INCOMPLETE BARRIERS TO HETEROSPECIFIC MATING AMONG SOMATOCHLORA SPECIES (ODONATA: CORDULIIDAE) AS REVEALED IN MULTI-GENE PHYLOGENIES By Jordy Hernandez A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Entomology – Master of Science 2024 ABSTRACT Somatochlora, commonly known as the striped emeralds, is an enigmatic genus whose systematics have lagged other Odonata genera, with the last revision done by Walker (1925). North American Somatochlora inhabit fens, bogs, and forest streams, with most closely related species sharing a sympatric range. As a result, Somatochlora males have elaborate claspers which are species-specific and provide a morphological barrier to heterospecific mating, but exceptions have been observed. The objective of this project was to investigate the occurrence of heterospecific mating between North American Somatochlora species as inferred from multi- gene phylogenies. We employed the use of two mitochondrial genes (COI and ND3) and two nuclear genes (EF1-α and ITS2) to construct well-substantiated phylogenies using a maximum parsimony optimality criterion. Compared to nuclear genes (nDNA), mitochondrial genes (mtDNA) have a high nucleotide substitution rate, thereby allowing for the genetic discrimination of populations and species. Monophyly of mtDNA lineages is expected for closely related species because ancestral mtDNA lineages go extinct after a speciation event four times faster than nDNA lineages. Observation of non-monophyletic mtDNA lineages but monophyletic nDNA lineages between Somatochlora sister-species would indicate mtDNA introgression and suggest heterospecific matings. Our results highlighted three instances of heterospecific mating in the following groups: 1) S. hineana + S. tenebrosa; 2) S. kennedyi + S. forcipata + S. franklini; 3) S. calverti + S. provocans + S. filosa. In addition, the recovered topology accurately reflected previous taxonomic understanding of the genus. These multi-gene phylogenies of North American Somatochlora are the first, providing a foundation for future ecological and evolution studies and knowledge for effective decision-making and public policy, which is especially important for endangered species, Somatochlora hineana. This thesis is dedicated to my family, friends, and the color brown. iii ACKNOWLEDGMENTS I would like to thank my major professor, Dr. Anthony Cognato, for his guidance and assistance in helping me complete my thesis; without him this wouldn’t have been possible. I thank my thesis committee members, Dr. Henry Chung and Dr. Mariah Meek, for their guidance and review. Thank you to Dr. Hannah Burrack for her guidance and positivity. Thank you to Julie Craves and Darrin O’Brien for their mentorship, advice, and enthusiasm for odonates. Thank you to those individuals and institutions for providing specimens for this study. Thank you to Peter B. Mills for permitting the usage of his beautiful Somatochlora illustrations. This project was supported in part by the American Museum of Natural History Theodore Roosevelt Memorial Funds and Michigan State University Department of Entomology Hutson Research Proposal Funds. I’m deeply indebted to my parents, Gloria and Francisco Hernandez, and my siblings for their enormous support and love; thank you for always being there for me. Thank you to my friends for the many laughs we’ve shared and for making this a gratifying experience. Finally, thank you to the dragonflies. You are more than the molecular sequences analyzed herein. Your unique life history and dazzling beauty whether swimming or flying continues to inspire me and many others. iv TABLE OF CONTENTS CHAPTER 1: INTRODUCTION ....................................................................................................1 CHAPTER 2: HETEROSPECIFIC MATING AMONG SOMATOCHLORA ................................5 Introduction ..........................................................................................................................5 Materials and Methods .......................................................................................................10 Results ................................................................................................................................18 Discussion ..........................................................................................................................27 CHAPTER 3: CONCLUSION ......................................................................................................32 REFERENCES ..............................................................................................................................34 v CHAPTER 1: INTRODUCTION Odonata (dragonflies and damselflies) have arisen as a unique ecological study group owing to their vagility, aquatic/terrestrial life history, unique sexual reproduction, and as bioindicators for a changing landscape of the Anthropocene (Bybee et al., 2016). Odonata comprise one of the oldest winged insect lineages, and many phylogenetic studies at the ordinal, familial, generic, and specific level have been carried out (Misof et al., 2014; Troast et al., 2016; Bybee et al., 2021). While Odonata systematics is developed, especially compared to other insect lineages, questions remain regarding basal relationships. Libelluloidea is a superfamily of dragonflies (infraorder Anisoptera) that is generally agreed to include families Synthemistidae, Macromiidae, Corduliidae, and Libellulidae (Carle et al., 2015). Previous studies have struggled to find monophyly of Corduliidae, with some showing the family is paraphyletic (Blanke et al., 2013) and a lack of resolution of the intra-familial relationships of Corduliidae (Carle et al., 2015; Bybee et al., 2021). A foundational systematic understanding of Odonata is essential for providing a framework for other causal branches of biology (e.g., conservation), especially for Corduliidae which has the greatest Species-of-Greatest-Conservation-Need among Odonata families (Bried and Mazzacano, 2010). Somatochlora dragonflies, commonly known as the striped emeralds, are an enigmatic group of odonates characterized by green compound eyes and pale thoracic markings. Their bodies are covered in dark metallic browns, greens, blues, and blacks, while their thoraxes have yellow spots or stripes. Somatochlora species are morphologically similar and often difficult to identify (Mills, 2015). Species identification involves analyzing the shape and position of claspers and genital plates. Somatochlora is the largest genus in the family Corduliidae; however, the systematic understanding of Somatochlora has lagged behind other Odonata genera primarily due to their remote haunts and unique ecological requirements (Walker, 1925; Mead, 2021). Previous phylogenetic studies involving Somatochlora have used mitochondrial DNA (Kohli et al., 2018; Walker et al., 2020) or genomic data (Bybee et al., 2021) of one or few species. The flight season of Southern species of Somatochlora begins in late August and continues into late September (Walker, 1925). Young adults will leave the vicinity after emergence and seek sheltered spots. Adults will typically fly at altitudes of 9-15m, although they will fly closer to the ground in search of prey or if wing musculature is not fully developed. 1 Imaginal life lasts approximately a month and a half. Adults will eat small insects, such as midges and black flies, while flying. Adults will seek a breeding place 2-3 weeks after emergence (Walker, 1925). Males will fly low, close to the surface of the water or bog where they alternate between rapid movements and hovering motionless. Females seldom fly low over water except when ovipositing. Mating rituals involve the males pouncing on females, both falling to the water, and separating. There are two types of ovipositors: rounded ovipositors directed towards the caudal end can be found in S. arctica and S. alpestris while spout-shaped ovipositors directed downwards can be found in S. linearis, S. tenebrosa, and S. hineana. Those with rounded ovipositors will strike the water with the end of their abdomen, releasing eggs into the water. Those with spout-shaped ovipositors will deposit their eggs near the water’s edge while in flight (Walker, 1925). Somatochlora nymphs’ stadium duration depends on food and temperature conditions. Nymphs grow exceedingly slow with 6-7 molts per season and an overall nymphal development cycle of 4-5 years (Walker, 1925; Pintor and Soluk, 2006). Nymphs are ambush predators but have low hunting success and will seldom attack prey from a distance. Early-stage nymphs will feed on protozoa (i.e., Euglena, Paramecium) while later stages feed on larger aquatic invertebrates such as Cyclops, Daphnia, and oligochaetes. Nymphs can often be found in breeding places of adults in the benthic zone near the shore, where they will be frequently covered in mud or slime. Somatochlora nymphs prefer habitats with cool summer temperatures ranging from 16°C-20°C; consequently, nymphs prefer deep, lotic water systems (Walker, 1925). Before emergence, nymphs will climb onto wet moss above the water’s edge. Somatochlora is a stenotopic genus; they are sensitive to anthropogenic disturbances and will often avoid dry streams and polluted waters. Somatochlora dragonflies predominantly inhabit Palearctic and Nearctic realms, specifically subarctic/subalpine habitats (Walker, 1925). Most North American Somatochlora can be found in bog habitats from Lake Superior to Hudson Bay (Walker, 1925), but there are specific differences in habitat, such as S. sahlbergi preferring the tundra and S. calverti preferring sandy-forest streams (Dunkle, 2004; Schröter et al., 2012). Of the 25 North American Somatochlora species, all have been designated a Species of Greatest Conservation Need in at least one U.S. state which may reflect the habitat degradation of fens and bogs characteristic of Somatochlora (Bried and Mazzacano, 2010). 2 The Hine’s Emerald, Somatochlora hineana (Williamson), is a federally endangered dragonfly species (U.S. Fish and Wildlife Service, 2001) which is morphologically similar to S. tenebrosa (clamp-tipped emerald) (Williamson 1931). Somatochlora hineana requires three ecological requirements for appropriate habitat: 1) calcareous fens; 2) crayfish burrows for nymphal development; 3) shaded and unshaded pastures for foraging (Walker et al., 2020). A fen is a type of wetland that takes millennia to develop and is difficult to restore from anthropogenic disturbance (Weixelman and Cooper, 2009). During drought periods, S. hineana nymphs will seek refuge in burrows created by the devil crayfish, Cambarus diogenes, which retain moisture when the heat dries the open channel (Pintor and Soluk, 2006). Nymphs are preyed upon mostly by Aeshna dragonflies (mosaic darners), dytiscids (predaceous diving beetles), and sialids (alderflies). Somatochlora hineana nymphs can and will be preyed upon by C. diogenes but seeking shelter in crayfish burrows leads to greater survivorship rates than risk of desiccation in open channels (Pintor and Soluk, 2006). Somatochlora hineana nymphs may occupy crayfish burrows despite flowing water (Pintor and Soluk, 2006). The main range of S. hineana is from Southern Ontario to Wisconsin, Michigan, and Illinois (Craves et al., 2022). A marginal population resides in the Ozarks in Missouri (Walker et al., 2020). This marginal population has greater genetic diversity (that is, comprising an even distribution of older and younger mitochondrial haplotypes) than the core population of S. hineana residing in the Great Lakes (Walker et al., 2020). Conservation of the marginal population may be critical for the preservation of ancestral haplotypes which may be beneficial for future adaptability (Walker et al., 2020). Geographically isolated populations are often genetically different (Avise, 2004). Population genetic architecture is influenced by a variety of forces including gene flow, random genetic drift, natural selection, mutational divergence, and genetic recombination. Gene flow via migration is important in vagile organisms with high mobility, such as insects. This is because genetic material can readily be exchanged between populations by movement of individuals or gametes. Gene flow may be affected by climate change, where species adapt to changing environmental conditions by changes in population range, thereby altering the likelihood of closely related species encountering (Arce-Valdés and Sánchez-Guillén, 2022). Interspecific hybridization may occur as a result of incomplete reproductive isolating barriers after secondary contact of closely related species (Arce-Valdés and Sánchez-Guillén, 2022). Genetic structure 3 assessments can elucidate the extent of introgression to a higher degree than morphological assessments alone (Avise, 2004), such as determining patterns of unidirectional hybridization (Solano et al., 2018). The aims of this study were: 1) to reconstruct multi-gene phylogenies using mitochondrial DNA (mtDNA) and nuclear DNA (nDNA) for North American Somatochlora species; and 2) to investigate mitochondrial introgression indicative of heterospecific mating among North American Somatochlora by evaluating incongruence between phylogenies informed by mtDNA and nDNA. With the increasing number of studies indicating heterospecific mating is a common phenomenon in Odonata (Solano et al., 2018; Kornová et al., 2024), and that introgression runs deep in Odonata evolutionary history (Suvorov et al., 2022), we seek evidence of Somatochlora interspecific hybridization or introgression. In addition, we use a novel molecular dataset which helps reevaluate the classification and relationships of North American Somatochlora. This study can provide a framework for future conservation studies, something especially pertinent to Somatochlora where all 25 North American species have been considered of great conservation need in at least one U.S. state (Bried and Mazzacano, 2010). 4 CHAPTER 2: HETEROSPECIFIC MATING AMONG SOMATOCHLORA Introduction Mating between related species is a common occurrence among insects (Andersen et al., 2019; San Jose et al., 2023). Heterospecific mating attempts and heterospecific matings have been documented in the laboratory and field setting for the major insect orders, namely Hymenoptera, Diptera, Lepidoptera, Coleoptera, Odonata, and Orthoptera (Gröning and Hochkirch, 2008). Heterospecific matings are more likely to occur when heterospecifics are abundant and conspecifics are rare – a phenomenon known as the Hubbs Principle (Hubbs, 1955). Incomplete reproductive isolating barriers during allopatry occur frequently since time between speciation events is often much shorter than the window for hybridization (Chan and Levin, 2005). The effects of introgression will vary according to reproductive isolating barriers (e.g., prezygotic vs. postzygotic) and modes of inheritance (e.g., maternal vs paternal) For example, maternally inherited mitochondrial DNA will introgress more rapidly through ineffective prezygotic reproductive isolating barriers than paternal or biparental modes of inheritance (Chan and Levin, 2005). Prezygotic reproductive isolating barriers are especially sensitive to the proportion of immigrants; as a result, even a low migration rate can lead to high levels of introgression (Chan and Levin, 2005). Odonata reproductive isolating barriers are mainly characterized by morphological and ethological barriers (Tennessen, 1982; Barnard et al., 2017). Odonates have highly developed compound eyes; thus, they rely on visual stimuli such as sexual dimorphism in the form of color patterns and UV reflectance for mate recognition (Futahashi et al., 2019). Other behaviors rely on tactile stimuli of genitalia (Tennessen, 1982). Males have elaborate claspers which are species specific (Figure 1) and thus are important for the identification of con- and heterospecifics when in copula (McPeek et al., 2008). Male odonates use their terminal appendages to clasp the female’s head. The females recognize conspecific males according to cerci morphology (Tennessen, 1982), but exceptions have been observed (Bick and Bick, 1981; Solano et al., 2018). Indeed, genetic introgression runs deep in the evolutionary history of Odonata (Suvorov et al., 2022). Heterospecific pairings have been recorded across families, mixed genera, and mixed species, including Somatochlora (Bick and Bick, 1981). However, evidence of heterospecific mating among Somatochlora is restricted to observational records and 5 morphological evidence of hybridization (Walker, 1925). For example, Walker noted a Somatochlora female of intermediate morphology; she had S. cingulata coloration and stature but with S. albicincta terminal appendage characteristics. In another instance, Bick and Bick (1981) observed tandem formation between S. albicincta (♂) and S. hudsonica (♀), but there was no direct evidence of hybridization. Somatochlora sahlbergi is known to hybridize with S. hudsonica and S. albicincta in northern Yukon where their ranges overlap (Cannings and Cannings, 1985). Potential S. hineana hybridization is restricted to accounts of Somatochlora specimens from the Missouri Ozarks whose identity was difficult to confirm (Monroe and Britten, 2014). 6 Figure 1 Illustrations of representative Somatochlora males used in this study and morphology of their cerci. The dorsal (left column) and lateral (center left column) view of the cerci are taken from Walker (1925) for all species except for S. hineana (Williamson, 1931). Illustrations were obtained from Mills (2015) with permission. 7 The use of mitochondrial genes for the discrimination of conspecifics and population genetic structure is well documented (Avise, 2004; Kohli et al., 2018). Despite its utility, several inherent properties of mtDNA limit conclusions drawn from the sole use of it. Mitochondrial genes can have more heterogeneity of site substitution variation than nuclear genes, leading to more homoplasy (Rubinoff and Holland, 2005). Nuclear genes have several qualities that are detrimental to their use in phylogenetics, such as heterozygosity, low substitution rates, low copy number, and paralogous loci. However, nuclear genes have less biased base composition than mitochondrial genes. Combined analysis of mtDNA and nDNA provides the most statistically robust and congruent phylogenies (Zhang and Hewitt, 2003; Rubinoff and Holland, 2005; Cameron, 2014). Two nDNA loci, Elongation Factor 1-α (EF1-α) and Internal Transcribed Spacer 2 (ITS2), as well as two mtDNA loci, NADH Dehydrogenase 3 (ND3) and Cytochrome C Oxidase I (COI) have been used for the discrimination of odonate species and other animal taxa (Pilgrim and Von Dohlen, 2008; Yao et al., 2010; Bergmann et al., 2013). MtDNA and nDNA have distinct modes of inheritance which can provide insight into the degree of reproductive isolation for introgression asymmetries influenced by mate choice mechanisms (Solano et al., 2018). MtDNA lineages go extinct after a speciation event four times faster than nuclear genes thus providing relative timing of population isolation and speciation events (Avise, 2004). Comparison of mtDNA and nDNA lineages can reveal four possible scenarios concerning speciation (Figure 2): (1) Non-monophyly for mtDNA and nDNA lineages suggests incomplete speciation; (2) monophyly of mtDNA and nDNA lineages suggests complete speciation; (3) monophyly of mtDNA but non-monophyly of nDNA lineages suggest recent speciation; (4) non-monophyly of mtDNA lineages but monophyly of nDNA lineages suggest gene flow after speciation. Observation of scenario (4) would provide phylogenetic evidence for heterospecific mating among dragonflies. 8 Figure 2 The four scenarios concerning speciation by comparison of mitochondrial- and nuclear based phylogenies. The objective of this study is to investigate heterospecific mating between North American Somatochlora species as inferred from multi-gene phylogenies. Somatochlora are charismatic dragonflies with brown, green, blue, or black metallic bodies marked by yellow spots or stripes. These species, including Somatochlora hineana (the only federal-listed endangered dragonfly), are indicators of aquatic habitat quality because many inhabit ecologically sensitive wetlands such as fens (Vogt and Cashatt, 1994; Spoelstra and Post, 2023). Given the anecdotal observations of mating between Somatochlora species, we hypothesize potential gene transfer among species. Observation of non-monophyletic mtDNA lineages but monophyletic nDNA lineages between Somatochlora sister-species would indicate mtDNA introgression and suggest heterospecific matings. In addition, because this is the first phylogenetic study of North American Somatochlora, we broadly discuss the species relationships in reference to previous morphologically based taxonomy. 9 Materials and Methods Specimens were obtained from museum collections and from the field. Most museum specimens were obtained from private and institutional collections. Field-collected specimens were enveloped, placed in acetone for 12-18 hours, and stored in a closed plastic container on silica gel. Vouchers were deposited in the A.J. Cook Arthropod Research Collection, MSU. In total, 108 specimens representing 31 Somatochlora species with 24/25 North American species and the remainder from Eurasia were included in this study (Table 1). Five taxa of Corduliidae (Neurocordulia yasmakanensis, Dorocordulia libera, Cordulia shurtlefii, Epitheca spinigera, and Epitheca prínceps) were selected as the outgroups. For DNA extraction, tissue from a meso- or meta- leg was processed with the Qiagen DNeasy Blood and Tissue Kit following manufacturer protocols. Purified DNA from each specimen was used to amplify four target genes – Cytochrome C Oxidase 1 (COI), NADH Dehydrogenase 3 (ND3), Elongation Factor 1-α (EF1-α), and Internal Transcribed Spacer 2 (ITS2). PCR primers were selected from previously published primers or designed for this study (Table 2). PCR cocktails contained a mixture of 17.25 µL ddH20, 2.5 µL 10X PCR buffer (Qiagen), 1.0 µL 25 mM MgCl2 (Qiagen), 0.5 µL dNTP mix (Qiagen), 2 µL template DNA, 0.25 µL HotStar Taq DNA polymerase (Qiagen), equating to a total volume of 25 µL. PCR was performed with a PTC-2000 MJ Research Thermocycler (MJ Research, Watertown, MA). Nuclear genes (EF1-α, ITS2) were initially denatured for 15 min at 95°C, followed by 36 cycles of denaturation at 94°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 60 s. Final extension was at 72°C for 5 min. Alternative primer pairs were used and with the same thermocycler settings as stated previously except for a shorter extension time for 30 s at 72°C for samples that did not yield sufficient PCR product for EF1-α. Mitochondrial genes (COI, ND3) were initially denatured for 15 min at 96°C, followed by 38 cycles of denaturation at 94°C for 30 s, annealing at 45°C for 30 s, and extension at 72°C for 45 s. Final extension was at 72°C for 5 min. PCR products were visualized using agarose gel electrophoresis and ethidium bromide illuminated under UV light. Following PCR visualization, samples were cleaned using ExoSAP- IT according to manufacturer protocol (Applied Biosytems by Thermo Fisher Scientific, Vilmus, Lithuania). Cleaned PCR products were sequenced (both strands via Sanger) at the Michigan State University Genomics Core Facility (East Lansing, MI). 10 Sense and antisense sequences were assembled and edited using Sequencher software version 5.0-7082 (Gene Codes Corporation). Consensus sequences were trimmed of primer sites and examined for ambiguous base calls and blasted In GenBank to check for potential contamination. Contamination or pseudogenes were not discovered. The resulting assembled sequences (base pairs: EF1-α = 618, ITS2 = 417, COI =487, ND3 = 541) were deposited in NCBI GenBank (Table 1). Protein coding genes were manually aligned given that nucleotide insertions/deletions and introns were not observed. ITS2 sequences were length variable, thus they were aligned using the EMBL-EBI Multiple Sequence Comparison by Log-Expectation program (MUSCLE) using the default settings (Madeira et al., 2022). A NEXUS file was created with the aligned sequences and use to infer phylogenies with PAUP* version 4.0a169 (Swofford, 2002). All DNA (the total dataset), mtDNA and nDNA data sets were analyzed. MtDNA and nDNA data were missing for some specimens and these were excluded from the genome specific phylogenetic analyses. Maximum parsimonious analyses consisted of 200 random stepwise addition heuristic searches with a tree-bisection-reconnection (TBR) branch-swapping algorithm. Gaps were treated as missing data. Jackknife branch support values (JK) were determined with 50%-character deletion using a 500 simple stepwise addition heuristic searches with a TBR branch swapping algorithm. Partition Bremer support was analyzed for the resulting strict consensus tree reconstructed for the total data set. TreeRot v2 (Sorenson, 1999) was used to build constraint trees for each node. Using the resulting constraint file, PAUP* was used to search for the most parsimony tree using the same search conditions explained above except branch swapping occurred on 500 best trees for each stepwise addition replicate. The averaged tree length for multiple trees found in each constraint tree search was subtracted from the partitioned lengths found in the unconstrained parsimony analysis. A negative value represented conflicting phylogenetic signal while a positive value represented supporting phylogenetic signal. All nodes unresolved in the strict consensus tree resulting from the simultaneous analysis of all data were zero. 11 Table 1. Voucher information for specimens used. “N/A” = missing sequence data. GenBank Accession No. COI ND3 EF1-α ITS2 N/A N/A Genus Epitheca Epitheca Species shurtleffii libera princeps spinigera Locality Voucher USA: Michigan: Marquette Co. COR_SHU1 Cordulia USA: Michigan: Marquette Co. DOR_LIB1 Dorocordulia USA: Michigan: Manistee Co. EPI_PRI1 EPI_SPI1 USA: Michigan: Marquette Co. NEU_YAM1 Neurocordulia yamaskanensis USA: Michigan: Marquette Co. SOM32 SOM91 SOM75 SOM76 SOM28 SOM112 SOM113 SOM114 SOM116 SOM33 SOM106 SOM107 SOM119 SOM120 SOM121 SOM35 SOM68 SOM8 SOM92 SOM93 SOM94 SOM104 PP749106 PP751515 PP757551 PP748920 PP749107 PP751516 PP757552 PP748921 PP751517 PP757553 PP748922 PP749108 PP751518 PP757554 PP748923 PP749109 PP751519 PP757555 PP748924 PP749110 PP748925 N/A Canada: British Columbia PP751520 PP757556 PP748926 N/A Canada: New Brunswick N/A PP751521 N/A N/A Norway PP748927 N/A N/A N/A Norway USA: Minnesota: Koochiching Co. PP749111 PP751522 PP757557 PP748928 PP749112 PP751523 PP757558 PP748929 USA: Florida: Nassau Co. PP749113 PP751524 PP757559 PP748930 USA: Florida: Gadsden Co. PP749114 PP751525 PP757560 PP748931 USA: Florida: Leon Co. PP749115 PP751526 PP757561 PP748932 USA: Florida: Liberty Co. PP749116 PP751527 PP757562 PP748933 Canada: Saskatchewan: Jade Lake PP749117 PP751528 PP757563 PP748934 USA: Michigan: Marquette Co. PP749118 PP751529 PP757564 PP748935 USA: Michigan: Marquette Co. PP749119 PP751530 PP757565 PP748936 USA: Michigan: Marquette Co. PP749120 PP751531 PP757566 PP748937 USA: Michigan: Marquette Co. PP749121 PP751532 PP757567 PP748938 USA: Michigan: Baraga Co. PP749122 USA: Wisconsin: Vilas Co. PP749123 PP751533 USA: Michigan: Mackinac Co. PP749124 PP751534 PP757569 USA: Michigan: Marquette Co. USA: New Hampshire: Grafton Co. USA: New York: Broome Co. USA: Maine: Somerset Co. USA: Michigan: Menominee Co. Somatochlora albicincta Somatochlora albicincta Somatochlora alpestris Somatochlora arctica Somatochlora brevicincta Somatochlora calverti Somatochlora calverti Somatochlora calverti Somatochlora calverti Somatochlora cingulata Somatochlora elongata Somatochlora elongata Somatochlora elongata Somatochlora elongata Somatochlora elongata Somatochlora elongata Somatochlora elongata Somatochlora elongata Somatochlora elongata Somatochlora elongata Somatochlora elongata Somatochlora ensigera PP751535 PP757570 PP748940 PP749125 PP751536 PP757571 PP748941 PP749126 PP751537 PP757572 PP748942 PP748943 N/A PP748939 N/A PP757568 N/A PP751538 N/A N/A N/A N/A 12 Table 1 (cont’d) GenBank Accession No. Voucher SOM105 SOM26 SOM115 SOM117 SOM68B SOM77 SOM102 SOM16 SOM37 SOM27 SOM41 SOM66 SOM14 SOM78 SOM163 SOM118 SOM1 SOM2 SOM54 SOM55 SOM96 SOM42 SOM97 SOM3 SOM60 SOM70 SOM95 N/A N/A ITS2 ND3 COI N/A Locality USA: Michigan: Menominee Co. USA: Minnesota: Red Lake Co. USA: Texas: Hardin Co. USA: Texas: Hardin Co. USA: Florida: Bay Co. Species Genus Somatochlora ensigera Somatochlora ensigera filosa Somatochlora filosa Somatochlora filosa Somatochlora flavomaculata Lithuania Somatochlora forcipata Somatochlora forcipata Somatochlora forcipata Somatochlora franklini Somatochlora franklini Somatochlora franklini Somatochlora Somatochlora franklini Somatochlora graeseri Somatochlora hineana Somatochlora hineana Somatochlora hineana Somatochlora hineana Somatochlora hineana Somatochlora hineana Somatochlora hineana Somatochlora hudsonica Somatochlora hudsonica incurvata Somatochlora incurvata Somatochlora incurvata Somatochlora incurvata Somatochlora EF1-α PP751539 PP757573 PP748944 PP749127 PP751540 PP757574 PP748945 PP749128 PP751541 PP757575 PP748946 PP749129 PP751542 PP757576 PP748947 PP749130 PP751543 PP757577 PP748948 PP757578 PP748949 PP749131 PP751544 PP757579 PP748950 USA: West Virginia: Tucker Co. PP749132 PP751545 PP757580 PP748951 USA: Wisconsin: Forest Co. PP748952 PP749133 PP751546 USA: Vermont: Essex Co. USA: Minnesota: Koochiching Co. PP749134 PP751547 PP757581 PP748953 PP748954 N/A USA: New Hampshire: Coos Co. PP749135 PP751549 PP757582 PP748955 USA: Michigan: Crawford Co. PP749136 PP751550 PP757583 PP748956 USA: Michigan: Alger Co. PP749137 PP751551 PP757584 PP748957 Russia: Sakhalin PP749138 PP751552 PP757585 PP748958 USA: Michigan: Mason Co. PP749139 PP751553 PP757586 PP748959 USA: Michigan: Mackinac Co. PP749140 PP751554 PP757587 PP748960 USA: Michigan: Oceana Co. PP749141 PP751555 PP757588 PP748961 USA: Michigan: Oceana Co. PP749142 PP751556 PP757589 PP748962 USA: Michigan: Mackinac Co. PP749143 PP751557 PP757590 PP748963 USA: Michigan: Mackinac Co. PP757591 PP748964 USA: Wisconsin: Door Co. PP748965 PP749144 PP751558 Canada: British Columbia PP749145 PP751559 PP757592 PP748966 USA: Colorado: Larimer Co. PP749146 PP751560 PP757593 PP748967 USA: Michigan: Chippewa Co. PP749147 PP751561 PP757594 PP748968 USA: Michigan: Mackinac Co. PP749148 PP751562 USA: Michigan: Mackinac Co. PP749149 PP751563 PP757595 PP748969 Canada: Nova Scotia PP751548 N/A N/A N/A N/A N/A N/A N/A 13 Table 1 (cont’d) Voucher SOM98 SOM100 SOM101 SOM17 SOM43 SOM57 SOM99 SOM24 SOM6 SOM129 SOM79 SOM109 SOM18 SOM44 SOM110 SOM71 SOM72 SOM80 SOM45 SOM46 SOM81 SOM82 SOM10 SOM11 SOM111 SOM12 SOM164 Locality Canada: Nova Scotia USA: Maine: Somerset Co. USA: Alaska USA: Wisconsin: Forest Co. USA: Wisconsin: Eau Claire Co. USA: Michigan: Mackinac Co. USA: Maine: Lake Co. USA: Texas: Jasper Co. USA: Michigan: Shiawassee Co. USA: Texas: Marion Co. Lithuania USA: Michigan: Alpena Co. USA: Wisconsin: Forest Co. Canada: British Columbia USA: Arkansas: Washington Co. USA: Oklahoma: McCurtain Co. USA: Mississippi: Stone Co. USA: Florida: Washington Co. Species Genus Somatochlora incurvata Somatochlora kennedyi Somatochlora kennedyi Somatochlora kennedyi Somatochlora kennedyi Somatochlora kennedyi Somatochlora kennedyi linearis Somatochlora linearis Somatochlora Somatochlora margarita Somatochlora metallica Somatochlora minor Somatochlora minor Somatochlora minor Somatochlora ozarkensis Somatochlora ozarkensis Somatochlora provocans Somatochlora provocans Somatochlora semicircularis Canada: British Columbia Somatochlora semicircularis Canada: British Columbia Somatochlora semicircularis USA: Idaho: Idaho Co. Somatochlora septentrionalis Canada: British Columbia Somatochlora Somatochlora Somatochlora Somatochlora Somatochlora USA: Michigan: Hillsdale Co. USA: Michigan: Benzie Co. USA: Indiana: Johnson Co. USA: Michigan: Benzie Co. USA: Michigan: Mason Co. tenebrosa tenebrosa tenebrosa tenebrosa tenebrosa 14 GenBank Accession No. COI ND3 EF1-α ITS2 N/A N/A N/A PP749150 PP751564 PP757596 PP748970 PP749151 PP751565 PP757597 PP748971 PP749152 PP751566 PP757598 PP748972 PP749153 PP751567 PP757599 PP748973 PP748974 PP749154 PP751568 PP749155 PP751569 PP748975 PP749156 PP751570 PP757600 PP748976 PP749157 PP751571 PP757601 PP749158 PP751572 PP757602 PP748977 PP749159 PP751573 PP757603 PP748978 PP749160 PP751574 PP757604 PP748979 PP749161 PP751575 PP757605 PP748980 N/A PP749162 PP751576 PP757606 PP749163 PP751577 PP748981 PP749164 PP751578 PP757607 PP748982 PP749165 PP751579 PP757608 PP751580 PP757609 PP749166 PP751581 PP757610 PP748983 PP748984 PP749167 PP751582 PP749168 PP751583 PP757611 N/A PP749169 PP751584 PP757612 PP748985 PP748986 PP749170 PP751585 PP757613 PP748987 PP749171 PP751586 PP757614 PP748988 PP749172 PP751587 PP757615 PP748989 PP748990 PP749173 PP751588 PP748991 PP749174 PP751589 N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A Table 1 (cont’d) Voucher SOM165 SOM19 SOM21 SOM22 SOM30 SOM31 SOM56 SOM83 SOM84 SOM85 SOM86 SOM9 SOM87 SOM88 SOM49 SOM69 SOM7 SOM89 SOM90 SOM122 SOM123 SOM52 SOM59 SOM64 SOM67 SOM73 Species tenebrosa tenebrosa tenebrosa tenebrosa tenebrosa tenebrosa tenebrosa tenebrosa tenebrosa tenebrosa tenebrosa tenebrosa Genus Somatochlora Somatochlora Somatochlora Somatochlora Somatochlora Somatochlora Somatochlora Somatochlora Somatochlora Somatochlora Somatochlora Somatochlora Somatochlora uchidai Somatochlora viridiaenea Somatochlora walshii Somatochlora walshii Somatochlora walshii Somatochlora walshii Somatochlora whitehousei Somatochlora williamsoni Somatochlora williamsoni Somatochlora williamsoni Somatochlora williamsoni Somatochlora williamsoni Somatochlora williamsoni Somatochlora williamsoni GenBank Accession No. N/A N/A COI N/A ND3 N/A EF1-α N/A Locality ITS2 PP748992 USA: Michigan: Mason Co. PP749175 PP751590 PP757616 PP748993 USA: New York: Broome Co. PP749176 PP751591 PP757617 USA: Oklahoma: McCurtain Co. PP749177 PP751592 PP757618 PP748994 USA: Oklahoma: McCurtain Co. PP749178 PP751593 PP757619 PP748995 USA: Vermont: Washington Co. PP749179 PP751594 PP757620 PP748996 USA: Wisconsin: Sauk Co. PP748997 PP749180 PP751595 USA: Michigan: Lenawee Co. USA: Pennsylvania: Huntingdon Co. PP749181 PP751596 PP757621 PP748998 PP749182 PP751597 PP757622 PP748999 USA: Tennessee: Franklin Co. PP749183 PP751598 PP757623 PP749000 USA: Kentucky: Carter Co. PP749184 PP751599 PP757624 PP749001 Canada: Nova Scotia PP749185 PP751600 PP757625 PP749002 USA: Michigan: Manistee Co. PP749186 PP751601 PP757626 PP749003 Japan PP749004 Japan PP749005 USA: Maine: Washington Co. N/A USA: Michigan: Mackinac Co. USA: Michigan: Marquette Co. Canada: Nova Scotia Canada: British Columbia USA: Michigan: Baraga Co. USA: Michigan: Marquette Co. Canada: Saskatchewan: Jade Lake USA: Michigan: Mackinac Co. USA: Michigan: Marquette Co. USA: Michigan: Mackinac Co. USA: Michigan: Mackinac Co. PP749187 PP751602 PP749188 PP751603 PP749189 PP751604 PP757627 PP749006 PP749190 PP751605 PP757628 PP749007 PP749191 PP757629 PP749008 PP749192 PP751606 PP757630 PP749009 PP749193 PP751607 PP757631 PP749010 PP749194 PP751608 PP757632 PP749011 PP749195 PP751609 PP757633 PP749196 PP751610 PP757634 PP749012 PP749197 PP751611 PP749198 PP751612 PP757635 N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A 15 Table 1 (cont’d) Voucher SOM74 Genus Somatochlora williamsoni Species Locality USA: Michigan: Shiawassee Co. COI ND3 PP749199 PP751613 EF1-α N/A ITS2 N/A GenBank Accession No. 16 Table 2. Primer sequences. Gene Primers Sequence Annealing Temp. Reference COI ND3 ITS2 COI - 1709F COI - 2191R TG-J-5584 TN-N-6160 CAS5p8sFc_Odon CAS28sB1d_Odon 5' TTCTTTTCCTCCSCTTAYTRATATGCTTAA 3' 5' TAATTGGAGGATTTGGAAATTG 3' 5' CCYGGTARAATTARAATRTARACTTC 3' 5' AGTATATTTGACTTCCAATC 3' 5' TCAATTATATCATTAACAGTGA 3' 5' TGAACATCGACATTTYGAACGCACAT 3' EF1-α EF1Rf_Odonate EF1Ra_Odonate Som EF1a A FW Som EF1a A REV Som EF1a B FW Som EF1a B REV 5' GGAGAATTCGAAGCTGGTATCTC 3' 5' GACACGTTCTTCACGTTGAAACC 3' 5' CACTCCTCGCTTTCACTCTT 3' 5' GCACTTTCCGTCAGCATTTC 3' 5' GATGGAAGGTGGAGCGTAAG 3' 5' CTCTTGGAGAGCTTCGTGATG 3' 45° 45° 45° 45° 55° 55° 55° 55° 55° 55° 55° 55° Kjer et al. (2001) Kjer et al. (2001) Beckenbach et al. (2008) Beckenbach et al. (2008) Ji et al. (2003) Ji et al. (2003) Pilgrim and Von Dohlen (2008) Pilgrim and Von Dohlen (2008) Designed in this study Designed in this study Designed in this study Designed in this study 17 Results Mitochondrial analysis included 1030 characters of which 282 (27.4%) were parsimony informative and recovered 100000 trees with a length of 969. The strict consensus tree was mostly resolved and recovered most species as monophyletic except for three instances (Figure 3). First, S. hineana was rendered non-monophyletic with respect to S. tenebrosa. Somatochlora hineana SOM1 was recovered in a supported clade (86 JK) with S. tenebrosa SOM11 and S. tenebrosa SOM12. This clade was sister to the rest of S. hineana with strong support (96 JK). The remaining S. tenebrosa specimens were recovered in a strongly supported (98 JK) clade sister to the S. hineana clade. Second, there was a lack of resolution within a clade comprised of S. kennedyi, S. franklini, S. forcipata, and S. semicircularis. These species were recovered in a polytomy. Within this polytomy occurred a strongly supported (93 JK) clade of S. incurvata. Third, S. provocans, S. calverti, and S. filosa were recovered as non-monophyletic. Somatochlora filosa was rendered paraphyletic with respect to S. provocans SOM80 and S. calverti SOM114. Somatochlora provocans SOM80 was recovered in a poorly supported clade (70 JK) with S. filosa and S. calverti SOM114. Jackknife support values were greater for internal branches and smaller for intraspecific clades. Nuclear analysis considered 992 included characters of which 271 (27.3%) were parsimony informative and recovered 55000 trees with a length of 928. PCR and sequencing of the nDNA loci had a lower success as compared to the mtDNA loci, with failure for 22 specimens. Specimens missing either EF1-α or ITS2 sequences were excluded from the analysis of nDNA loci. Most groups recovered as non-monophyletic in the mtDNA-informed phylogeny were recovered as monophyletic in the nDNA analysis (Figure 4). Somatochlora hineana was recovered as a monophyletic clade with poor support (66 JK). Somatochlora kennedyi, S. incurvata, S. franklini, and S. forcipata were each recovered as monophyletic with moderate to strong jackknife support values of 96, 84, 92, and 99, respectively. Somatochlora tenebrosa was recovered unresolved but separate from S. hineana, with no specimens resolving except for S. tenebrosa SOM9 which was sister to S. elongata SOM92 in an unsupported clade. Somatochlora filosa was recovered as monophyletic with strong support (87 JK). Somatchlora calverti was recovered as monophyletic in an unsupported clade (51 JK). Somatochlora provocans was recovered sister to S. calverti. Somatochlora walshii recovered as monophyletic in the mtDNA- informed phylogeny but non-monophyletic in the nDNA analysis. 18 Combined analysis of mitochondrial and nuclear loci yielded the greatest resolution, for a total of 2022 characters of which 559 (27.6%) were parsimony-informative. There were 79000 most parsimonious trees (length=1995). Partition Bremer support values showed COI and ITS2 to have the most positive clade support, while ND3 and EF1-α had less support (Table 3). In general, COI and ITS2 supported conspecific relationships, while ND3 and EF1-α supported heterospecific and more inclusive clade relationships. Most groups were recovered as monophyletic (Figure 5). Somatochlora hineana resolved as monophyletic (1 BS), and most of S. tenebrosa resolved in a clade (1 BS) sister to S. hineana. The southern clade of S. calverti + S. provocans + S. filosa + S. margarita + S. ozarkensis was recovered in a well-supported (10 BS) but unresolved clade, with S. filosa resolving as monophyletic (2 BS). Somatochlora linearis recovered in a clade (1 BS) sister to southern clade. Somatochlora ensigera recovered in a strongly supported (6 BS) clade outside of S. linearis. Somatochlora minor, S. elongata, S. williamsoni, and S. walshii were recovered as monophyletic with Bremer support values of 1, 1, 4, and 4, respectively. Somatochlora graeseri and S. uchidai were recovered as sister species with strong support (27 BS). Somatochlora incurvata and S. forcipata were recovered as sister each other (1 BS). Somatochlora franklini and S. kennedyi were recovered as sister to each other (4 BS). This clade of S. incurvata, S. forcipata, S. franklini, S. kennedyi, and S. semicircularis was recovered with strong support (17 BS). A clade containing S. brevicincta, S. albicincta, S. hudsonica, S. cingulata, S. whitehousei, S. septentrionalis, S. alpestris, and S. arctica was recovered with moderate support (3 BS). Somatochlora albicincta and S. hudsonica were recovered as monophyletic with strong support, with Bremer support values of 4 and 7, respectively. There were a few instances of nonmonophyly (Figure 5). Somatochlora hineana rendered S. tenebrosa paraphyletic, with two S. tenebrosa specimens (SOM11 & SOM 12) grouping with S. hineana at node 9. Mitochondrial genes (COI + ND3) provided conflicting or little support for node 9 (Table 3). ITS2 provided most support for this node. Somatochlora kennedyi, S. franklini, and S. forcipata each resolved as monophyletic. In the case of S. franklini and S. forcipata, ITS2 lent the most support (Table 3). In the case of S. kennedyi, COI and ND3 lent the most support. Somatochlora filosa resolved as monophyletic with EF1-α contributing most to this clade. S. provocans, S. calverti, S. margarita, and S. ozarkensis were recovered in a polytomy. 19 Figure 3 Rooted mitochondrial DNA (COI + ND3) strict consensus tree of 100000 most parsimonious trees for 100 Somatochlora specimens. Numbers above branches indicate jackknife support values greater than 50. Highlighted clades show mito-nuclear discordance and correspond to highlighted clades in Figure 4. 20 Figure 4 Rooted nuclear DNA (EF1-α + ITS2) strict consensus tree of 55000 most parsimonious trees for 86 Somatochlora specimens. Numbers above branches indicate jackknife support values greater than 50. Highlighted clades show mito-nuclear discordance and correspond to highlighted clades in Figure 3. 21 Figure 5 Rooted strict consensus tree of 79000 most parsimonious trees for 103 Somatochlora specimens using all data (COI + ND3 + EF1-α + ITS2). Numbers indicate nodes corresponding to Table 3. Colored clades indicate prior taxonomic groups following Walker’s (1925) group classification. 22 Figure 6 One of 79000 most parsimonious trees for 103 Somatochlora specimens using all data (COI + ND3 + EF1-α + ITS2). Colored clades on the left correspond to the enlarged clades on the right. Numbers indicate nodes corresponding to Table 3. 23 Table 3. Partition Bremer Support values. Node numbers correspond to Figure 5 and Figure 6. Nodes highlighted in grey correspond to zero net support. EF1-α 4.909091 -0.690909 -0.033117 -0.087879 -0.218182 0.118788 1.30101 0.085455 0.345455 -1.454545 -0.081212 0.098788 -0.034545 -0.004545 0.20101 0.045455 0.198788 0.020455 0.160839 0.175455 -0.374545 0.295455 0.127273 2.345455 -0.054545 -1.225974 -0.081818 0.345455 0.345455 0.345455 0.345455 3.066883 2.313102 0.345455 2.549621 1.738312 0.002597 -1.054545 COI -0.313636 0.722727 0.512662 1.259091 0.009091 0.575758 0.070202 1.109091 -0.990909 2.109091 0.699091 0.982424 1.424091 0.971591 0.903535 -0.457576 1.192424 0.880966 0.105245 0.594091 1.299091 1.484091 1.336364 0.259091 0.989091 1.201948 0.304545 0.086014 0.444091 0.569091 0.697552 4.548377 4.326738 0.909091 4.750758 3.694805 0.730519 -0.390909 ITS2 1.804545 -0.195455 -0.443506 0.385859 -0.045455 -0.50303 2.691414 -0.619697 1.363636 -0.60303 -0.41303 -0.779697 -0.561364 -0.523864 -0.536364 -0.10303 -0.766364 -0.433239 -0.263287 -0.236364 0.533636 -0.136364 -0.563636 -0.291919 -0.569697 -0.286364 -0.281818 -0.174825 -0.361364 -0.456364 -0.624825 -1.136364 -0.986364 0.363636 -1.219697 -0.993506 2.727922 1.363636 ND3 23.6 0.163636 -0.036039 -0.557071 0.254545 -0.191515 -3.062626 -0.574848 0.281818 -0.051515 -0.204848 -0.301515 -0.828182 -0.443182 -0.568182 0.515152 -0.624848 -0.468182 -0.002797 -0.533182 -0.458182 -0.643182 1.1 -0.312626 -0.364848 0.31039 0.059091 -0.256643 -0.428182 -0.458182 -0.418182 3.521104 -4.653476 0.381818 -5.080682 -3.43961 2.538961 1.081818 Total BS 30 0 0 1 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 1 1 2 2 0 0 0 0 0 0 0 10 1 2 1 1 6 1 Node Node 1 Node 2 Node 3 Node 4 Node 5 Node 6 Node 7 Node 8 Node 9 Node 10 Node 11 Node 12 Node 13 Node 14 Node 15 Node 16 Node 17 Node 18 Node 19 Node 20 Node 21 Node 22 Node 23 Node 24 Node 25 Node 26 Node 27 Node 28 Node 29 Node 30 Node 31 Node 32 Node 33 Node 34 Node 35 Node 36 Node 37 Node 38 24 Table 3 (cont’d) Node Node 39 Node 40 Node 41 Node 42 Node 43 Node 44 Node 45 Node 46 Node 47 Node 48 Node 49 Node 50 Node 51 Node 52 Node 53 Node 54 Node 55 Node 56 Node 57 Node 58 Node 59 Node 60 Node 61 Node 62 Node 63 Node 64 Node 65 Node 66 Node 67 Node 68 Node 69 Node 70 Node 71 Node 72 Node 73 Node 74 Node 75 Node 76 Node 77 COI 0.647552 0.569091 0.504545 1.479091 0.112424 0.009091 0.786869 2.909091 0.142424 0.459091 0.349091 0.615758 1.390909 1.993706 2.959091 0.759091 0.320629 1.286014 0.450758 0.609091 0.609091 -0.903409 2.006459 0.092424 0.674476 1.199091 1.124716 0.813636 0.245202 0.579679 3.040909 0.259091 1.509091 2.559091 3.092424 2.584091 0.809091 0.092424 1.959091 EF1-α 0.299301 0.078788 -0.263636 -0.874545 0.118788 0.152597 -0.276768 0.345455 -0.087879 -0.175974 0.125455 0.212121 0 0.176224 0.245455 -1.354545 0.153147 -0.185315 0.295455 0.306993 0.195455 0.320455 2.950718 0.185455 -0.085315 -0.107879 0.220455 0 -0.032323 -0.131016 -0.463636 -0.354545 -0.014545 0.345455 -0.981212 0.328788 0.034343 -0.132323 0.345455 ND3 0.627972 0.595152 0.131818 0.191818 -0.058182 0.049675 0.248485 -1.618182 0.248485 0.063961 -0.338182 -0.118182 -0.190909 -0.602797 0.431818 1.231818 -0.291259 -0.591259 -0.276515 -0.391259 -0.299432 0.394318 -1.420813 0.048485 -0.033566 4.295152 -0.068182 1.718182 -0.104293 0.014171 -1.213636 1.27028 0.021818 1.206818 -0.674848 -1.609848 0.192929 0.35404 -1.618182 25 ITS2 -0.574825 -0.24303 0.627273 -0.796364 -0.17303 -0.211364 0.241414 -0.636364 -0.30303 -0.347078 -0.136364 -0.709697 -0.2 -0.567133 0.363636 2.363636 -0.182517 3.490559 0.530303 2.475175 1.494886 4.188636 11.463636 -0.326364 -0.555594 -0.386364 -0.276989 3.468182 0.891414 -0.462834 0.636364 -0.174825 -0.516364 -0.111364 2.563636 -1.30303 -0.036364 0.685859 16.313636 Total BS 1 1 1 0 0 0 1 1 0 0 0 0 1 1 4 3 0 4 1 3 2 4 15 0 0 5 1 6 1 0 2 1 1 4 4 0 1 1 17 Table 3 (cont’d) Node Node 78 Node 79 Node 80 Node 81 Node 82 Node 83 Node 84 Node 85 Node 86 Node 87 Node 88 Node 89 Node 90 Node 91 Node 92 Total PBS COI 0.136364 0.609091 1.25 -1.540909 -0.390909 0.289091 0.630519 0.223377 -0.640909 1.696591 2.023797 13.956313 1.909091 2.426948 3.115758 105.993686 ND3 0.568182 8.081818 1.618182 0.431818 -0.172348 -0.238182 0.017532 0.124675 3.581818 7.419318 -1.397594 2.773485 -1.618182 -3.503896 -3.541515 24.808786 EF1-α 0.3 -1.054545 0.036364 1.245455 1.045455 -0.014545 -0.097403 0.288312 -0.054545 0.270455 -0.095722 3.739899 0.345455 0.199026 0.192121 24.690885 ITS2 2.995455 -0.636364 -0.904545 0.863636 0.517803 -0.036364 -0.550649 -0.636364 0.113636 -0.386364 -0.530481 6.530303 56.363636 9.877922 2.233636 114.506654 Total BS 4 7 2 1 1 0 0 0 3 9 0 27 57 9 2 26 Discussion This study provides the first molecular evidence of heterospecific mating among Somatochlora. Our results highlighted three instances that suggest gene flow after speciation: 1) S. hineana + S. tenebrosa; 2) S. kennedyi + S. forcipata + S. franklini; 3) S. calverti + S. provocans + S. filosa. Mitochondrial haplotypes go extinct four times faster than nuclear haplotypes, thus after a speciation event, we would expect monophyly of the mtDNA lineage if there is monophyly of the nDNA lineage (Avise, 2004). Instead, there is nonmonophyly of the mtDNA lineages which suggests mtDNA introgression due to heterospecific mating. Combined analysis of mtDNA and nDNA can resolve relationships that are not found in separate analyses (Rubinoff and Holland, 2005); however, each partition differs in its incongruence across clades within trees and in magnitude (Damgaard and Cognato, 2003). Previous studies have shown that multiple genes can interact to recover phylogenetic signal and resolve clades (Cognato et al., 2023). For unresolved nodes (BS = 0), COI provided the most support (27) while ITS2 (-17) and ND3 (-9) provided the most negative support. The large COI PBS values for unresolved nodes indicates support for non-monophyletic species while the negative values from ND3 and ITS2 indicate support for species monophyly. These partition incongruences suggest phylogenetic evidence of mitochondrial introgression, as exemplified by the positive support for unresolved nodes by mtDNA (COI) and negative support for these nodes by nDNA (ITS2). EF1-α provided the least support in resolved and unresolved nodes compared to the other genes, implying that EF1-α provided a general lack of phylogenetic information. In odonates, rapid divergence of mechanical and tactile incompatibilities lead to sexual isolation and ecological divergence (Barnard et al., 2017). Similar cerci morphology may prevent the reliable recognition of con- and heterospecifics (Paulson, 1974; Barnard et al., 2017). Somatochlora hineana and S. tenebrosa have similar cerci morphology (Williamson, 1931). Both species share the following cercal morphological traits: 1) distinctly arched shaped cerci, with a dorsal process at about midlength; 2) convergence of the cerci (obtuse for S. hineana and acute for S. tenebrosa) on the descent; 3) setae on the concave side of the cerci; 4) epiprocts approximately three-fourths the length of the cerci (Walker, 1925; Williamson, 1931). The central habitat of S. hineana is found in the Great Lakes Region with a marginal population in Missouri (Craves et al., 2022). Somatochlora tenebrosa is more common in eastern forests but can be found as far west as Missouri (Walker, 1925). Both species are partial to shady pastures 27 (Walker, 1925; Walker et al., 2020). The comparatively abundant S. tenebrosa versus the endangered S. hineana, coupled with a shared habitat range, promotes an ecological phenomenon known as the Hubbs Principle (Hubbs, 1955), where a rare species is more likely to mate with an abundant heterospecific. Our results indicate the mating of S. hineana females with S. tenebrosa males, and not the reciprocal pairing, given that S. hineana mitochondrial haplotypes were recovered from S. tenebrosa males. If the less abundant S. hineana female is confronted by more S. tenebrosa males than her conspecific males, the female may be more inclined to mate with heterospecific males. Dragonfly females ultimately decide whether to copulate, and her decision to mate with heterospecific males may be influenced by increased male harassment (Tennessen, 1982; Kornová et al., 2024). It is unknown if this asymmetrical mitochondrial gene flow, as observed for other insects (e.g. Cognato et al., 1999), is prevalent in all areas of S. hineana and S. tenebrosa sympatry. The polytomy within the mtDNA phylogeny and the monophyly within the nDNA phylogeny suggest gene flow between S. kennedyi, S. forcipata, and S. franklini. In general, this group + S. incurvata have forcipate cerci (in dorsal view) which curve inward in the distal third or beyond, with acute tips (in lateral view) (Needham et al., 2000). These species share some cercal morphological traits (Walker, 1918). Somatochlora kennedyi shares the following traits with S. franklini: 1) cerci about as long as abdominal segments 9 and 10 combined; 2) lateral carinae with a small external basal tooth, at a slight angular bend about the middle; 3) length of epiprocts slightly more than half of cerci length. Somatochlora forcipata has the same cerci length in proportion to S. kennedyi and S. franklini; however, S. forcipata has more arched cerci, and a ventral carina near the basal fourth of the cerci from which extends a large blunt ventral tooth. However, like S. kennedyi, S. forcipata has flattened apices that are bluntly pointed, with the apices of S. kennedyi turning inward at an acute angle and the apices of S. forcipata turning inward at an obtuse angle. Somatochlora kennedyi and S. franklini are distributed widely across Canada and northeastern United States, while S. forcipata and its sister species, S. incurvata, are mostly restricted to southeastern Canada and northeastern United States (www.gbif.org). The shared habitat ranges of S. kennedyi, S. franklini, S. forcipata, and S. incurvata thus lend themselves to chance encounters between this group of closely related species. Somatochlora incurvata was recovered as monophyletic and separate from S. kennedyi, S. forcipata, and S. franklini in both mtDNA and nDNA phylogenies. Considering we obtained 28 several specimens for each species (except S. forcipata) in relatively close proximity (Upper Peninsula of Michigan), it was surprising that mitochondrial introgression was not observed for S. incurvata given the species similar morphologies. In lateral view, the cerci of S. incurvata are less arched than those of S. forcipata, resembling those of S. kennedyi (Walker, 1925). Although the cerci are similar, they may act as a potential pre-mating barrier and hamper heterospecific mating among S. incurvata and its heterospecifics despite shared habitat range. Hamuli morphology also may serve as a premating barrier. The hamuli are a pair of male copulatory organs found on the venter of the second abdominal segment which contact the terminal reproductive organs of the female and function in species recognition (Watson, 1966). This group has similar morphology of the hamuli, especially between S. incurvata and S. forcipata (Walker, 1925). In general, the hamuli bend at almost a right angle, tapering abruptly to a blunt point. Local breeding habitat may better explain the monophyly of S. incurvata. Somatochlora kennedyi, S. forcipata, and S. franklini breed in fens with slow-flowing, spring-fed streams (Mead, 2021). Somatochlora incurvata breeding habitat is characterized by open sedge meadows where females prefer ovipositing in ephemeral pools (Mead, 2021; NatureServe, 2024). Consequently, S. incurvata may be found at breeding sites with comparatively lower amounts of water compared to its congeners (NatureServe, 2024), resulting in a decreased likelihood of heterospecific mating by local spatial isolation. The coastal group – S. calverti, S. provocans, and S. filosa – lacked resolution in the mtDNA phylogeny. In the nDNA phylogeny, S. filosa was the only clade of the coastal group with moderate support (87 JK). An unsupported (51 JK) clade of S. calverti was recovered sister to S. provocans. In the combined analysis, S. filosa was the only resolved clade (2 BS). The rest of the coastal group + S. margarita + S. ozarkensis were recovered in a polytomy. This lack of resolution for both mtDNA and nDNA phylogenies indicates a dearth phylogenetic signal for the resolution of these species relationships. Still, the greater resolution of the nDNA phylogeny (i.e., S. filosa resolving separate from S. calverti and S. provocans) versus the mtDNA phylogeny suggests potential mitochondrial introgression. The polytomic coastal group had specimens that were collected from the Florida panhandle, a sympatric range for the three species. This group has varied cerci morphology. Somatochlora provocans is the most dissimilar of the group, with divergent cerci that enlarge in the proximal half, converging and tapering in the distal half. Somatochlora calverti and S. filosa share the following cercal traits: 1) cerci rather close together 29 at base, curving gently inwards in proximal third and 2) subparallel along middle length, somewhat swollen (Walker, 1925; Williamson and Gloyd, 1933). The cerci of S. filosa differ from S. calverti, with S. filosa cerci obtusely curving upwards in profile (Walker, 1925), while S. calverti has a sharp lateral angulation at midlength (Williamson and Gloyd, 1933). This morphological dissimilarity suggests their cerci would serve as an effective reproductive isolating barrier, because greater species-specific cerci would allow for discrimination between con- and heterospecifics. However, S. filosa is the most abundant species of the three along the Florida coast (www.gbif.org), which may lend itself to increased harassment of heterospecific S. calverti and S. provocans females by S. filosa males. Although this study provides the first phylogenetic evidence for mitochondrial introgression among Somatochlora, heterospecific mating frequency and gene flow intensity remain unknown. In addition, the fitness consequences of introgression are not well understood. Understanding these factors is important for the conservation of the endangered S. hineana and other rare, range-restricted striped emeralds. These species may experience frequent mitochondrial and potential nuclear introgression because of increased sexual pressure to mate with heterospecifics thereby diluting the composition of genotypes. For S. hineana, the extirpation of populations from Ohio and Indiana further exacerbates this problem by severing gene flow between the Great Lakes and central US populations (Walker et al., 2020). Somatochlora kennedyi, S. forcipata, and S. franklini provide a compelling argument that cerci morphology is not an infallible prezygotic reproductive isolating barrier for this predominantly Canadian group. Quantifying the amount of gene flow between these species will provide insight into population genetic architecture and the long-term effects of introgression for Somatochlora dragonflies. There is a marked lack of ecological studies focusing on Somatochlora (except S. hineana), so although species appear sympatric, there may be small-scale habitat associations among stenotopic Somatochlora that are acting as isolating barriers (Sánchez-Guillén et al., 2012). Future evaluation of genetic variation at the genomic level will provide detailed measures of gene flow among Somatochlora as with other dragonflies (Higashikawa et al., 2023). Detailed genetic studies have indicated several cases of potential species collapse engendered by hybridization and introgression (Rhymer and Simberloff, 1996), such as with the candy darter (Gibson et al., 2019) and common raven (Kearns et al., 2018). Detailed genetic studies will deepen our understanding of the evolutionary and ecological factors that maintain cohesion of 30 Somatochlora species and provide data needed for effective decision-making for the conservation of the endangered S. hineana (Craves et al., 2022). This study reconstructs the most comprehensive phylogenies of North American Somatochlora to date. In the combined data set, most groups resolved as in Walker’s (1925) taxonomic revision of the genus. In this revision, Walker categorized North American Somatochlora into various groups according to similar morphology and distribution. There were six groups (Figure 5): 1) tenebrosa group – S. tenebrosa; 2) filosa group – S. provocans and S. filosa; 3) linearis group – S. linearis and S. ensigera; 4) metallica group – S. minor, S. elongata, S. williamsoni, and S. walshii; 5) arctica group – S. franklini, S. kennedyi, S. forcipata, S. incurvata, and S. semicircularis; 6) alpestris group – S. whitehousei, S. septentrionalis, S. sahlbergi, S. albicincta, S. hudsonica, and S. cingulata. In addition to these species, several Palearctic and more recently described species were included for phylogenetic analysis. For instance, Williamson (1931) described S. hineana and noted its close relation to S. tenebrosa. Bird (1933) and Donnelly (1962) described S. ozarkensis and S. margarita, respectively, and both authors concluded that their respective species belonged to the filosa group. Our results further corroborate Walker’s hypothesis of S. incurvata and S. forcipata as sister-species. The original description of S. brevicincta stated its close relatedness to S. albicincta (Robert, 1954). Contrary to Robert’s description, S. albicincta did not recover sister to S. brevicincta and was instead recovered sister to S. hudsonica. Instead of resolving with the arctica group like Walker (1925) predicted, S. arctica resolved with the alpestris group. Future phylogenetic study including genomic-level data and sampling of the world Somatochlora fauna will provide a more complete understanding of species relationships. 31 CHAPTER 3: CONCLUSION Somatochlora is the largest genus in the family Corduliidae whose breadth of systematic knowledge is limited. This study provides a framework for future research focused on North American Somatochlora and is the first to employ a novel dataset to reconstruct multi-gene phylogenies of the genus. There is phylogenetic evidence for heterospecific mating in the form of mitochondrial introgression among closely related Somatochlora species. However, much work needs to be done to fully understand the broad ecological impacts of long-term introgression. Firstly, future studies utilizing genomic datasets of Somatochlora can provide further insight into the extent and directionality of mitochondrial and potential nuclear introgression, as well as provide greater evidence of true heterogeneity across mitochondrial clines rather than phylogenetic artifacts associated with the use of a few molecular markers. Obtaining a greater number of specimens can lead to a more complete picture of the genetic architecture of Somatochlora in the form of measuring gene flow among the nonmonophyletic clades. The unfortunate absence of S. georgiana (coppery emerald) in this study leaves questions regarding its relationship with its congeners. The cerci of S. georgiana most closely resemble that of S. filosa, but the dull, brown, non-metallic coloration of S. georgiana may function as an adequate visual cue that prevents mating with co-occurring closely related species. Future studies can investigate if certain clades of Palearctic Somatochlora display evidence of heterospecific mating. Gene flow studies powered by genomic data can help determine what the long-term impacts of introgression may be, especially for clades of disparate population sizes (e.g., S. hineana and S. tenebrosa). Hybridization and introgression can threaten a rare species when it hybridizes with a common conspecific, a phenomenon common among animals (Rhymer and Simberloff, 1996). Phylogeographic studies would be especially informative as they can elucidate the historical range habitat of Somatochlora during the last glacial period. These types of studies would provide context for current species boundaries concerning both geographical distribution and morphological barriers. Our explanation of mito-nuclear discordance revolves around morphological differences in genitalia (i.e., cerci, hamules, epiproct) between closely related species. Detailed 32 morphometric analyses of these genitalia can quantify the architectural differences which would be useful for comparative studies of Somatochlora cerci. We used gbif.org occurrence data as a proxy for habitat range of specific Somatochlora; however, there may be fine-scale habitat differences associated with each species that may act as barriers to gene flow. At a more reductionist level, more accurate population surveys would lead to a more fine-tuned understanding of the population dynamics of this stenotopic genus, especially with climate- change-induced range fluctuations (Arce-Valdés and Sánchez-Guillén, 2022). For example, S. incurvata was previously thought to be rare, but more recent surveys revealed it was relatively common in its preferred habitat (Paulson, 2017). Previous studies have shown the caveats of solely using adult occurrences when making inferences about Odonata spatial distribution patterns, as often the breeding niches are more restricted than the niches of adults (Patten et al., 2015). Specific ethological differences may also play a role in sexual selection and could provide context for the recovered phylogenies. Indeed, S. incurvata males are noted to be aggressive in securing a mate, chasing off other striped emeralds (Mead, 2021). 33 REFERENCES Andersen, J.C., Havill, N.P., Broadley, H.J., Boettner, G.H., Caccone, A., Elkinton, J.S. 2019. Widespread hybridization among native and invasive species of Operophtera moths (Lepidoptera: Geometridae) in Europe and North America. Biological Invasions. 21, 3383-3394. doi:10.1007/s10530-019-02054-1 Arce-Valdés, L.R., Sánchez-Guillén, R.A. 2022. The evolutionary outcomes of climate-change- induced hybridization in insect populations. Current Opinion in Insect Science. 54, 100966. doi:10.1016/j.cois.2022.100966 Avise, J.C. 2004. Molecular Markers, Natural History, and Evolution, 2nd ed. ed. Sinauer Associates, Inc., Sunderland, Massachusetts. Barnard, A.A., Fincke, O.M., McPeek, M.A., Masly, J.P. 2017. Mechanical and tactile incompatibilities cause reproductive isolation between two young damselfly species. Evolution. 71, 2410–2427. doi:10.1111/evo.13315 Bergmann, T., Rach, J., Damm, S., DeSalle, R., Schierwater, B., Hadrys, H. 2013. The potential of distance-based thresholds and character-based DNA barcoding for defining problematic taxonomic entities by CO1 and ND1. Molecular Ecology Resources. 13, 1069–1081. doi:10.1111/1755-0998.12125 Bick G.H., Bick J.C. 1981. Heterospecific pairing among Odonata. Odonatologica. 10, 259–270. Blanke, A., Greve, C., Mokso, R., Beckmann, F., Misof, B. 2013. An updated phylogeny of Anisoptera including formal convergence analysis of morphological characters. Systematic Entomology. 38, 474–490. doi:10.1111/syen.12012 Bried, J.T., Mazzacano, C.A. 2010. National review of state wildlife action plans for Odonata species of greatest conservation need. Insect Conservation and Diversity. 3, 61–71. doi:10.1111/j.1752-4598.2010.00081.x Bybee, S.M., Kalkman, V.J., Erickson, R.J., Frandsen, P.B., Breinholt, J.W., Suvorov, A., Dijkstra, K.-D.B., Cordero-Rivera, A., Skevington, J.H., Abbott, J.C., Sanchez Herrera, M., Lemmon, A.R., Moriarty Lemmon, E., Ware, J.L. 2021. Phylogeny and classification of Odonata using targeted genomics. Molecular Phylogenetics and Evolution. 160, 107115. doi:10.1016/j.ympev.2021.107115 Cameron, S.L. 2014. Insect mitochondrial genomics: Implications for evolution and phylogeny. Annual Review of Entomology. 59, 95–117. doi:10.1146/annurev-ento-011613-162007 Cannings, S.G., Cannings, R.A. 1985. The larva of somatochlora sahlbergi Trybom, with notes on the species in the Yukon Territory, Canada (Anisoptera: Corduliidae. Odonatologica. 14, 319–330. 34 Carle, F., Kjer, K., May, M. 2015. A molecular phylogeny and classification of Anisoptera (Odonata). Arthropod Systematics & Phylogeny. 73, 281–301. doi:10.3897/asp.73.e31805 Chan, K.M.A., Levin, S.A. 2005. Leaky prezygotic isolation and porous genomes: rapid introgression of maternally inherited DNA. Evolution. 59, 720–729. doi:10.1111/j.0014- 3820.2005.tb01748.x Cognato, A.I., Seybold, S.J., Sperling, F.A.H. 1999. Genetic structure of Ips pini (Say) populations. United States Department of Agriculture Forest Service General Technical Report PNW. Cognato, A.I., Taft, W., Osborn, R.K., Rubinoff, D. 2023. Multi-gene phylogeny of North American clear-winged moths (Lepidoptera: Sesiidae): a foundation for future evolutionary study of a speciose mimicry complex. Cladistics. 39, 1–17. doi:10.1111/cla.12515 Craves, J., Cognato, A., O’Brien, D., Mahoney, M. 2022. A new locality and unexpected haplotypes of the federally-endangered Hine’s Emerald dragonfly, Somatochlora hineana (Odonata: Corduliidae) 13, 7–17. Damgaard, J., Cognato, A.I. 2003. Sources of character conflict in a clade of water striders (Heteroptera: Gerridae). Cladistics. 19, 512–526. doi:10.1111/j.1096- 0031.2003.tb00386.x Dunkle, S. 2004. Critical species of Odonata in North America. Worldwide Dragonfly Association. 7, 149–162. Futahashi, R., Yamahama, Y., Kawaguchi, M., Mori, N., Ishii, D., Okude, G., Hirai, Y., Kawahara-Miki, R., Yoshitake, K., Yajima, S., Hariyama, T., Fukatsu, T. 2019. Molecular basis of wax-based color change and UV reflection in dragonflies. eLife. 8, e43045. doi:10.7554/eLife.43045 Gibson, I., Welsh, A.B., Welsh, S.A., Cincotta, D.A. 2019. Genetic swamping and possible species collapse: tracking introgression between the native Candy Darter and introduced Variegate Darter. Conservation Genetics. 20, 287–298. doi:10.1007/s10592-018-1131-2 Gröning, J., Hochkirch, A. 2008. Reproductive interference between animal species. The Quarterly Review of Biology. 83, 257–282. doi:10.1086/590510 Higashikawa, W., Yoshimura, M., Nagano, A.J., Maeto, K. 2023. Conservation genomics of an endangered floodplain dragonfly, Sympetrum pedemontanum elatum (Selys), in Japan. Conservation Genetics. doi:10.1007/s10592-023-01595-2 Hubbs, C.L. 1955. Hybridization between fish species in nature. Systematic Biology. 4, 1–20. doi:10.2307/sysbio/4.1.1 35 Kearns, A.M., Restani, M., Szabo, I., Schrøder-Nielsen, A., Kim, J.A., Richardson, H.M., Marzluff, J.M., Fleischer, R.C., Johnsen, A., Omland, K.E. 2018. Genomic evidence of speciation reversal in ravens. Nature Communications. 9, 906. doi:10.1038/s41467-018- 03294-w Kohli, M.K., Sahlén, G., Kuhn, W.R., Ware, J.L. 2018. Extremely low genetic diversity in a circumpolar dragonfly species, Somatochlora sahlbergi (Insecta: Odonata: Anisoptera). Scientific Reports. 8, 15114. doi:10.1038/s41598-018-32365-7 Kornová, V., Bílková, E., Pyszko, P., Dolný, A., Ožana, S. 2024. What determines mate choices? Heterospecific mating in Sympetrum dragonflies. Freshwater Biology. n/a. doi:10.1111/fwb.14226 Madeira, F., Pearce, M., Tivey, A.R.N., Basutkar, P., Lee, J., Edbali, O., Madhusoodanan, N., Kolesnikov, A., Lopez, R. 2022. Search and sequence analysis tools services from EMBL-EBI in 2022. Nucleic Acids Research. 50, W276–W279. doi:10.1093/nar/gkac240 McPeek, M.A., Shen, L., Torrey, J.Z., Farid, H. 2008. The Tempo and Mode of Three‐ Dimensional Morphological Evolution in Male Reproductive Structures.. The American Naturalist. 171, E158–E178. doi:10.1086/587076 Mead, K. 2021. Dragonflies of the North Woods, 3rd ed. Kollath-Stensaas Pub., Duluth, Minnesota. Misof, B., Liu, S., Meusemann, K., Peters, R.S., Donath, A., Mayer, C., Frandsen, P.B., Ware, J., Flouri, T., Beutel, R.G., Niehuis, O., Petersen, M., Izquierdo-Carrasco, F., Wappler, T., Rust, J., Aberer, A.J., Aspöck, U., Aspöck, H., Bartel, D., Blanke, A., Berger, S., Böhm, A., Buckley, T.R., Calcott, B., Chen, J., Friedrich, F., Fukui, M., Fujita, M., Greve, C., Grobe, P., Gu, S., Huang, Y., Jermiin, L.S., Kawahara, A.Y., Krogmann, L., Kubiak, M., Lanfear, R., Letsch, H., Li, Yiyuan, Li, Z., Li, J., Lu, H., Machida, R., Mashimo, Y., Kapli, P., McKenna, D.D., Meng, G., Nakagaki, Y., Navarrete-Heredia, J.L., Ott, M., Ou, Y., Pass, G., Podsiadlowski, L., Pohl, H., von Reumont, B.M., Schütte, K., Sekiya, K., Shimizu, S., Slipinski, A., Stamatakis, A., Song, W., Su, X., Szucsich, N.U., Tan, M., Tan, X., Tang, M., Tang, J., Timelthaler, G., Tomizuka, S., Trautwein, M., Tong, X., Uchifune, T., Walzl, M.G., Wiegmann, B.M., Wilbrandt, J., Wipfler, B., Wong, T.K.F., Wu, Q., Wu, G., Xie, Y., Yang, S., Yang, Q., Yeates, D.K., Yoshizawa, K., Zhang, Q., Zhang, R., Zhang, W., Zhang, Yunhui, Zhao, J., Zhou, C., Zhou, L., Ziesmann, T., Zou, S., Li, Yingrui, Xu, X., Zhang, Yong, Yang, H., Wang, Jian, Wang, Jun, Kjer, K.M., Zhou, X. 2014. Phylogenomics resolves the timing and pattern of insect evolution. Science. 346, 763–767. doi:10.1126/science.1257570 Monroe, E.M., Britten, H.B. 2014. Conservation in Hine’s sight: the conservation genetics of the federally endangered Hine’s emerald dragonfly, Somatochlora hineana. Journal of Insect Conservation. 18, 353–363. doi:10.1007/s10841-014-9643-7 NatureServe. 2024. Somatochlora incurvata | NatureServe Explorer [WWW Document]. NatureServe Network Biodiversity Location Data. URL 36 https://explorer.natureserve.org/Taxon/ELEMENT_GLOBAL.2.109311/Somatochlora_in curvata (accessed 4.4.24). Needham, J.G., Westfall, M.J., May, M.L., Needham, J.G. 2000. Dragonflies of North America, Rev. ed. Scientific Publishers, Gainesville, FL. Patten, M.A., Bried, J.T., Smith-Patten, B.D. 2015. Survey data matter: predicted niche of adult vs breeding Odonata. Freshwater Science. 34, 1114–1122. doi:10.1086/682676 Paulson, D.R. 2017. IUCN Red List of Threatened Species: Somatochlora incurvata [WWW Document]. IUCN Red List of Threatened Species. URL https://www.iucnredlist.org/en (accessed 3.28.24). Paulson, D.R. 1974. Reproductive isolation in Damselflies. Systematic Biology. 23, 40–49. doi:10.1093/sysbio/23.1.40 Pilgrim, E.M., Von Dohlen, C.D. 2008. Phylogeny of the Sympetrinae (Odonata: Libellulidae): further evidence of the homoplasious nature of wing venation. Systematic Entomology. 33, 159–174. doi:10.1111/j.1365-3113.2007.00401.x Pintor, L.M., Soluk, D.A. 2006. Evaluating the non-consumptive, positive effects of a predator in the persistence of an endangered species. Biological Conservation. 130, 584–591. doi:10.1016/j.biocon.2006.01.021 Rhymer, J.M., Simberloff, D. 1996. Extinction by hybridization and introgression. Annual Review of Ecology, Evolution, and Systematics. 27, 83–109. doi:10.1146/annurev.ecolsys.27.1.83 Robert, F.A. 1954. Un nouveau Somatochlora subarctique (Odonates, Corduliidae). The Canadian Entomologist. 86, 419–422. doi:10.4039/Ent86419-9 Rubinoff, D., Holland, B.S. 2005. Between two extremes: Mitochondrial DNA is neither the panacea nor the nemesis of phylogenetic and taxonomic inference. Systematic Biology. 54, 952–961. doi:10.1080/10635150500234674 San Jose, M., Doorenweerd, C., Rubinoff, D. 2023. Genomics reveals widespread hybridization across insects with ramifications for species boundaries and invasive species. Current Opinion in Insect Science. 58, 101052. doi:10.1016/j.cois.2023.101052 Sánchez-Guillén, R.A., Wellenreuther, M., Cordero Rivera, A. 2012. Strong asymmetry in the relative strengths of prezygotic and postzygotic barriers between two damselfly sister species. Evolution. 66, 690–707. doi:10.1111/j.1558-5646.2011.01469.x Schröter, A., Schneider, T., Schneider, E., Karjalainen, S., Hämäläinen, M. 2012. Observations on adult Somatochlora sahlbergi – a species at risk due to regional climate change? (Odonata: Corduliidae). Libellula. 31, 41–60. 37 Solano, E., Hardersen, S., Audisio, P., Amorosi, V., Senczuk, G., Antonini, G. 2018. Asymmetric hybridization in Cordulegaster (Odonata: Cordulegastridae): Secondary postglacial contact and the possible role of mechanical constraints. Ecology and Evolution. 8, 9657–9671. doi:10.1002/ece3.4368 Sorenson, M.D., 1999. TreeRot, Version 2. Boston University, Boston, MA. Spoelstra, J., Post, R. 2023. Groundwater characterization of the eastern Minesing Wetlands in support of the endangered Hine’s emerald dragonfly (Somatochlora hineana). Wetlands Ecology and Management. 31, 309–327. doi:10.1007/s11273-023-09918-3 Suvorov, A., Scornavacca, C., Fujimoto, M.S., Bodily, P., Clement, M., Crandall, K.A., Whiting, M.F., Schrider, D.R., Bybee, S.M. 2022. Deep Ancestral Introgression Shapes Evolutionary History of Dragonflies and Damselflies. Systematic Biology. 71, 526–546. doi:10.1093/sysbio/syab063 Swofford, D.L. 2002. PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods). Tennessen, K.J. 1982. Review of reproductive isolating barriers in Odonata. Advances in Odonatology. 1, 251–265. Troast, D., Suhling, F., Jinguji, H., Sahlén, G., Ware, J. 2016. A global population genetic study of Pantala flavescens. PLOS ONE. 11, e0148949. doi:10.1371/journal.pone.0148949 U.S. Fish and Wildlife Service. 2001. Hine’s Emerald Dragonfly (Somatochlora hineana) Recovery Plan. Vogt, T.E., Cashatt, E.D. 1994. Distribution, habitat, and field biology of Somatochlora hineana (Odonata: Corduliidae). Annals of the Entomological Society of America. 87, 599–603. doi:10.1093/aesa/87.5.599 Walker, E.M. 1925. The North American Dragonflies of the Genus Somatochlora. University of Toronto. Walker, J., Mahoney, M., Templeton, A.R., McKenzie, P., Vogt, T.E., Cashatt, E.D., Smentowski, J., Day, R., Gillespie, R., Henry, B., Wiker, J., Braude, S., Landwer, B. 2020. Contrasting Ozark and Great Lakes populations in the endangered Hines emerald dragonfly (Somatochlora hineana) using ecological, genetic, and phylogeographic analyses. Conservation Science and Practice. 2, e162. doi:10.1111/csp2.162 Watson, J.A.L. 1966. Genital structure as an isolating mechanism in Odonata. Proceedings of the Royal Entomological Society of London. Series A, General Entomology. 41, 171–174. doi:10.1111/j.1365-3032.1966.tb00338.x Weixelman, D.A., Cooper, D.J. 2009. Assessing Proper Functioning Condition for Fen Area in the Sierra Nevada and Southern Cascade Ranges in California. US Department of Agriculture, US Forest Service. doi:10.13140/RG.2.1.3606.0246 38 Williamson, E.B. 1931. A new north american Somatochlora (Odonata-Cordulinae). doi:10.1126/science.46.1200.643 Williamson, E.B., Gloyd, L.K. 1933. A new Somatochlora from Florida (Odonata-Cordulinae). doi:10.1126/science.46.1200.643 Yao, H., Song, J., Liu, C., Luo, K., Han, J., Li, Y., Pang, X., Xu, H., Zhu, Y., Xiao, P., Chen, S. 2010. Use of ITS2 region as the universal DNA barcode for plants and animals. PLOS ONE. 5, e13102. doi:10.1371/journal.pone.0013102 Zhang, D.-X., Hewitt, G.M. 2003. Nuclear DNA analyses in genetic studies of populations: practice, problems and prospects. Molecular Ecology. 12, 563–584. doi:10.1046/j.1365- 294X.2003.01773.x 39