INSTITUTO NACIONAL DE PESQUISAS DA AMAZÔNIA

PROGRAMA DE PÓS-GRADUAÇÃO EM ECOLOGIA

SELEÇÃO SEXUAL EM MULTI-ESCALAS: EFEITOS DA MALÁRIA

AVIÁRIA SOBRE O COMPORTAMENTO DE LEQUE E
PREFERÊNCIA DA FÊMEA DO UIRAPURU-DE-COROA-AZUL
(Lepidothrix coronata, AVES: PIPRIDAE)

MARIANE BOSHOLN

Manaus, Amazonas
Março 2015
 MARIANE BOSHOLN

SELEÇÃO SEXUAL EM MULTI-ESCALAS: EFEITOS DA MALÁRIA

AVIÁRIA SOBRE O COMPORTAMENTO DE LEQUE E
PREFERÊNCIA DA FÊMEA DO UIRAPURU-DE-COROA-AZUL
(Lepidothrix coronata, AVES: PIPRIDAE)

DRA. MARINA ANCIÃES

Orientadora

DR. ALAN FECCHIO

Coorientador

Dissertação apresentada ao Instituto

Nacional de Pesquisas da Amazônia
como parte dos requisitos para obtenção
do título de Mestre em Biologia
(Ecologia).

Manaus, Amazonas
Março 2015

i
BANCA EXAMINADORA DA DEFESA ORAL PÚBLICA

Dr. Wanderli Pedro Tadei

(INPA)

Dr. Renato Cintra Soares

(INPA)

Dra. Thierry Ray Jehlen Gasnier

(UFAM)

Aprovada por unanimidade

i
B743 Bosholn, Mariane
Seleção sexual em multi-escalas: efeitos da malária aviária sobre o
comportamento de leque e preferência da fêmea do uirapuru-de-coroa-azul
(Lepidothrix coronata, Aves: Pipridae) / Mariane Bosholn. --- Manaus: [s.n.],
2015.
59 f. : il. color.

Dissertação (Mestrado) --- INPA, Manaus, 2013.

Orientador : Marina Anciães.
Coorientador : Alan Fecchio.
Área de concentração : Ecologia

1. Malária aviária. 2. Uirapuru-de-coroa-azul. I. Título.

CDD 631.4

Sinopse:

Estudamos os efeitos de Plasmodium spp. (Sporozoa: Haemosporida) sobre o

comportamento de machos e fêmeas do uirapuru-de-chapéu-azul Lepidothrix
coronata (Aves: Pipridae) na Amazônia Central. Para tanto, aspectos como
condição corporal, comportamento de vocalização e de dança, além da interação
social foram avaliados.

Palavras chaves: Comportamento social, aves, Pipridae, Lepidothrix coronata,

interação hospedeiro-parasito, hemoparasitos, Plasmodium, Amazônia, BR 319

ii
Aos tangarás e à Floresta Amazônica
iii
AGRADECIMENTOS

Ao Instituto Nacional de Pesquisas da Amazônia (INPA), professores, pesquisadores e

funcionários que tanto contribuíram para meu crescimento e formação acadêmica.

Aos meus orientadores, Dra. Marina Anciães e Dr. Alan Fecchio pela orientação, confiança,

disposição, paciência, e experiência despendida a mim na execução da presente pesquisa.

Espero poder retribuir toda a ajuda que tive durante todo este tempo.

À Fundação de Amparo à Pesquisa do Estado do Amazonas (FAPEAM)

(PRONEX/FAPEAM/CNPQ EDITAL n° 003/2009; e Programa de Desenvolvimento

Cientificio Regional –MCT/FAPEAM/CNPQ EDITAL n° 005/2007) e ao Conselho

Nacional de Desenvolvimento Científico e Tecnológico (CNPq) (UNIVERSAL –

MCTI/CNPq N º 14/2012 – Faixa C) pela bolsa de pesquisa e pelo financiamento dado para

que pudessemos realizar todas as expedições de campo e de laboratório.

Ao Instituto Chico Mendes de Conservação da Biodiversidade – ICMBio (Número: 41069-1)

e Centro Nacional de Pesquisa para Conservação das Aves Silvestres – CEMAVE (Número:

3767) conceder as licenças necessárias e anilhas metálicas para realizar as expedições de

campo e coletar o material biológico na área estudada.

À David McDonald, Ph.D.; Renata Durães, Ph.D; Erik Johnson, Ph.D; Dr. Igor Kaefer, Dr.

Fabrício Baccaro e Dra. Thais Bilalba, pelas contribuições que fizeram ao meu meu plano de

mestrado, pelas sugestões, e questionamentos que me direcionaram para o caminho certo.

À Dra. Érika Braga, que gentilmente me recebeu do Laboratório de Malária da Universidade

Federal de Minas Gerais (UFMG). Também agradeço a todos os membros do Laboratório de

Malária, e em especial à Dra. Patricia Silveira.

Aos colegas do Laboratório de Evolução e Comportamento Animal (Mari, Fernando, Thiago,

Angélica e Arielli), e aos integrantes do Laboratório Temático de Biologia Molecular

iv
(LTBM), que ajudaram direta ou indiretamente durante a execução desta pesquisa, pelos

momentos sérios e de descontração.

Ao Programa de Pesquisa em Biodiversidade (PPBio), pelo apoio logístico, e aos moradores

do “Ramal do 10” – Km 100 da BR 319, que tão gentilmente nos receberam durante as

excursões de campo.

Aos companheiros de campo (Mariana, Nayara, Marina, Gueviston, Priscila, Ricardo,

Guilherme, Ayres e Zelão) que auxiliaram na coleta de dados na BR 319. Agradeço por toda

ajuda durante o trabalho de campo, pela amizade, pela ótima convivência, e por serem

parceiros de todas as horas. Sem vocês as excursões de campo não teriam sido realizadas.

Aos colegas da Ecologia INPA 2013/2015, pelo convívio, amizade e todo apoio durante o

mestrado.

Aos amigos de Santa Maria (RS) e Manaus (AM), em especial ao “Bando Misto” (Tainara,

Dilson, Patricia e Jean) e ao Pedro, companheiros do meu dia-a-dia, parceiros de tantas horas

e há tantas horas. Pelos momentos que passamos juntos, divertidos ou difíceis. Enfim, pelo

apoio em simplesmente tudo, e por estarem sempre presentes, mesmo estando a quilômetros

de distância.

Aos meus pais, Ivan e Vera, e minha irmã, Carolina, pelo exemplo de família que são, pelo

carinho, ensinamentos, e por toda confiança que foi depositada em mim. Muito obrigada por

acreditar, torcer e apoiar na realização desta pesquisa; por toda estrutura familiar que me

deram, pelo incentivo e por não me deixarem desistir.

A todos que contribuíram de alguma maneira para a realização deste trabalho, muito

obrigada!

v
 “Eles passarão
Eu passarinho...”

(Mario Quintana)

vi
RESUMO

A infecção de aves por hemoparasitos do gênero Plasmodium pode afetar a sobrevivência e

reprodução, já que os hospedeiros podem apresentar alterações dos caracteres usados na

seleção sexual. Embora diferentes estudos ressaltem a associação do parasitismo com a

diminuição do sucesso reprodutivo dos hospedeiros, ainda é incerto quais seriam os

componentes deste afetados pelo parasitismo. Por isso, investigamos se a infecção por

Plasmodium spp. compromete as taxas de vocalização, dança e interação social em machos

de Lepidothrix coronata (uirapuru-de-chapéu-azul), e se a mesma altera a preferência sexual

das fêmeas, em leks e indivíduos de uma população da Amazônia Central. Observações do

comportamento e diagnóstico molecular para detecção de hemoparasitos revelaram que a

infecção por Plasmodium afetou a atividade dos machos, bem como a escolha das fêmeas. E,

mesmo visitando tanto machos infectados quanto não infectados, as fêmeas discriminam os

mesmos pela sua atividade, visitando machos que vocalizam e dançam mais, os quais, são

menos infectados. Assim, sugere-se o papel da atividade de vocalização e dança dos machos

como um sinal honesto de resistenciados mesmos na população estudada. Por isso, as

exibições dos machos seriam um importante sinalizador para a seleção sexual, uma vez que

tais características permitem às fêmeas obterem pistas sobre a probabilidade de um macho

estar, ou não, infectado por Plasmodium spp.. Trabalhos futuros que testem a herança das

preferências e das atividade dos machos auxiliarão no entendimento do papel evolutivo dos

comportamentos aqui apresentados para mecanismos de preferências de fêmeas com

sinalização honesta.

Palavras-chave: Comportamento social, lek, aves, Pipridae, Lepidothrix coronata, interação

hospedeiro-parasito, hemoparasitos, Haemosporideo, malária aviária, Amazônia.
vii
ABSTRACT
Sexual selection in multiscales: Effects of avian malaria on male behavior and female
preference in lekking Blue-crowned Manakins

Infection by blood parasites of the genus Plasmodium can affect survival and reproduction in

the host, and may affect the expression of sexual selection traits. Although different studies

highlight the association of parasitism with a decrease in reproductive success of hosts, it is

still uncertain which fitness components are affected by the parasites. Hence, few studies have

addressed the role of parasites in honest signalling among lekking species. Here, we

investigated if infection by Plasmodium spp. compromises rates of vocalization, dance and

social interaction of male Blue-crowned Manakins (Lepidothrix coronata), and if it affects

female preferences in a Central Amazon population. Behavioral observations coupled with

molecular diagnosis to detect haemosporidian parasites revealed that infection by Plasmodium

affected male activities, as well as female choice. Furthemore, despite similar visiting

frequencies to both infected and uninfected males, they were more frequent to males that

vocalized and danced more, which were more frequently not infected. Therefore, we suggest a

role of male vocalization and dance rates as honest signals for females, indicating male

immunity in the studied population. Thus, male exhibitions would function as an important

signal for sexual selection, used by females as cues about the probability of a male being

infected. Future studies testing the inheritance of female preferences and male behaviour rates

will add to the understanding of the role of such behaviors in the evolution of mechanisms of

female preferences with honest signals.

Key-Words: Social behavior, host-parasite interaction, sexual signaling, Haemosporidae,

Plasmodium spp., Amazon

viii
SUMÁRIO

RESUMO..................................................................................................................... VII

ABSTRACT ................................................................................................................ VII

LISTA DE TABELAS ....................................................................................................X

LISTA DE FIGURAS................................................................................................... XI

INTRODUÇÃO GERAL ............................................................................................... 1

OBJETIVO GERAL ...................................................................................................... 5

OBJETIVOS ESPECÍFICOS........................................................................................ 5

CAPÍTULO 1: Effects of avian malaria on male behavior and female visitation in

lekking Blue-crowned manakins .................................................................................. 6

COVER PAGE ................................................................................................................ 7

LAY SUMMARY ........................................................................................................... 9

ABSTRACT .................................................................................................................. 10

INTRODUCTION ........................................................................................................ 11

MATERIAL E METHODS ......................................................................................... 15

RESULTS ...................................................................................................................... 19

DISCUSSION ................................................................................................................ 21

ACKNOWLEDGEMENTS ......................................................................................... 27

REFERENCES ............................................................................................................. 27

FIGURE LEGENDS .................................................................................................... 35

TABLES ........................................................................................................................ 36

FIGURES ...................................................................................................................... 38

CONCLUSÃO............................................................................................................... 44

ANEXOS ....................................................................................................................... 45

ix
LISTA DE TABELAS

Tabela 01: Índices para condição corporal e comportamento dos machos adultos de
L.coronata.................................................................................................................................36

Tabela 02: Colinearidade entre os índices comportamentais (condição corporal, taxa de

vocalização, taxa de dança, taxa de interação, taxa de atividade total e frequência de machos
no lek), entre índices dos leks e individuais..............................................................................37

x
LISTA DE FIGURAS

Figura 01: Mapa da área de estudo, módulo de pesquisa “Manaquiri”, Amazonas, Brasil. O
quadrado representa o módulo amostrado, e os pontos em preto (blackdots) representam os
leks. Adaptado de: http://ppbio.inpa.gov.br/sitios/br319/infra/km100....................................38

Figura 02: Gráfico demonstrando frequência de indivíduos capturados nos leks de

L.coronata: machos adultos (preto); fêmeas (cinza) e machos jovens (branco).......................39

Figura 03: Gráficos relacionando o (a) log da taxa de vocalização total do lek por minuto; e
(b) taxa de dança total do lek por minuto; com a prevalência de malária no lek......................40

Figura 04: Gráficos comparando (a) o resíduo do log da taxa de vocalização por minuto; e (b)
resíduo da taxa de dança por minuto; entre machos adultos infectados e não-infectados por
Plasmodium spp........................................................................................................................41

Figura 05: Gráficos demonstrando a variação da frequência de visitas de fêmeas em função

da taxa de interação de L. coronata dentro dos leks.................................................................42

Figura 06: Gráficos demonstrando a relação entre o comportamento social dos machos de
L.coronata e frequência de visitas de fêmeas. a) Efeito da taxa de vocalização do indivíduo
por minuto (log) sobre a frequência de visitas de fêmeas aos machos adultos; b) efeito da taxa
de dança do indivíduo por minuto sobre a frequência de visitas de fêmeas aos machos adultos;
c) efeito da taxa interação do indivíduo sobre a frequência de visitas de fêmeas aos machos
adultos.......................................................................................................................................43

xi
INTRODUÇÃO GERAL

A vida em grupo traz diversos benefícios aos animais que apresentam comportamento

social (Alexander 1974, Breed et al. 2004, revisão em Kob e Heinze 2008, Jones et al. 2010).

Entretanto, a vida em grupo também facilita a transmissão de patógenos como o Plasmodium

spp. (Daviews et al. 1991, Tella 2002, Nunn e Heymann 2005), já que a formação de

agrupamentos promove maior exposição dos hospedeiros aos vetores, aumentando a

probabilidade de infecção (Caillaud et al. 2013). Indivíduos infectados com Plasmodium spp.

tem sua condição corporal afetada, pois perdem massa corporal e sofrem redução nos

estoques de gordura (revisão em Atkinson e Van Ripper III 1991, revisão em Valkiūnas

2005, Willians 2005). Infecções por Plasmodium spp. também pode modificar as

características fenotípicas dos hospedeiros, podendo ocasionar redução em seu fitness, já que

prejudica a expressão das exibições reprodutivas dos individuos infectados (Buchanan et al.

1999, Gilman et al. 2007, Knowles et al. 2010, Asghar et al. 2011). Estas exibições podem

servir como indicadores honestos de resistência genética dos machos a doenças e parasitos,

sendo um importante sinalizador para a seleção sexual, como previsto pela Hipótese dos

parasitos (Hamilton e Zuk 1982). Essa hipótese prevê tanto benefícios diretos quanto indiretos

para a fêmea e sua prole, sendo, portanto, determinante durante a escolha de um macho

(Hamilton e Zuk, 1982, Able 1996, revisão em Beltran-Bech e Richard 2014). Afinal, as

chances da fêmea ser infectada por patógenos durante visitas à machos saudáveis são

relativamente pequenas (Able 1996, revisão em Beltran-Bech e Richard 2014). Além disso, ao

escolherem se reproduzir com machos com características sexuais secundárias que sinalizem

honestamente sua maior qualidade genética, as fêmeas estão contribuindo para aumentar a

viabilidade de seus descendentes (Hamilton e Zuk 1982).

Em espécies com sistema de acasalamento de lek poligínico, os machos se agregam

em arenas para atrair fêmeas, facilitando a comparação e escolha entre machos para cópula
1
(Bradbury e Gibson 1983). Nesse sistema, não há defesa de recursos pelos machos, que

fornecem apenas o sêmen para as fêmeas, sem haver necessariamente uma associação entre a

preferência da fêmea e a qualidade genética do macho (Bradbury e Gibson 1983). Deste

modo, haveria pouca oportunidade para a evolução de sinalização honesta de benefícios

diretos ou indiretos dos machos para as fêmeas em leks. De fato, modelos teóricos preveem

que preferências arbitrárias das fêmeas seriam suficientes para a evolução de caracteres

sexuais secundários exagerados nos machos, através de correlação genética entre a

preferência herdada e característica preferida (Runway Selection de Fisher 1930, revisões em

Kirkpatrick e Ryan 1991, Andersson 1994, Kuijper et al. 2012; testado em Prum 1997). Tal

modelo não exclui, entretanto, a existência de outros mecanismos de preferência de fêmeas

em leks, conforme já testado em diversos trabalhos (Endler e McLellan 1988, Endler e Thery

1996, Heidl e Winkler 2003ab, Gomez e Théry 2004, Uy e Stein 2007, Anciães e Prum 2008),

incluindo a hipótese dos parasitos (Boyce 1990, Johnson e Boyce 1991, Höglund et al. 1992,

revisão em Kirkpatrick e Ryan 1991).

A hipótese dos parasitos (Hamilton e Zuk, 1982, revisão em Kirkpatrick e Ryan 1991)

propõe a existência de sinalizações diferenciadas entre machos dentro dos leks, e isso teria um

efeito sobre a preferência da fêmea, podendo haver visitação distinta a machos com

determinada característica comportamental associada a um grau de parasitismo. Poucos

trabalhos testaram esta hipótese em espécies com sistema de acasalamento de leks (Boyce

1990, Johnson e Boyce 1991, Höglund et al. 1992), e não houve consenso entre os resultados

encontrados (Höglund et al. 1992, Lebigre et al. 2013). Assim, não se sabe ao certo quais os

efeitos dos parasitos sobre os caracteres sexuais dos machos, a preferência de fêmeas de

espécies com sistema de acasalamento de lek e se, de fato, espécies deste sistema apresentam

esse mecanismo de preferência. Por isso, são necessários estudos a fim de compreender se

2
parasitos realmente comprometem o comportamento de machos, e se a escolha da fêmea é

afetada pela presença dos mesmos.

O uirapuru-de-chapéu-azul Lepidothrix coronata (Spix, 1825) (Aves: Pipridae) é uma

das espécies que possui sistema reprodutivo de leks poligínicos (Snow 1963, Sick 1967,

Skutch 1969, Snow 2004, Durães 2009a). Os machos realizam exibições reprodutivas dentro

de leks explodidos, que podem ter até sete indivíduos (Anciães et al. 2009,Durães 2009a,

revisão em Kirwan e Green 2012). Interações entre machos adultos são relativamente comuns

(Durães 2009a). Durante essas associações, além de executar exibições de dança, os machos

de L. coronata também vocalizam em dueto. Essas vocalizações são emitidas nos poleiros de

dança e são boas previsoras do sucesso reprodutivo dos indivíduos machos adultos, indicando

que as fêmeas preferem os machos que vocalizam mais dentro de cada lek (Durães et al. 2008,

2009b). Além disso, fêmeas preferem machos que interagem com outros indivíduos durante

duetos vocais e durante exibições de dança, já que isso pode promover tanto benefícios diretos

quanto indiretos para as mesmas (Durães et al. 2009b, revisão em Diaz-Munhoz 2014).

Como o gênero Plasmodium tem efeito negativo sobre o fitness de hospedeiros de

diferentes espécies (Hamilton e Zuk 1982, Schall 1983, Marzal et al. 2005), hipotetizamos

que o Plasmodium spp. também compromete o comportamento do L. coronata em

multiescalas, tanto em nível de lek quanto em nível individual. Em nível de lek, esperamos

que haja, no total, menos vocalização e dança em arenas com alta prevalência de Plasmodium

spp, havendo variação nestas atividades entre machos dentro dos leks. Ainda, dada a maior

prevalência de malária observada em maiores agrupamentos (Daviews et al. 1991, Tella 2002,

Nunn e Heymann 2005), esperamos que os indivíduos sejam mais infectados em leks onde

haja maior interação entre os machos. Os leks com alta prevalência de Plasmodium spp.

seriam, entretanto, os mais visitados pelas fêmeas, com elas preferindo leks com mais

interação total entre os machos (revisão em Diaz-Munhoz 2014). Em nível indivídual,

3
esperamos que a probabilidade de ser infectado pelo vetor seja maior para machos que

interajam mais. Além disso, esperamos que Plasmodium spp. tenham efeito negativo sobre o

comportamento dos indivíduos, com os infectados apresentando pior condição corporal e

vocalizando e dançando menos. Hipotetizamos ainda, baseado em estudos anteriores, maior

frequência de vistas das fêmeas a machos mais ativos e, uma vez que os comportamentos

utilizados como sinalizador para as fêmeas estariam comprometidos mediante incidência de

Plasmodium spp., as visitas das fêmeas também seriam afetadas pela infecção. Assim, as

fêmeas visitariam mais machos não-infectados por Plasmodium spp. Vale ressaltar que as

hipóteses apresentadas diferem daquelas testadas em estudos anteriores, que consideraram ou

somente a frequência de exibições reprodutivas individuais (Durães et al. 2008, 2009b et al.)

ou apenas o grau de interação entre machos independentemente da identidade dos leks (Diaz-

Munhoz et al. 2014) como previsores do comportamento das fêmeas. Afinal, a frequência de

exibições poderia ser independente do número de machos interagindo dentro do lek para a

população em estudo e, sobretudo, que a preferência de fêmeas pode não refletir o grau de

interação entre indivíduos, mas sim a ocorrência de Plasmodium spp.

OBJETIVO GERAL

Investigar se a infecção por Plasmodium spp. compromete o comportamento social em

machos, e se altera a preferência da fêmea do uirapuru-de-chapéu-azul (Lepidothrix

coronata), em multi-escalas (lek e indivíduo), em uma população de vida livre, na Amazônia

Central.

OBJETIVOS ESPECÍFICOS

4
Testar as hipóteses de que:

1) Em nível de lek, a taxa de interação total entre machos é correlacionada à

prevalência de Plasmodium spp., que por sua vez tem efeito negativo sobre a taxa

total de vocalização e sobre a taxa total de dança para cada lek;

2) Em nível indivídual, a ocorrência de Plasmodium spp. afeta negativamente a

condição corporal de machos e fêmeas, a taxa de vocalização, a taxa de dança dos

machos adultos, e a taxa de interação entre machos;

3) A frequência de visitas de fêmeas é maior em leks com alta taxa de interação total

(onde a prevalência de Plasmodium spp. é maior) e a machos não-infectados com

Plasmodium spp., sendo positivamente relacionada à taxa de vocalização, taxa de

dança, interação entre indivíduos ao todo para cada lek e para cada macho dentro

dos leks.

5
 Capítulo 1

Bosholn M, Fecchio A, Silveira P, Braga EM, Anciães M. Effects of avian malaria on

male behavior and female visitation in lekking Blue-crowned Manakins. Manuscrito

formatado para a revista Behavioral Ecology

6
 Effects of avian malaria on male behavior and female visitation in lekking Blue-

crowned Manakins

Mariane Bosholn1*, Alan Fecchio2,4,5, Patricia Silveira3, Érika M. Braga3, Marina

Anciães4

¹ Programa de Pós-Graduação em Ecologia, Instituto Nacional de Pesquisas da

Amazônia, Manaus, Amazonas, Brazil.

² Universidade Federal do Amazonas, Manaus, Amazonas, Brazil.

³ Laboratório de Malária, Departamento de Parasitologia, Universidade Federal de

Minas Gerais, Belo Horizonte, Minas Gerais, Brazil.

4
Laboratório de Evolução e Comportamento Animal; Coordenação de Biodiversidade

(CBio); Instituto Nacional de Pesquisas da Amazônia, Manaus, Amazonas, Brazil.

5
Present Addres: Ornithology Department: Academy of Natural Sciences of Drexel

University, Philadelphia, Pennsylvania, United States of America

*Corresponding author

E-mail Address: [email protected]

Telephone/Fax: (0xx92) 36431823

7
8
1 LAY SUMMARY

3 If you are sick, you shall not reproduce! Avian malaria causes serious damage to

4 sick individuals. This explains why male Blue-crowned Manakins infected with Malaria

5 sing and dance less than uninfected individuals. Because females like active males, who

6 are also healthy, they avoid visiting those who are sick. Healthy males are therefore

7 more atractive, and might be better for females and their offspring, since their risk of

8 becoming infected will be reduced.

9
10 Effects of avian malaria on male behavior and female visitation in lekking Blue-

11 crowned Manakins

12 ABSTRACT

13 Infection by blood parasites of the genus Plasmodium can affect survival and

14 reproduction in the host, and may affect the expression of sexual selection traits.

15 Although different studies highlight the association of parasitism with a decrease in

16 reproductive success of hosts, it is still uncertain which fitness components are affected

17 by the parasites. Hence, few studies have addressed the role of parasites in honest

18 signalling among lekking species. Here, we investigated if infection by Plasmodium

19 spp. compromises rates of vocalization, dance and social interaction of male Blue-

20 crowned Manakins (Lepidothrix coronata), and if it affects female preferences in a

21 Central Amazon population. Behavioral observations coupled with molecular diagnosis

22 to detect haemosporidian parasites revealed that infection by Plasmodium affected male

23 activities, as well as female choice. Furthemore, despite similar visiting frequencies to

24 both infected and uninfected males, they were more frequent to males that vocalized and

25 danced more, which were more frequently not infected. Therefore, we suggest a role of

26 male vocalization and dance rates as honest signals for females, indicating male

27 immunity in the studied population. Thus, male exhibitions would function as an

28 important signal for sexual selection, used by females as cues about the probability of a

29 male being infected. Future studies testing the inheritance of female preferences and

30 male behaviour rates will add to the understanding of the role of such behaviors in the

31 evolution of mechanisms of female preferences with honest signals.

32 KEY-WORDS: Social behavior, host-parasite interaction, sexual signaling,

33 Haemosporidae, Plasmodium spp., Amazon.
34

10
35 INTRODUCTION

36 Living in groups brings many benefits to animals (Alexander 1974, Breed et al.

37 2004; review in Korb and Heinze 2008, Jones et al. 2010). However, social behaviors

38 also favors the transmission of pathogens such as Plasmodium spp. (Daviews et al.

39 1991, Tella 2002, Nunn and Heymann 2005), because spatial aggregation of hosts

40 promotes higher exposition to vectors, increasing the probability of infection (Caillaud

41 et al. 2013). Individuals infected by Plasmodium spp. have their body condition

42 affected, since they lose body mass and suffer reductions in fat reserves (Williams 2005;

43 review in Atkinson and Van Ripper III 1991, Valkiūnas 2005). Species of the genus

44 Plasmodium also modify phenotypic characters of infected individuals, which may lead

45 to reduction in fitness, since it affects the expression of host courtship displays

46 (Buchanan et al. 1999, Gilman et al. 2007, Knowles et al. 2010, Asghar et al. 2011).

47 Such displays may serve as honest indicators of genetic resistance of males to diseases

48 and parasites, acting as an important signal for sexual selection, as predicted by the

49 Parasite Hypothesis (Hamilton and Zuk 1982). This hypothesis predicts both direct and

50 indirect benefits for the female and its offspring, hence being determinant when

51 choosing a male (Hamilton and Zuk 1982, Able 1996; review in Beltran-Bech and

52 Richard 2014). After all, chances of the female being infected by pathogens, while

53 visiting healthy males, are relatively small (Able 1996; review in Beltran-Bech and

54 Richard 2014). Besides, when choosing to reproduce with males presenting secondary

55 sex characteristics that honestly signals high genetic quality, females are contributing to

56 increase the viability of their offspring (Hamilton and Zuk 1982).

57 In species that present polygynous lek mating system, males aggregate in arenas

58 to attract females, facilitating comparison and choice between males for copulation

11
59 (Bradbury and Gibson 1983). In this system, males provide only semen for females, and

60 there is no resource defense by males. Therefore, there is no benefit necessarily

61 associated to female choice (Bradbury and Gibson 1983). According to this cencept,

62 there would be little opportunity for the evolution of honest signals indicating male

63 quality to females in leks. In fact, theoretical models predict that arbitrary female

64 preferences would be sufficient for the evolution of exaggerated secondary sexual

65 characteristics in males, through genetic correlation between inherited preferences and

66 preferred male traits (Runway Selection of Fisher 1915, Lande 1981 and Kirkpatrick

67 1982; reviews in Kirkpatrick and Ryan 1991, Andersson 1994, Kokko et al. 2006, Prum

68 2010; tested in Prum 1997). However, such model does not exclude the existence of

69 other mechanisms of female choice in leks (Prum 2010, 2012), accordingly tested in

70 several works (Endler and McLellan 1988, Endler and Thery 1996, Heindl and Winkler

71 2003ab, Gomez and Théry 2004, Uy and Stein 2007, Anciães and Prum 2008),

72 including the Parasite Hypothesis (Hamilton e Zuk 1982, Boyce 1990, Johnson and

73 Boyce 1991, Höglund et al. 1992; review in Kirkpatrick and Ryan 1991).

74 According to the parasite hypothesis (Hamilton and Zuk, 1982; review in

75 Kirkpatrick and Ryan 1991) there would be differential signalizations among males

76 within leks reflecting their degree of parasitism, which would affect female choice.

77 Only a few studies have tested this hypothesis in species with lek mating system, and

78 there is no consensus among their results (Boyce 1990, Gibson 1990, Johnson and

79 Boyce 1991, Höglund et al. 1992, Lebigre et al. 2013). Therefore, it is unclear what are

80 the effects of parasites on secondary sexual traits of males and on female choice in this

81 system and if, in fact, species in this system present such a preference mechanism.

82 Therefore, studies of lekking species focuising on the Parasite Hypothesis will provide

12
 83 further comprehension on the effect of parasites on male behavior and female choice in

84 leks.

85 Male blue-crowned manakins, Lepidothrix coronata (Spix, 1825) (Aves:

86 Pipridae), perform courtship displays in exploded leks (Snow 1963, 2004; Sick 1967,

87 Skutch 1969, Anciães et al. 2009, Durães 2009a; review in Kirwan and Green 2012),

88 and group displays with up to seven males, including individuals with both pre-

89 definitive and definitive plumage, are relatively commom. During these associations,

90 besides performing aerial and on-perch dispays, males vocalize from display perches as

91 part of their courtship behavior. These vocalizations were indeed good predictors of

92 male reproductive success within a population, since females preferred males with

93 highest vocalization rates on each lek (Durães et al. 2008, 2009b). Furthemore, in some

94 manakin species, females preferred males that interacted with other individuals in vocal

95 duetting and group displays, suggesting that these social behaviors can promote direct

96 and indirect benefits for females (review in Díaz-Muñoz 2014).

97 Because Plasmodium spp. have negative effect over fitness in different host

98 species (Hamilton and Zuk 1982, Schall 1983, Marzal et al. 2005), it may also

99 compromise the behavior of males and females L. coronata, both at lek and individual

100 levels. At the lek level, an increase in prevalence of Plasmodium spp. could reduce

101 vocalization and display rates, with variation in these activities among males within

102 leks. Still, given the higher prevalence or Plasmodium spp. observed in larger groups

103 (Daviews et al. 1991, Tella 2002, Nunn and Heymann 2005), a higher frequency of

104 individuals would be infected in leks with higher interaction among males. However,

105 considering female preferences for increased interaction among males (Díaz-Muñoz

106 2014), there would be higher frequency of female visitation to leks with higher

13
107 prevalences of Plasmodium spp.. At the individual level, the probability of being

108 infected by the vector would be higher for males that interact more with other males.

109 Besides, Plasmodium spp. would have a negative effect on the behavior of male

110 individuals, in which infected birds would present decreased corporal condition and

111 vocalization and display rates. Lastly, females would visit most active males more

112 often. If indeed these behaviors are preferred, and are compromised by incidence of

113 Plasmodium spp., then female visitation would be also affected by the infection and,

114 consequently, females would visit uninfected males more often. It is worth mentioning

115 that the presented hypotheses differ from those tested in previous studies, which

116 consider only either individual display frequencies (Durães et al. 2008, 2009b et al.), or

117 the degree of interaction among males, regardless of lek identity (Díaz-Muñoz et al.

118 2014), as cues for female behavior. Here, we consider the possibility that male display

119 frequencies could be independent of the number of males interacting inside a lek in the

120 studied population and, especially, that female choice may not reflect the degree of

121 interaction among male individuals, but whether males are infected by Plasmodium spp.

122 In this context, we investigated if the infection by Plasmodium spp.

123 compromises male social behavior of lekking blue-crowned manakin, and if female

124 choice varies in multiscales (leks and individuals) in a population from Central

125 Amazon. To do so, we specifically tested the hypotheses that: (i) at lek level, the rate of

126 total interaction among males is correlated to the prevalence of Plasmodium spp., which

127 in turn has a negative effect over the total rate of both vocalization and dance for each

128 lek; (ii) at individual level, the presence of Plasmodium spp. which also affect

129 negatively the body condition of males and females, as well as vocalization and display

130 rates, and rate of interaction among males; (iii) frequencies of female visitation are

14
131 higher in leks with high rates of total interaction among males, despite the predicted

132 higher prevalence of Plasmodium spp. in these leks, and to males that are not infected

133 by Plasmodium spp., being positively related to the total vocalization and display rates

134 at leks and for each individual.

135

136 MATERIAL AND METHODS

137 Study Area

138 We carried out the fieldwork in a research site from the Biodiversity Research

139 Program (PPBio) located aproximatedly 100 km from Manaus along the 319 Highway,

140 within the Purus-Madeira interfluvium and south of the Amazon River in Amazonas

141 State, Brazil (Figure 1). The study area lies within a well preserved terra-firme forest

142 composed of dense vegetation typical of ombrophilous forests with emerging canopy

143 (RADAM Brasil 1978). The climate in the region is characterized by high pluviosity

144 (2000 mm), with annual average temperature of 27°C, and the geomorphology consists

145 of large tabular interfluves with plain topography (RADAM Brasil 1978) and altitudes

146 ranging from 30 to 50 meters. The presence of micro-terrain with altitudes ranging from

147 one to three meters is frequent and favors the formation of temporary pools on the site

148 (PPBio 2014).

149

150 Studied Population and Individual Sampling

151 We sampled individuals from a population belonging to the form L. c. arimensis

152 from the L. c. coronata complex, in which male plumages present considerable amounts

153 of greenish plumage in the body, wings and tail feathers (Hellmayr 1929, Anciães et al.

154 2009, Kirwan and Green 2012), representing intermediary forms between the black-

15
155 bodied males typical of the species and the green-bodied forms from southwestern

156 Amazon. In spite of their green plumage feathers, definitive plumaged males can easily

157 be differentiated from those in pre-definitive plumages by the blackish plumage in the

158 upper body and their typical blue crown.

159 We collected data for individuals present in ten leks, during the reproductive

160 seasons of 2013 (October and November) and 2014 (June and July). Display perches of

161 adult males were located through continuous acoustic census and visual search along

162 trails. Lekking behavior was diagnosed by persistence of individuals in perches across

163 consecutive days and by observation of behavioral elements typical of their courtship

164 displays (Durães 2009a). After locating individual perches, we used approximately

165 fifteen mist nets with 12 meters in length by 3 meters in heigth in each lek, close to the

166 display perches, for two consecutive days. We captured 67 blue-crowned manakins

167 individuals, measured their body mass with a dynamometer (in grams) and their tarsus

168 length with a digital caliper (in millimeters), and verified presence of brood patch for

169 preliminary sexing of green-plumaged birds. We marked individuals with two colored

170 bands on the right tarsus and a metallic band on the left tarsus. We collected one blood

171 sample of approximately 50 ul from each individual through puncture of the brachial

172 vein, using a disposable hypodermic needle and microcapillaries. The blood was stored

173 in microtubes with approximately 1 ml of 95% ethanol until DNA extraction. Because

174 of the conspicuous sexual dimorphism typical of the species (Hellmayr 1929, Sick

175 1967, Snow 2004, Ryder and Durães 2005, Kirwan and Green 2012), we sexed

176 individuals in the field according to the color of their plumage. Males with olive-green

177 and black body plumage with a bright blue crown were classified as adult males in

178 definitive plumage, whereas individuals with predominantly green plumage were

16
179 classified as individuals of unkown sex. The gender of these individuals was determined

180 through molecular sexing using protocols described elsewhere (Ito et al. 2003). The

181 entire sampling was conducted with proper authorization from the Instituto Chico

182 Mendes de Conservação da Biodiversidade (Number: 41069-1) and Centro Nacional de

183 Pesquisa para Conservação das Aves Silvestres (Number: 3767).

184

185 Behavioral Observations

186 MB conducted bevavioral observations on male courts in the mornings

187 (07h00min – 12h00min), and afternoons (13h00min – 16h00min) through continuous

188 focal sampling with five-minute intervals (Altman 1974). The observation protocol

189 included, for each observation interval, sampling the total number of vocalizations and

190 emitted by each focal male, occurrence of courtship displays, number of adult males

191 interacting with the focal male during vocalizations and displays, and frequency of

192 observed females visits. The frequency of female visits was considered the proxy for

193 female choice. We considered as visiting events, those in which both banded and non-

194 banded females, and both infected and uninfected females visited the display perches.

195 Females do not frequently vocalize, when compared to young and adult males (Durães

196 2009a). That being the case, in every event in which a non-banded, green-feathered

197 individual visited a perch occupied by an adult male, and remained in front of them

198 without emitting vocalizations, that individual was also considered to be a female.

199 Vocalization, displays, and interaction behavior among young males were not

200 considered because these individuals present a nomadic behavior, and may visit several

201 leks during a single reproductive season (Durães 2009a).

202

17
203 Molecular diagnosis of blood parasites

204 DNA extractions were screened by Nested Polymerase Chain Reaction to detect

205 haemosporidian DNA following the protocol developed by Hellgren et al. (2004). This

206 protocol uses parasite-specific primers targeting a 479 bp fragment of the cytochrome b

207 gene (cyt b) of Plasmodium and Haemoproteus. Positive and negative controls were

208 included in all PCR runs. The product of the second PCR, from infected individuals,

209 was purified according to Sambrook et al. (2001). In order to identify the species of

210 parasite, the product of the PCR from individuals considered parasitized was sequenced

211 after purification, according to the protocol described in Lacorte et al. (2013).

212

213 Data analysis

214 We calculated several indexes to summarize the social behaviors of males and

215 females observed at individual perches and leks (Table 1). Vocalization rates are

216 presented for one-minute intervals. Collinearity among behavioral indexes were tested

217 through Pearson correlations (Table 2). Prevalence of infection by Plasmodium spp.

218 may vary seasonally (Bensch et al. 2007, Hellgren et al. 2013) and, for this reason, we

219 tested for similarity in prevalence of individuals captured infected between years (χ² 1 =

220 1.11, P = 0.34). Because there was no difference, we pooled the data for both years in

221 subsequent analyses.

222 We compared the effects of the occurrence of Plasmodium spp. over the body

223 condition of individuals using Student t-tests between infected and uninfected females

224 and adult males, separatedly for each sex. In order to verify if the prevalence of

225 Plasmodium spp. was related to vocalization and display behaviors at lek level, we used

226 simple linear regressions. The association between prevalence of Plasmodium spp. at

18
227 leks and the total interaction frequency at leks was tested with Pearson correlation. We

228 compared the behavior of infected individual males to that of uninfected males through

229 Student t-tests for vocalization rate, display and interaction frequencies, separately. The

230 prediction that the frequency of female visits to uninfected males was higher than to

231 infected males, was tested though the Student t-test. Finally, we used simple linear

232 regressions to test the predictions that the frequency of female visits was higher in leks

233 with increased prevalence of Plasmodium spp., being positively related to male

234 vocalization rate, display and interaction frequencies - across leks and among

235 individuals. Non-normal and/or heteroscedastic data was logarithmized. All of the

236 analyses were done through use of the software R version 3.1.1 (R Development Core

237 Team (2011).

238

239 RESULTS

240 We captured and observed males in ten leks during the reproductive seasons of

241 2013 and 2014. The number of females and males (juvenile and adults) varied among

242 leks (Figure 2). We captured a total of 67 individuals of blue-crowned manakins; 29

243 from which were adult males, 11 were juvenile males and 27 were females. From the

244 total of 29 adult males captured, 28 participated in leks with two to six individuals, and

245 only one participated in a solitary lek ( x +SE = 2.9± 1.44, N = 29). There was also

246 variation among leks in the number of females captured ( x +SE = 2.7± 1.41, N = 27).

247 The frequency of capture of juvenile males was low on average, when compared to the

248 total number of females and adult males captured ( x +SE = 1.1± 0.99, N = 11).

249 Although there were a few female recaptures between different leks, we did not

250 recapture juvenile or adult males across leks. The majority of observed females were

19
251 also banded. Each adult male captured was observed for two consecutive days ( x +SE =

252 10 hours ± 2.14). The observed adult males were present (vocalizing, displaying or

253 silent) in the leks for 65% of the total observation time.

254 The prevalence of infection by Plasmodium spp. for the captured birds was 35%

255 within adult males and 38% within females. Sequencing of the cyt b DNA fragment

256 revealed the presence of four blood parasite strains from the genus Plasmodium. Among

257 the sequences obtained for Plasmodium from 23 individuals, the Plasmodium sp. Lecor

258 01 strain was prevalent (N = 18), while the other three strains were observed in fewer

259 individuals (strains Plasmodium sp. HMA-2012 isolate BCRM2 (N = 3), Plasmodium

260 sp. G22 (N = 1), and Plasmodium sp. P-T138 isolate NK162353-T138 (N = 1)). The

261 colinearity among individual indexes were all significant, but not for the indexes at lek

262 level. Interaction frequencies at leks were correlated to the number of individuals at the

263 leks (Table 2) The infection caused by Plasmodium spp. did not affect the body

264 condition of adult males (t 28 = 0.61, P = 0.54) nor of females (t 26 = 0.96, P = 0.34). On

265 the other hand, we observed that in leks with high prevalence of infected males there

266 were, in total, less episodes of vocalization and displays (r² 9 = 0.34, P = 0.04 and r² 9 =

267 0.43, P = 0.02 respectively; Figures 3a, b), but the total interaction frequency in these

268 leks remained unaltered by the prevalence of lek infection (r 9 = 0.09, P = 0.78).

269 The frequency of interaction did not differ between infected and uninfected

270 males (t 26 = - 0.37, P = 0.35). However, since both individual vocalization rate and

271 display frequency were correlated with the individual interaction frequency, it is

272 possible that the social interaction had masked the effect of the infection on those

273 individual activities. Therefore, we used the residuals of each of those relations to

274 compare the vocalization rates and display frequencies, respectively, between infected

20
275 and uninfected males. We observed that, individually, uninfected males vocalized and

276 displayed more than infected males (t 26 = -2.31, P = 0.01 and t 26 = -2.37, P = 0.01

277 respectively; Figures 4a, b).

278 The prevalence of infection by Plasmodium spp. in the leks did affect frequency

279 of female visits to the leks (r²9 = -0.06, P = 0.53). The frequency of female visits was

280 also not correlated to the total vocalization rate (r²9 = -0.09, P = 0.66) nor to the total

281 display frequency (r²9 = 0.04, P = 0.26) in leks, when testing for each behavior

282 separately. On the other hand, the frequency of female visits to the display perches was

283 positively correlated to the total interaction frequency inside each lek (r² 9 = 0.39, P =

284 0.02) (Figure 5).

285 The frequency of female visits to infected individual males did not differ from

286 those to uninfected males (t26 = 0.39, P = 0.69). However, the frequency of visits was

287 related to the social behavior of adult males: individuals that vocalized and displayed

288 more (r²26 = 0.12, P = 0.04 and r² = 0.26, P < 0.01 respectively; Figures 6a, b) were the

289 ones preferred by the females. Besides, the higher the number of males with which an

290 individual interacted, the higher was the number of female visits that individual

291 received (r²26 = 0.23, P = 0.01, Figures 6c).

292 DISCUSSION

293 Our results show that infection by Plasmodium spp did not compromize the

294 body condition of neither males nor females. Our sampling scheme, with mist nets

295 placed close to display perches, may have favored the capture of individuals that were

296 healthy enough to engage in lekking activities, such as uninfected and chronically

297 infected males. In fact, this group of individuals did not present the clinical alterations

298 frequently observed during infection by Plasmodium spp., such as inactivity (review in

21
299 Atkinson e Van Ripper III 1991). However, inactive males would not contribute to

300 answer our hipotheses and, likewise, sampling outside leks would provide a random

301 sample of the prevalence and effects of the infection in the population, but would not

302 allow us to address explicitly the effect of the parasite on the social behavior of the

303 studied birds, which could only be done by comparing infected and not infected

304 individuals attending arenas and, to some extent, able to perform lekking activities. The

305 behavior of the sampled infected individuals suggests that these individuals were in the

306 chronic phase of infection, which is characterized by low parasitaemia of Plasmodium

307 spp. Additionally, it is possible that the Plasmodium spp. strains found in blue-crowned

308 manakins are pathogenic only during the acute phase of infection, in which the hosts

309 present high parasitaemia (high abundance of parasites in the bloodstream). During the

310 acute phase, infected individuals would remain quiescent for most of the time and, this

311 way (review in Atkinson e Van Ripper III 1991), would not fall in the mist nets.

312 Moreover, only during the initial acute phase of the disease (high parasitemia), in which

313 hosts are less active and anemic, the malaria parasites could have an effect on bird’s

314 body condition.

315 The presence of Plasmodium spp. affected the vocalization and display

316 behaviors of males both at lek and individual levels, corroborating our initial

317 hypothesis. In leks with high prevalence of malaria, uninfected males vocalized and

318 displayed more than infected ones. Such effects on host behavior might be explained by

319 the high pathogenicity of the Plasmodium spp. strains that were found in the studied

320 population, which are in majority phylogenetically related to the clade of morphospecies

321 of the subgenus Haemamoeba. This subgenus is composed by different Plasmodium

22
322 species that are considered highly pathogenic, such as P. tejerai (Silveira et al. 2013), P.

323 relictum, and P. cathemerium (Valkiūnas 2005).

324 Exposition to parasites stimulates the immunological response of hosts, resulting

325 in immunological response in the host (review in Ardia and Schat 2008). Such

326 investments in immunological response may compromize the investiment in breeding

327 activities, such as display behaviors within Pipridae, which are considered to be costly

328 (e.g. Manacus Vitellinus, Barske et al. 2014). Therefore, there might be a trade-off

329 between investiments in immune response and breeding activities, such as vocalization

330 and displays. Infected males could therefore invest either in survival or reproductive

331 success. Based on our results, we hypothesized that the infected males invested more in

332 fighting the infection by Plasmodium spp., in an attempt to increase their longevity.

333 Against our prediction, social interaction among males in each lek was not

334 correlated to malaria prevalence. This result may be explained by the fact that

335 Plasmodium spp. are parasites with indirect lifecycles and, therefore, requires both an

336 intermediate and final host to complete their cycles. Since the vectors of Plasmodium

337 spp. are highly mobile mosquitoes, transmission becomes less dependent on the

338 aggregation of individuals and, for that reason, infection patterns tend to be random

339 (Godfrey et al. 2006). We also did not find any relation between the frequency of

340 interaction among individuals and the occurrence of infection by Plasmodium spp. This

341 may also be explained by the fact that malaria transmission depends of the vector

342 mobility, without transmission among individuals. Therefore, when a male becomes

343 infected with Plasmodium spp., the other males that interact with this individual will not

344 necessarily become infected.

23
345 Despite the effect of Plasmodium spp. on the vocal rates and display frequencies

346 of male Blue-crowned Manakins, the courtship behavior of L. coronata males were not

347 regulated only by Plasmodium spp. infection. The frequency of social interaction also

348 affected these behaviors and, therefore, had masked the effect of the infection on male

349 activity when social interaction had not been controlled for – such that an infected male

350 could be as active as an unifected male depending on the level of interaction performed

351 with other males. Furthemore, despite the lack of statistical difference in social

352 interaction btween infected and non-infected males, and because more active males

353 interacted more, and were more frequently infected, our data suggest an indirect effect

354 of infection by Plasmodium spp. on social interaction, through the effect of male

355 activity. In fact, during behavioral sampling we observed that as soon as a male started

356 to vocalize, or even head to the display perch, other individuals started to interact with

357 that male, so that higher activity rates performed by one individual urged others to

358 interact. The frequency of social interaction derived for leks, in the other hand, was

359 dictated by lek size (number of males in the lek) and not associated to male activity.

360 These result differs from what Durães et al. (2009b) found for Blue-crowned Manakins

361 in Ecuador, where higher vocalization rates were observed in larger leks. Here, we

362 observed males with high activity and interaction rates, even in leks where social

363 interaction rates were low (i.e. in smaller leks). Therefore, we suggest that, in the

364 studied population, the individual level, and not the lek level like in Durães et al.

365 (2009b), regulates male activity..

366 In our study, females visited more frequently males in larger leks, where total

367 frequency of social interaction was high,, corroborating our hypothesis. This result is in

368 accordance to the female choice hypothesis (Beehler and Foster 1988, review in Diaz-

24
369 Munhoz 2014), which predicts that females prefer sites where many males interact,

370 making possible to compare their secondary sexual characteristics before selecting a

371 partner. The displays performed during these coalitions may allow females to evaluate

372 which male is the best, considering that counter-singing and group displays can serve as

373 an indicator of genetic quality of the male, showing how coordinated the individuals are

374 (Maynard et al. 2014). The total vocalization rate and total display frequencies derived

375 for leks, in turn, were not used as cues by the females observed in our study. At the

376 largest sampled lek, for instance, we captured six males and, although the total

377 frequency of social interaction was high in this lek, its total vocalization rate was similar

378 to that of smaller leks. This possibly occurred because not all individuals inside the lek

379 vocalized or displayed actively, with considerable variation among males within a

380 single lek. In turn, in the smallest sampled lek, only one male was captured and, despite

381 the zero social interaction in this lek, its vocalization rate was relatively high.

382 Our data suggest therefore that female choice was based only secondarily on

383 male secondary sexual characteristics, with more female visits to males that vocalized

384 and displayed more and that interacted with a higher number of males, on each lek.

385 Considering that uninfected males are more active, and that females visit them more

386 often, it is possible that vocalization and display behaviors are honest male signals,

387 reliable to indicate male health to females, thus helping in their choice when visiting

388 leks (Zahavi 1975). In turn, the frequency of social interaction of individual males,

389 although correlated to their vocalization rates and display frequencies would not be used

390 by females to distinguish between infected and uninfected males, because of the lack of

391 association between infection and social interaction frequencies.. Therefore, when

392 choosing to visit males that vocalized and displayed more on each lek, female Blue-

25
393 crowned Manakins would have a higher probability of choosing a healthy male, even if

394 the number of males interacting with it could not be used as a health indicator.

395 However, we cannot disregard that, even corroborating the parasite hypothesis

396 (Hamilton and Zuk 1982, Kirkipatrick and Ryan 1991), in this study we only considered

397 the effects of Plasmodium infection in short term, without considering immune and

398 hormonal factors, or other mechanisms that could be influencing female preferences.

399 For that reason, future studies about the effects of infection by Plasmodium spp. on

400 Blue-crowned Manakins shall explore variables directly related to the host fitness; other

401 characteristics related to parasitism, such as the identification of Plasmodium spp.

402 species; and the impact of the infection in the species over longer time-scales. Testing

403 the inheritance of female preferences and the activities of infected and uninfected males

404 will be crucial to the understanding of the role of the male behaviors presented here to

405 the evolution of mechanisms of female preferences based on honest signals.

406 Based on what was stated above, we concluded that the Plasmodium spp. has an

407 effect over social behavior of L. coronata males in multi-scales. The social behavior of

408 males also had an influence, at least indirectly, over female preference, although this

409 preference had no direct relation neither with Plasmodium spp. prevalence in the lek,

410 nor the occurrence of malaria in the individual. These results are partially in accordance

411 to the parasite hypothesis (Hamilton and Zuk 1982, Kirkipatrick and Ryan 1991), which

412 suggests that parasites affect the behavior of males and, consequently, female visits,

413 although they do not discriminate between infected and uninfected males. In this regard,

414 the male vocalization and display behaviors would be honest signals, capable of

415 indicating how resistant to infection the individual is. For that reason, such male

416 courtship behaviours would be important signals for sexual selection, allowing females

26
417 to obtain cues about the probability of a male being infected by Plasmodium spp. during

418 visits to their perches.

419 ACKNOWLEDGEMENTS

420 The project funding and grants were provided by Conselho Nacional de

421 Desenvolvimento Científico e Tecnológico ((UNIVERSAL – MCTI/CNPq 471092/

422 2012-6 to MA)) and Fundação de Amparo à Pesquisa do Estado do Amazonas

423 (PRONEX/FAPEAM/CNPQ EDITAL n° 003/2009; MCT/FAPEAM/CNPQ EDITAL

424 n° 005/2007 to AF). MB currently receives a fellowship from CNPq. AF receives a

425 post-doctoral fellowship from FAPEAM, and PS receives a post-doctoral fellowship

426 from CAPES. We thank Brazilian Biodiversity Research Programme (PPBio), for

427 logistical support; Mariana Tolentino, Nayara Soto, Marina Maximiano, Gueviston

428 Lima, Priscila Diniz, Guilherme Christo, Ricardo Gonçalves, and José Ferreira for field

429 work assistance; and Renata Durães for valuable comments on the manuscript.

430

431 REFERENCES

432 Able DJ. 1996. The contagion indicator hypothesis for parasite-mediated sexual

433 selection. PNAS. 93: 2229 – 2233.

434 Alexander RD. 1974. The evolution of social behavior. Ann Rev Ecol Syst. 5: 325-383.

435 Altmann J. 1974. Observational study of behavior: sampling methods. Behaviour.

436 49:227-67.

437 Anciães M, Durães RR, Cerqueira MC, Fortuna JR, Sohn N, Cohn-Haft M, Farias IP.

438 2009. Diversidade de piprídeos (Aves: Pipridae) amazônicos: seleção sexual, ecologia e

439 evolução. Oecol Bras. 13: 165-182.

27
440 Anciães M, Prum R. 2008. Manakin display and visiting behaviour: a comparative test

441 of sensory drive. Anim Behav. 75:783-790.

442 Andersson M. 1994. Sexual Selection. Princeton, NJ: Princeton University Press

443 Ardia DR, Schat KA. 2008. Ecoimmunology. In: Davison F, Kaspers B, Schat KA,

444 editors. 5nd ed. Avian Immunology. London: Academic Press, Elsevier. p. 421-441.

445 Asghar M, Hasselquist D, Bensch S. 2011. Are chronic avian haemosporidian infections

446 costly in wild birds?. J Avian Biol. 42: 530 – 537.

447 Atkinson CT, Van Ripper III C. 1991. Pathogenicity and epizootiology of avian

448 haematozoa: Plasmodium, Leucocytozoon, and Haemoproteus. In: Loye JE, Zuk M,

449 editors. Bird-Parasite interaction: Ecology, Evolution and Behavior. 1nd ed. New York:

450 Oxford. p. 19-48.

451 Barske J, Fusani L, Wikelski M, Feng NY, Santos M, Schlinger BA. 2014. Energetics

452 of the acrobatic courtship in male golden-collared manakins (Manacus vitellinus).

453 PNAS. B281: 20132482.

454 Beehler BM, Foster MS. 1988. Hotspots, and Female Preference in the Organization of

455 Lek Mating Systems. Am Nat. 131: 203-219.

456 Beltran-Bech S, Richard F. 2014. Impact of infection on mate choice. Anim Behav. 90:

457 159-170.

458 Bensch S, Waldenström J, Jonzén N, Westerdahl H, Hansson B, Sejberg D, Hasselquist

459 D. 2007. Temporal dynamics and diversity of avian malaria parasites in a single host

460 species. J Anim Ecol. 76:112–12.

461 Boyce MS. 1990. The Red QueenVisits Sage Grouse Leks. Amer Zool. 30:263-270.

28
462 Bradbury JW, Gibson RM. 1983. Leks and mate choice. In: Bateson P, editor. Mate

463 choice. Cambridge: Cambridge University Press. p. 109-138.

464 Breed MD, Guzmán-Novoa E, Hunt GJ. 2004. Defensive behavior of honey bees:

465 Organization, genetics, and comparisons with other bees. Annu Rev Entomol. 49: 271-

466 298.

467 Buchanan KL, Catchpole CK, Lewis JW, Lodge A. 1999. Song as an indicator of

468 parasitism in the sedge warbler. Anim Behav. 57:307–314.

469 Caillaud D, Craft ME, Meyers LA. 2013. Epidemiological effects of group size

470 variation in social species. J R Soc Interface. 10: 20130206.

471 Daviews CR, Ayres JM, Dye C, Deane LM. 1991. Malaria Infection Rate of Amazonian

472 Primates Increases with Body Weight and Group Size. Funct Ecol. 5: 655-662.

473 Díaz-Muñoz SL, DuVal EH, Krakauer AH, Lacey EA. 2014. Cooperating to compete:

474 altruism, sexual selection and causes of male reproductive cooperation. Anim Behav.

475 88:67-78.

476 Durães R, Loiselle BA, Blake JG. 2008. Spatial and temporal dynamics at manakin

477 leks: reconciling lek traditionality with male turnover. Behav Ecol Sociobiol. 62:1947–

478 1957.

479 Durães R. 2009a. Lek Structure and Male Display Repertoire of Blue-Crowned

480 Manakins in Eastern Ecuador. Condor 111:453-461.

481 Durães R, Loiselle BA, Parker PG, Blake JG. 2009b. Female mate choice across spatial

482 scales: influence of lek and male attributes on mating success of blue-crowned

483 manakins. Proc R Soc B. 276.

29
484 Endler JA, McLellan T. 1988. The process of evolution: toward a new synthesis. Ann

485 Rev Ecol Syst. 19: 395-421.

486 Endler JA, Théry M. 1996. Interacting effects of lek placement, display behavior,

487 ambient light, and color patterns in three neotropical forest-dwelling birds. Am Nat.

488 148:421-445.

489 Fisher RA. 1915.The evolution of sexual preference. Eugenics Rev. 7: 184-192.

490 Relationships between Blood Parasites, Mating Success and Phenotypic Cues in Male

491 Sage Grouse Centrocercus urophasianus. Amer Zool. 30:271-278.

492 Gilman S, Blumstein DT, Foufopoulos J. 2007. The Effect of Hemosporidian Infections

493 on White-Crowned Sparrow Singing Behavior. Ethol. 113:437–44.

494 Godfrey SS, Bull CM, Murray K, Gardner MG. 2006. Transmission mode and

495 distribution of parasites among groups of the social lizard Egernia stokesii. Parasitol

496 Res 99:223–230.

497 Gomez D, Théry M. 2004. Influence of ambient light on the evolution of colour signals:

498 comparative analysis of a neotropical rainforest bird community. Ecol Lett. 7:279-284.

499 Hamilton WD, Zuk M. 1982. Heritable True Fitness and Bright Birds: A Role for

500 Parasites?. Science. 218:384-387.

501 Heindl M, Winkler H. 2003a. Interacting effects of ambient light and plumage color

502 patterns in displaying wire-tailed manakins (Aves, Pipridae). Behav Ecol Sociobiol. 53:

503 153-162.

504 Heindl M, Winkler H. 2003b. Vertical lek placement of forestdwelling manakin species

505 (Aves, Pipridae) is associated with vertical gradients of ambient light. Biol J Linnean

506 Soc 80: 647-658

30
507 Hellgren O, Waldenström J, Bensch S. 2004. A new PCR assay for simultaneous studies

508 of Leucocytozoon, Plasmodium, and Haemoproteus from avian blood. J Parasitol

509 90:797-802.

510 Hellgren O, Wood MJ, Waldenström J, Hasselquist D, Ottosson U, Stervander M,

511 Bensch S. 2013. Circannual variation in blood parasitism in a sub-Saharan migrant

512 passerine bird, the garden warbler. J Evol Biol. 26:1047–1059.

513 Hellmayr CE. 1929. Catalogue of Birds of the Americas and the adjacent islands. Field

514 Museum Natural History Publication Zoological Series. 6nd ed. Chicago: Field

515 Museum of Natural History

516 Höglund J, Alatalo RV, Lundberg A. 1992. The effects of parasites on male ornaments

517 and female choice in the lek-breeding black grouse (Tetrao tetrix). Behav Ecol

518 Sociobiol. 30:71-76.

519 Ito H, Sudo-Yamaji A, Abe M, Murase T, Tsubota T. 2003. Sex Identification by

520 Alternative Polymerase Chain Reaction Methods in Falconiformes. Zool Sci. 20:339-

521 344.

522 Johnson LL, Boyce MS. 1991. Female choice of males with low parasite loads sage

523 grouse. In: Loye JE, Zuk M, editors. Bird-Parasite interaction: Ecology, Evolution and

524 Behavior. 1nd ed. New York: Oxford. p. 19-48.

525 Jones TC, Pruitt JN, Riechert S. 2010. Reproductive success in a socially polymorphic

526 spider: social individuals experience depressed reproductive success in isolation. Ecol

527 Entomol 35:684–690.

528 Kirkpatrick M. 1982. Sexual selection and the evolution of female choice. Evolution

529 82:1–12.

31
530 Kirkpatrick M, Ryan MJ. 1991. The evolution of mating preferences and the paradox of

531 the lek. Nature. 350: 33-38.

532 Kirwan GM, Green G. 2012. Cotingas and Manakins. Princeton University Press.

533 Knowles SCL, Palinauskas V, Sheldon BC. 2010. Chronic malaria infections increase

534 family inequaly and reduce parental fitness: experimental evidence from a wild bird

535 population. J Evol Biol. 23:557-569.

536 Kokko H, Jennions MD, Brooks R. 2006. Unifying and testing models of sexual

537 selection. Annu Rev Ecol Evol Syst. 37:43–66.

538 Korb J, Heinze J (2008) Ecology of Social Evolution. Verlag Berlin Heidelberg:

539 Springer.

540 Kuijper B, Pen I, Weissing FJ. 2012. A Guide to Sexual Selection Theory. Ann Rev

541 Ecol Syst. 43:287–311.

542 Lacorte GA, Félix GMF, Pinheiro RRB, Chaves AV, Almeida-Neto G, Neves FS, Leite

543 LO, Santos FR, Braga EM. 2013. Exploring the Diversity and Distribution of

544 Neotropical Avian Malaria Parasites – A Molecular Survey from Southeast Brazil.

545 PLoS ONE. 8(3): e57770.

546 Lande R. 1981. Models of speciation by sexual selection on polygenic traits. Proc Natl

547 Acad Sci. 78:3721–3725.

548 Lebigre C, Alatalo RV, Siitari H. 2013. Physiological costs enforce the honesty of lek

549 display in the black grouse (Tetrao tetrix). Oecol. 172:983–993.

550 Marzal A, Lope F, Navarro C, Møller AP. 2005. Malarial parasites decrease

551 reproductive success: an experimental study in a passerine bird. Oecol. 142:541–545.

32
552 Maynard DF, Ward KA, Doucet SM, Mennill DJ. 2014. Telemetric and video

553 assessment of female response to male vocal performance in a lek-mating manakin.

554 Behav Ecol 00:1–10.

555 Nunn CL, Heymann EW. 2005. Malaria infection and host behavior: a comparative

556 study of Neotropical primates. Behav Ecol Sociobiol. 59: 30–37.

557 PPBio (2014): http://ppbio.inpa.gov.br/sitios/br319/infra/km100.

558 Prum RO. 1997. Phylogenetic Tests of Alternative Intersexual Selection Mechanisms:

559 Trait Macroevolution in a Polygynous Clade (Aves: Pipridae). Am Nat. 149: 668-692.

560 Prum RO. 2010. The lande–kirkpatrick mechanism is the null model of evolution by

561 intersexual selection: implications for meaning, honesty, and design in intersexual

562 signals. Evolution. 64: 3085–3100.

563 Prum RO. 2012. Aesthetic evolution by mate choice: Darwin’s really dangerous idea.

564 Phil Trans R Soc B. 367:2253–2265.

565 R Development Core Team. 2011. R: A language and environment for statistical

566 computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-

567 07-0, URL http://www.R-project.org/.

568 RADAM Brasil. 1978. Folha SB.20 Purus; geologia, geomorfologia, pedologia,

569 vegetação e uso potencial da terra (ed Mineral, D.N.P.). p. 17-566.

570 Ryder TB, Durães R. 2005. It’s not easy being Green: using molt and morphological

571 criteria to age and sex Green-plumage manakin (Aves: Pipridae). Ornitol Neotrop. 16:

572 481-491.

573 Sambrook J, Russel DW, Sambrook J. 2001. Molecular Cloning: A Laboratory Manual.

574 New York: CHSL Press.

33
575 Schall JJ. 1983. Lizard malaria: cost to vertebrate host's reproductive success. Parasitol

576 87: 1-6.

577 Sick H. 1967. Courtship behavior in manakins (Pipridae): a review. Liv Bird. 6: 5-22.

578 Silveira P, Marin SYG. ; Moreira PA. ; Tocantins BB ; Lacorte G; Paixão TA ; Martins

579 NRS ; Braga EM. 2013. Interactions of Plasmodium juxtanucleare and chicken anaemia

580 virus: establishing a model. Parasitolology. 140: 1777-1788.

581 Skutch AF. 1969. Life histories of Central American birds. III. Pacific Coast Avifauna.

582 35: 1–580.

583 Snow DW. 1963. The Evolution of manakin courtship display. Proceedings of the

584 International Ornithological Congress. 13: 553-561

585 Snow DW. 2004. Family Pipridae (Manakins). In: Hoyo D, Elliott A, Christie DA.

586 Handbook of the Birds of the World. 9nd ed. Barcelona: Lynx Editions. p. 110-169.

587 Tella JL. 2002. The evolutionary transition to coloniality promotes higher blood

588 parasitism in birds. J Evol Biol. 15: 32-41.

589 Uy JAC, Stein AC. 2007. Variable visual habitats may influence the spread of colourful

590 plumage across an avian hybrid zone. J Evol Biol. 20:1847-1858.

591 Valkiūnas G. 2005. Avian malaria parasites and other haemosporida. CRC Press

592 Williams RB (2005) Avian malaria: clinical and chemical pathology of Plasmodium

593 gallinaceum in the domesticated fowl Gallus gallus. Avian Pathol. 34:29-47.

594 Zahavi A (1975) Mate selection: a selection for a handicap. J Theor Biol. 53: 205–214.

34
595 FIGURE LEGENDS

596 Figure 1: Map of the study area, and the research module in Manaquiri, Careiro da

597 Varzea municipality, Castanho, Amazonas state, Brazil. The open square represents the

598 location of the research module, not in scale. The black dots represent the sampled leks.

599 Adapted from http://ppbio.inpa.gov.br/sitios/br319/infra/km100

600 Figure 2: Frequency of Blue-crowned Manakins captured in leks: adult males (black);

601 females (gray), and juvenile males (white).

602 Figure 3: Effect of prevalence of Plasmodium spp. in leks on (a) total vocalization rate

603 and (b) total display frequency in leks.

604 Figure 4: Residuals (Mean + 1 SD) of regressions between (a) vocalization rate and (b)

605 display frequency and individual interaction frequency for infected and uninfected

606 males.

607 Figure 5: Correlation between frequency of female visits and total interaction

608 frequencies in leks.

609 Figure 6: Effects of social behaviors of adult males on frequency of female visits in leks

610 of Blue-crowned Manakins: a) male vocalization rates; b) male display frequencies and

611 c) male interaction frequencies.

612

613

614

615

616

617

618

35
619 Table 1.

Indexes Formula

Body condition (body mass of individual/ length of the tarsus)

Total vocalization frequency per [(∑ Nt of vocalization in lek/ ∑ Nintle)/5]

minute in the lek

Total display frequency in the lek [(∑ Nintle with displays in lek/∑ Nintle)/5]

Total interaction frequency in the lek (∑ Nintle vocalizing or displaying in lek/ ∑ Nintle)

Individual vocalization rate per minute [(∑ Nt of vocalization at individual perch/∑

Nintind)/5]

Individual display frequency [(∑ Nintind with displays at individual perch/ ∑

Nintind)/5]

Individual interaction frequency (∑ Nind vocalizing or displaying/ ∑ Nintind)

Occurrence of malaria in individual Presence (1) or absence (0) of malaria in individual

Malaria prevalence in lek (∑ Nind captured infected in lek/ ∑ Nind captured in

lek)

Frequency of females Nt of female visits observed at indivídual perch

Frequency of females in lek Nt of female visits observed in lek

620

621 Abbreviation: Nt = Total Number; Nintle = Total number of observation intervals in the

622 lek; Nind = Total number of individuals; Nintind = Total number of observation

623 intervals to focal perches; the frequency of interaction among males.

624

625 Table 1: Indexes for body condition and social behaviors of adult male Blue-crowned

626 Manakins.

627

36
628 Table 2.

Correlation indexes Results

Body condition and and total activity frequency per minute in

r 26 = 0.31, P = 0.08
lek

Total vocalization frequency per minute in lek and total

r 9 = 0.56, P = 0.08)
display frequency in the lek

Total vocalization frequency per minute in lek and total

r 9 = 0.26, P = 0.29
interaction frequency in the lek

Total display frequency in the lek and total interaction

r 9 = 0.30, P = 0.38
frequency in the lek

Individual vocalization frequency per minute and individual

r 26 = 0.71, P = 0.02
display frequency

Individual vocalization frequency per minute and individual

r 26 = 0.62, P < 0.01
interaction frequency

Dance rate of individual per minute and interaction rate of

r 26 = 0.65, P = 0.02
individual

Total interaction frequency in the lek and frequency of males

r²9 = 0.52, P = 0.01
in lek

629

630 Table 2: Collinearity among derived indexes for body condition and social behaviors of

631 lekking Blue-crowned Manakins.

37
Figure 1.

38
Figure 2.

39
Figure 3.

a) b)

40
Figure 4.

a) b)

41
Figure 5.

42
Figure 6.

a) b)

c)

43
CONCLUSÃO

Com base no exposto acima, concluimos que o Plasmodium spp. tem efeito sobre o

comportamento social dos machos de L. coronata em multi-escalas. O comportamento social dos

machos também teve influência, ao menos indireta, sobre a preferência da fêmea, embora essa

preferência não tenha relação direta nem com a prevalência de Plasmodium spp. no lek e nem com a

ocorrência de malária no indivíduo. Esses resultados estão parcialmente de acordo com a hipótese

dos parasitos (Hamilton e Zuk, 1982, Kirkipatrick e Ryan 1991), o qual sugere que os parasitos

afetam o comportamento dos machos e, consequentemente, a preferência da fêmea. Desse modo, os

comportamentos de vocalização e de dança dos machos seriam sinais honestos capazes de mostrar o

quão resistente à infecção é o indivíduo. Por isso, as exibições dos machos seriam um importante

sinalizador para a seleção sexual, uma vez que tais características permitem às fêmeas obterem

pistas sobre a probabilidade de um macho estar, ou não, infectado por Plasmodium spp. durante as

visitas aos poleiros de corte.

44
ANEXO 1 (Ata de qualificação)

45
ANEXO 2 (Ata de defesa)

46