Ana Luiza Anes Pimenta

Universidade Federal do Rio de Janeiro

Filogenia de Protoneurinae e taxonomia integrativa de Forcepsioneura Lencioni, 1999 (Odonata: Zygoptera: Coenagrionidae)

Rio de Janeiro

2019 II

III

TESE DESENVOLVIDA NO LABORATÓRIO DE ENTOMOLOGIA, DEPARTAMENTO DE ZOOLOGIA, INSTITUTO DE BIOLOGIA DA UNIVERSIDADE FEDERAL DO RIO DE JANEIRO

Sob orientação Profª. Drª. Daniela Maeda Takiya e Prof. Dr. Ângelo Parise Pinto

FICHA CATALOGRÁFICA

PIMENTA, Ana Luiza Anes. Filogenia de Protoneurinae e taxonomia integrativa de Forcepsioneura Lencioni, 1999 (Odonata: Zygoptera: Coenagrionidae) - Ana Luiza Anes Pimenta - Rio de Janeiro: UFRJ, 2018. pp. I-XIX, 1-95; 29,7 cm. Orientador: Profª. Drª. Daniela Maeda Takiya Coorientador: Dr. Ângelo Parise Pinto Tese (doutorado) – UFRJ/ Programa de Pós-graduação em Biodiversidade e Biologia Evolutiva, 2019 Referências Bibliográficas: 25–29; 68–71; 93–95. 1. Evolução. 2. Sistemática filogenética. 3. Morfologia. 4. Hipótese de evidência total. 5. Taxonomia. I. Takiya, Daniela Maeda II. Pinto, Ângelo Parise III. Universidade Federal do Rio de Janeiro, Programa de Pós-graduação em Biodiversidade e Biologia Evolutiva IV. Filogenia de Protoneurinae e taxonomia integrativa de Forcepsioneura Lencioni, 1999 (Odonata: Zygoptera: Coenagrionidae)

AGRADECIMENTOS

Sem amigos não somos ninguém. Portanto, gostaria de ressaltar algumas pessoas que a universidade me presenteou com amizades tão especiais. Elas que me ajudaram, me deram forças quando tudo estava quase perdido, inclusive a esperança (que ia e voltava), me lembravam de como todas nós importamos e que temos uma força absurda dentro de nós para continuar respirando, são elas: Beatriz Camisão, Juliana Kirchmeyer, Manuella Folly e Tainá Stauffer. Além dessas mulheres lindas, agradeço também ao meu amigo Pedro Fasura por ter me aturado pelos corredores do CCS. Agradeço também aos meus pais, que conviveram diariamente com o meu mau humor, estresse e falta de tempo pra uma atenção, uma palavra, um telefonema e que sempre me davam uma palavra positiva de que tudo iria dar certo. A todos os meus amigos que ao longo da vida fui encontrando e que me deram muita força em todos os momentos difíceis que eu vivi fora do doutorado. Muitas vezes deixei de estar com eles para poder terminar alguma parte da tese, alguma análise, trecho ou resultado. Sem amigos não somos ninguém (2x). Aos amigos do LABENT por todo o convívio, muitas vezes maçante, dando mais leveza para o estresse do dia-a-dia. Obrigada por toda a ajuda, sugestões, conversas e apoio. Aos meus orientadores Daniela Takiya e Ângelo Pinto, por aceitarem essa empreitada de 4 anos e não terem desistido de mim, apesar de eu ter merecido algumas vezes. Agradeço também todo o apoio logístico, financeiro e intelectual. E por fim, ao Programa de Pós-graduação em Biodiversidade e Biologia Evolutiva e a CAPES por fornecerem minha bolsa de estudos para a realização desta tese. VI

RESUMO

Protoneurinae (Protoneuridae s.s.) é composta por 122 espécies neotropicais com tamanho corporal e venação alar reduzidos. Análises filogenéticas baseadas tanto em dados morfológicos como moleculares, sustentam seu monofiletismo, no entanto, o relacionamento com outras subfamílias de Coenagrionidae permanece incerto. Apesar do baixo suporte, uma análise filogenética evidencia o clado Roppaneura, composto por seis gêneros. O relacionamento entre esses gêneros é inconclusivo e isso pode ser explicado em parte por diagnoses frágeis devido a caracteres não exclusivos. Forcepsioneura, um dos gêneros desse clado, inclui oito espécies muito semelhantes morfologicamente e de difícil determinação. A presente tese visou contribuir para o entendimento da taxonomia, filogenia e evolução de Protoneurinae, sendo dividida em dois capítulos. O primeiro inclui (1) descrição de duas espécies novas de Forcepsioneura, com o auxílio de uma abordagem integrativa, explorando a distância genética de três marcadores moleculares (COI mtDNA, 16S rDNA, e PRMT nDNA); e a (2) primeira hipótese de relacionamento filogenético do gênero com base em dados moleculares utilizando os mesmos marcadores. O segundo capítulo inclui a (1) primeira hipótese filogenética de Protoneurinae por meio de uma abordagem integrativa, utilizando caracteres morfológicos e três marcadores moleculares mitocondriais e nucleares; e a (2) proposta do relacionamento interno do clado Roppaneura baseada em evidência total. No Capítulo 1, foram descritas duas novas espécies de Forcepsioneura: Forcepsioneura gabriela sp. nov. proximamente relacionada a F. garrisoni Lencioni, 1999 e F. regua Pinto & Kompier, 2018 (grupo azul), e Forcepsioneura janeae sp. nov. proximamente relacionada à F. lucia Machado, 2000. Uma análise Bayesiana concatenada de dados moleculares inéditos utilizando sete das 10 espécies de Forcepsioneura foi realizada, e recuperou o seu monofiletismo. Esse estudo ressaltou a importância do uso de distâncias genéticas de sequências de COI a nível específico, porém não sustenta o uso do PRMT e 16S para esse grupo de Odonata a este nível. No Capítulo 2, a filogenia de Protoneurinae foi proposta utilizando 49 táxons terminais e seis famílias de Zygoptera. Dos 15 gêneros de Protoneurinae, 10 foram amostrados e 23 espécies terminais. O conjunto de dados reuniu três marcadores moleculares: COI, 16S rDNA e 18S rDNA e 46 caracteres morfológicos . Análises cladísticas com diferentes pesagens de caracteres e explorando diferentes valores de k, Máxima Verossimilhança e Inferência bayesiana foram conduzidas, explorando tanto os dados combinados quanto VII

somente moleculares. O monofiletismo de Protoneurinae foi recuperado em todas as árvores. A hipótese com maior resolução interna foi a obtida com a parcimônia que, apesar de não ter obtido altos suportes, os clados foram recuperados por análises com diferentes valores de k. O clado Roppaneura, observado em análises morfológicas anteriores, foi recuperado apenas na análise de parcimônia, enquanto Forcepsioneura é suportado como monofilético em todas as análises.

Palavras-chave: Evolução, Hipótese de evidência total, Libélulas, Morfologia, Sistemática filogenética, Taxonomia

VIII

ABSTRACT

Protoneurinae (Protoneuridae s.s.) is composed by 122 Neotropical species with reduced body size and venation. Phylogenetic analysis based on both morphological and molecular datasets recover the monophyly of the subfamily, however, the relationship with other Coenagrionidae subfamilies still uncertain. A recent phylogenetic analysis showed the Roppaneura clade, composed for six genera, but low supported. The relationships among these genera are inconclusive and can be explained, in parts, by fragile diagnosis due to non-exclusive characters. One of the genera of this clade, Forcepsioneura, includes eight species very morphologically similar and difficult to identify. This study aimed to contribute to the understanding of the taxonomy, phylogeny and evolution of Protoneurinae, being divided into two chapters. The first includes (1) description of two new species of Forcepsioneura, with the aid of an integrative approach, exploring the genetic distance of three molecular markers (COI mtDNA, 16S rDNA, e PRMT nDNA); and (2) the first molecular phylogenetic hypothesis of the genus using the same molecular markers. The second chapter includes: (1) the first phylogenetic hypothesis of Protoneurinae through an integrative approach using morphological characters and three molecular markers (mitochondrials and nuclear); and (2) the proposal of the internal relationship of the Roppaneura clade based on total evidence. In Chapter 1, two new species of Forcepsioneura were described: Forcepsioneura gabriela sp. nov. was more related to F. garrisoni Lencioni, 1999 and F. regua Pinto & Kompier, 2018 (light-blue group), and Forcepsioneura janeae sp. nov. was more related to F. lucia Machado, 2000. The combined Bayesian analysis of unpublished molecular data using seven of the 10 Forcepsioneura species was performed, and recovered the monophyly of the genus. This study highlighted the importance of using genetic distances from COI sequences at a specific level, but does not support the use of the PRTM and 16S markers for this group of Odonata in this level. In Chapter 2, Protoneurinae phylogeny was proposed using 49 terminal taxa and six Zygoptera families. Of the 15 genera of Protoneurinae, 10 were sampled and 23 terminal taxa. The datasets gathered three molecular markers: COI, 16S rDNA and 18S rDNA and 46 morphological characters. Cladistic analyses with different character weights and exploring different k values, maximum likelihood and Bayesian inference were conducted, exploring both combined and molecular dataset only. The monophyly of Protoneurinae was recovered in all trees. The hypothesis with higher resolution IX

hypothesis was the one obtained with parsimony, that, despite not having obtained high bootstrap values, the clades were recovered by analysis with different k values. The Roppaneura clade, a group with six genera observed in previous morphological analyses, was recovered as monophyletic only in parsimony analyses, while Forcepsioneura is supported as monophyletic in all analysis. X

Sumário

FICHA CATALOGRÁFICA IV AGRADECIMENTOS V RESUMO VI ABSTRACT VIII ÍNDICE DE FIGURAS XIII FIGURAS SUPLEMENTARES XV ÍNDICES DE TABELAS XVII TABELAS SUPLEMENTARES XVIII 1. Introdução 1 1.1 Ordem Odonata 1 1.2 Coenagrionidae 2 1.3 Protoneuridae s.l. 3 1.4 Protoneurinae 4 1.5 Clado Roppaneura 5 1.6 Forcepsioneura 5 1.7 Taxonomia integrativa 7 2. Objetivos 8 Capítulo 1 9 ABSTRACT 9 1. INTRODUCTION 11 2. MATERIAL AND METHODS 12 2.1. Taxon sampling 12 2.2. DNA extraction, amplification and sequencing 13 2.3. Genetic distances and phylogenetic analysis 13 2.4. Morphological analysis and terminology 14 3. RESULTS 14 3.1. Genetic distances 14 3.2. Proposal of new taxa 15 3.3. Taxonomy 15 3.3.1. Forcepsioneura gabriela sp.n. 15 3.3.2. Forcepsioneura janeae sp.n. 19 XI

3.4. Phylogeny of Forcepsioneura 22 4. DISCUSSION 22 4.1. Species delimitation 22 4.1.1. Morphological data 22 4.1.2. Genetic distances 22 4.2. Forcepsioneura phylogeny 23 5. CONCLUSIONS 24 6. ACKNOWLEDGEMENTS 24 7. REFERENCES 25 FIGURE FILES 31 TABLES FILES 40 SUPPLEMENTARY MATERIAL 46 FIGURES 46 TABELS 49 Capítulo 2 52 ABSTRACT 52 1. INTRODUCTION 53 2. MATERIAL AND METHODS 55 2.1 Taxon sampling 55 2.2 DNA extraction, purification and sequencing 56 2.3 Phylogenetic analyses 57 2.3.1 Cladistic analysis 57 2.3.2 Maximum likelihood analysis 58 2.3.3 Bayesian Inference analysis 59 2.4 Selection of preferred hypothesis of relationship 59 3. RESULTS 60 3.1 Morphological data 60 3.2 Datasets for phylogenetic analyses 63 3.3 Cladistic analysis 63 3.4 Bayesian Inference analysis 64 3.5 Maximum likelihood analysis 64 3.6 Morphological character optimizations 65 4. DISCUSSION 65 4.1 Coenagrionidae 65 XII

4.2 Position of Protoneurinae 66 4.3 Internal relationships of Protoneurinae and Roppaneura clade 67 5. CONCLUSIONS 67 6. REFERENCES 68 FIGURE FILES 72 TABLE FILES 76 SUPPLEMENTARY MATERIALS 81 FIGURES 81 TABLE FILE 84 List of specimens of Protoneurinae and outgroups studied for morphological character coding and DNA extractions. 86 3. Conclusões Gerais 92 4. Referências gerais 93

XIII

ÍNDICE DE FIGURAS

Figura 1. Representação das hipóteses filogenéticas de Odonata. Retirada de Kim et al. (2014). A–F. hipóteses baseadas em caracteres morfológicos, G –J. hipóteses baseadas em dados moleculares, K–L. hipóteses baseadas em dados moleculares e morfológicos. Subordens indicadas entre parênteses (A, Anisoptera; Z, Zygoptera e AZ, Anisozygoptera). Pontos acima dos nós indicam o monofiletismo das subordens...... 2

Capítulo 1. INTEGRATIVE TAXONOMY AND PHYLOGENY OF THE DAMSELFLY GENUS FORCEPSIONEURA (ODONATA: COENAGRIONIDAE: PROTONEURINAE) WITH DESCRIPTION OF TWO NEW SPECIES FROM THE BRAZILIAN ATLANTIC FOREST

Figure 1. Neighbor-joining tree based on K2P distances of COI sequences of Forcepsioneura and outgroup taxa. Node-associated values refer to bootstrap percentages higher than 50%...... 31

Figure 2. Neighbor-joining tree based on uncorrected distances of 16S sequences of Forcepsioneura and outgroup taxa. Node-associated values refer to bootstrap percentages higher than 50%...... 32

Figure 3. Neighbor-joining tree based on uncorrected distances of PRMT sequences of Forcepsioneura and outgroup taxa. Node-associated values refer to bootstrap percentages higher than 50%...... 33

Figure 4. Bayesian post-burn-in consensus of mixed-model analysis of the concatenated dataset (COI, 16S, and PRMT) of Forcepsioneura and outgroup taxa. Node-associated values are posterior probabilities higher than 50%...... 34

Figure 5. Forcepsioneura gabriela sp.n. holotype ♂ (Brazil, Bahia: Reserva Biológica de Una, DZRJ 3555). A: head, dorsal view; B: prothorax, lateral view. Forcepsioneura garrisoni ♂ (Brazil, Rio de Janeiro: Praia de Tarituba, DZRJ 0325). C: head, dorsal view; D: prothorax, dorsolateral view. – Scale bars: 1 mm...... ………………………35 XIV

Figure 6. Caudal appendages of Forcepsioneura gabriela sp.n. holotype ♂ (Brazil, Bahia: Reserva Biológica de Una, DZRJ 3555) (A–C) and Forcepsioneura garrisoni (Brazil, Rio de Janeiro: Praia de Tarituba, DZRJ 0325) (D–F). A, D: lateral view; B, E: dorsolateral view; C, F: posterior view. – Scale bars: 1 mm…………………………36

Figure 7. A–D, Forcepsioneura janeae sp.n. holotype ♂ (Brazil, Espírito Santo: Estação Biológica de Santa Lúcia, DZUP 499056). A: head, dorsal view; B: prothorax, lateral view; C: caudal appendages, lateral view; D: caudal appendages, dorsolateral view. E–H, Forcepsioneura lucia paratype ♂ (Brazil, Minas Gerais, Parque Estadual do Rola Moça, DZUP 499902); E: prothorax, dorsal view; F: prothorax, lateral view; G: caudal appendages, lateral view; H: caudal appendages, dorsolateral view. – Scale bars: 1 mm...... 37

Figure 8. Genital ligula of Forcepsioneura species. Forcepsioneura gabriela sp.n. holotype ♂ (Brazil, Bahia: Reserva Biológica de Una, DZRJ 3555) in ventral (A) and lateral (B) views. Forcepsioneura janeae sp.n. paratype ♂ (Brazil, Espírito Santo: Estação Biológica de Santa Lúcia, MNRJ 0141) in ventral (C) and lateral (D) views. – Scale bars: 1 mm...... 38

Figure 9. Habitus and habitat of Forcepsioneura species.: Forcepsioneura gabriela sp.n. A: holotype ♂ (Brazil, Bahia: Reserva Biológica de Una, DZUP 498858). B: type locality. C: paratype ♂ (same as holotype,DZRJ 3555). D: Forcepsioneura janeae sp.n. type locality at Brazil, Espírito Santo: Estação Biológica de Santa Lúcia. Photos: (A–C) Ângelo P. Pinto, (D) Jane P. E. Buss...... 39

Capítulo 2. PHYLOGENY OF PROTONEURINAE DAMSELFLIES (ODONATA: COENAGRIONIDAE) BASED ON MORPHOLOGICAL AND MOLECULAR DATA

Figure 1. Strict consensus of 5 most parsimonious trees obtained with implied weighting (Reference K = 3) of the combined datasets of Protoneurinae and outgroups (Fit = 240.969; L = 3,313; CI = 0.27; and RI = 0.40). Rectangles below the nodes correspond to the value of k where the node is also obtained, according to the legend. Values in the branches correspond respectively to absolute Bootrstap, Jackknife and Bremer above 50%...... 72 XV

Figure 2. Bayesian consensus of the combined morphological and molecular datasets of Protoneurinae and outgroups, using the GTR+I+G for COI and 16S partitions and the HKY+I+G model for 18S. Numbers above branches are posterior probabilities higher than 50%...... 73

Figure 3. Maximum likelihood tree (lnL= -14528.01391) found for the combined morphological and molecular datasets of Protoneurinae and outgroups, using the GTR+I+G for COI and 16S partitions and the HKY+I+G model for 18S. Numbers above branches are bootstrap values above 50%..……………………………..………74

Figure 4. Most parsimonious tree found with implied weighting (K=3) with morphological dataset of Protoneurinae (Length=1070, CI=0.172 and RI=0.425) showing character state changes……………………………………………………….75

FIGURAS SUPLEMENTARES

Capítulo 1. INTEGRATIVE TAXONOMY AND PHYLOGENY OF THE DAMSELFLY GENUS FORCEPSIONEURA (ODONATA: COENAGRIONIDAE: PROTONEURINAE) WITH DESCRIPTION OF TWO NEW SPECIES FROM THE BRAZILIAN ATLANTIC FOREST

Figure S1. Bayesian post-burn-in consensus of COI dataset of Forcepsioneura and outgroup taxa. Node-associated values are posterior probabilities higher than 50%...... 46

Figure S2. Bayesian post-burn-in consensus of 16S dataset of Forcepsioneura and outgroup taxa. Node-associated values are posterior probabilities higher than 50%...... 47

Figure S3. Bayesian post-burn-in consensus of PRMT dataset of Forcepsioneura and outgroups. Node-associated values are posterior probabilities higher than 50%...... 48

Capítulo 2. PHYLOGENY OF PROTONEURINAE DAMSELFLIES (ODONATA: COENAGRIONIDAE) BASED ON MORPHOLOGICAL AND MOLECULAR DATA

XVI

Figure S1. Most parsimonious tree obtained with implied weighting of the molecular datasets of Protoneurinae and outgroups (Fit = 219.731291, L = 3.067, CI = 0.273, RI = 0.391). Numbers above branches are bootstrap values above 50…………………….81

Figure S2. Bayesian consensus of molecular datasets of Protoneurinae and outgroups, using the GTR+I+G for COI and 16S partitions and the HKY+I+G model for 18S. Numbers above branches are posterior probabilities higher than 50%...... 82

Figure S3. Maximum likelihood tree found for the molecular datasets of Protoneurinae and outgroups, using the GTR+I+G for COI and 16S partitions and the HKY+I+G model for 18S. Numbers above branches are bootstrap values above 50%...... 83

XVII

ÍNDICES DE TABELAS

Capítulo 1. INTEGRATIVE TAXONOMY AND PHYLOGENY OF THE DAMSELFLY GENUS FORCEPSIONEURA (ODONATA: COENAGRIONIDAE: PROTONEURINAE) WITH DESCRIPTION OF TWO NEW SPECIES FROM THE BRAZILIAN ATLANTIC FOREST

Table 1. Species included in the phylogenetic analysis of the genus Forcepsioneura with voucher specimen code and collection locality in Brazil, and GenBank accession codes for molecular markers sequenced in this study (except one marked with *).………………………………...………………………………………….………….40 Table 2. Range (and mean) of intraspecific genetic variation of the three molecular markers (COI, 16S, and PRMT) sequenced for Forcepsioneura, including numbers of individuals (N) analyzed……………………………………………..…………………41

Table 3. Range (and mean) of interspecific K2P distances between COI sequences of Forcepsioneura species………………………….…………………………….……….41

Table 4. Range (and mean) of uncorrected interspecific distances between 16S sequences of Forcepsioneura species……………..……………………………………41

Table 5. Range (and mean) of uncorrected interspecific distances between PRMT sequences of Forcepsioneura species…………………………………………………..42

Table 6. Diagnostic morphological characteristics for species of Forcepsioneura…….43

Table 7. Intraspecific K2P distances of COI sequences of some Zygoptera species published in the literature. * denotes uncorrected values…………….…………………………………………………….……………….45

XVIII

Capítulo 2. PHYLOGENY OF PROTONEURINAE DAMSELFLIES (ODONATA: COENAGRIONIDAE) BASED ON MORPHOLOGICAL AND MOLECULAR DATA

Table 1. Representatives of Protoneurinae and outgroups included in the molecular and morphological datasets. Molecular markers with bold X were sequenced herein, and those sequences marked with an asterisk refer from which similarly marked specimen voucher they were obtained……………………………………………………………76

Table 2. Primers used for amplification and sequencing of molecular markers for the phylogeny of Protoneurinae……………………………………………………………78

Table 3. Summary of values of parameters obtained with tested values of k of the parsimony analysis with implied weighting of the combined dataset. F: fit intervals; K: concavity constant; L: length of the most parsimonious trees found; T: number of most parsimonious trees found; Fit: values of fit obtained; CI: consistency index; RI: retention index. Values in bold represent the selected K value……….……………….79

TABELAS SUPLEMENTARES

Capítulo 1. INTEGRATIVE TAXONOMY AND PHYLOGENY OF THE DAMSELFLY GENUS FORCEPSIONEURA (ODONATA: COENAGRIONIDAE: PROTONEURINAE) WITH DESCRIPTION OF TWO NEW SPECIES FROM THE BRAZILIAN ATLANTIC FOREST

Table S1. K2P pairwise distances of COI sequences of Forcepsioneura and outgroup taxa. Intraspecific distances of F. aff. lucia in gray, F. sancta in blue, F. garrisoni in red, and F. gabriela sp.n. in green …………………………………………...……...... 49

XIX

Table S2. Uncorrected genetic distances (p-distance) of 16S sequences of Forcepsioneura and outgroup taxa. Intraspecific distances of F. aff. lucia in gray, F. sancta in blue, F. regua in orange, and F. gabriela sp.n. in green.…….……..………..50

Table S3. Uncorrected genetic distances (p-distance) of the nuclear gene PRMT of Forcepsioneura and outgroup taxa. Intraspecific distances of F. aff. lucia in gray, F. sancta in blue, F. garrisoni in red, and F. gabriela sp.n. in green………………..……51

Capítulo 2. PHYLOGENY OF PROTONEURINAE DAMSELFLIES (ODONATA: COENAGRIONIDAE) BASED ON MORPHOLOGICAL AND MOLECULAR DATA

Table S1. Morphological data matrix of Protoneurinae and outgroups with 49 terminals, and 46 characters. ? = missing character...... 84

1. Introdução

1.1 Ordem Odonata

A ordem Odonata Fabricius, 1793 pertence a uma das linhagens de insetos alados mais antigas, datada de cerca de 250 milhões de anos, no período Permiano (Grimaldi & Engel 2005). Compreende cerca de 6.300 espécies (Schorr & Paulson 2019) ocorre em todas as regiões biogeográficas, com destaque para a Região Neotropical a qual abriga cerca de 30% de toda a diversidade de libélulas conhecidas (von Ellenrieder 2009). Caracterizam-se por serem insetos anfibióticos, ambas as fases de vida são predadoras e geralmente possuem um voo preciso e veloz (Corbet 1980). Suas espécies são distribuídas em três subordens: Anisoptera (cosmopolita, ca. 3000 spp.), Zygoptera (cosmopolita, ca. 3000 spp.) e Anisozygoptera (oriental, 4 spp.) (Dijkstra et al. 2013). O corpo mais robusto, olhos que ocupam dorsalmente grande parte da região da cabeça e asas posteriores alargadas na base diferem morfologicamente as espécies de Anisoptera de Zygoptera, que por sua vez, apresentam corpo esguio, cabeça alongada transversalmente ao corpo, olhos amplamente separados (distância entre os olhos ≥ 2x a largura de um olho), asas anteriores e posteriores similares e maioria pecioladas. As quatro espécies de Anisozygoptera apresentam características intermediárias e similares as duas subordens acima: corpo robusto e asas similares e pecioladas. O relacionamento filogenético das subordens já foi amplamente explorado (Figura 1). Atualmente, com o avanço de novas metodologias e o uso de dados moleculares, o monofiletismo das três subordens é bem suportado (e.g., Bybee et al. 2008, Dumont et al. 2010, Dijkstra et al. 2014). 2

1.2 Coenagrionidae

Das 27 famílias que compõem a subordem Zygoptera, Coenagrionidae é a mais diversa com pelo menos 1.267 espécies em 114 gêneros (Dijkstra et al. 2013) e seu monofiletismo é bem suportado (Hasegawa & Kasuya 2006, Dumont et al. 2010, Dijkstra et al. 2014). A subdivisão tradicional e aceita por longas décadas de Coenagrionidae até então consistia em seis subfamílias: Agriocnemidinae, Argiinae, Coenagrioninae, Ischnurinae, Leptobasinae e Pseudagrioninae (Davis & Tobin 1984). No entanto, uma extensa análise filogenética baseada em caracteres morfológicos já estabelecidos e outros 3

codificados, demonstrou um relacionamento impreciso entre as subfamílias. A única subfamília recuperada como monofilética foi Agriocnemidinae, enquanto as demais foram reconhecidas como polifiléticas (O‘Grady & May 2003). Esse cenário inconclusivo pode ser explicado pelo fato de que muitas características morfológicas utilizadas para separar as subfamílias são imprecisas e não exclusivas. Análises filogenéticas recentes reconhecem dois grandes grupos dentro de Coenagrionidae: um grupo caracterizado por seus representantes com fronte arredondada (Core-Coenagrionidae), com distribuição predominantemente holártica, sustentado com alto suporte tanto na análise de Dijkstra et al. (2014) quanto em análises anteriores (Bybee et al. 2008; Carle et al. 2008; Dumont et al. 2010) e outro grupo com representantes com fronte angulada (ridged-face Coenagrionidae). Segundo a classificação mais recente de Zygoptera, o primeiro grupo inclui as sufamílias Agriocnemidinae, Ischnurinae e Pseudagrioninae e o segundo grupo Protoneurinae, Pseudostigmatinae e Teinobasinae e o gênero Argia (Bybee et al. 2008; Dijkstra et al. 2013, 2014).

1.3 Protoneuridae s.l.

Formada para abrigar espécies que apresentam venação alar reduzida, Protoneuridae s.l., gênero-tipo Protoneura Selys in Sagra, 1857, passou por diversas mudanças taxonômicas ao longo do tempo. Tillyard (1917) já havia apontado diferenças entre as espécies que ocorrem no Velho e Novo Mundo. Posteriormente, Fraser (1957) propôs quatro subfamílias dentro de Protoneuridae s.l., a saber: Protoneurinae (Novo Mundo), Caconeurinae (Subcontinente Indiano), Disparoneurinae (Velho Mundo) e Isostictinae (Australasia). Posteriormente, baseado em caracteres larvais, Lieftinck (1975) elevou Isostictinae a família. Watson (1992) reanalisou as três subfamílias restantes e constatou que os caracteres morfológicos utilizados para separar essas subfamílias não são exclusivos e não suportam essa subdivisão. Desse modo, Protoneuridae deveria conter somente uma única subfamília Protoneurinae incluindo Caconeurinae e Disparoneurinae, proposta que contribuiu para que Protoneuridae reunisse espécies distantemente relacionadas. As hipóteses filogenéticas baseadas em caracteres morfológicos e moleculares que foram sendo publicadas (Rehn 2003, O‘Grady & May 2003, Bybee et al. 2008, Pessacq 4

2008) demonstraram recorrentemente o monofiletismo de Protoneurinae contudo restrito aos representantes Americanos (Protoneuridae s.s.).

1.4 Protoneurinae

Atualmente, Protoneurinae reúne 122 espécies em 15 gêneros, são exclusivos da Região Neotropical (Schorr & Paulson 2019, Pimenta et al. 2019) e que apresentam o tamanho do corpo e venação alar reduzidos. Ocorrem em ambientes florestais, com voo fraco, são geralmente encontrados em regiões sombreadas associados a corpos de água corrente (Pinto & Kompier 2018). Apesar da baixa amostragem taxonômica, as análises filogenéticas baseadas em caracteres morfológicos e moleculares suportam o monofiletismo da subfamília (Rehn 2003, O‘Grady & May 2003, Rehn 2003, Bybee et al. 2008, Pessacq 2008, Dijkstra et al. 2014). As características morfológicas que sustentam o grupo são (1) venação alar reduzida (e.g., uma intercalar entre o RP1 e RP2, ramos cubitais e anais reduzidos ou indistintos), (2) superfície dorsal do antenífero carenada e (3) subárculo proximal ou na divergência entre a RP e MA (Rehn 2003, Pessacq 2008). O relacionamento filogenético de Protoneurinae com as demais subfamílias de Coenagrionidae ainda é incerto. As hipóteses variam de acordo com a amostragem taxonômica e da metodologia aplicada. Rehn (2003), por exemplo, utilizou dois gêneros da subfamília em uma análise cladística com 122 caracteres morfológicos a qual resultou na hipótese de Protoneurinae como grupo-irmão de Isostictidae, suportada por uma única sinapomorfia que é a redução ou ausência da CuA. Na análise bayesiana apresentada por Bybee et al. (2008) dois gêneros de Protoneurinae foram amostrados e suportados como grupo-irmão de representantes de Argiinae e Pseudostigmatinae. Por outro lado, Pessacq (2008) em uma análise cladística com a maior amostragem taxonômica da subfamília (treze gêneros) e com 47 caracteres morfológicos, Protoneurinae foi demonstrado como grupo-irmão de um grupo que inclui Isostictidae, Platycnemididae e os protoneurídeos paleotropicais Carle et al. (2008), baseado em genes nucleares, propôs Protoneurinae como proximamente relacionado a Argia Rambur, 1842, gênero extremamente diverso que apresentam a fronte arredondada.

1.5 Clado Roppaneura

Como ressaltado em estudos anteriores (e.g., Pessacq 2008) os gêneros de Protoneurinae apresentam diagnoses conflitantes. A monotipia de alguns gêneros (Junix Racenis, 1968; Lamproneura De Marmels, 2003 e Roppaneura Santos, 1966) e de descrições genéricas inconsistentes evidenciaram hipóteses de relacionamento intergenérico imprecisas. Apesar disso, Pessacq (2008) em uma análise filogenética morfológica com pesagem diferencial de caracteres demonstrou um clado que foi recorrentemente recuperado (com baixo suporte) composto por seis gêneros: Amazoneura Machado, 2004, Forcepsioneura Lencioni, 1999, Lamproneura De Marmels 2003, Phasmoneura, Williamson, 1916, Psaironeura Williamson, 1915 e Roppaneura Santos, 1966. Esses gêneros não apresentam caracteres exclusivos, como por exemplo, o antenífero não carenado, que é encontrado também nas espécies de Idioneura Selys, 1860. Pode-se citar outro exemplo, um processo lateroposterior do protórax desenvolvido, um caráter utilizado para distinguir as espécies de Forcepsioneura de Amazoneura, processo, o qual, também é observado no gênero monotípico Lamproneura e possivelmente em Roppaneura e Psaironeura.

1.6 Forcepsioneura

Desde a década de 1980, Machado (1985) reconheceu que espécies de Phasmoneura deveriam ser transferidas a um gênero próprio (cf. Pinto & Kompier 2018). Lencioni (1999), analisando um exemplar de Protoneurinae proveniente do estado de São Paulo verificou que esse se assemelhava a Phasmoneura ciganae, porém com diferenças nos cercos (forma dos processos médio- e ventro-basais) e na lígula genital (dobra interna alongada e margem laterodistal com dois filamentos longos). Por essas diferenças, propôs o gênero Forcepsioneura, para abrigar três espécies previamente descritas em Phasmoneura (P. ephippigera Selys, 1886; P. ciganae Santos, 1968 e P. itatiaiae Santos, 1970) e descreveu uma nova espécie, Forcepsioneura garrisoni Lencioni, 1999, a qual designou como espécie-tipo. 6

Simultaneamente, Machado (1999) propôs que Protoneura sancta Hagen in Selys, 1860 descrita somente por um exemplar fêmea de Lagoa Santa, Minas Gerais, a qual estava alocada no gênero Psaironeura Williamson, 1915 e apresentava uma diagnose problemática e sem ilustrações, era sinônimo sênior de F. ciganae. Posteriormente, Machado (2000) propôs mais três espécies: F. lucia Machado, 2000, F. haerteli Machado, 2001 e F. westfalli Machado, 2001. A curiosa distribuição das espécies de Forcepsioneura, com espécies ocorrendo na Região Amazônica e outras nos domínios da Mata Atlântica já havia sido ressaltada por Machado (1985). Além disso, ele também distinguiu dois grupos de espécies: F. sancta e F. ephippigera (Machado, 2000). O primeiro inclui representantes que apresentam um tubérculo póstero-lateral no lobo mediano do protórax bem desenvolvido e a região posterior da cabeça de coloração clara, além de ocorrerem na Mata Atlântica. Nesse grupo estão incluídas F. garrisoni, F. haerteli, F. itatiaiae, F. lucia e F. sancta. Já o outro grupo, abriga F. ephippigera e F. westfalli, com distribuição amazônica, incluindo representantes com tubérculo póstero-lateral do protórax pouco desenvolvido e a região posterior da cabeça de coloração escura (Machado 2000a, 2000b, 2001). Diante dessa distinção e da descoberta de uma terceira espécie com as características do grupo F. ephippigera, Machado (2004) criou o gênero Amazoneura, para abrigar as espécies desse grupo amazônico e descreveu Amazoneura juruaensis (Machado, 2004). Dessa forma, Forcepsioneura é tratado como um gênero endêmico da Mata Atlântica, nos remanescentes da região Sul e Sudeste do Brasil, com exceção de F. sancta, também registrada para o Cerrado do país (Pessacq et al. 2012). São espécies associadas a ambientes aquáticos de planícies e regiões montanhosas, voam próximos á superfície de corpos d‘água com pouca correnteza que são parte integrantes e secundárias de riachos (Pinto & Kompier 2018, Pimenta et al. 2019). Atualmente onze espécies são incluídas em Forcepsioneura: F. gabriela Pimenta, Pinto et al., 2019, F. garrisoni, F. grossiorum Machado, 2001, F. haerteli (Machado, 2005), F. itatiaiae, F. janeae Pimenta et al., 2019, F. lucia, F. regua Pinto & Kompier, 2018, F. sancta e F. serrabonita Pinto & Kompier, 2018. Dois grupos de espécies de Forcepsioneura podem ser distinguidos baseados em caracteres morfológicos e hábitat (Pinto & Kompier 2018). O primeiro grupo é composto por espécies de menor porte, com coloração predominantemente verde- azulada, cercos delgados com um longo processo ventrobasal e estão associadas a 7

hábitats de planície (F. gabriela, F. garrisoni, F. haerteli, F. regua e F. sancta). O segundo grupo é composto por espécies de maior porte, de coloração predominantemente laranja-esverdeada, cercos robustos com processo ventro-basal curto e estão associadas à habitats montanhosos (F. grossiorum, F. itatiaiae, F. janeae, F. lucia e F. serrabonita). O monofiletismo desses grupos foi investigado por Pimenta et al. (2019) e apenas o primeiro foi considerado monofilético.

1.7 Taxonomia integrativa

Documentar as espécies através de estudos comparativos em taxonomia é fundamental para o conhecimento da biodiversidade do planeta. Atualmente, nomear, classificar e caracterizar os seres vivos ainda tem um papel importante para a biologia, dando suporte para projetos de conservação e ecologia (Schlick-Steiner et al 2010). Delinear e identificar indivíduos muito similares morfologicamente sempre é um grande desafio (Dayrat 2005). Além disso, as linhagens evoluem separadamente dentro das populações e somente quando essa separação se torna morfologicamente visível, os limites das espécies se tornam mais claros (Padial et al 2010). Porém, mesmo com a expressão morfológica, muitas vezes essas diferenças são tão sutis que ainda assim torna-se a distinção entre as entidades difícil. Agregar variadas disciplinas utilizando diferentes fontes de dados para delimitar espécies é a principal proposta da taxonomia integrativa (Dayrat 2005). Quando essas diferentes fontes associadas mostram o mesmo resultado, temos uma maior evidência da separação das histórias evolutivas dos táxons (Schlick-Steiner et al 2010). Análises realizadas com dados morfológicos e moleculares, normalmente apresentam uma hipótese filogenética mais robusta. Além disso, o uso de marcadores moleculares com taxas evolutivas diferentes também podem fornecer hipóteses mais informativas e com isso identificar entidades que ainda não foram separadas morfologicamente (Padial et al. 2010). No primeiro capítulo, duas novas espécies são descritas baseadas em uma minuciosa descrição morfológica, acrescentando também evidências moleculares, de variação genética, dando maior robustez à delimitação das novas entidades.

2. Objetivos

A presente tese visou contribuir para o entendimento da taxonomia, filogenia e evolução de Protoneurinae (Protoneuridae s.s.) com a implementação de métodos integrativos, baseados em caracteres morfológicos e moleculares. O primeiro capítulo desta tese teve como objetivo principal a descrição de duas espécies novas de Forcepsioneura, com o auxílio de uma abordagem integrativa, explorando a distância genética dos indivíduos utilizando três marcadores moleculares (Pimenta et al. 2019). Além disso, foi proposta a primeira hipótese de relacionamento filogenético do gênero com base em dados moleculares. O segundo capítulo visou propor a primeira hipótese filogenética de Protoneurinae, bem como testar seu monofiletismo e o relacionamento de seus gêneros, incluindo uma ampla amostragem taxonômica, diferentes metodologias e fontes de dados. Utilizando caracteres morfológicos e três marcadores moleculares foi possível explorar a relação entre os gêneros de Protoneurinae e seu posicionamento em Coenagrionidae. Além disso, foi investigado o relacionamento interno do clado Roppaneura, bem como testar o monofiletismo dos gêneros incluídos.

Capítulo 1

Integrative taxonomy and phylogeny of the damselfly genus Forcepsioneura Lencioni, 1999 (Odonata: Coenagrionidae: Protoneurinae) with description of two new species from the Brazilian Atlantic Forest

ANA LUIZA ANES PIMENTA 1, 2, ÂNGELO PARISE PINTO 3, * & DANIELA MAEDA TAKIYA 2

1 Graduate Program in Biodiversity and Evolutionary Biology (PPGBBE), Instituto de Biologia, Universidade Federal do Rio de Janeiro, Rio de Janeiro, RJ, Brazil — 2 Laboratório de Entomologia, Departamento de Zoologia, Instituto de Biologia, Universidade Federal do Rio de Janeiro, Caixa Postal 68044, 21941-971 Rio de Janeiro, RJ. Brazil — 3 Laboratório de Sistemática de Insetos Aquáticos (LABSIA), Departamento de Zoologia, Universidade Federal do Paraná, P.O. Box 19020, 81531- 980 Curitiba, PR, Brazil. — * Corresponding author

Accepted 07.08.2019. Published online at www.arthropod-systematics.de on 00.xx.20??. doi

Editors in charge: Gavin Svenson & Klaus-Dieter Klass

ABSTRACT Forcepsioneura Lencioni, 1999 is a small genus of eight forest-dependent damselfly species endemic to the Brazilian Atlantic Forest domain. Some of its species are difficult to identify due to their strong morphological similarities. Thus, the use of DNA sequences for taxonomic purposes is warranted. This study examined the diversity among mitochondrial COI and 16S and nuclear PRMT markers in Forcepsioneura, identified discrete evolutionary units based on morphological and molecular characters, and described two new species using an integrative approach to propose species-level hypotheses. The first molecular phylogeny of Forcepsioneura species, including seven of the 10 valid species, is presented. Forcepsioneura gabriela sp.n. and Forcepsioneura janeae sp.n. are described and illustrated based on males. Forcepsioneura gabriela sp.n. is closely related to F. garrisoni Lencioni, 1999 and F. regua Pinto & Kompier, 2018 and was included in the light blue group, but was recovered with high K2P COI divergence values relative to F. garrisoni. PRMT and ribosomal 16S rDNA sequences were too conservative to distinguish this new species from others of the light blue group. Nevertheless, F. gabriela sp.n. can be distinguished from other Forcepsioneura 10

by its coloration and shape and length of the ventrobasal process of cercus and MP vein. On the other hand, we were unable to get COI sequences for F. janeae sp.n., but morphological diagnostic characters, such as, coloration and shape of the posterior lobe of the prothorax and ventrobasal process of cercus supported its proposal as a new species. A concatenated Bayesian analysis of all markers supported the monophyly of both Forcepsioneura and the light blue group of species. This study affirmed the value of COI sequence variation for species-level studies but did not support the use of PRMT and 16S for this group of damselflies, as there was very little interspecific variation between some closely related species.

Key words. DNA-barcoding, dragonfly, molecular phylogeny, Protoneuridae, Zygoptera.

¹ Manuscrito aceito para publicação em 5 de agosto de 2019 na revista Arthropod Systematics & Phylogeny (CAPES Qualis A2, FI (2017) = 1,703. 11

1. INTRODUCTION

Recent studies on Odonata phylogeny based on morphological and molecular characters have resulted in a comprehensive family-level phylogenetic hypothesis for the suborder Zygoptera (damselflies) (DIJKSTRA et al. 2014) and suggested Protoneuridae as a polyphyletic group (BYBEE et al. 2008; CARLE et al. 2008; PESSACQ 2008; DIJKSTRA et al. 2014). These studies, however, recognize all protoneurid species occurring in the New World, including the type-genus Protoneura Selys, 1857, as a monophyletic group. This group is currently treated as a subfamily of Coenagrionidae, with Protoneurinae damselflies comprising 123 species distributed in 15 genera (GARRISON & von Ellenrieder 2016). Forcepsioneura Lencioni, 1999 is a small genus of forest-dependent damselflies endemic to the Brazilian Atlantic Forest domain, except for F. sancta (Hagen in Selys, 1860), also recorded from the Cerrado of the Central Brazilian plateau. Species in the genus occur in Atlantic Forest remnants from South and Southeastern Brazil associated with specific habitats (PINTO & KOMPIER 2018). Additionally, an unknown species from the State of Rio Grande do Norte, at the northern boundary of the Atlantic Forest domain, was placed in Forcepsioneura (IRUSTA & LENCIONI 2015), but it was most likely misidentified (J. Irusta pers. comm.). Eight species are known for the genus, two of which (F. serrabonita Pinto & Kompier, 2018 and F. regua Pinto & Kompier, 2018) were discovered from material recently collected in the Brazilian states of Bahia and Rio de Janeiro, and more species are yet to be described (see PINTO & KOMPIER 2018; PINTO 2019). According to PINTO & KOMPIER (2018), two informal groups of species can be recognized in Forcepsioneura based on morphological characters and habitat preferences: the light blue group, comprising smaller species with slender cerci and pale areas predominantly bluish-green, which includes F. garrisoni Lencioni, 1999 (type- species), F. haerteli Machado, 2001, F. regua Pinto & Kompier, 2018, and F. sancta (Hagen in Selys, 1860); and the orange-black group, comprising larger species with robust cerci and pale areas orange-green, which includes F. grossiorum Machado, 2005, F. itatiaiae (Santos, 1970), F. lucia Machado, 2000, and F. serrabonita. Species in the first group are associated with lowland habitats and have slender cerci with a comparatively long ventrobasal process, whereas species in the second group are associated with montane habitats and have robust cerci with a short ventrobasal process. In addition, PINTO & KOMPIER (2018) highlighted the morphological similarity of species in the light blue group, especially between F. garrisoni and F. regua and a putative undescribed species collected in southern Bahia. These species are difficult to identify due to the strong morphological similarities that exist in diagnostic characters. Finally, another interesting case that needs to be carefully addressed is that of populations found in the states of Espírito Santo and Rio de Janeiro identified as F. lucia, even though the latter was originally described from the state of Minas Gerais (PINTO & KOMPIER 2018). Any morphological variation, however subtle it may be, that is eventually found between these two populations may indicate that they actually represent distinct species, albeit very closely related to F. lucia. 12

Although morphology has been the basis for taxonomic work, in some cases data from other areas of biology may also provide invaluable information on the species delimitation and phylogeny of a group of species. PADIAL et al. (2010), based on the concept of integrative taxonomy, stressed the importance of using new methods and protocols for species delimitation. With the advance of molecular techniques, molecular data have been increasingly used in taxonomic studies to describe and delimit species and to advance more robust species-level hypotheses. Integrating data from the fields of ecology, geography, population genetics, and behavior into taxonomic studies, as earlier advocated by DAYRAT (2005), has consistently refined species delimitation proposals. Different types of analytical methods and molecular data have been used to investigate species delimitation in Odonata. For example, genetic distance- and character-based methods supported the description of new species from different regions of Africa (DIJKSTRA et al. 2015), whereas phylogeographic studies helped distinguish levels of polymorphism in specimens of Zygoptera species (FERREIRA et al. 2014a) and identify a large-scale barcoding gap using the ABGD method (KOROIVA & KVIST 2017). In Odonata, the three molecular markers most commonly used in taxonomic and phylogenetic studies are the genes encoding mitochondrial cytochrome c oxidase subunit I (COI), ribosomal 12S rDNA, and 16S rDNA (WARE et al. 2007; BALLARE & WARE 2011; SÁNCHEZ HERRERA et al. 2010; DIJKSTRA et al. 2014, 2015; KOROIVA et al. 2017). COI is widely used in studies that address genetic identity in animals, intra- and interspecific distances, and DNA-barcoding methods. Additionally, the gene encoding arginine methyltransferase (PRMT), a nuclear marker, has been proposed by FERREIRA et al. (2014a) as suitable for species level studies in several Zygoptera but was also used in a study of endangered populations of a new species of Gomphidae (Anisoptera) (FERREIRA et al. 2014b). The aims of this study were to examine the genetic diversity among Forcepsioneura species, to identify discrete evolutionary units based on morphological and molecular characters, and to describe the new species identified. Additionally, the first molecular phylogeny of the genus, including seven of the 10 described species, is presented.

2. MATERIAL AND METHODS

2.1. Taxon sampling

Thirty-four specimens of seven species of Forcepsioneura, including two new species that are described herein, were used to investigate inter- and intraspecific genetic distances and infer the genus phylogeny based on the molecular markers COI, 16S, and PRMT (Table 1). Specimens examined are deposited in the following collections: DZRJ – Coleção Entomológica Prof. José Alfredo Pinheiro Dutra, Departamento de Zoologia, Instituto de Biologia, Universidade Federal do Rio de Janeiro, Rio de Janeiro, RJ, Brazil; DZUP – Coleção Entomológica Pe. Jesus Santiago Moure, Departamento de Zoologia, Setor de Ciências Biológicas, Universidade Federal do Paraná, Curitiba, PR, 13

Brazil; MNRJ – Coleção Entomológica do Museu Nacional, Universidade Federal do Rio de Janeiro, Rio de Janeiro, RJ, Brazil; MZSP – Serviço de Entomologia, Museu de Zoologia, Universidade de São Paulo, SP, Brazil. In addition, outgroup species including representatives of the Protoneurinae, ―Teinobasinae‖, Pseudostigmatinae s.l. (all ridge-faced Coenagrionidae), Ischnurinae (core-Coenagrionidae), and Lestidae (Table 1) were selected to represent different levels of the Zygoptera phylogeny following the most comprehensive molecular hypothesis (i.e., DIJKSTRA et al. 2014).

2.2. DNA extraction, amplification and sequencing

DNA was extracted from one leg and muscle bundles of adult specimens using the DNeasy Blood & Tissue Kit following the manufacturer‘s protocol but without macerate the samples (QIAGEN, Hilden, Germany). Fragments of the genes encoding mitochondrial cytochrome c oxidase subunit I (COI, 658 bp), subunit 16S of rDNA (16S, 524 bp), and nuclear arginine methyltransferase (PRMT) were amplified and sequenced using the following primers: LCO1490 (5‘ GGTCA ACAAA TCATA AAGAT ATTGG 3‘) and HCO2198 (5‘ TAAAC TTCAG GGTGA CCAAA AAATC A 3‘) (Folmer et al. 1994); LR-J-12887 (5‘ CCGGT YTGAA CTCAR ATCA 3‘) and LR-N-13398 (5‘ CRMCT GTTTA WCAAA AACAT 3‘) (TAKIYA et al. 2006); and ARG_F2 (5‘ TGCCG CCAAG GCTGG AGCAT C 3‘) and ARG_R3 (5‘ CCGGA ACTCT ATGTA CCACA AC 3‘) (FERREIRA et al. 2014a); respectively. The protocol for amplification of COI and 16S consisted of 3 min at 94°C followed by 35 cycles of 1 min at 94°C, 1 min at 50°C, 2 min at 72°C, and a final extension period of 7 min at 72°C. For PRMT, the amplification protocol consisted of 5 min at 94°C followed by 40 cycles of 30 sec at 92°C, 30 sec at 54°C, 45 sec at 72°C, and a final extension period of 10 min at 72°C. Amplified products were separated on a 1% agarose gel and stained with GelRed™ (Biotium, Inc., Fremont, CA, USA). Purification and sequencing of each amplicon (both strands) was performed by Macrogen (Seoul, Korea). Consensus sequences were generated based on electropherograms with GeneStudio™ Professional Edition v. 2.2.0.0 (Genestudio, Inc., Suwanee, GA, USA). Consensus sequences (and a single PRMT sequence from Neoneura amelia; GenBank KM276629) were aligned with MUSCLE 6 (EDGAR 2004) implemented in MEGA 7 (KUMAR et al. 2016) for COI and PRMT, and MAFFT 6 (KATOH et al. 2005) using the online server (http://mafft.cbrc.jp/alignment/server/) with the Q-INS-I algorithm for 16S.

2.3. Genetic distances and phylogenetic analysis

Pairwise genetic distances (p-distances) (NEI & KUMAR 2000) of specimens were calculated for both 16S and PRMT. However, for COI, distances were modeled under the Kimura two-parameter (K2P) model (KIMURA 1980) for comparability with other barcoding studies. In addition, cluster analysis was conducted for each gene separately using the neighbor-joining (NJ) distance method (SAITOU & NEI 1987), whereas group 14

support was calculated with 1,000 bootstrap (BS) pseudoreplicates. P-distances and NJ analyses were calculated in MEGA 7 (KUMAR et al. 2016). Bayesian inference analyses were performed based on separate gene alignments and the concatenated dataset under a mixed-model strategy. The most appropriate evolutionary model for COI and 16S markers (GTR+I+G) and PRMT (GTR+I) was selected based on the Akaike information criterion (AKAIKE 1974), implemented in jModeltest2 v.2.1.7 (DARRIBA et al. 2012). Four independent Monte Carlo Markov Chain (MCMC) simulations were run in MrBayes 3.2 (RONQUIST et al. 2012) with four chains for 5,000,000 generations with a sample frequency of 1,000 generations in the concatenated analysis and, for individual-gene analysis, four chains for 1,000,000 generations with a sample frequency of 2,000 generations. Convergence and mixing of sampled parameters was checked in Tracer 1.6 (RAMBAUT et al. 2013) with 10% of trees discarded as ―burn- in‖. Branch support was assessed by posterior probabilities of clades (PP, RONQUIST & HUELSENBECK 2003).

2.4. Morphological analysis and terminology

Morphological terminology and general procedures follow PINTO & KOMPIER (2018). The following morphological abbreviations were used in the text: Ax = antenodal crossvein; Fw = fore wings; GL = genital ligula; Hw = hind wings; MBP = mediobasal process; Px = postnodal crossvein; Pt = pterostigma; S1‒10 = abdominal segments; and VBP = ventrobasal process. Measurements and photographs were taken with a Leica DFC 500 digital camera mounted on Leica MZ16 and M205C stereomicroscopes. Multiple focal plane images were compiled using LAS MONTAGE v.4.7 and LAS CORE v.4.6 software.

3. RESULTS

3.1. Genetic distances

Pairwise genetic distances among sequences of all three markers studied are shown in Supplementary Tables S1–S3. Maximum intraspecific variation of COI sequences (Table 2) varied among Forcepsioneura species, from 0% in F. gabriela sp.n. and F. garrisoni and 1.1% in F. aff. lucia to 5.0% in F. sancta. Other markers studied showed no intraspecific variation among sequences, except for 16S (1.1%) and PRMT (1.5%) in F. sancta. Interspecific distances among Forcepsioneura species ranged from 3.8 to 18.4% for COI, 0.0 to 5.2% for 16S, and 0.0 to 3.0% for PRMT (Tables 3‒5). Interestingly, no genetic variation was found between 16S sequences of F. regua and F. gabriela sp.n. and PRMT sequences of F. regua and F. garrisoni.

3.2. Proposal of new taxa

Both new species proposed herein have diagnostic morphological characteristics that justify their description as new taxa (see Table 6). In addition, DNA sequence divergences were assessed to support their distinction. The new species F. gabriela sp.n., described herein, was not recovered as closely related to and with high divergence (3.8%) from the most similar species, F. garrisoni, in K2P distances and the neighbor joining tree of COI sequences (Fig. 1, Table 3). However, 16S and PRMT failed to distinguish the new species from the closely related F. garrisoni and F. regua (Figs. 2, 3). Nevertheless, we believe that COI results in addition to morphological evidence are sufficient to propose this new taxon. Unfortunately, only a single individual of the other new species proposed, F. janeae sp.n., was available for DNA extraction and only 16S was successfully sequenced. Nonetheless, the p-distance between the new species and its most similar species, F. aff. lucia (Fig. 4), was 0.6% (Table 4), which was higher than the variation among some species of the light blue group. Thus, the morphological evidence gathered and the additional variation in 16S sequences was found to be sufficient to propose this new taxon.

3.3. Taxonomy

Coenagrionidae Kirby, 1890 Protoneurinae Yakobson & Bianchi, 1905

3.3.1. Forcepsioneura gabriela sp.n. (Figs. 5A,B, 6A–C, 8A,B, 9A–C)

Material examined. Type material: Holotype ♂, BRAZIL. Bahia State, Una municipality, Reserva Biológica de Una, Expedição Gabriela Cravo e Canela II, first order stream [?], after Fazenda Piedade (15°09′36.2″S 39°10′31.1″W, 100 m a.s.l), 14–15.vi.2014, A.P. Pinto leg. (DZUP 498858, DNA voucher ENT2784). Paratype ♂ same data as holotype but Expedição Gabriela Cravo e Canela IV, 07.viii.2016, A.P. Pinto, A.P.M. Santos, D.M. Takiya & P.M. Souto leg. (DZRJ 3555; DNA voucher ENT3445).

Etymology. Specific name in apposition after the strong female character of the famous novel ―Gabriela, cravo e canela‖ by Brazilian writer Jorge Amado. The novel is set in the region of the type locality at the beginning of the 20th century, when the southern coast of Bahia prospered from the exploitation of cacao trees.

Diagnosis. A small, dorsally brown with metallic green reflections, lateroventrally light blue and yellow protoneurid. Frons angulated, rear of head pale; CuA&AA indistinct; and with genital ligula (GL) with long inner fold and distal lateral lobe elongated into a flagellum according to current concept of Forcepsioneura (see Pinto & Kompier 2018). The long ventrobasal process of cercus (VBP), with length ≥ 0.65 of total cercus 16

length, distinguishes the new species from F. grossiorum, F. lucia, F. itatiaiae and F. serrabonita (as long as 0.55 in F. itatiaiae and ≤ 0.4 the length of cercus in the others); MP short, reaching distally at the level of the vein descending from subnodus, distinguishes it from F. haerteli and F. sancta (MP reaches distally at 0.3‒0.5 the vein descending from subnodus). Forcepsioneura gabriela sp.n. is very similar to F. garrisoni and F. regua by their blue coloration of the lateral portion of the synthorax and shape of the cercus. However, the well-defined tubercle-like process on the posterolateral margin of the prothorax distinguishes it from F. regua (prothorax with an ill-defined process), whereas the robust VBP, with apex not reaching the ventral margin of S10 distally and curved ventrally allows separation from F. garrisoni (apex of VBP reaching distally the ventral margin of S10 and curved mesally).

Description of male holotype. Head (Fig. 5A): Labium, visible parts of maxilla and mandibles ivory-yellow, except apex brown. Genae ivory-yellow. Labrum black encircled by yellow that separates into two large lateral and small mesal spots, yellow occupying ventrally 0.25 of labrum length. Anteclypeus ivory-yellow with mesal wide C-shaped black spot; postclypeus shining black. Antefrons, postfrons, and remainder of epicranium black with shining bluish-green reflections, except by a pair of ill-defined blue spots on antefrons. Antennifer and scape black; posterior surface of pedicel yellow; distal apex of scape ivory-yellow; flagellum dark brown. Posterior region of cranium ivory-yellow, dorsal part 0.40 brownish-black extending ventrally close to occipital foramen. Thorax (Figs. 5B, 9A): Prothorax black, whitish-blue stripe laterally on notum interrupted at posterior margin of median lobe; lateral of anterior lobe, posterolateral margin of median lobe, and ventral 0.4 of propleuron yellow; posterolateral margin of median lobe strongly projected into a tubercle-like process; posterior lobe convex, rectangular, narrower than median lobe, whitish-blue on lateral apex. Synthorax dorsally dark brown to black with bluish-green metallic reflections; mesepisternum entirely black with bluish-green metallic reflections, anterior 0.25 with whitish-blue wedge-shaped spot; mesepimeron and metepisternum dark brown with wide light blue longitudinal stripe running from posterolateral angle of mesinfraepisternum to antealar process, occupying maximum 0.33 of mesepimeron to 0.9 of metepisternum width; metepimeron ivory-yellow; metapostepimeron ivory-yellow with black spot at lateroventral angle. Legs ivory-yellow with irregular dark brown to black areas on dorsal surface of femora and tibia; articulations of femur-tibia, tarsal segments, apices of pretarsal claws and spurs black, except by scale-like ivory-yellow proximal femoral spurs and tibial comb of prothoracic legs; femora with anteroventral surface armed with four long and robust spurs; femora with posteroventral surface with short and thinner spurs, 4 on pro-, 6 on meso-, and 7 on metathoracic leg; tibiae with anteroventral surface armed with 10 spurs on pro- (5 of tibial comb), 5 on meso-, and 6 on metathoracic leg; tibiae with posteroventral surface with 8 on pro-, 7 on meso-, and 12 on metathoracic leg. Wing: Membrane hyaline; venation dark brown to black; Pt black, quadrangular, encircled by thin hyaline line; MP reaches anal margin at level of vein descending from 17

subnodus or very slightly distal, covering 2 cells on all wings; Px on Fw 12; Hw 10; RP2 originating at Px 6 on Fw, at Px 5 on Hw. Abdomen (Figs. 6A–C, 8A, B, 9A, B): S1–10 tergites dark brown to black dorsally, lateroventrally light brown to ivory-yellow, pale areas of S1–3 with blue shade, darker in S8–10, dorsal carina of S1–8 with a very thin pale line along; sternites similar in color to adjacent areas of tergites; pale longitudinal stripe occupying about 0.5 ventral of S1–7 tergites laterally, gradually narrowing to ca 0.2 in S8, a narrow line in S9 and ill-defined areas in S10; S3–7 with anterior pale ring ≤ 0.1 of total length of segment, separated dorsally in S7; S1–7 cylindrical; S8–10 distinctly wider than others segments (S7 width 0.7 of posterior part of S8); S9–10 dorsally covered by whitish-grey pruinosity, less amount on S10; posterior margin of S10 with slight concavity. Secondary genitalia (Fig. 8A,B) typical of Coenagrionoidea; anterior lamina with deep and acute incision; anterior hamule dark brown, quadrangular, with anteroventral angle acutely projected; posterior hamule almost entirely internalized with curved thumb- shape; VS longer than wide, maximum width 0.3 of total length in ventral view. Genital ligula (Fig. 8A,B) with L1 smooth, without any kind of special setae; L2 with posterolateral portion of flexure projected distally beyond median region, making its margin slightly concave in ectal view, distal margin (tip of ligula) with mesal concavity; lateral margins prolonged into two curved long flagella, in ectal view basally almost perpendicular, posteriorly twisted; internal fold proximal to flexure, long, ca 0.4 of L2 length in lateral view; no sclerotized tubercle at flexure. Epiproct reduced to membranous-like plate. Cercus (Fig. 6A–C) brown to dark brown, apex of MBP and VBP black; in lateral view slightly directed obliquely dorsad, gradually tapering distally; VBP in lateral view perpendicular to dorsal branch, length ca 0.7 of cercus, apex stoutly rounded, at distal 0.35 distinctly curved posteriorly, distal edge of apex curved ventrally, reaching distally 0.8 from distance of VBP base to margin of S10; MBP not visible in lateral view; tip of cercus blunt; in dorsal view forcipate, proximally wide, slender distally; lateral margin almost straight, internal margin very slightly curved; apexes converging (Fig. 6C); MBP as an acute fin-shaped plate dorsally, apex strongly directed anteriorly, positioned at basal 0.25, in posterior view directed ventrally obliquely; apex of VBP broadly rounded and abruptly curved inwardly with apical edge ventrally in posterior view. Paraproct light brown with dorsal margin black, plate-like.

Measurements (mm): Total length (incl. caudal appendages) 35; abdomen length (excl. caudal appendages) 29.7; head maximum width 2.9; Fw length 19.5; Hw length 18; Fw maximum width 3.4; Hw maximum width 3.2; Pt length on Fw 0.5 and on Hw 0.52; length of metathoracic femur 1.8; metathoracic tibia 1.7; length of S9+10 in lateral view 1.1; length of cercus (dorsal branch) in lateral view 0.45; length of VBP in lateral view 0.37. Variation of male paratype. The single male paratype is very similar to the holotype, but is generally lighter in coloration, with more extensive pale and darker areas well defined. Minor differences are described below. 18

Head: Labrum lighter, dark areas brown; labrum with ventral 0.4 yellow. Antefrons yellow, including base of antennifer. Posterior region of the cranium ivory-yellow; 0.2 dorsal brownish-black, not extending ventrally close to occipital foramen. Thorax: Prothorax black, lighting to brown and yellow ventrally; posterior lobe with small mesal concavity. Mesepisternum, about half of length, with rounded light brown lateral area close to mesopleural suture; whitish-blue wedge-shaped spot at anterior 0.3 larger; mesepimeron and metepisternum with dark spots larger and black. Legs with femora with anteroventral surface armed with 3 on pro- and 4 spurs on meso- and metathoracic legs long and robust; posteroventral surface spurs 3 on pro-, 4 on meso-, and 6–7 on metathoracic leg; anteroventral surface of tibiae armed with 10 spurs (4 of tibial comb) on pro-, 5 on meso-, and 5–6 on metathoracic leg, posteroventral surface with 9 on pro-, 10 on meso- and 11–14 on metathoracic leg. Wings: MP reaches anal margin at vein descending from subnodus on Fw and very slightly distal on Hw; Px on Fw 12–13; Hw 10; RP2 originating at Px 6 on Fw, at Px 4– 5 on Hw. Abdomen: Pale areas larger than in holotype. Measurements (mm): Total length (incl. caudal appendages) 35.2; abdomen length (excl. caudal appendages) 31; head maximum width 3.0; Fw length 19.7; Hw length 18.5; Fw maximum width 3.5; Hw maximum width 3.3; Pt length on Fw and on Hw 0.5; length of metathoracic femur 1.9; metathoracic tibia 1.9; length of S9+10 in lateral view 1.1; length of cercus (dorsal branch) in lateral view 0.45; length of VBP in lateral view 0.4. Female. Unknown. Larva. Unknown.

Ecology and behavior. Specimens were collected at a slow-running, non-perennial, forested swamp next to a small stream densely covered by aquatic and semiaquatic plants with a soft mud-silt-clay bottom (Fig. 9B). The collection site, at a secondary forest at 100 m a.s.l., was visited twice under very different conditions in the rainy and dry seasons when dramatic changes in the length of the water body occurred. The mesohabitat is very similar to places where F. regua and F. garrisoni have been collected, i.e., a slow-running stream with shallow water column and fine substrate (silt and clay) under shaded, forested Atlantic Forest formations. Males were seen flying in sunflecks at low height, close to ground level and water surface among dense herbaceous-shrubby vegetation. Other species associated with shaded and forested habitats such as Perilestes eustaquioi Machado, 2015, Metaleptobasis selysii Santos, 1956, Idioneura ancilla Selys, 1860, and a new species of Heteragrion were collected at the same site.

Remarks. Forcepsioneura gabriela sp.n. was recovered forming a group of closely related species together with F. garrisoni and F. regua (Fig. 4). K2P interspecific distances for COI between F. gabriela sp.n. and its closest species and F. garrisoni was 3.8%, which is generally considered high for intraspecific divergences. This result and the morphological evidence gathered above are considered sufficient for supporting the 19

hypothesis that individuals of F. gabriela sp.n. represent a species distinct from F. garrisoni and F. regua.

3.3.2. Forcepsioneura janeae sp.n. (Figs. 7A–D, 8C,D, 9D)

Forcepsioneura lucia nec Machado, 2000: — LENCIONI (2005: 192, in part, misidentification from specimen from Espírito Santo State, Brazil); — PESSACQ et al. (2012: 64, record from Espírito Santo State based on LENCIONI 2005); — LENCIONI (2017: 203, in part, replication of the record from Espírito Santo State, Brazil based on LENCIONI 2005).

Material examined. Type material: Holotype ♂, BRAZIL, Espírito Santo State, Santa Teresa municipality, [Biological Station of Santa Lúcia], collecting point 16, limpo 21º [19°56′55″S 40°32′23″W, 796 m a.s.l], 04.iii.2014, Jane Peter Egert Buss & Wander Antônio Martinelli leg. (DZUP 499056). Paratype ♂. same data as holotype, but no additional data further than point 16 (MNRJ 0141; DNA voucher ENT 3403).

Etymology. Specific name of feminine gender, in the genitive form, dedicated to the biologist and fellow colleague Jane Peter Egert Buss, who kindly sent type specimens for study.

Diagnosis. A medium-sized, dorsally brown with bluish-green metallic reflections, laterally to ventrally orange-yellow protoneurid. Frons angulated; rear of the head pale; CuA&AA indistinct; GL with long inner fold and distal lateral lobe elongated into a flagellum according to current concept of Forcepsioneura (see PINTO & KOMPIER 2018). Ventrobasal process (VBP) short, with length < 0.3 the length of cercus, distinguishes the new species from F. gabriela sp.n., F. garrisoni, F. haerteli, F. itatiaiae, F. regua and F. sancta (VBP length ≥ 0.55). Forcepsioneura janeae sp.n. is similar to F. lucia, F. grossiorum, and F. serrabonita by the general orange-black coloration and robust cercus. The new species differs from F. grossiorum by the apex of VBP slightly acute (strongly curved ventrally in F. grossiorum) and from F. lucia by the ratio of VBP and cercus length ≤ 0.3 (as long as ≤ 0.4 in F. lucia) and rectangular posterior lobe of prothorax (rounded convex in F. lucia and F. serrabonita).

Description of male holotype. Head (Fig. 7A): Labium, visible parts of maxilla and mandibles orangish-yellow, except apex brown. Genae ivory-yellow. Labrum black, expect by transversal orangish-brown stripe occupying 0.25 ventral of labrum length. Anteclypeus black, irregularly spotted with ivory-yellow ventrally; postclypeus black. Antefrons shining black with a pair of elongated pale spots; postfrons and remainder of epicranium opaque black with weak greenish-copper luster. Antenna black; distal apex of scape and posterior surface of pedicel ivory-yellow. Posterior region of cranium (―postgena‖ plus ―occiput‖) dark yellow to pale-brown, probably ivory-yellow in life; dorsal part 0.40 brownish-black. Thorax (Fig. 7B): Prothorax black, lightening to yellow laterally; anterior margin of anterior lobe yellow; propleura almost yellow with irregular dark areas; median lobe 20

with lateral margin black, posterolateral margin strongly projected into tubercle-like process; posterior lobe convex, rectangular, width similar to median lobe, posterior margin almost straight, ca 0.13 lateral folded ventrally. Synthorax dorsally dark brown to black with bluish-green metallic reflections; lateral to ventrally orangish-yellow; mesepisternum entirely black with metallic reflections; mesepimeron dark-brown to black with orangish-yellow longitudinal stripe running from of mesinfraepisternum to ca 0.8 posterior, narrowing posteriorly, occupying maximum 0.33 of mesepimeron; metepisternum ivory-yellow with brown longitudinal stripe anterior to metapleural suture, running from to metinfraespisternum to antealar carina; metepimeron ivory- yellow; metapostepimeron ivory-yellow with black spot at lateroventral angle. Legs ivory-yellow with irregular dark-brown to black areas on dorsal surface of femora and tibia; articulations of femur-tibia, tarsal segments, apex of pretarsal claws and spurs black, except scale-like ivory-yellow proximal femoral spurs and tibial comb of prothoracic leg; femora with anteroventral surface armed with long and robust spurs, 3 on pro- and 4 on meso- and metathoracic legs; posteroventral with short and thinner spurs, 4 on all legs; tibiae with anteroventral surface armed with 8 spurs on pro- (3 of tibial comb), 5–6 on meso-, and 6–7 on metathoracic leg; tibiae with posteroventral surface with 8 on pro-, 5–6 on meso- and 11–13 on metathoracic leg. Wing: Membrane hyaline; venation black, light brown at base; Pt black, quadrangular; MP reaches anal margin at distal 0.25 on Fw, 0.33–0.40 on Hw to vein descending from subnodus, covering 2 cells on all wings; Px on Fw 14; Hw 11–12; RP2 originating at 0.5 distal to Px 6 on Fw, slightly distal at Px 4 on Hw. Abdomen (Fig. 7C,D, 8C,D): S1–10 tergites dark brown to black dorsally, lateroventrally brown to orange, darker in S8–10, pale areas of S1–3, sternites similar in color to adjacent areas of tergites; pale longitudinal stripe occupying about 0.5 ventral of S1–7 tergites laterally, gradually narrowing to ca 0.2 in S8, narrow line in S9 and ill- defined areas in S10; S3–7 with anterior pale ring ≤ 0.1 of total length of segment, separated dorsally in S3 and S5–7; S1–7 cylindrical, S8–10 distinctly wider than others segments; posterior margin of S10 with slight concavity in dorsal view. Secondary genitalia (based on paratype, Fig. 8C, D) typical of Coenagrionoidea; anterior lamina with deep and acute incision; anterior hamule dark brown, quadrangular, with anteroventral angle acutely projected; posterior hamule almost entirely internalized with a curved thumb-shape; VS longer than wide, maximum width 0.3 of total length in ventral view. Genital ligula (based on paratype, Fig. 8C,D) rectangular in ectal view with L1 smooth, without any kind of special setae; L2 with posteromedial portion of flexure projected distally beyond median region, making its margin slightly convex in ectal view, distal margin (tip of ligula) with mesal concavity; lateral margins prolonged into two curved long flagella, in ectal view basally almost perpendicular; internal fold proximal to flexure, long, ca 0.4 of L2 length in lateral view; no sclerotized tubercle at flexure. Epiproct reduced to membranous-like plate. Cercus (Fig. 7C,D) orange, dark brown on VBP and ventral margin of MBP; in lateral view slightly directed obliquely dorsad, gradually tapering distally; VBP process in lateral view perpendicular to dorsal branch, short, length ca 0.32 of cercus, apex slightly acute, distal edge less than half distance from ventral margin of S10; MBP largely visible in lateral view such as 21

rounded flat tubercle; tip of cercus blunt, in dorsal view forcipate, wide and robust proximally, slender distally; lateral margin curving at ca 0.5 to apex, internal margin slightly curved; apexes converging; MBP positioned at basal 0.3 of cercus, in posterior view directed ventrally obliquely and projected from a dilated area of cercus; apex of VBP broadly rounded and curved inwardly. Paraproct black brown, plate-like. Measurements (mm): Total length (incl. caudal appendages) 37.6; abdomen length (excl. caudal appendages) 34; head maximum width 3.3; Fw length 22.7; Hw length 21.3; Fw maximum width 3.7, Hw maximum width 3.6; Pt length on Fw and Hw 0.5; length of metathoracic femur 2.1; metathoracic tibia 1.9; length of S9+10 in lateral view 1.3; length of cercus (dorsal branch) in lateral view 0.4; length of VBP in lateral view 0.13 Variation of male paratype. The single male paratype is very similar to the holotype. Minor differences are described below. Wings: MP reaches anal margin at distal 0.25–0.3 on Fw, 0.40–0.45 on Hw to vein descending from subnodus, covering 2 cells on all wings; Px on Fw 12–13; Hw 11; RP2 originating at Px 8 vein on Fw, at Px 7 on Hw. Abdomen: Epiproct reduced to membranous-like plate. L1 smooth, without any kind of special setae; VBP in lateral view perpendicular to dorsal branch, length ca 0.2 of cercus; apex of VBP broadly rounded and abruptly curved inwardly in posterior view. Measurements (mm): Total length (incl. caudal appendages) 38; abdomen length (excl. caudal appendages) 32.7; head maximum width 3.2; Fw length 22; Hw length 20.4; Fw maximum width 3.4, Hw maximum width 3.3; Pt length on Fw 0.5 and Hw 0.55; no legs, length of S9+10 in lateral view 1.27; length of cercus (dorsal branch) in lateral view 0.4; length of VBP in lateral view 0.1. Female. Unknown. Larva. Unknown.

Ecology and behavior. The two males were collected near a small first order stream with muddy bottom under secondary forest of a typical Atlantic Forest remnant of Southeastern Brazil at 796 m a.s.l. Like other montane species in the genus, it is most likely associated with rocky seepages rather than man-made dams with muddy substrate such as the one at the type-locality (Fig. 9D).

Remarks. Interspecific p-distances for 16S between the paratype of F. janeae sp.n. and specimens of its genetically closest species, F. aff. lucia, was 0.6%, a value that apparently supports the erection of F. janeae sp.n. as a new species based on molecular data. The two males are in poor condition, with loss of color and poorly preserved abdominal segments, including S10, and eyes slightly to severely smashed/crushed.

3.4. Phylogeny of Forcepsioneura

The current analysis supports Forcepsioneura as a monophyletic genus based on the concatenated (Fig. 4) and individual gene (Supplementary Figs. 1‒3) Bayesian inference trees. In the concatenated tree, two main groups were recovered with support: one included F. janeae sp.n. and Forcepsioneura aff. lucia (PP = 97%), and the other included all other Forcepsioneura species examined (PP = 50%). This subclade includes F. serrabonita as sister to the light blue group (PP = 96%) containing F. sancta (PP = 89%) and a group of species closely related to F. garrisoni (PP = 99%). Resulting Bayesian and NJ trees based on COI and 16S also recovered a monophyletic light blue group (F. sancta, F. garrisoni, F. regua, and F. gabriela sp.n.) with moderate support. A well-supported monophyletic group of three very closely related species was identified (F. garrisoni, F. regua, and F. gabriela sp.n.), despite the low support (BS = 64%) in the PRMT NJ tree.

4. DISCUSSION

4.1. Species delimitation

4.1.1. Morphological data

Taxonomy of lower categories, hence species delimitation, in Protoneurinae is largely based on the caudal appendages and genital ligula of males (see discussion in PINTO & KOMPIER 2018). Often species-level is distinguished based on minor differences of caudal appendages. Several Protoneurinae genera group species with great similarity in general appearance thus making species identification a hard task (e.g., VON ELLENRIEDER & GARRISON 2008, ANJOS-SANTOS & PESSACQ 2013). The two new species herein proposed were erected after a careful study of the external morphology that allowed observation of convincing, although slight, differences used for distinction from their congeners. In F. gabriela we have detected minor differences in caudal appendages, wing venation, and prothorax from the two other very similar species of the light blue group, while in F. janeae equally minor differences in the caudal appendages and prothorax of males from the orange-black group. All these differences combined with genetic distances of COI (F. gabriela) and 16S rDNA (F. janeae) supported the erection of these species in an integrative approach due to congruence between these two sources of data.

4.1.2. Genetic distances

With the exception of F. sancta (K2P distances up to 5%), intraspecific COI diversity in Forcepsioneura species was in agreement with previous studies that reported maximum intraspecific divergence values < 2% for odonate species from the suborder Zygoptera (Table 7). However, a recent meta-analysis of the barcoding gap in 497 Zygoptera 23

species reported a level of intraspecific variation ranging between 0‒22%, with only ca. 10% of species having divergence values > 3% (KOROIVA & KVIST 2017). The high intraspecific diversity for F. sancta relative to other Forcepsioneura species may be explained by the large sample size (14 specimens), which better captured the geographical diversity of the species and provided evidence for the possibility that some populations represent cryptic species and should be properly investigated. Nuclear markers, including PRMT, have been successfully used to identify high levels of polymorphism in specimens of Coenagrion mercuriale (Charpentier, 1840) and to describe a new species of Onychogomphus Selys, 1854 (FERREIRA et al. 2014a, b). However, nuclear PRMT and ribosomal 16S rDNA were apparently too conservative to reveal any differences between the very similar F. gabriela sp.n., F. garrisoni, and F. regua. Nevertheless, COI distances resulted in an intraspecific divergence value of 3.8% between F. gabriela sp.n. and F. garrisoni, which is well above the average intraspecific divergence for Zygoptera (Table 7). Moreover, detailed and rigorous morphological comparisons with other species (see diagnosis and Table 6) provided robust evidence to propose F. gabriela sp.n. as a new species. Evidence for the distinction between F. janeae sp.n. and F. aff. lucia, in addition to morphological characters, was also based in a small variation (interspecific distance 0.6%) of 16S rDNA (Table 4). Although the 16S region is not widely used in species delimitation studies in insects, the distance found between these two species was higher than no variation found between other related pair of species in Forcepsioneura. In insects, the 16S has a lower rate of substitution in comparison to COI, thus it has been neglected as a marker used in species delimitation studies, but Misof et al. (2002) show that 16S substitution rates are uneven across insect clades, thus, a significant interspecific distance is strongly clade dependent. Finally, a large dataset of COI barcode sequences of Zygoptera (KOROIVA & KVIST 2017) failed to reveal any differences between intraspecific and interspecific values of genetic distances (barcoding gap). This finding shows the importance of integrating morphological and molecular data in taxonomic studies, especially in highly similar and closely related species, as was the case of F. gabriela sp.n.

4.2. Forcepsioneura phylogeny

The current analysis recovered Forcepsioneura as a monophyletic group with high support. However, we acknowledge that we did not include probably highly related genera such as Psaironeura Williamson, 1915 for a more robust test of its monophyly. Two main groups of species were recovered in the Bayesian inference analysis. However, these groups apparently do not agree with the two species groups tentatively suggested by PINTO & KOMPIER (2018). In our analysis, the orange-black group is not monophyletic (with low support) because F. serrabonita is recovered as sister group to the monophyletic light blue group. This species was expected to be more closely related to F. janeae sp.n. (herein included in the orange-black group) and F. aff. lucia, the latter of which is very similar to F. lucia. Even though PINTO & KOMPIER (2018) tentatively 24

erected these groups and recognized them as potentially not reflecting phylogenetic relationships, the division is premature. In all analyses conducted herein, F. garrisoni, F. regua, and F. gabriela sp.n. were recovered as a monophyletic group. Species in this group are morphologically very similar, sharing the blue coloration of the lateral portion of the synthorax and slender cercus with comparatively long VBP. Unfortunately, gaps in gene sampling and the highly conserved markers used prevented us from confidently establishing interspecific relationships among these species. As an example, F. regua specimens did not group together in the concatenated analysis, which is possibly a result of these factors.

5. CONCLUSIONS

This was the first attempt to establish phylogenetic relationships among species of a South American damselfly genus from the Atlantic Forest using DNA sequences. The phylogeny of Forcepsioneura presented herein supported the monophyly of the genus and its internal relationships. In most cases, the genetic distance method was able to distinguish intra- and interspecific divergence within Forcepsioneura. This study confirms the value of COI sequence variation in species-level studies, which recovered high intraspecific divergences in one case, but does not support the use of PRMT, as previously advocated, as there was very little to no variation among specimens of the same species. Additionally, very low interspecific variation was found between PRMT and 16S sequences in some closely related species pairs examined. Of the molecular markers used in a preliminary study, 16S rDNA provided the best information to resolve relationships among species with good clade support. Nevertheless, this study has some limitations, including unbalanced molecular sampling, missing data, and the use of few not very informative genes. Finally, using an integrative approach based on morphological and molecular data, we described two new species from the states of Bahia and Espírito Santo. We suggest the proposal of new taxa in this group be based on evidence from both morphological and molecular data to accurately identify relationships among Forcepsioneura species, especially complexes of highly similar species.

6. ACKNOWLEDGEMENTS

Angelo B.B. Machado and Jane P.E. Buss sent specimens vital for study. We thank the curators of the DZRJ (Nelson Ferreira-Jr) and MNRJ (Leonardo H.G. Azevedo) for providing access to their respective collections. Additional photographic and microscopic equipment used in this study was made available by Rede Paranaese de Coleções Biológicas – TAXONLINE (Universidade Federal do Paraná), Departamento de Entomologia and Laboratório de Biologia e Sistemática de Odonata Universidade 25

Federal do Rio de Janeiro (MNRJ). Collection of specimens examined was made possible through permits 25034-1, 14591-3, 14591-4, 25034-3, and 41194-1 from Instituto Chico Mendes de Conservação da Biodiversidade (ICMBio). Paulo Cruz and Ivan Leão (ICMBio) provided assistance in the collection of specimens at Reserva Biológica de Una. A previous version of the manuscript benefited from comments by Paulo C. Paiva, Alcimar L. Carvalho, Beatriz Mello, Julianna F. Barbosa (Universidade Federal do Rio de Janeiro), and Allan P.M. Santos (Universidade Federal do Estado do Rio de Janeiro). This paper is part of the requirements for obtaining a D. Sc. Degree of the Programa de Pós-graduação em Biodiversidade e Biologia Evolutiva (PPGBBE) at UFRJ. Ana Luiza Anes Pimenta acknowledges doctoral scholarships from Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES, finance code 001) and Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) from PROTAX project (proc. 440564/2015-8). Ângelo Parise Pinto acknowledges a post- doctoral fellowship from CNPq (PDJ proc. 157592/2015-4). Daniela Maeda Takiya is a research productivity fellow from CNPq (proc. 313677/2017-4) and holds a Jovem Cientista do Nosso Estado fellowship from Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ, proc. E-26/202.786/2015). This study was partially funded by a Universal grant from CNPq (proc. 423821/2016-4).

7. REFERENCES

ANJOS-SANTOS D., PESSACQ P. 2013 Peristicta guarellae sp. nov. from Brazil (Odonata: Protoneuridae). –International Journal of Odonatology 16(4): 293–299. doi:10.1080/13887890.2013.845066

AKAIKE H. 1974. A new look at the statistical model identification. – IEEE Transactions on Automatic Control 19(6): 716–723.

BALLARE E.F., WARE J.L. 2011. Dragons fly, biologists classify: an overview of molecular odonate studies and our evolutionary understanding of dragonfly and damselfly (Insecta: Odonata) behavior. – International Journal of Odonatology 14(2): 137–147. doi: 10.1080/13887890.2011.579538 BEATTY C.D., SÁNCHES-HERRERA M., SKEVINGTON J.H., RASHED A., GOSSUM H.A.,

KELSO S., SHERRAT T.N. 2017. Biogeography and systematics of endemic island damselflies: The Nesobasis and Melanesobasis (Odonata: Zygoptera) of Fiji. – Ecology and Evolution 7(17): 7117–7129. doi: 10.1002/ece3.3175

BYBEE S.M., ODGEN T.H., BRANHAM M.A., WHITING M.F. 2008. Molecules, morphology and fossils: a comprehensive approach to odonate phylogeny and the 26

evolution of the odonate wing. – Cladistics 24(4): 477–514. doi: 10.1111/j.1096- 0031.2007.00191.x

CARLE F.L., KJER K.M., MAY M.L. 2008. Evolution of Odonata, with Special Reference to Coenagrionidae (Zygoptera). – Arthropod Systematics & Phylogeny 66(1): 37–44. DARRIBA D., TABOADA G.L., DOALLO E., POSADA D. 2012. jModelTest 2: more models, new heuristics and parallel computing. – Nature Methods 9(8): 772. doi: 10.1038/nmeth.2109 DAYRAT B. 2005. Towards integrative taxonomy. – Biological Journal of the Linnean Society 85: 407–415. DIJKSTRA K-D.B., KALKMAN V.J, DOW R.A., STOKVIS F.R., VAN TOL J. 2014. Redefining the damselfly families: the first comprehensive molecular phylogeny of Zygoptera (Odonata). – Systematic Entomology 39: 68–96. doi: 10.1111/syen.12035 DIJKSTRA K.-D.B., KIPPING J., MÉZIÈRE N. 2015. Sixty new dragonfly and damselfly species from Africa (Odonata). – Odonatologica 44(4): 447–678. doi: 10.5281/zenodo.35388

EDGAR R.C. 2004. MUSCLE: a multiple sequence alignment method with reduced time and space complexity. – BMC Bioinformatics 5(1): 113. doi: 10.1186/1471-2105-5- 113

FERREIRA S., LORENZO-CARBALLA M.O., TORRES-CAMBAS Y., CORDERO-RIVERA A., THOMPSON D.J., WATTS P.C. 2014a. New EPIC nuclear DNA sequence markers to improve the resolution of phylogeographic studies of coenagrionids and other odonates, – International Journal of Odonatology 17(2–3): 135–147. doi: 10.1080/13887890.2014.950698

FERREIRA S., VELO-ANTÓN G., BROCHARD C., VIEIRA C., ALVES P.C., THOMPSON D.J., WATTS P.C., BRITO J.C. 2014b. A critically endangered new dragonfly species from Morocco: Onychogomphus boudoti sp. nov. (Odonata: Gomphidae). – Zootaxa 3856(3): 349–365. doi: 10.11646/zootaxa.3856.3.3 FOLMER O., BLACK M., HOEH W., LUTZ R., VRIJENHOEK R. 1994. DNA primers for amplification of mitochondrial cytochrome C oxidase subunit I from diverse metazoan invertebrates. – Molecular Marine Biology and Biotechnology 3: 294–299.

GARRISON R.W., VON ELLENRIEDER N., LOUTON J.A. 2010. Damselfly genera of the New World: an illustrated and annotated key to the Zygoptera. – The John Hopkins University Press, Baltimore. xiv + 490 pp. 27

GARRISON R.W., VON ELLENRIEDER N. 2016. A synonymic list of the New World Odonata – revised version of 18 February 2016. [Updated electronic version of Garrison, R.W. 1991. A synonymic list of the New World Odonata. Argia 3(2): 1–30] – Available on [accessed 28 March 2018] IRUSTA J.B., LENCIONI F.A.A. 2015. First record of Pseudostigmatidae (Insecta: Odonata) in the Northeast Region of Brazil. – Checklist 11(2): 1565. doi: 10.15560/11.2.1565 KALKMAN V., ORR A.G. 2016. A description of the larva and discussion of radiation in th phytotelm breeding damselfly genus Papuagrion in New Guinea (Odonata: Zygoptera: Coenagrionidae). – International Journal of Odonatology 19(3): 169–182. doi: 10.1080/13887890.2016.1203363

KATOH K., KUMA K., TOH H., MIYATA T. 2005. MAFFT version 5: improvement in accuracy of multiple sequence alignment. – Nucleic Acids Research 33: 511–518. doi: 10.1093/nar/gki198 KIMURA M. 1980. Simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. – Journal of Molecular Evolution 16: 111–120. KOROIVA R., PEPINELLI M., RODRIGUES M.E., ROQUE F.O, LORENZ-LEMKE A.P., KVIST S. 2017. DNA barcoding of odonates from the Upper Plata basin: Database creation and genetic diversity estimation. – PloS ONE 12(8): e0182283. doi: 10.1371/journal.pone.0182283

KOROIVA R., KVIST S. 2017. Estimating the barcoding gap in a global dataset of cox1 sequences for Odonata: close, but no cigar. – Mitochondrial DNA, Part A: 2–7. doi: 10.1080/24701394.2017.1357709 KUMAR S., STECHER G., TAMURA K. 2016. MEGA7: Molecular Evolutionary Genetics Analysis version 7.0 for bigger datasets. – Molecular Biology and Evolution 33: 1870– 1874. doi: 10.1093/molbev/msw054 MARINOV M., AMAYA-PERILLA C., HOLVWELL G.I, VARSANI A., VAN BYSTERVELD K.,

KRABERG S., STAINTON D., DAYARAMA A., CURTIS N., CRUICKSHANK R.H., PATERSON A. 2016. Geometric morphometrics and molecular systematics of Xanthocnemis sobrina (McLachlan, 1873) (Odonata: Coenagrionidae) and comparison to its congeners. – Zootaxa 4078(1): 084–120. doi: 10.11646/zootaxa.4078.1.9 28

MISOF B., ANDERSON C.L., BUCKLEY T.R., ERPENBECK D., RICKERT A., MISOF K. 2002. An empirical analysis of mt 16S rRNA covarion-like evolution in insects: site-specific rate variation is clustered and frequently detected.– Journal of Molecular Evolution 55(4), 460–469. doi: 10.1007/s00239-002-2341-1

NEI M., KUMAR S. 2000. Molecular Evolution and Phylogenetics. – Oxford University Press, New York. 352 pp.

NOLAN L., HOGG I.D., SUTHERLAND D.L., STEVENS M.I., SCHNABEL K.E. 2007. Allozyme and mitochondrial DNA variability within the New Zealand damselfly genera Xanthocnemis, Austrolestes, and Ischnura (Odonata). – New Zealand Journal of Zoology 34:4, 371–380. doi: 10.1080/03014220709510097

PESSACQ P. 2008. Phylogeny of neotropical Protoneuridae (Odonata: Zygoptera) and a preliminary study of their relationship with related families. – Systematic Entomology 33: 511–528. doi: 10.1111/j.1365-3113.2007.00414.x PINTO A.P. 2019. First report on the dragonflies from Parque Estadual da Ilha do Cardoso, state of São Paulo, Brazil, with notes on the morphology and behavior of Lauromacromia picinguaba (Odonata: Corduliidae s.l.). – Studies on Neotropical Fauna and Environment 54(1): 48–60. doi: 10.1080/01650521.2018.1538656

RAMBAUT A. 2006–2009. FigTree v1.3.1. – URL . RAMBAUT A., SUCHARD M., DRUMMOND A. 2013. Tracer v1.6 – URL . RONQUIST F., TESLENKO M., VAN DER MARK P., AYRES D.L., DARLING A., HÖHNA S.,

LARGET B., LIU L., SUCHARD M.A., HUELSENBECK J.P. 2012. MrBayes 3.2: Efficient Bayesian Phylogenetic Inference and Model Choice Across a Large Model Space. – Systematic Biology 61(3): 539–542. doi: 10.1093/sysbio/sys029

RONQUIST F., HUELSENBECK J.P. 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. – Bioinformatics 19(12): 1572–1574. doi: 10.1093/bioinformatics/btg180

SAITOU N., NEI M. 1987. The Neighbor-joining Method: A new method for reconstruction phylogenetic trees. Molecular Biology and Evolution 4(4): 406–425. 29

SÁNCHES HERRERA M., REALPE E., SALAZAR C. 2010. A neotropical polymorphic damselfly shows poor congruence between genetic and traditional morphological characters in Odonata. – Molecular Phylogenetics and Evolution 57: 912–917. doi: 10.1016/j.ympev.2010.08.016

VON ELLENRIEDER N., GARRISON R.W. 2008 Drepanoneura gen. nov. for Epipleoneura letitia and Protoneura peruviensis, with descriptions of eight new Protoneuridae from South America (Odonata: Protoneuridae). – Zootaxa 1842: 1–34. doi:10.5281/zenodo.183222 WARE J., MAY M., KJER K. 2007. Phylogeny of the higher Libelluloidea (Anisoptera: Odonata): An exploration of the most speciose superfamily of dragonflies. – Molecular Phylogenetics and Evolution 45: 289–310. doi: 10.1016/j.ympev.2007.05.027

Authors’ contributions

A.L. Pimenta gathered and analyzed the data. A.L. Pimenta and A.P. Pinto conducted the descriptions and the taxonomic study. A.P. Pinto and D.M. Takiya verified analytical methods. All authors discussed the results and contributed to the final version of the manuscript.

Electronic Supplement File at http://www.senckenberg.de/arthropod-systematics

File 1: PimentaEtAl-ZygopteraForcepisoneura-ElectronicSupplement.doc — Table S1. K2P pairwise distances of COI sequences of Forcepsioneura and outgroup taxa. Intraspecific distances of F. aff. lucia in gray, F. sancta in blue, F. garrisoni in red, and F. gabriela sp.n. in green. — Table S2. Uncorrected genetic distances (p-distance) of 16S sequences of Forcepsioneura and outgroup taxa. Intraspecific distances of F. aff. lucia in gray, F. sancta in blue, F. regua in orange, and F. gabriela sp.n. in green. — Table S3. Uncorrected genetic distances (p-distance) of the nuclear gene PRMT of Forcepsioneura and outgroup taxa. Intraspecific distances of F. aff. lucia in gray, F. sancta in blue, F. garrisoni in red, and F. gabriela sp.n. in green. — Fig. S1. Bayesian post-burn-in consensus of COI dataset of Forcepsioneura and outgroup taxa. Node- associated values are posterior probabilities higher than 50%. — Fig. S2. Bayesian post- burn-in consensus of 16S dataset of Forcepsioneura and outgroup taxa. Node- associated values are posterior probabilities higher than 50%. — Fig. S3. Bayesian post- 30

burn-in consensus of PRMT dataset of Forcepsioneura and outgroup taxa. Node- associated values are posterior probabilities higher than 50%.

Zoobank Registrations at http://zoobank.org

Present article: http://zoobank.org/urn:lsid:zoobank.org:pub:80BDBD84-30D8-492E- BB13-ADE554D0D661 Forcepsioneura gabriela Pimenta, Pinto & Takiya, 201?: http://zoobank.org/urn:lsid:zoobank.org:act:FC484A31-7CEE-45BC-A7E5- 55E2D5A19A4B Forcepsioneura janeae Pimenta, Pinto & Takiya, 201?: http://zoobank.org/urn:lsid:zoobank.org:pub:80BDBD84-30D8-492E-BB13- ADE554D0D661

FIGURE FILES

Figure 1. Neighbor-joining tree based on K2P distances of COI sequences of Forcepsioneura and outgroup taxa. Node-associated values refer to bootstrap percentages higher than 50%.

Figure 2. Neighbor-joining tree based on uncorrected distances of 16S sequences of Forcepsioneura and outgroup taxa. Node-associated values refer to bootstrap percentages higher than 50%.

Figure 3. Neighbor-joining tree based on uncorrected distances of PRMT sequences of Forcepsioneura and outgroup taxa. Node-associated values refer to bootstrap percentages higher than 50%.

Figure 4. Bayesian post-burn-in consensus of mixed-model analysis of the concatenated dataset (COI, 16S, and PRMT) of Forcepsioneura and outgroup taxa. Node- associated values are posterior probabilities higher than 50%.

TABLES FILES

Family/Subfamily/Species Voucher Locality COI 16S PRMT INGROUP Protoneurinae Epipleoneura venezuelensis ♂ ENT 4355 Una - BA MN058172 MN080560 Forcepsioneura garrisoni ♂ ENT 4344 Antonina - PR MN058173 MN080577 Forcepsioneura garrisoni ♂ ENT 4345 Antonina - PR MN058174 MN080573 Forcepsioneura gabriela sp. nov. ♂ ENT 2784 Una – BA MN058175 MN023110 MN080574 Forcepsioneura gabriela sp. nov. ♂ ENT 3445 Una – BA MN058176 MN023111 MN080575 Forcepsioneura janeae sp. nov. ♂ ENT 3403 Santa Teresa – ES MN023112 Forcepsioneura aff. lucia ♂ ENT 2786 Nova Friburgo – RJ MN058177 MN023113 MN080572 Forcepsioneura aff. lucia ♂ ENT 3404 Itatiaia – RJ MN058178 MN023114 MN080576 Forcepsioneura aff. lucia ♀ ENT 3405 Itatiaia – RJ MN058179 MN023115 Forcepsioneura aff. lucia ♂ ENT 3606 Nova Friburgo – RJ MN058180 MN023116 Forcepsioneura aff. lucia ♂ ENT 3615 Itatiaia – RJ MN058181 MN023117 MN080581 Forcepsioneura aff. lucia ♂ ENT 3616 Itatiaia – RJ MN058182 MN023118 MN080578 Forcepsioneura regua ♂ ENT 2854 Cachoeiras de Macacu – RJ MN023119 MN080570 Forcepsioneura regua ♂ ENT 3608 Cachoeiras de Macacu – RJ MN023120 Forcepsioneura sancta ♂ ENT 2365 Petrópolis – RJ MN058183 MN023121 MN080563 Forcepsioneura sancta ♂ ENT 2366 Rio de Janeiro – RJ MN023122 MN094792 Forcepsioneura sancta ♂ ENT 2369 Nova Friburgo – RJ MN058184 MN023123 MN080564 Forcepsioneura sancta ♂ ENT 2785 Nova Friburgo – RJ MN058185 MN023124 MN080582 Forcepsioneura sancta ♀ ENT 3506 Itatiaia – RJ MN058186 MN023125 MN080565 Forcepsioneura sancta ♂ ENT 3507 Itatiaia – RJ MN058187 MN023126 Forcepsioneura sancta ♂ ENT 3508 Itatiaia – RJ MN058188 MN023127 MN080562 Forcepsioneura sancta ♂ ENT 3509 Itatiaia – RJ MN058189 MN080571 Forcepsioneura sancta ♂ ENT 3510 Itatiaia – RJ MN058190 MN023128 Forcepsioneura sancta ♂ ENT 3511 Itatiaia – RJ MN058191 MN023129 Forcepsioneura sancta ♂ ENT 3512 Itatiaia – RJ MN058192 MN023130 MN080569 Forcepsioneura sancta ♂ ENT 3612 Rio de Janeiro – RJ MN023131 Forcepsioneura sancta ♀ ENT 4340 Curitiba - PR MN058193 MN080561 Forcepsioneura sancta ♂ ENT 4341 Antonina - PR MN058194 Forcepsioneura sancta ♂ ENT 4343 Itatiaia - RJ MN058195 Forcepsioneura sancta ♂ ENT 4346 Antonina - PR MN058196 MN080566 Forcepsioneura sancta ♂ ENT 4347 Antonina - PR MN080579 Forcepsioneura sancta ♂ ENT 4348 Antonina - PR MN080580 Forcepsioneura sancta ♂ ENT 4350 Antonina - PR MN080567 Forcepsioneura sancta ♂ ENT 4354 Itatiaia - RJ MN080568 Forcepsioneura serrabonita ♂ ENT 2857 Camacan – BA MN058197 MN023132 Idioneura ancilla ♂ ENT 3447 Una – BA MN058198 MN023133 Neoneura amelia - Na-3 - KM276629* Peristicta aeneoviridis ♂ ENT 4342 Maquiné - RS MN058199 Roppaneura beckeri ♀ ENT 4337 Curitiba - PR MN058200 OUTGROUP Lestidae: Lestes forficula ♂ ENT 2789 Nova Friburgo – RJ MN058201 MN023134 Ischnurinae: Acanthagrion aepiolum ♀ ENT 3398 Brasília – DF MN023135 Pseudostigmatinae: Leptagrion andromache ♂ ENT 2788 Parati – RJ MN058202 MN023136 Pseudostigmatinae: Leptagrion elongatum ♂ ENT 3408 Cachoeiras de Macacu – RJ MN023137 Pseudostigmatinae: Mecistogaster amalia ♂ ENT 2860 Rio de Janeiro – RJ MN023138 Pseudostigmatinae: Mecistogaster asticta ♂ ENT 3406 Itatiaia – RJ MN023139 ― Teinobasinae‖: Metaleptobasis selysii ♂ ENT 3446 Una – BA MN058203 MN023140

Table 2. Range (and mean) of intraspecific genetic variation of the three molecular markers (COI, 16S, and PRMT) sequenced for Forcepsioneura, including numbers of individuals (N) analyzed.

Species N K2P distances (COI) N p-distances (16S) N p-distances (PRMT) F. gabriela sp. nov. 2 0.0 2 0.0 2 0.0 F. garrisoni 2 0.0 0 - 2 0.0 F. aff. lucia 6 0–0.011 (0.004) 6 0.0 4 0.0-0.015 (0.008) F. regua 0 - 2 0.0 1 - F. sancta 14 0–0.050 (0.025) 11 0–0.011 (0.004) 14 0–0.015 (0.008)

Table 3. Range (and mean) of interspecific K2P distances between COI sequences of Forcepsioneura species.

F. gabriela sp. nov. F. garrisoni F. aff. lucia F. serrabonita F. garrisoni 0.038 (0.038) F. aff. lucia 0.112– 0.122 0.126–0.136 (0.117) (0.132) F. serrabonita 0.162 0.158 0.173–0.184 (0.162) (0.158) (0.178) F. sancta 0.079–0.096 0.089–0.095 0.105–0.133 0.140–0.158 (0.091) (0.092) (0.122) (0.153)

Table 4. Range (and mean) of uncorrected interspecific distances between 16S sequences of Forcepsioneura species.

F. gabriela sp F. aff. lucia F. janeae sp F. regua F. sancta nov. nov. F. aff. lucia 0.036 (0.036) F. janeae sp nov. 0.030 0.006 (0.030) (0.006) F. regua 0.000 0.036 0.030 (0.000) (0.036) (0.030) F. sancta 0.025–0.033 0.030–0.039 0.025–0.033 0.025–0.033 (0.029) (0.035) (0.029) (0.029) F. serrabonita 0.052 0.041 0.036 0.052 0.039–0.047 (0.052) (0.041) (-) (0.052) (0.043)

Table 5. Range (and mean) of uncorrected interspecific distances between PRMT sequences of Forcepsioneura species.

F. gabriela sp F. garrisoni F. aff. lucia F. regua nov. F. garrisoni 0.008 (0.008) F. aff. lucia 0.000–0.015 0.008–0.023 (0.004) (0.012) F. regua 0.008 0.000 0.008–0.023 (0.008) (0.000) (0.012) F. sancta 0.000–0.015 0.008–0.023 0–0.030 0.008–0.023 (0.011) (0.019) (0.012) (0.019)

Table 6. Diagnostic morphological characteristics for species of Forcepsioneura. Taxa 1. Process of 2. Posterior lobe 3.Stripes on 4. Length of 5. Internal 6. VBP/cercus 7. Length of 8. Apex of VBP 9. Shape of MPB 10. MBP in 11. median lobe of prothorax synthorax MP vein in Hw fold of GL length ratio in VBP of cercus in lateral view lateral view Anteromesal of prothorax lateral view margin of MBP in dorsal view F. gabriela sp. nov. Well-defined Concave mesally Bluish-green Short, reaches Thin and ≥ 0.5 Short, not Broadly rounded Small fin-shaped Not visible Acute anal margin oblique reaching S10 and abruptly plate with an acute usually at the curved inwardly and strongly vein descending anteriorly margin from subnodus (rare distally) F. garrisoni Well-defined Sinuous Bluish-green Short, reaches Thin and ≥ 0.5 Long, exceeding Strongly curved Small fin-shaped Not visible Acute anal margin oblique S10 inwardly plate with an acute usually at the and strongly vein descending anteriorly margin from subnodus (rare distally) F. regua Ill-defined Asymmetrically Bluish-green Short, reaches Thin and ≥ 0.5 Short, not Strongly curved Small fin-shaped Not visible Acute convex anal margin oblique reaching S10 inwardly plate with an acute usually at the and strongly vein descending anteriorly margin from subnodus (rare distally) F. haerteli Well-defined Convex, with Greenish-red – – ≥ 0.5 Long, reaching Rounded, slightly Small slightly Not visible Acute lateral margins S10 curved inwardly rounded plate with straight na acute anteriorly margin F. sancta Well-defined Straight or slightly Bluish-green Long, reaches – ≥ 0.5 Short, not Rounded Small Rounded Not visible Rounded* concave anal margin reaching S10 plate with a distally 0.3‒0.5 rounded anteriorly from the vein margin descending from subnodus F. itatiaiae Well-defined Sinuous, almost Greenish- Long, reaches – ≥ 0.5 Short, not Truncate Large fin-shaped Visible Truncate straight orange anal margin reaching S10 plate with a truncate distally 0.3‒0.5 anteriorly margin from the vein descending from subnodus 44

Table 6. Continued.

Taxa 1. Process of 2. Posterior lobe 3.Stripes on 4. Length of 5. Internal 6. VBP/cercus 7. Length of 8. Apex of VBP 9. Shape of MPB 10. MBP in 11. median lobe of prothorax synthorax MP vein fold of GL length ratio in VBP of cercus in lateral view lateral view Anteromesal of prothorax lateral view margin of MBP in dorsal view F. grossiorum Well-defined Strongly sinuous. Greenish- Long, reaches – ≤ 0.4 Short, not Strongly curved Large, slightly Visible Acute Two broad orange anal margin reaching S10 inwardly rounded plate with concavities with distally 0.3‒0.5 an acute anteriorly convex median from the vein margin elevation, lateral descending from margin forming a subnodus flap F. janeae sp. nov. Well-defined Rectangular Orange- Long, reaches Large and ≤ 0.3 Short, not Slightly acute Large, rounded and Visible Acute yellow anal margin upright reaching S10 flat plate with an distally 0.3‒0.5 acute and strongly from the vein anteriorly margin descending from subnodus F. lucia Well-defined Convex, with Greenish- Long, reaches – ≤ 0.4 Short, not Strongly curved Large, rounded and Visible Acute lateral margin orange anal margin reaching S10 inwardly flat plate with an slightly acute distally 0.3‒0.5 acute and strongly from the vein anteriorly margin descending from subnodus F. serrabonita Well-defined Laterally as small Orange- Long, reaches Large and < 0.4 Short, not Rounded Small, rounded Visible Rounded flat processes yellow anal margin upright reaching S10 plate with a distally 0.3‒0.5 rounded anteriorly from the vein margin descending from subnodus * Sensu Machado (2001)

Table 7. Intraspecific K2P distances of COI sequences of some Zygoptera species published in the literature. * denotes uncorrected values.

# of Intraspecific Family Species References individuals distances Amphipterygidae Pentaphlebia mangana 7 0–0.009 Dijkstra et al. 2015 Coenagrionidae Xanthocnemis zealandica 151 0–0.016* Nolan et al. 2007 Xanthocnemis zealandica 51 0–0.032 Marinov et al. 2016 Acanthagrion aepiolum 2 –0.003 Koroiva et al. 2017

Acanthagrion dorsale 5 0.006 Koroiva et al. 2017

Argia reclusa 4 0.007 Koroiva et al. 2017

Argia smithiana 2 0.0015 Koroiva et al. 2017

Argia tamoyo 4 0 Koroiva et al. 2017

Homeoura nepos 3 0.0015 Koroiva et al. 2017

Oxyagrion 5 0 Koroiva et al. 2017 sulmatogrossense Oxyagrion terminale 2 0.0154 Koroiva et al. 2017

Telebasis willinki 4 0.003 Koroiva et al. 2017

Ceriagrion banditum 6 0–0.003 Dijkstra et al. 2015

Ceriagrion obfuscans 10 0–0.011 Dijkstra et al. 2015

Pseudagrion aureolum 6 0–0.002 Dijkstra et al. 2015

Pseudagrion dactylidium 3 0–0.005 Dijkstra et al. 2015 Pseudagrion pacale 5 0–0.002 Dijkstra et al. 2015 Pseudagrion 7 0–0.026 Dijkstra et al. 2015 tanganyicum Nesobasis brachycera 2 0.004 Beatty et al. 2017 Papuagrion Kalkman & Orr 3 0–0.003 marijanmatoki 2016 Kalkman & Orr Papuagrion occipitale 4 0.002–0.015 2016 Chlorocyphidae Africocypha varicolor 7 0–0.023 Dijkstra et al. 2015 Chlorocypha flammea 2 0 Dijkstra et al. 2015

Heteragrionidae Heteragrion triangulare 3 0–0.001 Koroiva et al. 2017 Megapodagrionidae Rhinagrion atripes 4 0.02 Casas et al. 2017 Rhinagrion flammea 2 0.013 Casas et al. 2017

Rhinagrion fulgifrons 3 0.006 Casas et al. 2017

Rhinagrion tendipes 4 0.015 Casas et al. 2017

Platycnemididae Coeliccia angustior 2 0.067 Casas et al. 2017 Elattoneura aurifex 5 0.003–0.012 Dijkstra et al. 2015 Elattoneura lapidaria 7 0–0.011 Dijkstra et al. 2015

Elattoneura 2 0–0.002 Dijkstra et al. 2015 tambotonorum

SUPPLEMENTARY MATERIAL

FIGURES

Figure S1. Bayesian post-burn-in consensus of COI dataset of Forcepsioneura and outgroup taxa. Node-associated values are posterior probabilities higher than 50%

Figure S2. Bayesian post-burn-in consensus of 16S dataset of Forcepsioneura and outgroup taxa. Node-associated values are posterior probabilities higher than 50%.

Figure S3. Bayesian post-burn-in consensus of PRMT dataset of Forcepsioneura and outgroups. Node-associated values are posterior probabilities higher than 50%.

TABELS

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 1 Lestes forficula ENT 2789 2 Leptagrion andromache ENT 2788 0.220 3 Peristicta aeneoviridis ENT 4342 0.143 0.201 4 Metaleptobasis selysi ENT 3446 0.185 0.178 0.113 5 Epipleoneura venezuelensis ENT 4355 0.204 0.218 0.189 0.173 6 Idioneura ancilla ENT 3447 0.211 0.187 0.185 0.185 0.201 7 Roppaneura beckeri ENT 4337 0.197 0.213 0.171 0.161 0.215 0.182 8 Forcepsioneura serrabonita ENT 2857 0.248 0.213 0.202 0.218 0.237 0.198 0.175 9 Forcepsioneura aff. lucia ENT 2786 0.233 0.194 0.172 0.172 0.256 0.184 0.153 0.173 10 Forcepsioneura aff. lucia ENT 3404 0.229 0.197 0.175 0.175 0.256 0.188 0.149 0.176 0.003 11 Forcepsioneura aff. lucia ENT 3405 0.229 0.197 0.175 0.175 0.256 0.188 0.149 0.176 0.003 0.000 12 Forcepsioneura aff. lucia ENT 3606 0.233 0.194 0.172 0.172 0.256 0.184 0.153 0.173 0.000 0.003 0.003 13 Forcepsioneura aff. lucia ENT 3615 0.245 0.205 0.179 0.183 0.256 0.188 0.164 0.184 0.009 0.011 0.011 0.009 14 Forcepsioneura aff. lucia ENT 3616 0.245 0.205 0.179 0.183 0.256 0.188 0.164 0.184 0.009 0.011 0.011 0.009 0.000 15 Forcepsioneura sancta ENT 2365 0.232 0.210 0.179 0.190 0.252 0.180 0.165 0.147 0.105 0.109 0.109 0.105 0.116 0.116 16 Forcepsioneura sancta ENT 2369 0.228 0.202 0.164 0.171 0.231 0.184 0.161 0.147 0.109 0.112 0.112 0.109 0.119 0.119 0.035 17 Forcepsioneura sancta ENT 2785 0.228 0.202 0.161 0.171 0.223 0.184 0.154 0.140 0.106 0.109 0.109 0.106 0.116 0.116 0.041 0.009 18 Forcepsioneura sancta ENT 3506 0.244 0.218 0.172 0.186 0.235 0.196 0.176 0.158 0.126 0.129 0.129 0.126 0.133 0.133 0.050 0.014 0.017 19 Forcepsioneura sancta ENT 3507 0.244 0.222 0.183 0.197 0.261 0.188 0.176 0.158 0.116 0.119 0.119 0.116 0.116 0.116 0.009 0.038 0.044 0.047 20 Forcepsioneura sancta ENT 3508 0.244 0.218 0.172 0.186 0.235 0.196 0.176 0.158 0.126 0.129 0.129 0.126 0.133 0.133 0.050 0.014 0.017 0.000 0.047 21 Forcepsioneura sancta ENT 3509 0.244 0.218 0.172 0.186 0.235 0.196 0.176 0.158 0.126 0.129 0.129 0.126 0.133 0.133 0.050 0.014 0.017 0.000 0.047 0.000 22 Forcepsioneura sancta ENT 3510 0.244 0.218 0.172 0.186 0.235 0.196 0.176 0.158 0.126 0.129 0.129 0.126 0.133 0.133 0.050 0.014 0.017 0.000 0.047 0.000 0.000 23 Forcepsioneura sancta ENT 3511 0.244 0.218 0.172 0.186 0.235 0.196 0.176 0.158 0.126 0.129 0.129 0.126 0.133 0.133 0.050 0.014 0.017 0.000 0.047 0.000 0.000 0.000 24 Forcepsioneura sancta ENT 3512 0.244 0.218 0.172 0.186 0.235 0.196 0.176 0.158 0.126 0.129 0.129 0.126 0.133 0.133 0.050 0.014 0.017 0.000 0.047 0.000 0.000 0.000 0.000 25 Forcepsioneura sancta ENT 4340 0.224 0.202 0.164 0.179 0.235 0.184 0.157 0.150 0.115 0.118 0.118 0.115 0.122 0.122 0.041 0.017 0.026 0.026 0.041 0.026 0.026 0.026 0.026 0.026 26 Forcepsioneura sancta ENT 4341 0.224 0.202 0.164 0.179 0.235 0.184 0.157 0.150 0.115 0.118 0.118 0.115 0.122 0.122 0.041 0.017 0.026 0.026 0.041 0.026 0.026 0.026 0.026 0.026 0.000 27 Forcepsioneura sancta ENT 4343 0.244 0.226 0.183 0.197 0.261 0.192 0.176 0.158 0.116 0.119 0.119 0.116 0.116 0.116 0.011 0.035 0.044 0.044 0.003 0.044 0.044 0.044 0.044 0.044 0.038 0.038 28 Forcepsioneura sancta ENT 4346 0.220 0.198 0.168 0.183 0.231 0.180 0.153 0.154 0.118 0.122 0.122 0.118 0.126 0.126 0.044 0.020 0.029 0.029 0.044 0.029 0.029 0.029 0.029 0.029 0.003 0.003 0.041 29 Forcepsioneura garrisoni ENT 4344 0.249 0.217 0.191 0.209 0.215 0.195 0.168 0.158 0.133 0.136 0.136 0.133 0.126 0.126 0.095 0.095 0.095 0.092 0.092 0.092 0.092 0.092 0.092 0.092 0.092 0.092 0.092 0.089 30 Forcepsioneura garrisoni ENT 4345 0.249 0.217 0.191 0.209 0.215 0.195 0.168 0.158 0.133 0.136 0.136 0.133 0.126 0.126 0.095 0.095 0.095 0.092 0.092 0.092 0.092 0.092 0.092 0.092 0.092 0.092 0.092 0.089 0.000 31 Forcepsioneura gabriela sp. nov. ENT 2784 0.253 0.206 0.194 0.221 0.244 0.196 0.168 0.162 0.118 0.122 0.122 0.118 0.112 0.112 0.095 0.085 0.088 0.096 0.092 0.096 0.096 0.096 0.096 0.096 0.082 0.082 0.089 0.079 0.038 0.038 32 Forcepsioneura gabriela sp. nov. ENT 3445 0.253 0.206 0.194 0.221 0.244 0.196 0.168 0.162 0.118 0.122 0.122 0.118 0.112 0.112 0.095 0.085 0.088 0.096 0.092 0.096 0.096 0.096 0.096 0.096 0.082 0.082 0.089 0.079 0.038 0.038 0.000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 1 Lestes forficula ENT 2789 2 Acanthagrion aepiolum ENT 3398 0.179 3 Leptagrion andromache ENT 2788 0.168 0.143 4 Leptagrion elongatum ENT 3408 0.163 0.163 0.055 5 Mecistogaster amalia ENT 2860 0.187 0.168 0.121 0.129 6 Mecistogaster asticta ENT 3406 0.171 0.171 0.124 0.127 0.074 7 Metaleptobasis selysi ENT 3446 0.165 0.160 0.102 0.107 0.105 0.107 8 Idioneura ancilla ENT 3447 0.157 0.146 0.135 0.154 0.149 0.152 0.135 9 Forcepsioneura serrabonita ENT 2857 0.174 0.168 0.140 0.140 0.143 0.160 0.124 0.143 10 Forcepsioneura janeae sp. nov. ENT 3403 0.174 0.163 0.129 0.135 0.138 0.149 0.110 0.127 0.036 11 Forcepsioneura aff. lucia ENT 2786 0.179 0.168 0.135 0.140 0.140 0.154 0.116 0.132 0.041 0.006 12 Forcepsioneura aff. lucia ENT 3404 0.179 0.168 0.135 0.140 0.140 0.154 0.116 0.132 0.041 0.006 0.000 13 Forcepsioneura aff. lucia ENT 3405 0.179 0.168 0.135 0.140 0.140 0.154 0.116 0.132 0.041 0.006 0.000 0.000 14 Forcepsioneura aff. lucia ENT 3606 0.179 0.168 0.135 0.140 0.140 0.154 0.116 0.132 0.041 0.006 0.000 0.000 0.000 15 Forcepsioneura aff. lucia ENT 3615 0.179 0.168 0.135 0.140 0.140 0.154 0.116 0.132 0.041 0.006 0.000 0.000 0.000 0.000 16 Forcepsioneura aff. lucia ENT 3616 0.179 0.168 0.135 0.140 0.140 0.154 0.116 0.132 0.041 0.006 0.000 0.000 0.000 0.000 0.000 17 Forcepsioneura sancta ENT 2365 0.176 0.165 0.140 0.140 0.154 0.163 0.127 0.135 0.041 0.028 0.033 0.033 0.033 0.033 0.033 0.033 18 Forcepsioneura sancta ENT 2366 0.176 0.163 0.140 0.140 0.154 0.163 0.127 0.135 0.044 0.030 0.036 0.036 0.036 0.036 0.036 0.036 0.008 19 Forcepsioneura sancta ENT 2369 0.174 0.163 0.138 0.138 0.152 0.160 0.124 0.132 0.039 0.025 0.030 0.030 0.030 0.030 0.030 0.030 0.003 0.006 20 Forcepsioneura sancta ENT 2785 0.174 0.163 0.138 0.138 0.152 0.160 0.124 0.132 0.039 0.025 0.030 0.030 0.030 0.030 0.030 0.030 0.003 0.006 0.000 21 Forcepsioneura sancta ENT 3506 0.176 0.163 0.140 0.140 0.154 0.163 0.127 0.135 0.044 0.030 0.036 0.036 0.036 0.036 0.036 0.036 0.008 0.000 0.006 0.006 22 Forcepsioneura sancta ENT 3507 0.176 0.165 0.140 0.140 0.154 0.163 0.127 0.135 0.041 0.028 0.033 0.033 0.033 0.033 0.033 0.033 0.000 0.008 0.003 0.003 0.008 23 Forcepsioneura sancta ENT 3508 0.176 0.163 0.140 0.140 0.154 0.163 0.127 0.135 0.044 0.030 0.036 0.036 0.036 0.036 0.036 0.036 0.008 0.000 0.006 0.006 0.000 0.008 24 Forcepsioneura sancta ENT 3510 0.176 0.163 0.140 0.140 0.154 0.163 0.127 0.135 0.047 0.033 0.039 0.039 0.039 0.039 0.039 0.039 0.011 0.003 0.008 0.008 0.003 0.011 0.003 25 Forcepsioneura sancta ENT 3511 0.176 0.163 0.140 0.140 0.154 0.163 0.127 0.135 0.044 0.030 0.036 0.036 0.036 0.036 0.036 0.036 0.008 0.000 0.006 0.006 0.000 0.008 0.000 0.003 26 Forcepsioneura sancta ENT 3512 0.176 0.163 0.140 0.140 0.154 0.163 0.127 0.135 0.044 0.030 0.036 0.036 0.036 0.036 0.036 0.036 0.008 0.000 0.006 0.006 0.000 0.008 0.000 0.003 0.000 27 Forcepsioneura sancta ENT 3612 0.176 0.163 0.140 0.140 0.154 0.163 0.127 0.135 0.044 0.030 0.036 0.036 0.036 0.036 0.036 0.036 0.008 0.000 0.006 0.006 0.000 0.008 0.000 0.003 0.000 0.000 28 Forcepsioneura regua ENT 2854 0.190 0.174 0.146 0.146 0.157 0.160 0.129 0.146 0.052 0.030 0.036 0.036 0.036 0.036 0.036 0.036 0.028 0.030 0.025 0.025 0.030 0.028 0.030 0.033 0.030 0.030 0.030 29 Forcepsioneura regua ENT 3608 0.190 0.174 0.146 0.146 0.157 0.160 0.129 0.146 0.052 0.030 0.036 0.036 0.036 0.036 0.036 0.036 0.028 0.030 0.025 0.025 0.030 0.028 0.030 0.033 0.030 0.030 0.030 0.000 30 Forcepsioneura gabriela sp. nov. ENT 2784 0.190 0.174 0.146 0.146 0.157 0.160 0.129 0.146 0.052 0.030 0.036 0.036 0.036 0.036 0.036 0.036 0.028 0.030 0.025 0.025 0.030 0.028 0.030 0.033 0.030 0.030 0.030 0.000 0.000 31 Forcepsioneura gabriela sp. nov. ENT 3445 0.190 0.174 0.146 0.146 0.157 0.160 0.129 0.146 0.052 0.030 0.036 0.036 0.036 0.036 0.036 0.036 0.028 0.030 0.025 0.025 0.030 0.028 0.030 0.033 0.030 0.030 0.030 0.000 0.000 0.000

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 1 Neoneura amelia KM276629 2 Epipleoneura venezuelensis ENT 4355 0.075 3 Forcepsioneura aff. lucia ENT 2786 0.075 0.053 4 Forcepsioneura aff. lucia ENT 3404 0.075 0.053 0.000 5 Forcepsioneura aff. lucia ENT 3615 0.075 0.053 0.015 0.015 6 Forcepsioneura aff. lucia ENT 3616 0.075 0.053 0.000 0.000 0.015 7 Forcepsioneura sancta ENT 2365 0.060 0.038 0.015 0.015 0.015 0.015 8 Forcepsioneura sancta ENT 2366 0.060 0.038 0.015 0.015 0.015 0.015 0.000 9 Forcepsioneura sancta ENT 2369 0.060 0.038 0.015 0.015 0.015 0.015 0.000 0.000 10 Forcepsioneura sancta ENT 2785 0.060 0.038 0.015 0.015 0.015 0.015 0.000 0.000 0.000 11 Forcepsioneura sancta ENT 3506 0.060 0.038 0.015 0.015 0.015 0.015 0.000 0.000 0.000 0.000 12 Forcepsioneura sancta ENT 3508 0.060 0.038 0.015 0.015 0.015 0.015 0.000 0.000 0.000 0.000 0.000 13 Forcepsioneura sancta ENT 3509 0.060 0.038 0.015 0.015 0.015 0.015 0.000 0.000 0.000 0.000 0.000 0.000 14 Forcepsioneura sancta ENT 3512 0.060 0.038 0.015 0.015 0.015 0.015 0.000 0.000 0.000 0.000 0.000 0.000 0.000 15 Forcepsioneura sancta ENT 4340 0.060 0.038 0.015 0.015 0.030 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.015 16 Forcepsioneura sancta ENT 4346 0.075 0.053 0.000 0.000 0.015 0.000 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.015 17 Forcepsioneura sancta ENT 4347 0.075 0.053 0.000 0.000 0.015 0.000 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.000 18 Forcepsioneura sancta ENT 4348 0.075 0.053 0.000 0.000 0.015 0.000 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.000 0.000 19 Forcepsioneura sancta ENT 4350 0.075 0.053 0.000 0.000 0.015 0.000 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.000 0.000 0.000 20 Forcepsioneura sancta ENT 4354 0.060 0.038 0.015 0.015 0.015 0.015 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.015 0.015 0.015 0.015 0.015 21 Forcepsioneura garrisoni ENT 4345 0.083 0.060 0.008 0.008 0.023 0.008 0.023 0.023 0.023 0.023 0.023 0.023 0.023 0.023 0.023 0.008 0.008 0.008 0.008 0.023 22 Forcepsioneura garrisoni ENT 4344 0.083 0.060 0.008 0.008 0.023 0.008 0.023 0.023 0.023 0.023 0.023 0.023 0.023 0.023 0.023 0.008 0.008 0.008 0.008 0.023 0.000 23 Forcepsioneura regua ENT 2854 0.083 0.060 0.008 0.008 0.023 0.008 0.023 0.023 0.023 0.023 0.023 0.023 0.023 0.023 0.023 0.008 0.008 0.008 0.008 0.023 0.000 0.000 24 Forcepsioneura gabriela sp. nov. ENT 2784 0.075 0.053 0.000 0.000 0.015 0.000 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.000 0.000 0.000 0.000 0.015 0.008 0.008 0.008 25 Forcepsioneura gabriela sp. nov. ENT 3445 0.075 0.053 0.000 0.000 0.015 0.000 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.015 0.000 0.000 0.000 0.000 0.015 0.008 0.008 0.008 0.000

Capítulo 2

Phylogeny of Protoneurinae damselflies (Odonata: Coenagrionidae) based on morphological and molecular data

ANA LUIZA ANES PIMENTA 1, 2, ÂNGELO PARISE PINTO 3, * & DANIELA MAEDA TAKIYA 2

ABSTRACT

This paper included phylogenetic analyses based on combined datasets and a large Protoneurinae sampling to investigate the subfamily relationship with other Coenagrionidae subfamilies and its intergeneric relationships. Also, the status of the Roppaneura clade was investigated, a group of six genera of Protoneurinae recovered as monophyletic in previous phylogenetic analysis based on morphological data. We conducted parsimony, maximum likelihood, and Bayesian analysis. The monophyly of Protoneurinae was recovered in all resulted trees. The hypothesis with higher resolution was that based on the parsimony shows support in analysis with different k values. Bayesian analysis recovered the subfamily with maximum support (PP = 100), while in maximum likelihood the bootstrap was above 50%, but a polytomy was found in relationships among genera. In addition, the Roppaneura clade was recovered only in parsimony results, which also recovered the monophyly of the most largely sampled genus Forcepsioneura.

Key words: Neotropical dragonfly, phylogenetic analysis, total evidence, Zygoptera 53

1. INTRODUCTION

Species of damselflies (Zygoptera) currently assigned to Coenagrionidae compose a large group of the Odonata, which along with dragonfly (Anisoptera) families Gomphidae and Libellulidae, add up to around 55% of the entire order richness (Dijkstra et al. 2013). Classification and phylogenetic hypotheses for this family have been controversial especially in lineages with reduced venation, which traditionally have been treated as Protoneuridae (Tillyard & Fraser 1938, Rehn 2003, Bybee et al. 2008, Carle et al. 2008, Dumont et al. 2009, Pessacq 2008, Dijkstra et al. 2014). The concept of Protoneuridae s.l. has undergone several changes throughout its taxonomic history and is largely recognized as a polyphyletic taxon (Watson 1992, Pessacq 2008, Dijkstra et al. 2014, Pinto & Kompier 2018). Tillyard (1917) in one of the earliest systems of classification for the order had already highlighted significant differences between Old and New World protoneurids. Four decades later, in an updated version of the system of Tillyard & Fraser (1938, 1939, 1940), Fraser (1957), based on wing venation and caudal appendages of males, recognized Protoneuridae as divided into four subfamilies: Protoneurinae (Protoneuridae s.s., species exclusively from the New World), Caconeurinae (Indian subcontinent), Disparoneurinae (Old World), and Isostictinae (Australasia). However, Watson (1992) reanalyzed Fraser‘s (1957) proposal and concluded that characters used to separate the subfamilies were not unique nor diagnostic, thus restricted Protoneuridae to the New World species of the subfamily Protoneurinae. The putative polyphyly of Protoneuridae received additional support with proposals of phylogenetic hypotheses based on morphological and molecular data (Rehn 2003, O‘Grady & May 2003, Rehn 2003, Bybee et al. 2008, Pessacq 2008, Dijkstra et al. 2014). However, in all of them, Protoneurinae (Protoneuridae s.s.) have been recovered as monophyletic. The following morphological characters support this group: angulated frons (Rehn, 2003), dorsal surface of the antennifer carinated, and subarculus proximal or at the divergence of RP-MA bifurcation (Rehn 2003, Pessacq 2008). Currently, Protoneurinae includes 116 species in 15 genera from the Neotropical region (Garrison et al. 2010, Pinto & Kompier 2018, Pimenta et al. 2019) with reduced venation and body size and usually found in forested areas. Some species fly near the 54

surface of small bodies of running water, and others are associated to rivers with abundant water volumes (Pinto & Kompier 2018). Despite the monophyly of Protoneurinae, phylogenetic relationships of the subfamily are still uncertain (O‘Grady & May 2003, Bybee et al. 2008, Pessacq 2008). Current hypotheses were based on few character sets (e.g., mitochondrial genes, morphology etc.) or low taxon sampling for protoneurine and were investigated with few methods of inference, hence pending corroboration. Rehn (2003) included two protoneurine genera in a cladistic analysis of 122 morphological characters of extant and extinct Zygoptera, which resulted in the hypothesis of Protoneurinae + Isostictidae (composed by oriental species), supported by a single synapomorphy: the CuA reduced and ending on the subdiscoidal vein or absent. In contrast, Pessacq (2008) in a study with higher taxonomic sampling with 13 protoneurine genera and 47 morphological characters, supported Protoneurinae are sister group to a clade including Isostictidae, Platycnemidae, and paleotropical protoneurids. The sister group relationship of Protoneurinae is also not consistent in analyses of molecular data. Carle et al. (2008), in a relationship hypothesis based on a Bayesian inference of two nuclear markers (28S and 18S rDNA) including a single Protoneurinae, the subfamily was found most closely related to Argia Rambur, 1842 (Argiinae), a large genus of Coenagrionidae that has a rounded frons. Finally, a much more comprehensive analysis, focusing on the Zygoptera, was conducted by Dijkstra et al. (2014), which included four Protoneurinae genera (Drepanoneura, Epipleoneura, Neonera, and Protoneura) and three molecular markers (COI, 16S, and 28S rDNA). In this study, two analyses were shown, a Bayesian inference of all combined molecular markers with a lower specimen sampling (295 specimens) and a maximum likelihood analysis of only the ribosomal markers with a higher specimen sampling (356 specimens). In the former, Protoneurinae were recovered as sister group to Argia and, in the latter, as sister group to Telebasis + Aeolagrion (Teinobasinae), both coenagrionid genera with angulated frons. The conflicting phylogenetic position of Protoneurinae has also influenced the monophyly of two large divisions of the current concept of Coenagrionidae (Dijkstra et al. 2014). Carle et al. (2008) had previously drawn attention to two clades within Coenagrionoidea, a clade with representatives with a rounded and another with an angulated frons, the latter comprising not only those coenogrionid genera with angulated frons (including Protoneurinae), but also Argia. Similarly, in Dijkstra et al. (2014), Coenagrionidae was also mostly divided into two clades, the core- and ridge- 55

faced Coenagrionidae. However, although in both analyses the core-Coenagrionidae were recovered as monophyletic, in one of the analysis, the ridge-faced (referring to the angulated frons) group was not, due to the clade formed by Protoneurinae + Argia being sister group to the core-Coenagrionidae. As far as relationships among genera of Protoneurinae, only Pessacq (2008) included a suitable sampling of genera (13 out of 15) to draw any conclusions about them. His morphological analysis evidenced a clade, with low branch support, composed of six South American genera: Amazoneura Machado 2004, Forcepsioneura Lencioni, 1999, Lamproneura De Marmels 2003, Phasmoneura, Williamson, 1916, Psaironeura Williamson, 1915, and Roppaneura Santos, 1966, herein treated as clade Roppaneura. Although this clade has been recovered in all analyses conducted by Pessacq (2008), characters that support it are shared with species of other Protoneurinae genera, e.g., dorsal surface of the antennifer smooth, not carinated (shared with Idioneura ancilla). Another problem with generic circumscriptions of the Roppaneura clade is that diagnostic characters are not exclusive, e.g., the lateroposterior process of the prothorax, which is very distinct and variable in Forcepsioneura, is poorly developed or absent in Amazoneura and in the monotypic Lamproneura. Possibly because of the high morphological similarity among genera of the Roppaneura clade, relationships within the Roppaneura clade were inconclusive, and therefore showed low support in Pessacq (2008) analyses. Finally, considering divergences among the previously proposed hypotheses of phylogenetic relationships of Protoneurinae, the objective of this paper is to provide a more comprehensive and integrative analysis based on both morphological and molecular datasets to propose a clearer hypothesis of the subfamily position in Coenagrionidae and its internal relationships. To attain this goal, we have newly generated DNA data for a higher amount of representatives of Protoneurinae and combined these data with a revised morphological dataset.

2. MATERIAL AND METHODS

2.1 Taxon sampling A total of 49 terminal taxa of 6 families of Zygoptera were used (Table 1). As outgroups, we included nine terminal taxa of Lestidae (1 sp.), Heteragrionidae (1 sp.), Calopterygidae (1 sp.), Isostictidae (2 spp.), and Platycnemididae (4 spp.). The ingroup 56

included 40 Coenagrionidae representatives, including 6 species with rounded frons (core Coenagrionidae clade) and 32 species with angulated frons (ridge-faced complex of genera). Of all 15 genera of Protoneurinae, 10 genera were sampled. Twenty-three terminal taxa were used to represent the subfamily including terminals representing four of the six genera of the Roppaneura clade: Amazoneura, Psaironeura, and Roppaneura (1 species each), and Forcepsioneura (8 species). Among the Forcepsioneura species sampled, there is one unidentified individual (Forcepsioneura sp.), which fits the current diagnosis of the genus, but has a few distinctive features indicating that it may belong to a new genus. In addition, sequences from a total of 21 taxa were obtained from Genbank, as specified in Table 1. Specimens studied are deposited in the following collections: DZRJ – Coleção Entomológica Prof. José Alfredo Pinheiro Dutra, Departamento de Zoologia, Instituto de Biologia, Universidade Federal do Rio de Janeiro, Rio de Janeiro, RJ, Brazil; DZUP – Coleção Entomológica Pe. Jesus Santiago Moure, Departamento de Zoologia, Setor de Ciências Biológicas, Universidade Federal do Paraná, Curitiba, PR, Brazil; MNRJ – Coleção Entomológica do Museu Nacional, Universidade Federal do Rio de Janeiro, Rio de Janeiro, RJ, Brazil; MZSP – Serviço de Entomologia, Museu de Zoologia, Universidade de São Paulo, SP, Brazil. Voucher and other specimens used to code the morphological data are listed in Supplementary File 1.

2.2 DNA extraction, purification and sequencing

DNA was extracted from a leg or thoracic muscle samples of the specimens using DNeasy Blood & Tissue Kit (QIAGEN, Venlo, The Netherlands). Fragments of three molecular markers were amplified and sequenced: the mitochondrial cytochrome c oxidase subunit I (COI) and subunit 16S rRNA (16S), and the nuclear 18S rDNA (18S). Primers used for PCR amplifications are in Table 2. Cycling protocol consisted of initial denaturation of 3 min at 94°C followed by 35 cycles of 1 min at 94°C, 1 min at 50°C, 2 min at 72°C, and a final extension period of 7 min at 72°C. Amplified products were separated on a 1% agarose gel and stained with GelRed™ (Biotium, Inc., Fremont, CA, 57

USA). Purification and sequencing of each amplicon (both strands) was performed by Macrogen (Seoul, Korea). Consensus sequences were generated from electropherograms of both strands using GeneStudio™ Professional Edition v. 2.2.0.0 (Genestudio, Inc., Suwanee, GA, USA). Alignments of the 16S and 18S sequences were made using the online server (http://mafft.cbrc.jp/alignment/server/) of MAFFT v.7 (Katoh et al. 2009) with the algorithm L-INS-i and COI sequences using the default settings of Muscle (Edgar 2004) in MEGA 6 (Tamura et al. 2013). The most appropriate evolutionary models were chosen using the Akaike information criteria (Akaike 1974) in jModeltest v. 2.1.4 (Darriba et al. 2012).

2.3 Phylogenetic analyses

2.3.1 Cladistic analysis Parsimony analyses of the separate or combined morphological and molecular datasets were conducted in TNT 1.5 (Goloboff & Catalano 2016) using distinct schemes of character weighting. All characters were treated as unordered (Fitch 1971) and gaps as missing data. Tree searches under the parsimony criterion are sensible to the amount of missing data commonly observed in molecular datasets. Considering the amount of missing entries in our datasets, three schemes of character weighting were adopted as follows: equal weights (EW), implied weighting (IW, Goloboff 1993), and extended implied weighting (IEW, Goloboff 2014). Differential weighting schemes have been demonstrated to be superior to equal weights in empirical studies and weighting characters inversely to its amount of homoplasy is consistent with the parsimony criterion (for details see Goloboff 1993, 1997, Goloboff et al. 2008). Selection of the value of the constant of concavity (K) should be based on the investigated dataset (Goloboff 1993, Goloboff et al. 2008). Some strategies were developed to select the best value or values of K for a dataset (e.g., Mirande 2009). The strategy adopted here was to explore the congruence among trees recovered by distinct K values as a sensitivity analysis (Wheeler 1995; Goloboff et al. 2008, Goloboff 2014). The preferred K value (K = 14.9732) was selected based on the combined dataset and calculated using the script setk.run (J. S. Arias, unpublished). For both IW and IEW, the preferred K value was adopted, but additional values, 3, 5, 10, 15, and 20 were also investigated. In 58

EIW, molecular data was weighted by the average homoplasy of each partition (COI, 16S, and 18S) using default parameters with distinct K reference values as follows xpiwe(*0.5 <5/K-reference value), while morphological data was weighted individually based on the K value (excluded from collectively weighted partitions). Search strategies should be defined based on the complexity of data sets (i.e., number of terminals, amount of missing data, etc.), computational time consumed, and mainly, guarantee of having investigated globally optimal solutions. Search strategies for most parsimonious trees were first investigated mixing algorithms of ―traditional search‖ (RAS + TBR) and ―new technology‖ (Tree-Drifting, Tree-fusing, Ratchet and Sectorial Search) (Swofford, 1990, Goloboff 1999, Nixon 1999) with user-defined parameters. In addition, a driven search (DS, machine-defined parameters) using the strongest level, suggested for hard datasets (Goloboff et al. 2008), was undertaken with the command line ―xmult= level 10 hits 10 checklevel 2‖. After comparing results and computational time consumed of all these initial searches strategies, DS using new technology algorithms was the preferred search strategy for the hypothesis selection based on similar number and resolution of the resulting trees, and less computational time consumed. Two resampling techniques were conducted to evaluate clade stability, Poisson Bootstrapping and Symmetric Jackknife (Goloboff et al. 2003), with 100 pseudoreplications and DS with level 5 and hits 5 due time consumed. Relative Bremer support (Goloboff & Farris 2001) was also calculated with heuristic search with 100 random addition sequences, TBR strategy, and saving 10 trees by replicate. All analyses were carried out with the molecular dataset only or combined with morphological dataset (combined analysis). Also, morphological characters were optimized over the preferred tree found with the selected value of K using the software WinClada 1.00.08 (Nixon, 2002).

2.3.2 Maximum likelihood analysis

The software jModeltest 2.1.4 (Posada 2008) was used to choose the appropriate model of evolution for each molecular partition, using the Akaike information criteria (Akaike, 1974). Evolution of morphological characters was treated under the Mkv model (Lewis 2001). 59

Maximum likelihood analyses (ML) were performed using GARLI 2.0 (Zwickl 2006) based on the combined dataset with independent models of molecular evolution for each partition, with 100 replicates and 5,000,000 generations. Clades stability was assessed by 1,000 bootstrap pseudoreplicates (Felsenstein 1988).

2.3.3 Bayesian Inference analysis

Bayesian Inference (BI) of the combined morphological and molecular datasets was performed under a mixed-model strategy partitioned by molecular marker (exactly as in the ML analysis). Four independent Monte Carlo Markov Chain (MCMC) simulations were run in MrBayes 3.2 (Ronquist et al. 2012) each with four chains for 2,000,000 generations with a sample frequency of 500 generations. Convergence and mixing of sampled parameters were checked in Tracer 1.6 (Rambaut et al. 2013) with 10% of trees discarded as ―burn-in‖. Posterior probabilities of clades was adopted as value of clade confidence (PP, Ronquist & Huelsenbeck 2003). Resulting trees were read using FigTree 1.4 (Rambaut 2012) and edited using the software Adobe Illustrator CC.

2.4 Selection of preferred hypothesis of relationship We investigated the entire data sets using four different criteria of inference (optimality criteria), adoption of distinct strategies of search and matrixes. The preferred hypothesis was that attained four criteria: (1) resolution, (2) large correspondence with previous analyses, (3) higher values of branches stability indexes, and (4) highest number of nodes shared with other trees in our analyses. These criteria were also used to choose the best cladistic tree from the combined data of different k values (Table 3). For each method, analyses were performed with molecular datasets only and combined data (Table 4). Only the trees generated with total evidence datasets will be presented and discussed.

3. RESULTS

3.1 Morphological data Most characters for the phylogenetic analyses were taken or modified from Pessacq (2008). New characters proposed herein are marked with an asterisk. A list of characters and states is given below.

Head 1. Dorsal surface of antennifer: (0) not carinated; (1) carinated (modified from Pessacq, 2008) 2. Frons: (0) rounded; (1) angulated (Pessacq 2008). 3. Premental cleft: (0) well developed; (1) poorly developed (Pessacq 2008). 4. Postfrontal suture: (0) distinct; (1) indistinct (Pessacq 2008). 5. Ecdysial cleavage line: (0) distinct; (1) indistinct (Pessacq 2008).

Thorax 6. Tubercle-like process of median lobe of prothorax: (0) absent; (1) present (modified from Pessacq 2008). 7. Lobes of posterolateral margin of prothorax: (0) poorly developed; (1) well developed (modified from Pessacq 2008). 8. Female mesostigmal plate: (0) without projections; (1) with one central projection; (2) with two lateral projections (Pessacq 2008). 9. Tibial spurs: (0) long; (1) short (modified from Pessacq 2008). 10. Tarsal claw‘s tooth: (0) well developed; (1) absent or vestigial (Pessacq 2008).

Wings 11. CuA: (0) long, extending several cells beyond its divergence from MP-CuA; (1) short, not extending beyond MP-CuA (Pessacq 2008). 12. CuA&AA: (0) distinct; (1) not distinct (modified from Pessacq 2008). 13. Range of MP: not surpassing the cross vein descending from subnodus; (1) surpassing the cross vein descending from subnodus (modified from Pessacq 2008). 14. Length of MP beyond cross vein descending from subnodus: (0) slightly surpassing or reaching wing margin at half of the first cell beyond cross vein 61

descending from subnodus; (1) reaching wings margin at 2 cells and half beyond cross vein descending from subnodus; (2) reaching wings margin five cells or more beyond cross vein descending from subnodus (modified from Pessacq 2008). 15. Origin of IRP2: (0) at cross vein descending from subnodus; (1) distal to cross vein descending from subnodus (Pessacq 2008) 16. Numbers of additional veins on wings margin between RP1 and RP2 (supplementary veins): (0) one; (1) two; (2) none (modified from Pessacq 2008). 17. Pterostigma: (0) rectangular (width>height); (1) quadrangular (width = height) (modified from Pessacq 2008). 18. Divergence of RP1-RP2 (0) coincident with second postnodal; (1) distal to second postnodal (Pessacq 2008). 19. FW discoidal cell*: (0) quadrangular; (1) rectangular. 20. HW discoidal cell*: (0) quadrangular; (1) rectangular. 21. Subarculus position: (0) distal to RP-MA bifurcation; (1) at or proximal to RP- MA bifurcation. 22. Short cross vein separation IR2 and RP beyond its origin: (0) present; (1) absent or extremely reduced. 23. Numbers of antenodals*: (0) three; (1) more than three. 24. Antenodal space 1*: (0) equal or slightly longer than space 2; (1) longer than twice the space 2.

Genital ligula 25. Internal fold: (0) present; (1) absent (Pessacq 2008) 26. Segment 1 spines: (0) present; (1) absent (Pessacq 2008) 27. Segment II postero-lateral lobes: (0) absent; (1) present (Pessacq 2008) 28. Segment II lateral lobes: (0) present, short, not longer than segment II; (1) present, long, at least twice as long as segment II; (2) absent (Pessacq 2008). 29. Segment II apical lobes: (0) absent; (1) present (Pessacq 2008). 30. Terminal fold: (0) absent; (1) present (Pessacq 2008).

Caudal appendages 31. Cercus: (0) not forcipated; (1) forcipated (Pessacq 2008). 62

32. Dorsal branch shape: (0) inner distal half with concave or flat surface, not rounded in section; (1) inner distal half with convex surface, rounded in section; (2) broadly foliate. 33. Median process of cercus: (0) absent; (1) present (modified from Pessacq 2008). 34. Position of accessory structure: (0) basal in dorsal branch; (1) distal in dorsal branch. 35. Hair concentration on the inner margin of dorsal branch: (0) absent; (1) present. 36. Ventrobasal process of cercus: (0) absent; (1) present, ventrally directed and filiform; (2) present, directed and truncate; (3) present, directed posteriorly (Pessacq 2008) 37. Epiproct: (0) vestigial; (1) well developed (Pessacq 2008).

Ovipositor 38. Ovipositor length: (1) not surpassing posterior apex of S10; (1) surpassing posterior apex of S10 (Pessacq 2008).

Coloration 39. Head: (0) dark with lights markings; (1) uniformly dark, with no light markings; (2) uniformly pale, with no light markings (Pessacq 2008). 40. Synthorax: (0) with well-defined light and dark areas; (1) without well-defined light areas, uniformly dark; (2) without well-defined light areas, uniformly pale (Pessacq 2008). 41. Dark mesothoracic colour: (0) not extending ventrally behind coxae; (1) extending ventrally behind coxae (Pessacq 2008). 42. Iridescence: (0) absent; (1) present (Pessacq 2008).

Larvae 43. Caudal lamellae nodus: (0) poorly developed; (1) well developed (Pessacq 2008). 44. Caudal lamellae shape: (0) flat; (1) sacooid (Pessacq 2008). 45. Labial setae number: (0) more than two pairs; (1) one or two pairs; (2) zero (Pessacq 2008). 63

46. Premental palp distal margin: (0) with an inner curved tooth and several smaller external teeth; (1) with an inner curved tooth and one slightly smaller tooth (Pessacq 2008).

3.2 Datasets for phylogenetic analyses A total of 46 morphological characters were coded for 49 terminal taxa (Supplementary Table 1). The concatenated dataset for DNA sequences has a total of 1,515 bp from three molecular markers: COI (455 bp), 16S (444 bp), and 18S (568 bp). For mixed-model probabilistic analyses, 16S and COI were modeled with GTR+I+G and 18S with HKY+I+G. The NEXUS file including the morphological and molecular characters analyzed in MrBayes is available in Supplementary File 2.

3.3 Cladistic analysis The implied weighting (IW) analysis with K value = 3 was selected as the preferred parsimony tree for the combined dataset, based on the criteria presented in section 2.4. The strict consensus of the combined parsimony analysis (MP) (Figure 1) with implied weighting (IW) found five most parsimonious trees with Fit = 240.969; L = 3,313; CI = 0.27; and RI = 0.40. In general, branch supports were lower than 50%. Coenagrionidae was recovered as monophyletic. The core Coenagrionidae were recovered as monophyletic and the ridge-faced as paraphyletic groupings. Protoneurinae was recovered as a monophyletic group, but with bootstrap support below 50%. The sister group of Protoneurinae was a clade containing the others ridge- faced species (Teinobasinae + Pseudostigmatinae) with only the latter as a monophyletic group, and also the core Coenagrionidae representatives. In addition, the monophyly of Protoneurinae was supported by most K values. Finally, the Roppaneura clade (herein represented by Amazoneura, Forcepsioneura, Psaironeura, and Roppaneura) was recovered as monophyletic, although the clade was not found in any other analysis with different K values. Assuming the putative undescribed species Forcepsioneura sp. as really from another genus, Forcepsioneura has also been recovered as a monophyletic group. 64

3.4 Bayesian Inference analysis

The Bayesian Inference analysis (BI, Figure 2) with combined dataset supports the monophyly of Coenagrionidae (PP = 90), however the core Coenagrionidae was not recovered monophyletic, because Agriocnemidinae was recovered as sister group to the monophyletic ridge-faced Coenagrionidae (PP = 76). Protoneurinae is monophyletic with high support (PP = 100) and sister to all remaining ridge-faced complex representatives. However, relationships among protoneurine genera are uncertain, because of a large polytomy. The only intergeneric relationships supported were Idioneura + Psaironeura (PP = 52) and Roppaneura + Forcepsioneura (PP = 99). Nevertheless, the Roppaneura clade did not appear as monophyletic because of the sister group relationship of Psaironeura tenuissima and Idioneura species. Finally, all genera sampled with more than one terminal were recovered as monophyletic, and assuming that Forcepsioneura sp. is really from another genus, Forcepsioneura has also been recovered as a monophyletic group and appears as sister to Roppaneura beckeri + Forcepsioneura (Gen. sp.) (PP = 99).

3.5 Maximum likelihood analysis The maximum likelihood (ML) tree recovered based on the combined molecular and morphological datasets (lnL= -14528.01391) is shown in Figure 3. The monophyly of Coenagrionidae was recovered with low 50% BS, as well as the core and ridge-faced Coenagrionidae (BS < 50). Argiinae was recovered as sister group to the ridge-faced Coenagrionidae (BS < 50). Protoneurinae was supported as monophyletic with high bootstrap value (BS = 75) and sister group to the remaining ridge-faced Coenagrionidae. The position of Psaironeura tenuissima as closely related to Idioneura species resulted on the non- monophyly of the Roppaneura clade. Most genera relationships presented low support, and only Idioneura and Neoneura show supports above 50% (BS = 97% and 95% respectively). Finally, as in the BI tree, all tested genera were recovered as monophyletic, and again Forcepsioneura (Gen. sp.) is closer to Roppaneura beckeri (BS = 00).

3.6 Morphological character optimizations

Character state changes are presented in the IW combined tree (Figure 4). Coenagrionidae was recovered with a single homoplastic synapomorphy the presence of a short cross vein separating IR2 and RP beyond its origin (char.22[0]), which is shared with species of Isostictidae and Platycnemididae. Five non-exclusive synapomorphies were recovered for Protoneurinae: angulated frons (char.2[1]), premental cleft (char.3[0]), CuA and MP short (char.11[1] and char.14[0] respectively), and median process of cercus (char.33[1]). The Roppaneura clade was supported by four homoplastic synapomorphies: CuAA indistinct (char.12[1]) shared with Aceratobasis macilenta, IRP2 rising at crossvein descending from subnodus (char.15[1]), pterostigma quadrangular (char.17[1]) shared with the clade core Coenagrionidae + ridge-faced non Protoneurinae, and antenodal space 1 longer than twice space 2 (char.24[1]) shared with Heteragrion aurantiacum and A. macilenta.

4. DISCUSSION Phylogenetic analyses based on morphological and molecular data conducted in this study helped obtain a better relationship hypothesis among taxonomic groups sampled here.

4.1 Coenagrionidae Coenagrionidae has long been recovered by previous phylogenetic hypotheses based on morphological data, as a non-monophyletic group (Rehn 2003, O‘Grady & May 2003, Pessacq 2008), because of inconsistencies of many morphological characters which support its subfamilies. However, the monophyly of family has been consistently recovered in a recent molecular study (Dijkstra et al 2014) and is corroborated herein by most analyses conducted, in BI (PP=90) and with clade support low in ML (BS < 50%). We recover the monophyly of core-Coenagrionidae in cladistic and maximum likelihood analyses, although bootstrap values were below 50%. However, it was recovered as paraphyletic in the Bayesian analysis. This result contrasts to Dijkstra et al. (2014), who recover this clade with PP = 100 and 90, depending on the analysis. The monophyly of this group is well supported by previous analysis (Bybee et al. 2008, Carle et al. 2008, Dumont et al. 2010, Dijkstra et al 2014). 66

All phylogenetic hypotheses presented herein support the ridge-faced clade (MP, BSP = 00; BI, PP = 96; ML, BS < 50%), similarly to results found by Carle et al. (2008), Bybee et al. (2008), and Dijkstra et al (2014). This group is also recovered as monophyletic by a cladistic analysis that used only morphological data (Rehn 2003).

4.2 Position of Protoneurinae In previous analyses, depending on the methodology used, the subfamily was found related to different groups, indicating its uncertain positioning. The morphological analysis by Rehn (2003), including three genera and five species of Protoneurinae recovered the subfamily as sister group of Isostictinae, an Old World subfamily, supported by a single synapomorphy: CuA reduced or absent, which was also recovered in this study. These two subfamilies were once a single family in the past. Also performed with only morphological data, Pessacq (2008) recovered Protoneurinae as a sister group of Platycnemidinae + Isostictinae, this clade as a sister group of Telebasis. Protoneurinae was recovered with 2 synapomorphies: dorsal surface of antennifer carinated and subarculus at or proximal to RP-MA bifurcation. This was the phylogenetic hypothesis based on the most extensive taxon sampling of Protoneurinae, with thirteen genera and 29 species of Protoneurinae. In the present study, the last character state appeared in almost all sampled taxa, with exception of H. aurantiacum, Elattoneura glauca, Metaleptobasis selysi, and Protoneura species. Finally, the two representatives of Protoneurinae appear as a sister group of Argia, core and ridge-faced coenagrionids in Bybee et al. (2008). Based on molecular datasets only, Protoneurinae was found related to Argia, a Coenagrionidae species (Carle et al. 2003, Dijkstra et al. 2014) and recovered as sister of Telebasis + Aeolagrion, ridge-faced coenagrionids in ML analyses (Dijkstra et al 2014). Our work is the first to present a hypothesis of phylogenetic relationship of Protoneurinae with an integrative methodology. All topologies confirm and recovered Protoneurinae as a monophyletic group and are in agreement with previous hypotheses. Bayesian and maximum likelihood analyses showed high support values (PP = 100 and BS = 75, respectively), although in the cladistic analysis BS was < 50%. 67

4.3 Internal relationships of Protoneurinae and Roppaneura clade The preferred phylogenetic hypothesis for Protoneurinae internal relationships was presented by parsimony, and is similar to the latest Protoneurinae phylogeny presented, based on morphological characters (Pessacq 2008). Neoneura, for example, was also recovered as sister group of all protoneurines genera. This hypothesis is also observed in other molecular analyses (Dijkstra et al. 2014), but not with our molecular data. Other published analyses did not include enough taxon sampling to investigate relationships among genera. Because of that, results regarding internal genera relationships of protoneurines will be compared to Pessacq (2008). The combined parsimony analysis with implied weighting was the only that recovered the Roppaneura clade, as originally supported by Pessacq (2008). One synapomorphy for this clade cited in his work was the carinated antennifer. In our study this character state was found to have evolved at the node of protoneurines, with some terminals losing the carina. In our probabilistic results, the Roppaneura clade proved to be paraphyletic or even polyphyletic. We included four of the six genera contained in that clade, however, we only managed to obtain COI sequences for Amazoneura, while for Psaironeura and Roppaneura we did not obtain sequences of 16S. We believe that with a more complete molecular sampling of all genera included in the Roppaneura clade, the monophyly of this group could be more suitable tested. The genus with the higher amount of terminal taxa was Forcepsioneura, sampled 7 of 10 described species. This is the most comprehensive analysis of the genus, and it was recovered with high support in the Bayesian analysis (PP = 99). In parsimony analyses, although not having a bootstrap value above 50%, the clade was recovered in all trees with all investigated K values, showing a robust phylogenetic hypothesis for the genus.

5. CONCLUSIONS The present study provided the most integrative phylogeny of Protoneurinae using morphological and molecular datasets. The subfamily recovered as monophyletic in all analyses and showed a closely relation to the other ridge-faced coenagrionids. Most phylogenetic relationships observed in the combined dataset analyses were consistent with phylogenetic hypotheses presented in the literature based on different datasets. We herein provide the most comprehensive in terms of character sampling and the only integrative analysis of Protoneurinae phylogenetic relationships. 68

Although morphological characters are much more laborious and require more time to obtain, and the matrix is with many missing data for some taxa, adding molecular datasets from different markers were able to recover the monophyly of several groups that were not recovered with molecular data only. But we believe that by including and encoding more morphological characters and molecular fragments we can improve the robustness of some clades that have been recovered but with low branch support.

6. REFERENCES

AKAIKE H. 1974. A new look at the statistical model identification. – IEEE Transactions on Automatic Control 19(6): 716–723.

BYBEE S.M., ODGEN T.H., BRANHAM M.A., WHITING M.F. 2008. Molecules, morphology and fossils: a comprehensive approach to odonate phylogeny and the evolution of the odonate wing. – Cladistics 24(4): 477–514. doi: 10.1111/j.1096- 0031.2007.00191.x CARLE F.L., KJER K.M., MAY M.L. 2008. Evolution of Odonata, with Special Reference to Coenagrionidae (Zygoptera). – Arthropod Systematics & Phylogeny 66(1): 37–44. DARRIBA D., TABOADA G.L., DOALLO E., POSADA D. 2012. jModelTest 2: more models, new heuristics and parallel computing. – Nature Methods 9(8): 772. doi: 10.1038/nmeth.2109

DIJKSTRA K.-D.B., BECHLY G., BYBEE S.M. ET AL. 2013. The classification and diversity of dragonflies and damselflies (Odonata). Animal Biodiversity: An Outline of Higher-level Classification and Survey of Taxonomic Richness, (ed. by Z.-Q. Zhang). Zootaxa in press. DIJKSTRA K-D.B., KALKMAN V.J, DOW R.A., STOKVIS F.R., VAN TOL J. 2014. Redefining the damselfly families: the first comprehensive molecular phylogeny of Zygoptera (Odonata). – Systematic Entomology 39: 68–96. doi: 10.1111/syen.12035

DUMONT H.J., VIERSTRAETE A., VANFLETEREN J.R. 2010. A molecular phylogeny of the Odonata (Insecta). – Systematic Entomology 35: 6–18. 69

EDGAR R.C. 2004. MUSCLE: a multiple sequence alignment method with reduced time and space complexity. – BMC Bioinformatics 5(1): 113. doi: 10.1186/1471-2105- 5-113 FRASER F.C. 1957. A Reclassification of the Order Odonata. Royal Zoological Society of New South Wales, Syndey. KATOH, K., ASIMENOS, G. & TOH, H. (2009) MULTIPLE ALIGNMENT OF DNA SEQUENCE

WITH MAFFT. METHODS IN MOLECULAR BIOLOGY, 537, 39–64.

FITCH W.M. 1971. Toward defining the course of evolution: minimum change for a specific tree topology. – Systematic Zoology 20:406–416.

GARRISON, R.W., VON ELLENRIEDER, N., LOUTON, J.A. 2010. Damselfly Genera of the New World: An Illustrated and Annotated Key to the Zygoptera. The Johns Hopkins University Press, Baltimore, Maryland. xiv + 490 p.

GOLOBOFF P.A., FARRIS J.S. 2001. Methods for quick consensus estimation. – Cladistics 17: S26–S34.

GOLOBOFF P.A. 1993. Estimating character weights during tree search. – Cladistics 9: 83–91.

GOLOBOFF P.A. 1999. Analyzing large data sets in reasonable times: solutions for composite optima. – Cladistics 15: 415–428. GOLOBOFF P.A. 2014. Extended implied weighting. – Cladistics 30(3): 260–272. doi: 10.1111/cla.12047 GOLOBOFF P.A., CARPENTER J., ARIAS J., MIRANDA‐ ESQUIVEL D. 2008. Weighting against homoplasy improves phylogenetic analysis of morphological data sets. – Cladistics 24:1–16. doi: 10.1111/j.1096-0031.2008.00209.x GOLOBOFF P.A., CATALANO, S.A. 2016. TNT version 1.5, including a full implementation of phylogenetic morphometrics. – Cladistics 32(3): 221–238. doi:10.1111/cla.12160

GOLOBOFF P.A., FARRIS J.S., KÄLLERSJÖ M., OXELMANN B., RAMÍREZ M.J., SZUMIK

C.A. 2003. Improvements to resampling measures of group support. – Cladistics 19: 324–332

GOLOBOFF P. 1997. Self-weighted optimization: tree searches and character state reconstructions under implied transformation costs. – Cladistics 13: 225–245. Grimaldi, D., Engel, M.S. 2005. Evolution of the Insects. Cambridge University press. xiv+754 p. 70

MIRANDE M. 2009. Weighted parsimony phylogeny of the Family Characidae (Teleostei: Characiformes). – Cladistics 25: 574–613

NIXON, K. C. 1999. The parsimony ratchet, a new method for rapid parsimony analysis. Cladistics 15: 407–414.

O‘GRADY E.W., MAY M.L. 2003. A phylogenetic reassessment of the subfamilies of Coenagrionidae (Odonata: Zygoptera). – Journal of Natural History 37: 2807– 2834

PESSACQ P. 2008. Phylogeny of neotropical Protoneuridae (Odonata: Zygoptera) and a preliminary study of their relationship with related families. – Systematic Entomology 33: 511–528. doi: 10.1111/j.1365-3113.2007.00414.x PIMENTA, A.L., PINTO, A.P, TAKIYA, D.M. (2019). Integrative taxonomy and phylogeny of the damselfly genus Forcepsioneura (Odonata: Coenagrionidae: Protoneurinae) with description of two new species from the Brazilian Atlantic Forest. – Arthropod Systematics & Phylogeny. In press.

PINTO A.P., KOMPIER T. 2018. In honor of conservation of the Brazilian Atlantic Forest: description of two new damselflies of the genus Forcepsioneura discovered in private protected areas (Odonata: Coenagrionidae). – Zoologia 35 (e21351): 1–19. doi: 10.3897/zoologia.35.e21351 POSADA, D. 2008. jModelTest: Phylogenetic Model Averaging. – Molecular Biology and Evolution 25: 1253–1256 RAMBAUT A., SUCHARD M., DRUMMOND A. 2013. Tracer v1.6 – URL .

REHN A.C. 2003. Phylogenetic analysis of higher-level relationships of Odonata. – Systematic Entomology 28:181–239

RONQUIST F., TESLENKO M., VAN DER MARK P., AYRES D.L., DARLING A., HÖHNA S., LARGET B., LIU L., SUCHARD M.A., HUELSENBECK J.P. 2012. MrBayes 3.2: Efficient Bayesian Phylogenetic Inference and Model Choice Across a Large Model Space. – Systematic Biology 61(3): 539–542. doi: 10.1093/sysbio/sys029 RONQUIST F., HUELSENBECK J.P. 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. – Bioinformatics 19(12): 1572–1574. doi: 10.1093/bioinformatics/btg180

SWOFFORD D.L. 1999. PAUP: Phylogenetic Analysis Using Parsimony, version 4.04b4a (Champaign, IL: Illinois Natural History Survey) 71

TAMURA K., STECHER G., PETERSON D., FILIPSKI A., KUMAR S. 2013. MEGA6: Molecular Evolutionary Genetics Analysis Version 6.0. – Molecular Biology and Evolution 30: 2725–2729.– URL TILLYARD R.J. 1917. The Biology of Dragonflies. Cambridge Zoological Series. University Press, Cambridge; xii+321. TILLYARD R.J., FRASER F.C. 1938–1940. A reclassification of the order Odonata. – Australian Zoologist 9: 125–169, 195–221, 359–390.

WATSON J.A.L. 1992. The subfamilies of Protoneuridae (Zygoptera). – Odonatologica, 21: 195–201.

FIGURE FILES

Figure 1. Strict consensus of 5 most parsimonious trees obtained with implied weighting (Reference K = 3) of the combined morphological and molecular datasets of Protoneurinae and outgroups (Fit = 240.969; L = 3,313; CI = 0.27; and RI = 0.40) Rectangles below the nodes correspond to the values of k where the node is also obtained, according to the legend.Values in the branches correspond respectively to Bootstrap, Jackknife, and relative Bremer above 50%

Figure 4. Most parsimonious tree found with implied weighting (K = 3) with combined dataset of Protoneurinae (Length=1070, CI=0.172 and RI=0.425) showing character state changes.

TABLE FILES

Family Subfamily Species Voucher COI 16S 18S Lestidae Lestinae Lestes forficula Rambur, 1842 ENT 2789 MN058201 MN023134 X Heteragrionidae - Heteragrion aurantiacum Selys, 1862 ENT 4608 X X X Calopterygidae Hetaerininae Hetaerina americana (Fabricius, 1798) - KM383854.1 AF170951.1 FJ010010.1 Isostictidae Isostictinae Isosticta robustior Ris, 1915 - - EU055113.1 EU055206.1 Isostictidae Isostictinae Selysioneura capreola Lieftinck, 1932 - KF369541.1 KF369900.1 - Platycnemididae Disparoneurinae Elattoneura glauca (Selys, 1860) - KU566048.1 KF369713.1 FN356083.1 Platycnemididae Disparoneurinae Nososticta solida Hagen in Selys, 1860 - - EU055060.1 FJ009983.1 Platycnemididae Platycnemidinae Copera marginipes (Rambur, 1842) - MG518616.1 KF369679.1 FJ009988.1 Platycnemididae Platycnemidinae Platycnemis pennipes (Pallas, 1771) - KF369498.1 KF369849.1 FJ009986.1 Coenagrionidae Agriocnemidinae Agriocnemis pygmaea (Rambur, 1842) - AB708464.1 AB707520.1 EU055149.1 Coenagrionidae Argiinae Argia mollis Hagen in Selys, 1865 ENT 4606 X X X Coenagrionidae Argiinae Argia moesta (Hagen, 1861) - JN419311.1 - FJ009997.1 Coenagrionidae Ischnurinae Acanthagrion aepiolum Tenessen, 2004 ENT 3398 - X X Enallagma cyanthigerum (Charpentier, Coenagrionidae Ischnurinae - KC912310.1 KF855858.1 AF461237.1 1840) Coenagrionidae Ischnurinae Ischnura elegans (Vander Linden, 1820) - KF369415.1 KF369750.1 AF461239.1 Coenagrionidae Ischnurinae Oxyagrion terminale Selys, 1876 ENT 4611 X - X Coenagrionidae Pseudagrioninae Pseudagrion pilidorsum (Brauer, 1868) - AB708545.1 AB707601.1 - Leptagrion andromache Hagen in Selys, Coenagrionidae Pseudostigmatinae ENT 2788 MN058202 MN023136 1876 X Coenagrionidae Pseudostigmatinae Leptagrion elongatum Selys, 1876 ENT 3408 - MN023137 X Mecistogaster amalia (Burmeister, Coenagrionidae Pseudostigmatinae ENT 2860 - MN023138 - 1839) Coenagrionidae Pseudostigmatinae Platystigma asticta (Selys, 1860) ENT 3406* KC912403.1 MN023139* X* Coenagrionidae Teinobasinae Aceratobasis macilenta (Rambur, 1842) ENT 4607 X - X Coenagrionidae Teinobasinae Metaleptobasis selysi Santos, 1956 ENT 3446 MN058203 MN023140 X Coenagrionidae Teinobasinae Teinobasis scintillans Lieftinck, 1932 - KF369563.1 KF369923.1 - Coenagrionidae Teinobasinae Telebasis griffinii (Martin, 1896) ENT 4612 X - X

Coenagrionidae Teinobasinae Telebasis obsoleta (Selys, 1876) - KF369567.1 KF369927.1 -

Coenagrionidae Protoneurinae Amazoneura ephippigera (Selys, 1886) - KF369292 - - Drepanoneura muzoni von Ellenrieder Coenagrionidae Protoneurinae - KF369365.1 KF369695.1 - & Garrison, 2008 Coenagrionidae Protoneurinae Epipleoneura lamina Williamson, 1915 - KF369385.1 KF369716.1 - Coenagrionidae Protoneurinae Epipleoneura metallica Rácenis, 1955 ENT 2790 X X X Epipleoneura venezuelensis Rácenis, Coenagrionidae Protoneurinae ENT 4357 1955 X X X Forcepsioneura gabriela Pimenta, Pinto Coenagrionidae Protoneurinae ENT 2784 MN058175 MN023110 - & Takiya, 2019 Forcepsioneura garrisoni Lencioni, ENT 4344 Coenagrionidae Protoneurinae MN058173 X* X 1999 ENT 2787* Forcepsioneura janeae Pimenta, Pinto & Coenagrionidae Protoneurinae ENT 3403 MN023112 - Takiya, 2019 X Coenagrionidae Protoneurinae Forcepsioneura aff. lucia ENT 3616 MN058182 MN023118 - 77

Table 1. Continued.

Family Subfamily Species Voucher COI 16S 18S Forcepsioneura regua Pinto & Coenagrionidae Protoneurinae ENT 2854 - MN023119 - Kompier, 2018 Forcepsionera sancta (Hagen in Selys, Coenagrionidae Protoneurinae ENT 2369 MN058184 MN023123 1860) X Forcepsioneura serrabonita Pinto & Coenagrionidae Protoneurinae ENT 2857 MN058197 MN023132 Kompier, 2018 X Coenagrionidae Protoneurinae Forcepsioneura sp. ENT 3812 X X - Coenagrionidae Protoneurinae Idioneura ancilla Selys, 1860 ENT 3447 MN058198 MN023133 X Coenagrionidae Protoneurinae Idioneura celioi Lencioni, 2009 ENT 2791 - X X Neoneura confundens Wassher & Van Coenagrionidae Protoneurinae ENT 3399 ‗t Bosch, 2013 X X X Coenagrionidae Protoneurinae Neoneura fulvicollis Selys, 1886 - KF369455.1 KF369795.1 - Coenagrionidae Protoneurinae Neoneura maria (Scudder, 1866) - - EU055087.1 EU055182.1 ENT 2858 Coenagrionidae Protoneurinae Peristicta aeneoviridis Calvert, 1909 MN058199* ENT 4342* X X* ENT 2859 Coenagrionidae Protoneurinae Psaironeura tenuissima (Selys, 1886) - ENT 4613* X X* Coenagrionidae Protoneurinae Protoneura capillaris (Rambur, 1842) ENT 4614 X - X Coenagrionidae Protoneurinae Protoneura scintilla Gloyd, 1939 - - KF369872.1 - Coenagrionidae Protoneurinae Roppaneura beckeri Santos, 1966 ENT 4338 X - X

Table 2. Primers used for amplification and sequencing of molecular markers for the phylogeny of Protoneurinae.

Primers Direction Sequence 5‘ to 3‘ Reference LCO-1490 (COI+) Foward GGTCAACAAATCATAAAGATATTGG Folmer et al. 1994 C1-J-1718 (COI+a) Foward GGAGGATTTGGAAATTGATTAGTTCC Simon et al. 1994 HCO-2198 (COI-) Reverse TAAACTTCAGGGTGACCAAAAAATCA Folmer et al. 1994 LR-J-12887 (16S+) Foward CCGGTYTGAACTCARATCA Takiya et al. 2006 LR-N-13398 (16S-) Reverse CRMCTGTTTAWCAAAAACAT Takiya et al. 2006 18Sf Foward AGGGCAAGTCTGGTGCCAGC Dias et al. 2019 18Sr Reverse TTTCAGCTTTGCAACCATAC Dias et al. 2019

K L T Fit CI RI 1 3 3313 5 240.969858 0.270 0.406 2 5 3281 5 195.789743 0.272 0.414 3 7 3278 5 165.984261 0.273 0.415 4 10 3278 5 135.848110 0.273 0.415 5 15 3268 7 14.826885 0.274 0.417 6 20 3258 5 85.500387 0.274 0.419

Table 4. Summary of results found for each taxon studied with different phylogenetic methods. Mol: molecular dataset only. Morph+Mol: combined morphological and molecular dataset. Taxon Method Dataset Status Coenagrionidae Parsimony Mol Polyphyletic Morph+Mol Monophyletic Bayesian Inference Mol Polyphyletic Morph+Mol Monophyletic Maximum Likelihood Mol Monophyletic Morph+Mol Monophyletic core Coenagrionidae Parsimony Mol Polyphyletic Morph+Mol Monophyletic Bayesian Inference Mol Monophyletic Morph+Mol Paraphyletic Maximum Likelihood Mol Monophyletic Morph+Mol Monophyletic ridge-faced Coenagrionidae Parsimony Mol Polyphyletic Morph+Mol Paraphyletic Bayesian Inference Mol Polyphyletic Morph+Mol Monophyletic Maximum Likelihood Mol Monophyletic Morph+Mol Monophyletic Ischnurinae Parsimony Mol Monophyletic Morph+Mol Monophyletic Bayesian Inference Mol Monophyletic Morph+Mol Monophyletic Maximum Likelihood Mol Monophyletic Morph+Mol Monophyletic Teinobasinae Parsimony Mol Paraphyletic Morph+Mol Paraphyletic Bayesian Inference Mol Polyphyletic Morph+Mol Polyphyletic Maximum Likelihood Mol Paraphyletic Morph+Mol Paraphyletic Pseudostigmatinae Parsimony Mol Monophyletic Morph+Mol Monophyletic Bayesian Inference Mol Paraphyletic Morph+Mol Polyphyletic Maximum Likelihood Mol Polyphyletic Morph+Mol Monophyletic Protoneurinae Parsimony Mol Polyphyletic Morph+Mol Monophyletic Bayesian Inference Mol Paraphyletic Morph+Mol Monophyletic Maximum Likelihood Mol Monophyletic Morph+Mol Monophyletic Roppaneura clade Parsimony Mol Polyphyletic Morph+Mol Monophyletic Bayesian Inference Mol Paraphyletic Morph+Mol Polyphyletic Maximum Likelihood Mol Polyphyletic Morph+Mol Polyphyletic

SUPPLEMENTARY MATERIALS

FIGURES

Figure S1. Most molecular parsimonious tree obtained with implied weighting (K = 3) of the molecular datasets of Protoneurinae and outgroups (Fit = 219.731291, L = 3.067, CI = 0.27, RI = 0.39). Values in the branches correspond respectively to Bootstrap, Jackknife, and relative Bremer above 50%.

TABLE FILE

Table S1. Morphological data matrix of Protoneurinae and outgroups with 49 terminals, and 46 characters. ? = missing character.

Species 1-10 11-20 21-30 31-40 41-46 Acanthagrion aepiolum 0010?0??00 0012011?00 1000110000 0????00?00 00?0?? Aceratobasis macilenta 011??0??00 0112001?00 1?01110001 000??00??? ???0?? Agriocnemis pygmaea 001??0???0 00???0???? ??0?1????? ?????????? ?????? Amazoneura ephippigera 000010?001 1110101011 100?010100 1110020011 11?01? Argia moesta 001??0??10 001200?100 1000?????? 0????0???? ???0?? Argia mollis 011??0??1? 0?120???00 1000?????? 0????0???? ???0?? Copera marginipes 00110??010 001?0100?? 1100010001 000?000000 00???? Drepanoneura muzoni 110??0?201 ?1001?1111 ??0010???? 0?1??000?? ???0?? Elattoneura glauca 00110??211 101?00?00? 000?010010 000?000011 00?0?? Enallagma cyanthigerum 0010?0??00 0012011?00 1000?????? 0?0???0??? ???0?? Epipleoneura lamina 110000?001 1110000011 1000111200 0110001011 01?01? Epipleoneura metallica 1100?0?001 1110020111 1000101200 001000101? ???0?? Epipleoneura venezuelensis 110000?001 1110000111 1000111200 0110001011 01?01? Forcepsioneura aff. lucia 1100?11?01 11??1??111 1100011??? 1110010110 01?0?? Forcepsioneura gabriela 1100?11?01 11101?1111 1101011100 1110020110 01?0?? Forcepsioneura garrisoni 1100?11?01 110?12?111 1101011?00 1110010110 01?0?? Forcepsioneura janeae 1100?11?01 111?1?1111 1101011100 1110010110 01?0?? Forcepsioneura regua 1100?0??01 11011??111 1101011100 1110020110 01?0?? Forcepsioneura sancta 0100111001 1100101111 1001010100 1110010011 01?01? Forcepsioneura serrabonita 1100?11?01 11101??111 1101011100 1110010110 01?0?? Forcepsioneura sp. 1100????01 1???1?1??? ??0???1??? 1?????0??? ???0?? Hetaerina americana 0000?0???? 001211?100 111??110?? 1011?00??0 01?0?? Heteragrion aurantiacum 010??0?000 0012100?00 01010??0?? 110?000000 00?1?? Idioneura ancilla 010000?110 1010000111 1001110200 0011010111 010000 Idioneura celioi 0100?0?001 1010020111 10011102?? 0110010100 01?0?? Ischnura elegans 0010????00 0012011?00 1?000001?? 0?0??00??? ???0?? Isosticta robustior 0000?0?0?0 001?000??? ??0??????? 0?1010?110 ?1?0?? Leptagrion andromache 011??0?000 0012001?00 1000010??1 01???0010? ?1?0?? Leptagrion elongatum 011??0?000 0012001?00 1000010??1 01??000110 00?0?? Lestes forficula 0010?0?0?? 0?12?00?00 11?0?1???? 110?000000 00?0?? Mecistogaster amalia 0110?0???0 01120???00 1100010?00 0?0?000?10 00?0?? Platystigma asticta 0110?0???0 01120???00 1100??0?00 010?000010 00?0?? Metaleptobasis selysi 0110?0?200 001200?100 010?110200 110?000120 10?0?? Neoneura confundens 1100?0?001 1110020000 100001???? 0110020000 00?0?? Neoneura fulvicollis 110000?000 1010010100 100001???? 0010030022 00?01? Neoneura maria 1100?0?001 1110020000 100001???? 0110000000 00?0?? Nososticta solida 001?????11 101???0??? ??0??????? ?????????? ?????? Oxyagrion terminale 001??0??00 00120?1??? 1000010210 010?000000 ???0?? Peristicta aeneoviridis 110000?000 1011010111 1000010100 1111110111 010010 Platycnemis pennipes 0010?0???0 001??1???? ??0??????? ?????????? ?????? 85

Protoneura capillaris 110000?001 1010000111 0000010210 0011000011 011010 Protoneura scintilla 1100?0?001 1010020111 0000010210 0010000010 0??0?? Psaironeura tenuissima 0100?00001 110?121011 1001010210 0011030110 00?0?? Pseudagrion pilidorsum 0010?????0 00???1???? ??0??????? ?????????? ?????? Roppaneura beckeri 010000?000 1111001111 1000010210 1110020000 01???? Selysioneura capreola 00000??001 100?10?10? 100?000000 00??000000 002111 Teinobasis scintillans 001??0???0 00???1?1?? ??0??????? ?????????? ?????? Telebasis griffinii 011??0?011 00120?1?00 1000110201 0?0?000013 00?0?? Telebasis obsoleta 011??0?011 0012001?00 1000110201 000?0000?? ?1?0??

List of specimens of Protoneurinae and outgroups studied for morphological character coding and DNA extractions.

Lestidae

Lestinae

Lestes forficula Rambur, 1842: ♂, BRAZIL. Rio de Janeiro State, Nova Friburgo municipality, "Distrito" de Salinas, Fazenda Campestre, Expedição Aracnobiosis - EAB_01 - Sede Fazenda Campestre (22°21'32.45"S 42°40'38.01"W, 1074 m a.s.l.), 10– 11.iv.2015, A.P. Pinto, A.B. Kury, A. Garcia & M.A. Medrano leg. (DZRJ 3542, DNA voucher ENT 2789).

Heteragrionidae

Heteragrion aurantiacum Selys, 1862: ♂, BRAZIL. Paraná State, Antonina municipality, RPPN Reserva Natural Guaricica (SPVS), Rio do Turvo - Trilha dos Fornos (LABIA-074) (25°17'22.13"S, 48°39'12.19"W, 88 m a.s.l.), 18.iv.2018, A.P. Pinto leg. (DNA voucher ENT 4608)

Coenagrionidae

Argiinae

Argia mollis Hagen in Selys, 1865: ♂, BRAZIL. Distrito Federal, Brasília municipality, Reserva Ecológica do IBGE (RECOR), Recor_1 - Fonte (Represa e Riacho) (15°56'55.22"S 47°52'07.23"W, 1119 m a.s.l.), 11–14.x.2013, A.P. Pinto & J.G. Da Silva leg. (DZRJ 280, DNA voucher ENT 4606)

Ischnurinae

Acanthagrion aepiolum Tenessen, 2004: ♂, BRAZIL. Distrito Federal, Brasília municipality, Reserva Ecológica do IBGE (RECOR), Recor_1 - Fonte (Represa e Riacho) (15°56'55.22"S 47°52'07.23"W, 1119 m a.s.l.), 11–14.x.2013, A. P. Pinto & J.G. Da Silva leg. (DZRJ 2290, DNA voucher ENT 3398).

Oxyagrion terminale Selys, 1876: ♂, BRAZIL. Distrito Federal, Brasília municipality, Reserva Ecológica do IBGE (RECOR), Recor_1 - Fonte (Represa e Riacho) (15°56'55.22"S 47°52'07.23"W, 1119 m a.s.l.), 11–14.x.2013, A. P. Pinto & J.G. Da Silva leg. (DZRJ 2264, DNA voucher ENT 4611)

Pseudostigmatinae

Leptagrion andromache Hagen in Selys, 1876: ♀, BRAZIL. Rio de Janeiro State, Parati municipality, Praia de Tarituba, prática de campo Ecologia, UEZO luminosa e ativa (23°02‘22.90‖S 44°35‘56.49‖W, 35 m a.s.l.), 19–21.ix.2013, A.P. Pinto & R.A. Carvalho leg. (DZRJ 317, DNA voucher ENT 2788)

Leptagrion elongatum Selys, 1876: ♂, BRAZIL. Rio de Janeiro State, Itatiaia municipality, Parque Nacional do Itatiaia (PNI), BIOTA FAPERJ Coleta-05, Ativa, Setor Lago Azul, PNI-M3, riachos, canais artificias, áreas de represamento e poças entre M3A e B (22°27'01.10"S 44°36'55.30"W, 830 m a.s.l.), 03.xi.2015, A.P. Pinto leg. (MNRJ 155, DNA voucher ENT 3408)

Mecistogaster amalia (Burmeister, 1839): ♂, BRAZIL. Rio de Janeiro State, Rio de Janeiro municipality, Parque Nacional da Floresta da Tijuca, Trilha da Cova da Onça (22°57'27.24"S 43°17'23.69"W, 510 m a.s.l.), 12.ii.2011, A.P. Pinto leg. (DZRJ 1344, DNA voucher ENT 2860)

Platystigma asticta (Selys, 1860): ♂, BRAZIL. Rio de Janeiro State, Itatiaia municipality, Parque Nacional do Itatiaia (PNI), Coleta-04, Malaise, Complexo do 88

Maromba, Travessia Ruy Braga, PNI-M2B (22°26'07.50"S 44°37'33.20"W, 1234 m a.s.l.), 02.x–02.xi.2015, BIOTA FAPERJ leg. (MNRJ 143, DNA voucher ENT 3406)

Teinobasinae

Aceratobasis macilenta (Rambur, 1842): ♀, BRAZIL. Paraná State, Antonina municipality, RPPN Reserva Natural Guaricica, trilha da Guaricica (LABIA 037) (25°18'53"S, 48°41'46"W, 6 m a.s.l.), 10.iii.2018, A.P. Pinto leg. (DZUP 499419, DNA voucher ENT 4607)

Metaleptobasis selysi Santos, 1956: ♂, BRAZIL. Bahia State, Una municipality, Reserva Biológica de Una, Expedição Gabriela Cravo e Canela IV, first order stream [?], after Fazenda Piedade (15°09′36.2″S, 39°10′31.1″W, 100 m a.s.l.), 07.viii.2016, A.P. Pinto, A.P.M. Santos, D.M. Takiya & P.M. Souto leg. (DZUP 500171) (DNA voucher ENT 3446)

Telebasis griffinii (Martin, 1896): ♂, BRAZIL: Paraná State, Antonina municipality, RPPN Reserva Natural Guaricica (SPVS), Estrada (DNA GU 009) (25°18'48"S, 48°40'37"W, 13 m a.s.l.), 24.x.2018, A.P. Pinto leg. (DZUP 500164, DNA voucher ENT 4612)

Protoneurinae

Epipleoneura metallica Rácenis, 1955: ♂, BRAZIL. Maranhão State, [Mirador] municipality, [Parque Estadual Mirador, Riacho Pindaíba], Malaise (6°39'05"S 45°1'39"W), 25–27.ix. 2014, W.R.M. Souza & T.T. Andrade leg. (DZRJ 3775, DNA voucher ENT 2790)

Epipleoneura venezuelensis Rácenis, 1955: ♂, BRAZIL. Bahia State, Una municipality, Reserva Biológica de Una, "Expedição Gabriela Cravo e Canela III", RUNA_02 - Rio Maruim ponto jangada (15°10'29.43"S 39°03'23.98"W, 23 m a.s.l.), 12–13.vi.2015, A.P. Pinto leg. (DZRJ 3792, DNA voucher ENT 4357) 89

Forcepsioneura gabriela Pimenta, Pinto & Takiya, 2019: Holotype ♂, BRAZIL. Bahia State, Una municipality, Reserva Biológica de Una, Expedição Gabriela Cravo e Canela II, first order stream [?], after Fazenda Piedade (15°09′36.2″S 39°10′31.1″W, 100 m a.s.l.), 14–15.vi.2014, A.P. Pinto leg. (DZUP 498858, DNA voucher ENT 2784). Paratype ♂, same data as holotype but Expedição Gabriela Cravo e Canela IV, 07.viii.2016, A.P. Pinto, A.P.M. Santos, D.M. Takiya & P.M. Souto leg. (DZRJ 3555; DNA voucher ENT 3445).

Forcepsioneura garrisoni Lencioni, 1999: ♂, BRAZIL. Rio de Janeiro State, Parati municipality, Praia de Itaituba, Prática de Campo Ecologia, UEZO Luminosa e ativa, Ponto luminosa (23°02′22.90″S 44°35′56.49″W, 35 m a.s.l.), 19–21.xi.2013, A.P. Pinto & R.A. Carvalho leg. (DZRJ 0325, DNA voucher ENT 2787). ♂. Paraná State, Antonina, RPPN Reserva Natural Guaricica (SPVS), Rio do Turvo - Trilha dos Fornos (25°17'22.13"S, 48°39'12.19"W, 88 m a.s.l.), 23-27.x.2017, A.P. Pinto leg (DZUP 500169, DNA voucher ENT 4344)

Forcepsioneura janeae Pimenta, Pinto & Takiya, 2019: Paratype ♂, BRAZIL. Espírito Santo State, Santa Teresa municipality, [Biological Station of Santa Lúcia], (MNRJ 0141; DNA voucher ENT 3403).

Forcepsioneura aff. lucia: ♂, BRAZIL. Rio de Janeiro State, Itatiaia municipality, Parque Nacional do Itatiaia (PNI), BIOTA FAPERJ Coleta-06, Ativa, Complexo do Maromba (PNI-M2), pequenas nascentes sobre afloramento rochoso, formando pequenos riachos na trilha para a Cachoeira Itaporani, Rio Campo Belo (22°25′38.27″S 44°37‘12.79‖W, 1150 m a.s.l.), 06.i.2016, A.P. Pinto leg. (MNRJ 210, DNA voucher ENT 3616). Same data but Parque Nacional do Itatiaia (PNI), BIOTA FAPERJ Coleta- 05, Ativa, Setor Lago Azul, PNI-M3, riachos, canais artificias, áreas de represamento e poças entre M3A e B (22°27′01.10″S 44°36′55.30″W 830 m a.s.l.), 02.xi.2015, A.P. Pinto leg. (MNRJ 0147, DNA voucher ENT 3404)

Forcepsioneura regua Pinto & Kompier, 2018:2 Paratype ♂, BRAZIL. Rio de Janeiro State, Cachoeiras de Macacu municipality, Reserva Ecológica de Guapiaçu (REGUA), 90

forest fragment (22°28′04″S, 42°45′32″W, 70 m a.s.l.), 13.i.2014, T.M.F. Kompier leg. (DZRJ 2251, DNA voucher ENT 2854). Same data but açude com macrófitas (22°27′10.11″S 42°46′13.01″W, 34 m a.s.l), 03.xii.2009, A.L. Carvalho & Equipe Disc. Téc. Coleta leg. (DZRJ 315, DNA voucher ENT 3609)

Forcepsionera sancta (Hagen in Selys, 1860): ♂, BRAZIL. Rio de Janeiro State, Nova Friburgo municipality, [Fazenda] Campestre, Salinas (22°22'07.42"S 42°40'43.39"W), 11.i.2014, A.P. Pinto & T.M.F. Kompier leg. (DZRJ 2453, DNA voucher ENT 2369)

Forcepsioneura serrabonita Pinto & Kompier, 2018: Paratype ♂, BRAZIL. Bahia State, Camacan municipality, [Reserva Particular do Patrimônio Natural (RPPN) Serra Bonita], first trickle on road beyond managers house (Quarry Stream) (15°23'19"S, 39°33'57"W, 710 m a.s.l. [822 m]), 20.ii.2015, C.M. Flint & O.S. Flint Jr. leg. (MZSP ODO-492, DNA voucher ENT 2857).

Forcepsioneura sp.: ♂, BRAZIL: Rio Grande do Sul State, Maquiné municipality, Distrito de Barra do Ouro, Rio Maquiné - diversos pontos ao longo da estrada e trilhas para a Cascata Forqueta (29°32'00.24"S, 50°12'36.45"W, 200 m a.s.l.), 22.xii.2016, A.P. Pinto & G.P. Cassel leg. (DZUP 500165, DNA voucher ENT 3812).

Idioneura ancilla Selys, 1860: ♂, BRAZIL: Paraná State, Antonina municipality, RPPN Reserva Natural Guaricica, trilha da Guaricica (SPVS-004) (25°18'53"S, 48°41'46"W, 6 m a.s.l.), 10.iii.2017, A.P. Pinto leg. (DZUP 500165, DNA voucher ENT 3447)

Idioneura celioi Lencioni, 2009: ♂, BRAZIL. São Paulo State, Cananéia municipality, Parque Estadual da Ilha do Cardoso (PEIC_Aeschnosoma_1) (25°05'07.38"S 47°55'39.91"W, 12 m a.s.l.), 20.xi.2011, A.P. Pinto leg. (DZRJ 500166, DNA voucher ENT 2791)

Neoneura confundens Wassher & Van ‗t Bosch, 2013: ♂, BRAZIL. Bahia State, Una municipality, Reserva Biológica de Una, "Expedição Gabriela Cravo e Canela III", RUNA_02 - Rio Maruim ponto jangada (15°10'29.43"S 39°03'23.98"W, 23 m a.s.l.), 12–13.vi.2015, A.P. Pinto leg. (DZRJ 3791, DNA voucher ENT 3399) 91

Peristicta aeneoviridis Calvert, 1909: ♂, BRAZIL. Rio Grande do Sul State, Cambará do Sul municipality, Parque Nacional dos Aparados da Serra (PNAS), EGA 05 - Malaise, Arroio Preá, a montante da ponte, riacho de 1° ordem (29°09'49.66"S 50°05'51.13"W, 924 m a.s.l.), 08–10.ii.2014, A.P. Pinto & J.G. Da Silva leg. (DZRJ 3545, DNA voucher ENT 2858). ♂ BRASIL: Rio Grande do Sul State, Maquiné municipality, Distrito de Barra do Ouro, Rio Maquiné - diversos pontos ao longo da estrada e trilhas para a Cascata Forqueta (29°32'00.24"S, 50°12'36.45"W, 200 m a.s.l.), 22.xii.2016, A.P. Pinto & G.P. Cassel leg. (DZUP 500170, DNA voucher ENT 4342).

Psaironeura tenuissima (Selys, 1886): ♂, BRAZIL. Amazonas State, Manaus municipality, ZF2 Km 14, 2°35,24‖S, 60°06‘55‖W, 6-18.VIII.2016, malaise pequena Igarapé perto torre, J.A. Rafael & F.F. Xavier leg. (LABIA 106) (DZUP 500167 DNA voucher ENT 4613)

Protoneura capillaris (Rambur, 1842): ♂, CUBA. Santiago de Cuba State (20.01°N, 75.82°W), López leg. (DZUP 500168, DNA voucher ENT 4614)

Roppaneura beckeri Santos, 1966: ♂, BRAZIL. Paraná State, Curitiba municipality, Campus Centro Politécnico, borda de bosque na área de drenagem dos campos de futebol da Educação Física, área alagadiça com Apiacae, Eryngium sp. (25°27'11"S, 49°14'06"W, 914 m a.s.l.), 14.X.2017, A.P. Pinto leg. (DZUP 500083, DNA voucher ENT 4338). 92

3. Conclusões Gerais

A presente tese conseguiu alcançar seus principais objetivos e além disso, contribuiu para o conhecimento da filogenia de Protoneurinae. Dados moleculares inéditos foram obtidos e uma hipótese de relacionamento da subfamília foi proposta. Esses resultados foram possíveis através da inclusão de métodos integrativos com diferentes origens de dados. No primeiro capítulo foram propostas duas espécies novas de Forcepsioneura, delimitadas de uma forma integrada com base em dados morfológicos e moleculares. Foi proposta a primeira hipótese filogenética molecular do gênero, abrangendo sete das dez espécies descritas. Essa análise, utilizando diferentes marcadores moleculares, resultou em uma hipótese sobre o relacionamento filogenético de Forcepsioneura e evidenciou a existência de dois clados internos de espécies, corroborando com observações publicadas anteriormente. Baseada em evidência total, o segundo capítulo apresentou uma hipótese filogenética integrativa de Protoneurinae. Foram explorados a relação entre os gêneros de Protoneurinae e também o seu posicionamento em Coenagrionidae. O relacionamento interno dos gêneros de Protoneurinae foi discutido, incluindo a maior amostragem de caracteres (morfológicos e moleculares) já realizada até então. Devido ao uso de diferentes metodologias foi possível explorar possíveis relacionamentos alternativos e hipóteses de homologia dos caracteres morfológicos. Além disso, com a nossa amostragem taxonômica, foi possível testar o monofiletismo de grupos de Coenagrionidae, como os clados de espécies de fronte angulada e fronte arredondada e também do clado Roppaneura que foi evidenciado por análises publicadas posteriormente.

4. Referências gerais

BYBEE, S.M., ODGEN, T.H., BRANHAM, M.A., WHITING, M.F. 2008. Molecules, morphology and fossils: a comprehensive approach to odonate phylogeny and the evolution of the odonate wing. – Cladistics 24(4): 477–514. doi: 10.1111/j.1096- 0031.2007.00191.x

CARLE, F.L., KJER, K.M., MAY, M.L. 2008. Evolution of Odonata, with Special Reference to Coenagrionidae (Zygoptera). – Arthropod Systematics & Phylogeny 66(1): 37–44.

CORBET, P.S. 1980. Biology of Odonata. – Annual. Review of Entomololy 25:189–217 Davies, D.A.L., and P. Tobin. 1984. The dragonfiies of the world: a systematic list of the extant species of Odonata. Volume 1. Zygoptera, Anisozygoptera. Societas Internationalis Odonatologica Rapid Communications 3: ix + 127 p. DAYRAT, B. 2005. Towards integrative taxonomy. – Biological Journal of the Linnean

Society 85: 407–415. DIJKSTRA, K.-D.B., BECHLY, G., BYBEE, S.M. ET AL. 2013. The classification and diversity of dragonflies and damselflies (Odonata). Animal Biodiversity: An Outline of Higher-level Classification and Survey of Taxonomic Richness, (ed. by Z.-Q. Zhang). Zootaxa in press.

DIJKSTRA, K-D.B., KALKMAN, V.J, DOW, R.A., STOKVIS, F.R., VAN TOL, J. 2014. Redefining the damselfly families: the first comprehensive molecular phylogeny of Zygoptera (Odonata). – Systematic Entomology 39: 68–96. doi: 10.1111/syen.12035 DUMONT, H.J., VIERSTRAETE A., VANFLETEREN J.R. 2010. A molecular phylogeny of the Odonata (Insecta). – Systematic Entomology 35: 6–18. FRASER, F.C. 1957. A Reclassification of the Order Odonata. Royal Zoological Society of New South Wales, Syndey.

HASEGAWA, E., KASUYA, E. 2006. Phylogenetic analysis of the insect order Odonata using 28S and 16S rDNA sequences: a comparison between data sets with different evolutionary rates. – Entomological Science 9: 55–66 LENCIONI, F.A.A.1999. The genus Phasmoneura, with description of Forcepsioneura gen. nov. and two new species. – Odonatologica 28(2): 127–137 94

LIEFTINCK, M.A. 1975. The dragonfly (Odonata) of New Caledonia and the Loyalty Island Par. 1. Imagines. Cahiers de l‘Office de la Recherche Scientifique et Technique Outre-Mer. – Seire Hydrobiologie 9: 127–166 MACHADO, A.B.M. 1985. Studies on neotropical Protoneuridae. 5. Redescription of the type of Phasmoneura ephippigera (Selys, 1886) (Zygoptera). – Odonatologica 14(4): 363–368

MACHADO, A.B.M. 1999. Studies on neotropical Protoneuridae 9. Phasmoneura ciganae Santos, conspecific with Phasmoneura sancta (Hagen) comb. nov. (Zygoptera). – Notulae Odonatologica 5(3): 37–38

MACHADO, A.B.M. 2004. Studies on Neotropical Protoneuridae. 15. Amazoneura gen. nov. with description of A. juruaensis sp. nov. (Odonata: Zygoptera). – Revista Brasileira de Zoologia 21(2): 333–336. doi: 10.1590/S0101-81752004000200029

O‘GRADY, E.W., MAY, M.L. 2003. A phylogenetic reassessment of the subfamilies of Coenagrionidae (Odonata: Zygoptera). – Journal of Natural History 37: 2807– 2834 PADIAL, J.M., MIRALLES, A., DE LA RIVA, I., VENCES, M. 2010. The integrative future of taxonomy. – Frontiers in Zoology 7: 1–16

PESSACQ, P. 2008. Phylogeny of neotropical Protoneuridae (Odonata: Zygoptera) and a preliminary study of their relationship with related families. – Systematic Entomology 33: 511–528. doi: 10.1111/j.1365-3113.2007.00414.x PESSACQ, P., SANTOS, T.C., COSTA, J.M. 2012. Checklist and update distribution of Protoneuridae from Brazil. – International Journal of Odonatology 15: 59–73

PIMENTA, A.L., PINTO, A.P, TAKIYA, D.M. (2019). Integrative taxonomy and phylogeny of the damselfly genus Forcepsioneura (Odonata: Coenagrionidae: Protoneurinae) with description of two new species from the Brazilian Atlantic Forest. – Arthopod Systematics & Phylogeny. In press.

PINTO, A.P., KOMPIER, T. 2018. In honor of conservation of the Brazilian Atlantic Forest: description of two new damselflies of the genus Forcepsioneura discovered in private protected areas (Odonata: Coenagrionidae). – Zoologia 35 (e21351): 1–19. doi: 10.3897/zoologia.35.e21351 REHN, A.C. 2003. Phylogenetic analysis of higher-level relationships of Odonata. – Systematic Entomology 28:181–239

SCHLICK-STEINER, B.C., FLORIAN, S.M., SEIFERT, B., STAUFFER, C., CHRISTIAN E., CROZIER, R.H. 2010. Integrative Taxonomy: A Multisource Approach to 95

Exploring Biodiversity Annual Review of Entomology 55:421–38. doi:10.1146/annurev-ento-112408-085432

SCHORR, M., PAULSON, D. 2019. World Odonata List. – URL

WATSON, J.A.L. 1992. The subfamilies of Protoneuridae (Zygoptera). – Odonatologica, 21: 195–201.