Second, there are high levels of ambiguous sites in the transcriptome sequences. The prevalence of ambiguous sites for the Ef1-a alignment is over 3%, and so this locus would have been discarded from the concatenated nuclear matrix. Such high levels of ambiguous sites may be the result of multiple paralogs or transcript variants assembled together on the reference and summarized as a single consensus sequence. Ambiguous sites are most concentrated in exon 3, and this exon’s best BLAST hit is a Salticidae that is not Habronattus. The 1st and 2nd exon also have higher levels of ambiguous sites in the transcriptome specimens compared to the Sanger specimens, but their best BLAST hits are Habronattus Ef1-a. Even with the 3rd exon removed (as in the final alignment), transcriptome specimens have a tendency to group together, particularly in the “middle” portion of the phylogeny (i.e. groups pugillis, hallani, altanus, decorus, banksi, fallax). Unlike other transcriptome specimens, H. festus-GLR088, 094 does not have any ambiguous sites, resulting in a long branch for that specimen. The absence of ambiguous sites in only this sequence is further indication that multiple variants are most likely summarized as a single Ef1-a sequence. There are two Ef1-a variants known so far, one with 3 exons (Sanger specimens included here), and another with 2 exons (Hedin and Maddison 2000). The concentration of ambiguous sites in the 3rd exons suggests that other variants may exist.  3.6 Introgression 3.6.1 Bayesian concordance analysis We only report CFs with credibility intervals that do not overlap 0 (and as a result are considered to be significant, Ané 2011). We only included CFs with credibility intervals overlapping 0.1 or higher for inclusions in Figures 9 and 10. CFs with credibility intervals ! &-!overlapping values as low as 0.05 are also listed in the Appendix 3 (americanus group) and 4 (coecatus group). Lower CFs (under 0.05) that did not overlap 0 were also prevalent, although we do not report them all because low CF values are likely to be overestimated (Ané 2011) and could be especially vulnerable to false positives due to uninformative genes causing noisy data.  americanus group – The BUCKy analysis of 679 americanus group loci (105 tree topologies sampled, 25 distinct splits) converged with an average SD of mean sample-wide CF of 3.24 x 10-5. The primary concordance phylogeny (Figure 9) reflects the concatenated nuclear phylogeny and is well supported by most gene trees. There is a number of secondary CFs supported by a large portion of the data. H. ophrys is linked to H. americanus and H. sansoni with strong support (CF = 0.272 - 0.339). This high CF is more than twice the CF shared for the equivalent pairing H. tarsalis and H. sansoni/H. americanus (CF = 0.112 - 0.162), suggesting that ILS and shared ancestry can only account for approximately half of that shared genetic history. H. sp. (ESTU) is substantially linked with H. tarsalis (CF = 0.112 - 0.162). However, it is not significantly linked (CF = 0) with H. ophrys, another indication of possible introgression. In addition to those clades that are listed in Appendix 3, there are 12 clades with CFs under 0.05 that are significant (i.e., they do not overlap 0) but which are not listed. VCC group - The BUCKy analysis for 517 VCC group loci (4,795,750 tree topologies sampled, 8,177 distinct splits) converged with an average SD of mean sample-wide CF of 0.003. The longer time to convergence and lower CFs overall compared to the americanus group could be indicative of a higher level of discordance in the VCC group. Low informativeness of genes could also explain low CFs and difficulties with convergence, as well as the large number (4,795,750) of tree topologies sampled. The primary concordance phylogeny (Figure 10) reflects the concatenated phylogeny except for the positions of H. virgulatus, which is a part of the ! '.!Southern coecatus clade  (i.e., H. pyrrithrix and H. sp. (BLNDI)) rather than the Northern clade in the primary concordance tree. H. sp. (ROBRT) has two strong secondary CFs: it is grouped with the clypeatus/coecatus clade (CF = 0.174 – 0.222) and with the clypeatus group (CF = 0.17 – 0.205). Support linking H. sp. (ROBRT) to the clypeatus group is unanticipated based on their phylogenetic distance, and is unlikely to be due entirely to ILS given a very low CF (0.039 – 0.06) with the equidistant viridipes group. There is some support  (CF=0.087 – 0.122) that ties viridipes and coecatus together as a group. The species of the coecatus group have many secondary CFs amongst them, possibly because of widespread ILS. H. virgulatus has a low primary CF (CF=0.101 – 0.137) and 7 secondary CFs >  0.05 within the coecatus group, the most genetic discordance of any species. H. sp. (BLNDI) and H. pyrrithrix are the other coecatus species most often implicated in secondary signals. In addition to those clades that are listed in Appendix 4, there are 181 clades with CFs under 0.05 that are significant (i.e., they do not overlap 0). Of these 181 clades, 35 clades with CFs below 0.04 (average near 0.01) group together species from the clypeatus and the coecatus group.  3.6.2 Patterson’s D statistic and DFOIL Allele counts are listed for DFOIL and D-statistics in Appendices 5 and 6. D values and p values are reported in Appendices 7 and 8. (1) Among species of the americanus group (Figure 11)    Directional introgression is detected twice within the americanus group: from the common ancestor of H. sansoni and H. americanus to H. ophrys (Figure 11A, DFO=0.267 p <  10-12, DIL=0.235 p=0, DFI=0.019 p=0.737, DOL=0.075 p=0.178) and between H. sp. (ESTU) and H. tarsalis (Figure 11B, D= 0.146, p < 10-12) ! '%!(2) Between the coecatus and the clypeatus group (Figure 12)  Introgression is detected only once between the coecatus group and the clypeatus group. It is detected between H. aztecanus and H. clypeatus and H. pyrrithrix when the fourth species included is the distant coecatus group H. mexicanus (Figure 12C, DFO=0.146 p < 10-12, DIL=0.155 p < 10-12, DFI=-0.015 p=0.634, DOL=-0.026 p=0.391). The same signal is weakened due to a DFO = 0 when the less distant coecatus group species is H. borealis (Figure 12F, DFO=‐0.094 p=0.001, DIL=‐0.13164 p=4 x 10‐6, DFI=‐0.011 p=0.7334, DOL=‐0.055 p=0.07757). This signature (DFO=0 DIL=- DFI=0 DOL=0) is not indicative of any introgression, although it does hint at introgression between H. clypeatus and H. pyrrithrix. Similarly, reciprocal introgression is detected between H. clypeatus and H. pyrrithrix when species 1 is H. sp. (SLCTY) and species 4 is either H. mexicanus (Fig.12D, DFO=-0.096 p=0.003, DIL=-0.167 p < 10-12, DFI=-0.011 p=0.814, DOL=-0.166 p < 10-12) or H. borealis (Fig12E, DFO=‐0.0079 p=0.802 DIL=‐0.14145 p=6 x 10-6 DFI=0.113 p=0.01104 DOL=‐0.157 p=0.00043). Support for introgression between H. pyrrithrix and the clypeatus group disappears when H. sp. (BLNDI) is the fourth species (Fig12A-B), ruling out H. pyrrithrix-specific introgression. The signal is most likely from a ghost taxon from the Southern clade that is closely related to H. pyrrithrix.   (3) Between H. sp. (ROBRT) and the VCC clade (Figure 13)  H. sp. (ROBRT) may have hybridized with both the coecatus and the clypeatus group. Introgression is detected between H. sp. (ROBRT) and H. sp. (SLCTY) when the third species is H. festus (Fig.13B D=0.360, p < 10-12) or H. jucundus (Fig.13C, D=-0.405, p < 10-12). Introgression is also detected between H. festus and H. sp. (ROBRT) when the clypeatus group is excluded (Fig.13D, D=-0.166, p < 10-12).   ! '&!(4) Between H. decorus and the americanus group (Figure 14)  Introgression is detected with both D-statistics (Fig.14A, D= 0.17 p < 10-12) and DFOIL (Fig.14A, DFO=-0.164 p < 10-12, DIL=-0.116 p < 10-12, DFI=-0.004 p=0.883, DOL=0.117 p < 10-12) between H. decorus and H. ophrys when compared to H. zapotecanus. When the sister to H. decorus is a more distant species (H. cambridgei, H. oregonensis, H. jucundus, H. festus), the introgression signal between H. decorus and H. ophrys disappears (Fig.14B-E), and D-statistics introgression signals emerge instead between H. ophrys and the more distant species H. oregonensis (Fig.14C, D= -0.11 p=1.6 x 10-11), H. jucundus (Fig.14D, D=-0.15 p < 10-12), and H. festus (Fig.14E, D=-0.1 p=2 x 10-12). Introgression is also detected between H. oregonensis and the americanus group (H. ophrys/H. sp. (ESTU)) using DFOIL (Fig.14C, DFO=0.104 p < 2 x 10-10, DIL=0.111 p < 10-11, DFI=0.068- p=0.008, DOL=0.085 p < 9 x 10-4). DFOIL tests do not detect introgression between H. ophrys and H. jucundus or H. festus. These DFOIL signatures are inconsistent with a lack of introgression (expected signature: DFO=0 DIL=0 DFI=0 DOL=0), and they could be caused by multiple introgressions (DFO=+ DIL=+ DFI=0 DOL=+). Unexpected introgression detected between the americanus group and distant species H. festus, and H. jucundus (represented by DFO=+ DIL=+) may conceal introgression between H. decorus and H. ophrys detected using D-statistics (represented as the positive DOL=+).       ! ''!4. Discussion  Using transcriptomes, we were able to reconstruct a phylogeny for 34 species of Habronattus with high confidence at most nodes using 1,884 orthologous loci. While previous phylogenies for Habronattus were poorly resolved and limited to a few mitochondrial and nuclear genes (Maddison and Hedin 2003), this improved phylogeny will permit further research on character evolution, sexual selection, and hybridization, within a robust phylogenetic framework. We also integrated Ef1-a and 16SND1 transcripts with Sanger sequencing data from Maddison and Hedin (2003), clarifying some of their results and placing our data in the context of a more complete phylogeny. Based on differences in the mitochondrial and nuclear phylogenies, we infer several possible instances of mitochondrial introgression. We identified extensive nuclear genetic discordance indicative of ILS or introgression within the VCC group and the americanus group using BCA, and we identified several instances of nuclear introgression across the transcriptome phylogeny using D-statistics and DFOIL.   4.1. Habronattus phylogeny The concatenated nuclear phylogeny is generally in agreement with previous work, finding strong support for the monophyly of the AAT, VCC, and americanus group (Maddison and Hedin 2003; Griswold 1987). The positions of H. hallani, pugillis group, altanus/decorus/banksi groups, H. sp. (ROBRT), as well as the internal relationships of the VCC group and the americanus group, are also clarified with this phylogeny. The VCC clade has very low dominant CFs in the BCA primary concordance tree, potentially indicating that the group is still in the early stages of divergence with widespread incomplete lineage sorting and possibly also ongoing hybridization. The americanus group on the other hand, has stronger genetic concordance (high ! '(!dominant CF values) even though it is also very recently diverged (BEAST analysis, Maddison and Leduc-Robert 2013). There is only mild support for the monophyly of Habronattus in the concatenated nuclear phylogeny, and there is considerable variation in the position of Pellenes across analyses. The possibility of non-monophyly of Habronattus has been previously raised based on similarly variable positions for Pellenes specimens within a set of phylogenetic analyses (Maddison and Hedin 2003). However, non-monophyly of Habronattus is unexpected, because the group is well defined morphologically. For instance, the elbowed tegular apophysis (a component of the male palp), a trait unique to Habronattus and shared by most species (with the exception of secondary loss in the coecatus group), is not found in any Pellenes species. This trait appears to be a synapomorphy consistent with monophyly of the genus; however, the molecular boundary delineating Habronattus and Pellenes was difficult to resolve with the specimens we sampled. Pellenes cf. levii-GLR106 (transcriptome), a North American specimen, when it is integrated with 16SND1 data from additional species (Fig. 7), pairs with the ingroup Pellenes specimens (Sanger specimens: HA093, HA510, HA430, HanMN105, MAA2005), also from North America. These same Pellenes (Sanger) specimens are outgroups in the Ef1-a phylogeny (Fig. 9). Because Pellenes are also distributed in Asia (Logunov 1999) and Europe (Fiser and Azarkina 2005; Buchholz 2007), a broader global sample of Pellenes specimens and other closely related groups (e.g., Harmochirus) should be included to better tease apart relationships at the base of the Habronattus tree and determine if the genus is monophyletic. With a better-resolved phylogeny, we can more accurately map Y chromosome origins onto the Habronattus phylogeny. The ancestral sex chromosome state for Habronattus species is XXO!/XXXX" (Maddison 1982). An X-autosome fusion producing an XXY!/XXXX" ! ')!karyotype, and a second autosome-autosome fusion producing an XXXY!/XXXXXX" karyotype, are also found in multiple species (Maddison and Leduc-Robert 2013). We reduce the maximum number of origins from 8 to 15 (Maddison and Leduc-Robert 2013) to 8 to 14. Because there is strong evidence that the altanus (XXY!) and decorus (XXXY!) groups are sisters and that altanus/decorus group is sister to the banksi group (XXY!), we infer a single origin of Y chromosomes for this group and an additional fusion in the decorus group to yield an XXXY! from XXY! (reducing the total from three possible to two possible fusions). H. hallani (XXY!) is sister to the pugillis group (XXO! and XXY!); thus, we can either infer a single origin of Y in H. hallani, followed by a loss in H. pugillis, or two independent origins with no losses. Other inferred origins are unchanged by the new phylogeny; we still count 2 to 4 origins in the coecatus group, 1 to 3 origins in the viridipes group, one origin in H. cf. paratus, one in H. sp. (MACHAL), and a single origin in the dorotheae group.   4.2. Introgression in Habronattus We examined nuclear and mitochondrial phylogenetic signals discordant to the species tree (the concatenated nuclear phylogeny, Fig. 3) for signs of possible introgressive hybridization. Because the mitochondrial genome is haploid and maternally inherited, mitochondrial introgression signals are undiluted by recombination from backcrossing and carry signatures of directional introgression from important processes such as sex-biased hybridization (Avise 2001). To assess nuclear genetic discordance for introgression, we first documented nuclear signals discordant with the species tree using a BCA, and then explicitly tested for introgression using D-statistics. BCA is a good first step to assess the extent of genetic discordance and the degree of divergence between species (Ané et al. 2007; Baum 2007) insofar ! '*!as it quantifies all of nuclear gene discordance from both ILS and hybridization. This information can be useful to interpret introgression in the context of overall discordance, and to estimate how much of the genome may have introgressed (estimated as a CF credibility interval; Baum 2007). Alternatively, D-statistics explicitly test for introgression and can determine directionality. D-statistics are also more sensitive to weak signals of introgression because they are based on shared alleles rather than gene trees. Gene trees, especially from genes with low informativeness due to low variability (a problem arising when working with closely related species like Habronattus), may create considerable noise in the BCA resulting in wider CF credibility intervals and possibly overestimated low CF values (Chung and Ané 2011). Only a small fraction of orthologs meet the criteria for gene tree construction and inclusion in the BCA (i.e., they are long enough to be informative and present in all species), while D-statistics can include all nuclear sequencing data because it does not depend on a gene tree construction step. Simulations indicate that introgression as low as 1% of the genome can yield positive D-statistic results (Good et al. 2015). Consequently, relative to BCA, D-statistics are expected to better detect ancient introgression that may result in only short introgressed genomic regions (Sankararaman et al. 2012) and in infrequent hybridization that may have not resulted in widespread nuclear introgression. Such low levels of introgression may be overinflated as CFs in the BCA, obscured during the gene tree construction process, or not sampled at all using a gene tree approach. D-statistics, while highly sensitive to introgression, can also be skewed by other factors that are difficult to account for like multiple introgression events, ancient population structure, and selection (Durand et al. 2011). Multiple introgression events could reduce the power of the test if introgression occurred between sister taxa (Durand et al. 2011), or if multiple introgression ! '+!events are detected together and conflict (Pease and Hahn 2015). Patterns identical to introgression can be produced as a result of ancient population structure (Slatkin and Pollack 2008), and possibly also selection (Durand et al. 2011). A lineage with a much higher substitution rate might increase the likelihood of shared alleles with distant species (Pease and Hahn 2015). Lineages undergoing parallel selection in similar environments could also have a greater than expected number of shared alleles as a result of convergence. A larger sample of loci should reduce the effects of selection, since selection differences are unlikely to be substantial across the genome.  Because D-statistics are the only analyses to explicitly test for introgression, ILS cannot be ruled out as the cause of a discordant signal when only mito-nuclear discordance and BCA results is used to assess discordance. However, there are several indications that introgression may be the cause of a discordant signal rather than ILS without explicit testing. We infer that the mitochondrial grouping of morphospecies that are phylogenetically distant in the species tree is strongly suggestive of introgression because the likelihood of ILS is decreased with increasing phylogenetic distance due to deeper coalescence times (Pamilo and Nei 1988). Sympatric species grouping together in the mitochondrial but not the nuclear phylogeny may also be an indication of ongoing hybridization (Funk and Omland 2003). BCA CF credibility intervals can also be compared to determine if they are equivalently supported as expected under ILS, or if one contradictory clade has a greater CF, as expected under introgression. Non-overlapping CF credibility intervals for two contradictory secondary CFs can be thought of as comparable to an excess of ABBA or BABA pattern resulting in a D-statistic significantly different from 0. By focusing more broadly on hybridization across the Habronattus phylogeny, and by using sensitive detection methods such as D-statistics, we were able to determine whether ! ',!introgression occurs between distantly related Habronattus clades. Most hints of hybridization detected in Habronattus prior to our analysis were restricted to closely related species and subspecies (amicus group - Hedin and Lowder 2009, coecatus group - Maddison and Hedin, unpubl., the pugillis group - Maddison and McMahon 2000, H. americanus - Maddison and Blackburn 2015), with the exception of more distant mitochondrial introgression detected between the clypeatus group and the coecatus group (Maddison and Hedin 2003). We detected both introgression between closely related americanus group species and between several more distantly related clades (e.g., clypeatus group and coecatus group, H. sp. (ROBRT) and the clypeatus group, americanus group and H. decorus). These results are unexpected based on levels of divergence in courtship behaviour and morphology between clades. Nuclear introgression did not always co-occur with possible mitochondrial introgression. Significant admixture highlighted by both the BCA and D-statistics was almost exclusively detected in closely related species that are currently sympatric, while more ancient introgression signals between species that do not share a contact zone had more minimal nuclear introgression better captured by D-statistics.  4.2.1. Introgression within species groups americanus group -  We identified two instances of introgression in the americanus group using D-statistics: between H. sp. (ESTU) and H. tarsalis and between H. ophrys and H. americanus/H. sansoni (Fig. 11). The BCA (Fig. 9) identifies both of these instances of introgression as substantial CFs as well. While there is no obvious co-occurring mitochondrial introgression for H. sp. (ESTU) and H. tarsalis, there is mitochondrial introgression detected between the sympatric H. ophrys and H. americanus. They are grouped together in the ! '-!mitochondrial phylogeny (Fig. 6, 100% mitochondrial bootstrap support), but not in the nuclear phylogeny (Fig. 3, 100% nuclear bootstrap support). DFOIL tests indicate that introgression occurred between H. americanus/H. sansoni and H. ophrys (rather than H. ophrys and H. americanus only), and this is in accordance with BCA results which link H. americanus/H. sansoni and H. ophrys together (CF = 0.272-0.339). These results could be explained by introgression prior to H. sansoni/H. americanus speciation, or by multiple introgressions (between H. sansoni and H. ophrys and between H. americanus and H.ophrys). Introgression could also explain why H. ophrys and H. americanus are resolved as sister taxa by many nuclear data subsets (Fig. 5). Morphological and behavioural differences in the americanus group are relatively few compared to other clades like the decorus group (Maddison and Hedin 2003). While each americanus group species has a distinct courtship display, they all have similar visual, acoustic and vibratory elements (W.P. Maddison, pers. comm.). Americanus group male palps are also distinctive in that the tibial apophysis is bifurcated, and the embolus and tegular apophysis are consistent in length and shape (Griwold 1987, Fig. 184-186). Among americanus group species, male ornaments exhibit the most variation. For example, ornaments range in colour from green (H. ophrys) to blue and red (H. americanus). It is possible that introgression could have influenced the rapid evolution of sexually selected male americanus group ornaments. For example, tufts above the front pair of eyes, a male ornament unique to the americanus group, is found only in H. sansoni and close relatives (H. kubai, some H. americanus populations), and more distant H. ophrys, with which our data suggests a shared history of hybridization.   ! (.!coecatus group -  There are a total of 15 secondary CFs greater than 0.05 detected within the coecatus group (and many more under 0.05), indicating high levels of nuclear discordance within the clade. This number would likely increase with a complete sample of coecatus group species (we only sampled 7 of 16 described species). Because the origin of the coecatus group is predicted to be very recent (see BEAST analysis,  Maddison and Leduc-Robert 2013), ILS and hybridization are both likely contributors to the high levels of discordance detected. ILS cannot be ruled out as the only cause of discordance because we did not explicitly test for introgression using D-statistics. However, there are non-overlapping secondary CFs for contradictory clades, indicating that introgression may have occurred. For example, there is a significant CF linking H. festus and H. captiosus (CF = 0.130–0.172) and no significant CF for the contradictory clade H. festus and H. borealis, which could be indicative of introgression between H. festus and H. captiosus. There are also several indications of possible mitochondrial introgression. Pairs H. sp. (BLNDI)/H. captiosus, and H. mexicanus/H. pyrrithrix are sister taxa in the mitochondrial phylogeny, but not in the concatenated nuclear phylogeny.  It is difficult to say if courtship behaviour and morphology is divergent enough between coecatus group species to completely prevent hybridization. Short branch lengths in the 16SND1 tree and low primary CFs in the BCA indicate low genetic divergence within the group. Morphological differences are concentrated in male ornaments, which vary widely in colours and styles, while coecatus group genitalic morphology is more consistent across the clade (Griswold 1987). Coecatus group species all share the loss of an elbowed tegular apophysis, and all species have similar lengths, shapes, and rotations of the embolus and tibial apophysis (Griswold 1987). Courtship is very complex and unique in each species, although they are all variants of a display composed of similar motifs of display elements performed in a similar order (Elias et al. 2012). ! (%!The possibility of widespread hybridization within the coecatus group despite these differences should be explicitly tested using D-statistics with a complete coecatus group phylogeny.  4.2.2. Distant introgression coecatus and clypeatus group -  There is a strong signature of mitochondrial introgression between the coecatus group and H. clypeatus (Fig. 6), as well as a corresponding nuclear signal of introgression detected by DFOIL (Fig. 12). DFOIL results indicate that introgression must have occurred more recently than the H. borealis/H. pyrrithrix split within the coecatus group, and therefore does not pre-date divergence of the clypeatus and coecatus clades. H. clypeatus also falls within the coecatus group (rather than the clypeatus group as expected) in the concatenated mitochondrial phylogeny with 100% bootstrap support. There are some very low and significant CFs that also link H. clypeatus with coecatus group species together in a clade: H. clypeatus/H. mexicanus (CF = 0.006-0.017), H. clypeatus/H. virgulatus (CF = 0.004 - 0.017), H. clypeatus/H. pyrrithrix (CF = 0.002 - 0.012). In addition, there are 32 other CFs averaging 0.01 link species from the clypeatus group and coecatus group. While ILS or noisy gene trees could explain low-level discordance, some of these CFs could be a result of introgression. Patterns of mitochondrial introgression detected from H. sp. (CHIH)-HA292 and H. velivolus-HA659 in Maddison and Hedin (2003) are also replicated in the 16SND1 tree (Fig. 7). H. sp. (CHIH)-HA292 (clypeatus group member) falls within the coecatus group (and H. clypeatus-GLR227 is its sister), while the other specimen of H. sp. (CHIH)-HA272 falls clearly within the clypeatus group. H. velivolus-HA659 (clypeatus group) falls within the coecatus group as well, while H. velivolus-HA661 falls within the clypeatus group.  ! (&!These results, coupled with those of Maddison and Hedin (2003), could be an indication of frequent and recent introgression between the clypeatus and coecatus group. This is unexpected because these clades differ substantially and consistently in morphology and courtship behaviour (Elias et al. 2012; Elias, Mason, and Maddison 2003; Maddison and Hedin 2003). Both groups have highly complex ornamentations and courtship displays. The coecatus group courtship display is composed of more stages, and its motions are different from those of the clypeatus group (Elias et al. 2012). For example, early stages of the courtship display have the clypeatus group moving with their front legs oriented downward, while the coecatus group keep their legs up (W.P. Maddison, pers. comm.). Vibratory and acoustic signals have been shown to be very important components of the display (Elias et al. 2003), and these signals are also very different between the two groups (Elias et al. 2003; Elias et al. 2012). Genitalia morphology also differ considerably between the two groups. Coecatus group species all share the loss of an elbowed tegular apophysis that is present in all clypeatus group species (Griswold 1987). The tibial apophyses of the clypeatus group are thinner and hook-like while they are thicker and more triangular in the coecatus group. The female epigyna of each group also have correspondingly different forms (Griswold 1987; W.P. Maddison, pers. comm.), consistent with a lock and key coevolutionary process. The possibility of hybridization despite these substantial clade differences is worth noting. Because of an evolutionary lag between male sexually selected traits and female preferences (Schluter and Price 1993), species-specific differences in male courtship traits do not always result in behavioural isolation. However, clypeatus group females are very likely to have evolved preference to clypeatus rather than coecatus group displays (and vice versa). Introgression between these two groups strongly suggests that distant hybridization is ! ('!possible despite morphological and behavioural differences and could have resulted in introgression of distant genes.  clypeatus group and H. sp. (ROBRT) -  D-statistics indicate that the clypeatus group has a strong signal of nuclear introgression with H. sp. (ROBRT) (Fig. 13). This signal is also reflected in a larger CF for the clypeatus group/H. sp. (ROBRT) clade (CF = 0.1705-0.205) compared to the coecatus group/H. sp. (ROBRT) clade (CF = 0.027-0.058). Introgression detected between H. sp. (ROBRT) and the coecatus group is also detected using D-statistics. However, because introgression with the coecatus group is only detected in the absence of the clypeatus group, and because of the large difference in CFs found by the BCA, the introgression signal detected between H. sp. (ROBRT) and the coecatus group is most likely a result of introgression with the unsampled clypeatus group, which is acting a ghost lineage. There is not a clear and comparable mitochondrial introgression signal linking H. sp. (ROBRT) and the clypeatus group. Instead, H. sp. (ROBRT) is sister to the clypeatus and coecatus group in the mitochondrial phylogeny, and all four H. sp. (ROBRT) specimens in the 16SND1 tree fall within the coecatus group (Fig. 7). The BCA found substantial genomic support for H. sp. (ROBRT)/clypeatus group/coecatus group as a clade (CF = 0.174-0.222) compare to the contradictory clade H. sp. (ROBRT)/viridipes group, which has much less genomic support (CF = 0.039 – 0.068). ILS, introgression with an ancestral coecatus group/clypeatus group lineage, and introgression with the clypeatus group only, could all have contributed to these estimated CFs. More ancestral introgression could explain why there is strong genomic support (large CFs) for H. sp. (ROBRT) and the entire clypeatus group, or clypeatus/coecatus group, rather than with one or a few particular clypeatus/coecatus species specifically. However, the possibility of introgression in ! ((!ancestral lineages that pre-date divergence of the clypeatus and coecatus group clade requires further testing. It is also possible that denser sampling within the clypeatus group could uncover a lineage with recent or ongoing introgression with H. sp. (ROBRT). The phylogenetic position of H. sp. (ROBRT) has been historically difficult to resolve based on molecular data, morphology, and courtship behaviour (Maddison and Hedin 2003). It is similar to both the viridipes and clypeatus groups in many ways, and introgression could be in part responsible for this unique combination of viridipes and clypeatus group traits. H. sp. (ROBRT) has a raised-V setae ridge on the male carapace and both 1st and 3rd legs are modified like the viridipes group (Maddison and Hedin 2003). On the other hand, its ventral and dorsal abdominal stripes are very similar to the clypeatus group (W.P. Maddison, pers.comm.). Males from some populations of H. sp. (ROBRT) have red-purple bumps on their 3rd legs, a trait found otherwise only in clypeatus group species H. sp. (CHIH), H. formosus, and H. velivolus (a species sympatric with H. sp. (ROBRT)) (W.P. Maddison, pers. comm.). H. sp. (ROBRT) and clypeatus group  males (H. velivolus, H. aztecanus) also share a unique zig-zag pattern emerging from within the frontal eyes (W.P. Maddison, pers.comm.).   americanus group and distant clades -  Based on phylogenetic distance, mitochondrial and nuclear introgression detected between the americanus group and the decorus/altanus/banksi group, as well as nuclear introgression with H. oregonensis, H. jucundus, and H. festus, were most unexpected. There is a strong signal of distant mitochondrial introgression between the americanus group and the decorus/altanus/banksi group, which together form a clade in the mitochondrial transcriptome phylogeny (Fig. 6) and the 16SND1 phylogeny (Fig. 7) but are not a clade and distant from each other in the nuclear phylogeny (Fig. 3). It is unlikely that ILS is the ! ()!cause of this topology, because the common ancestor of H. ophrys and H. decorus is relatively ancient and because the phylogenetic signal is strong across different mitochondrial data subsets. Moreover, D-statistics show a weak signal of nuclear introgression between H. decorus, but not H. zapotecanus (banksi group) (Fig. 14). Ancient introgression at the base of the americanus group with the decorus/altanus/banksi group may have resulted in negligible nuclear introgression and fixation of a single mitochondrial type, which was followed by diversification of extant species groups. Because introgression would have been ancestral to the diversification of these groups, the ancestral hybridizing decorus group and the americanus group members may have looked and behaved quite differently from extant species. However, extant americanus group and decorus group species have very different morphology (Griswold 1987) and behaviour (W.P. Maddison, pers. comm.) that would predict reproductive isolation. For instance, the americanus group male palp has an unrotated embolus and a thin bifurcated tibial apophysis, while decorus group palps have thicker and more triangular tibial apophyses and a rotated embolus (Griswold 1987).  The americanus group and the coecatus group, clypeatus group, and oregonensis group are even more genetically, morphologically, and behaviourally different (Griswold 1987, Maddison and Hedin 2003), yet introgression between these groups also emerged from D-statistics and DFOIL tests. While it is not out of the question that introgression has frequently occurred between the americanus group and the rest of the Habronattus clade, signals of introgression between H. ophrys and H. oregonensis, H. jucundus, and H. festus could all be the result of a single ghost lineage belonging to a clade distant to H. ophrys. It is also possible that introgression occurred in an ancestral lineage prior to clade diversification. More extensive tests ! (*!for distant introgression would be required to clarify the specific origin and timing of these signals.   4.2.3. Potential drivers of hybridization Our results suggests that distant introgression in Habronattus may be much more pervasive than we initially predicted based on divergent male traits. We not only found introgression between the clypeatus and coecatus groups (like in Maddison and Hedin (2003)), but also between H. sp. (ROBRT) and the clypeatus group, and between the americanus group and distant species. Sexual selection is an important process driving diversification and the evolution of male traits in these Habronattus groups (Masta and Maddison 2002). These differences in courtship behaviours and male ornaments are expected to result in behavioural reproductive isolation because they should generally coevolve with female preferences for those traits (Panhuis et al. 2001). Directionality of introgression, determined by DFOIL and the mitochondrial phylogeny, hint at what processes may be driving hybridization despite divergent morphology and behaviour. H. clypeatus falls within the coecatus group in the mitochondrial phylogeny, and several species of the clypeatus group are polymorphic for coecatus group mitochondria in Maddison and Hedin (2003). Nuclear introgression however is reciprocal. Directionality of mitochondrial introgression but not nuclear introgression could be the result of sex-biased hybridization, driven by a difference in female discrimination (Avise 2001), causing coecatus females to sometimes hybridize with clypeatus group males but not the other way around. H. pugillis females have been shown to sometimes choose foreign males from divergent populations (Hebets and Maddison 2005), a preference that could be driven by a bias for vibratory signals ! (+!that are more complex (Elias 2006). This could be a consequence of a co-evolutionary arms race between the sexes, where males evolve more complex signals in response to females evolving resistance to those traits (Holland and Rice 1998). A female evolving under sexual antagonism could potentially mate with an allospecific male if their courtship display exploits female sensory biases or if the female is more likely to respond to novel signals from foreign males to whom she has not yet evolved resistance (Elias et al. 2012). Both of these scenarios could promote hybridization and would predict that the introgressing species with more complex courtship behaviour would be more likely to be the “donor”. In fact, these predictions are consistent with previous observations of courtship and introgression in Habronattus. The “donor” species has a more complex courtship than the “recipient” species in instances of mitochondrial introgression detected in the coecatus group (Maddison and Hedin 2003) and the amicus group (Hedin and Lowder 2009).  Demographic characteristics of hybridizing species, such as differences in population sizes (Lepais et al. 2009) and dispersal behaviours (Funk and Omland 2003), could also cause asymmetric introgression. The abundant species is more likely to be the “donor” during nuclear introgression (Levin et al. 1996), which could explain the direction of nuclear introgression from H. sansoni/H. americanus to H. ophrys. Differences in population sizes can also have the reverse effect on mitochondrial introgression because females of the rarer species are more likely to mate with males of the abundant species, thus becoming the mitochondrial “donor” (Funk and Omland 2003). Female-biased dispersal could also produce a similar pattern (Wirtz 1999). Consequences of abundance differences between species on introgression directionality extend to biological invasions and range expansions (Currat et al. 2008). As a result, introgression patterns can provide insights on historical species distributions and changes. However, because most of our ! (,!introgression tests were not species-specific, inferences into historical events that caused introgression are difficult to make; thus, we can only describe the extent of historical introgression as opposed to particular events and movements of species in the past. More in-depth understanding about the natural history, relative species abundances, and geographic distributions of Habronattus would be useful to determine if demographic effects are major drivers of introgression. However, general trends of introgression observed here are consistent with limited information on the geographic proximity among introgressing species. In the VCC clade, for example, a lack of introgression involving the viridipes group could be attributed to their northern and eastern distribution. Most introgression that we detected in the VCC clade is constrained to species from the southern US and Mexico, where the clypeatus group and most of the coecatus group is distributed.  Many extant Habronattus live in sympatry (Griswold 1987), and these species could theoretically come into contact and hybridize. Species nearer to the Northern (e.g., Yukon) and Southern (e.g., Costa Rica) edges of the Habronattus range may be more geographically isolated, and have fewer opportunities for hybridization. On the other hand, species from the arid southwestern desert of the United States and northern Mexico, where Habronattus are most concentrated, usually have overlapping distributions with multiple other Habronattus species (Griswold 1987), or else are closely distributed so that a contact zone in the recent past could be inferred (e.g., H. pugillis, Maddison and McMahon 2000).  Divergent genitalia morphology may not prevent successful introgression between phylogenetically distant species. For example, our data indicates a history of hybridization between the clypeatus group and coecatus group, which possess very different genitalia.  ! (-!Similarly, while reduced hybrid viability has been detected in divergent populations of H. pugillis (Masta and Maddison 2000), hybrids could be viable and backcross, resulting in distant introgression.    4.2.4. Evolutionary consequences of introgression Introgression in Habronattus may have been frequent enough to result in detectable nuclear admixture, but infrequent enough so as not to behave as a homogenizing force reducing genetic and morphological diversity (Slatkin 1985). These findings agree with other studies in subpopulations of H. americanus (Maddison and Blackburn 2015) and H. pugillis (Maddison and McMahon 2000) showing strong selection for male ornaments and courtship differences in the face of gene flow. Distantly related species (e.g., the clypeatus and coecatus groups, H. sp. (ROBRT)), and closely related species with substantial admixture (e.g. H. ophrys and H. americanus/H. sansoni), appear to have retained morphological distinctiveness despite a history of hybridization. Historical introgression in the species-rich americanus group (11 species) and the VCC group (38 species) does not appear to have slowed down diversification of these clades, both of which are composed of many species and have remarkable diversities of male ornaments. While introgression may not have substantially reduced divergence among Habronattus species, it is unclear whether it has played a creative evolutionary role, either by promoting diversification or by influencing the phylogenetic distribution of traits. Distant introgression, which we detected in several clades, is more likely to have adaptive effects on lineages because novel and potentially adaptive genetic combinations are more likely to form as a result of introgression when there is more time to accumulate genetic differences (Stelkens and Seehausen 2009). Introgression could also create adaptive potential by increasing the standing variation of ! ).!hybridizing lineages, which could facilitate subsequent diversification (Seehausen 2013). There are an increasing number of documented cases of adaptive introgression (Huerta-Sánchez et al. 2014; Norris et al. 2015; Heliconius Consortium 2012) and introgression-facilitated diversification in animals (Streicher et al. 2014; Seehausen 2004). Given the strength of sexual selection in Habronattus (Masta and Maddison 2002), loci implicated in sexually selected traits could be under strong selection if they were exchanged between hybridizing species. There are several examples of morphological traits under sexual selection in Habronattus whose phylogenetic distributions could be explained by introgression (many are mentioned in sections 4.2.1 and 4.2.2). In particular, the red-purple bumps and dark spots only found on the 3rd legs of males of some populations of H. sp. (ROBRT) as well as in the clypeatus group member H. velivolus (with which H. sp. (ROBRT) is sympatric) and H. sp. (CHIH), and the distinctive “eyebrows” of H. ophrys and H. sansoni have phylogenetic distributions that match predictions based on introgression rather than convergence. What is most evident and interesting upon observation of these traits is the details of the trait shared amongst distantly related species. While almost all VCC group species have colourful ornaments on their 3rd legs (Griswold 1987), the similarity between the purple-red bumps and black spots found on the 3rd legs of clypeatus group species and H. sp. (ROBRT) is striking. It seems unlikely that this similarity evolved as a result of such refined convergence in female preferences. A possible explanation is that the trait was exchanged through introgression, and that the trait’s details were retained in the recipient lineage. Hybridization has been previously invoked as the best explanation for homoplasy of male ornaments in hybridizing populations of H.pugillis, in part because homoplasy was found in the details of male ornaments between hybridizing populations (Maddison and McMahon 2000).  ! )%!4.2.5. Future research Sparse taxonomic sampling, both at the species and population level, limits our abilities to make inferences from D-statistics (Eaton et al. 2015). We did not include more than a single population per species and instead focused on constructing a more complete species tree. While we were able to produce a broad phylogeny of the group, all introgression signals are potentially affected by ghost taxa; therefore they cannot be attributed to specific species in most cases (Eaton et al. 2015). With a more complete phylogeny, the temporal and spatial sequence of introgression and speciation events can be more carefully mapped to the phylogeny, and historical biogeographical patterns and examples of adaptive introgression may emerge more readily (examples: Streicher et al. 2014; Eaton and Ree 2013; Cui et al. 2013). Increasing sampling at the population level would also test for effects of population choice (Pease and Hahn 2015; Eaton et al. 2015) and help identify biases arising from ancient population structure (Eriksson and Manica 2012) and population bottlenecks (Durand et al. 2011). This could be especially important for species of Habronattus with significant population structure (pugillis group: Maddison and McMahon 2000; americanus group: Maddison and Blackburn 2015). More in-depth sampling and testing for introgression between distant clades, particular between H. sp. (ROBRT) and the clypeatus group and between the americanus group and distant clades, could also help identify the timing of introgression relative to divergence of clades. This would clarify whether hybridization actually occurred between species with very different courtship behaviours, or if hybridization mostly occurred in ancestral lineages with less divergent courtship traits. With introgression more fully characterized across Habronattus, future research should determine the extent of introgression’s contribution to adaptation and diversification. ! )&!Widespread introgression in the highly diverse VCC clade, particularly between more distant groups H. sp. (ROBRT), coecatus and clypeatus, could be indicative of a correlation between introgression and diversification. Identifying the genetic basis of ornaments that may have introgressed is the next step to determine if hybridization played a role in their evolution. Because loci in this study have unknown genomic positions and low variability, it is unclear if the signal of introgression is spread throughout the genome, or if it is concentrated at particular regions. Understanding how the introgression signal is distributed and if there are signs of selection at introgressed loci would clarify whether introgression has been adaptive (Hedrick 2013).               ! )'!5. Conclusions We produced a highly resolved phylogeny for Habronattus and determined the relative contributions of hybridization and incomplete lineage sorting to genetic discordance in the group. We found that hybridization has been historically common in Habronattus, and has resulted in significant nuclear introgression in some instances (e.g., the americanus group, H. sp. (ROBRT)) and minimal nuclear introgression accompanied by strong mitochondrial introgression in others (e.g., the coecatus and clypeatus groups). Widespread introgression between both distant and closely related species indicates that only partial reproductive isolation has evolved across the Habronattus phylogeny. Frequent introgression may continue to occur between Habronattus species groups with divergent male ornaments and courtship behaviours (e.g. between the clypeatus group and the coecatus group), although ancestral introgression that precedes diversification of extant courtship behaviours and ornaments is also a possibility for some groups (e.g., H. sp. (ROBRT) and the clypeatus group, americanus group and distant clades).          !"Species 1 Species 2 Species 3 Species 4 Outgroup Corresponding figure(1A) H. sansoni H. americanus H. ophrys H. tarsalis H. signatus 11A(1B) H. ophrys H. tarsalis H. sp.   (ESTU) - H. signatus 11B(1C) H. sansoni H. americanus H. sp.   (ESTU) - H. signatus 11C(2A) H. aztecanus H. clypeatus H. mexicanus H. pyrrithrix H. ophrys 12A(2B) H. sp. (BLNDI) H. pyrrithrix H. clypeatus H. aztecanus H. ophrys 12B(2C) H. sp. (SLCTY) H. clypeatusH. sp. (BLNDI) H. pyrrithrix H. ophrys 12C(2D) H. sp. (SLCTY) H. clypeatus H. mexicanus H. pyrrithrix H. ophrys 12D(2E) H. sp. (SLCTY) H. clypeatus H. borealis H. pyrrithrix H. ophrys 12E(2F) H. aztecanus H. clypeatus H. borealis H. pyrrithrix H. ophrys 12F(3A) H. sp. (SLCTY) H. jucundusH. sp. (ROBRT) - H. ophrys 13A(3B) H. festus H. sp. (SLCTY)H. sp. (ROBRT) - H. ophrys 13B(3C) H. festus H. jucundus H. sp. (ROBRT) - H. ophrys 13C(4A) H. ophrys H. sp.   (ESTU) H. zapotecanus H. decorus H. signatus 14A(4B) H. ophrys H. sp.   (ESTU) H. cambridgei H. decorus H. signatus 14B(4C) H. ophrys H. sp.   (ESTU) H. oregonensis H. decorus H. signatus 14C(4D) H. ophrys H. sp.   (ESTU) H. jucundus H. decorus H. signatus 14D(4E) H. ophrys H. sp.   (ESTU) H. festus H. decorus H. signatus 14E(4F) H. zapotecanus H. decorus H. ophrys - H. signatus 14A(4G) H. cambridgei H. decorus H. ophrys - H. signatus 14B(4H) H. oregonensis H. decorus H. ophrys - H. signatus 14C(4I) H. jucundus H. decorus H. ophrys - H. signatus 14D(4J) H. festus H. decorus H. ophrys - H. signatus 14ETables and figuresTable 1 Species tested for introgression with Patterson’s D statistic (4 species at a time) or DFOIL  (5 species at a time). Results are shown in Figures 11-14. (3) H. sp. (ROBRT) and VCC clade(2) clypeatus and coecatus group(1) americanus group(4) H.decorus and americanus groupFigure 1. Summary of H. ophrys de novo transcriptome assembly and filtering steps (1. Abundance 2. Redundancy 3. Paralogy 4. Coverage 5. Coding Regions).55Figure 2. Flow chart describing how data was partitioned into matrices (green circles) for RAxML phylogenetic analyses. 56EvarchaPellenesH.conjunctusH.hirsutusH.signatusH.ustulatusH.geronimoiH.sp. (ESTU)H.americanusH.sansoniH.ophrysH.tarsalisH.altanusH.decorusH.sp. (CHMLA)H.zapotecanusH.cambridgeiH.icenogleiH.oregonensisH.hallaniH.pugillisH.sp. (SUNGL)H.sp. (ROBRT)H.calcaratusH.jucundusH.aztecanusH.clypeatusH.sp.(SLCTY)H.mexicanusH.sp.(BLNDI)H.pyrrithrixH.virgulatusH.festusH.borealisH.captiosus100100100100100100100100100100100100100100100100100100100831001001001001009710010010097****100*viridipes  groupclypeatus groupcoecatus groupVCC groupamericanus groupAAT groupFigure 3. Concatenated nuclear Maximum likelihood phylogeny (lnL = -8016699). The matrix analyzed is a concatenated set of 1884 concatenated genes over 200 bp present in 25 or more species(total length = 2.41 Mbp). Branch lengths are proportional to change.  Constructed in RAxML with 20 search replicates, 1000 bootstrap replicates, GTR+G+I substitution rate, partitioned as: codon pos1,pos2, pos3, noncoding. * indicate uncertain nodes that varied across nuclear ML analyses (i.e., phylogenies constructed from matrices composed of noncoding sites, coding sites, contigs present in 15-24 species, contigs without coding regions, subsets 1-8 matrices (described in Figure 4 and 5).57(A) (B)(C) (D)Pellenes cf. levii100100100 100EvarchaH.conjunctusH.hirsutusH.signatusH.ustulatusPellenesother HabronattusEvarchaPellenesH.conjunctusH.hirsutusH.signatusH.ustulatusother HabronattusEvarchaH.conjunctusPellenesH.hirsutusH.signatusH.ustulatusother HabronattusEvarchaPellenesH.conjunctusH.hirsutusH.signatusH.ustulatusother Habronattus(E) (F)H. icenoglei, H. oregonensis, and H. cambridgei100100100(G) H.cambridgeiH.icenogleiH.oregonensisH.hallaniH.pugillisH. sp. (SUNGL)VCC cladeH.cambridgeiH.icenogleiH.oregonensisH.hallaniH.pugillisH. sp. (SUNGL)VCC cladeH.cambridgeiH.oregonensisH.hallaniH.pugillisH. sp. (SUNGL)H.icenogleiVCC cladeH.icenogleiH.cambridgeiH.oregonensisH.hallaniH.pugillisH. sp. (SUNGL)VCC clade(H)97CN subsetsCN ML treeCN BootstrapsCSNCSMS matrix UG matrix676666*Figure 4. Common topologies at nodes that varied across maximum likelihood phylogenetic analyses. A dark square represents support from that particular matrix listed in the legend to the left: concatenated nuclear (CN) 3.18kb subsets, Concatenated Nuclear (CN) matrix ML tree and CN Bootstrap tree, Non-CodingSites (NCS), Coding Sites (CS), Missing Species (15-24 species, MS), Unidentified Genes (no coding regions, UG).CN bootstrap values are at nodes. Phylogenies A-D show the position of Pellenes cf. levii relative to basal Habronattus species. Phylogenies E-G show the relationships of H. icenoglei and H. oregonensis to each other and to the “right” side of the phylogeny. The * by H. oregonensis at (H) indicates variable positions for this species amongst subsets. 58(E) (F)(D)(A) (B)(C)H. sp. (ROBRT)H. virgulatus10010010010010097H.mexicanusH.virgulatusH.sp. (BLNDI)H.pyrrithrixH.festusH.borealisH.captiosusH.mexicanusH.virgulatusH.sp. (BLNDI)H.pyrrithrixH.festusH.borealisH.captiosusH.mexicanusH.sp. (BLNDI)H.pyrrithrixH.virgulatusH.festusH.borealisH.captiosusH. sp. (ROBRT)viridipes groupcoecatus groupclypeatus groupviridipes group H. sp. (ROBRT)clypeatusgroupcoecatus groupviridipes groupH.sp. (ROBRT)clypeatus groupcoecatus group(G) (H)H. ophrys and H. tarsalisH.sp. (ESTU)H.tarsalisH.ophrysH.americanusH.sansoniH.sp. (ESTU)H.ophrysH.tarsalisH.americanusH.sansoni10010010010010010083Figure 5. Common topologies at nodes that varied across maximum likelihood phylogenetic analyses. A dark square represents support from that particular matrix listed in the legend: Concatenated Nuclear (CN) 3.18 kb subsets, Concatenated Nuclear (CN) matrix ML tree, Non-Coding Sites (NCS), and Coding Sites (CS), Missing Species (MS), Unidentified genes (UG).CN bootstrap values are at nodes. Phylogenies A-C show the different placement of H. sp. (ROBRT) relative to the viridipes, clypeatus, and coecatus groups. Phylogenies D-F show the relationship of H. ophrys and H. tarsalis to the rest of the americanus group. Phylogenies G-H show the possible positions of H. virgulatus within the coecatus group. CN subsetsCN ML treeCN BootstrapsCSNCSMS matrix UG matrix59EvarchaPellenesH.conjunctusH.hirsutusH.signatusH.ustulatusH.paratusH.geronimoiH. sp. (CHMLA)H.zapotecanusH.decorusH.altanusH.sansoniH.sp. (ESTU)H.tarsalisH.americanusH.ophrysH.cambridgeiH.oregonensisH.icenogleiH.pugillisH.sp. (SUNGL)H.hallaniH.calcaratusH.jucundusH.aztecanusH.sp. (SLCTY)H.sp. (ROBRT)H.festusH.virgulatusH.mexicanusH.pyrrithrixH.clypeatusH.borealisH.sp. (BLNDI)H.captiosus997191100537010099561001001009710010086 1001005252 10010078100100100661001001009299americanus group“VCC” clade“AAT” clade100Figure 6. Concatenated Mitochondrial Maximum Likelihood phylogeny (12,233 bases, lnL = -3429108). Constructed in RAxML with 24 partitions (coding genes, rRNA, codon positions 1,2,3), 20 search replicates, 1000 bootstraps, GTR+G+I substitution rate. Branch lengths are proportionalto change. Stars indicate nodes that were not recovered in the bootstrapConsensus tree (bootstrap consensus topology is shown in inset A).**1001005 21005 2H. hallaniH. sp. (SUNGL)H. pugillisH. calcaratus H. jucunduscoecatus groupclypeatus groupH. sp. (ROBRT)AA60Figure 7. Maximum likelihood tree for the 16SND1 mitochondrial region (1047 bp). 160 Sanger sequenced specimens and all transcriptome specimens (marked by black circles) are included. The tree was constrained to the phylogenetic structure of the CM phylogeny (nodes with >90% bootstrap support only). Branch lengths are proportional to change. Constructed in RAxML with 20 search replicates, GTR+G+I substitution model. lnL = -25861.Evarcha Proszynskii GLR135Evarcha Proszynskii d096Bianor MRB055Sibianor aemulus S238Bianor Maculatus d017Harmochirus brachiatus MRB054Harmochirus MRB058Havaika cruciata HavH110Havaika cf. pubens HavK112Havaika S127H.cognatus HA617H.cognatus HA294*Habronattus mataxus 602AH.elegans HA213H.peckhami HA382H.alachua HA527H.contigens HA620H.conjunctus HA179H.conjunctus GLR234Pellenes apacheus HA093Pellenes longimanus HA510Pellenes GLR106Pellenes shoshonensis HA430H.rufescens  HanMN105Habronattus sp. MAA 2006H.hirsutus HA321*H.hirsutus GLR080H.hirsutus HA565H.tranquillus HA675/676H. sp. (LKHSTY) HA498H. sp. (SCOTSD) HA492H.amicus HA387H.ustulatus HA570 H.ustulatus GLR601H.ustulatus HA210H.signatus GLR600H.signatus HA684H.ustulatus HA551H.paratus 85AM/086H.paratus GLR363H. sp. (HUAST) HA624H. sp. (CHUAST) HA623*H. sp. (MACHAL) HA088H.dorotheae HA473H.dorotheae HA621H.geronimoi HA061H.geronimoi GLR267H.banksi HA104H.pochtecanus HA632H.zapotecanus HA441H.zapotecanus GLR339H. sp. (YESOS) HA638H. sp. (CHMLA) HA639H. sp. (CHMLA) GLR352H.iviei HA641H. sp. (CTARA) HA554H.tarascanus HA643H. sp. (SPLEND) HA525 H.carolinensis HA134H.ocala HA612H.cockerelli HA108H.venatoris HA119H.decorus HA051H.decorus GLR132H.texanus HA494H.altanus HA127*H.altanus HA645H.altanus.GLR180H.altanus HA325H.altanus HA124H.americanus HA316*H.tuberculatus HA464H.sansoni GLR066H. sp. (ESTU) GLR287H.sansoni HA368* HA349H.waughi HA129H.ophrys GLR015/023H.kawini HA592H.americanus HA345H.tarsalis GLR297H.sansoni HA540H.americanus.GLR014H.ophrys HA352H.bulbipes HA536H.tarsalis HA591H.mustaciata HA365H.tarsalis HA377H.cambridgei HA651H.cambridgei GLR351H.cambridgei HA649H.cambridgei HA443*H.fallax HA288H.fallax HA166H.fallax HA647H.fallax HA648H.oregonensis HA005H.oregonensis GLR049H.delectus HA467H.icenoglei HA196H.icenoglei GLR283H. sp. (YUCUN) HA635H.sugillatus HA486H.pugillis (Galiuro) M-29GLH. sp. (SUNGL) d436H.pugillis (Sierra Madre) M-10SMXH. sp. (SUNGL) GLR218H.pugillis HA009H.pugillis (Sierrita) M-30SIH.pugillis (Cerro Colorado) M-1CCH.pugillis (Black) M-1BH.pugillis (Galiuro) M-55GLH.pugillis (Sierra Madre) M-5SMXH.pugillis (Tumacacori) M-1TH.pugillis (Atascosa) M-5AH.pugillis GLR236H.pugillis (Santa Catalina) M-20SCH.pugillis HA008H.pugillis (Mule) M-11MH.pugillis (Santa Rita) M-1RH.pugillis (Empire) M-1EH.pugillis (Huachuca) M-27HH.pugillis (Whetstone) M-2WH. sp. (NAYAR) HA281H.hallani 037H.hallani GLR209H.calcaratus calcaratus HA524H.notialis HA471H.calcaratus agricola HA295H.calcaratus maddisoni HA06H.calcaratus maddisoni GLR321H. cf.calcaratus HA497H.moratus HA484/485H.trimaculatus HA559H.viridipes HA044H.orbus HA101/027H.viridipes HA470H. jucundus HA001*H.jucundus HA395H.jucundus GLR320H. sp. (CHIH) HA272H. sp. (CNCTY) HA468H.forticulus HA289H.velivolus HA661*H.aztecanus.GLR347H.aztecanus HA218H.clypeatus HA221H.divaricatus HA563H.clypeatus HA039H.dossenus HA274H. sp. (SLCTY) HA285H. sp. (SLCTY) AS56/GLRH.formosus HA007H.dossenus HA531H.divaricatus HA600H.californicus HA380H.anepsius HA557H.pyrrithrix HA578 H.anepsius HA576 H.ammophilus HA556*H.ammophilus HA580 H.festus HA392H.festus GLR094/088H.schlingeri HA160H.ballatoris HA383H.klauserii HA390H. sp. (ROBRT) HA667H. sp. (ROBRT) HA662H. sp. (ROBRT) HA228*H. sp. (ROBRT) GLR346H.pyrrithrix HA010H.velivolus HA659H.mexicanus HA670 H.tlaxalanus HA674H.zebraneus HA442H.mexicanus GLR353H.pyrrithrix GLR304H.virgulatus HA167H.virgulatus HA682H.virgulatus GLR205H.borealis HA130H.borealis GLR040H. sp. (CHIH) HA292H.clypeatus GLR227H. sp. (NMEX) HA517H.mexicanus HA496H.anepsius HA282*H. sp. (BLNDI) GLR282H.borealis d207H.brunneus HA543H.cuspidatus HA550H.klauserii HA502*H.coecatus HA096H.virgulatus HA652H.mexicanus HA654H.captiosus GLR35661Figure 8. Maximum Likelihood tree for nuclear gene Ef1-a (819 bp). 105 Sanger sequenced specimens and all 33 Habronattus transcriptome specimens (marked by black circles) are included. Branch lengths are proportional to change. The tree was constrained to the phylogenetic structure of the concatenated nuclear phylogeny (nodes with >90% bootstrap support only). Constructed in RAxML with 20 search replicates and GTR+G+I substitution model, lnL = -25861.Pellenes apacheus HA093Pellenes longimanus HA510Pellenes shoshonensis HA430H.conjunctus GLR234H.cognatus HA294*H.alachua HA527H.peckhami HA382H.conjunctus HA179H.elegans HA213H.hirsutus GLR080H.amicus HA387H.ustulatus GLR601H.signatus HA684H.signatus GLR600H.ustulatus HA570 H.ustulatus HA551H.ustulatus HA210H. sp. (CHUAST) HA623*H.dorotheae HA621H.geronimoi GLR267H.geronimoi HA061H. sp. (ESTU) GLR287H.ophrys GLR023,GLR015H.tarsalis GLR297H.mustaciata HA365H.ophrys HA352H.tuberculatus HA464H.kawini HA592H.americanus GLR014H.tarsalis HA377H.tarsalis HA591H.sansoni HA368*H.sansoni GLR066H.waughi HA129H.bulbipes HA536H.americanus HA316*H.sansoni HA540H.americanus HA349H. sp. (YESOS) HA638H. sp. (CHMLA) GLR352H. sp. (CHMLA) HA639H.zapotecanus GLR339H.banksi HA104H. sp. (CTARA) HA554H.tarascanus HA643H.decorus GLR132H.carolinensis HA134H.venatoris HA119H. sp. (SPLEND) HA525 H.cockerelli HA108H.ocala HA612H.pugillis HA459H.cambridgei HA443*H.fallax HA648H.fallax HA647H.fallax HA288H.altanus HA325H.altanus HA124H.texanus HA494H.altanus GLR180H.altanus HA645H. sp. (YUCUN) HA635H. sp. (NAYAR) HA281H.icenoglei HA196H.oregonensis HA005H.jucundus HA395H.calcaratus calcaratus HA524H.notialis HA471H.cambridgei GLR351H.icenoglei GLR283H.oregonensis GLR149H.hallani GLR209H.hallani HA037H.pugillis GLR236H. sp. (SUNGL) GLR218H. sp. (ROBRT) GLR346H.calcaratus maddisoni GLR321H.jucundus HA285H.dossenus AS56/GLRH.clypeatus GLR227H.aztecanus HA218H.dossenus HA274H.dossenus HA531H.sp. (ROBRT) HA667H. sp. (ROBRT) HA611H. sp. (ROBRT) HA662H.aztecanus GLR347H.forticulus HA289H.virgulatus HA652H.zebraneus HA442H.divaricatus HA600H.divaricatus HA563H.velivolus HA659H.velivolus HA661*H.orbus HA101H.trimaculatus HA559H.viridipes HA470H. calcaratus maddisoni HA497H.moratus HA484H.mexicanus GLR353H. sp. (CNCTY) HA468H. sp. (BLNDI) GLR282H.pyrrithrix GLR304H.tlaxalanus HA674H.mexicanus HA670 H.ammophilus HA580 H.pyrrithrix HA578 H.pyrrithrix HA010*H.festus GLR088,GLR094H.brunneus HA543H.borealis GLR040H.captiosus GLR356H.virgulatus HA682H.virgulatus GLR205H.virgulatus HA167H.ballatoris HA383H.anepsius HA282*H.schlingeri HA160H.ammophilus HA556*H.anepsius HA557H.anepsius HA576 H.mexicanus HA496H.borealis HA130H.klauserii HA502*H.coecatus HA096H.cuspidatus HA517H.klauserii HA39062H. signatusH. sp. (ESTU)H. ophrysH. tarsalisH. sansoniH. americanus0.89-0.920.48-0.540.46-0.54Figure 9. BUCKy Primary Concordance tree for the americanus group, based on 679 genes longer than 1 kb. Node values are the average estimated primary CFs for each clade. Diagonal lines represent non-dominant relationships with CFs credibility intervals overlapping 0.1. * indicates relationships alsodetected as introgression using D-statistics/DFOIL. The thickness of the diagonal line indicates the size of the CF (range from 0.084 to  0.305) and lines reaching an ancestral branch indicate introgression with all descendants. All CFs and their credibility intervals are listed in Appendix 3. **63H. ophrysH. sp. (ROBRT) H. calcaratusH. jucundusH. aztecanusH. clypeatusH. sp.(SLCTY)H. mexicanusH. virgulatusH. sp. (BLNDI)H. pyrrithrixH. festusH. captiosusH. borealis0.55-0.590.52-0.580.47-0.510.46-0.510.36-0.410.31-0.350.25-0.290.24-0.290.21-0.250.36-0.41Figure 10. BUCKy Primary Concordance tree for the VCC clade, based on 517 geneslonger than 1 kb. Node values are the CF credibility intervals for each clade. Diagonallines represent non-dominant relationships with a CFs credibility intervals overlapping0.1.  The thickness of the diagonal line indicates the size of the CF (range from 0.196 to0.084) and lines reaching an ancestral branch indicate introgression with all descendants.* indicates relationships detected as introgression using D-statistics/DFOIL (species tested listed in Table 1). All CFs and their credibility intervals are listed in Appendix 4.*0.1-0.1464ABBA BABA(B)(C)*ABBA BABA H.tarsalis + H.sp.(ESTU) H.ophrys + H.sp.(ESTU)H.americanus + H.sp.(ESTU)H.sansoni + H.sp.(ESTU)H.sansoni +H.ophrysH.sansoni +H.tarsalisH.americanus +H.ophrysH.americanus +H.tarsalisH.sansoni +H.ophrysH.americanus +H.ophrysH.sansoni +H.tarsalisH.americanus +H.tarsalis*DFO DIL DFI DOLFigure 11. In (A), the total biallelic pattern counts for all DFOIL tests for introgression between americanus group species.  In (B) and (C), ABBA vs. BABA allele patterns counts for D statistic tests for introgression americanus group species. Detected introgression is summarized on the phylogeny: dashed line represents support from D-statistics, solid line from DFOIL. The phylogenetic positions of each set of species are listed in Table 1.1A-C. p <0.00074 (indicated by *, 95% significance adjusted with Bonferroni correction for 68 tests) are from a chi-square binomial test and indicate that the left and right terms are significantly different. Allele counts are listed in Appendix 5 and 6, Dstatistics, DFO, DIL, DFI, DOL, and their p-values are listed in Appendix 7 and 8.H .signatus (outgroup)H. sp. (ESTU)H .tarsalisH. ophrysH. sansoniH. americanus(A)CountCountCountCountCount0200400600800Count020040060080002004006008001000*(A)(B)65* ** ****(A)(B)(C)(D)(E)(F)H.aztecanus +H.mexicanusH.aztecanus +H.pyrrithrixH.clypeatus +H.pyrrithrixH.clypeatus +H.mexicanusH.aztecanus +H.mexicanusH.clypeatus +H.mexicanusH.aztecanus +H.pyrrithrixH.clypeatus +H.pyrrithrixH.sp.(BLNDI) +H.clypeatusH.sp.(BLNDI) +H.aztecanusH.pyrrithrix+H.clypeatusH.pyrrithrix+H.aztecanusH.sp.(BLNDI)+H.clypeatusH.pyrrithrix+H.clypeatusH.sp.(BLNDI) +H.aztecanusH.pyrrithrix+H.aztecanusH.sp.(SLCTY) +H.sp.(BLNDI)H.clypeatus + H.sp.(BLNDI)H.clypeatus +H.pyrrithrixH.sp.(SLCTY) +H.sp.(BLNDI)H.clypeatus + H.sp.(BLNDI)H.sp.(SLCTY) +H.pyrrithrixH.clypeatus +H.pyrrithrixH.sp.(SLCTY) +H.mexicanusH.sp.(SLCTY) +H.pyrrithrixH.clypeatus +H.mexicanusH.clypeatus +H.pyrrithrixH.sp.(SLCTY) +H.mexicanusH.clypeatus +H.mexicanusH.sp.(SLCTY) +H.pyrrithrixH.clypeatus +H.pyrrithrixH.sp.(SLCTY) +H.borealisH.sp.(SLCTY) +H.pyrrithrixH.clypeatus +H.borealisH.clypeatus +H.pyrrithrixH.sp.(SLCTY) +H.borealisH.clypeatus +H.borealisH.sp.(SLCTY)+H.pyrrithrixH.clypeatus +H.pyrrithrixH.aztecanus +H.borealisH.aztecanus +H.pyrrithrixH.clypeatus +H.borealisH.clypeatus +H.pyrrithrixH.aztecanus +H.borealisH.clypeatus +H.borealisH.aztecanus +H.pyrrithrixH.clypeatus +H.pyrrithrixDFO DIL DFI DOLH.ophrysH.aztecanusH.sp.(SLCTY)H.clypeatusH.mexicanusH.borealisH.pyrrithrixH.sp. (BLNDI)H.ophrysH.aztecanusH.sp.(SLCTY)H.clypeatusH.mexicanusH.borealisH.pyrrithrixH.sp. (BLNDI)H.ophrysH.aztecanusH.sp.(SLCTY)H.clypeatusH.mexicanusH.borealisH.pyrrithrixH. sp. (BLNDI)H.ophrysH.aztecanusH.sp.(SLCTY)H.clypeatusH.mexicanusH.borealisH.pyrrithrixH.sp. (BLNDI)H.ophrysH.aztecanusH.sp.(SLCTY)H.clypeatusH.mexicanusH.borealisH.pyrrithrixH.sp. (BLNDI)H.ophrysH.aztecanusH.sp.(SLCTY)H.clypeatusH.mexicanusH.borealisH.pyrrithrixH.sp. (BLNDI)Figure 12. Total biallelic pattern counts for all DFOIL tests for introgression between the coecatus and clypeatus groups. p <0.00074 (indicated by *, 95% significance adjusted with Bonferroni correction for 68 tests) are from a chi-squarebinomial test and indicate that the left and right terms are significantly different. The phylogenetic positions of each setof species are listed in Table 1.2A-F. Detected introgression is summarized on the phylogeny (green branches represent the species being tested).  Counts are listed in Appendix 5 and DFO, DIL, DFI, DOL, and their p-values are listed in Appendix 7.  Count02004006008001000Count02004006008001000Count02004006008001000Count02004006008001000Count02004006008001000Count0200400600800100066ABBA BABAABBA BABAABBA BABA***H. sp. (SLCTY) + H. sp. (ROBRT)H. jucundus + H. sp. (ROBRT) H. festus + H. sp. (ROBRT)H. jucundus + H. sp. (ROBRT)H. festus + H. sp. (ROBRT)H. sp. (SLCTY) + H. sp. (ROBRT)(A)(B)(C)Figure 13. ABBA vs. BABA allele patterns counts for Partitioned D statistic tests for introgression betweenH. sp. (ROBRT) and the VCC clade. * indicate ABBA vs. BABA counts are significantly different (95% significance adjusted with Bonferroni correction for 68 tests to p <0.00074) from chi-square binomial test. Detected introgression is summarized on the phylogeny (dark branches represent the species being tested, light gray branches are excluded from the test). The phylogenetic positions of each set of species are listed in Table 1.1a-c. Counts are listed in Appendix 6, and D statistic and p values are listed in Appendix 8.H.ophrysH.sp.(ROBRT)H.jucundus(viridipes group)H.sp.(SLCTY) (clypeatus group)H.festus (coecatus group)Count050015002500Count050015002500Count050015002500H.ophrysH.sp.(ROBRT)H.jucundus(viridipes group)H.sp.(SLCTY) (clypeatus group)H.festus (coecatus group)H.ophrysH.sp.(ROBRT)H.jucundus(viridipes group)H.sp.(SLCTY) (clypeatus group)H.festus (coecatus group)67* ***H.ophrys +H.zapotecanusH.ophrys+H.decorusH.sp.(ESTU) +H.decorusH.sp.(ESTU) +H.zapotecanusH.zapotecanus +H.ophrysH.zapotecanus +H.sp.(ESTU)H.decorus +H.ophrysH. +decorusH.sp.(ESTU)H. ophrys +H.cambridgeiH.sp.(ESTU) +H.cambridgeiH.canbridgei +H.ophrysH.cambridgei +H.sp.(ESTU)H.ophrys +H.oregonensisH.sp.(ESTU) +H.oregonensisH.oregonensis +H.ophrysH.oregonensis +H.sp.(ESTU)H.ophrys +H.jucundusH.sp.(EsTU) +H.jucundusH.jucundus +H.ophrysH.jucundus +H.sp.(ESTU)H.ophrys +H.festusH.sp.(ESTU) +H.festusH.festus +H.ophrysH.festus +H.sp.(ESTU)H.decorus +H.ophrysH. +decorusH.sp.(ESTU)H.decorus +H.ophrysH. +decorusH.sp.(ESTU)H.decorus +H.ophrysH. +decorusH.sp.(ESTU)H.decorus +H.ophrysH. +decorusH.sp.(ESTU)H.ophrys+H.decorusH.ophrys+H.decorusH.ophrys+H.decorusH.ophrys+H.decorusH. sp.(ESTU) +H.decorusH. sp.(ESTU) +H.decorusH. sp.(ESTU) +H.decorusH. sp.(ESTU) +H.decorus(A)(B)(C)(D)(E)H.signatusH. sp. (ESTU)H.ophrysH.decorusH.zapotecanusH.cambridgeiH.oregonensisH.jucundusH.festusH.signatusH. sp. (ESTU)H.ophrysH.decorusH.zapotecanusH.cambridgeiH.oregonensisH.jucundusH.festusH.signatusH. sp. (ESTU)H.ophrysH.decorusH.zapotecanusH.cambridgeiH.oregonensisH.jucundusH.festusH.signatusH. sp. (ESTU)H.ophrysH.decorusH.zapotecanusH.cambridgeiH.oregonensisH.jucundusH.festusH.signatusH. sp. (ESTU)H.ophrysH.decorusH.zapotecanusH.cambridgeiH.oregonensisH.jucundusH.festusFigure 14. The total biallelic pattern counts for all DFOIL tests between americanus group, H. decorus,and species H. zapotecanus, H. cambridgei, H. oregonensis, H. jucundus, H. festus. Detected introgression is represented on the phylogeny: dashed line represents support from D-statistics, solid line from DFOIL. Green branches represent the species being tested. The phylogenetic positions of each set of species are listed in Table 1.4. p <0.00074 (indicated by *, 95% significance adjusted with Bonferroni correction for 68 tests) are from a chi-square binomial test and indicate that the left and right terms are significantly different. The black rectangle indicates the comparisonssupported by D-statistic tests (tests excluding H. sp. (ESTU)). Allele counts are listed in Appendix 5 and 6, D-statistics, DFO, DIL, DFI DOL, and their p-values are listed in Appendix 7 and 8.Count050010001500200025003000Count050010001500200025003000050010001500200025003000050010001500200025003000Count050010001500200025003000DFO DIL DFI DOL* ** *** **68! *-!Bibliography  Abbott, R., D. Albach, S. Ansell, J. W. Arntzen, S. J. E. Baird, N. Bierne, J. Boughman, A. Brelsford, C. A. Buerkle, R. Buggs, et al. 2013. Hybridization and speciation. J. Evol. Biol. 26:229–246. Anderson, E. 1949. Introgressive hybridization. John Wiley and Sons, Inc. Andrews, S. 2010. FastQC: A quality control tool for high throughput sequence data.; Ané, C., B. Larget, D. Baum, S. D. Smith, and A. Rokas. 2007. Bayesian estimation of concordance among gene trees. Mol. Biol. Evol. 24:412–26. Ané, C. 2011. Reconstructing concordance trees and testing the coalescent model. In Knowles, L. L., & Kubatko, L. S. (Eds.). Estimating species trees: practical and theoretical aspects. John Wiley and Sons. Aronesty, E. 2011. ea-utils: Command-line tools for processing biological sequencing data.; Avise, J. 2001. Cytonuclear genetic signatures of hybridization phenomena: rationale, utility, and empirical examples from fishes and other aquatic animals. Rev. Fish Biol. Fish. 10:253–263. Baack, E. J., and L. H. Rieseberg. 2007. A genomic view of introgression and hybrid speciation. Curr. Opin. Genet. Dev. 17:513–8. Baum, D. A. 2007. Concordance trees , concordance factors , and the exploration of reticulate genealogy. 56:417–426. Blackburn, G. S., and W. P. Maddison. 2014. Stark sexual display divergence among jumping spider populations in the face of gene flow. Mol. Ecol. 23:5208–5223. Boeckmann B., A. Bairoch, R. Apweiler, M. Blatter, A. Estreicher, E. Gasteiger, M. J. Martin, K. Michoud, C. O'Donovan, I. Phan, et al. 2003. The SWISS-PROT protein knowledgebase and its supplement TrEMBL in 2003. Nucleic Acids Res. 31:365-370. Buchholz, S. 2007. A First Contribution to the Arachnofauna (Arachnida: Araneae) of the Nestos Delta (NE Greece). Acta Zool. Bulg. 59:241–252. Chung, Y., and Ané, C. 2011. Comparing two Bayesian methods for gene tree/species tree reconstruction: simulations with incomplete lineage sorting and horizontal gene transfer. Systematic Biology, 60:261-275. ! +.! Cui, R., M. Schumer, K. Kruesi, R. Walter, P. Andolfatto, and G. G. Rosenthal, 2013. Phylogenomics reveals extensive reticulate evolution in Xiphophorus fishes. Evolution, 67:2166-2179 Currat, M., M. Ruedi, R. J. Petit, and L. Excoffier. 2008. The hidden side of invasions: massive introgression by local genes. Evolution 62:1908–20. Degnan, J., and N. Rosenberg. 2009. Gene tree discordance, phylogenetic inference and the multispecies coalescent. Trends Ecol. Evol. 24:332–340. Donnelly, M. J., J. Pinto, R. Girod, N. J. Besansky, and T. Lehmann. 2004. Revisiting the role of introgression vs shared ancestral polymorphisms as key processes shaping genetic diversity in the recently separated sibling species of the Anopheles gambiae complex. Heredity. 92:61–68. Dunn, C. W., A. Hejnol, D. Q. Matus, K. Pang, W. E. Browne, S. A Smith, E. Seaver, G. W. Rouse, M. Obst, G. D. Edgecombe, et al. 2008. Broad phylogenomic sampling improves resolution of the animal tree of life. Nature 452:745–749. Durand, E., N. Patterson, D, Reich, and M. Slatkin. 2011. Testing for ancient admixture between closely related populations. Mol. Biol. Evol. 28:2239–2252. Eaton, D. A. R., and R. H. Ree. 2013. Inferring phylogeny and introgression using RADseq data: an example from flowering plants (Pedicularis: Orobanchaceae). Syst. Biol. 62:689–706. Eaton, D., A. Hipp, A. González‐Rodríguez, and J. Cavender‐Bares. 2015. Historical introgression among the American live oaks and the comparative nature of tests for introgression. Evolution. 69:2587–2601. Elias, D. O. 2006. Female preference for complex/novel signals in a spider. Behav. Ecol. 17:765–771. Elias, D. O., W. P. Maddison, C. Peckmezian, M. B. Girard, and A. C. Mason. 2012. Orchestrating the score: complex multimodal courtship in the Habronattus coecatus group of Habronattus jumping spiders (Araneae: Salticidae). Biol. J. Linn. Soc. 105:522–547. Elias, D., A. Mason, and W. Maddison. 2003. Seismic signals in a courting male jumping spider (Araneae: Salticidae). J. Exp. Biol. 206:4029–4039. Eriksson, A., and A. Manica. 2012. Effect of ancient population structure on the degree of polymorphism shared between modern human populations and ancient hominins. Proc. Natl. Acad. Sci. U. S. A. 109:13956–13960. ! +%!Fiser, C., and G. N. Azarkina. 2005. A Contribution to the Knowledge of the Jumping Spiders (Salticidae: Araneae) of the Republic of Macedonia. Acta Zool. Bulg. 57:299–304. Fu, L., B. Niu, Z. Zhu, S. Wu, and W. Li. 2012. CD-HIT: accelerated for clustering the next-generation sequencing data. Bioinformatics 28:3150–3152. Funk, D. J., and K. E. Omland. 2003. Species-Level Paraphyly and Polyphyly: Frequency, Causes, and Consequences, with Insights from Animal Mitochondrial DNA. Annu. Rev. Ecol. Evol. Syst. 34:397–423. Good, J., D. Vanderpool, S. Keeble, and K. Bi. 2015. Negligible nuclear introgression despite complete mitochondrial capture between two species of chipmunks. Evolution. 69:1961–1972. Grant, P. R., B. R. Grant, and K. Petren. 2005. Hybridization in the recent past. Am. Nat. 166:56–67. Gregory, T. R., and D. P. Shorthouse. 2003. Genome Sizes of Spiders. J. Hered. 94:285–290. Griswold, C. 1987. A revision of the jumping spider genus Habronattus FOP-Cambridge (Araneae; Salticidae), with phenetic and cladistic analyses. Univ. Calif. Publ. Entomol. 107:1–344. Haas, B. J., A. Papanicolaou, M. Yassour, M. Grabherr, P. D. Blood, J. Bowden, M. B. Couger, D. Eccles, B. Li, M. Lieber, et al. 2013. De novo transcript sequence reconstruction from RNA-seq using the Trinity platform for reference generation and analysis. Nat. Protoc. 8:1494–512. Hebets, E., and W. P. Maddison. 2005. Xenophilic mating preferences among populations of the jumping spider Habronattus pugillis Griswold. Behav. Ecol. 16:981–988. Hedin, M., and M. Lowder. 2009. Phylogeography of the Habronattus amicus species complex (Araneae: Salticidae) of western North America, with evidence for localized asymmetrical mitochondrial. Zootaxa 2307:39–60. Hedin, M. C., and W. P. Maddison. 2000. Phylogenetic Utility and Evidence for Multiple Copies of Elongation Factor-1! in the Spider Genus Habronattus (Araneae#: Salticidae). 18:1512–1521. Hedrick, P. 2013. Adaptive introgression in animals: examples and comparison to new mutation and standing variation as sources of adaptive variation. Mol. Ecol. 22:4606–4618. Heliconius Genome Consortium. 2012. Butterfly genome reveals promiscuous exchange of mimicry adaptations among species. Nature. 487:94-98.  ! +&!Holland, B., and W. Rice. 1998. Perspective: chase-away sexual selection: antagonistic seduction versus resistance. Evolution. 52:1–7. Huerta-Sánchez H., X. Jin, A. and Z. Bianba, B.M. Peter, N. Vinckenbosch, Y. Liang, X. Yi, M. He, M. Somel, P. Ni, et al. 2014. Altitude adaptation in Tibetans caused by introgression of Denisovan-like DNA. Nature. 512:194-197. Joly, S., P. A. McLenachan, and P. J. Lockhart. 2009. A statistical approach for distinguishing hybridization and incomplete lineage sorting. Am. Nat. 174:54–70. Katoh, K., and D. Standley. 2013. MAFFT multiple sequence alignment software version 7: improvements in performance and usability. Mol. Biol. Evol. 30:3059–3066. Kronforst, M., L. Young, L. Blume, and L. Gilbert. 2006. Multilocus analyses of admixture and introgression among hybridizing Heliconius butterflies. Evolution. 60:1254–1268. Kubatko, L. S., and J. H. Degnan. 2007. Inconsistency of phylogenetic estimates from concatenated data under coalescence. Syst. Biol. 56:17–24. Lanfear, R., B. Calcott, S. Ho, and S. Guindon. 2012. PartitionFinder: combined selection of partitioning schemes and substitution models for phylogenetic analyses. Mol. Biol. Evol. 29:1695–1701. Larget, B., S. Kotha, C. Dewey, and C. Ané. 2010. BUCKy: gene tree/species tree reconciliation with Bayesian concordance analysis. Bioinformatics 26:2910–2911. Lemmon, A. R., and E. M. Lemmon. 2012. High-throughput identification of informative nuclear loci for shallow-scale phylogenetics and phylogeography. Syst. Biol. 61:745–61. Lepais, O., R. J. Petit, E. Guichoux, J. E. Lavabre, F. Alberto, A. Kremer, and S. Gerber. 2009. Species relative abundance and direction of introgression in oaks. Mol. Ecol. 18:2228–42. Levin, D. A., J. Francisco-Ortega, and R. K. Jansen. 1996. Hybridization and the Extinction of Rare Plant Species. Conserv. Biol. 10:10–16. Li, B., and C. Dewey. 2011. RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinformatics 12:323. Logunov, D. 1999. A review of the genus Pellenes in the fauna of Central Asia and the Caucasus (Araneae, Salticidae). J. Nat. Hist. 33:89–148. Maddison, W. P. 1982. XXXY sex chromosomes in males of the jumping spider genus Pellenes (Araneae: Salticidae). Chromosoma 85:23–37. Maddison, W. P. 1997. Gene trees in species trees. Systematic biology, 46:523-536.  ! +'!Maddison, W. P., and G. S. Blackburn. 2015. Insights to the mating strategies of Habronattus americanus jumping spiders from natural behaviour and staged interactions in the wild. Behaviour 152:1169–1186.  Maddison, W. P., and M. Hedin. 2003. Phylogeny of Habronattus jumping spiders (Araneae#: Salticidae), with consideration of genitalic and courtship evolution. Syst. Entomol. 28:1–22. Maddison, W. P., and G. Leduc-Robert. 2013. Multiple origins of sex chromosome fusions correlated with chiasma localization in Habronattus jumping spiders (Araneae: Salticidae). Evolution 67:2258–72. Maddison, W. P., and D. Maddison. 2015. Mesquite: a modular system for evolutionary analysis. Version 2.75. 2011. Maddison, W. P., and M. McMahon. 2000. Divergence and Reticulation among Montane Populations of a Jumping Spider (Habronattus pugillis Griswold). Syst. Biol. 49:400–421. Mallet, J. 2005. Hybridization as an invasion of the genome. Trends Ecol. Evol. 20:229–37. Martin, S., J. Davey, and C. Jiggins. 2015. Evaluating the use of ABBA-BABA statistics to locate introgressed loci. Mol. Biol. Evol. 32:244–257. Masta, S. E., and J. L. Boore. 2004. The complete mitochondrial genome sequence of the spider Habronattus oregonensis reveals rearranged and extremely truncated tRNAs. Mol. Biol. Evol. 21:893–902. Masta, S. E., and W. P. Maddison. 2002. Sexual selection driving diversification in jumping spiders. Proc. Natl. Acad. Sci. U.S.A. 99:4442–4447. Miller, J. R., S. Koren, and G. Sutton. 2010. Assembly algorithms for next-generation sequencing data. Genomics 95:315–327. Norris, L.C., B.J. Main, Y. Lee, T.C. Collier, A. Fofana, A.J. Cornel, and G.C. Lanzaro. 2015. Adaptive introgression in an African malaria mosquito coincident with the increased usage of insecticide-treated bed nets. Proc. Natl. Acad. Sci. USA. 112:815-820. Pamilo, P., and M., Nei. 1988. Relationships between gene trees and species trees. Mol. Biol. Evol. 5:568-583. Panhuis, T. M., R. Butlin, M. Zuk, and T. Tregenza. 2001. Sexual selection and speciation. Trends Ecol. Evol. 16:364–371. Pease, J., and M. Hahn. 2015. Detection and Polarization of Introgression in a Five-taxon Phylogeny. Syst. Biol. 64:651–662. ! +(!Peckham, G., and E. Peckham. 1889. Observations on sexual selection in spiders of the family Attidae. Ornamentation. Occas. Pap. Nat. Hist. Soc. Wisconsin. 1:117-151.  Peckham, G., and E. Peckham. 1890. Additional Observations on Sexual Selection in Spiders of the Family Attidae: With Some Remarks on Mr. Wallace’s Theory of Sexual Ornamentation. Occas. Pap. Nat. Hist. Soc. Wisconsin. 1:3-60.  Rieseberg, L., O. Raymond, D. Rosenthal, and Z. Lai. 2003. Major ecological transitions in wild sunflowers facilitated by hybridization. Science. 301:1211–1216. Ronquist, F., and J. Huelsenbeck. 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19:1572–1574. Ronquist, F., J. Huelsenbeck, and P. van der Mark. 2005. MrBayes 3.1 Manual. Sanggaard, K.W., J.S Bechsgaard, X. Fang, J. Duan, T.F. Dyrlund, V. Gupta, X. Jiang, L. Cheng, D. Fan,  Y. Feng, et al. 2014. Spider genomes provide insight into composition and evolution of venom and silk. Nature communications 5:3765. Sankararaman, S., N. Patterson, H. Li, S. Pääbo, and D. Reich. 2012. The date of interbreeding between Neandertals and modern humans. PLoS Genet. 8:e1002947.  Scheidemantel, D. D. 1997. Behavioral and natural history studies of the jumping spider Habronattus oregonensis and inquiry based secondary laboratory lesson development stemming from university research.  PhD. Diss., The University of Arizona, Arizona. Schluter, D., and T. Price. 1993. Honesty, perception and population divergence in sexually selected traits. Proc. R. Soc. London B Biol. Sci. 253:117–122. Schmieder, R., and R. Edwards. 2011. Quality control and preprocessing of metagenomic datasets. Bioinformatics 27:863–4. Seehausen, O. 2004. Hybridization and adaptive radiation. Trends Ecol. Evol. 19:198–207. Seehausen, O. 2013. Conditions when hybridization might predispose populations for adaptive radiation. J. Evol. Biol. 26:279–81. Slatkin, M. 1985. Gene flow in natural populations. Annu. Rev. Ecol. Syst. 393–430. Smith, S., M. Moore, and J. Brown. 2015. Analysis of phylogenomic datasets reveals conflict, concordance, and gene duplications with examples from animals and plants. BMC Evol. Biol. 15:150. Stamatakis, A. 2006. RAxML-VI-HPC: maximum likelihood-based phylogenetic analyses with thousands of taxa and mixed models. Bioinformatics 22:2688–2690. ! +)!Stelkens, R., and O. Seehausen. 2009. Genetic distance between species predicts novel trait expression in their hybrids. Evolution 63:884–97. Streicher, J. W., T. J. Devitt, C. S. Goldberg, J. H. Malone, H. Blackmon, and M. K. Fujita. 2014. Diversification and asymmetrical gene flow across time and space: lineage sorting and hybridization in polytypic barking frogs. Mol. Ecol. 23:3273–3291. Toews, D. P. L., and A. Brelsford. 2012. The biogeography of mitochondrial and nuclear discordance in animals. Mol. Ecol. 21:3907–30. Turelli, M., N. H. Barton, and J. A. Coyne. 2001. Theory and speciation. Trends Ecol. Evol. 16:330–343. Twyford, A. D., and R. A. Ennos. 2012. Next-generation hybridization and introgression. Heredity (Edinb). 108:179–189. Wirtz, P. 1999. Mother species-father species: unidirectional hybridization in animals with female choice. Anim. Behav. 58:1–12. Yu, Y., C. Than, J. H. Degnan, and L. Nakhleh. 2011. Coalescent histories on phylogenetic networks and detection of hybridization despite incomplete lineage sorting. Syst. Biol. 60:138–49. Zink, R. M., & Barrowclough, G. F. 2008. Mitochondrial DNA under siege in avian phylogeography. Molecular ecology, 17:2107-2121. Zurek, D. B., T. W. Cronin, L. A. Taylor, K. Byrne, M. L. G. Sullivan, and N. I. Morehouse. 2015. Spectral filtering enables trichromatic vision in colorful jumping spiders. Curr. Biol. 25:R403–R404.  !"AppendicesSpecies Voucher # Sex Locality Latitude , LongitudeSpecies karyotype (male sex chromosomes only)H. festus GLR094 M Hayne's Lease Ecological Reserve, near Osoyoos, BRITISH COLUMBIA, CANADA 49.0813, -119.5181 XXOH. festus GLR088 F Hayne's Lease Ecological Reserve, near Osoyoos, BRITISH COLUMBIA, CANADA 49.0813, -119.5181 XXOH. virgulatus GLR205 M Mt. Hopkins Road, Santa Cruz Co., ARIZONA, U.S.A. 31.689, -110.975 XXOH. borealis GLR040 M Hamilton Beach Strip, Burlington, ONTARIO, CANADA 43.33, -79.8 XXXYH. mexicanus GLR353 M  Rancho Primavera, El Tuito, JALISCO, MEXICO 20.341, -105.350 XXYH. pyrrithrix GLR304 M Levee Road, Sunrise Point Park, Yuma County., ARIZONA, U.S.A. 32.731, -114.612 XXOH. captiosus GLR356 F shore of Little Smoky River at HWY 49, S of Guy, ALBERTA, CANADA 55.4505, -117.1440 XXXYcoecatus groupAppendix 1 Specimens from which transcriptomes were obtained, their localities, and their sex chromsomes (if they are known. Only mitochondrial data was gathered for H.paratus.!!Species Voucher # Sex Locality Latitude , LongitudeSpecies karyotype (male sex chromosomes only)H. sp. (BLNDI) GLR282 M Estero Morúa, Puerto Peñasco, SONORA, MEXICO 31.293, -113.452 uncertainH. ophrys GLR023 M Iona Beach, Richmond, BRITISH COLUMBIA, CANADA 49.221, -123.214 XXOH. ophrys GLR015 F Iona Beach, Richmond, BRITISH COLUMBIA, CANADA 49.221, -123.214 XXOH. americanus GLR014 M Iona Beach, Richmond, BRITISH COLUMBIA, CANADA 49.221, -123.214 XXOH. tarsalis GLR297 M Levee Road, Sunrise Point Park,Yuma Co., ARIZONA, U.S.A. 32.731, -114.612 XXOH. sansoni GLR066 M Borgata Lodge, Kelowna 49.954, -119.398 XXOH. sp. (ESTU) GLR287 F Estero Cerro Prieto, SONORA, MEXICO 31.418, -113.626 uncertainH. clypeatus GLR227 M Mt. Hopkins Road, Santa Cruz Co., ARIZONA, U.S.A. 31.686, -110.975 uncertainH. aztecanus GLR347 M  Bocanegra Beach, Puerto Vallarta, JALISCO, MEXICO 20.670, -105.274 XXOH. sp. (SLCTY) AS56 F Silver City, NEW MEXICO, U.S.A. n/a uncertainamericanus groupclypeatus group!"Species Voucher # Sex Locality Latitude , LongitudeSpecies karyotype (male sex chromosomes only)H. signatus GLR600 n/a Ocotillo, CALIFORNIA, U.S.A. 32.7421, -115.9949 XXOH. ustulatus GLR601 n/a Cleveland National Forest, CALIFORNIA, U.S.A. 32.7302, -116.4607 uncertainH. jucundus GLR320 M Bolan lake, Joseph County, OREGON, U.S.A. 42.024, -123.461 XXOH. calcaratus maddisoni GLR321 MWest Road near HWY 11, Hailey Byw W. , Haileybury, ONTARIO, CANADA 47.45, -79.708 XXXYH. oregonensis GLR149 M Tantalus Lookout Road, Squamish, BRITISH COLUMBIA, CANADA 49.8465, -123.1452 XXOH. icenoglei GLR283 M Puerto Peñasco, Estero Morúa, E of Playa Encanto, SONORA, MEXICO 31.273, -113.361 XXOH. pugillis GLR236 M Mt. Hopkins Road, Santa Cruz Co., ARIZONA, U.S.A. 31.689, -110.975 XXOH. sp. (SUNGL) GLR218 F Amateur Astronomy Vista, Mt. Hopkins Rd, Santa Cruz Co., ARIZONA, U.S.A. 31.6759, -110.9289 uncertainH. zapotecanus GLR339 M  La Viuda restaurant, Chamela, JALISCO, MEXICO 19.5316, -105.0707 XXYH. sp. (CHMLA) GLR352 M Estación de Biología Chamela, 400-650 m on Calandria Trail, JALISCO, MEXICO 19.5038, -105.0334 uncertainamicus groupviridipes grouporegonensis grouppugillis groupbanksi group!"Species Voucher # Sex Locality Latitude , LongitudeSpecies karyotype (male sex chromosomes only)Pellenes cf. levii GLR106 M Mt. Baldy Road, BRITISH COLUMBIA, CANADA 49.1135, -119.2103 XXOEvarcha prozjinski GLR135 M Silvermere Lake, Mission, BRITISH COLUMBIA, CANADA 49.166, -122.409 XXOH. conjunctus GLR234 M Madera Canyon, near Proctor Road, Pima Co., ARIZONA, U.S.A. 31.7417, -110.8847 XXOH. altanus GLR180 M East of Smoky Lake, ALBERTA, CANADA 54.112, -112.198 XXYH. sp. (ROBRT) GLR346 M Estación de Biología Chamela, Chachalaca Trail, JALISCO, MEXICO 19.496, -105.042 XXOH. decorus GLR132 M Silvermere Lake, Mission, BRITISH COLUMBIA, CANADA 49.166, -122.409 XXXYH. geronimoi GLR267 M Miller Canyon,Huachuca Mountains., Cochise Co., ARIZONA, U.S.A. 31.416, -110.276 XXYH. cambridgei GLR351 M Bocanegra beach, Puerto Vallarta, JALISCO, MEXICO 20.670, -105.274 XXOH. hirsutus GLR080 M Mt. Kobau Road, BRITISH COLUMBIA, CANADA 49.095, -119.610 XXOaltanus groupsp. (ROBRT) groupdecorus groupdorotheae groupfallax grouptranquillus groupoutgroupsagilis group!"Species Voucher # Sex Locality Latitude , LongitudeSpecies karyotype (male sex chromosomes only)H. hallani GLR209 M near Arivaca, Pima Co., ARIZONA, U.S.A. 31.668, -111.245 XXYH. paratus GLR363 M Starfish Beach, Punta Galeta,  Isla Colon, BOCAS DEL TORO, PANAMA 9.40376, -79.8635 XXOparatus grouphallani group!"Species # of raw reads # of reads after trimming % of reads left after trimming% of reads assembled to nuclear% of reads assembled to mtDNA%GC# contigs 200-500bp# contigs 500-1000bp# contigs                > 1000bpAverage nuclear coverageAverage mtDNA coverageTotal contigsH. festus  156,826,220 123,414,758 78.7 67.0 15.9 33 20,398 4,376 4,072 174 n/a 28,846H. pyrrithrix 30,979,896 23,954,560 77.3 65.9 15.4 36 6,037 2,899 1,892 93 23,939    10,828H. virgulatus 13,823,012 10,967,254 79.3 60.1 18.8 33 4,349 2,155 1,167 60 13,256    7,671H. borealis 31,803,862 29,404,408 92.5 64.0 16.6 38 8,494 3,304 2,194 89 31,830    13,992H. mexicanus 11,390,510 10,318,886 90.6 59.5 21.1 36 5,179 1,995 1,018 50 14,828    8,192H. sp. (BLNDI) 11,052,058 9,278,850 84.0 59.0 19.6 35 8,439 2,132 1,095 57 11,904    11,666H. captiosus 11,515,080 10,026,878 87.1 62.8 16.8 35 5,679 2,519 1,503 46 11,492    9,701H. ophrys 169,057,044 123,988,153 77.6 73.1 15.9 33 37,287 8,804 5,052 111 n/a 51,143H. americanus 10,974,358 10,107,232 92.1 70.3 11.9 37 5,656 2,348 1,000 53 7,721      9,004H. tarsalis 10,563,210 9,497,728 89.9 63.6 11.5 37 4,562 2,117 992 52 7,095      7,671H. sansoni 10,963,598 9,383,970 85.6 65.2 15.1 33 5,412 2,392 1,207 44 9,010      9,011H. sp. (ESTU) 22,859,706 18,199,030 79.6 66.1 14.0 35 7,357 3,170 1,933 63 16,711    12,460Appendix 2 Summary of RNA-sequencing and transcriptome assemblies for all species (except DNA-sequencing for H.paratus): total number of reads sequenced and reads remaining after trimming and filtering, % of reads mapped to mitochondrial and nuclear references, number of transcripts, and average coverage for all de novo and reference-based transcriptome assemblies.coecatus groupamericanus group!"Species # of raw reads # of reads after trimming % of reads left after trimming% of reads assembled to nuclear% of reads assembled to mtDNA%GC# contigs 200-500bp# contigs 500-1000bp# contigs                > 1000bpAverage nuclear coverageAverage mtDNA coverageTotal contigsH. clypeatus 19,732,586 18,838,386 95.5 64.8 15.4 37 5,984 2,881 1,689 78 19,106    10,554H. aztecanus 6,974,386 6,223,514 89.2 92.5 0.2 35 3,464 1,526 777 49 7,293      5,767H. sp. (SLCTY) 15,954,538 15,953,730 100.0 70.0 8.7 37 9,935 3,378 2,251 60 13,666    15,564H. signatus 150,582,550 126,114,806 83.8 64.8 10.0 37 11,329 3,955 3,052 87 n/a 18,336H. ustulatus 137,650,334 120,525,796 87.6 69.5 7.0 37 11,039 4,086 3,211 82 n/a 18,336H. jucundus 14,017,246 12,015,998 85.7 65.1 16.0 37 5,841 2,227 1,108 50 12,504    9,176H. calcaratus maddisoni 9,522,580 8,691,730 91.3 56.0 25.3 35 3,583 1,770 768 54 14,496    6,121H. icenoglei 16,345,750 13,819,638 84.5 71.7 9.3 37 3,623 2,438 1,132 90 7,720      7,193H. oregonensis 13,721,754 10,598,794 77.2 62.9 16.0 37 5,197 2,288 1,215 54 11,821    8,700oregonensis groupclypeatus groupamicus     group  viridipes    group!"Species # of raw reads # of reads after trimming % of reads left after trimming% of reads assembled to nuclear% of reads assembled to mtDNA%GC# contigs 200-500bp# contigs 500-1000bp# contigs                > 1000bpAverage nuclear coverageAverage mtDNA coverageTotal contigsH. sp. (SUNGL) 23,875,266 21,128,070 88.5 71.2 8.9 38 6,185 2,607 1,698 89 12,169    10,490H. pugillis 14,485,868 13,625,478 94.1 68.4 16.0 37 4,890 2,363 1,241 74 13,985    8,494H. sp. (CHMLA) 10,253,210 8,896,604 86.8 70.8 13.8 36 4,887 2,183 1,220 48 8,382      8,290H. zapotecanus 23,804,988 22,550,394 94.7 50.0 10.3 40 7,446 2,912 1,581 65 15,759    11,939Pellenes cf. levii 22,692,970 18,291,042 80.6 55.7 15.8 36 5,464 2,403 1,559 65 18,840    9,426Evarcha prozjinski 24,090,418 21,543,358 89.4 38.5 21.5 37 3,316 1,503 597 60 30,009    5,416H. conjunctus 29,102,382 25,453,304 87.5 63.7 15.3 35 6,045 2,629 1,680 94 25,324    10,354H. altanus 8,321,456 7,546,466 90.7 61.0 13.7 36 5,225 1,050 1,035 45 6,632      7,310pugillis      groupbanksi       groupoutgroupsagilis       groupaltanus     group!"Species # of raw reads # of reads after trimming % of reads left after trimming% of reads assembled to nuclear% of reads assembled to mtDNA%GC# contigs 200-500bp# contigs 500-1000bp# contigs                > 1000bpAverage nuclear coverageAverage mtDNA coverageTotal contigsH. sp. (ROBRT) 6,051,598 5,308,674 87.7 63.1 17.1 35 2,255 745 746 43 6,188      3,746H. decorus 14,842,184 11,274,830 76.0 56.7 20.3 34 5,059 2,501 1,187 48 15,069    8,747H. geronimoi 18,321,104 15,484,070 84.5 67.8 13.7 36 5,027 2,545 1,261 79 13,571    8,833H. cambridgei 13,076,744 12,432,290 95.1 64.9 16.9 36 4,837 2,596 1,366 65 14,220    8,799H. hirsutus 8,551,110 7,444,104 87.1 58.1 14.9 36 3,834 1,610 794 42 7,517      6,238H. hallani 16,041,252 13,578,488 84.6 63.2 16.5 36 5,679 2,589 1,501 57 14,366    9,769H.sp. (ROBRT) groupdecorus       groupdorotheae     groupfallax       grouptranquillus  grouphallani      group!"Species # of raw reads # of reads after trimming % of reads left after trimming% of reads assembled to nuclear% of reads assembled to mtDNA%GC# contigs 200-500bp# contigs 500-1000bp# contigs                > 1000bpAverage nuclear coverageAverage mtDNA coverageTotal contigsH. paratus 24,000,000 23,926,440 99.7 31.6 0.0004 30 8,437 561 111 64 60           9,109paratus     group!"Appendix 3 Significant Concordance Factors (CFs > 0.05) and their credibility intervals for the americanus clade.Bipartition Supported clade Sample-wide CF 95% credibilityGenome-wide CF 95% credibilityIn primary concordance tree?SD{1,2,3,4|5,6} H. sansoni + H.  americanus 0.906 0.887-0.923 0.9 0.869-0.928 Y 0{1,2|3,4,5,6} H. ophrys + H. tarsalis + H. sansoni + H. americanus0.508 0.476-0.541 0.505 0.456-0.555 Y 0{1,2,5,6|3,4} H. ophrys + H. tarsalis 0.501 0.464-0.536 0.498 0.446-0.550 Y 0{1,2,4|3,5,6}  H. ophrys + H. sansoni/ H. americanus 0.305 0.272-0.339 0.303 0.257-0.352 N 0{1,5,6|2,3,4} H. sp. (ESTU) + H. ophrys/H. tarsalis 0.19 0.165-0.216 0.19 0.152-0.230 N 0{1,3,5,6|2,4} H. sp. (ESTU) + H. tarsalis 0.136 0.112-0.162 0.136 0.102-0.174 N 0{1,2,3|4,5,6} H. tarsalis + H. sansoni/H. americanus 0.136 0.112-0.162 0.136 0.102-0.174 N 0{1,3,4|2,5,6} H. sp. (ESTU) + H. sansoni/H. americanus 0.087 0.068-0.108 0.087 0.060-0.118 N 0{1,4|2,3,5,6}H. sp. (ESTU) + H. ophrys + H. sansoni + H. americanus0.055 0.038-0.074 0.056 0.034-0.082 N 0Bipartitions number codes:  1= H. signatus, 2= H. sp. (ESTU),  3= H. ophrys,  4= H. tarsalis,  5= H. sansoni,  6= H. americanus!"Appendix 4 Significant Concordance Factors (CFs > 0.05) and their credibility intervals for the VCC clade.Bipartition Supported clade Sample-wide CF95% credibilityGenome-wide CF95% credibilityIn primary concordance tree?SD{1,2,3,4,8,9,10,11,12,13,14|5,6,7} clypeatus group 0.571 0.553-0.588 0.571 0.525-0.617 Y 0.001{1,2,3,4,5,8,9,10,11,12,13,14|6,7} H. clypeatus + H. sp.  (SLCTY) 0.552 0.518-0.582 0.552 0.498-0.605 Y 0{1,4,5,6,7,8,9,10,11,12,13,14|2,3} H. calcaratus +H. jucundus 0.494 0.472-0.515 0.494 0.446-0.542 Y 0.002{1,2,3,4,5,6,7,8,9,10,11,12|13,14} H. captiosus + H. borealis 0.487 0.460-0.513 0.487 0.437-0.538 Y 0.003{1,2,3,4,5,6,7|8,9,10,11,12,13,14} coecatus group 0.389 0.360-0.408 0.389 0.339-0.437 Y 0.005{1,2,3,4,5,6,7,8,11,12,13,14|9,10} H. sp. (BLNDI) + H. pyrrithrix 0.385 0.360-0.408 0.385 0.337-0.434 Y 0.005{1,2,3,4,5,6,7,8,9,10,11|12,13,14} H. festus + H. borealis + H. captiosus 0.331 0.311-0.350 0.331 0.287-0.377 Y 0.004{1,4|2,3,5,6,7,8,9,10,11,12,13,14} VCC clade, excluding H.sp. (ROBRT) 0.267 0.246-0.292 0.267 0.223-0.314 Y 0.001{1,2,3,4,5,6,7,8|9,10,11,12,13,14} coecatus group except H. mexicanus 0.267 0.244-0.288 0.267 0.224-0.312 Y 0{1,2,3,4|5,6,7,8,9,10,11,12,13,14}  clypeatus + coecatus group 0.23 0.213-0.248 0.23 0.191-0.271 Y 0.005{1,2,3,4,5,6,7,8,12,13,14|9,10,11}H. sp. (BLNDI) + H.pyrrithrix + H.virgulatus0.117 0.101-0.137 0.117 0.086-0.153 Y 0{1,2,3|4,5,6,7,8,9,10,11,12,13,14} H. sp. (ROBRT) +  clypeatus/coecatus group0.196 0.174-0.222 0.196 0.155-0.240 N 0.011{1,2,3,8,9,10,11,12,13,14|4,5,6,7} H. sp. (ROBRT) + clypeatus group 0.188 0.170-0.205 0.188 0.151-0.227 N 0.006{1,2,3,4,7,8,9,10,11,12,13,14|5,6} H. aztecanus + H. sp. (SLCTY) 0.175 0.149-0.201 0.175 0.134-0.218 N 0.002Bipartitions number codes: 1=H. ophrys, 2=H. calcaratus, 3=H. jucundus, 4= H. sp. (ROBRT), 5= H. aztecanus, 6= H. clypeatus, 7= H. dossenus, 8= H. mexicanus, 9=H. sp. (BLNDI), 10= H. pyrrithrix, 11= H. virgulatus, 12= H. festus, 13= H. captiosus, 14= H. borealis!!Bipartition Supported clade Sample-wide CF95% credibilityGenome-wide CF95% credibilityIn primary concordance tree?SD{1,2,3,4,5,6,7,8,9,10,11,14|12,13} H. festus + H. captiosus 0.153 0.130-0.172 0.153 0.116-0.191 N 0.001{1,2,3,4,6,8,9,10,11,12,13,14|5,7} H. aztecanus + H. clypeatus 0.115 0.091-0.141 0.115 0.081-0.155 N 0.001{1,2,3,4,5,6,7,8,9,10|11,12,13,14}  H. virgulatus + H. festus + H. captiosus + H. borealis0.113 0.093-0.130 0.113 0.082-0.147 N 0.001{1,2,3,4,5,6,7,11,12,13,14|8,9,10} H. mexicanus + H .sp. (BLNDI) + H. pyrrithrix 0.104 0.087-0.120 0.104 0.075-0.136 N 0{1,2,3,4,5,6,7,8,10|9,11,12,13,14} H. sp. (BLNDI) + H. virgulatus + H. festus + H. borealis + H. captiosus0.101 0.087-0.116 0.101 0.073-0.132 N 0.005{1,4,5,6,7|2,3,8,9,10,11,12,13,14} viridipes + coecatus groups 0.103 0.087-0.122 0.103 0.073-0.137 N 0.004{1,2,3,4,5,6,7,8,10,12,13,14|9,11} H. sp. (BLNDI) + H. virgulatus 0.099 0.083-0.116 0.099 0.071-0.132 N 0{1,2,3,4,5,6,7,8,9,10,13,14|11,12} H. festus + H. virgulatus 0.093 0.077-0.110 0.093 0.065-0.125 N 0.001{1,2,3,4,5,6,7,8,11|9,10,12,13,14} H. sp. (BLNDI) + H. pyrrithrix + H. festus + H. borealis +H. captiosus0.089 0.070-0.110 0.088 0.059-0.124 N 0.008{1,2,3,4,5,6,7,8,10,11|9,12,13,14}H. sp. (BLNDI) + H. festus + H. borealis + H .captiosus0.086 0.066-0.104 0.086 0.055-0.119 N 0.009{1,2,3,4,5,6,7,8,9,12,13,14|10,11} H. pyrrithrix + H. virgulatus 0.084 0.064-0.104 0.084 0.054-0.117 N 0.007{1,2,3,4,5,6,7,8,10,11,13,14|9,12} H. sp. (BLNDI) +H. festus 0.083 0.066-0.099 0.083 0.056-0.113 N 0.001{1,2,3,4,5,6,7,8,9,10,11,13|12,14} H. festus + H. borealis 0.081 0.066-0.099 0.081 0.054-0.112 N 0.002{1,2,3,4,5,6,7,9,11,12,13,14|8,10} H. mexicanus + H. pyrrithrix 0.071 0.056-0.085 0.071 0.046-0.099 N 0.002!"Bipartition Supported clade Sample-wide CF95% credibilityGenome-wide CF95% credibilityIn primary concordance tree?SD{1,3,4|2,5,6,7,8,9,10,11,12,13,14} H. calcaratus + clypeatus group and coecatus group0.068 0.052-0.083 0.068 0.043-0.096 N 0.001{1,2,3,4,5,6,7,8,13,14|9,10,11,12}H. sp. (BLNDI) + H. pyrrithrix + H. virgulatus + H. festus 0.061 0.048-0.074 0.061 0.039-0.086 N 0.003{1,3|2,4,5,6,7,8,9,10,11,12,13,14} H. calcaratus + H. sp. (ROBRT) + clypeatus group + coecatus group0.06 0.048-0.075 0.06 0.038-0.087 N 0.005{1,2,4|3,5,6,7,8,9,10,11,12,13,14} H. jucundus + clypeatus group + coecatus group 0.059 0.044-0.074 0.059 0.037-0.086 N 0{1,2|3,4,5,6,7,8,9,10,11,12,13,14} H. jucundus + H. sp.  (ROBRT) + clypeatus group + coecatus group0.058 0.044-0.074 0.058 0.035-0.086 N 0.006{1,2,3,6,7,8,9,10,11,12,13,14|4,5} H. sp. (ROBRT) + H. aztecanus 0.056 0.043-0.070 0.056 0.035-0.082 N 0{1,2,3,4,5,6,7,8,11,13,14|9,10,12} H. sp. (BLNDI) + H. pyrrithrix + H. festus 0.054 0.041-0.068 0.054 0.032-0.080 N 0{1,2,3,4,8|5,6,7,9,10,11,12,13,14}clypeatus group and coecatus group, excluding H. mexicanus0.054 0.041-0.066 0.054 0.032-0.079 N 0.002{1,2,3,4,5,6,7,9,10,12,13,14|8,11} H. mexicanus +H. virgulatus 0.052 0.039-0.066 0.052 0.031-0.078 N 0.002{1,5,6,7,8,9,10,11,12,13,14|2,3,4} viridipes group + H. sp. (ROBRT) 0.052 0.039-0.068 0.052 0.031-0.078 N 0.001!"Appendix 5  Counts of shared alleles between Habronattus species used for DFOIL tests Species 1 Species       2Species    3Species     4Outgro-up AAAAA BBBBA BBAAA AABBA ABABA ABBAA ABBBA BAABA BABAA BABBA BBABA BBBAA BAAAAABAAAAAABA AABAAVCC clade(A) aztecanus clypeatus mexicanus pyrrithrix ophrys 1261335 16612 1950 1550 105 91 300 120 102 279 569 358 3017 2014 2951 4534(B) sp.       (BLNDI) pyrrithrix clypeateus aztecanus ophrys 1305813 17312 2452 2108 55 55 248 37 52 188 300 346 2368 1846 3122 2023(C) sp. (SLCTY) clypeatussp. (BLNDI) pyrrithrix ophrys 1505981 17856 2576 2813 42 33 160 21 30 124 231 176 1192 1235 1711 2373(D) sp. (SLCTY) clypeatus mexicanus pyrrithrix ophrys 1399985 16507 2264 1709 68 38 145 34 40 99 471 333 1072 1140 2806 4400(E) sp. (SLCTY) clypeatus borealis pyrrithrix ophrys 1598152 18920 2600 2072 80 39 144 31 65 127 439 350 1208 1324 3004 3711(F) sp. (SLCTY) clypeatus borealis pyrrithrix ophrys 1344100 17806 2068 1787 89 78 340 75 88 312 490 338 3218 2121 2860 3650americanus group(G) sansoni american-us ophrys tarsalis signatus 1144994 33651 1257 678 34 49 74 21 37 63 266 468 375 562 1700 1325decorus group with americanus group(H) ophrys sp. (ESTU)zapotecan-us decorus signatus 1433854 31209 6391 2939 189 225 347 231 167 431 1853 1313 3424 6131 9970 14493(I) ophrys sp. (ESTU)cambridg-ei decorus signatus 1401567 31044 6024 2502 176 211 361 210 157 444 1491 1561 3395 6013 10165 10913(J) ophrys sp. (ESTU)oregonen-sis decorus signatus 1389314 31290 5940 2377 172 184 360 177 176 478 1303 1703 3370 6048 10296 9094(K) ophrys sp. (ESTU) jucundus decorus signatus 1389556 31429 5766 2391 171 211 355 177 191 483 1232 1784 3346 5971 10300 9529(L) ophrys sp. (ESTU) festus decorus signatus 1475505 33310 6145 2575 191 252 392 194 229 513 1523 1973 3595 6462 11029 12866!"Appendix 6 Counts of alleles shared between Habronattus species used for Patterson’s D statistic tests. Species 1 Species 2 Species 3 Outgroup ABBA BABA AAAA AABA ABAA BAAA BBAA BBBAamericanus group (H.sp.(ESTU))(A) ophrys tarsalis sp. (ESTU) signatus 604 437 1460667 6699 2314 2323 2062 43518(B) sansoni americanus sp. (ESTU) signatus 106 98 1314158 5972 763 465 2939 37177decorus group with americanus group(C) zapotecanus decorus ophrys signatus 2276 1620 1498519 10627 10931 16117 3547 34071(D) cambridgei decorus ophrys signatus 1868 1883 1466423 10183 11122 12084 3076 33903(E) oregonensis decorus ophrys signatus 1637 2058 1454931 10087 11288 10140 2970 34321(F) jucundus decorus ophrys signatus 1537 2138 1449870 9819 11189 10558 2961 34214(G) festus decorus ophrys signatus 1906 2401 1551936 10659 12187 14479 3233 36798H. sp. (ROBRT) and VCC clade(H) sp.  (SLCTY) jucundussp. (ROBRT) ophrys 531 1255 1128821 4047 4228 3236 1025 12437(I) festus jucundus sp. (ROBRT) ophrys 1031 1473 1315652 6191 5550 7773 1985 16703(J) festus sp. (SLCTY)sp. (ROBRT) ophrys 1051 476 1136547 4356 2939 5551 1541 13284!"Species 1 Species 2 Species 3 Species 4 Outgroup total # sites DFO DFO     p DIL DIL    p DFIDFI       p DOL DOL    p introgression(A) sp. (BLNDI) pyrrithrix clypeateus aztecanus ophrys 1411233 0.061 6.E-02 0.027 4.E-01 -0.070 0.068 -0.116 2.E-03 none(B) sp. (SLCTY) clypeatussp. (BLNDI) pyrrithrix ophrys 1564822 -0.086 4.E-02 -0.122 4.E-03 -0.058 0.227 -0.105 3.E-02 none(C) aztecanus clypeatus mexicanus pyrrithrix ophrys 1363153 -0.146 4.E-08 -0.155 7.E-09 -0.015 0.624 -0.026 4.E-01 aztecanus/clypeatus          <-> pyrrithrix(D) sp. (SLCTY) clypeatus mexicanus pyrrithrix ophrys 1454728 -0.096 3.E-03 -0.167 2.E-07 -0.011 0.814 -0.166 4.E-04 clypeatus <-> pyrrithrix(E) sp. (SLCTY) clypeatus borealis pyrrithrix ophrys 1660568 -0.008 8.E-01 -0.141 6.E-06 0.113 0.011 -0.157 4.E-04 clypeatus <-> pyrrithrix(F) aztecanus clypeatus borealis pyrrithrix ophrys 1453126 -0.094 1.E-03 -0.132 4.E-06 -0.011 0.733 -0.055 8.E-02 none(G) sansoni americanus ophrys tarsalis signatus 1221642 0.267 6.E-16 0.235 1.E-12 0.019 0.737 -0.075 2.E-01 sansoni/americanus  -> ophrysdecorus and americanus group(H) ophrys sp. (ESTU) zapotecanus decorus signatus 1559881 -0.161 0.E+00 -0.111 3.E-12 -0.010 0.688 0.116 4.E-06 ophrys -> decorus(I) ophrys sp. (ESTU) cambridgei decorus signatus 1522609 -0.005 8.E-01 0.042 1.E-02 -0.003 0.899 0.110 1.E-05 none(J) ophrys sp. (ESTU) oregonensis decorus signatus 1511334 0.104 2.E-10 0.111 1.E-11 0.068 0.008 0.085 9.E-04 ophrys/sp. (ESTU) <-> oregonensis (K) ophrys sp. (ESTU) jucundus decorus signatus 1513001 0.140 0.E+00 0.153 0.E+00 0.064 0.010 0.097 1.E-04 none(L) ophrys sp. (ESTU) festus decorus signatus 1608166 0.097 1.E-10 0.109 6.E-13 0.054 0.024 0.083 5.E-04 noneclypeatus and coecatus groupamericanus groupAppendix 7 Summary of DFOIL results. p-value threshold for significance adjusted using a Bonferroni correction for multiple comparisons p < 0.00074 (95% significance for 68 tests)!"Group Species 1 Species 2 Species 3 Outgroup total # sites D p introgressionamericanus group (H. sp. (ESTU))(A) ophrys tarsalis sp. (ESTU) signatus 1526228 0.146 2.E-06 tarsalis <-> sp. (ESTU)(B) sansoni americanus sp. (ESTU) signatus 1357252 0.115 8.E-02 nonedecorus group with americanus group(C) zapotecanus decorus ophrys signatus 1527650 0.174754707 0.E+00 decorus <-> ophrys(D) cambridgei decorus ophrys signatus 1490550 0.003888889 8.E-01 none(E) oregonensis decorus ophrys signatus 1478768 -0.113111986 2.E-11 oregonensis <-> ophrys(F) jucundus decorus ophrys signatus 1547434 -0.149 0.E+00 jucundus <-> ophrys(G) festus decorus ophrys signatus 1660536 -0.103 3.E-12 festus <-> ophrysH. sp. (ROBRT) and VCC clade(H) sp. (SLCTY) jucundus sp. (ROBRT) ophrys 1143143 -0.405 0.E+00 sp. (SLCTY) <-> sp. (ROBRT)(I) festus jucundus sp. (ROBRT) ophrys 1415913 -0.166 0.E+00 festus <-> sp. (ROBRT)(J) festus sp. (SLCTY) sp. (ROBRT) ophrys 1182634 0.360 0.E+00 sp. (SLCTY) <-> sp. (ROBRT)Appendix 8 Summary of Patterson’s D statistic results. p-value threshold for significance adjusted using a Bonferroni correction for multiple comparisons p < 0.00074 (95% significance for 68 tests)


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items