Chiroptera, Mammalia): Alometria, Taxonomia E Morfologia Funcional

UNIVERSIDADE FEDERAL DO RIO DE JANEIRO

Instituto de Biologia

Nathália Siqueira Veríssimo Louzada

Análise comparada do fêmur em Yangochiroptera (Chiroptera, Mammalia): alometria, taxonomia e morfologia funcional

Rio de Janeiro Março de 2019

UNIVERSIDADE FEDERAL DO RIO DE JANEIRO

Análise comparada do fêmur em Yangochiroptera (Chiroptera, Mammalia): alometria, taxonomia e morfologia funcional

Nathália Siqueira Veríssimo Louzada

Tese de doutorado apresentada ao Programa de Pós- graduação em Ciências Biológicas (Biodiversidade e Biologia Evolutiva), Instituto de Biologia, Universidade Federal do Rio de Janeiro, como requisito parcial à obtenção do título de Doutor em Ciências Biológicas (Biodiversidade e Biologia Evolutiva).

Orientadora: Dra. Leila Maria Pessôa Coorientador: Dr. Marcelo Rodrigues Nogueira

Rio de Janeiro Março de 2019 i

Louzada, Nathália Siqueira Veríssimo Análise comparada do fêmur em Yangochiroptera (Chiroptera, Mammalia): alometria, taxonomia e morfologia funcional/ Nathália Siqueira Veríssimo Louzada. Rio de Janeiro: UFRJ/IB, 2019. xvii + 107 fls. Orientadores: Dra. Leila Maria Pessôa e Dr. Marcelo Rodrigues Nogueira.

Tese (Doutorado) – Universidade Federal do Rio de Janeiro, Instituto de Biologia, Programa de Pós-graduação em Ciências Biológicas (Biodiversidade e Biologia Evolutiva), 2019. Referências bibliográficas: f. 101-107.

1. Anatomia. 2. Membros posteriores. 3. Métodos comparados. 4. Morfologia funcional. 5. Robustez do fêmur. I. Pessôa, Leila Maria; Nogueira, Marcelo Rodrigues. II. Universidade Federal do Rio de Janeiro, Instituto de Biologia, Programa de Pós-graduação em Ciências Biológicas (Biodiversidade e Biologia Evolutiva). III. Análise comparada do fêmur em Yangochiroptera

(Chiroptera, Mammalia): alometria, taxonomia e morfologia funcional.

iii

It means no worries

For the rest of your days

It's our problem-free philosophy

Hakuna Matata!

The Lion King

À minha amada mãe, Maria Christina, e ao meu querido irmão, Ronaldo Júnior.

AGRADECIMENTOS

Agradeço a todos os meus familiares, que sempre estiveram presentes nessa jornada e que constituem uma família pequena, porém unida e repleta de amor. Agradeço à minha mãe, Maria Christina Siqueira Veríssimo Louzada, que esteve me dando forças em todos os momentos do curso, que sorriu comigo a cada conquista e me reergueu nos momentos mais difíceis, sempre demonstrando muito amor e carinho; que, apesar de achar estranho estudar morcegos, sempre esteve preocupada em entender a importância dos mesmos; que sempre teve orgulho de dizer “A minha filha é bióloga, futura doutora”. Nem todas as palavras boas do mundo seriam suficientes para expressar meu amor por você. Obrigada por tudo!

Ao meu irmão, Ronaldo de Miranda Louzada Júnior, o “manymol” mais chato do mundo! Te amo tanto! Obrigada por ser uma pessoa tão maravilhosa que, mesmo em meio a momentos de aflição enfrentados ao longo do caminho, demonstrava apoio e, ao mesmo tempo, fazia brincadeiras, tornando tudo mais leve e descontraído. Sua força e determinação sempre foram uma inspiração para mim, um exemplo a seguir. Obrigada pelo carinho e preocupação comigo, pelo apoio aos meus objetivos e por se mostrar sempre presente na minha vida.

À minha orientadora, Dra. Leila Maria Pessôa, que está comigo desde 2009, quando me recebeu no laboratório, e desde então tem me ensinado tanto. Passamos por muita coisa juntas, graduação, JIC, mestrado, seminários, congressos, publicações, doutorado. Obrigada por ter sido excepcional durante esses dez anos, por ter me ensinado tantas coisas, por ter me apoiado nos momentos difíceis e ter compartilhado as alegrias dos momentos bons, que foram muitos. Obrigada por confiar em mim, na minha capacidade e no meu trabalho. Obrigada por me ajudar a crescer profissionalmente e a atingir tantas metas. Obrigada por sempre ter me incentivado a melhorar e evoluir. Tenho um carinho imenso por você, que sempre me inspirou e me ajudou a chegar tão longe. Obrigada por compartilhar comigo mais esse momento.

Ao meu coorientador, Dr. Marcelo Rodrigues Nogueira, por ter me ensinado tanto nesses quatro anos. O conhecimento que adquiri nessa trajetória é imensurável. Obrigada pelo apoio nas decisões difíceis, pelo incentivo a uma jornada numa área nova (anatomia) e pelo aprendizado de tantas técnicas e análises. Obrigada pelos dias no museu, por compartilhar a empolgação com a anatomia do fêmur, pelas conversas sobre morfologia funcional. Obrigada por sempre estar disponível para trocar conhecimento, conversar sobre morcegos, tirar dúvidas

e dar dicas. Obrigada por você, que sempre foi uma inspiração para mim, ter compartilhado comigo as conquistas oriundas dessa orientação.

Agradeço imensamente às minhas primas, Mariana Siqueira Veríssimo Palatnic e Iris Maria Deschaumes, pela parceria, pela união. Vocês sempre estiveram ao meu lado durante essa trajetória, conferindo um apoio imensurável nas partes mais difíceis dela e compartilhando os melhores momentos comigo. Obrigada por me amarem, mesmo eu sendo a prima estranha que pega e tenta salvar todos os insetos que aparecem em qualquer lugar, que gosta de levar pra trilhas e aventuras, às vezes não tão fáceis. Obrigada pelas noites de conversas, pelos dias de comer besteira. Com vocês, até uma simples reunião no sofá com mil doces e um filme aleatório, se torna mais que especial! Amo vocês! À Mariana por ser sempre uma amiga verdadeira, uma irmã. Por conversar comigo nos momentos bons e ruins, por me ajudar a crescer e ser o que sou, por me dar alegria apenas por estar comigo, por me ajudar a me levantar nos momentos difíceis e, sem dúvida, por chegar até aqui. Te amo, jeguinha! Mesmo você não tendo molhado o cabelo naquele dia da piscina e não ter apoiado a ideia de se vestir de fantasma para assustar os outros. À Iris, agradeço pelo carinho incondicional. Obrigada por ser uma pessoa tão boa, tão fofa, tão tudo. Aposto que, se unicórnios existissem, estariam disfarçados de você! Obrigada por todos os doces, brownies e bolos, que fizeram diversos momentos mais alegres. Obrigada pelas risadas e momentos especiais vividos. Obrigada por ser uma musa ecológica e ter tanto carinho pela natureza, pela conservação; sempre pensando nas atitudes que podem minimizar os danos ao ambiente, me inspirando a ser uma pessoa melhor. Agradeço por toda ajuda com a revisão do inglês, por me ensinar tantas coisas e por se disponibilizar em momentos tão importantes da minha vida! Amo vocês duas incondicionalmente, para sempre.

Agradeço a todos os meus tios e tias. Em especial, agradeço ao meu padrinho, Paulo Francisco Siqueira Veríssimo, que me dava amor e carinho, mesmo de longe, e que, acima de tudo, tinha orgulho de mim e do que eu fazia. Tenho certeza de que agora ele está olhando por mim e muito feliz por mais essa conquista. À minha madrinha, Maria Inês Siqueira Veríssimo, por ter muito orgulho de mim e demonstrar isso a todo momento; por me apoiar em momentos difíceis e demonstrar preocupação a cada passo meu; por desde pequena, ter me levado para fazendinhas e ter me colocado em contato com a natureza, lugar que atualmente amo estar. Ao meu primo Vinícius (Brico) e à princesa da Lulu, que é muito comunicativa e transborda alegria por onde passa. À Margarida Maria Siqueira Veríssimo por ter sempre aberto o lar para mim e para a família, onde compartilhamos diversos momentos únicos; por ser uma pessoa muito

vii

animada, organizando reuniões familiares, ou ajudando em outros eventos. Ao Luis Antônio Siqueira Veríssimo, tio Luck, à Rita de Cássia Rodrigues de Castro Veríssimo, tia Rita, e ao meu primo Guigui (Luiz Guilherme), sempre interessados e preocupados com os animais, mostrando curiosidade pelos mesmos, inclusive morcegos. À tia Maria dos Anjos Coelho que, mesmo morando longe, está presente no meu coração. Obrigada pelas conversas sobre biologia, sobre o interesse pela área e por proporcionar momentos únicos e felizes, quando juntas. À Gabriela Coelho Veríssimo, prima Gabi, que tem o sorriso mais lindo desse mundo e transborda good vibes; obrigada pelo compartilhamento de momentos únicos e engraçados, tenho certeza que mais e mais virão! Amo todos vocês.

À minha cunhada, Lidia Sabanef, que faz parte da nossa família, que sempre foi tão boa e carinhosa com todos; que proporcionou momentos especiais mostrando bichos que caíram na sua varanda ou que apareceram no seu condomínio, sempre procurando me informar de tudo que ocorria no mundo animal. Que, além de cunhada, é uma amiga especial, que ajudou muito e cuidou de mim em diversos momentos. Obrigada por sempre ter tanto carinho comigo e com nossa família, amo você!

Ao meu pai, Ronaldo de Miranda Louzada, que conversou muito comigo durante essa trajetória, que se preocupou com meu futuro e se disponibilizou a ajudar no alcance das minhas metas. Obrigada pelas conversas e carinho.

Agradeço à Anne Caruliny do Monte Lima, amiga e irmã, que me deu apoio durante toda a jornada, que me ajudou nos momentos de aflição e que foi paciente ao ler diversas vezes artigos ou partes dos trabalhos que eu, preocupada, enviava; por ter compartilhado comigo o prazer de escrever dois capítulos sobre morcegos, por todos os momentos felizes e únicos, pelos momentos engraçados, pelos risos e sorrisos, por todas as longas conversas sobre assuntos acadêmicos ou não, por ser minha ecóloga favorita, por sempre cuidar de mim e estar presente nos momentos mais importantes da minha vida, sejam eles bons ou ruins. Nos ruins, por me ajudar a ver o lado bom de tudo; a crescer, vencer e superar. Nos bons, por estar comigo e compartilhar suas experiências mais incríveis. Espero e quero ainda dividir muitos momentos com você!

Agradeço ao Felipe de Melo Barreto Pereira, pelo apoio durante os momentos mais difíceis que enfrentei nesses quatro anos, por estar presente e me ajudar a resolvê-los. Pela

viii

companhia nos momentos bons, pelo estímulo à participação de eventos e congressos, por ajuda durante a logística de tais eventos.

Obrigada aos pesquisadores que me ajudaram de forma incisiva durante esses quatro anos de Doutorado. Ao Dr. João Alves Oliveira, por facilitar o acesso à coleção mastozoológica do Museu Nacional do Rio de Janeiro, pela confiança durante o uso do material e espaço, e pela disponibilidade constante em ajudar. Ao Dr. Leandro Oliveira Salles, que permitiu o acesso a parte dos espécimes, com os quais eu desenvolvi parte fundamental do trabalho. Ao Dr. Adriano Lúcio Peracchi pela doação de três espécimes utilizados no presente estudo. Ao Carlos Rodrigues de Moraes Neto com a ajuda na preparação de alguns fêmures de morcegos, bem como na elaboração do protocolo de extração dos mesmos. Ao Dr. Leandro Monteiro, pela contribuição nas análises multivariadas.

Agradeço aos membros da pós-graduação em Biodiversidade e Biologia Evolutiva, que sempre estiveram disponíveis a ajudar e melhorar o curso. À Dr. Claudia Russo e ao Sr. Heber Borges de Araújo, que foram solícitos ao ajudar e solucionar imprevistos relacionados à tese e à pós-graduação. Pela paciência e boa vontade em todos os momentos do curso. A todos os professores da pós-graduação, em especial João Alves Oliveira e Ricardo Moratelli Mendonça da Rocha que, com suas disciplinas, me ensinaram novos conteúdos, que contribuíram para o desenvolvimento da tese.

Um obrigada especial aos colegas de laboratório e do Museu Nacional, pela companhia e compartilhamento de momentos excepcionais. Ao Dr. William Corrêa Tavares, que sempre esteve disposto a ajudar, e que muito me ensinou, com seminários ou conversas. Ao colega Aldo Caccavo, pelos muitos momentos de conversas, sérias ou não, pelos campos, pela lembrança e carinho ao trazer bibliografias do AMNH. À Gisela Sobral, pelas conversas e empolgações em relação a projetos diferentes, pelo apoio e confiança em relação a congressos, projetos e artigos. Em especial, à Ana Pantaleão, Maíra Laeta e Natália Boroni, que com muita alegria, irreverência, companheirismo, alto astral e felicidade transformaram estes anos de Doutorado em anos mais leves e inesquecíveis.

Deixo expressa a minha gratidão a todos os amigos, melhores amigos, que estiveram unidos nos momentos bons e ruins, de alegrias e tristezas, ajudando uns aos outros. Destaco meus companheiros desde a graduação, Anne Caruliny do Monte Lima, Bárbara Cristina Francisco, Bruna Carla, Felipe Alvarenga Machado, Felipe Soares Coelho e Tamires Azamor. ix

Obrigada por fazerem parte da minha vida. Obrigada também ao Diogo Ribeiro, querido amigo, que me ajudou com a revisão do texto e que compartilhou tantos momentos de alegria. E à Aline Ferrari, pela amizade, pelas conversas, pelo incentivo e parceria.

Agradeço a todos os pesquisadores de todas as instituições que foram citadas neste trabalho, a todos os funcionários e professores da Universidade Federal do Rio de Janeiro (UFRJ) e do Museu Nacional (MN), que permitiram a visitação de suas coleções. E aos órgãos de fomento, que forneceram apoio financeiro. À Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ) e à Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES).

Agradeço à banca examinadora por aceitar a leitura e avaliação do presente trabalho.

E por fim, mas não menos importante, agradeço a Deus por permitir a realização deste trabalho, me dar fé e força nos momentos difíceis e por colocar pessoas boas e amigos verdadeiros no meu caminho.

RESUMO

Análise comparada do fêmur em Yangochiroptera (Chiroptera, Mammalia): alometria, taxonomia e morfologia funcional

Nathália Siqueira Veríssimo Louzada

Orientadores: Leila Maria Pessôa e Marcelo Rodrigues Nogueira

Resumo de tese de doutorado submetido ao Programa de Pós-graduação em Ciências Biológicas (Biodiversidade e Biologia Evolutiva), Instituto de Biologia, Universidade Federal do Rio de Janeiro, como requisito parcial à obtenção do título de Doutor em Ciências Biológicas (Biodiversidade e Biologia Evolutiva).

Mais conhecidos pela notável morfologia dos membros anteriores, os morcegos também são únicos entre os mamíferos em relação aos seus membros posteriores — suas pernas são rotacionadas em até 180º, geralmente reduzidas em tamanho. O fêmur é o principal osso da perna, mas até hoje poucos estudos com morcegos consideraram sua morfologia em detalhe. Yangochiroptera é a maior ordem de Chiroptera e é altamente diversificada em relação à morfologia, ecologia e comportamento, representando um bom modelo para análises comparadas. A presente tese teve como objetivo descrever a anatomia do fêmur em uma grande amostra de Yangochiroptera (125 espécies, 70 gêneros e 12 famílias), e explorar as principais tendências de variação morfológica e padrões alométricos neste osso. Parte da amostra foi obtida através da extração do fêmur de espécimes de coleções fluidas, cujo protocolo foi descrito em detalhe. Na descrição anatômica, foram utilizados 13 caracteres categóricos e, nas análises quantitativas, que incluíram a análise de componentes principais filogenética (ACPf) e regressões filogenéticas, cinco dimensões lineares. Com a utilização dos dados categóricos, as 12 famílias foram diagnosticadas e as da região neotropical foram incluídas em uma chave de identificação. A ACPf, mostrou que, além do tamanho, os principais eixos de variação no fêmur estão relacionados à robustez e à morfologia de sua cabeça. As regressões filogenéticas mostraram que, apesar do comprimento do fêmur apresentar alometria negativa em relação ao tamanho corporal, a largura e a robustez, apresentaram, respectivamente, isometria e alometria positiva. Isto refuta, então, um padrão geral esperado de redução dos membros posteriores em morcegos. Em nível de espécie, o resultado mais notável foi relacionado à Myotis simus, que apresentou o fêmur mais robusto para seu tamanho. A anatomia do fêmur permitiu o

agrupamento de algumas famílias com base no significado funcional de seus caracteres. Molossidae, Mystacinidae e Desmodontinae, por exemplo, possuem tubérculos e cristas posteriores no trocânter maior e cristas laterais longas ou deslocadas medial/distalmente na haste, características que permitem uma maior inserção e ação da musculatura do quadril e são importantes para a locomoção quadrúpede. Nossos resultados demonstraram que a morfologia do fêmur é muito variável em Yangochiroptera, reforçando seu potencial taxonômico e funcional.

Palavras-chave: chave de identificação; membros posteriores; métodos comparados; quadrupedalismo; robustez óssea.

Rio de Janeiro Março de 2019

xii

ABSTRACT

Análise comparada do fêmur em Yangochiroptera (Chiroptera, Mammalia): alometria, taxonomia e morfologia funcional

Nathália Siqueira Veríssimo Louzada

Orientadores: Leila Maria Pessôa e Marcelo Rodrigues Nogueira

Abstract de tese de doutorado submetido ao Programa de Pós-graduação em Ciências Biológicas (Biodiversidade e Biologia Evolutiva), Instituto de Biologia, Universidade Federal do Rio de Janeiro, como requisito parcial à obtenção do título de Doutor em Ciências Biológicas (Biodiversidade e Biologia Evolutiva).

Better known by their remarkable forelimb morphology, bats are also unique among mammals in respect to their hindlimbs—its legs are rotated up to 180º, generally reduced in size. The femur is the main leg bone, but to date few studies with bats have considered its morphology in detail. Yangochiroptera is the largest order of Chiroptera and is highly diversified in relation to morphology, ecology, and behavior, representing a good model for comparative analyzes. This study aimed to describe the anatomy of the femur in a large sample of yangochiropteran bats (125 species, 70 genera, and 12 families), and explore major trends of morphological variation and scaling patterns in this bone. Part of the sample was obtained by extracting the femur from specimens preserved in fluid, which protocol was described in detail. Thirteen categorical characters were used in the anatomical description and five linear dimensions were used in the quantitative analyses, which included a phylogenetic principal component analysis (pPCA) and phylogenetic regressions. Based on categorical data, the 12 families were diagnosed and those from the neotropical region were included in an identification key. The pPCA showed that, in addition to size, major axes of variation in bat femur are related to robusticity and head morphology. The phylogenetic regressions showed that, although the femur length exhibited negative allometry with body mass, the width and robusticity showed respectively isometry and positive allometry. This refutes an expected general pattern of hindlimbs reduction in bats. At species level, the most remarkable finding was related to Myotis simus, which presented the most robust femur for its size. The femur anatomy allowed the grouping of some families based on the functional meaning of their characters. Molossidae, Mystacinidae, and Desmodontinae, for example, have tubercles and posterior ridges on the

xiii

greater trochanter, and lateral ridges that are long or medially/distally displaced at the shaft, features that allow a greater insertion and action of the hip musculature, and are important to quadrupedal locomotion. Our results proved that the morphology of the femur is very variable in Yangochiroptera, reinforcing its taxonomic and functional potential.

Key words: comparative methods; hindlimbs; identification key; bone robusticity; quadrupedalism.

Rio de Janeiro Março de 2019

xiv

SUMÁRIO

1. Introdução ...... 1 1.1 Yangochiroptera ...... 2 1.2 O estudo do fêmur ...... 5 1.3 Taxonomia ...... 7 1.4 Alometria ...... 7 1.5 Morfologia funcional ...... 8 2. Objetivos ...... 12 2.1 Objetivo Geral ...... 13 2.2 Objetivos específicos, apresentados sob a forma de capítulos ...... 13 3. Capítulos ...... 14 Capítulo 1 ...... 15 Protocol of femur extraction from bats in fluid-preserved collections Capítulo 2 ...... 24 Comparative morphology and scaling of the femur in yangochiropteran bats 4. Discussão ...... 89 4.1 Morfologia do fêmur ...... 90 4.2 Alometria ...... 91 4.3 Predições funcionais ...... 92 4.4 Perspectivas evolutivas no estudo do fêmur ...... 94 5. Conclusões ...... 99 6. Referências bibliográficas ...... 101

LISTA DE FIGURAS

Figura 1: Relacionamento filogenético entre as famílias de Chiroptera, de acordo com o método de Máxima Verossimilhança (modificado de Amador et al., 2016). Note o posicionamento de Rhinolophoidea, como grupo-irmão de Pteropodidae...... 3

Figura 2: Diversidade de morcegos na subordem Yangochiroptera. (A) Mystacina tuberculata (Mystacinidae), (B) Thyroptera tricolor (Thyropteridae), (C) Furipterus horrens (Furipteridae), (D) Noctilio leporinus (Noctilionidae), (E) Pteronotus parnellii (Mormoopidae), (F) Myzopoda aurita (Myzopodidae), (G) Desmodus rotundus (Desmodontinae, Phyllostomidae), (H) Glossophaga soricina (Glossophaginae, Phyllostomidae), (I) Nycteris thebaica (Nycteridae), (J) Peropteryx macrotis (Emballonuridae), (K) Natalus stramineus (Natalidae), (L) Eumops perotis (Molossidae), (M) Miniopterus schreibersii (Miniopteridae), (N) Cistugo seabrai (Cistugidae) e (O) Lasiurus cinereus (Vespertilionidae). Crédito das imagens: (A) e (K) Bruce Thomson; (B) e (I) MerlinTuttle.org; (C-E, G, H, J, L) Roberto L. M. Novaes; (F) Daniel Riskin; (M) Alice Hughes; (N) Manuel Ruedi; (O) Justin Lindsay...... 4

Figura 3: Distribuição aproximada das famílias de Yangochiroptera. Note que algumas famílias possuem uma ampla distribuição (ex. Molossidae, Emballonuridae e Vespertilionidae), enquanto outras são mais restritas (ex. Cistugidae, Miniopteridae). Mystacina tuberculata, a única espécie vivente de Mystacinidae, é endêmica da Nova Zelândia. Modificado de The IUCN Red List of Threatened Species, http://www.iucnredlist.org...... 5

Figura 4: Características anatômicas do fêmur de Chiroptera. (A) Visão posterior do fêmur de Mystacina tuberculata (AMNH 160269); (B) Epífise proximal; (C) Epífise distal...... 6

Figura 5: Morcegos quadrúpedes especializados: (A) Desmodus rotundus (Desmodontinae); (B) Cheiromeles torquatus (Molossidae); (C) Mystacina tuberculata (Mystacinidae). Crédito das imagens: (A) José Cañas; (B) MerlinTuttle.org; (C) David Mudge...... 9

Figura 6: Exemplos de variação na orientação dos membros posteriores de morcegos. (A) morcegos não quadrúpedes; (B) morcegos capazes de rastejar; (C) morcegos capazes de locomoção quadrúpede ágil e especializada. Modificado de Schutt & Simmons (2006)...... 10

Figura 7: Uso dos membros posteriores para (A) captura de presas através do arrasto na superfície da água (Noctilio leporinus), (B) captura de presas durante o voo, com o auxílio do xvi

uropatágio (Myotis daubentonii), (C) juntamente com o uropatágio, gerar impulso durante a decolagem (Myotis evotis). Crédito das imagens: (A) Giancarlo Zorzin; (B) Lasse Jakobsen & Coen Elemans; (C) Adams et al., 2012 (modificada)...... 11

xvii

1. Introdução

1.1 Yangochiroptera

A ordem Chiroptera representa uma das maiores e mais diversificadas radiações de mamíferos, englobando um quinto das espécies conhecidas no mundo (ca. 1.300 spp.; Amador et al., 2016). Tradicionalmente, ela era dividida em duas subordens, Microchiroptera e Megachiroptera (Koopman, 1994; Simmons, 1998; Simmons & Geisler, 1998). Esta divisão era baseada principalmente em dados morfológicos e paleontológicos, mas também destacava as diferenças no modo dominante de percepção sensorial, ultrassom e visão, respectivamente (Teeling et al., 2012). Estudos com dados moleculares, entretanto, mostraram que o monofiletismo de Microchiroptera não se sustentava, estando a superfamília Rhinolophoidea mais relacionada aos Megachiroptera (Teeling et al., 2000, 2005, 2012; Van Den Bussche & Hoofer, 2004). Essa nova configuração, fortemente corroborada por estudos recentes (Amador et al., 2016), considera a divisão de Chiroptera na subordem Yinpterochiroptera (Rhinolophoidea + Pteropodidae), representada por 7 famílias, 63 gêneros e 375 espécies; e, na subordem Yangochiroptera, representada por 14 famílias, 158 gêneros e 866 espécies (Figura 1; Amador et al., 2016).

Além de englobar a maior parte da diversidade de Chiroptera (ca. 70% das espécies), Yangochiroptera inclui famílias com uma ampla variedade de formas (Figura 2), traduzidas na coloração da pelagem, no tamanho da orelha, na forma do crânio, na presença de apêndices faciais, na forma da asa e do uropatágio, entre outras características externas e do esqueleto (Fenton & Simmons, 2014; Norberg & Rayner, 1987; Vaughan, 1970). Há também uma grande variação de tamanho, desde a pequena Thyroptera tricolor (3-5g) até a enorme Vampyrum spectrum (130-235g) (Fenton & Simmons, 2014). Algumas famílias são amplamente distribuídas na região neotropical (ex. Phyllostomidae) e outras apresentam uma distribuição mais extensa, panglobal (ex. Molossidae e Vespertilionidae; Figura 3). Essa subordem inclui representantes de todas as categorias de dieta reconhecida nos morcegos — insetivoria, frugivoria, nectarivoria, carnivoria, piscivoria e hematofagia (Fenton & Simmons, 2014) —, que adotam diferentes estratégias de forrageamento para obtenção do alimento: voam em espaço aberto, borda de mata, ou fechado em busca de insetos; forrageiam acima de superfícies de água (rios, lagoas) na busca por insetos ou peixes e pequenos crustáceos e usam as pernas para o arraste e captura dos mesmos; coletam presas do substrato de um modo passivo, através da visão ou da geração de sons das próprias presas; encontram e se alimentam de flores ou frutos em espaços estreitos ou fechados; entre outras (Denzinger et al., 2016). O uso do abrigo

também é amplo e variado, incluindo espécies que se abrigam em cavernas e troncos/ocos de árvores, e espécies que usam abrigos mais especializados, como frestas, pelos molossídeos, e folhas, por Thyropteridae e Myzopodidae (Feldhamer et al., 2007; Fenton & Simmons, 2014). Yangochiroptera representa, portanto, importante modelo para estudos morfológicos comparados e funcionais.

rotundus (Desmodontinae, Phyllostomidae), (H) Glossophaga soricina (Glossophaginae, Phyllostomidae), (I) Nycteris thebaica (Nycteridae), (J) Peropteryx macrotis (Emballonuridae), (K) Natalus stramineus (Natalidae), (L) Eumops perotis (Molossidae), (M) Miniopterus schreibersii (Miniopteridae), (N) Cistugo seabrai (Cistugidae) e (O) Lasiurus cinereus (Vespertilionidae). Crédito das imagens: (A) e (K) Bruce Thomson; (B) e (I) MerlinTuttle.org; (C-E, G, H, J, L) Roberto L. M. Novaes; (F) Daniel Riskin; (M) Alice Hughes; (N) Manuel Ruedi; (O) Justin Lindsay.

hyropteridae Mormoopidae Natalidae hylostomidae Nycteridae Noctilionidae Molossidae Miniopteridae uripteridae espertilionidae Myzopodidae mballonuridae istugidae

Mystacinidae

1.2 O estudo do fêmur

As primeiras descrições do fêmur de morcegos datam dos séculos XVIII e XIX (ex. Allen, 1893; Buffon, 1760; Dobson, 1878; Flower & Gadow, 1885), mas caracterizações mais detalhadas vieram posteriormente com os trabalhos de Vaughan (1959, 1966, 1970). De acordo com esses trabalhos, o fêmur de morcegos é um pouco maior que a tíbia, geralmente fino e com pouco volume muscular, e pode ser dividido em três regiões: epífise proximal, haste e epífise distal (Figura 4). Na epífise proximal está a cabeça do fêmur e dois trocânteres, o maior e o menor, que são áreas de inserção muscular. A cabeça do fêmur é larga e levemente deslocada do eixo longitudinal da haste, o que permite uma grande liberdade de movimentos na articulação do quadril. Os trocânteres podem ser pouco desenvolvidos ou alongados. A haste

do fêmur geralmente é reta e possui duas cristas, uma lateral e uma medial, que também são áreas de inserção muscular. Essas cristas variam de posicionamento — podendo se iniciar ainda no trocânter menor ou mais abaixo na haste — e tamanho, podendo ser curtas ou longas. A epífise distal inclui dois côndilos, o lateral e o medial, separados pela fossa intercondilar.

Figura 4: Características anatômicas do fêmur esquerdo de Chiroptera. (A) Visão posterior do fêmur de Mystacina tuberculata (AMNH 160269); (B) Epífise proximal; (C) Epífise distal.

Os trabalhos mais detalhados em relação à anatomia do fêmur são os de Vaughan (1959) e de Smith (1972). Vaughan (1959) caracterizou o fêmur de três espécies de morcegos — Eumops perotis (Schinz, 1821) (Molossidae), Macrotus californicus Baird, 1858 (Phyllostomidae) e Myotis velifer (J.A. Allen, 1890) (Vespertilionidae). Ele observou variações na robustez e na forma da haste, no desenvolvimento e na posição dos trocânteres e das cristas laterais e mediais, na largura da epífise distal e no desenvolvimento da fossa intercondilar. Smith (1972), utilizando mais táxons — ao todo de nove famílias —, destacou diferenças na epífise proximal do fêmur, em relação ao tamanho dos trocânteres e à posição da cabeça. 6

Segundo ele, alguns táxons, como Natalus e Furipterus, possuem os trocânteres reduzidos em tamanho, enquanto outros (ex. Desmodus) possuem os trocânteres bem desenvolvidos, características que podem estar relacionadas a aspectos funcionais desses grupos (Smith, 1972).

1.3 Taxonomia

Quando aplicada à taxonomia, a morfologia do fêmur tem sido pouco explorada. Os estudos de Vaughan (1959) e Smith (1972) mostram que alguns caracteres da epífise proximal e da haste do fêmur são importantes para destacar diferenças interfamiliares em Chiroptera. Entretanto, a despeito desse potencial informativo, poucos trabalhos consideram formalmente tais caracteres como taxonômicos, diagnósticos ou em uma chave de identificação, tendo como exemplos os de Simmons (1998), Simmons & Geisler (1998), Gunnel & Simmons (2005), que utilizam poucos caracteres femorais para diagnose de táxons.

Aliado à taxonomia, a análise do esqueleto pós-craniano é extremamente relevante no estudo da paleofauna de morcegos (ex. Hand et al., 2009; Salles et al., 2014; Velazco et al., 2013; Ziegler et al., 2016), seja na identificação de fósseis de uma determinada localidade (ex. Czaplewski et al., 2005) ou em trabalhos mais aprofundados, envolvendo morfologia funcional e evolução. Hand et al. (2009), por exemplo, notou que diversas características da parte distal do úmero de Mystacina tuberculata Gray, 1843, que estão relacionadas à locomoção terrestre, são compartilhadas com um mistacinídeo extinto, Icarops aenae Hand et al., 1998, sugerindo que essa espécie do Mioceno já apresentava locomoção terrestre, o que permitiu um melhor entendimento da evolução desse hábito nessa família. Além do úmero, o crânio, a escápula e o fêmur são ossos que aparecem bem representados no registro fóssil (veja Velazco et al., 2013). Entretanto, vários táxons estão sub-representados ou mesmo ausentes nas coleções (Czaplewski et al., 2005) e nenhuma referência com muitos táxons e ilustrações de ossos está disponível para comparação (para uma contribuição recente com o esqueleto axial de filostomídeos, veja Gaudioso et al., 2017). O desenvolvimento nesta área, portanto, ajudará a aumentar nosso conhecimento sobre o registro fóssil de morcegos e sobre a evolução deste diversificado grupo de mamíferos (Gunnel & Simmons, 2005; Hand et al., 2009; Morgan & Czaplewski, 2012; Simmons & Geisler, 1998).

1.4 Alometria

Estudos com morfometria comparada mostram que as formas animais não são conservativas com o aumento do tamanho (ex. Christiansen, 2001; Madan et al., 2017; Schmidt & Fischer, 2009). Para Chiroptera, muitos estudos enfatizam a redução considerável dos membros posteriores em relação aos membros anteriores (ex. rádio e úmero) e em relação aos membros posteriores de mamíferos de tamanho semelhante (Simmons et al., 2008; Swartz, 1997; Swartz & Middleton, 2008). Essa tendência de redução estaria relacionada à pressão evolutiva para redução do peso corporal, essencial para evolução de um voo eficiente (Howell & Pylka, 1977; Riskin et al., 2005, 2016). O fêmur de morcegos apresenta alometria negativa no comprimento, mas uma tendência reversa — alometria positiva — em relação ao diâmetro (Swartz, 1997; Swartz & Middleton, 2008). Portanto, embora geralmente caracterizado como frágil, quando comparado a outros mamíferos (Howell & Pylka, 1977; Vaughan, 1959), parece haver uma tendência de morcegos maiores apresentarem fêmures mais robustos. Howell & Pylka (1977) propuseram que a robustez seria importante para as espécies quadrúpedes (ex. Desmodus rotundus (É. Geoffroy, 1810)) suportarem o estresse de compressão gerado durante a locomoção terrestre. Riskin et al. (2005), entretanto, refutaram essa hipótese ao constatarem que as pernas dos desmodontíneos não suportam forças maiores que as pernas mais finas de outros morcegos menos ágeis na locomoção quadrúpede (ex. Pteronotus parnellii Gray, 1843), sugerindo que outro mecanismo, como a miologia da cintura peitoral, limite a habilidade quadrúpede na maioria dos morcegos. Sendo assim, aspectos funcionais relacionados à alometria do fêmur ainda precisam ser elucidados.

1.5 Morfologia funcional

Morcegos usam seus membros posteriores para outros propósitos além de se pendurar de cabeça para baixo durante o repouso, e demandas funcionais associadas à locomoção terrestre (Lawrence, 1969; Schutt & Simmons, 2006) e à captura de presas (ex. insetos ou peixes; Neuweiler, 2000; Vaughan, 1966, 1970) também parecem ter desempenhado um papel importante na evolução dessas estruturas, eventualmente restringindo sua redução em algumas linhagens (Howell & Pylka, 1977; Riskin et al., 2005).

Em sua primeira monografia sobre locomoção de morcegos, Vaughan (1959) forneceu importantes dados associando características esqueléticas e musculares dos membros posteriores ao quadrupedalismo. Ele demonstrou que Eumops perotis, quando comparado a Macrotus californicus e Myotis velifer, possui músculos mais volumosos inseridos nas cristas

da haste do fêmur, o que confere a ele maior poder de mobilidade dos membros inferiores durante o quadrupedalismo. Este autor também propôs que os morcegos podem ser agrupados em três categorias de acordo com sua mobilidade quadrúpede: morcegos que não caminham, morcegos que caminham desajeitadamente, e morcegos que caminham bem (Vaughan, 1970).

Morcegos que não possuem habilidade quadrúpede ficam pendurados no abrigo e, quando colocados em uma superfície horizontal, preferem alçar o voo imediatamente; são exemplos natalídeos, muitos filostomídeos e alguns mormoopídeos (Riskin et al., 2016; Schutt & Simmons, 2006; Vaughan, 1970). Os morcegos que possuem uma capacidade limitada de caminhar usam movimentos desordenados e a superfície ventral do corpo confronta o chão periodicamente; são exemplos alguns mormoopídeos e vespertilionídeos (Riskin et al., 2006, 2016). Os quadrúpedes especializados são ágeis e conseguem manter o corpo afastado do chão, forrageando mais eficientemente; são representantes de Desmodontinae, Mystacinidae e Molossidae (Figura 5) (Riskin et al., 2006, 2016). Determinantes da habilidade quadrúpede parecem ser a posição dos membros posteriores, direcionados caudalmente (rotacionados cerca de 180°) no primeiro grupo (Figura 6A), direcionados lateralmente (cerca de 90°) no segundo grupo (Figura 6B) e direcionados antero-lateralmente nos quadrúpedes especializados (Figura 6C) (Schutt & Simmons, 2006). A postura nesses últimos, onde o fêmur está direcionado antero-lateralmente e a tíbia verticalmente, permite uma mobilidade maior e mais eficiente dos membros durante a locomoção (Schutt & Simmons, 2006; Vaughan, 1959).

Além do posicionamento dos membros, algumas características anatômicas do fêmur têm sido associadas ao quadrupedalismo. Smith (1972), por exemplo, notou que os sulcos e cristas presentes nos fêmures de Desmodus rotundus são modificações aparentemente relacionadas à agilidade quadrúpede, por permitirem uma maior inserção muscular. Schutt & Simmons (2001) destacaram que em Cheiromeles torquatus Horsfield, 1824, um dos molossídeos mais especializados no quadrupedalismo, a epífise proximal do fêmur é rotacionada em um ângulo maior — 30° — que nos outros molossídeos — 10° —, o que permitiria uma maior eficiência durante a locomoção quadrúpede e arbórea. Apesar de algumas associações entre forma e função serem mais claras, outras ainda permanecem desconhecidas, como destacado por Smith (1972) em relação aos trocânteres pouco desenvolvidos dos mormoopídeos.

Morcegos geralmente capturam as presas diretamente com a boca, mas também podem fazê-lo através das pernas (Anderson & Racey, 1991). Noctilio leporinus (Linnaeus, 1758), por exemplo, captura suas presas, principalmente peixes, na superfície da água usando seus pés e garras bem desenvolvidos (Figura 7A), e empregam o uropatágio para ajudá-lo na transferência da presa dos pés para a boca (Altenbach, 1979; Bloedel, 1955). Em outros casos, o próprio uropatágio é usado como uma espécie de concha para capturar a presa (insetos) durante o voo (Figura 7B). Os ossos que controlam essa estrutura, junto ao calcâneo, são de grande importância (Anderson & Racey, 1991; Webster & Griffin, 1962). Mesmo quando apenas o desempenho do voo é considerado, o uropatágio tem exercido um papel importante na geração de impulso e ascensão durante a decolagem a partir de superfícies horizontais a uma baixa

velocidade de voo (Adams et al., 2012; Figura 7C). Sendo assim, os membros posteriores desempenham papel importante nesses mecanismos, colapsando ou estendendo a membrana da cauda (Adams et al., 2012; Vaughan, 1959).

Figura 7: Uso dos membros posteriores para (A) captura de presas através do arrasto na superfície da água (Noctilio leporinus), (B) captura de presas durante o voo, com o auxílio do uropatágio (Myotis daubentonii), (C) juntamente com o uropatágio, gerar impulso durante a decolagem (Myotis evotis). Crédito das imagens: (A) Giancarlo Zorzin; (B) Lasse Jakobsen & Coen Elemans; (C) Adams et al., 2012 (modificada).

2. Objetivos

2.1 Objetivo Geral

Estudar qualitativa e quantitativamente a morfologia do fêmur em uma amostra significativa de Yangochiroptera, abordando aspectos taxonômicos, alométricos e funcionais desta estrutura.

2.2 Objetivos específicos, apresentados sob a forma de capítulos

Capítulo 1 – Descrever um protocolo de extração do fêmur a partir de espécimes preservados em coleções fluidas.

Capítulo 2 – (I) Descrever e comparar características anatômicas do fêmur nas famílias de Yangochiroptera, elaborar uma diagnose para cada uma delas e compor uma chave de identificação para as famílias neotropicais;

(II) Identificar as principais tendências na variação morfológica quantitativa do fêmur em Yangochiroptera através de análises multivariadas filogeneticamente informadas;

(III) Descrever padrões alométricos nas principais dimensões do fêmur de Yangochiroptera a partir de regressões filogenéticas.

3. Capítulos

Capítulo 1

Protocol of femur extraction from bats in fluid-preserved collections

Nathália Siqueira Veríssimo Louzada, Carlos Rodrigues de Moraes Neto, Marcelo Rodrigues Nogueira & Leila Maria Pessôa

Boletim de Mastozoologia

Submetido

Protocol of femur extraction from bats in fluid-preserved collections

Extraction of bats’ femur

Nathália Siqueira Veríssimo Louzada1,2, Carlos Rodrigues de Moraes Neto3, Marcelo Rodrigues Nogueira4 & Leila Maria Pessôa2

1Programa de Pós-graduação em Biodiversidade e Biologia Evolutiva, Instituto de Biologia, Universidade Federal do Rio de Janeiro (UFRJ), Rio de Janeiro, Rio de Janeiro, Brazil.

2Laboratório de Mastozoologia, Departamento de Zoologia, Instituto de Zoologia, Universidade Federal do Rio de Janeiro (UFRJ), Rio de Janeiro, Rio de Janeiro, Brazil.

3Setor de mamíferos, Departamento de Vertebrados, Museu Nacional, Universidade Federal do Rio de Janeiro (UFRJ), Rio de Janeiro, Rio de Janeiro, Brazil.

4Programa de Pós-graduação em Ecologia e Recursos Naturais, Laboratório de Ciências Ambientais, CBB, Universidade Estadual do Norte Fluminense (UENF), Campos dos Goytacazes, Rio de Janeiro, Brazil.

*Corresponding-author: [email protected]

Abstract The anatomy of appendicular bones has been demonstrated as informative to taxonomic, paleontological, and functional studies on bats, but its use is limited by the few skeletons available in scientific collections. Here we describe a protocol for the extraction of the femur from bats that are preserved in fluid. This protocol was tested successfully on a large sample, including 58 species in 43 genera and 9 families (total of 183 specimens), and is minimally invasive, requiring only two incisions in the leg, at disarticulation points (knee and coxofemoral joints). This method provides material to morphological research with minimal damage to external morphology.

Key-words Chiroptera; bones; method; taxonomic collection.

Protocolo de extração do fêmur de morcegos conservados em meio líquido

Resumo A anatomia dos ossos apendiculares tem se mostrado informativa para estudos taxonômicos, paleontológicos e funcionais com morcegos, mas seu uso é limitado pelos poucos

esqueletos disponíveis em coleções científicas. Descrevemos aqui um protocolo para extração do fêmur de morcegos que estão preservados em meio líquido. Esse protocolo foi aplicado com sucesso em uma grande amostragem, incluindo 58 espécies de 43 gêneros e 9 famílias (total de 183 espécimes), e é minimamente invasivo, requerendo somente duas incisões na perna, em pontos de desarticulação (joelho e junção coxofemoral). Esse método fornece material para pesquisa morfológica gerando danos mínimos à morfologia externa.

Palavras-chave Chiroptera; ossos; método; coleções taxonômicas.

A taxonomic collection is an ordered set of preserved specimens prepared for scientific studies (Martins, 1994). In the case of mammals, three main preparation methods have been used: “skin and skull”, “fluid”, and “skeleton” (Yates et al., 1996). Bats are increasingly represented in collections (e.g. Dunnum et al., 2018) and are frequently preserved as “fluid” specimens (i.e. fixation of the entire animal in formalin 10% and storage in alcohol 70%). As emphasized by Simmons & oss (2009), although fur color fades in fluid preservatives, eventually hampering the identification of banding patterns, this method has several advantages over the traditional “skin and skull”, including faster execution and better preservation of the external morphology, mainly face and ears. Not only the external morphology, but all internal structures are preserved in fluid specimens, maximizing their usefulness in subsequent studies (Simmons & oss, 2009).

luid-preserved specimens are (or at least should be) always fixed with their mouth open, allowing examination of dental morphology without the need of skull extraction (Simmons & oss, 2009). urthermore, if a more detailed dental analysis is required, or if skull structures need to be examined—both common practices in taxonomic studies (e.g. Gardner, 2008; Gregorin & addei, 2002; elazco et al., 2010)—, the procedure of keeping the specimen with the mouth open will greatly facilitate skull removal (Simmons & oss, 2009). When examination of post-cranial bones of fluid-preserved specimens are necessary, however, researchers found limited material in collections (e.g. only nearly 8% of the 63,444 bat records in the American Museum of Natural History mammal database represent skeletons; query on January 2019) and no formal procedure describing bone removal is available in the literature. Skeleton preparation is a lengthy process and, in many cases, involves scarifying pelage and internal organs (Simmons & oss, 2009). onsidering the wide spectrum of studies in areas such as taxonomy, paleontology and ecomorphology (e.g. zaplewski et al., 2005; Gaudioso et

al., 2017; Hand et al., 2009; elazco et al., 2013; Ziegler et al., 2016) that could benefit from an increase in the number of taxa and specimens in their samples, techniques allowing bone extraction from fluid-preserved material with minimum damage would be useful. Here we describe a protocol for the extraction of the femur of fluid-preserved bats. his protocol was tested on a sample of 183femora, removed from specimens belonging to 9 families, 43 genera, and 58 species of bats (Appendix). Nomenclature of bones and ligaments followed König & Liebich (2016), and values of body weight (g) were obtained from Reis et al. (2017).

The first step is the skin hydration process, where the specimens are removed from fluid (let the excess of alcohol drain) and placed in a container with 40% alcohol. The time each specimen needs to be soaked in this stage varies according to its size: small bats (3.5–20 g), such as Furipterus horrens and Natalus macrourus, should be immersed for 15–20 minutes; medium-sized bats (20–50 g), such as Molossus rufus and Chiroderma villosum, for 30–40 minutes; and larger bats (>50 g), such as Noctilio leporinus and Chrotopterus auritus, for 40– 50 minutes. The quantity of bats immersed should be calculated according to their weight: a maximum of 100 g of bats for each liter of 40% alcohol (e.g. two 50g specimens of Noctilio leporinus immersed in 1L of alcohol). This step helps to rehydrate the specimen, softening its leg articulations (Simmons, 2014). Since excessive exposure to this solution may result in soft tissue deterioration (Neto et al., 2015), the specimens should be observed every 10 (smaller ones) to 15 (larger ones) minutes. After the soaking time, the specimen is removed from the bath and, carefully, the femoral joints at the pelvis and knee are checked, making circular and top-down movements, respectively. If the joints are less rigid, the next step is proceeded, and the specimen is dried with absorbent paper, to facilitate the handling of the skin. If not, the specimen should be returned to the 40% bath until softening the joins.

The femoral extraction is proceeded with the specimen in dorsal decubitus. Two incisions need to be made, one at the knee joint and another at the coxofemoral joint (Figure 1A). With a scalpel, a slight incision at the knee joint are made to carefully disarticulate the femur. In this region the skin is very thin and the muscles and tendons are easily cut. The second incision is made from the lateral of the pelvic region to the middle of the body, until reaches the tail or uropatagium area. At this region, there is usually a lot of musculature, especially in molossids, which requires extra care to avoid damaging the femur. With the aid of a clip, the muscles that involves the proximal region of the femur are carefully cut until to locate the head of this bone, which is articulated to the pelvis. Once found, a cut at the ligament (ligamentum

capitis ossis femoris) joining these bones are made, which detach the femur of the acetabulum. Once free of both joints, the femur can be carefully pulled out by the incision made in the area of the pelvis, with minimal damage on the specimen’ skin ( igure 1B). After the extraction, the specimen is placed in a new immersion of alcohol 70% but, as the specimen are less concentrate, the immersion should be checked periodically until the alcohol concentration stabilize at 70%, completing with alcohol whenever necessary. When the concentration stabilizes at 70%, the specimens are transferred to their original pots.

The extracted femur is cleaned using the mechanical cleaning and the treatment with Dermestes beetles (larvae and adults). The excess of musculature is removed with a clip and the femur is exposed under a lamp light during a day to dry. Once dried, the femur is stored with a resistant tag (we used Rotex tape) on a recipient with two to four Dermestes larvae. The time these beetles take to clean femora varies from two to five days, so on this phase the check must be daily. Once cleaned, the femur is ready to be studied (Figure 1C).

This protocol was successfully applied on bat specimens kept in fluid for years (the oldest specimen dates from 1942), demonstrating its effectiveness, contributing to the increase of the femoral sample of the Chiroptera collection of the Museu Nacional (Rio de Janeiro), and providing material to the development of comparative and anatomical studies in this poorly- known area (e.g. Louzada et al., 2019). Although our focus has been on the femur, the technique described here can also work for other bones, once the peculiarities of each one are considered.

Acknowledgements

The authors are thankful to João Alves de Oliveira for providing access to the Museu Nacional (Rio de Janeiro) material studied here. This paper is a part of the D. Sc. requirements of Nathália Siqueira Veríssimo Louzada of the Programa de Pós-graduação em Biodiversidade e Biologia Evolutiva (PPGBBE), of the Universidade Federal do Rio de Janeiro. Financial support was provided by Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ) and Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES) – Finance Code 001. MRN is supported by the PNPD/CAPES fellowship program. LMP is supported by the CNPQ fellowship program (308505/2016-6).

Appendix

Specimens studied: Thyropteridae (1 genus, 1 species): Thyroptera tricolor (ALP 2912, 2914, 2920); Furipteridae (1 genus, 1 species): Furipterus horrens (MN 36053, 36287, 78115, 78116, 78117); Noctilionidae (2 genera, 2 species): Noctilio albiventris (MN 64141, 64144, 64145, 64152), and Noctilio leporinus (MN 71319, 71321, 71577, 71580); Mormoopidae (1 genus, 2 species): Pteronotus gymnonotus (MN 68061, 68073, 68074, 68085), Pteronotus parnellii (MN 80542, 80553, 80564, 80569, 80623); Phyllostomidae: Micronycterinae (1 genus, 3 species): Micronycteris megalotis (MN 36158, 36159), Micronycteris microtis (MN 80534, 80573), and Micronycteris sanborni (MN 75194, 79755, 80572); Desmodontinae (2 genera, 2 species): Diaemus youngii (MN 71029, 71037, 71379, 77875, 79877), and Diphylla ecaudata (MN 68033, 68034, 68035, 68036); Lonchorhininae (1 genus, 1 species): Lonchorhina aurita (MN 79798, 79802, 80540, 80541); Phyllostominae (9 genera, 12 species): Chrotopterus auritus (MN 70862), Gardnerycteris crenulatum (MN 71390, 71391, 75189, 80583), Lophostoma brasiliense (MN 80495), Lophostoma carrikeri (MN 71404), Lophostoma silvicola (MN 71351, 71486), Macrophyllum macrophyllum (MN 70599, 70600, 70601, 70662), Mimon bennettii (MN 79816, 79827, 79891, 80537), Phylloderma stenops (MN 70594, 70861), Phyllostomus discolor (67723, 67724, 67725), Phyllostomus elongatus (MN 70545, 70546), Tonatia saurophila (MN 70248, 70860), and Trachops cirrhosus (MN 71426, 71346, 71469, 80536); Lonchophyllinae (1 genus, 1 species): Lonchophylla dekeyseri (MN 80497, 80525, 80563); Rhinophyllinae (1 genus, 2 species): Rhinophylla fischerae (MN 70252, 70254, 70255, 70323), and Rhinophylla pumilio (MN 70808, 70809, 70811, 70845); Stenodermatinae (6 genera, 7 species): Chiroderma villosum (MN 64518, 64519, 71370, 71372), Mesophylla macconnellii (MN 71372), Pygoderma bilabiatum (MN 81269, 81275), Sturnira lilium (MN 36189, 36192, 36313, 36314), Uroderma bilobatum (MN 70877, 70881, 70888, 70918), Uroderma magnirostrum (MN 70302, 70655, 70885), and Vampyriscus bidens (MN 70908, 70909, 70910, 70921); Emballonuridae (4 genera, 4 species): Diclidurus isabella (MN 70449, 70644), Peropteryx macrotis (MN 79767, 79768, 79797, 79809), Rhynchonycteris naso (MN 70428, 70429, 70431), and Saccopteryx bilineata (MN 70445, 70611, 70914, 71594); Natalidae (1 genus, 1 species): Natalus macrourus (MN 67863, 67868, 67873, 81448); Molossidae (8 genera, 12 species): Cynomops planirostris (MN 70276, 70287, 70288, 70301), Eumops glaucinus (MN 71472), Eumops perotis (MN 64370, 64747, 64750, 64787, 64789, 71287), Molossops temminckii (MN 71350, 71355, 71415, 71418, 71421, 71434), Molossus currentium (MN 71560, 71563), Molossus molossus (MN 71561, 71562, 71564), Molossus pretiosus (MN 71551, 71552, 71566), Molossus rufus (MN 71568, 79894), Neoplatymops

mattogrossensis (MN 36029, 36030, 36031, 37322), Nyctinomops macrotis (MN 49583, 49584, 49586, 49587, 49588, 49594), Promops nasutus (MN 64677, 64678, 64746, 64762, 64763, 71480), and Tadarida brasiliensis (MN 6562, 6564, 6565, 6566); Vespertilionidae (4 genera, 7 species): Eptesicus brasiliensis (MN 71417), Histiotus velatus (MN 46491), Lasiurus blossevillii (MN 71304, 71459), Lasiurus ega (MN 70593), Myotis nigricans (MN 71530, 71532), Myotis riparius (MN 71311, 71589), and Myotis simus (MN 71451, 71458).

References

Czaplewski NJ, Rincón AD, Morgan GS. 2005. Fossil bat (Mammalia: Chiroptera) remains from Inciarte Tar Pit, Sierra de Perijá, Venezuela. Caribbean Journal of Science 41: 768-781.

Dunnum JL, McLean BS, Dowler RC. 2018. Mammal collections of the Western Hemisphere: a survey and directory of collections. Journal of Mammalogy 99(6): 1307-1322. https://doi.org/10.1093/jmammal/gyy151.

Gardner AL. 2008. Mammals of South America, volume 1: marsupials, xenarthrans, shrews, and bats. University of Chicago Press, Chicago.

Gaudioso PJ, Díaz MM, Barquez RM. 2017. Morphology of the axial skeleton of seven bat genera (Chiroptera: Phyllostomidae). Anais da Academia Brasileira de Ciências 89(3): 2341- 2358. http://dx.doi.org/10.1590/0001-3765201720170076.

Gregorin R, Taddei VA. 2002. Chave artificial para a identificação de molossídeos brasileiros (Mammalia, Chiroptera). Mastozoología Neotropical 9(1): 13-32.

Hand SJ, Weisbecker V, Beck RM, Archer M, Godthelp H, Tennyson AJ, Worthy TH. 2009. Bats that walk: a new evolutionary hypothesis for the terrestrial behavior of New Zealand's endemic mystacinids. BMC Evolutionary Biology 9(1): 169. https://doi.org/10.1186/1471- 2148-9-169.

König HE, Liebich HG. 2016. Anatomia dos Animais Domésticos: Texto e Atlas Colorido. Artmed Editora, Porto Alegre.

Louzada NSV, Nogueira MR, Pessôa LM. 2019. Comparative morphology and scaling of the femur in yangochiropteran bats. Journal of Anatomy. DOI: 10.1111/joa.12996.

Martins UR. 1994. A coleção taxonômica. Pp. 19-43, in Papavero N (Ed.), Fundamentos Práticos de Taxonomia Zoológica. Editora da Universidade Estadual Paulista, São Paulo.

Neto CRM, Lazar A, Pessôa, LM. 2015. Taxidermia de pequenos mamíferos não-voadores previamente preservados em meio líquido. Boletim da Sociedade Brasileira de Mastozoologia 73: 92-94.

Reis NR, Peracchi AL, Batista CB, de Lima IP, Pereira AD. 2017. História natural dos morcegos brasileiros: chave de identificação de espécies. Technical Books Editora, Rio de Janeiro.

Simmons NB, Voss RS. 2009. Collection, preparation, and fixation of specimens and tissues. Ecological and Behavioral Methods for the Study of Bats. Johns Hopkins University Press, Baltimore.

Simmons JE. 2014. Fluid preservation: a comprehensive reference. Rowman & Littlefield, Maryland.

Velazco PM, Gardner AL, Patterson BD. 2010. Systematics of the Platyrrhinus helleri species complex (Chiroptera: Phyllostomidae), with descriptions of two new species. Zoological Journal of the Linnean Society 159(3): 785-812. https://doi.org/10.1111/j.1096- 3642.2009.00610.x.

Velazco PM, O'Neill H, Gunnell GF, Cooke SB, Rimoli R, Rosenberger AL, Simmons NB. 2013. Quaternary bat diversity in the Dominican Republic. American Museum Novitates 3779: 1-20.

Yates L, Jones , ook JA. 1996. reservation of voucher specimens. p 265-274, in Wilson D , ole R, Nichols JD, Rudran R, oster MS ( ds.) Measuring and monitoring biological diversity: standard methods for mammals. Smithsonian Institution ress, Washington.

Ziegler AC, Howarth FG, Simmons NB. 2016. A second endemic land mammal for the Hawaiian Islands: a new genus and species of fossil bat (Chiroptera: Vespertilionidae). American Museum Novitates 3854: 1-52.

Figure 1 Fluid-preserved Noctilio leporinus (MN 71321) showing the pelvic and knee incisions (arrows) for subsequent femur extraction (A) and the hindlimb region after the bone extraction (B). Note that the plagiopatagium and the tail membrane were not damaged. In (C), cleaned femora of Diaemus youngii (MN 71037), Phyllostomus discolor (MN 67724), Sturnira lilium (MN 36192), Lasiurus blossevillii (MN 72304), and Cynomops planirostris (MN 70276), respectively. Scale bar: 2 mm.

Capítulo 2

Comparative morphology and scaling of the femur in yangochiropteran bats

Nathália Siqueira Veríssimo Louzada, Marcelo Rodrigues Nogueira & Leila Maria Pessôa

Journal of Anatomy

Provisional accept – Minor revisions

Femur morphology in bats

Comparative morphology and scaling of the femur in yangochiropteran bats

Nathália Siqueira Veríssimo Louzada1,2, Marcelo Rodrigues Nogueira3, and Leila Maria Pessôa2

1Programa de Pós-graduação em Biodiversidade e Biologia Evolutiva, Instituto de Biologia, Universidade Federal do Rio de Janeiro, Avenida Brigadeiro Trompowski, 221, Rio de Janeiro, RJ, CEP 21941–590, Brazil

2Laboratório de Mastozoologia, Departamento de Zoologia, Instituto de Zoologia, Universidade Federal do Rio de Janeiro, Avenida Brigadeiro Trompowski, 221, Rio de Janeiro, RJ, CEP 21941– 590, Brazil

3Programa de Pós-graduação em Ecologia e Recursos Naturais, Laboratório de Ciências Ambientais, CBB, Universidade Estadual do Norte Fluminense, Av. Alberto Lamego 2000, Campos dos Goytacazes, RJ, CEP 28013–602, Brazil

Abstract

Better known by their remarkable forelimb morphology, bats are also unique among mammals in respect to their hindlimbs. heir legs are rotated through 180º, generally reduced in size, and in some extant taxa particular bones (e.g. fibula) can even be absent. he femur is the main leg bone, but to date few bat studies have considered its morphology in detail, none in a wide-scale comparative study. Yangochiroptera is the largest bat taxon, spans nearly three orders of magnitude in body mass, and is highly diverse both in ecology and behavior, representing a good model for comparative analyses. Here we describe the anatomy of the femur in a large sample of yangochiropteran bats (125 species, 70 genera, and 12 families), and explore major trends of morphological variation and scaling patterns in this bone. We used 13 categorical characters in the anatomical description and five linear dimensions in the quantitative analyses. Based on the categorical data, each family studied here was diagnosed, and those from the Neotropical region were included in an identification key. rom the phylogenetic principal component analysis (p A) we showed that, in addition to size, major axes of variation in bat femur are related to robusticity and head morphology, features that are clearly distinct among some families. We also generated a phylomorphospace based on p A scores, highlighting

convergences in femur shape. Molossidae, Mystacinidae, and Desmodontinae were grouped based on their greater robusticity, a pattern that was also recovered from categorical data. In these families we found anatomical features (e.g. presence of tubercles and posterior ridges on the greater trochanter, long or medially/distally displaced lateral ridges on the shaft) that are well-known by their functional link with quadrupedal locomotion. Using phylogenetic regressions, we found out that compared to body mass, femur length scaled with negative allometry, as expected, but that femur width scaled isometrically, counter to expectations. As a result, robusticity index (the ratio of width to length), scaled with positive allometry—larger bats tended to have more robust hindlimbs. At species level, our most remarkable finding was related to Myotis simus, which presented the most robust femur (for its size) among yangochiropterans. Our results reinforce the informative potential of the chiropteran femur from both taxonomic and functional perspectives. urthermore, the allometric trends seen in this bone may help understand the strategies adopted by flying vertebrates to deal with the high energetic cost of flight and, at the same time, evolve diversified foraging behaviors.

Key words: allometry; bone robusticity; comparative methods; hindlimbs; identification key; quadrupedalism.

Introduction

Bats are well known by their unique forelimbs, modified into wings, but their hindlimbs have also passed through noticeable evolutionary changes. Onychonycteris finneyi Simmons et al. 2008, one of the oldest fossil species of bats, was found to have a robust fibula (Simmons et al. 2008), while in most recent species this bone is poorly developed or absent (Vaughan, 1970). This tendency for reduction is also evident for other components of the hindlimbs (Howell & Pylka, 1977; Swartz, 1997) and is in agreement with an expected overall reduction of body mass in these mammals, which would reduce the energy required for flying (Vaughan, 1959; Swartz, 1997). The posture adopted by most bats while resting and feeding—hung upside-down (Howell & Pylka, 1977)—seems also to have evolved in association with flight and does not require robust leg bones or developed muscles (Fenton & Simmons, 2014).

The first descriptions of the bat femur date from the 18th/19th centuries (e.g. Buffon, 1760; Allen, 1864; Dobson, 1878; Flower & Gadow, 1885), but since then relatively few studies have been published on this subject. Functional aspects were examined based on categorical characters by Vaughan (1959) and Schutt & Simmons (2001), with a focus on Cheiromeles 26

(Molossidae) and quadrupedal locomotion in the latter study. From a quantitative perspective, scaling analyses were performed by Howell & Pylka (1977), Swartz (1997), and Riskin et al. (2005), with two of these studies emphasizing vampire bats (Desmodontinae) and a possible association between the robusticity of their femur and functional demands related to walking. Based on a large taxon sampling, Swartz (1997) searched for more general patterns and found out that the bat femur shows negative allometry in length, but a reverse trend in diameter. Therefore, although generally characterized as gracile, when compared to other mammals (Vaughan, 1959; Howell & Pylka, 1977), there seems to be a tendency for larger bats to present more robust femora.

When applied to taxonomy, the femur morphology has also been poorly explored. Smith (1972) found striking interfamilial differences when comparing mormoopids to other Neotropical families, highlighting, among other aspects, the peculiar morphology of Desmodontinae and Molossidae. The size and position of the main elements in the proximal end of the femur (head and trochanters) were useful characters in this study, as was the width of the shaft in interspecific comparisons in Desmodus (Morgan, 1991; Suárez, 2005). The study of paleofaunas has particularly benefited from postcranial analyses (Hand et al. 2009; Velazco et al. 2013; Salles et al. 2014; Ziegler et al. 2016), but several taxa are underrepresented or even absent in collections (Czaplewski et al. 2005) and no taxon-wide published reference with bone illustrations is available (for a recent contribution in phyllostomids, see Gaudioso et al. 2017). Development in this area, therefore, will help to increase our knowledge on the bat fossil record and on the evolution of this rich mammalian group (e.g. Simmons & Geisler, 1998; Gunnel & Simmons, 2005; Hand et al. 2009; Morgan & Czaplewski, 2012).

Herein we present a comparative study of the bat femur using a wide range of families (12), genera (70) and species (125) of Yangochiroptera. We expected to find high variation in the morphology of this bone not only as a result of the large taxonomic sampling performed here (the whole suborder comprises 14 families, 158 genera, and 866 species; Amador et al. 2016), but also due to the ecological diversity reported for some of its components (Gaudioso et al. 2017). More specifically, our aims in this study are: (i) to describe femoral anatomical features and diagnose bat families; (ii) to explore major quantitative trends in the morphology of the femur; and (iii) to revisit allometric patterns in femur main dimensions. In respect to this analysis of allometry, we investigate how femur length, width, and robusticity are related to size in bats. Considering the general expectation of an adaptive scaling leading to body mass

reduction within the group (Swartz, 1997), and that even species with less robust hindlimbs are able to deal with compressive forces related to walking (Riskin et al. 2005), we predict that robusticity will show negative allometry.

Materials and methods

Specimens studied

We examined the femur of 415 specimens representing 12 families, 70 genera, and 125 species of extant yangochiropteran bats (Table 1). These specimens are deposited in the following Institutions: Museu Nacional (MN), Rio de Janeiro; Universidade Federal Rural do Rio de Janeiro (Coleção Adriano Lúcio Peracchi – ALP), Rio de Janeiro; Museu de Zoologia da USP (MZUSP), São Paulo; and American Museum of Natural History (AMNH), New York (Appendix S1). They were recognized as adults based on the closure of basicranial sutures and metacarpal-phalangeal joints (Kunz & Anthony, 1982). The primary source to assign species names to these specimens were their tags but, in many cases, we revised those identifications considering more recent taxonomic assessments (Gregorin & Taddei, 2002; Gardner, 2008; Velazco et al. 2010; Gregorin et al. 2011; Moratelli et al. 2011; Tejedor, 2011; Medina et al. 2014; Moratelli & Dias, 2015; Pavan & Marroig, 2016). Classification is in accordance with Amador et al. (2016) and subfamilial names adopted for Phyllostomidae follows Baker et al. (2016).

Anatomical Description

The terminology used in the anatomical description of the femur followed Vaughan (1959), Schutt & Simmons (2001), and König & Liebich (2016). The femur can be divided into three regions (Fig. 1): proximal, with the head (caput ossis femoris); body (shaft) (corpus ossis femoris); and distal, with lateral and medial condyles (condylus lateralis et medialis). At the proximal end, the neck of the femur is reduced and the head of the femur is slightly displaced from the main axis of the shaft. Laterally and medially to the head (posterior view) there are two processes, the greater and the lesser trochanter, respectively, which are separated by the trochanteric fossa. On the shaft there are two ridges, one lateral and another medial. The condyles at the distal end are separated by the intercondylar fossa (posterior view).

Table 1 Species, abbreviations and sample size of yangochiropteran bats analyzed for their femur morphology in this study.

Taxon Abr. Sample size Superfamily Noctilionoidea Family Mystacinidae (1 genus, 1 species) MYS Mystacina tuberculata Mtub 1 Family Thyropteridae (1 genus, 1 species) THY Thyroptera tricolor Ttri 7 Family Furipteridae (1 genus, 1 species) FUR Furipterus horrens Fhor 6 Family Noctilionidae (1 genus, 2 species) NOC Noctilio albiventris Nalb 6 Noctilio leporinus Nlep 7 Family Mormoopidae (2 genera, 6 species) MOR Mormoops blainvillei Mbla 1 Pteronotus gymnonotus Pgym 4 Pteronotus macleayi Pmal 1 Pteronotus rubiginosus Prub 9 Pteronotus psilotis Ppsi 1 Pteronotus quadridens Pqua 1 Family Phyllostomidae (11 subfamilies) PHY Subfamily Macrotinae (1 genus, 1 species) Macrotus waterhousii Mwat 1 Subfamily Micronycterinae (1 genus, 3 species) Micronycteris megalotis Mmeg 2 Micronycteris microtis Mmic 3 Micronycteris sanborni Msan 3 Subfamily Desmodontinae (3 genera, 3 species) Desmodus rotundus Drot 19 Diaemus youngii Dyou 5 Diphylla ecaudata Deca 8 Subfamily Lonchorhininae (1 genus, 1 species) Lonchorhina aurita Laur 5 Subfamily Phyllostominae (10 genera, 15 species) Chrotopterus auritus Caur 1 Gardnerycteris crenulatum Gcre 5 Lophostoma brasiliense Lbra 2 Lophostoma carrikeri Lcar 1 Lophostoma schulzi Lsch 1 Lophostoma silvicola Lsil 3

Table 1 (cont.) Species, abbreviations, and sample size of yangochiropteran bats analyzed for their femur morphology in this study.

Taxon Abr. Sample size Macrophyllum macrophyllum Mmac 5 Mimon bennettii Mben 4 Phylloderma stenops Pste 3 Phyllostomus discolor Pdis 8 Phyllostomus elongatus Pelo 4 Phyllostomus hastatus Phas 9 Tonatia saurophila Tsau 3 Trachops cirrhosus Tcir 5 Vampyrum spectrum Vspe 1 Subfamily Glossophaginae (7 genera, 9 species) Anoura caudifer Acau 3 Anoura geoffroyi Ageo 3 Brachyphylla cavernarum Bcav 1 Choeroniscus minor Cmin 1 Erophylla sezekorni Esez 1 Glossophaga commissarisi Gcom 1 Glossophaga soricina Gsor 13 Leptonycteris yerbabuenae Lyer 1 Monophyllus redmani Mred 1 Subfamily Lonchophyllinae (2 genera, 2 species) Hsunycteris thomasi Htho 1 Lonchophylla dekeyseri Ldek 4 Subfamily Carolliinae (1 genus, 4 species) Carollia brevicauda Cbre 8 Carollia castanea Ccas 1 Carollia perspicillata Cper 5 Carollia sowelli Csow 1 Subfamily Glyphonycterinae (1 genus, 1 species) Trinycteris nicefori Tnic 1 Subfamily Rhinophyllinae (1 genus, 2 species) Rhinophylla fischerae Rfis 4 Rhinophylla pumilio Rpum 7 Subfamily Stenodermatinae (11 genera, 27 species) Ametrida centurio Acen 1 Artibeus concolor Acon 1 Artibeus fimbriatus Afim 3 Artibeus jamaicensis Ajfa 1

Table 1 (cont.) Species, abbreviations, and sample size of yangochiropteran bats analyzed for their femur morphology in this study.

Taxon Abr. Sample size Artibeus lituratus Alit 9 Artibeus obscurus Aobs 4 Artibeus planirostris Apla 8 Chiroderma salvini Csal 1 Chiroderma trinitatum Ctri 1 Chiroderma villosum Cvil 5 Dermanura cinerea Dcin 5 Dermanura gnoma Dgno 2 Mesophylla macconnellii Mman 3 Platyrrhinus brachycephalus Pbra 1 Platyrrhinus incarum Pinc 1 Platyrrhinus lineatus Plin 6 Pygoderma bilabiatum Pbil 3 Sturnira erythromos Sery 1 Sturnira ludovici Slud 1 Sturnira oporaphilum Sopo 1 Sturnira lilium Slil 7 Sturnira tildae Stil 2 Uroderma bilobatum Ubil 5 Uroderma magnirostrum Umag 4 Vampyressa brocki Vbro 1 Vampyressa thyone Vthy 1 Vampyriscus bidens Vbid 4 Superfamily Emballonuroidea Family Nycteridae (1 genus, 1 species) NYC Nycteris macrotis luteola Nmar 1 Family Emballonuridae EMB Subfamily Emballonurinae (5 genera, 8 species) Cormura brevirostris Cbrv 1 Diclidurus isabella Disa 2 Peropteryx kappleri Pkap 2 Peropteryx macrotis Pmac 6 Peropteryx trinitatis Ptri 1 Rhynchonycteris naso Rnas 6 Saccopteryx bilineata Sbil 5 Saccopteryx leptura Slep 1 Subfamily Tophozoinae (1 genus, 2 species)

Table 1 (cont.) Species, abbreviations, and sample size of yangochiropteran bats analyzed for their femur morphology in this study.

Taxon Abr. Sample size Taphozous georgianus Tgeo 1

Taphozous melanopogon Tmel 1 Superfamily Vespertilionoidea Family Natalidae (2 genera, 3 species) NAT Natalus macrourus Nmac 6

Natalus major Nmaj 1 Nyctiellus lepidus Nled 1 Family Molossidae MOL

Subfamily Molossinae (11 genera, 20 species) Chaerophon plicatus Cpli 1

Cheiromeles torquatus Ctor 1

Cynomops greenhalli Cgre 1 Cynomops planirostris Cpla 6 Eumops glaucinus Egla 2

Eumops perotis Eper 7 Eumops trumbulli Etru 1

Molossops temminckii Mtem 7

Molossus coibensis Mcoi 1 Molossus currentium Mcur 2 Molossus molossus Mmol 4

Molossus pretiosus Mpre 5 Molossus rufus Mruf 2

Mops leucostigma Mleu 4

Neoplatymops mattogrossensis Nmat 4 Nyctinomops femorosaccus Nfem 1 Nyctinomops laticaudatus Nlat 3

Nyctinomops macrotis Nmai 7 Promops nasutus Pnas 6

Tadarida brasiliensis Tbra 5

Family Miniopteridae (1 genus, 1 species) MIN Miniopterus magnater Mmag 1 Family Vespertilionidae VES

Subfamily Vespertilioninae (3 genera, 5 species) Eptesicus brasiliensis Ebra 1

Histiotus montanus Hmon 1

Histiotus velatus Hvel 3 Lasiurus blossevillii Lblo 2

Table 1 (cont.) Species, abbreviations, and sample size of yangochiropteran bats analyzed for their femur morphology in this study. Taxon Abr. Sample size Lasiurus ega Lega 1

Subfamily Myotinae (1 genus, 6 species) Myotis albescens Malb 2 Myotis levis Mlev 1 Myotis nigricans Mnig 3 Myotis riparius Mrip 3 Myotis ruber Mrub 1 Myotis simus Msim 2

From these anatomical features, 13 characters, described below, were used to characterize the families studied. These characters are summarized in Table S1 and demonstrated in the Figures 2-5. We also provide an identification key based on these characters, to help taxonomic and paleontological studies, but due to the little material available for some families, it includes only Neotropical groups (Appendix 1). Unless stated otherwise, characters should be visualized from the posterior view.

(1) Development of the medial ridge (Fig. 2): poorly developed (the ridge is thin and long; sometimes its beginning is slightly projected medially), moderately developed (the ridge is clearly noticeable; may be short or long, but it is not greatly pronounced medially), or well developed (the ridge is short and greatly pronounced medially, or long, extending down to the shaft).

(2) Proximal end of the medial ridge (Fig. 2): on the lesser trochanter (the ridge begins on the lesser trochanter and extends to the shaft, sometimes giving a straight appearance to the proximal epiphysis), or on the shaft (the ridge usually begins on the medial side of the shaft, sometimes just distal to the lesser trochanter; in some groups it is slightly displaced and begins on the anterior side of the femur).

Remarks: because the ridge can be displaced and begins on the anterior side, we recommend to look at this character also from the anterior view.

(3) Development of the trochanters (Fig. 3): poorly developed (the trochanters are small and usually rounded, with a narrow gap between them); or well developed (the trochanters are robust and elongated, and a conspicuous gap separates them). 33

(4) Proximal level of the greater trochanter relative to the lesser trochanter (Fig. 3): aligned or with subtle difference (the proximal end of the two trochanters reaches nearly the same proximal level), higher (the greater trochanter projects further proximally than the lesser), or lower (the lesser trochanter projects further proximally than the greater).

(5) Tubercle on the greater trochanter (Fig. 3): absent (the greater trochanter is simple, with no bone protuberance), poorly developed (there is a small bone protuberance), or well developed (the greater trochanter presents a great projection at the distal part, that extends laterally).

Remarks: when well developed, the tubercle can be rounded or hook-shaped.

(6) Posterior ridge on the greater trochanter (Fig. 3): absent (the ridge is absent, the greater trochanter is smooth), poorly developed (the ridge is low), or well developed (the ridge is high and well defined).

Remarks: the posterior ridge is more easily visualized on the lateral or medial view of the greater trochanter.

(7) Trochanteric fossa (Fig. 2): poorly developed (the trochanters are joined or, if separated, the fossa is minute), shallow (the fossa is semicircular), or deep (the fossa is “U” shaped).

(8) Femur shape (Fig. 4A,B): straight (the shaft is almost rectilinear and the proximal end is only slightly displaced laterally related to the distal end) or sinusoidal (the upper part of the shaft is markedly twisted sideways).

Remarks: for better visualization of this character, the femur should be positioned at the posterior view, and the position of the proximal epiphysis should be compared with the distal one.

(9) Position of the proximal epiphysis relative to the shaft (Fig. 4C,D,E): aligned (the main axis of the shaft, if extended proximally, separates the trochanters nearly equally and does not pass through the lesser trochanter), slightly twisted (the main axis of the shaft passes through the lesser trochanter but does not extend beyond it), or markedly twisted (the main axis of the shaft extends beyond most of the lesser trochanter).

(10) Notch between the lesser trochanter and the medial ridge (Fig. 2): absent (the proximal end of the medial ridge begins at the proximal end of the lesser trochanter, and no notch is seen), poorly developed (a little notch is seen), or well developed (a deep notch is seen).

(11) Development of the lateral ridge (Fig. 5): poorly developed (the ridge is long or short, but always low, sometimes imperceptible) or well developed (the ridge is usually short and prominent, or long and extends down through the shaft).

(12) Proximal end of the lateral ridge (Fig. 5): on the proximal region of the shaft (the ridge begins on the proximal half of the shaft) or on the distal region (the ridge begins on the distal half of the shaft).

Remarks: the best view to observe this character and the previous one is the posterior view, but the lateral ridges of some groups (e.g. Mormoopidae) are anteriorly displaced and we suggest turning laterally the femur to find the ridge and its proximal end.

(13) Intercondylar fossa (Fig. 4F,G): narrow (the gap between the condyles is narrow) or broad (the gap between the condyles is wide).

Quantitative analysis

Five characters of the femur (Fig. 1) were selected according to Riskin et al. (2005) and Coutinho et al. (2013): femur length (FL), from the proximal femoral head to the distal end of the condyles; antero-posterior width (APW), from the anterior to the posterior edge of the midshaft; latero-medial width (LMW), from the lateral to the medial edge of the midshaft; head proximo-distal height (HH), from the proximal to the distal end of the head; and head latero- medial width (HW), from the lateral to the medial edge of the head. We also calculated a robusticity index, as the ratio of the antero-posterior width to the femur length (APW/FL). This latter variable has been used in studies of the hindlimb morphology in bats (e.g. Howell & Pylka, 1977; Riskin et al. 2005, 2016) and other flying vertebrates (Madan et al. 2017). Width measurements were taken on the middle of the shaft and did not include lateral or medial ridges, except in cases where the ridges extended distally through the shaft. All measurements were obtained using digital calipers (accuracy of 0.01 mm) and, whenever possible, we used the left femur to standardize data collection. When more than one specimen was available per species, mean values were computed and used in subsequent analyses.

Fig. 1 Quantitative characters and anatomical features of the left femur of Mystacina tuberculata (AMNH 160269). (A) Posterior view: FL (femur length); LMW (latero-medial width); HW (head latero-medial width). (B) Medial view: APW (antero-posterior width) and HH (head proximo-distal height). (C) Proximal and (D) distal epiphyses in posterior view. Scale bars: 1 mm.

Fig. 2 Anatomical characters of the femur (1, 2, 7 and 10). Posterior view of the proximal epiphysis of A: Eumops trumbulli (MN 6486); B: Histiotus velatus (MN 46491); C: Carollia brevicauda (MN 36941); D: Natalus macrourus (MN 81448). The medial ridge (arrows) is long, poorly developed and begins on the shaft, just below the lesser trochanter in Molossidae (A); long, moderately developed and begins on the lesser trochanter in Vespertilionidae (B); short, well developed and begins on the shaft in Phyllostomidae (C); short, well developed, and begins on the anterior side of the shaft in Natalidae (D). The trochanteric fossa (dashed black lines) is shallow in Molossidae (A) and Phyllostomidae (C), deep in Vespertilionidae (B), and poorly developed in Natalidae (D). 36

The notch between the lesser trochanter and the medial ridge (dashed white lines) is absent in vespertilionids (B), poorly developed in molossids (A) and natalids (D), and well developed in some phyllostomids (C). Note the presence of a deep notch between the lesser trochanter and the medial ridge, followed by a short and well- developed medial ridge in C. brevicauda (C). Scale bars: 1 mm.

Fig. 3 Anatomical characters of the femur (3 to 6). Posterior view of the proximal epiphysis of A: Pteronotus gymnonotus (MN 68073); B: Mystacina tuberculata (AMNH 160269); C: Diaemus youngii (MN 71037); D: Mops leucostigma (AMNH 170640); E: Eumops perotis (MN 64750); F: Miniopterus magnater (AMNH 235587). The level of the trochanters is compared by the horizontal dashed white lines. The trochanters are poorly developed (rounded) and reach the same level proximally in Mormoopidae (A), well developed and reach the same level proximally in Mystacinidae (B), well developed with the greater projecting further proximally than the lesser in Desmodontinae (C) and Molossidae (D), and well developed with the lesser projecting further proximally than the greater in Miniopteridae (F). The tubercle (arrows) and the posterior ridge (dotted black lines) are absent in mormoopids (A) and Miniopterus (F), poorly developed in desmodontines (C), and well developed in M.

tuberculata (rounded; B), and Molossidae (hook-shaped; D, E). Note that the posterior ridge in E. perotis (E) extends distally by the shaft. Scale bars: 1 mm.

Fig. 4 Anatomical characters of the femur (8, 9 and 13). Posterior view of the femur of A: Noctilio albiventris; MN 64141) and B: Nalatus macrourus (MN 81448). Proximal epiphysis of C: Eptesicus brasiliensis (MN 71417); D: Carollia brevicauda (MN 36941); E: Furipterus horrens (MN 78114). Distal epiphysis of F: Histiotus montanus (AMNH 205649) and G: Vampyrum spectrum (MN 46499). The femur is straight in Noctilionidae (A), and sinusoidal in Natalidae (B). The proximal epiphysis is aligned (dashed vertical white lines) with the shaft in Vespertilionidae (C), slightly twisted in some Phyllostomidae (D), and markedly twisted in Furipteridae (E). The intercondylar fossa is narrow (dashed horizontal black lines) in vespertilionids (F) and broad in phyllostomids (G). Note that the condyles in H. montanus (F) are elongated, while in V. spectrum (G) are rounded. Scale bars: 2 mm.

Fig. 5 Anatomical characters of the femur (11 and 12). Posterior view of the femur of A, E (proximal epiphysis): Myotis ruber (MN 80299); B, G (proximal epiphysis): Pteronotus gymnonotus (MN 68073); C, F (distal epiphysis): Neoplatymops mattogrossensis (MN 36030). Anterior view of the femur of D: Desmodus rotundus (MN 36574). The lateral ridge (arrows) is poorly developed and begins on the proximal region of the shaft in some Vespertilionidae (just below the greater trochanter; A, E), and in Mormoopidae (at the shaft, displaced laterally; B, G). In Molossidae (C, F) the lateral ridge is well developed and begins on the distal region of the shaft. In Desmodontinae (D), the lateral ridge is well developed, begins on the proximal region, and extends distally. Scale bars: 2 mm.

To explore major trends in the morphology of the femur, accounting at the same time for the phylogenetic relatedness among species (they are not independent observations), we used a phylogenetic principal component analysis (pPCA) (Revell, 2009). We used a pruned version of the Chiroptera phylogeny proposed by Amador et al. (2016) as source of distance data among species (branch lengths). For a few genera, we had species in our sample that were not in the original tree. We solved this problem in two ways: replacing a congeneric that is not in our sample or adding our species using the same branch length found in the closest congeneric available in the original tree. In the first case, we performed the following replacements: Micronycteris minuta (Gervais, 1856) by Micronycteris sanborni Simmons, 1996; Lonchophylla mordax Thomas, 1903 by Lonchophylla dekeyseri Taddei, Vizotto & Sazima, 1983; Pteronotus personatus (Wagner, 1843) by Pteronotus psilotis (Dobson, 1878); Natalus tumidirostris Miller, 1900 by Natalus macrourus (Gervais, 1856); Cynomops abrasus

(Temminck, 1827) by Cynomops greenhalli (Goodwin, 1958); Cynomops paranus (Thomas, 1901) by Cynomops planirostris (Peters, 1865); Promops centralis Thomas, 1915 by Promops nasutus (Spix, 1823); Histiotus macrotus (Poeppig, 1835) by Histiotus montanus (Philippi & Landbeck, 1861), and Histiotus magellanicus (Philippi, 1866) by Histiotus velatus (I. Geoffroy, 1824). Phylogenetic affinities supporting the replacements in Micronycteris, Lonchophylla, Natalus, and Cynomops were obtained, respectively, from Simmons (1996), Rojas et al. (2016), Dávalos (2005), and Moras et al. (2016); Promops and Histiotus were represented in the original tree by one and two species, respectively, and did not require phylogenetic information to guide the replacements. When adding species, the position of insertion was defined according to Gregorin (2009), for Eumops trumbulli (Thomas, 1901), Loureiro (2014) for Molossus pretiosus Miller, 1902, and Lindsey & Ammerman (2016) for Molossus coibensis J.A. Allen, 1904 and Molossus currentium Thomas, 1901. The final tree (Fig. 6) was generated with the R package phytools (Revell, 2012).

We log-transformed all measurements before performing the pPCA analysis, and the matrix C, used to introduce the phylogenetic information in the calculation of the phylogenetic covariance matrix (matrix R), was generated adopting the lambda parameter (λ) (Revell, 2009). This parameter is obtained by an optimization process in a maximum likelihood analysis and is bounded from 0 to 1. The closer to 1 the value is, the stronger is the evidence that the data are phylogenetically structured under a Brownian motion model of evolution, justifying their treatment under such a model (in , if λ=1, distance between the last common ancestor of each pair of taxa and the root will be as derived from the phylogeny adopted). Conversely, if values near zero better fit to the data, these distances will be reduced, as the result of their multiplication by λ. It was not the case here (the recovered most likely λ was 0.88), but if λ is equal to zero (no phylogenetic pattern indicated by the data), the pPCA will work as a non- phylogenetic PCA (Monteiro, 2013). We also mapped a phylogenetic tree with the PCs provided by the pPCAs. This procedure creates a phylomorphospace that is useful as a visualization tool, highlighting convergences (Sidlauskas, 2008; Monteiro, 2013). As adopted here, however, the term convergence only reflects the evolution of similar phenotypes in independent lineages, does not requiring a particular process (e.g. adaptation) (Stayton, 2015). These analyses were performed using the R package phytools (Revell, 2012).

Fig. 6 Phylogenetic relationships among Yangochiroptera bats studied here, based on Amador et al. (2016). The original tree was pruned for species absent in our sample and a few species were added (see text). Cistugidae and Myzopodidae were not accessed.

To compare families of Yangochiroptera in respect to median values of key variables revealed as informative in the pPCA analyses (e.g. head of the femur and robusticity), we used notched boxplots graphs (Chambers et al. 1983) prepared in R (R Core Team, 2018). If two

boxes' notches do not overlap, there is ‘strong evidence’ (95% confidence) their medians differ (Chambers et al. 1983). In these comparisons, we kept Desmodontinae separated from Phyllostomidae due to the peculiar morphology of their femur (Howell & Pylka, 1977; Swartz, 1997). Phyllostomidae, highly diverse both taxonomically and ecologically, were also analyzed in detail.

Scaling patterns in femur dimensions were accessed via phylogenetic generalized least squares (PGLS) regressions, using the R package Caper (Orme, 2013). In these analyses, isometry in linear dimensions was related to a scaling exponent (b) of 0.33, and in the case of robusticity the value considered was 0 (Madan et al. 2017). Confidence limits for slopes were calculated using model parameters and the t-distribution, and predictive limits plotted on graphs were calculated according to Smaers & Rohlf (2016). The strength of the correlation (R2) between variables and the statistical significance of those correlations (p-value) are also displayed. As in the case of the p A, the agel’s lambda (λ) was adopted here (values varied from 0.174 to 0.774), and we also log-transformed all variables excepting the Robusticity index (Madan et al. 2017). We used diagnostic plots in R to check assumptions of the PGLS models (normality and homogeneity of the residuals; Mundry, 2014) and found no substantial violation. Species average body mass were obtained from literature (e.g. Kumirai & Jones, 1990; Rodríguez-Durán & Kunz, 1992; Giannini & Barquez, 2003; Cole & Wilson, 2006; Reis et al. 2017).

Results

Anatomical characterization

The femur of yangochiropteran bats can be generally described as presenting two trochanters, a developed and rounded head, slightly anteriorly displaced, and two ridges on the shaft. We found remarkable differences among the families studied here (Fig. 7, Table S1), which allowed us to develop the diagnoses presented below and an identification key (Appendix 1). Intergeneric and interspecific variation is also reported, including when expressed by mensural data.

Fig. 7 Posterior view of the femur in yangochiropteran bats. (A) Emballonuridae (Saccopteryx bilineata; MN 70914), (B) Nycteridae (Nycteris macrotis; AMNH 187320), (C) Mystacinidae (Mystacina tuberculata; AMNH 160269), (D) Phyllostomidae (Desmodus rotundus; MN 36574), (E) Mormoopidae (Pteronotus gymnonotus; MN 68073), (F) Noctilionidae (Noctilio albiventris; MN 64141), (G) Furipteridae (Furipterus horrens; MN 78115), (H) Thyropteridae (Thyroptera tricolor; ALP 2912), (I) Natalidae (Natalus macrourus; MN 81448), (J) Molossidae (Molossus pretiosus; MN 71566), (K) Vespertilionidae (Myotis levis; MN 80299), and (L) Miniopteridae (Miniopterus magnater; AMNH 235587). Scale bar: 2 mm.

Family Mystacinidae Dobson, 1875 (Fig. 8D,F)

Mystacina tuberculata Gray, 1843 has a sturdy and straight femur, with a medial ridge moderately developed, and trochanters that reach the same level proximally. The greater trochanter is remarkable in this family, presenting a well-developed tubercle and posterior ridge. The lateral ridge is short and well developed, rising on the proximal region, nearly at the middle of the shaft.

Family Thyropteridae Miller, 1907 (Fig. 8C,E)

Thyroptera tricolor Spix, 1823 has a straight femur with a moderately-developed medial ridge that begins on the lesser trochanter or immediately bellow it. The lateral ridge is short and well developed, and there is a shallow trochanteric fossa. The distal condyles are rounded and the intercondylar fossa is narrow.

Family Furipteridae Gray, 1866 (Fig. 9E)

Furipterus horrens (F. Cuvier, 1828) presents a slender femur, with one of the smallest diameters among all bats studied here (LMW: 0.50–0.56 mm). Its femur is characterized by the presence of poorly-developed and rounded trochanters, where the lesser projects slightly more proximally than the greater, and by a short and moderately-developed medial ridge. The femur is sinusoidal, with the proximal epiphysis markedly twisted. The lateral ridge is displaced toward the anterior side and is poorly developed.

Remarks: the femur morphology is similar to that of mormoopids, but in the latter the shaft is more robust (LMW>0.67 mm), the trochanters usually reach the same level proximally, and the medial ridge is longer, extending onto the shaft.

Family Noctilionidae Gray, 1821 (Fig. 8A,B)

Noctilionids have a straight femur, with trochanters that are well developed (the greater being more robust than the lesser) and separated by a shallow fossa. he greater trochanter projects further proximally than the lesser, especially in Noctilio albiventris Desmarest, 1818. he medial ridge is moderately developed in both species, but it is more pronounced in Noctilio leporinus (Linnaeus, 1758). his ridge begins on the lesser trochanter in N. leporinus or on the shaft in N. albiventris. he lateral ridge also differs in these species, being usually poorly developed in N. albiventris and well developed in N. leporinus. Size is also clearly useful, being the femur in N. albiventris consistently smaller ( L: 18.11–21.02 mm; LMW: 1.31–1.54 mm) than in N. leporinus ( L: 27.48–30.80 mm; LMW: 1.86–2.05 mm).

Family Mormoopidae Saussure, 1860 (Fig. 9A–C)

Mormoopids have a very differentiated femur morphology, compared to other bats. The trochanters are poorly developed and rounded, usually reaching the same level proximally, and the notch between them is very small (poorly-developed trochanteric fossa). The femur is sinusoidal with the proximal epiphysis strongly twisted. The medial ridge is long and moderately developed. The lateral ridge is displaced caudally, rising on the proximal end of the shaft, and is well developed in Pteronotus gymnonotus Wagner, 1843, Pteronotus davyi Gray, 1838, P. psilotis, and Mormoops blainvillei Leach, 1821.

Family Phyllostomidae Gray, 1825 (Figs 10–13)

Members of this family generally present a sinusoidal femur, well-developed trochanters, and a broad intercondylar fossa. A conspicuous feature present in most members is the presence of a deep notch between the lesser trochanter and the medial ridge, followed by a short and well- developed medial ridge. Below we present data for each subfamily, grouping only those for which the main features are shared, regardless their phylogenetic proximity.

Subfamily Desmodontinae Wagner, 1840 (Fig. 10)

Desmodontines have unique femoral characters among all bats studied here: the shaft is straight and antero-posteriorly flattened, which is more conspicuous in Desmodus rotundus (É. Geoffroy, 1810) and Diaemus youngii (Jentink, 1893) than in Diphylla ecaudata Spix, 1823, and the lateral and medial ridges are long and form grooves that are easily seen in lateral and anterior views. From other phyllostomids, Desmodontinae can also be differentiated by presenting a greater trochanter that projects further proximally than the lesser and bears a small tubercle, in addition to a poorly-developed posterior ridge. The degree of development of the tubercle and posterior ridge may vary among and within species. Diaemus youngii has the widest femur (LMW: 2.38–2.51 mm) among desmodontines, being followed by D. rotundus (LMW: 2.09–2.42 mm) and D. ecaudata (LMW: 1.42–1.87 mm).

Subfamilies Macrotinae Van Den Bussche, 1992, Micronycterinae Van Den Bussche, 1992, Lonchorhininae Gray, 1866, Phyllostominae Gray, 1825, and Glyphonycterinae Baker et al. 2016 (Fig. 11)

In these subfamilies, the femur is typically sinusoidal, the medial ridge is short and well developed, and the lesser trochanter projects further proximally than the greater. The lateral ridge begins on the proximal region of the shaft, but its position varies among genera. While this ridge appears in the first fifth of the shaft in most species, in Phyllostomus spp., Phylloderma stenops Peters, 1865, Macrophyllum macrophyllum (Schinz, 1821), and Vampyrum spectrum (Linnaeus, 1758) it appears in the second or third fifth of the shaft. Gardnerycteris crenulatum (É. Geoffroy, 1810) (Phyllostominae) and Trinycteris nicefori Sanborn, 1949 (Glyphonycterinae) share a unique condition in this group, with the medial ridge beginning on the proximal region of the lesser trochanter (begins on the shaft in other species), giving a straight appearance to the proximal epiphysis.

Subfamilies Glossophaginae Bonaparte, 1845 (Fig. 12A–G) and Lonchophyllinae Griffiths, 1982 (Fig. 12H–I)

The femur is small and slender in almost all species, with the exceptions of Leptonycteris yerbabuenae Martínez & Villa-R, 1940, Brachyphylla cavernarum Gray, 1834, and Erophylla sezekorni (Gundlach, 1860). The shaft is markedly sinusoidal, the medial ridge is well developed, and the lesser trochanter projects further proximally than the greater. There is an intraspecific variation in the development of the lateral ridge in Glossophaga and Anoura, being poorly or well developed. Brachyphylla cavernarum and Choeroniscus minor (Peters, 1868) present a well-developed lateral ridge.

Subfamily Carolliinae Miller, 1924 (Fig. 13L)

Carollia spp. have a typical sinusoidal phyllostomid femur, with a well-developed medial ridge that begins on the shaft and a poorly-developed lateral ridge that in some specimens is almost imperceptible.

Subfamily Rhinophyllinae Baker et al. 2016 (Fig. 13M)

Rhinophylla spp. have the proximal epiphysis aligned with the shaft, giving a straighter appearance for the femur. The medial ridge is long and the lateral ridge is long and well developed.

Subfamily Stenodermatinae Gervais, 1855 (Fig. 13A–K)

Similar to desmodontines and rhinophyllines, stenodermatines have the proximal epiphysis aligned or slightly twisted in relation to the shaft. There is an interspecific variation in the development of the lateral ridge, but it always begins on the proximal region of the shaft. Sturnira spp. are noticeably variable: the medial ridge may begin on the lesser trochanter; the femur is straight; the proximal epiphysis is aligned with the shaft; the notch between the lesser trochanter and the medial ridge is poorly developed; and the shaft is broad, especially in the proximal region.

Family Nycteridae Van der Hoeven, 1855 (Fig. 14A)

This monotypic family is represented here by a single species, Nycteris macrotis Dobson, 1876, and can be recognized by a well-developed medial ridge that begins on the shaft of the femur, a short and well-developed lateral ridge, and well-developed trochanters, which nearly reach the same level proximally. The femur is greatly sinusoidal and the torsion of the proximal epiphysis is the most conspicuous of all families considered here.

Remarks: the femur morphology is similar to that of some phyllostomids, but it can be distinguished by the position of the trochanters, by the well-developed lateral ridge, and by the torsion of the proximal epiphysis. Additionally, in phyllostomids the head of the femur is rounded and well defined, reaching a more proximal level than the trochanters, whereas in N. macrotis the head is at the level of the trochanters.

Family Emballonuridae Gervais, 1856 (Fig. 14B–G)

Emballonurids have a slender and straight femur, with a subtle lateral inclination in the proximal epiphysis. The medial ridge is moderately developed and short, the trochanters are well developed, and the trochanteric fossa is shallow. Intergeneric variation is seen on the proximal epiphysis: it is aligned with the shaft in Cormura brevirostris (Wagner, 1843), Diclidurus isabella (Thomas, 1920), Peropteryx spp. and Taphozous spp., while in Saccopteryx spp. and Rhynchonycteris naso (Wied-Neuwied, 1820) varies from aligned to slightly twisted. Taphozous spp. and R. naso have the lesser trochanter projecting just slightly more proximally than the greater, whereas in other genera the difference is more conspicuous.

Family Natalidae Gray, 1866 (Fig. 9D)

A sinusoidal femur, well-developed and short medial ridge, poorly-developed trochanters and trochanteric fossa, and a marked twist in the proximal epiphysis relative to the shaft are the characters that best describe natalids. Nyctiellus lepidus (Gervais, 1837) is consistently smaller in femur length than Natalus macrourus and Natalus major Miller, 1902 (12.88 mm, 18.89– 21.19 mm, and 22.22 mm, respectively).

Remarks: in Natalidae the femur morphology is similar to that of Furipteridae, but in the former both medial and lateral ridges are more developed, and the trochanters are more elongate (rounded in furipterids).

Family Molossidae Gervais, 1856 (Fig. 15) 47

he main characteristics of molossids’ femur concern to the greater trochanter, which presents a well-developed hook-shaped tubercle—“hook-like process” (Simmons, 1998) and a well- developed posterior ridge. Moreover, the medial ridge is poorly developed and long, with the proximal region slightly pronounced in some specimens. The femur is straight and the trochanters are well developed, the greater projecting further proximally than the lesser. The lateral ridge is usually well developed and, although in all genera it appears on the distal region, there is noticeable variation among them: it appears in the third fifth of the femur in Eumops and Promops; in the fourth fifth in Chaerophon, Cynomops, Molossops, Mops, Neoplatymops, and Tadarida; and in the last fifth in Cheiromeles torquatus. In Molossus, the lateral ridge appears between the third and fourth fifth of the femur. The genus Nyctinomops presents interspecific variation: it appears almost on the mid-shaft in Nyctinomops macrotis (Gray, 1840); in the third fifth in Nyctinomops femorosacus (Merriam, 1899), and between the third and fourth fifth in Nyctinomops laticaudatus (É. Geoffroy, 1805).

Family Miniopteridae Dobson, 1875 (Fig. 16E,F)

Miniopterus magnater Sanborn, 1931 has a straight femur with a well-developed medial ridge, that begins on the lesser trochanter, which projects further proximally than the greater. A narrow intercondylar fossa is shared with vespertilionids and thyropterids. The distal condyles are rounded.

Family Vespertilionidae Gray, 1821 (Fig. 16A–D)

Vespertilionids have a straight femur, with well-developed trochanters that are separated by a deep fossa. The medial ridge is moderately developed and the proximal end usually begins on the lesser trochanter (except in Lasiurus). The intercondylar fossa is narrow and the distal condyles are elongated (markedly compressed mediolaterally). The position of the trochanters varies among the species analyzed: in Eptesicus brasiliensis (Desmarest, 1819) and Histiotus spp. the trochanters nearly reach the same level proximally; in Lasiurus spp. the lesser trochanter projects further proximally than the greater; and in Myotis spp. there is intraspecific variation, with the trochanters reaching the same level proximally or the greater projecting slightly more proximally than the lesser. The tubercle on the greater trochanter, when present, is poorly developed in Myotis spp.; in other genera it is absent. A poorly-developed posterior ridge in the greater trochanter is present in E. brasiliensis and in some species of Myotis. The lateral ridge is well developed and long in Myotis spp. and poorly developed in other genera. 48

Lasiurus spp. have a poorly-developed notch between the lesser trochanter and the medial ridge, while the other genera have no such notch.

Fig. 8 Posterior view of the femur in Noctilionidae [A: Noctilio leporinus, MN 71580 (right femur mirrored); B: Noctilio albiventris, MN 64141], Thyropteridae (C, E: Thyroptera tricolor, AMNH 266361) and Mystacinidae (D, F: Mystacina tuberculata, AMNH 160269). The femur is straight in these groups and the proximal epiphysis is aligned with the shaft. The medial ridge (lined arrows) is more developed and begins at the lesser trochanter in N. leporinus (A) and T. tricolor (C) while in N. albiventris (B) and M. tuberculata (D) it begins at the shaft. The lateral ridge (dashed arrows) is poorly developed in N. albiventris (B) and well developed in other species. Note the tubercle and posterior ridge in the greater trochanter (dotted arrows) of M. tuberculata (D). The intercondylar fossa (dashed black lines) is narrow in T. tricolor (E; distal view) and broad in M. tuberculata (F; distal view). Scale bars: 1 mm.

Fig. 9 Posterior view of the proximal epiphysis of the femur in Mormoopidae (A: Pteronotus rubiginosus, MN 80564; B: Pteronotus gymnonotus, MN 68073; C: Mormoops blainvillei, AMNH 238144), Natalidae (D: Natalus macrourus, MN 81448), and Furipteridae (E: Furipterus horrens, MN 78115). The proximal epiphysis is twisted laterally and the poorly-developed trochanters are rounded or elongated (in N. macrourus, D). Note that the medial ridges (lined arrows) are long in Mormoopidae (A-C) and short in Natalidae (D) and Furipteridae (E), and the lateral ridges (dashed arrows) are more developed in P. gymnonotus (B), M. blainvillei (C), and N. macrourus (D). Scale bars: 1 mm.

Fig. 10 Posterior (A-C), medial (D), and anterior (E) views of the femur in Desmodontinae (A, D, E: Desmodus rotundus, MN 36574; B: Diaemus youngii, MN 71037; C: Diphylla ecaudata, MN 68035). The dashed and lined arrows indicate the lateral and medial ridges, respectively. Note that the ridges are well developed and begin on the proximal part of the shaft and extend until the distal part, forming grooves easily seen in the lateral view (D). The dotted arrow (B) points at the small tubercle on the greater trochanter. Note that the robusticity varies from D. youngii (B) to D. ecaudata (C). Scale bars: 2 mm.

Fig. 11 Posterior view of the proximal epiphysis of the femur in Phyllostominae (A: Vampyrum spectrum, MN 46499 (right femur mirrored); B: Chrotopterus auritus, MN 70862; C: Phyllostomus hastatus, MN 56238; D: Phylloderma stenops, MN 70861; E: Trachops cirrhosus, MN 71469; F: Tonatia saurophila, MN 70860; G: Lophostoma silvicola, MN 79730; H: Macrophyllum macrophyllum, MN 70600; I: Mimon bennettii, MN 79827; J: Gardnerycteris crenulatum, MN 75189), Glyphonycterinae (K: Trinycteris nicefori, AMNH 267878), Micronycterinae (L: Micronycteris microtis, MN 80534), Lonchorhininae (M: Lonchorhina aurita, MN 79802), and Macrotinae (N: Macrotus waterhousii, MN 236658). The proximal epiphysis of these subfamilies is usually laterally twisted, the medial ridge (lined arrows) is short, well developed, and pronounced. The lateral ridge (dashed arrows) may begin at the first fifth of the femur (B, E, F, G, J) or more distally (C, D). Note the difference in the position of the medial ridge—it begins at the shaft in M. bennettii (I) and M. microtis (L) and at the lesser trochanter in G. crenulatum (J) and T. nicefori (K). Scale bars: 2 mm (A-H, M-N); 1 mm (I-L).

Fig. 12 Posterior view of the proximal epiphysis of the femur in Glossophaginae (A: Brachyphylla cavernarum, AMNH 188238 (right femur mirrored); B: Erophylla sezekorni, AMNH 176294; C: Leptonycteris yerbabuenae, AMNH 189692; D: Monophyllus redmani, AMNH 39523; E: Anoura geoffroyi, MN 37356; F: Glossophaga soricina, MN 56284; G: Choeroniscus minor, AMNH 267153) and Lonchophyllinae (H: Lonchophylla dekeyseri, MN 80525; I: Hsunycteris thomasi, AMNH 266103). They have a well-developed medial ridge (lined arrows). Note the larger femur of B. cavernarum (A), E. sezekorni (B), and L. yerbabuenae (C) and the well-developed lateral ridges (dashed arrows) in B. cavernarum (A) and A. geoffroyi (E). Scale bars: 2 mm.

Fig. 13 Posterior view of the proximal epiphysis of the femur in Stenodermatinae (A: Artibeus obscurus, MN 36525; B: Chiroderma villosum, MN 71370; C: Platyrrhinus lineatus, MN 56213; D: Sturnira lilium, MN 36192; E: Uroderma bilobatum, MN 70918; F: Pygoderma bilabiatum, MN 81269; G: Dermanura gnoma, MZUSP

22544; H: Vampyriscus bidens, MN 70921; I: Mesophylla macconnellii, MN 70887 (right femur mirrored); J: Ametrida centurio, AMNH 264276; K: Vampyressa brocki, AMNH 268567), Carolliinae (L: Carollia brevicauda, MN 36941), and Rhinophyllinae (M: Rhinophylla fischerae, MN 70252). Note that the femur of stenodermatines and rhinophyllines are straighter than in other phyllostomids, and the proximal epiphysis is straight (D, M) or slightly twisted (others). Note the robusticity of the femur of S. lilium (D). The medial ridge (lined arrows) is poorly (A, D, F, J, M) or well (B, C, E, G-I, K, L) pronounced laterally. Note the poorly-developed lateral ridge (dashed arrows) in C. brevicauda (L) and the long and well-developed lateral ridge in R. fischerae (M). Scale bars: 1 mm.

Fig. 14 Posterior view of the proximal epiphysis of the femur in Nycteridae (A: Nycteris macrotis, AMNH 187320) and Emballonuridae (B: Taphozous melanopogon, AMNH 235571; C: Diclidurus isabella, MN 70644; D: Saccopteryx bilineata, MN 70914; E: Cormura brevirostris, AMNH 267827 70887 (right femur mirrored); F: Peropteryx macrotis, MN 79797; G: Rhinchonycteris naso, MN 70430). Note the well-developed medial (lined arrow) and lateral (dashed arrow) ridges in N. macrotis (A) and a head that does not extend proximally the level of the trochanters. The medial ridge is moderately developed in Emballonuridae, the lateral ridge is poorly developed, and the lesser trochanter projects further proximally than the greater in most species but reaches nearly the same level in T. melanopogon (B) and R. naso (G). Saccopteryx bilineata (D) has the proximal epiphysis more laterally displaced than others emballonurids. Scale bars: 1 mm.

Fig. 15 Posterior view of the femur in Molossidae (A: Cheiromeles torquatus, AMNH 247583; B: Eumops glaucinus, MN 71472; C: Nyctinomops macrotis, MN 49583; D: Promops nasutus, MN 64677; E: Tadarida brasiliensis, MN 6562; F: Mops leucostigma, AMNH 170640 (right femur mirrored); G: Chaerophon plicatus, AMNH 27372; H: Molossus pretiosus, MN 71566; I: Cynomops planirostris, MN 70276; J: Molossops temminckii, MN 71355; K: Neoplatymops mattogrossensis, MN 36030). Note the robusticity of the femur of C. torquatus (A) and how the lateral ridge (dashed arrows) are distally displaced and varies among the genera. The lined arrows point at the hook-shaped tubercle in some species. Scale bars: 2 mm.

Fig. 16 Posterior view of the femur in Vespertilionidae (A: Lasiurus blossevillii, MN 71304; B: Histiotus velatus, MN 46491; C: Eptesicus brasiliensis, MN 71417; D: Myotis nigricans, MN 71530) and Miniopteridae (E; 54

proximal epiphysis, F; distal epiphysis: Miniopterus magnater, AMNH 235587). The shaft is straight in these groups and the proximal end of the medial ridge (lined arrows) begins on the lesser trochanter, except for L. blossevillii (A), where the ridge begins more distally, on the shaft. Note the narrow intercondylar fossa (B and F; black line and dotted arrows) and the elongated condyles in vespertilionids, while in miniopterids the latter are more rounded. The lateral ridge (dashed arrows) is usually poorly developed, except in some Myotis (D). Scale bars: 1 mm.

Quantitative analysis

The first three phylogenetic principal components accounted for nearly 98% of the variance in our femur dataset (loadings in Table 2). The first component (pPC1; Fig. 17), expressing most of this variation (91.6%), is related to size, arranging species mainly according to the length of the femur (larger species on the left; e.g. V. spectrum and Cheiromeles torquatus Horsfield, 1824). The second component (pPC2) accounted for 3.8% of the variation and is associated with the robusticity of the femur, with length and width measurements of this bone influencing with opposite signals (Table 2). Along this axis (Fig. 18), less robust species are on the right (e.g. Furipteridae, Mormoopidae) and more robust on the left (e.g. Desmodontinae, Molossidae). Surprisingly, Myotis simus Thomas, 1901 appeared among Desmodontinae, distant from other vespertilionids. The third component (pPC3; Fig. 18), accounted for 2.9% of the variation and is related to variables obtained from the head of the femur. In this axis molossids depart from Desmodontinae and are approached only by a few vespertilionids and by Sturnira erythromos (Tschudi, 1844). When the major axes of shape variation were examined in a phylomorphospace formed by pPC2 and 3 (Fig. 19), the independent colonization of particular regions of the shape space became evident, as recorded for some vespertilionids, clearly associated to Desmodontinae and Molossidae.

Table 2 Loadings on the first three components of a phylogenetic principal components analysis based on femur measurements of yangochiropteran bats. See Material and Methods for variable abbreviations.

Variables PC1 PC2 PC3 FL -0.9043 0.3898 0.1629 APW -0.9637 -0.2152 0.0864 LMW -0.9705 -0.1008 0.1558 HH -0.9556 0.0381 -0.2772 HW -0.9789 0.0233 -0.1095

Fig. 17 Scatterplot of scores for the two first components of a phylogenetic principal components analysis (pPCA) based on femur measurements of yangochiropteran bats. pPC1 depicts general size (larger bats on the left side) and pPC2 is arranging species according to femur robusticity (more robust species on the bottom). Convex hulls group species per families (MOL, Molossidae, green; MOR, Mormoopidae, pink; VES, Vespertilionidae, blue; EMB, Emballonuridae, yellow; NAT, Natalidae, orange; PHY, Phyllostomidae, dashed line) or subfamily, in the case of Desmodontinae (DES, red). The following symbols indicate families represented by one or two species: Mystacinidae ( ), Noctilionidae ( ), Miniopteridae ( ), Furipteridae ( ), Nycteridae ( ) and Thyropteridae ( ). Complete species names corresponding to the abbreviations are present in Table 1.

Fig. 18 Scatterplot of scores for the second and third components of a phylogenetic principal components analysis (pPCA) based on femur measurements of yangochiropteran bats. pPC2 is arranging species according to femur robusticity (more robust species on the left side) and pPC3 according to the size of the femur head (bigger heads on the bottom). Convex hulls group species per families (MOL, Molossidae, green; MOR, Mormoopidae, pink; VES, Vespertilionidae, blue; EMB, Emballonuridae, yellow; NAT, Natalidae, orange; PHY, Phyllostomidae, dashed line) or subfamily, in the case of Desmodontinae (DES, red). The following symbols indicate families represented by one or two species: Mystacinidae ( ), Noctilionidae ( ), Miniopteridae ( ), Furipteridae ( ), Nycteridae ( ) and Thyropteridae ( ). Complete species names used here are present in Table 1. Mystacina, vespertilionids, and molossids are separated along the pPC3. The purple polygon remarks Sturnira, a group that seems to be an outlier with respect to Stenodermatinae.

Fig. 19 Phylomorphospace projection based on a pruned version of the phylogeny generated from Amador et al. (2016) for yangochiropteran families. The projection is performed on the second and third axes of a phylogenetic principal components analysis (pPCA) based on femur measurements. MOL, Molossidae, green; MOR, Mormoopidae, pink; VES, Vespertilionidae, blue; EMB, Emballonuridae, yellow; NAT, Natalidae, orange; PHY, Phyllostomidae, grey; DES, Desmodontinae, red. Sturnira, a genus that seems to be an outlier of Phyllostomidae, is in purple. The following symbols indicate families represented by one or two species: Mystacinidae ( ), Noctilionidae ( ), Miniopteridae ( ), Furipteridae ( ), Nycteridae ( ) and Thyropteridae ( ). Note the occupation of several species of espertilionidae in molossids’ and desmodontines’ morphospace.

The robusticity of the femur is highly variable among Yangochiroptera families, and achieved the highest values in Molossidae, Mystacinidae, Desmodontinae, and Noctilionidae (Fig. 20). Vespertilionidae and Phyllostomidae form the second group in development of this index, but M. simus is a noticeable outlier. The lowest robusticity values were found in Miniopteridae, Emballonuridae, Mormoopidae, Nycteridae, Natalidae, Thyropteridae, and Furipteridae. Phyllostomids exhibited a great range in robusticity (Fig. 21), with Desmodontinae standing out with the highest values and Lonchorhininae with the lowest. Stenodermatinae appeared as the second most robust group.

In a further analysis of the femur head morphology, useful in the quantitative separation of molossids from other families (Fig. 18), we evaluated among-family differences using ratios of head length and head width to femur length. Femur head width was the most informative in this case, and highlighted Mystacinidae, in addition to Molossidae, from all other bats (Fig. 22). Different from femur robusticity results, in this analysis Desmodontinae showed values even lower than Vespertilionidae, a family that also surpassed noctilionids.

Fig. 20 Notched boxplot showing variation in femur robusticity (ratio of antero-posterior width to length) among yangochiropteran families (Desmodontinae separated from Phyllostomidae to highlight its high values). MOL: Molossidae; MYS: Mystacinidae; DES: Desmodontinae; NOC: Noctilionidae; VES: Vespertilionidae; PHY: Phyllostomidae; MIN: Miniopteridae; EMB: Emballonuridae; MOR: Mormoopidae; NYC: Nycteridae; NAT: Natalidae; THY: Thyropteridae; FUR: Furipteridae.

Fig. 21 Notched boxplot showing variation in femur robusticity (ratio of antero-posterior width to length) among phyllostomid subfamilies. DES: Desmodontinae; STE: Stenodermatinae; RHI: Rhinophyllinae; GLI: Gliphonycterinae; CAR: Carolliinae; PHY: Phyllostominae; GLO: Glossophaginae; LON: Lonchophyllinae; MIC: Micronycterinae; MAC: Macrotinae; LOC: Lonchorhininae.

Fig. 22 Notched boxplot showing variation in the ratio of femur head latero-medial width to femur length among yangochiropteran families (Desmodontinae separated from Phyllostomidae). MOL: Molossidae; MYS: Mystacinidae; DES: Desmodontinae; NOC: Noctilionidae; VES: Vespertilionidae; PHY: Phyllostomidae; MIN: Miniopteridae; EMB: Emballonuridae; MOR: Mormoopidae; NYC: Nycteridae; NAT: Natalidae; THY: Thyropteridae; FUR: Furipteridae.

All femur variables analyzed in the PGLS regressions presented statistically significant scaling relationships with body mass (Table 3 and Figs 23–25), but the strength of these relationships was highly variable. The width of the femur showed the highest proportion of explained variation (R2=0.82) and did not deviate from isometry (b=0.3510±0.0292), whereas femur length exhibits negative allometry (b=0.2466±0.0423) and was only moderately explained in the model (R2=0.51). Robusticity showed the least amount of explained variation (R2=0.17) and is under weak positive allometry (b=0.0045±0.0018). In this model, C. greenhalli and C. torquatus were found to be more robust than expected for their size, and several other molossids approached the upper limit of the predictive interval (Fig. 25). The most robust bat for its size, however, was M. simus.

Table 3 Results of phylogenetic least squares regressions of femur dimensions to log body mass of yangochiropteran bats. CI: confidence interval for the slope.

Femur Intercept (a) Slope (b) Lower CI Upper CI Allometry p Adjusted R² dimensions Length 2.2726 0.2466 0.2043 0.2890 Negative < 2.2e-16 0.5152 Width -1.0652 0.3510 0.3218 0.3802 Isometry < 2e-16 0.8198 Robusticity 0.0370 0.0045 0.0028 0.0063 Positive 1.057e-03 0.17

Fig. 23 Phylogenetic generalized least squares regression of femur length plotted against log body mass (g) in yangochiropteran bats (MOL, Molossidae, green; MOR, Mormoopidae, pink; VES, Vespertilionidae, blue; EMB, Emballonuridae, yellow; NAT, Natalidae, orange; PHY, Phyllostomidae, grey; DES, Desmodontinae, red). The following symbols indicate families represented by one or two species: Mystacinidae ( ), Noctilionidae ( ), Miniopteridae ( ), Furipteridae ( ), Nycteridae ( ) and Thyropteridae ( ). Complete species names used here are present in Table 1. Dashed lines indicate predictive limits.

Fig. 24 Phylogenetic generalized least squares regression of femur width plotted against log body mass (g) in yangochiropteran bats (MOL, Molossidae, green; MOR, Mormoopidae, pink; VES, Vespertilionidae, blue; EMB, Emballonuridae, yellow; NAT, Natalidae, orange; PHY, Phyllostomidae, grey; DES, Desmodontinae, red). The following symbols indicate families represented by one or two species: Mystacinidae ( ), Noctilionidae ( ), Miniopteridae ( ), Furipteridae ( ), Nycteridae ( ) and Thyropteridae ( ). Complete species names used here are present in Table 1. Dashed lines indicate predictive limits.

Fig. 25 Phylogenetic generalized least squares regression of femur robusticity index plotted against log body mass (g) in yangochiropteran bats (MOL, Molossidae, green; MOR, Mormoopidae, pink; VES, Vespertilionidae, blue; EMB, Emballonuridae, yellow; NAT, Natalidae, orange; PHY, Phyllostomidae, grey; DES, Desmodontinae, red). The following symbols indicate families represented by one or two species: Mystacinidae ( ), Noctilionidae ( ), Miniopteridae ( ), Furipteridae ( ), Nycteridae ( ) and Thyropteridae ( ). Complete species names used here are present in Table 1. Dashed lines indicate predictive limits. Note the position of some molossids and Myotis simus, out of the predictive limits.

Discussion

Femur morphology

The bat femur has been reported as unique among mammals by presenting a globular head nearly aligned (vs. at an angle) with the shaft, in addition to a reduced or even absent neck (Flower & Gadow, 1885; Walton & Walton, 1970; Simmons, 1994). These characters are associated to a lateral, 90 degrees or more, rotation of the hindlimbs in bats (Wible & Novacek, 1988; Cheney et al. 2014), and were confirmed in all species studied here. With respect to phylogenetic relationships within Chiroptera, the femur has also proved to be informative, with 64

variation on the shaft (if straight or bended) allowing the separation of all fossil and most extant families from Nycteridae, Megadermatidae, Rhinolophidae, and some Phyllostomidae (Simmons & Geisler, 1998). No longer recognized as a close relative of Megadermatidae and Rhinolophidae (both yinpterochiropterans; Amador et al. 2016), Nycteridae was highlighted in our sample as presenting the strongest bending on the shaft of the femur among yangochiropterans. Shafts with this trend (reported here as sinusoidal) favor the raise of the knee joint even higher than the hip joint (Simmons & Geisler, 1998), and were also cited here for Furipteridae, Mormoopidae, and Natalidae, in addition to Nycteridae and Phyllostomidae.

Still in a systematic context, but dealing only with extant bats, Smith (1972) worked on a large number of femur characters and highlighted the proximal epiphysis as the most informative region of this bone. Our analysis also largely relied on the proximal epiphysis, where we found seven out of the thirteen diagnostic characters used here. Some of these characters are at the greater trochanter, a region that can be well developed and bears a tubercle and a posterior ridge, important taxonomic features. For example, the tubercle on the greater trochanter was considered by Simmons (1998) as a synapomorphy of Molossidae— “greater trochanter of femur with a hooklike process”. In our interpretation this character is also present in M. tuberculata and in some Desmodontinae, although in these groups it is a rounded process, not a hooklike one (here we refer to this process as a “tubercle”). rochanters are also important in the diagnose of Mormoopidae (Simmons & Conway, 2001), and Smith (1972) noticed that the reduced condition in this group is also present in Natalidae and Furipteridae. We reinforce this association here, highlighting that these three families present the most differentiated and simpler femur among yangochiropterans. In addition to the reduced trochanters, they also share a proximal epiphysis markedly displaced laterally and a trochanteric fossa poorly developed. Smith (1972) commented on the difficulties of discussing the possible functional meaning of reduced trochanters in these groups, but the features shared by Molossidae, Mystacina, and Desmodontinae deserves functional considerations, as discussed in the next section.

Analyzed from the perspective of categorical data, the ecologically diversified phyllostomid subfamilies (Rex et al. 2010) were conservative in several characters (e.g. sinusoidal femur, well-developed medial ridge). he morphology of G. crenulatum ( hyllostominae) and T. nicefori (Glyphonycterinae), however, are noteworthy. he former was only recently placed in a separate genus, apart from Mimon (Hurtado & acheco, 2014), and our data support this arrangement, showing a conspicuous difference in respect to Mimon

bennettii (Gray, 1838). he distinction between T. nicefori and Micronycteris spp. relied on the same character, and also supports the current arrangement. Trinycteris, with its sole species, was described as a subgenus of Micronycteris (Sanborn, 1949; Simmons & oss, 1998), and nowadays each genus is placed in a separate subfamily ( irranello et al. 2016). As reported for other characters above, the straight appearance of the proximal epiphysis in G. crenulatum and T. nicefori did not derive from common ancestry (see the phylogeny in Amador et al. 2016).

Quantitatively, our results pointed to three main directions in the femur morphological space. Size is the major source of variation, and the large overlap among families makes it difficult to use mensural data in taxonomic assessments at this level (e.g. Appendix 1). For some genera and species, however, there seems to be a potential for these measurements to be used in taxonomic differentiation (e.g. in Natalidae and Desmodontinae; see also Morgan & Czaplewski, 2003 and Suárez, 2005). Thereby, preparation of complete skeletons to increase samples in museum collections is highly recommended. With respect to robusticity and femur head morphology, the convergent occupation of the shape space suggests, considering some anatomical data discussed above, functional connections related to quadrupedalism (Schutt & Simmons, 2001). The grouping of taxa such as M. simus and Noctilio spp. with well-known good walkers (molossids, Mystacina, and Desmodontinae), however, may be related to additional functional demands (see next section).

Our scaling analyses confirmed the results from Swartz (1997) for femur length, supporting an adaptive trend of reduction in this trait. This author discussed the advantage of this reduction in the context of saving energy for the expensive activities of flight and echolocation, and Arévalo et al. (2018) have shown how these activities constrained the evolution of body size in bats. The positive allometry reported for femur width by Swartz (1997), however, runs against the body mass-reduction hypothesis, and the same can be said about our result for both width (isometry) and robusticity (positive allometry). Below we discuss how some possible functional demands (e.g. quadrupedalism, trawling behavior), may account for this pattern, constraining the overall reduction expected for the hindlimbs.

Functional predictions

When diagnosing the bat families and searching for general trends in shape space, a common theme revealed here was the grouping of not closely related taxa under morphologies with an apparent functional meaning (Gaudioso et al. 2017). This was well-illustrated by the proximity 66

between Desmodontinae, Mystacinidae, and Molossidae, from both categorical and quantitative (robusticity) data. These taxa can be functionally grouped based on their quadrupedal abilities, and their robusticity may be important to provide greater area for muscle insertion (Vaughan, 1959) and improve walking behavior (Riskin et al. 2005). Moreover, because several muscles related to the movements of the femur attach to the proximal epiphysis (m. psoas major, m. gluteus medius, m. adductor brevis, m. adductor magnus, m. obturator externus, and m. gemmelus [“m.” for “musculus”]), this bone part is particularly relevant to crawling ( aughan, 1959; Smith, 1972). Quadrupedal bats, as reported here, present well-developed trochanters, with a tubercle and a posterior ridge on the greater one, both providing a larger area for the attachment of muscles that assist in the forward walk (Vaughan, 1959; Grassé, 1967). These bats also have a straight femur (the proximal epiphysis is aligned with the shaft), which helps them maintain the sprawling posture during terrestrial walk (Vaughan, 1970).

The femur head morphology, here convergently allying Molossidae and Mystacina, but not Desmodontinae, is also noteworthy in respect to quadrupedalism. Previous studies have suggested that the head of the femur is directly related to the mobility of the hindlimbs (e.g. Vaughan, 1959, 1970; Smith, 1972; Schutt & Simmons, 2001). A more developed and rounded head, which is also shifted from the axis of the femur, may allow an improved articulation with the hip, and a more efficient movement during quadrupedal locomotion (Vaughan, 1970). Contrary to what could be expected, however, Desmodontinae did not stand out in relation to femur head variables. It appears that the head in vampires are latero-medially flattened (oval), which may be related to the unique habit of running seen in these bats (Riskin & Hermanson, 2005). Grassé (1967) noted that the head of the femur is flattened in some jumping mammals, such as Macropus (Diprotodontia) and Dipus (Rodentia). Bouma et al. (2013) proposed that the flattened head morphology seen in some mammals is important to “better withstand the high forces at the postero-superior head-neck junction” in runners and jumpers. If head measurements are indeed a good marker for quadrupedal ability, we predict that some vespertilionids positioned close to Molossidae and Mystacina in the morphospace defined in our ordination (Fig. 18) will prove to be good walkers.

Still in respect to the quadrupedal bats, another feature that deserves highlight is the position of the lateral ridges. The long (Desmodontinae), medially (Mystacina) and distally (Molossidae) displaced lateral ridges in quadrupedal bats may provide a better action of the m. gluteus maximus and m. caudofemoralis, helping to produce the extension at the hip joint and

to maintain the dorsoventral stability of the hindlimb during quadrupedal activity (Vaughan, 1959). Within the molossids, aughan (1959) noted that “relative to the total mass of the muscles of the thigh”, the m. gluteus maximus is larger in Eumops perotis (Schinz, 1821) than in the other species he studied (Macrotus californicus Baird, 1858 and Myotis velifer (Allen, 1890)). A remarkable displacement of the lateral ridge (located in the distal fifth of the shaft, near to the lateral condyle) is seen in C. torquatus, a molossid that also has a great femur robusticity (Fig. 25). These characteristics, together with the ones described by Schutt & Simmons (2001), suggest that this species is the most quadrupedal molossid.

An unexpected result from our data is the association between Noctilionidae and quadrupedal bats based on femur robusticity. Despite their crawling capacity observed in some studies (e.g. Bloedel, 1955; Blood, 1987), noctilionids are not known to be good walkers. For them, however, an alternative functional hypothesis worth of testing could involve demands related to trawling behavior. The two species in this family gaff their prey, fish or insects, from the water surface (Norberg & Rayner, 1987; Kalko et al. 1998; Aizpurua & Alberdi, 2018). In this case, more robust legs could favor the counteraction of resistance forces from water or even the management of the preys, which, after the capture, must be transferred from the feet to the mouth (Bloedel, 1955; Blood, 1987). If the functional link between trawling behavior and leg robusticity holds true, an interesting prediction that could be derived is that other bat groups foraging in a similar manner would present a similar morphological trend. This might be the case for Myotis simus, which appears in our data as an outlier among Vespertilionidae, presenting a robusticity arguably high even for a quadrupedal bat. Unfortunately, little is known about the foraging activities of this species (see Fenton & Bogdanowicz, 2002), beyond the fact that it is recognized as a water bat (i.e. forages over water; Myers & Wetzel, 1983; Findley, 1993). Another remarkable result was observed in Sturnira spp., which were highlighted in our sample by their high femur robusticity (and also anatomy—straight femur, greatly robust at the proximal end). These bats, however, are known to forage on the ground, where they feed on fallen fruits and occasionally are captured in rodent traps (Eisenberg & Redford, 1989; Muylaert et al. 2014). To these species, therefore, further ecomorphological investigations should focus on an improved crawling ability, not recorded in other stenodermatines studied so far (e.g. Jones & Hasiotis, 2018).

Considering our allometric results, quadrupedalism, in addition to other possible functional demands (e.g. trawling behavior), may account for the positive allometry found in

robusticity, constraining the overall reduction expected for the hindlimbs. The position of molossids and M. simus in respect to the regression line (Fig. 25) clearly points to this direction. Particularly in respect to quadrupedalism, although the lack of robusticity is not a limiting factor for the capacity of walking (Riskin et al. 2005), possible advantages of developed legs may be expressed in performance variables such as maneuverability, time spending foraging on the ground, and speed of movement (Vaughan, 1966; Riskin et al. 2005, 2006). Improved performance on more challenging tasks on the ground may require larger muscles, and more robust bones could provide surface area for their attachment (Vaughan, 1959; Grassé, 1967).

Conclusions

The categorical and quantitative features of the femur vary in Yangochiroptera, from taxa with a simpler morphology (e.g. Mormoopidae, Furipteridae, and Natalidae) to those with a more complex one (e.g. Molossidae, Mystacinidae, Desmodontinae). The proximal epiphysis is the most informative region of the femur and, from the quantitative assessment, robusticity and head morphology are the features that better explain variation among yangochiropterans. Anatomical characters allow diagnosing families, supporting their use in the taxonomy of this group. The study of allometry shows that the general size reduction trend on the hindlimbs expected for Chiroptera, as a result of the high energetic cost of flight, is not a pervasive pattern among yangochiropterans.

Acknowledgements

The authors are thankful to João Alves de Oliveira for helping during the development of this work and for providing access to specimens deposited in the mammal collection at Museu Nacional (Rio de Janeiro); to Leandro Salles for also allowing access to specimens from Museu Nacional and specimens on loan to him from the American Museum of Natural History to Adriano Lúcio Peracchi (Universidade Federal Rural do Rio de Janeiro) for donating three specimens now deposited at the Museu Nacional; to Leandro Rabello Monteiro for helping us with R codes and comparative methods; to Lucila Amador (Fundacion Miguel Lillo - CONICET) for sending the files of the Chiroptera phylogeny used in our comparative analyses; to Carlos Rodrigues de Moraes Neto (Museu Nacional) for helping in the preparation of some femora; to professors of the Department of Herpetology (Museu Nacional) for allowing us to use the photographic equipment; to Gilson Yack-Ximenes for helping us with the references; and to Iris Maria Deschaumes for reviewing the English. We are also indebted to Ely Amson 69

(Museum fur Naturkunde) and an anonymous reviewer for their many insightful comments and edits that greatly improved earlier versions of the manuscript.. This paper is a part of the D. Sc. requirements of Nathália Siqueira Veríssimo Louzada of the Programa de Pós-graduação em Biodiversidade e Biologia Evolutiva (PPGBBE), of the Universidade Federal do Rio de Janeiro. Financial support was provided by Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ) and Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES) – Finance Code 001. MRN is supported by the PNPD/CAPES fellowship program. LMP is supported by the CNPQ fellowship program (308505/2016-6). The authors declare no conflict of interest.

References

Aizpurua O, Alberdi A (2018) Ecology and evolutionary biology of fishing bats. Mammal Rev. 10.1111/mam.12136.

Allen H (1864) Monograph of the bats of North America. Smithsonian Misc Coll 7, 1–182.

Amador LI, Arévalo RLM, Almeida FC, Catalano SA, Gianinni NP (2016). Bat systematics in the light of unconstrained analyses of a comprehensive molecular supermatrix. J Mammal Evol 25, 37–70.

Arévalo RLM, Amador LI, Almeida FC, Giannini NP (2018) Evolution of body mass in bats: insights from a large supermatrix phylogeny. J Mamm Evol 1–16. 10.1007/s10914-018- 9447-8.

Baker RJ, Solari S, Cirranello A, Simmons NB (2016) Higher level classification of phyllostomid bats with a summary of DNA synapomorphies. Acta Chiropt 18, 1–38.

Bloedel P (1955) Hunting methods of fish-eating bats, particularly Noctilio leporinus. J Mammal 36, 390–399.

Blood BR (1987) Convergent hind limb morphology and the evolution of fish-catching in the bats Noctilio leporinus and Myotis (Pizonyx) vivesi (Mammalia: Chiroptera). Ph.D. thesis, pp. 186. University of Southern California, Los Angeles.

Bouma HW, De Boer SF, De Vos J, Van Kampen PM, Hogervorst T (2013) Mammal hip morphology and function: coxa recta and coxa rotunda. Anat Rec 296, 250–256.

Buffon GLL (1760) L’Histoire Naturelle, générale et particulière, avec la description du Cabinet du Roi, pp. 398. ome III. L’imprimerie Royale, aris.

Chambers JM, Cleveland WS, Tukey PA (1983) Graphical Methods for Data Analysis, pp. 395. Wadsworth International Group, Michigan.

Cheney JA, Ton D, Konow N, et al. (2014) Hindlimb motion during steady flight of the lesser dog-faced fruit bat, Cynopterus brachyotis. PloS one 9, e98093.

Cirranello A, Simmons NB, Solari S, Baker RJ (2016) Morphological diagnoses of higher- level phyllostomid taxa (Chiroptera: Phyllostomidae). Acta Chiropt 18, 39-71.

Cole FR, Wilson DE (2006) Leptonycteris yerbabuenae. Mamm Species 797, 1–7.

Coutinho LC, de Oliveira JA, Pessôa, LM (2013) Morphological variation in the appendicular skeleton of Atlantic forest sigmodontine rodents. J Morphol 274, 779–792.

Czaplewski NJ, Rincón AD, Morgan GS (2005) Fossil bat (Mammalia: Chiroptera) remains from Inciarte Tar Pit, Sierra de Perijá, Venezuela. Caribbean J Sci 41, 768–781.

Dávalos LM (2005) Molecular phylogeny of funnel-eared bats (Chiroptera: Natalidae), with notes on biogeography and conservation. Mol Phylogenet Evol 37, 91–103.

Dobson GE (1878) Catalogue of the Chiroptera in the collection of the British Museum, pp. 567. British Museum of Natural History, London.

Eisenberg JF, Redford KH (1989) Mammals of the Neotropics, Volume 2: The Southern Cone: Chile, Argentina, Uruguay, Paraguay, pp. 460. University of Chicago Press, Chicago.

Fenton MB, Bogdanowicz W (2002) Relationships between external morphology and foraging behavior: bats in the genus Myotis. Can J Zool 80, 1004–1013.

Fenton MB, Simmons NB (2014) Bats, a world of science and mystery, pp. 303. University of Chicago Press, Chicago.

Findley JS (1993) Bats: a community perspective, pp. 179. Cambridge University Press, Cambridge.

Flower WH, Gadow H (1885) An Introduction to the Osteology of the Mammalia, 3rd ed, pp. 406. Macmillan and CO, London.

Gardner AL (2008) Mammals of South America, volume 1: marsupials, xenarthrans, shrews, and bats, pp. 690. University of Chicago Press, Chicago.

Gaudioso PJ, Díaz MM, Barquez RM (2017) Morphology of the axial skeleton of seven bat genera (Chiroptera: Phyllostomidae). An Acad Bras Cienc 89, 2341–2358.

Giannini NP, Barquez RM (2003) Sturnira erythromos. Mamm Species 729, 1–5.

Grassé PP (1967) Traité de zoologie: Anatomie, systématique, biologie: Mammifères: Téguments et squelette, pp. 1162. Masson, Paris.

Gregorin R, Taddei VA (2002) Chave artificial para a identificação de molossídeos brasileiros (Mammalia, Chiroptera). Mastozool Neotrop 9, 13–32.

Gregorin R (2009) Phylogeny of Eumops (Chiroptera: Molossidae) using morphological data. Acta Chiropt 11, 247–258.

Gregorin R, Tahara AS, Buzzato DF (2011) Molossus aztecus and other small Molossus (Chiroptera: Molossidae) in Brazil. Acta Chiropt 13, 311–317.

Gunnell GF, Simmons NB (2005) Fossil evidence and the origin of bats. J Mamm Evol 12, 209–246.

Hand SJ, Weisbecker V, Beck RM, et al. (2009) Bats that walk: a new evolutionary hypothesis for the terrestrial behavior of New Zealand's endemic mystacinids. BMC Evol Biol 9, 169.

Howell DJ, Pylka J (1977) Why bats hang upside down: a biomechanical hypothesis. J Theor Biol 69, 625–631.

Hurtado-Miranda NE, Pacheco-Torres V (2014) Análisis filogenético del género Mimon Gray, 1847 (Mammalia, Chiroptera, Phyllostomidae) con la descripción de un nuevo género. Therya 5, 751–791.

Jones MF, Hasiotis ST (2018) Terrestrial behavior and trackway morphology of Neotropical bats. Acta Chiropt 20, 229–250.

Kalko EK, Schnitzler HU, Kaipf I, Grinnell AD (1998) Echolocation and foraging behavior of the lesser bulldog bat, Noctilio albiventris: preadaptations for piscivory? Behav Ecol Sociobiol 42, 305–319.

König HE, Liebich HG (2016) Anatomia dos Animais Domésticos: Texto e Atlas Colorido, 6th ed, pp. 824. Artmed Editora, Porto Alegre.

Kumirai A, Jones JK (1990) Nyctinomops femorosaccus. Mamm Species 349, 1–5.

Kunz TH, Anthony EL (1982) Age estimation and post-natal growth in the bat Myotis lucifugus. J Mammal 63, 23–32.

Lindsey LL, Ammerman LK (2016) Patterns of genetic diversification in a widely distributed species of bat, Molossus molossus. Occ Pap Texas Tech Univ 339, 1–16.

Loureiro LO (2014) Sistemática de Molossus (Mammalia: Chiroptera: Molossidae) com ênfase nas espécies ocorrendo no Brasil. MSc. dissert., 119 pp. Universidade Federal de Minas Gerais, Brasil.

Madan MA, Rayfield EJ, Bright JA (2017) Scaling and functional morphology in strigiform hind limbs. Sci Rep 7, 1–9.

Medina CE, Gregorin R, Zeballos H, Zamora HT, Moras LM (2014) A new species of Eumops (Chiroptera: Molossidae) from southwestern Peru. Zootaxa 3878, 19–36.

Monteiro LR (2013) Morphometrics and the comparative method: studying the evolution of biological shape. Hystrix 24, 25–32.

Moras LM, Tavares VC, Pepato AR, Santos FR, Gregorin R (2016) Reassessment of the evolutionary relationships within the dog-faced bats, genus Cynomops (Chiroptera: Molossidae). Zool Scr 45, 465–480.

Moratelli R., Peracchi AL, Dias D, de Oliveira JA (2011) Geographic variation in South American populations of Myotis nigricans (Schinz, 1821) (Chiroptera, Vespertilionidae), with the description of two new species. Mamm Biol 76, 592–607.

Moratelli R, Dias D (2015) A new species of nectar-feeding bat, genus Lonchophylla, from the Caatinga of Brazil (Chiroptera, Phyllostomidae). Zoo Keys 514, 73.

Morgan GS (1991) Neotropical Chiroptera from the Pliocene and Pleistocene of Florida. Bull Amer Mus Nat Hist 206, 176–213.

Morgan GS, Czaplewski NJ (2003) A new bat (Chiroptera: Natalidae) from the early Miocene of Florida, with comments on natalid phylogeny. J Mammal 84, 729–752.

Morgan GS, Czaplewski NJ (2012) Evolutionary history of the Neotropical Chiroptera: the fossil record. In: Evolutionary history of bats: fossils, molecules, and morphology (ed Gunnel GF, Simmons NB), pp. 105–161. Cambridge University Press, Cambridge.

Mundry R (2014) Statistical issues and assumptions of phylogenetic generalized least squares. In: Modern phylogenetic comparative methods and their application in evolutionary biology (ed Garamszegi LZ), pp. 131–153. Springer, Berlin, Heidelberg.

Muylaert RL, Matos DMS, Mello MAR (2014) Interindividual variations in fruit preferences of the yellow-shouldered bat Sturnira lilium (Chiroptera: Phyllostomidae) in a cafeteria experiment. Mammalia 78, 93–101.

Myers P, Wetzel RL (1983) Systematics and zoogeography of the bats of the Chaco Boreal. Misc Publ Mus Zool Univ Mich 165, 1–59.

Norberg UM, Rayner JM (1987) Ecological morphology and flight in bats (Mammalia; Chiroptera): wing adaptations, flight performance, foraging strategy and echolocation. Phil Trans R Soc Lond B 316, 335–427.

Orme D (2013) The caper package: comparative analysis of phylogenetics and evolution in R. R package version 5, 1–36.

Pavan AC, Marroig G (2016) Integrating multiple evidences in taxonomy: species diversity and phylogeny of mustached bats (Mormoopidae: Pteronotus). Mol Phylogenet Evol 103, 184– 198.

R Core Team (2018) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.

Reis NR, Peracchi AL, Batista CB, de Lima IP, Pereira AD (2017) História natural dos morcegos brasileiros: chave de identificação de espécies, pp. 480. Technical Books Editora, Rio de Janeiro.

Revell LJ (2009). Size-correction and principal components for interspecific comparative studies. Evolution 63, 3258–3268.

Revell LJ (2012) Phytools: an R package for phylogenetic comparative biology (and other things). Methods Ecol Evol 3, 217–223.

Rex K, Czaczkes BI, Michener R, Kunz TH, Voigt CC (2010) Specialization and omnivory in diverse mammalian assemblages. Ecoscience 17, 37–46.

Riskin DK, Hermanson JW (2005) Biomechanics: independent evolution of running in vampire bats. Nature 434, 292.

Riskin DK, Bertram JEA, Hermanson JW (2005) Testing the hindlimb-strength hypothesis: Non-aerial locomotion by Chiroptera is not constrained by the dimensions of the femur or tibia. J Exp Biol 208, 1309–1319.

Riskin DK, Parsons S, Schutt WA Jr., Carter GG, Hermanson JW (2006) Terrestrial locomotion of the New Zealand Short-tailed Bat Mystacina tuberculata and the Common Vampire Bat Desmodus rotundus. J Exp Biol 209, 1725–1736.

Riskin DK, Bertram JEA, Hermanson JW (2016) The evolution of terrestrial locomotion in bats: The bad, the ugly and the good. In: Understanding Mammalian Locomotion: Concepts and Applications. (ed Bertram JEA), pp. 307–323. John Wiley & Sons, Hoboken, New Jersey. 75

Rodríguez-Durán A, Kunz TH (1992) Pteronotus quadridens. Mamm Species 395, 1–4.

Rojas D, Warsi OM, Dávalos LM (2016) Bats (Chiroptera: Noctilionoidea) challenge a recent origin of extant neotropical diversity. Syst Biol 65, 432–448.

Salles LO, Arroyo-Cabrales J, Lima ACDM, et al. (2014) Quaternary bats from the Impossível-Ioiô cave system (Chapada Diamantina, Brazil): humeral remains and the first fossil record of Noctilio leporinus (Chiroptera, Noctilionidae) from South America. Am Mus Novit 3798, 1–32.

Sanborn CC (1949) Bats of the genus Micronycteris and its subgenera. Fieldiana Zool 31, 215–233.

Schutt WA Jr., Simmons NB (2001) Morphological specializations of Cheiromeles (naked bulldog bats; Family Molossidae) and their role in quadrupedal locomotion. Acta Chiropt 3, 225–236.

Sidlauskas B (2008) Continuous and arrested morphological diversification in sister clades of characiform fishes: a phylomorphospace approach. Evolution 62, 3135–3156.

Simmons NB (1994) The case for chiropteran monophyly. Am Mus Novit 3103, 1–54.

Simmons NB (1996) A new species of Micronycteris (Chiroptera, Phyllostomidae) from northeastern Brazil: with comments on phylogenetic relationships. Am Mus Novit 3158, 1–34.

Simmons NB (1998) A Reappraisal of Interfamilial Relationships of Bats. In: Bat Biology and Conservation (ed Kunz TH, Racey PA), pp. 3–26. Smithsonian Institution Press, Washington.

Simmons NB, Geisler JH (1998) Phylogenetic relationships of Icaronycteris, Archaeonycteris, Hassianycteris, and Palaeochiropteryx to extant bat lineages, with comments on the evolution of echolocation and foraging strategies in Microchiroptera. Bull Am Mus Nat Hist 235, 1–182.

Simmons NB, Voss RS (1998) The Mammals of Paracou, French Guiana: a Neotropical lowland rainforest fauna. Part 1. Bats. Bull Am Mus Nat Hist 237, 1–219.

Simmons NB, Conway TM (2001) Phylogenetic relationships of mormoopid bats (Chiroptera: Mormoopidae) based on morphological data. Bull Am Mus Nat Hist 258, 1–97.

Simmons NB, Seymour KL, Habersetzer J, Gunnell GF (2008) Primitive Early Eocene bat from Wyoming and the evolution of flight and echolocation. Nature 451, 818–822.

Smaers JB, Rohlf FJ (2016) esting species’ deviation from allometric predictions using the phylogenetic regression. Evolution 70, 1145–1149.

Smith JD (1972) Systematics of the chiropteran: family Mormoopidae. Misc Publ Univ Kans Mus Nat Hist 56, 1–132.

Stayton CT (2015) What does convergent evolution mean? The interpretation of convergence and its implications in the search for limits to evolution. Interface Focus 5, 20150039.

Suárez, W (2005) Taxonomic status of the cuban vampire bat (Chiroptera: Phyllostomidae: Desmodontinae: Desmodus). Caribbean J Sci 41, 761–767.

Swartz SM (1997) Allometric patterning in the limb skeleton of bats: implications for the mechanics and energetics of powered flight. J Morphol 234, 277–294.

Tejedor A (2011) Systematics of funnel-eared bats (Chiroptera: Natalidae). Bull Am Mus Nat Hist 353, 1–140.

Vaughan TA (1959) Functional morphology of three bats: Eumops, Myotis, Macrotus. Publ Mus Nat Hist Univ Kans 12, 1–153.

Vaughan TA (1966) Morphology and flight characteristics of molossid bats. J Mammal 47, 249–260.

Vaughan TA (1970) The skeletal system. In: Biology of bats. (ed Wimsatt WA), pp. 97–138. Academic Press, New York.

Velazco PM, Gardner AL, Patterson BD (2010) Systematics of the Platyrrhinus helleri species complex (Chiroptera: Phyllostomidae), with descriptions of two new species. Zool J Linn Soc 159, 785–812.

Velazco PM, O'Neill H, Gunnell GF, et al. (2013) Quaternary bat diversity in the Dominican Republic. Am Mus Novit 3779, 1–20.

Ziegler AC, Howarth FG, Simmons NB (2016) A second endemic land mammal for the Hawaiian Islands: a new genus and species of fossil bat (Chiroptera: Vespertilionidae). Am Mus Novit 3854, 1–52.

Walton DW, Walton GM (1970) Post-cranial osteology of bats. In: About Bats: a Chiropteran Symposium. (ed Slaughter BH, Walton DW), pp. 93–126. Southern Methodist University Press, Dallas.

Wible JR, Novacek MJ (1988) Cranial evidence for the monophyletic origin of bats. Am Mus Novit 2911, 1–19.

Appendix 1

Identification key to Neotropical families of Yangochiroptera (Chiroptera, Mammalia) based on the femur morphology

1. Greater trochanter projects further proximally than the lesser trochanter; greater trochanter with a hook-like process; medial ridge long and poorly developed (Figs 2A, 3D, 15) .……..………………………………………………………..…………..…..…….. Molossidae

Greater trochanter projects equal, more, or less proximally than the lesser trochanter; the greater trochanter is smooth or bear a poorly-developed tubercle; medial ridge short or long, moderately or well developed ……………………………………..……………………………...……...... 2

2. Narrow intercondylar fossa; straight femur …...... 3

Broad intercondylar fossa; straight or sinusoidal femur ……..……..…...….…..…...... 4

3. Lateral and medial condyles more elongated than broad; lateral ridge long and sometimes imperceptible (Figs 4F, 16A-D) ....…………...... …...... …… espertilionidae

Condyles nearly as elongated as broad; lateral ridge short and well developed (Fig. 8C,E) ……...………………………………………..………………..………....……… hyropteridae

4. Trochanters poorly developed, with little or no notch between them; sinusoidal femur ………………………….……………………………………………....………..……………. 5

Trochanters long and well developed, with remarkable notch between them; straight or sinusoidal femur ……………………………………………………..…………….…..…….... 7

5. Medial ridge long, beginning on the proximal region of the shaft and extending to its proximal half (Fig. 9A-C) ...... Mormoopidae

Medial ridge short ………………………………………………….………...…….…...……. 6

6. Elongated trochanters; medial ridge well developed; lateral ridge poorly developed, but perceptible ( ig. 9D) ……...... ……… Natalidae

Rounded trochanters; medial ridge moderately developed; lateral ridge poorly developed, almost imperceptible ( ig. 9 ) …..…...…………..…….…….…………..……….. uripteridae

7. Shaft conspicuously flattened dorsoventrally, with the anterior side almost straight (medial view; ig. 10D) …………....……..…....…………………..... Phyllostomidae (Desmodontinae)

Shaft cylindrical …………………………………..……………..……………….………....… 8

8. Proximal part of the shaft twisted sideways; medial ridge short and usually greatly pronounced (Figs 11–13) ………..………………………..………….…..……. hyllostomidae

Proximal part of the shaft aligned or slightly twisted; medial ridge moderately developed ………………………………..…………………………..…….………..……………….….... 9

9. Femur sturdy; lateral ridge usually short, poorly or well developed; lesser trochanter, which is more robust than the greater, projecting slightly more proximally than the greater; notch between the lesser trochanter and the medial ridge poorly developed (Fig. 8A-B) …………………………………………………..……..…………………....…… Noctilionidae

Femur slender; lateral ridge long and poorly developed; trochanters similar in size, with the lesser usually projecting further proximally than the greater; notch between the lesser trochanter and the medial ridge well developed (Fig. 14B-G) ………………… …..….… mballonuridae

Supporting information

Additional Supporting Information may be found in the online version of this article:

Table S1. Characterization of the femur morphology in 12 families of yangochiropteran bats based on 13 anatomical characters.

Appendix S1. Specimens examined.

Comparative morphology and scaling of the femur in yangochiropteran bats

Supporting Information

Nathália Siqueira Veríssimo Louzada, Marcelo Rodrigues Nogueira, and Leila Maria Pessôa

Material and Methods

Table S1 Characterization of the femur morphology in 12 families of yangochiropteran bats based on 13 anatomical characters.

Appendix S1 Specimens examined.

Table S1 Characterization of the femur morphology in 12 families of yangochiropteran bats based on 13 characters. Asterisks (*) indicate that the character varies among the genera. See the methodology for the description of the characters and their variations.

Family/Characters 1 2 3 4 5 6

Mystacinidae Moderately developed At the shaft Well developed Aligned or subtle Well developed Well developed difference (rounded process)

Thyropteridae Moderately developed At the shaft* Well developed Higher Absent Poorly developed

Furipteridae Moderately developed At the shaft Poorly developed Lower Absent Absent

Noctilionidae Moderately developed At the shaft* Well developed Lower Absent Poorly developed

Mormoopidae Moderately developed At the shaft Poorly developed Aligned or subtle Absent Absent difference

Phyllostomidae Well developed At the shaft* Well developed Lower* Absent* Absent*

Nycteridae Well developed At the shaft Well developed Aligned or subtle Absent Poorly developed difference

Emballonuridae Moderately developed At the shaft Well developed Lower* Absent Absent

Natalidae Well developed At the shaft Poorly developed Lower Absent Absent

Molossidae Poorly developed At the shaft Well developed Higher Well developed (hook- Well developed like process)

Miniopteridae Well developed Lesser Well developed Lower Absent Absent trochanter Lesser Aligned or subtle Vespertilionidae Moderately developed Well developed Poorly developed* Poorly developed* trochanter* difference*

Table S1 (cont.) Characterization of the femur morphology in 12 families of yangochiropteran bats based on 13 characters. Asterisks (*) indicate that the character varies among the genera. See the methodology for the description of the characters and their variations.

Family/Characters 7 8 9 10 11 12 13

Mystacinidae Deep Straight Aligned Poorly developed Well developed Proximal* Broad

Thyropteridae Shallow Straight Aligned Poorly developed Well developed Proximal Narrow

Furipteridae Poorly developed Sinusoidal Markedly twisted Poorly developed Poorly developed Proximal Broad

Noctilionidae Shallow Straight Aligned Poorly developed Well developed* Proximal Broad

Mormoopidae Poorly developed Sinusoidal Markedly twisted Poorly developed Poorly developed* Proximal Broad

Phyllostomidae Shallow Sinusoidal* Markedly twisted* Well developed* Poorly developed* Proximal Broad

Nycteridae Shallow Sinusoidal Markedly twisted Poorly developed Well developed Proximal Broad

Emballonuridae Shallow Straight* Aligned* Well developed Poorly developed Proximal Broad

Natalidae Poorly developed Sinusoidal Markedly twisted Poorly developed Poorly developed Proximal Broad

Molossidae Shallow Straight Aligned Poorly developed Well developed* Distal* Broad

Miniopteridae Deep Straight Slightly twisted Absent Poorly developed Proximal Narrow

Vespertilionidae Deep Straight Aligned Absent* Poorly developed Proximal Narrow

Appendix S1

Specimens examined (ALP: Coleção Adriano Lúcio Peracchi, Universidade Federal Rural do Rio de Janeiro, Rio de Janeiro, Brazil (Total: 8); AMNH: American Museum of Natural History, New York, USA (Total: 86); MN: Museu Nacional, Rio de Janeiro, Brazil (Total: 314); MZUSP: Museu de Zoologia da USP, São Paulo, Brazil (Total: 3). The sex of the specimens is specified after their museum number (♀: female, ♂: male, and SU: sex unknown).

Noctilionoidea

Mystacinidae (1 genus, 1 species): Mystacina tuberculata (AMNH 160269 ♀).

Thyropteridae (1 genus, 1 species): Thyroptera tricolor (ALP 2912 ♀, 2914 ♂, 2920 ♀; AMNH 266361 ♂, 267216 ♀, 267217 ♀, 268577 ♂).

Furipteridae (1 genus, 1 species): Furipterus horrens (MN 36053 ♂, 36057 ♀, 36287 ♀, 78115 ♀, 78116 ♂, 78117 ♀).

Noctilionidae (2 genera, 2 species): Noctilio albiventris (AMNH 210594 ♀; ALP 2880 SU; MN 64141 ♀, 64144 ♀, 64145 ♂, 64152 ♂), and Noctilio leporinus (AMNH 256528 ♀; MN 47199 ♂, 47204 ♀, 71319 ♂, 71321 ♀, 71577 ♀, 71580 ♂).

Mormoopidae (2 genera, 7 species): Mormoops blainvillei (AMNH 238144 ♂), Pteronotus davyi (AMNH 203565 ♀), Pteronotus gymnonotus (MN 68061 ♂, 68073 ♀, 68074 ♂, 68085 ♂), Pteronotus macleayi (AMNH 60917 ♂), Pteronotus parnellii (MN 36636 ♀, 37108 ♀, 37110 ♂, 80542 ♂, 80553 ♂, 80569 ♂, 80564 ♀, 80621 ♂, 80623 ♀), Pteronotus psilotis (AMNH 178468 SU), and Pteronotus quadridens (AMNH 39405 ♀).

Phyllostomidae

Macrotinae (1 genus, 1 species): Macrotus waterhousii (AMNH 236658 ♀).

Micronycterinae (1 genus, 3 species): Micronycteris megalotis (MN 36158 ♀, 36159 ♂), Micronycteris microtis (AMNH 266024 ♂; MN 80534 ♀, 80573 ♀), and Micronycteris sanborni (MN 75194 ♀, 79755 ♂, 80572 ♀).

Desmodontinae (3 genera, 3 species): Desmodus rotundus (MN 36573 ♂, 36574 ♀, 36586 ♂, 36587 ♀, 44877 SU, 56281 ♂, 67982 SU, 67985 SU, 67986 SU, 67988–67992 SU, 68025 SU, 68040 SU, 80589 ♀, 80590 ♂, 80612 ♀), Diaemus youngii (MN 71029 ♂, 71037 ♀, 71379 ♀, 77875 ♂, 79877 ♂), and Diphylla ecaudata (MN 47304 ♂, 68033 ♀, 68034 ♀, 68035 ♂, 68036 ♂, 77755 ♀, 77756 ♀; MZUSP 22676 SU).

Lonchorhininae (1 genus, 1 species): Lonchorhina aurita (MN 36251 ♂, 79798 ♂, 79802 ♂, 80540 ♂, 80541 ♀).

Phyllostominae (10 genera, 15 species): Chrotopterus auritus (MN 70862 ♂), Gardnerycteris crenulatum (MN 36684 ♂, 71390 ♀, 71391 ♀, 75189 ♂, 80583 ♂), Lophostoma brasiliense (AMNH 267917 ♂; MN 80495 ♀), Lophostoma carrikeri (MN 71404 ♀), Lophostoma schulzi (AMNH 267920 ♂), Lophostoma silvicola (MN 71351 ♂, 71486 ♀, 79730 ♂), Macrophyllum macrophyllum (MN 37202 ♂, 70599 ♀, 70600 ♀, 70601 ♀, 70662 ♂), Mimon bennettii (MN 79816 ♂, 79827 ♀, 79891 ♂, 80537 ♀), Phylloderma stenops (AMNH 267890 ♂; MN 70594 ♂, 70861 ♂), Phyllostomus discolor (AMNH 267121 ♀; MN 37329 ♀, 42733 ♂, 42734 ♀, 42737 ♂, 67723 SU, 67724 ♂, 67725 ♂), Phyllostomus elongatus (AMNH 267897 ♀; MN 70545 ♂, 70546 ♂; MZUSP 22570 SU), Phyllostomus hastatus (AMNH 99607 ♀, 209334 ♂, 230171 SU, 267901 ♀; MN 25708 ♂, 37330 ♂, 47235 ♂, 47236 ♀, 56238 ♀), Tonatia saurophila (AMNH 266045 ♀; MN 70248 ♂, 70860 ♂), Trachops cirrhosus (AMNH 267932 ♂; MN 71426 ♀, 71346 ♂, 71469 ♀, 80536 ♂), and Vampyrum spectrum (MN 46499 ♀).

Glossophaginae (7 genera, 9 species): Anoura caudifer (MN 43403 ♂, 43867 ♂, 43868 SU), Anoura geoffroyi (AMNH 230217 ♂; MN 37356 ♂, 56252 ♀), Brachyphylla cavernarum (AMNH 188238 ♂), Choeroniscus minor (AMNH 267153 ♀), Erophylla sezekorni (AMNH 176294 SU), Glossophaga commissarisi (AMNH 189600 ♀), Glossophaga soricina (AMNH 254618 ♂, 264585 ♂, 278304 ♂; MN 6499 SU, 26443 SU, 36109 ♂, 56284 ♂, 64784 ♂, 80597 ♂, 80600 ♂, 80606 ♀, 80611 ♀, 80619 ♂), Leptonycteris yerbabuenae (AMNH 189692 ♀), and Monophyllus redmani (AMNH 39523 ♂).

Lonchophyllinae (2 genera, 2 species): Hsunycteris thomasi (AMNH 266103 ♂), and Lonchophylla dekeyseri (MN 36846 ♀, 80497 ♀, 80525 ♀, 80563 ♀).

Carolliinae (1 genus, 4 species): Carollia brevicauda (ALP 7294 SU; AMNH 264612 ♀; MN 36933 ♀, 36941 ♀, 80587 ♀, 80604 ♀, 80608 ♀, 80613 ♂), Carollia castanea (AMNH 264943

♂), Carollia perspicillata (AMNH 264955 ♂; MN 36939 ♂, 40897 SU, 40898 SU, 80605 ♀), and Carollia sowelli (AMNH 278328 ♀).

Glyphonycterinae (1 genus, 1 species): Trinycteris nicefori (AMNH 267878 ♀).

Rhinophyllinae (1 genus, 2 species): Rhinophylla fischerae (MN 70252 ♂, 70254 ♂, 70255 ♂, 70323 ♂), and Rhinophylla pumilio (AMNH 266192 ♂; MN 46486 ♀, 47451 ♀, 70808 ♀, 70809 ♀, 70811 ♂, 70845 ♂).

Stenodermatinae (11 genera, 27 species): Ametrida centurio (AMNH 267276 ♀), Artibeus concolor (AMNH 267982 ♀), Artibeus fimbriatus (MN 43858 SU, 46341 ♂, 51700 ♂), Artibeus jamaicensis (AMNH 260244 ♂), Artibeus lituratus (AMNH 278314 ♂; MN 26442 ♂, 36682 ♂, 36783 ♂, 36825 ♀, 56175 ♀, 79762 ♂, 80609 ♂, 80614 ♂), Artibeus obscurus (AMNH 268640 ♂; MN 36480 ♂, 36525 ♀, 36841 ♂), Artibeus planirostris (AMNH 209607 ♂; MN 36207 ♀, 56176 ♂, 79736 ♂, 79738 ♀, 79765 ♂, 80616 ♂, 80622 ♀), Chiroderma salvini (AMNH 261668 ♀), Chiroderma trinitatum (AMNH 268531 ♀), Chiroderma villosum (MN 37213 ♀, 64518 ♀, 64519 ♂, 71370 ♀, 71372 ♀), Dermanura cinerea (AMNH 266320 ♂; MN 30624 ♀, 47053 ♂, 47057 SU, 79747 ♂), Dermanura gnoma (AMNH 267993 ♂; MZUSP 22544), Mesophylla macconnellii (AMNH 209577 ♀; MN 70887 ♂, 80526 ♂), Platyrrhinus brachycephalus (AMNH 210798 ♀), Platyrrhinus incarum (MN 80598 ♀), Platyrrhinus lineatus (MN 36492 ♀, 36697 ♂, 56213 ♀, 79728 ♂, 80593 ♂, 80596 ♀), Pygoderma bilabiatum (AMNH 248339 ♀; MN 81269 SU, 81275 SU), Sturnira erythromos (AMNH 268604 ♂), Sturnira ludovici (AMNH 254640 ♂), Sturnira oporaphilum (AMNH 264673 ♂), Sturnira lilium (AMNH 278326 ♀; MN 36013 ♀, 36189 ♂, 36192 ♂, 36313 ♀, 36314 ♀, 56251 ♀), Sturnira tildae (AMNH 266251 ♀; MN 36685 ♂), Uroderma bilobatum (AMNH 246609 ♂; MN 70877 ♀, 70881 ♂, 70888 ♀, 70918 ♀), Uroderma magnirostrum (AMNH 209426 ♂; MN 70302 ♂, 70655 ♀, 70885 ♀), Vampyressa brocki (AMNH 268567 ♀), Vampyressa thyone (AMNH 278333 ♂), and Vampyriscus bidens (MN 70908 ♂, 70909 ♀, 70910 ♂, 70921 ♂).

Emballonuroidea

Nycteridae (1 genus, 1 species): Nycteris macrotis (AMNH 187320).

Emballonuridae, Emballonurinae (4 genera, 7 species): Cormura brevirostris (AMNH 267827 ♂), Diclidurus isabella (MN 70449 ♀, 70644 ♀), Peropteryx kappleri (AMNH 265996 ♀; MN 43495 ♀), Peropteryx macrotis (AMNH 91237 ♀; MN 56150 ♀, 79767 ♀, 79768 ♂, 79797 ♀, 79809 ♂), Peropteryx trinitatis (MN 66312 ♀), Rhynchonycteris naso (AMNH 209212 ♀; MN 36310 ♂, 70428 ♂, 70429 ♂, 70431 ♀), Saccopteryx bilineata (AMNH 267064 ♂; MN 70445 ♀, 70611 ♂, 70914 ♀, 71594 ♂), Saccopteryx leptura (AMNH 210514 ♀), Taphozous georgianus (AMNH 197177 SU), and Taphozous melanopogon (AMNH 235571 ♂).

Vespertilionoidea

Natalidae (2 genera, 3 species): Natalus macrourus (MN 43104 ♀, 59843 SU, 67863 ♀, 67868 ♀, 67873 ♂, 81448 ♂), Natalus major (AMNH 238149 ♀), and Nyctiellus lepidus (AMNH 167130 SU).

Molossidae, Molossinae (11 genera, 20 species): Chaerophon plicatus (AMNH 27372 ♂), Cheiromeles torquatus (AMNH 247583 ♀), Cynomops greenhalli (MN 43856 ♀), Cynomops planirostris (MN 47350 SU, 46498 SU, 70276 ♀, 70287 ♂, 70288 ♀, 70301 ♂), Eumops glaucinus (AMNH 179948 ♂; MN 71472 ♀), Eumops perotis (MN 6646 ♂, 64370 ♂, 64747 ♂, 64750 ♂, 64787 ♀, 64789 ♀, 71287 ♀), Eumops trumbulli (MN 6486 ♂), Molossops temminckii (MN 36379 ♀, 71350 ♂, 71355 ♂, 71415 ♂, 71418 ♀, 71421 ♀, 71434 ♀), Molossus coibensis (MN 29028 ♂), Molossus currentium (MN 71560 ♂, 71563 ♀), Molossus molossus (MN 47147 ♀, 71561 ♂, 71562 ♂, 71564 ♀), Molossus pretiosus (MN 6672 ♀, 6597 ♂, 71551 ♀, 71552 ♀, 71566 ♂), Molossus rufus (MN 71568 ♀, 79894 ♂), Mops leucostigma (AMNH 170634 ♂, 170635 ♂, 170636 ♂, 170640 ♂), Neoplatymops mattogrossensis (MN 36029 ♂, 36030 ♀, 36031 ♀, 37323 ♀), Nyctinomops femorosaccus (AMNH 183172 SU), Nyctinomops laticaudatus (AMNH 209794 ♀; MN 49644 ♀, 69158 ♂), Nyctinomops macrotis (MN 49583 ♀, 49584 ♀, 49586 ♀, 49587 ♀, 49588 ♀, 49591 ♂, 49594 ♀), Promops nasutus (MN 64677 ♀, 64678 ♀, 64746 ♀, 64762 ♀, 64763 ♀, 71480 ♂), and Tadarida brasiliensis (AMNH 268660; MN 6562 ♀, 6564 ♀, 6565 ♀, 6566 ♀).

Miniopteridae (1 genus, 1 species): Miniopterus magnater (AMNH 235587 ♀).

Vespertilionidae

Vespertilioninae (3 genera, 5 species): Eptesicus brasiliensis (MN 71417 ♀), Histiotus montanus (AMNH 205649 ♀), Histiotus velatus (MN 3373 SU, 46450 ♂, 46491 SU), Lasiurus blossevillii (MN 71304 ♀, 71459 ♀), and Lasiurus ega (MN 70593 ♀).

Myotinae (1 genus, 6 species): Myotis albescens (ALP 6183 SU; AMNH 234361 ♀), Myotis levis (MN 80362 ♂), Myotis nigricans (ALP 6524 SU; MN 71530 ♀, 71532 ♀), Myotis riparius (ALP 2004 ♀; MN 71311 ♀, 71589 ♀), Myotis ruber (MN 80299 ♂), and Myotis simus (MN 71451 ♂, 71458 ♂).

4. Discussão

Para um grupo tão amplamente influenciado pela evolução do voo verdadeiro (Arita & Fenton, 1997; Fenton & Simmons, 2014), não é de surpreender que os membros posteriores tenham historicamente recebido menos atenção dos pesquisadores do que os membros anteriores. Ao estudar assimetria flutuante em Myotis lucifugus (Le Conte, 1831), Gummer & Brigham (1995) enfatizaram que os ossos posteriores deveriam ser funcionalmente menos relevantes para os morcegos do que os ossos das asas — e, portanto, mais sujeitos a assimetria — principalmente porque estes são usados para forragear — mais importante para o fitness — e aqueles para manter a postura de repouso (de cabeça para baixo) no abrigo. Esta avaliação pode ser justificada para a maioria dos morcegos, mas este é um grupo altamente diversificado (21 famílias, cerca de 1.400 spp.; Burgin et al., 2018; ver também Simmons & Conway, 2003), com muitas espécies pouco conhecidas; e padrões ecomorfológicos novos ou mais detalhados podem emergir quando análises comparativas em amplas escalas taxonômicas (e ecológicas) são feitas. Os resultados aqui apresentados, com base em dados quantitativos e categóricos, mostram que o fêmur é um osso muito informativo, tanto do ponto de vista taxonômico, permitindo a caracterização e identificação das famílias de Yangochiroptera, quanto do ponto de vista funcional, ajudando na interpretação de convergências e permitindo o desenvolvimento de predições funcionais.

4.1 Morfologia do fêmur

Nossos resultados apontaram três direções principais no espaço morfológico quantitativo do fêmur. O tamanho é a principal fonte de variação, mas a grande sobreposição entre as famílias torna difícil o uso de dados mensurais em avaliações taxonômicas a este nível. Para alguns gêneros e espécies, no entanto, parece haver um potencial para que essas medidas sejam usadas na diferenciação taxonômica (por exemplo, em Natalidae e Desmodontinae; ver também Morgan & Czaplewski, 2003 e Suárez, 2005). As medidas de largura e da cabeça do fêmur são as que se mostraram mais informativas para separação das famílias, e também foram importantes para o desenvolvimento das predições funcionais. O estudo detalhado da robustez foi importante porque permitiu a observação de que, a despeito da redução do tamanho dos membros posteriores esperada como adaptação ao voo (diminuição do peso corporal e menor gasto energético; Howell & Pylka, 1977), muitas linhagens de Chiroptera apresentam um fêmur robusto, o que também pode estar associado a aspectos funcionais, que teriam restringido a redução do fêmur nessas linhagens (Riskin et al., 2005). Foi observado ainda que, além das linhagens quadrúpedes, previamente notadas por terem um fêmur mais largo (ex.

Desmodontinae; Smith, 1972), outras linhagens apareceram próximas, com altos valores de robustez (ex. Noctilionidae), o que nos permitiu inferir que além do quadrupedalismo, outras demandas associadas ao uso das pernas, podem requerer fêmures robustos.

A caracterização anatômica se mostrou útil para separação das famílias, e nos permitiu a elaboração de uma chave de identificação que, potencialmente, pode ser aplicada na identificação de peças fósseis e auxiliar no estudo paleofaunístico de Chiroptera (veja Czaplewski et al., 2005). O fêmur também se mostrou informativo na separação de algumas subfamílias de Phyllostomidae (ex. Desmodontinae e Stenodermatinae) e até na caracterização de gêneros ou espécies (ex. Sturnira, Trinycteris nicefori Sanborn, 1949 e Gardnerycteris crenulatum (É. Geoffroy, 1810)). Esse potencial informativo também pode ser explorado do ponto de vista funcional. Diversos músculos associados à movimentação dos membros posteriores se inserem no fêmur, sendo os trocânteres e as cristas (medial e lateral) importantes áreas de inserção (Vaughan, 1959). O desenvolvimento e o posicionamento dessas estruturas, então, podem influenciar na ação muscular durante as atividades que exigem o uso das pernas (ex. quadrupedalismo, captura de presas, movimentação do uropatágio) e as convergências observadas em algumas linhagens, merecem interpretações funcionais.

4.2 Alometria

De forma geral, acreditava-se que as pernas de morcegos eram finas e delicadas, representando uma importante adaptação para o voo, pois permitiria a redução do peso corporal e consequentemente do gasto energético (Swartz, 1997). Essas expectativas de redução do tamanho foram correspondidas com o comprimento do fêmur, que apresentou alometria negativa em relação ao tamanho corporal. Por outro lado, os resultados para largura (isometria) e robustez (alometria positiva), contrariam a hipótese de redução do peso, indicando que outras demandas funcionais, além do voo, possam ter atuado na evolução dos membros posteriores. Howell & Pylka (1977) destacaram que os membros posteriores poderiam ser modificados, mas de forma limitada. Não poderiam, por exemplo, ser drasticamente encurtados, pois formam suporte necessário para a inserção e controle do uropatágio, membrana importante para decolagem e estabilidade do voo, e frequentemente usada como uma cesta para captura de alimentos (Adams et al., 2012; Howell & Pylka, 1977; Vaughan, 1970; Webster & Griffin, 1962). Nesse sentido, as pernas longas servem à função principal de manter a membrana firme e controlar sua forma, através da extensão ou retração, e servem como a origem dos músculos

filiformes que efetuam ajustes mais sutis (Howell & Pylka, 1977). Uma alternativa para redução do peso, no entanto, seria ter pernas mais finas (Howell & Pylka, 1977), o que aparentemente ocorreu em algumas linhagens aqui estudadas (ex. Natalidae, Furipteridae). Apesar disso, a alometria positiva observada para robustez aponta para a importância funcional de membros robustos em algumas linhagens, como os quadrúpedes ágeis de Molossidae, que aqui se destacaram com fêmures muito mais robustos do que esperado para o tamanho. Nesse caso, apesar das pernas não serem finas, parece haver uma redução no tamanho (fêmures mais curtos), o que poderia compensar o aumento do peso dado pela robustez. Dessa forma, a redução do tamanho do fêmur não prejudicaria a inserção do uropatágio, que nas espécies quadrúpedes é reduzido (Desmodontinae) ou controlado (estendido/retraído) pela cauda e calcâneo (Molossidae e Mystacina; Happold & Happold, 2013; Riskin et al., 2016), nem prejudicaria o desempenho do voo, que pode ser obtido através de características da asa (Gardiner et al., 2011). Sendo assim, além de quadrúpedes ágeis, esses morcegos também apresentam um voo eficiente (Riskin et al., 2016), mostrando que a mudança da forma do fêmur (e outras características morfológicas) não prejudicou a ocupação de nenhum dos dois nichos, pelo contrário, se adequou ao uso dos dois. Apesar da robustez não ser essencial para o quadrupedalismo, tendo em vista que mesmo espécies com fêmur mais fino conseguem caminhar — ineficientemente — no chão, ela pode ser importante para o desempenho durante esse tipo de locomoção, seja para percorrer maiores distâncias, atingir maiores velocidades, permitir uma maior capacidade de manobras, ou até mesmo para ficar mais tempo forrageando nessa postura (Riskin et al., 2005).

4.3 Predições funcionais

“Onde os parâmetros de um organismo diferem de tais modelos, como as dimensões das pernas de morcegos, podemos procurar por adaptações ecológicas” (Howell & ylka, 1977). As características mais marcantes do fêmur de morcegos, que se diferenciaram do padrão geral esperado (simples, fino e delicado), foram relacionadas à robustez, e a características da epífise proximal e da haste do fêmur. Nossos resultados mostraram que, apesar de terem evoluído de forma independente, os morcegos quadrúpedes (Desmodontinae, Molossidae e Mystacina) compartilham algumas características, dentre elas, um fêmur robusto e reto, com a epífise proximal alinhada ao eixo principal; trocânteres bem desenvolvidos, com o maior comportando uma crista e um tubérculo; cristas laterais localizadas ou se estendendo mais distalmente na haste do fêmur. Essas características proporcionam uma área mais ampla para inserção de

músculos mais volumosos, potencializando então a locomoção quadrúpede (Vaughan, 1959). Apesar das características compartilhadas, diferenças morfológicas também foram observadas em níveis interespecíficos. A variação da robustez entre as três espécies de Desmodontinae, por exemplo, pode estar relacionada a diferenças de estratégias alimentares adotadas por elas (Schutt & Simmons, 2006). Diphylla ecaudata é a linhagem mais basal de Desmodontinae (Amador et al., 2016) e é considerada uma caçadora exclusivamente arbórea, não havendo registro de locomoção terrestre durante a alimentação (Schutt & Simmons, 2006). Diferentemente de Desmodus rotundus e Diaemus youngii, D. ecaudata possui um calcar digitiforme que funciona como um sexto dígito, permitindo que esses morcegos se agarrarem em galhos, o que reflete uma especialização adicional em seu modo de alimentação único (sangue de aves). Além disso, os membros posteriores são mais robustos em Desmodus e Diaemus (Schutt, 1998), o que também foi observado para o fêmur no presente estudo. Sendo assim, uma hipótese que vale ser pensada é que a robustez dos membros posteriores tenha surgido no ancestral de Desmodus + Diaemus, permitindo a eles uma especialização maior no quadrupedalismo e no uso de ambientes terrestres para alimentação (Schutt, 1998; Schutt & Simmons, 2006). Uma outra evidência nesse sentido seria a presença da fíbula, osso associado a locomoção quadrúpede, nessas linhagens — incompleta em Diphylla, mas completa em Desmodus e Diaemus (Schutt, 1998).

Altos valores de robustez também foram observados em Noctilionidae, uma linhagem de morcegos que usam as pernas ativamente para o arraste na superfície de rios/lagos para captura de insetos e peixes (Aizpurua & Alberdi, 2018; Kalko et al., 1998; Norberg & Rayner, 1987), e que também possuem fêmures retos e trocânteres desenvolvidos. Um dos fêmures mais robustos em relação ao tamanho corporal é o de Myotis simus Thomas, 1901 (Vespertilionidae), uma espécie de morcego insetívoro aéreo com poucos dados ecológicos disponíveis, sendo apenas conhecido como um morcego que forrageia sobre a água (veja Fenton & Bogdanowicz, 2002). Se a robustez for realmente um bom marcador do uso das pernas, como temos visto com outras linhagens, é provável que essa espécie esteja forrageando de forma semelhante aos Noctilio spp. e a robustez, em ambos os casos, seria importante para suportar as forças de resistência da água durante o arraste (Bloedel, 1955; Blood, 1987). O uso das pernas para controle do uropatágio ou para a captura e posterior transferência da presa para a boca durante o voo também são características comuns em algumas espécies de Myotis (Adams et al., 2012; Fenton & Bogdanowicz, 2002) e, estudos mais aprofundados, comparando guildas ecológicas, podem ser úteis. Dessa forma, nosso estudo demonstrou que não só as linhagens quadrúpedes 93

possuem uma morfologia diferenciada no fêmur, como também linhagens que não evoluíram para um quadrupedalismo especializado, mas usam as pernas para captura de presas ou movimentação do uropatágio, comportamentos fundamentais para sua sobrevivência.

4.4 Perspectivas evolutivas no estudo do fêmur

Reconstruir o comportamento de animais extintos é notoriamente difícil (Fenton & Simmons, 2014), porém a comparação da morfologia de espécies recentes com espécies extintas, representadas por fósseis bem preservados, tem se mostrado um bom caminho. Algumas características da cabeça e do pescoço relacionadas à audição e à vocalização nos morcegos viventes (ex. tamanho da cóclea, conexão da orelha com a garganta via o osso stylohyal) são associadas à ecolocalização (Fenton & Simmons, 2014). A percepção desses caracteres em espécies fósseis pode funcionar como evidência clara de que eles eram ecolocalizadores (Fenton & Simmons, 2014). A maioria dos fósseis de morcegos, entretanto, não está bem preservada e, muitas vezes, o que temos num sítio fossilífero são partes do esqueleto (ex. crânio, mandíbula, fêmur, úmero), que eventualmente estão quebradas (Czaplewski et al., 2005; Hand et al., 2009; Velazco et al., 2013). O estudo morfológico detalhado desses ossos, consequentemente, seria relevante para identificação do táxon e posterior interpretação de aspectos funcionais. Metodologias como a de reconstrução ancestral também podem ser úteis para inferir as prováveis características de táxons fósseis pouco conhecidos, o que também permite a inferência de seus hábitos (Fenton & Simmons, 2014).

A maioria dos autores concordam que os morcegos evoluíram o voo a partir de um ancestral arbóreo que, por sua vez, evoluiu a torção das pernas e do hábito de se pendurar, antes do voo (Riskin et al., 2016; Schnitzler et al., 2003; Simmons & Geilser, 1998; Simmons et al., 2008). Um dos fósseis mais antigos de Chiroptera (ca. 50 m.a.), Onychonycteris finneyi Simmons et al. (2008), possui características únicas que condizem com essa hipótese evolutiva: proporções dos membros diferenciadas de todos os outros morcegos (alongamento pouco pronunciado do rádio e membros posteriores proporcionalmente mais longos); fíbula robusta; e retenção de garras na asa. Essas características sugerem que O. finneyi provavelmente era capaz de uma locomoção arbórea mais ágil que outros morcegos, incluindo táxons do Eoceno, e pode ter incorporado o quadrupedalismo e a suspensão embaixo de galhos em seu comportamento de locomoção e de abrigamento (Simmons et al., 2008). Nesse sentido, uma descrição mais detalhada do fêmur de O. finneyi, aliada aos dados disponibilizados no presente

estudo, pode ser importante para ajudar no entendimento das estratégias adotadas por essa espécie no uso do abrigo e durante o forrageamento. Estudos detalhados do fêmur em espécies fósseis, entretanto, são escassos. Simmons & Geisler (1998) mostraram que algumas características do fêmur de gêneros fósseis (Archaeonycteris, Icaronycteris, Hassianycteris e Palaeochiropteryx) são compartilhadas com as linhagens recentes e que a haste reta seria uma condição primitiva em Chiroptera. O mesmo é observado para variáveis quantitativas, onde os dados mais informativos estão no estudo de Simmons et al. (2008), que disponibilizou o comprimento do fêmur de oito espécies de morcegos extintos.

De acordo com Riskin et al. (2006), “em casos onde a morfologia animal simultaneamente atende aos requisitos de mais de uma forma de movimento, estudos de forma e função assumem outra dimensão de complexidade”. Acredita-se que ao mesmo tempo em que os morcegos se tornaram mais adaptados para voar, perderam a habilidade de correr devido às inúmeras modificações sofridas no esqueleto apendicular (Howell & Pylka, 1977; Vaughan, 1959). De fato, a maioria das linhagens recentes de Yangochiroptera (ex. Furipteridae, Natalidae, Emballonuridae, Mormoopidae) possui habilidade quadrúpede ausente ou limitada, o que, entre outros fatores, está relacionado ao posicionamento dos membros posteriores (Riskin et al., 2016; Schutt & Simmons, 2006; Vaughan, 1959). A adaptação a nichos e estratégias diferenciadas, entretanto, teria permitido o reaparecimento do quadrupedalismo em algumas linhagens, como observado em Desmodontinae, Molossidae, e Mystacinidae, que são quadrúpedes ágeis (Riskin et al., 2006, 2016). No caso de Desmodus rotundus (É. Geoffroy, 1810) sugere-se que o hábito terrestre confere benefício energético e vantagem seletiva na busca e captura de presas, que pode ocorrer furtivamente e, ao mesmo tempo, evita injúria ou predação no chão (Riskin et al., 2006). Em relação a Mystacina, entretanto, acreditava-se que a ausência de predadores ou competidores na Nova Zelândia teria permitido a exploração do nicho terrestre por esses pequenos morcegos, mas uma nova hipótese surgiu a partir da comparação da morfologia umeral de uma espécie extinta do Mioceno, Icarops aenae, com a vivente Mystacina tuberculata Gray, 1843, que mostrou que ela já era adaptada à locomoção terrestre, apesar da existência de muitos mamíferos predadores (Hand et al., 2009). Sendo assim, a vantagem seletiva (ex. exploração de uma maior área de recursos alimentares; Riskin et al., 2016) seria a força evolutiva por trás da habilidade terrestre observada nesses morcegos (Hand et al., 2009), tendo o úmero contribuído de forma excepcional para o entendimento da evolução do hábito no grupo.

Com respeito aos Molossidae, as circunstâncias nas quais o quadrupedalismo é empregado — em locais estreitos, como de frestas de rochas — sugerem cenários distintos para evolução desse hábito (Schutt & Simmons, 2006). Segundo Nolte et al. (2009), a habilidade de caminhar eficientemente nos abrigos seria vantajosa para escapar de predadores. Molossídeos possuem asas estreitas e longas que lhes conferem a habilidade de voar rápido e cobrir longas distâncias, mas limitam sua capacidade de alçar voo de uma superfície plana, fazendo com que eles necessitem de um momento de queda livre para iniciar o voo (Nolte et al., 2009; Norberg & Rayner, 1987; Schutt & Simmons, 2006; Vaughan, 1959). Dessa forma, uma vez no chão, esses morcegos se tornariam alvos fáceis de predadores, caso não conseguissem “correr” com eficiência; o quadrupedalismo aqui seria fundamental para a sobrevivência deles, permitindo que escapem depois de uma queda acidental (Riskin et al., 2016). Um segundo cenário para evolução do quadrupedalismo em Molossidae envolveria vantagem seletiva na competição por abrigos (Vaughan, 1959). Em paralelo às modificações do esqueleto axial observadas nos molossídeos (achatamento do crânio; esterno longo e com quilha reduzida; alongamento da escápula; Vaughan, 1959, 1966), poderia-se esperar modificações no esqueleto apendicular que potencializariam o hábito quadrúpede, permitindo uma maior mobilidade e acesso a esses ambientes estreitos (Altringham, 2011). Para morcegos, assim como para diversos outros grupos de animais, abrigos podem representar um recurso limitado (Voss et al., 2016) e características que lhes permitam empregar de maneira mais eficiente esses recursos podem ser alvos de seleção.

Os dois cenários acima apresentados podem ser testados através da comparação de uma variável preditiva do quadrupedalismo com variáveis preditivas da capacidade de alçar voo do chão e do uso de frestas. A robustez do fêmur, que no presente estudo demonstrou forte relação com as espécies quadrúpedes, pode ser uma boa variável preditiva desse hábito. Alguns índíces alares, como a carga alar e a taxa de aspecto, por outro lado, podem funcionar como variáveis preditivas da capacidade de alçar voo do chão — quanto maiores esses índices, mais difícil é para o morcego alçar o voo (Norberg & Rayner, 1987). Por sua vez, o uso de frestas pode ser estimado através de caracteres previamente considerados como importantes para evolução desse hábito, como o achatamento craniano e o alongamento do esterno — quanto mais achatado o crânio e mais alongado o esterno, mais especializada ao uso frestas estreitas a espécie está (Vaughan, 1959). Dessa forma, um estudo comparado com uma grande amostra de espécies de molossídeos pode ser útil no entendimento dessas relações — morcegos com alta carga alar e taxa de aspecto teriam uma dificuldade maior de alçar o voo, como observado em 96

Molossus rufus (Nolte et al., 2009); nesse caso, a maior eficiência na locomoção quadrúpede (fêmures mais robustos) seria importante para evitar a predação desses indivíduos. Da mesma forma, morcegos mais especializados ao uso de frestas, com achatamento craniano muito conspícuo, como visto em Neoplatymops mattogrossensis (Vieira, 1942), Platymops setiger Peters, 1878 e Sauromys petrophilus (Roberts, 1917) (Gregorin & Cirranello, 2016), tenderiam a apresentar fêmures mais especializados, o que permitiria o uso eficiente de abrigos mais estreitos. Uma análise à priori, entretanto, seria avaliar se de fato a robustez estaria relacionada a uma maior eficiência durante a locomoção quadrúpede, o que pode ser testado através de estudos biomecânicos, semelhantes aos realizados por Riskin et al. (2005, 2006), mas comparando uma maior amostra de espécies, incluindo molossídeos.

Associado à análise quantitativa, os dados categóricos aqui levantados também podem ser úteis no entendimento da variação da capacidade quadrúpede nos molossídeos. Apesar de compartilharem diversas características (ex. tubérculo em forma de gancho, crista posterior do trocânter maior desenvolvida), os molossídeos apresentam variação em relação ao posicionamento da crista lateral, o que pode ser importante para o desempenho durante a caminhada. Nessa região estão inseridos dois músculos importantes da perna, o m. gluteus maximus e o m. tensor faciae latae, responsáveis pela abdução e flexão do fêmur (Vaughan, 1959); as diferenças encontradas entre os gêneros estudados podem então ter relação com uma habilidade quadrúpede mais ou menos eficiente, o que merece considerações mais detalhadas.

Outra associação funcional que deve ser considerada é em relação ao voo. Os molossídeos são únicos por terem desenvolvido capacidade de voo rápido e duradouro (Norberg & Rayner, 1987), ao mesmo tempo em que apresentam um quadrupedalismo eficiente (Vaughan, 1959, 1966). Consequentemente, eles oferecem bons exemplos de um único caráter morfológico servindo a diversos aspectos funcionais (Vaughan, 1966). A flexão posterior das primeiras falanges dos dígitos três e quatro, por exemplo, permite que a parte distal da asa se dobre em um pacote que facilita a ação lateral dos membros anteriores durante a locomoção quadrúpede, o que permite que os molossídeos se locomovam rápida e facilmente dentro das frestas estreitas; da mesma forma, essa estratégia auxilia nos momentos de mergulho durante o voo (Vaughan, 1959; 1966). Do ponto de vista femoral, a robustez do fêmur e a presença de cristas ou tubérculos desenvolvidos facilitam a inserção dos fortes músculos adutores e flexores do quadril e também podem representar características com um propósito duplo. Além de importante para o quadrupedalismo, como discutido anteriormente, o fêmur pode atuar como

uma âncora rígida para a parte posterior do plagiopatagium, o que é importante para suportar a tração lateral e dorsal exercida pela membrana durante a descida da asa no voo (Vaughan, 1966).

O presente estudo destaca o grande potencial que a morfologia do fêmur, em uma abordagem comparada, pode apresentar. Diversos táxons de Yangochiroptera merecem destaque e são válidos de serem intimamente estudados, mas ressaltamos principalmente o potencial morfofuncional presente na família Molossidae. Por tratar-se de um grupo muito diversificado (16 gêneros, 110 espécies; Amador et al., 2016), molossídeos tem grande potencial para investigação de cenários acerca da evolução do hábito quadrúpede, principalmente quando se considera que há neles evidências de forte variabilidade em caracteres funcionais ligados a esse hábito (Schutt & Simmons, 2001, 2006; Vaughan, 1959, 1966, 1970).

5. Conclusões

A anatomia do fêmur se mostrou muito informativa, permitindo a caracterização das famílias de Yangochiroptera e a elaboração de uma chave de identificação para os táxons neotropicais. As dimensões do fêmur se mostraram importantes na caracterização das espécies no espaço morfológico quantitativo, sendo o diâmetro e a cabeça do fêmur as variáveis que melhor permitiram a visualização de grupos e a identificação de convergências na ocupação do morfoespaço. O estudo da alometria mostrou que a tendência de redução dos membros posteriores esperada para os morcegos só é observada em relação ao comprimento do fêmur, indicando que outros aspectos funcionais, além do voo, podem ter atuado na evolução da forma desse osso. Em conjunto, os resultados aqui apresentados permitiram a caracterização de táxons, a identificação de convergências e a interpretação de aspectos funcionais na morfologia do fêmur, ressaltando a importância do estudo de ossos apendiculares para o entendimento evolutivo, ecológico e taxonômico do grupo.

100

6. Referências bibliográficas

101

Adams RA, Snode ER, Shaw JB (2012) Flapping tail membrane in bats produces potentially important thrust during horizontal takeoffs and very slow flight. PLoS One 7, e32074.

Aizpurua O, Alberdi A (2018) Ecology and evolutionary biology of fishing bats. Mammal Review 10.1111/mam.12136.

Allen HMD (1893) A monograph of the bats of North America. In: Smithsonian Miscellaneous Collections, pp. 110. Smithsonian Institution, Washington.

Altenbach JS (1979) Locomotor morphology of the vampire bat, Desmodus rotundus. Special Publications of American Society of Mammologists 6, 1–137.

Altringham JD (2011) Bats: from evolution to conservation. Oxford University Press, New York.

Amador LI, Arévalo RLM, Almeida FC, Catalano SA, Gianinni NP (2016). Bat systematics in the light of unconstrained analyses of a comprehensive molecular supermatrix. Journal of Mammalian Evolution 25, 37–70.

Anderson ME, Racey PA (1991) Feeding behavior of captive brown long-eared bats, Plecotus auritus. The British Journal of Animal Behaviour 42, 489–493.

Arita HT, Fenton MB (1997) Flight and echolocation in the ecology and evolution of bats. Trends in Ecology & Evolution 12, 53–58.

Bloedel P (1955) Hunting methods of fish-eating bats, particularly Noctilio leporinus. Journal of Mammalogy 36, 390–399.

Buffon GLL (1760) L’Histoire Naturelle, générale et particulière, avec la description du Cabinet du Roi, pp. 398. ome III. L’imprimerie Royale, aris.

Burgin CJ, Colella JP, Kahn PL, Upham NS (2018) How many species of mammals are there? Journal of Mammalogy 99, 1–11.

102

Christiansen P (2002) Mass allometry of the appendicular skeleton in terrestrial mammals. Journal of Morphology 251, 195–209.

Czaplewski NJ, Rincón AD, Morgan GS (2005) Fossil bat (Mammalia: Chiroptera) remains from Inciarte Tar Pit, Sierra de Perijá, Venezuela. Caribbean Journal of Science 41, 768–781.

Denzinger A, Kalko EK, Tschapka M, Grinnell AD, Schnitzler HU (2016) Guild structure and niche differentiation in echolocating bats. In: Bat bioacoustics, pp. 141–166. Springer, New York.

Dobson GE (1878) Catalogue of the Chiroptera in the collection of the British Museum, pp. 567. British Museum of Natural History, London.

Feldhamer GA, Drickamer LC, Vessey SH, Merritt JF, Krajewski C (2007) Mammalogy: adaptation, diversity, ecology. Johns Hopkins University Press, Baltimore.

Fenton MB, Bogdanowicz W (2002) Relationships between external morphology and foraging behavior: bats in the genus Myotis. Canadian Journal of Zoology 80, 1004–1013.

Fenton MB, Simmons NB (2014) Bats, a world of science and mystery, pp. 303. University of Chicago Press, Chicago.

Flower WH, Gadow H (1885) An Introduction to the Osteology of the Mammalia, 3rd ed, pp. 406. Macmillan and CO, London.

Gardiner JD, Codd JR, Nudds RL (2011) An association between ear and tail morphologies of bats and their foraging style. Canadian Journal of Zoology 89, 90–99.

Gaudioso PJ, Díaz MM, Barquez RM (2017) Morphology of the axial skeleton of seven bat genera (Chiroptera: Phyllostomidae). Anais da Academia Brasileira de Ciências 89, 2341– 2358.

Gummer DL, Brigham RM (1995) Does fluctuating asymmetry reflect the importance of traits in little brown bats (Myotis lucifugus)? Canadian Journal of Zoology 73, 990–992.

Gunnell GF, Simmons NB (2005) Fossil evidence and the origin of bats. Journal of Mammalian Evolution 12, 209–246.

103

Hand SJ, Weisbecker V, Beck RM, et al. (2009) Bats that walk: a new evolutionary hypothesis for the terrestrial behavior of New Zealand's endemic mystacinids. BMC Evolutionary Biology 9, 169.

Happold M, Happold D (2013) Mammals of Africa Volume IV - Hedgehogs, Shrews and Bats, pp. 800. Bloomsbury, London.

Howell DJ, Pylka J (1977) Why bats hang upside down: a biomechanical hypothesis. Journal of Theoretical Biology 69, 625–631.

Kalko EK, Schnitzler HU, Kaipf I, Grinnell AD (1998) Echolocation and foraging behavior of the lesser bulldog bat, Noctilio albiventris: preadaptations for piscivory? Behavioral Ecology and Sociobiology 42, 305–319.

Koopman KF (1994) Chiroptera: systematics. In: Handbook of Zoology, vol. 8, Mammalia (ed. Niethammer J, Schliemann H, Starck D), pt. 60. Walter de Gruyter, Berlim.

Lawrence MJ (1969) Some observations on non‐volant locomotion in vespertilionid bats. Journal of Zoology 157, 309–317.

Madan MA, Rayfield EJ, Bright JA (2017) Scaling and functional morphology in strigiform hind limbs. Scientific Reports 7, 1–9.

Morgan GS, Czaplewski NJ (2003) A new bat (Chiroptera: Natalidae) from the early Miocene of Florida, with comments on natalid phylogeny. Journal of Mammalogy 84, 729–752.

Neuweiler G (2000) The biology of bats, pp. 310. Oxford University Press, New York.

Nolte MJ, Hockman D, Cretekos CJ, Behringer RR, Rasweiler IV JJ (2009) Embryonic staging system for the black mastiff bat, Molossus rufus (Molossidae), correlated with structure‐ function relationships in the adult. Anatomical Record (Hoboken) 292, 155–168.

104

Riskin DK, Parsons S, Schutt WA Jr., Carter GG, Hermanson JW (2006) Terrestrial locomotion of the New Zealand Short‐tailed Bat Mystacina tuberculata and the Common Vampire Bat Desmodus rotundus. Journal of Experimental Biology 209, 1725–1736.

Schmidt M, Fischer MS (2009) Morphological integration in mammalian limb proportions: dissociation between function and development. Evolution 63, 749–766.

Schnitzler HU, Moss CF, Denzinger A (2003) From spatial orientation to food acquisition in echolocating bats. Trends in Ecology & Evolution 18, 386–394.

Schutt WA Jr. (1998) Chiropteran hindlimb morphology and the origin of blood feeding in bats. In: Bat biology and conservation. (ed Kunz TH, Racey PA), pp. 157–168. Smithsonian Institution Press, Washington and London.

Schutt WA Jr., Simmons NB (2001) Morphological specializations of Cheiromeles (naked bulldog bats; Family Molossidae) and their role in quadrupedal locomotion. Acta Chiropterologica 3, 225–236.

105

Schutt WA Jr., Simmons NB (2006) Quadrupedal bats: form, function, and evolution. In: Functional and evolutionary ecology of bats. (ed Zubaid A, McCracken GF, Kunz TH), pp. 145–159. Oxford University Press, New York.

Simmons NB (1998) A Reappraisal of Interfamilial Relationships of Bats. In: Bat Biology and Conservation (ed Kunz TH, Racey PA), pp. 3–26. Smithsonian Institution Press, Washington.

Simmons NB, Conway TM (2003) Evolution of ecological diversity in bats. In: Bat ecology. (ed Kunz TH, Fenton MB), pp. 493–535. University of Chicago Press, Chicago.

Simmons NB, Seymour KL, Habersetzer J, Gunnell GF (2008) Primitive Early Eocene bat from Wyoming and the evolution of flight and echolocation. Nature 451, 818–822.

Smith JD (1972) Systematics of the chiropteran: family Mormoopidae. Miscellaneous Publication – University of Kansas, Museum of Natural History 56, 1–132.

Suárez, W (2005) Taxonomic status of the cuban vampire bat (Chiroptera: Phyllostomidae: Desmodontinae: Desmodus). Caribbean Journal of Science 41, 761–767.

Swartz SM (1997) Allometric patterning in the limb skeleton of bats: implications for the mechanics and energetics of powered flight. Journal of Morphology 234, 277–294.

Swartz SM, Middleton KM (2008) Biomechanics of the bat limb skeleton: scaling, material properties and mechanics. Cells Tissues Organs 187, 59–84.

Teeling EC, Scally M, Kao DJ, et al. (2000) Molecular evidence regarding the origin of echolocation and flight in bats. Nature 403, 188–192.

Teeling EC, Springer MS, Madsen O, et al. (2005) A molecular phylogeny for bats illuminates biogeography and the fossil Record. Science 307, 580–584.

106

Teeling EC, Dool S, Springer MS (2012) Phylogenies, fossils and functional genes: the evolution of echolocation in bats. In: Evolutionary history of bats: Fossils, molecules and morphology (ed Gunnel GF, Simmons NB), pp. 1–21. Cambridge University Press, New York.

Vaughan TA (1959) Functional morphology of three bats: Eumops, Myotis, Macrotus. University of Kansas Publications, Museum of Natural History 12, 1–153.

Vaughan TA (1966) Morphology and flight characteristics of molossid bats. Journal of Mammalogy 47, 249–260.

Vaughan TA (1970) The skeletal system. In: Biology of bats. (ed Wimsatt WA), pp. 97–138. Academic Press, New York.

Van Den Bussche RA, Hoofer SR (2004) Phylogenetic relationships among recent chiropteran families and the importance of choosing appropriate out-group taxa. Journal of Mammalogy 85, 321–330.

Velazco PM, O'Neill H, Gunnell GF, et al. (2013) Quaternary bat diversity in the Dominican Republic. American Museum Novitates 3779, 1–20.

Voss RS, Fleck DW, Strauss RE, Velazco PM, Simmons NB (2016) Roosting ecology of Amazonian bats: evidence for guild structure in hyperdiverse mammalian communities. American Museum Novitates 3870, 1–44.

Webster FA, Griffin DR (1962) The role of the flight membranes in insect capture by bats. Animal Behaviour 10, 332–340.

Ziegler AC, Howarth FG, Simmons NB (2016) A second endemic land mammal for the Hawaiian Islands: a new genus and species of fossil bat (Chiroptera: Vespertilionidae). American Museum Novitates 3854, 1–52.

107